World Bank Reprint Series: Number 299 Shujiro Urata Factor Inputs and Japanese Manufacturing Trade Sotructure Reprinted with permission from The Review of Econiomics aned Statistics, vol. 65, no. 4 (November I983), pp. 678-84. World Bank Reprints No. 258. Christiaan Grootaert, "The Conceptual Basis of Measures of Household Welfare and Their Implied Survey Data Requirements," The Review of Incomle lanfd Wealthl No. 259. Guy Pfeffermann and Richard Webb, "Poverty and Income Distribution in Brazil," The Rev)iew of Incomle anid WeAalth No. 260. Pradeep K. Mitra, "A Theory of Interlinked Rural Transactions," Journal of PulNic Economnics No. 261. David L. Lindauer and Richard H. Sabot, "The Public/Private Wage Differential in a Poor Urban Economy," Jouinral of Developtmient Econonoics No. 262. J. B. Knight and R. H. Sabot, "Labor Market Discrimination in a Poor Urban Economy," Joturnial of Developmlenft Studies No. 263. Carl Dahlman and Larry Westphal, "Technical Effort in Industrial Development: An Interpretative Survey of Recent Research," The Econiomiiics of New Teclmologil in Developing Countries No. 264. Michael Bamberger, "The Role of Self-Help Housing in Low-Cost Shelter Programs for the Third World," Built Environmient No. 265. Bela Balassa, "The Adjustment Experience of Developing Economies after 1973," IMF Conditionality No. 266. Bela Balassa, "Outward Orientation and Exchange Rate Policy in Developing Countries: The Turkish Experience," The Middle Eaist Joirniial No. 267. Dipak Mazumdar, "Segmented Labor Markets in LDCs," Amlericanl Ecotlonoic Reviewo No. 268. Stephen P. Heyneman and William A. Loxley, "The Effect of Primary-School Quality on Academic Achievement across Twenty-nine High- and Low-Income Countries," The American Journal of Sociology No. 269. James R. Follain, Jr., Gill-Chin Lim, and Bertrand Renaud, "Housing Crowding in Developing Countries and Willingness to Pay for Additional Space: The Case of Korea," Jouirnatil of Developmenit Econiomics No. 270. Bela Balassa, "Policy Responses to External Shocks in Sub-Saharan African Countries," Joturnzal of PolictI Modeling No. 27I. Jaime de Melo and Sherman Robinson, "Trade Adjustment Policies and Income Distribution in Three Archetype Developing Economies," Journal of Development Econiomics No. 272. J. B. Knight and R. H. Sabot, "The Role of the Firm in Wage Determination: An African Case Study," Oxford Ecolonoic Papers No. 273. William G. Tyler, "The Anti-Export Bias in Commercial Policies and Export Perfor- mance: Some Evidence from Recent Brazilian Experience," Weltwirtschaftliches Arclliv No. 274. Ron Duncan and Ernst Lutz, "Penetration of Industrial Country Markets by Agricul- tural Products from Developing Countries," World Development No. 275. Malcolm D. Bale, "Food Prospects in the Developing Countries: A Qualified Optimistic View," The Anmericant Economic Review (with Ronald C. Duncan) and "World Agricultural Trade and Food Security: Emerging Patterns and Policy Directions," Wiscotnsitn International Law Jouirnital (with V. Roy Southworth) No. 276. Sweder van Wijnbergen, "Interest Rate Management in LDCs," Joirnazzl of k\loi'etarl Economics No. 277. Oli Havrylyshyn and Iradj Alikhani, "Is There Cause for Export Optmism? An Inquiry into the Existence of a Second Generation of Successful Exporters," IMtt'lhwirft51 iC lics Archliiv Reprinted fromn THE REVIEW OF ECONOMICS AND STATISTICS Puiblished for Harvard Unilersity by the North-Holland Publishing Company Copyright, 1983, by the President and Fellowvs of Harvard College Vol. LXV, No. 4, November, 1983 FACTOR INPUTS AND JAPANESE MANUFACTURING TRADE STRUCTURE Shujiro Urata* I. Introduction the energy variable. Unlike the previous studies of Japan cited above, which examined the full range of tradable Two different approaches are often used to explain commodities, only trade in manufactured goods is the structure of trade (Stern (1975)). The first, devel- analyzed in this paper. Analysis is undertaken for the oped by Leontief (1953), compares the relative total years 1967 and 1975, which bracket major monetary (direct and indirect) factor intensities of exports and and energy shocks, and for which detailed sectoral data imports. Tatemoto and Ichimura (1959) used this are readily available. After performing separate regres- method to analyze the 1951 pattern of Japanese trade, sions on data for each year, the changes in trade struc- where they found an interesting dualism with respect to ture between 1967 and 1975 are examnined. Finally, the capital-labor intensity. Exports to developed countries results are analyzed in light of the changes in patterns of were labor intensive, while exports to developing coun- factor usage that occurred in Japan and her trading tries were capital intensive. Applying the same method partners. to trade during the 1955-1968 period, Heller (1976) found support for this dualism for both capital-labor II. Methodology and skill intensities, but also observed a substantial narrowing of the differential in the factor intensities of Two different sets of regressions have been carried exports to developed and developing countries. out in analyzing the Japanese trade structuredfn manu- The second approach, which is used in this paper to factured goods based on alternative measures of human analyze the Japanese manufacturing trade structure in capital: the skill measure (section A) and the wage 1967 and 1975, regresses interindustry variations in net differential measure (section B). One explanatory note exports, exports and imports on sectoral factor usage. on the wage differential measure that is utilized below: The explanatory variables include energy inputs, as well Although this type of wage differential measure is often as the more traditional primary inputs of labor, capital used as a human capital variable in analyzing the trade and human capital (Baldwin (1971), Branson and structure of the United States,' it may not be ap- Monoyios (1977)). Although energy itself is not a propriate as an explanatory variable for the Japanese primary factor of production, Griffin and Gregory's trade structure for several reasons. First, wage payments (1976) demonstration of substitutability between energy in Japan traditionally include not only regular salary, and value added provides a rationale for inclusion of but also a sizeable bonus payment, which is usually paid twice a year based on the performance of the company. Therefore, wage payments are highly correlated with the Received for publication July 30, 1981. Revision accepted for company's profit but not with the characteristics of the publication January 5,1983. employees. Second, the relative immobility of labor * The World Bank. A part of the research was done while the author was a across industries and the significant amount of non-wage research associate at the Brookings Institution. The author compensation such as company housing makes wage wishes to thank Jaime de Melo, Kenneth Flamm, Robert comparison difficult. Lawrence, Jeffrey Lewis, Kenneth Oye and two referees for helpful comments. Neither th-y, the Brookings Institution nor the World Bank are responsible for any views and interpreta- l See, for example, Branson and Monoyios (1977), and Stern tions expressed in this paper. and Maskus (1981). NOTES 679 A. Skill Measure where The regression equation takes the following form: TL = total labor (UL + SL) WD = human capital measured by wage differential. NX The notations for the rest of the variables are as given X ao + aMUL + a2SL + a3CA + a4EX earlier. where III. The Data NX= net exports X= exports Sixty-four SITC (5-8) 2- and 3-digit commodities are M = imports analyzed for 1967 and 1975 in 1975 prices. The data on UL = unskilled labor (production workers) trade variables are taken from OECD's Trade by Com- SL skilled labor (technical and admninistrative modities, Series B for the respective years. The conver- workers) sion of trade data to constant 1975 prices is based on CA capital (book value of depreciable assets) export and import price indices derived from the Price EN = energy (cost of fuels and electricity). Indexes Manual (1970 and 1975), published by the Bank of Japan. The data on factor usage and other explana- The regression coefficients are estimated separately tory variables and the value of shipments (VS) are for Japan's trade in 1967 and 1975 with (1) the world; 2 compiled from the Census of Manufactures (1967 and (2) the OECD countries; and (3) developing countries. 1975), published by the Ministry of International Trade The problem of heteroscedasticity is treated using the and Industry in Japan. Unskilled labor (UL) and skilled generalized least squares method following Branson and labor (SL) are the numbers of production workers and Monoyios (1977) which in turn followed the Goldfeld technical and administrative workers, respectively. The and Quandt and Glejser procedures outlined in John- human capital measure based on wage differentials ston (1972). According to this approach, when hetero- (WD) is calculated by discounting the excess of the scedasticity is detected, the absolute value of the calcu- average wage in each industry over the average wage lated residuals from ordinary least squares is regressed earned in 1975 by a worker under age 174 (taken to on VS (the value of shipments), VS2 and VS112 with a represent unskilled labor) and by multiplying this excess constant term. Then the equation which gives the largest wage by employment in the industry.5 Capital (CA) is multiple correlation coefficient (R2) is used as a weight the book value of depreciable assets. The value of for the generalized least squares regression.3 energy (EN) is the cost of fuels and electricity. The manufacturing wage index taken from the Japan B. Wage Differential Measure Statistical Yearbook (1972 and 1978) is used as a defla- tor for the computation of the wage differential mea- Similar regression analyses to those of section A are sure. The wholesale prices of capital goods and energy conducted using the human capital variable measured from the Price Indexes Manual (1978) are utilized for by wage differentials. The following equation is used for computing input values at constant prices. The pro- the regression analysis: ducers' price of manufactured goods from the Price X = a + a1TL + a2WD + a3CA - a4EN 4The average wage earned by a worker under age 17 in 1975 MX ao+aT 2D+aC 4was obtaned from the Census of Wages, vol 1, 1975, Ministry M of Labor, Japan. The definition of skilled labor used here is based on workers' age rather than workers' educational back- 2 The world here consists of the OECD countries, developing ground, which was used in earlier studies on the U.S. trade countries and other countries, where the last category is com- structure (e.g., see Branson and Monoyios (1977)). While the prised mainly of the socialist countries. unavailability of data on Japanese workers' educational back- 3The estimated value of the dependent variable is used for ground precluded use of the latter definition, the definition weights if both coefficients on the constant term and the based on workers' age may in fact be more applicable to Japan. variable are statistically significant. Otherwise, the value of the Japanese workers usually gain their professional skills through independent variable is used as a weight. Although Branson on-the-job training, so that close relation exists between and Monoyios (1977) attempted the same procedure described workers' skill level and their duration of employment at a in the text, they rnisspecified the weighted regression equation particular company. While data on duration of employment are by includi,ig a constant term as pointed out by Stern and not available, the immobility of Japanese labor markets means Maskus (1981). On the other hand, Maskus (1981) uses the reai that the workers' age may be a good proxy for employment value of shipment as a scaling factor. Therefore, the factor-out- duration and hence skill level. put ratios can be interpreted as the measure of factor intensities 5See Branson and Monoyios (1977), p. 114 for a more in the long run. complete description of this method. 680 THE REVIEW OF ECONOMICS AND STATISTICS TABLE 1.-ESTIMATES OF REGRESSIONS FOR 1967: SKILL MEASURE Independent Vzariables Dependent Equation Variables C UL SL CA Ei R2 Number World NX -22.32a 2.97b -2.06 -0.04 -0.13 0.269a 1.1 (-3.16) (2.61) (-0.53) (-0.09) (- 0.10) X 2.99 2.34a -1.22 0.34 -0.34 0.537a 1.2 (0.64) (3.11) (-0.48) (1.27) (-0.40) M 23.99a -0.43 0.37 0.35a -0.35 0.353a 1.3 (2,96) (- 1.56) (0.38) (3.13) (-1.02) OECD NX -11.75a 1.79a -2.55 -0.11 0.12 0.226a 1.4 (- 3.29) (3.12) (-1.30) (-0.52) (0.18) X 3.58 1.27a -0.64 0.03 -0.006 0.414a 1.5 (1.31) (2.88) (-0.42) (0.18) (- 0.01) M 18.07a -0.36b 1.03b 0.16a -0.22 0.510a 1.6 (3.91) (-2.36) (2.08) (2.97) (-1.02) Developing NX .-3.01 0.84b 0.23 0.14 -0.20 0.429a 1.7 Countries (-1.30) (2.24) (0.18) (1.02) (- 0.47) X -0.62 0.85b -0.15 0.21 -0.24 0.478a 1.8 (-0.27) (2.32) (- 0.12) (1.61) (-0.58) M 2.39a 0.01 -0.38 0.07b -0.04 0.223a 1.9 (3.82) (0.10) (- 1.10) (1.03) (-0.38) Notes: C should be interpreted as VS-112 in all equations except 1.3 and 1.6, where it should be interpreted as (23.761 + 0.035 VS)-' and (11.518 + 0.704 VS'/2)-', respectively. t-values are in parentheses. eSignificant at the 1% level. bSignificant at the 5% level. Indexes Manual (1970, 1975) is used as a deflator for the This inconsistency with previous studies may be at- value of shipments. tributable to differences in the measure of factor usage IV. Results' applied and differences in the range of commodities * Reuexamined. As was mentioned in the introduction, the previous studies on the Japanese trade structure analyzed A. Trade Structure in 1967 and 1975 indirect as well as direct factor inputs whereas in this paper only direct factor inputs are examined. Although e 1 s the total measure (indirect and direct inputs) should be skilled labor is positively related to net exports and used in the analysis as suggested by Deardorff (1982), exports to all three regions with at least a 5% signifi lack of input-output tables at a suitable disaggregation cance level, and shows a significant negative relation l wih.mprt fro th OEC contie., Th cofiin level and for the desired years precludes us from utiliz- on illreds laor is pOst and significas.T conl for ing tne total measure. However, the results do not seem onskie lfrom the OECD sitive s,nCapital is positively to be affected much by employing only direct measures, imports fDcountries, since the correlation coefficient between the direct and associated with imports in all the cases. From these t results Japan appears to have exported unskilled labor t m 7 n goods i (significant at 1% level). The previous studies analyzed intesivre good adeve porte catitsas ienasiv E trade in the primary and services sectors as well as itsntraes withdevelycopntering uthes wvidellf duas E manufactured goods, while our study analyzes only cuntres, tearle counter. manufactured goods. Therefore, Japan might be shown to have exported capital intensive goods and imported 6 Only the results based on the skill measure are reported here labor intensive goods from developing countries as the because of two reasons: (1) the results from the two approaches previous studies indicated if labor intensive primary are very similar and (2) the skill measure seems more ap- propriate to describe human capital in Japan than the wage 'The "Sources of Growth and Structural Change" (RPO differential measure as explained above. However, the results of 671-32) project data bank at the World Bank. 1965 is the regressions based on the wage differential measure are available closest year to the period analyzed here for which the data are from the author. available. NOTES 681 TABLE 2.-ESTIMATES OF REGRESSION EQUATIONS FOR 1975: SKILL MEASURE Independent Variables Dependent - Equation Variables C UL SL >A EN R2 Number World NX -20.61a -0.15 0.10 0.90a -1.78' 0.517a 2.1 (-2.87) (-0.22) (0.05) (3.64) (-1.98) X 3.48 -0.0001 0.47 0.92a -1.92b 0.638a 2.2 (0.54) (-0.0002) (0.24) (4.14) (-2.39) M 18.75a 0.34b 0.52 -0.06 0.13 0.565a 2.3 (3.25) (2.21) (1.24) (- 1.10) (0.64) OECD NX -13.25a 0.36 0.01 0.18 -0.31 0.338' 2.4 (3.77) (1.08) (0.01) (1.46) (-0.71) X 2.36 0.16 1.08 0.19c -0.41 0.512a 2.5 (0.76) (0.56) (1.18) (1.78) (-1.07) M 12.36' -0.07 1.01a -0.01 0.008 0.527a 2.6 (3.01) (-0.60) (3.11) (-0.36) (0.05) Developing NX -0.65 -0.44 -0.10 0.63' -1.41' 0.506a 2.7 Countries (-0.16) (-1.12) (-0.08) (4.41) (-2.70) X 1.21 -0.10 0.53 0.61' -1.38a 0.548a 2.8 (0.30) (-0.25) (-0.43) (4.36) (-2.70) M 2.26 0.35a -0.34' -0.05a 0.11 0.5658 2.9 (1.32) (7.20) (-2.47) (- 2.39) (1.61) Notes: C should be interpreted as VS-'/2 in all equations except 2.3, 2.6, 2.9, where it should be interpreted as (20.776 + 0.299 VS'I4)-', (12.616 + 0.272 VS'I/2)' and (4.651 + 0.102 VS'/2')-, respectively. t-values are in parentheses. 'Significant at the 1% level. b,Significanat at the 5% level. 'Significant at the 10% level. goods were included in the analysis, especially since our equation, there remain some manufactured goods Japan experienced a large import surplus in its trade in whose major inputs are natural resources (e.g., wood primary goods with developing countries.8 pulp into the paper sector). This would explain the However, the 1967 results remain a paradox since one positive, significant capital coefficient on manufactured might expect the trade structure of manufactures to imports from the developing countries, since due to its reflect the pattern of factor usage in trading countries. poor endowment of natural resources, Japan would In other words, being relatively well endowed with import natural resource intensive manufactured goods capital relative to labor within the manufacturing sector, whose production in Japan would require use of a Japan would be more likely to export capital intensive capital intensive technology,10 goods and import labor intensive goods in trade with The estimates for 1975 are shown in table 2. Un- the developing countries. There are several possible skilled labor shows a positive and significant relation- explanations for the paradox. The most appropriate ship with imports from the world and developing explanation seems to be the effect of inclusion of natu- countries. The coefficients on skilled labor are signifi- ral resource intensive commodities in our sample.9 Al- cantly positive and negative for imports from the OECD though the effect of energy related natural resources and developing countries, respectively. Capital is posi- such as oil should be captured by the energy variable in tive and significant for net exports and exports to the 8In for which the data are available from the "Sources world and developing countries and marginally signifi- In1965, frwihtedtaravialfrmte'Sucs canlt to the OECD countries, For the imports from the of Growth and Structural Change" project at the World Bank, labor-capital ratios for the primary and for the manufacturing t0The regression result excluding natural resource intensive sectors are 1.41 and 0.66. sho that re excint natal resourcesintensive 9 Other possible explanations include factor intensity reversal, goods shows that the coefficient on capital for imports from the domestic demand bias and distortions. In addition, Leamer and developing countries is not significantly different from zero. Bowen (1981) have recently demonstrated that analyzing the M = 1.74 + 0.009 UL -0.27 SL + 0.04 CA + 0.17 EN structures of trade and production in one country does not (8.43) (0.28) (-2.12) (1.07) (0.86) necessarily allow robust conclusions on the factor endowment of that country in relation to its trading partners, except in the R2 = 0.712 highly restrictive two factor case. (t-values are in parentheses.) 682 THE REVIEW OF ECONOMICS AND STATISTICS TABLE 3.--ESTIMATES OF REGRESSION EQUATIONS FOR TEE 1967-75 PERIOD: SKILL MEASURE Independent Variables Dependent - Equation Variables C UL DUL SL DSL CA DCA EN DEN R2 Number World NX -22.03a 2.96a -3.09b -2.06 2.19 -0.03 0.93C -0.13 -1.64 0.333' 3.1 (-4.42) (3.37) (-2.15) (-0.69) (0.46) (-0.11) (1.77) (-0.13) (-0.90) X 3.07 2.33a -*2.33b -1.22 1.70 0.34 0.57 -0.34 -1.58 0.573a 3.2 (0.87) (3.76) (-2.29) (-0.57) (0.51) (1.53) (1.55) (-0.48) (-1.22) M 21.46 - 0.55b 0.85a 0.09 0.35 0.47a -0.51a -0.35 0.41 0G554a 3.3 (4.00) (2.58) (2.64) (0.14) (0.36) (6.16) (-4.39) (- 1.11) (0.87) OECD NX -12.00a 1.80a -1.45b -2.55C 2.54 -0.11 0.29 0.12 -0.44 0.253a 3.4 (-4 79) (4.08) (-2.01) (-1.69) (1.07) (-0.68) (1.08) (0.24) (-0.48) x 3.37c 1.27a -1.12C -0.64 1.71 0.03 0.16 -0.004 -0.42 0.441a 3.5 (1.71) (3.67) (--1.97) (-0.54) (0.91) (0.22) (0.79) (-0.01) (-0.58) M 15.11a -0.32a 0.23 0.93b 0.05 0.16a -0.17a -0.19 -0.19 0.556a 3.6 (4.71) (-2.68) (1.28) (2.52) (0.10) (3.69) (-2.65) (-0.01) (0.68) Developing NX -2.62 0.83b -1.25b 0.23 -0.29 0.14 0.49b -0.20 -1.19C 0.466a 3.7 Countries (-1.37) (2.46) (-2.26) (0.20) (-0.16) (1.15) (2.43) (-0.53) (-1.70) X -0.32 0.84b -0.92C -0.15 -0.35 0.21C 0.40b -0.24 -1.12 0.511a 3.8 (-0.17) (2.55) (-1.71) (-0.13) (-0.20) (1.79) (2.03) (-0.65) (-1.64) M 2.30a 0.01 0.32b -0.38 -0.06 0.07b -0.09C -0.04 0.07 0.273a 3.9 (5.01) (0.16) (2.44) (-1.37) (-0.14) (2.52) (--1.88) (-0.46) (0.41) Notes: C should be interpreted as VS-112 in all equations except 3.3 and 3.6, where it should be interpreted as (19.542 + 0.721 VS112)-l and (14.324 + 0.359 VS1/2)-' respectively. t-values are in parentheses. uSignificant at the 1% level. bSignificant at the 5% level. CSignificant at the 10% level. developing countries, capital is negative and significant. DEN are dummy variables for UL, SL, CA and EN, The coefficient on energy is significantly negative for the respectively. net exports and exports to the world and the developing Compared to 1967 values, the coefficients on un- countries. skilled labor in 1975 are significantly lower for net In 1975, Japan exported capital intensive goods to exports and exports to all three regions, while they are both developing and OECD countries, while Japan im- significantly higher for imports from the world and ported unskilled labor intensive goods from the world developing countries. The opposite picture emerges for and especially from the developing countries. Japan the changes in coefficients on capital. The coefficients imported skilled labor intensive goods from the OECD on capital are significantly lower for imports from all countries, but she imported less of them from the devel- three areas in 1975, whereas they are significantly higher oping countries. Less energy intensive goods were ex- for net exports to the world and developing countries in ported to the world and the developing countries. These that year. The changes in coefficients on skilled labor findings on unskilled and skilled labor intensities indi- and energy are generally not significant, except that the cate that dualism did exist in trade with the OECD and coefficient on energy declined significantly for net ex- developing countries. ports to the developing countries. In short, Japanese exports shifted from unskilled labor intensive goods to B. Changes in Trade Structure between 1967 and 1975 capital intensive goods, while imports showed the op- posite shift from capital intensive goods to unskilled The differences in the estimated coefficients between labor intensive goods.1' 1967 and 1975 that appear in the previous sections An analysis of the changes in factor usage may help strongly suggest that the structure of the Japanese trade explain the shift in Japanese trade structure. Table 4 changed substantially during the 1967-75 period. In shows the changes in factor intensity with respect to order to examine the importance of changes in the unskilled labor for manufacturing industries in Japan, coefficients on independent variables, a series of dummy ,riables (which take the value of unity for 1975 and ItA similar trend regarding unskilled labor and human capital zero for 1967) are introduced into the estimated equa- has been observed for the U.S. manufacturing trade (see Maskus tions. Table 3 show the results. DUL, DSL, DCA and (1981)). NOTES 683 TABLE 4. -CHANGES IN FACTOR INTENSITY IN MANUFACTURING IV. Summary and Conclusions INDUSTRY BETWEEN 1967 AND 1975 FOR JAPAN, THE U.S. AND KOREA (%) The Japanese international trade stracture in manu- CA/UL SL/UL EN/UL factured goods has been analyzed using cross-section regression analysis in which labor, capital, human capital Japan 91.9 27.7 92.0 and energy are regressed on Japanese net exports, ex- United States 34.3 00 1 9 ports and imports with the world, OECD and develop- Koa 4ing countries for 1967 and 1975. Two different weasures Notes: U.S. data on employment are obtatned from the Cenus of Manufc. (skill and wage differential measures) have been utiized tures (1967) and the 4nnual Surrey of Manujautures (1975). Capital in 1972 pnces come from Fixed onresi-dential Business and Resuientaln Capital in the for the measuring of h Iman capital. Contrary to the United State,. J925-7.5t Data on energy are taken from the Stistslal Abhtract dualism found in earlier studies, the estimated results of the L nued States tIt979t uls on nere tuis h siae eut Korea.: Data, except energy, are taken from Hong t1976). As the data on show that in 1967 Japan exported unskilled labor inten- unskilled and skliled labor are not available. total number of employees IS used. Growth rates of unskilled and sklled labor are assumed to be identical. Energy Sive goods and imported capital intensive goods with data come from Korean Stotistirol Yearbook (1971 and 1978) published by the respect to the developing countries as well as the OECD Bureau of Stati,tics. Economic Planning Board, countries. In 1975, Japan exported capital intensive goods to all regions and imported unskilled labor inten- the United States, and Korea between 1967 and 1975. sive goods from the world and the developing countries. Figures for the United States and Korea are reasonable Skilled labor intensive goods were imported from the approximations of those for the OECD and developing OECD countries in both years. Less energy intensive country groups since their trade shares with Japan are goods were exported to the world and the developing the largest within the respective groups. Table 4 il- countries in 1975. Thf.._ does not seem to exist any lustrates that Japan's capital-unskilled labor intensity strong econometric evidence for choosing between the increased more than that of the United States and skill and wage differential measures of human capital in Korea, while the skilled labor-unskilled labor intensity explaining the Japanese trade structure. also increased, albeit less markedly.'2 The observation Drastic changes in the Japanese trade structure are regarding changes in the capital-unskilled factor inten- observed between 1967 and 1975. Japanese exports sity agrees with the shift in the Japanese trade shifted from the unskilled labor intensive goods to structure."3 Although energy usage increased most capital intensive goods while her imports shifted from among the factor inputs in Japan, Japan exported less capital intensive goods to unskilled labor intensive energy intensive goods to the developing countries in goods. This result conforms to the changes in unskilled 1975 than in 1967. One possible reason for this is that labor and capital usage in Japan and her major trading Korea may not represent a significant portion of devel- partners- oping countries' trade with Japan in energy intensive goods. In other words, if the factor usage pattern of the oil producing trading partners such as Indonesia were REFERENCES included, this apparent paradox might disappear. Baldwin, Robert E., "Determinants of Commodity Structure of The analysis above indicates that the changes in the U.S. Trade," The American Economic Review 61 (Mar. Japanese trade structure between 1967 and 1975 broadl 1971), 126-146. ly Bowen, Harry P., "Shifts in the International Distribution of reflected the shift in factor usage with respect to trading Resources and Their Impacts on U.S. Comparative Ad- partners for both unskilled labor and capital. vantage," mimeo, U.S. Department of Labor (1981). Branson, William H., and Nikolaos Monoyios, "Factor Inputs in U.S. Trade," Journlal of International Economics 7 '2Jorgenson and Nishimizu (1978) and Bowen (1981) also (May 1977), 111-131. found the gap in capital-labor ratios between the United States Deardorff, Alan V., "The General Validity of the Heckscher- and Japan becoming smaller over the 1967-75 period. How- Ohlin Theorem," The American Economic Review 72 ever, unlike our finding, Bowen found that the capital-labor (Sept. 1982), 683-694. ratio in Korea increased faster than that in Japan between 1963 Griffin, James M., and Paul R. Gregory, "An Intercountry and 1975. This difference may be attributed to the differences Translog Model of Energy Substitution Responses," The in sectoral coverage of the sample. Our study only examines the American Economic Review 66 (Dec. 1976), 845-857. manufacturing sector, whereas Bowen's figure seems to include Heller, Peter S., "Factor Endowment Change and Comparative other sectors such as agriculture and mining. Capital formation Advantage: The Case of Japan, 1956-1969," this RE- in those sectors in Korea in those periods might have been viEw 58 (Aug. 1976), 283-292. higher than t,.at in Japan. Hong, Wontack, Factor Supply and Factor Intensily of Trade in t3Heller (1976) found this relation in his analysis of the Korea (Seoul: Korea Development Institute, 1976). Japanese trade between 1955 and 1968. However, Heller did Johrston, John, Econometric Methods (New York: McGraw- not consider the changes in factor usage of Japan's trading Hill, 1972). partners explicitly. Bowen (1981) also found this association Jorgenson, Dale W., and Mieko Nishimizu, "U S. and Japanese between trade structure and resource structure from a sample Economic Growth, 1952-1974: An International Com- of thirty-four countries. parison," Economic Journal 88 (Dec. 1978), 707-726. 684 THE REVIEW OF ECONOMICS AND STATISTICS Leamer, Edward E., and Harry P. Bowen, "Cross-Section Tests Stern, Robert M., "Testing Trade Theories," in P. B. Kenen of the Heckscher-Ohlin Theorem: Comment," The (ed), International Trade and Finance (Cambrie'ge: Cam- American Economic Review 71 (Dec. 1981), 1040-1043. bridge University Press, 1975), 3-49. Leontief, Wassily W., "Domestic Production and Foreign Stern, Robert M., and Keith E. Maskus, "Determinants of the Trade: The American Capital Position Re-examined," Structure of U.S. Foreign Trade," Journal of Interna- Proceedings of the Americatn Philosophical Society 97 ional Economics 11 (May 1981), 207-224. (Sept. 1953), 332-349. Tatemoto, Masahiro, and Shinichi Ichimura, "Factor Propor- Maskus, Keith E., "Evidence on Shifts in the Determinants of tions and Foreign Trade: The Case of Japan," this the Structure of U.S. Foreign Trade," Discussion Paper REVIEW 41 (Nov. 1959), 442-446. No. 159, University of Colorado (1981). No. 278, Oli Havrylyshyn and Martin Wolf, "Recent Trends in Trade among Developing Countries," European ELo IIoI ii Reviezzw No. 279. Nancy Birdsall, "Fertility and Economic Change in Eighteenth and Nineteenth Century Europe: A Comment," PTopiluhfio', and Developnnent Reviewv No. 280. Walter Schaefer-Kehnert and John D. Von Pischke, "Agricultural Credit Policy in Developing Countries," translated from Handbuchti der Landwvirtschaft uitind Ernliihrtok' it den1 Enlvicklungsliindern (includes original German text) No. 281. Bela Balassa, "Trade Policy in Mexico," Vvbrld Dev,elopment No. 281a. Bela Balassa, "La politica de comercio exterior de M6xico," Comercio Exterior No. 282. Clive Bell and Shantayanan Devarajan, "Shadow Prices for Project Evaluation under Altemative Macroeconomic Specifications," The Quatirterli, Jouirnial of Econolics No. 283. Anne 0. Krueger. "Trade Policies in Developing Countries," Handlbook of International Econotmiic-s No. 284. Anne 0. Krueger and Baran Tuncer, "An Empirical Test of the Infant Industry Argument," American Economic Review No. 285. Bela Balassa, "Economic Policies in Portugal," Economia No. 286. F. Bourguignon, G. Michel, and D. Miqueu, "Short-run Rigidities and Long-run Adjustments in a Computable General Equilibrium Model of Income Distribution and Development," Journal of Development Econonuics No. 287. Michael A. Cohen, "The Challenge of Replicability: Toward a New Paradigrn 'or Urban Shelter in Developing Countries," Regional Development Dialogue No. 288. Hollis B. Chenery, "Interaction between Theory and Observation in Development," tAjorld De-eloneloent No. 289. J. B. Knight and R. H. Sabot, "Educational Expansion and the Kuznets Effect," Thle Amnerican Economiic Revieuw No. 290. Malcolm D. Bale and Ulrich Koester, "Maginot Line of European Farm Policies," Thie W'vorld Econontul No. 291. Danny M. Leipziger, "Lending versus Giving: The Economnics of Foreign Assistance," World Development No. 292. Gregory K. Ingram, "Land in Perspective: Its Role in the Structure of Cities," World Co1kgress ont Lanid Policu, 1980 No. 293. Rakesh Mohan and Rodrigo Villamizar, "Tihe Evolution of Land Values in the Context of Rapid Urban Growth: A Case Study of Bogota and Cali, Colombia," World Congress on Lanzd Pblicy, 1980 No. 294. Barend A. de Vries, 'Intemational Ramifications of the Extemal Debt Situation," Tlhe ANIEX Bank Revliew Special Papers No. 295. Rakesh Mohan, "The Morphology of Urbanisation in India," Economic and Political VWeekltl No. 296. Dean T. Jamison and Peter R. Moock, "Farmer Education and Farm Efficiency in Nepal: The Role of Schooling, Extension Services, and Cognitive Skills," World Developmenleti No. 297. Sweder van Wijnbergen, "The 'Dutch Disease': A Disease after All?" The Economic Journal No. 298. Ame Drud and Wafik M. Grais, "Macroeconomic Adjustment in Thailand: Demand Management and Supply Conditions," Journal of Rolicyu X fodelilig Issues of the World Bank Reprint Series are available free of charge from the address on the bottom of the back cover.