ABSTRACT NATIONAL CHARACTERISTICS AND THE COMMODITY COMPOSITION OF TRADE IN MANUFACTURED GOODS By Sohrab Behdad This study is a test of the Heckscher-Ohlin theory and alternative theories of international trade. By applying the most recent input-output tables and the industry input coefficients of the United States, the United Kingdom, and South Korea to their bilateral trade structure, it is found that the Heckscher- Ohlin theory is capable of explaining the commodity composition of trade among deve10ped and less developed economies. This result becomes more pronounced when trade only in manufactured goods is considered. But the theory fails to predict the pattern of trade between -developed and less developed economies. Similarly, bilateral trade of the same three countries are examined to test the human capital approach and neotechnology theories of trade. The results indicate that while these theories can explain commodity flows among deve10ped economies, they all fail in their attempt to explain trade among less deve10ped countries. Sohrab Behdad The existence of a systematic relationship between national characteristics and the commodity com— position of trade is tested for the manufacturing trade of twenty-three countries, according to each hypothesis. It is found that with the exception of the scale economies hypothesis all theories perform satisfactorily. However, among them the human capital approach, the stage of production theory, and the product cycle theory indicate the strongest showing. Considering the results of the bilateral tests and the regression analysis of the trade of twenty-three countries, the study concludes that the various existing theories are useful in explaining different segments of international trade flows. The Heckscher-Ohlin theory may be considered relevant for explaining trade among less deve10ped countries, while the product cycle and human capital models are best suited for predicting trade among developed countries and between deve10ped and less developed economies. NATIONAL CHARACTERISTICS AND THE COMMODITY COMPOSITION OF TRADE IN MANUFACTURED GOODS By Sohrab Behdad A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Economics 1973 «*6 00’} ACKNOWLEDGEMENTS (9‘ I wish to express my deepest gratitude to Professor Mordechai Kreinin whose continuing encouragement along with his insightful criticisms have been invaluable throughout my graduate program and at all stages of this study's development. Sincere thanks are extended to Professors Anthony Koo and Paul Strassmann, members of my thesis committee for their valuable comments and criticisms. ii Chapter TABLE OF CONTENTS INTRODUCTION. HECKSCHER- OHLIN THEORY OF INTERNATIONAL TRADE The Theory and its Assumptions. The Relative Price Definition of Factor Abundance. The Relative Supply Definition of lactor Abundance. . . . . Definition of Factor Abundance and the Demand Assumption Empirical Verification of the H-0 Theory: Leontief's Tests . Other Tests of the H-0 Theory Reexamination of the H-0 Theory and itsChitical Assumptions . . Factor Intensity Reversal. Production Functions and Factor Qualities O O O O O O Identical Demand. Natural Resources Trade Barriers . . . . . . ALTERNATIVE THEORIES OF TRADE Introduction . . . . . . . iii Page 10 IS 17 25 27 28 36 39 41 43 50 SO C11a[)t(:r 2.2 Ncofactor Theories: The Human Capital Approach Neotechnology Theories Scale Economies. Stage of Production. Technological Gap. Product Cycle. An Evaluation of Alternative Theories of Trade EMPIRICAL TEST OF THE HECKSCHER-OHLIN THEORY: A BILATERAL TRADE STUDY . Introduction Methodology. The Commodity Composition of U. 8. Trade . . . . . . . . . The Commodity Composition of U.K. Trade . . . . . . . . . The Commodity Composition of South Korean Trade. . . . . . Conclusion. THE HUMAN CAPITAL THEORY OF INTERNATIONAL TRADE: A BILATERAL TRADE STUDY . . . . . . . Introduction and Methodology. The Commodity Composition of U. 8. Trade . . . . . . . . . . . . The Commodity Composition of U. K. Trade . . . . . . . iv Page 51 58 59 61 62 66 67 74 74 75 80 87 92 .99 105 105 109 112 Chapter 4. 7 APPENDIX 4 The Commodity Composition of South Korean Trade Conclusion . NEOTECHNOLOGY THEORIES OF INTERNATIONAL TRADE: A BILATERAL TRADE STUDY . - - - - ' ° - Introduction. Scale Economies Stage of Production Technological Gap Product Cycle . . . Conclusion. THEORIES OF INTERNATIONAL TRADE: A STUDY OF TRADE PATTERN OF TWENTY- THREE COUNTRIES . . . Introduction Selection of Countries Methodology. Results and Conclusion SUMMARY AND CONCLUSIONS~ BIBLIOGRAPHY.. Page 115 117 121 121 121 135 140 146 147 157 157 157 158 162 175 190 225 No. LIST OF TABLES Domestic Capital and Labor Requirements Per Million Dollars of U.S. Exports and Competitive Import Replacements (Of Average 1947 Composition)- - - Keesings Estimate of Skill Intensity of U. S. Trade (1957) . . . . . . Capital/Labor Endowment Ratio of Selected Developed Economies (1964). . . Capital and Labor Requirements Per Million Dollars of U.S. Exports and Competitive Import Replacements (All Commodities, All Inputs Included). . . . . . . . . . Capital and Labor Requirements Per Million Dollars of U.S. Exports and Competitive Import Replacements (Inputs of Service Sectors are Excluded) Capital and Labor Requirements Per Million Dollars of U.S. Exports and Competitive Import Replacements (Non-Manufacturing Trade is Excluded). Capital and Labor Requirements Per Million Dollars of U.S. Exports and Competitive Import Replacements (Non-Manufacturing Inputs and Trade are Excluded) Capital/Labor Endowment Ratio of Selected Developed Economies (1964). Capital and Labor Requirements Per Million Dollars of U.K. Exports and Competitive Import Replacements (All Commodities, A11 Inputs Included). . . . . . . . . . . . . Capital and Labor Requirements Per Million Dollars of U.K. Exports and Competitive Import Replacements (Inputs of Service Sectors are Excluded). vi Page 18 57 82 83 85 86 88 91 93 94 No. 3-10 4—3 4-4 Capital and Labor Requirements Per Million Dollars of U.K. Exports and Competitive Import Replacements (Non-Manufacturing Trade is Excluded). Capital and Labor Requirements Per Million Dollars of U.K. Exports and Competitive Import Replacements (Non-Manufacturing Inputs and Trade are Excluded). Capital/Labor Endowment Ratio of Selected Less Developed Countries (1964) Capital and Labor Requirements Per Million Dollars of South Korean Exports and Competitive Import Replacements (All Commodities, All Inputs Included). Capital and Labor Requirements Per Million Dollars of South Korean Exports and Competitive Import Replacements (Non- Manufacturing Inputs and Trade are Ex- cluded). . . Skill Endowment Ratio of Selected DeveIOped Economies (1961- 1964) . . . . . Coefficients of Relative Skill Intensity of Manufacturing Trade of the United States Skill Endowment Ratio of Selected DeveIOped Economies and Economic Regions (1961-1964) Coefficients of Relative Skill Intensity of Manufacturing Trade of the United Kingdom. Skill Endowment Ratio of Selected Less Developed Countries and Economic Regions (1961-1964) . Coefficients of Relative Skill Intensity of Manufacturing Trade of South Korea . Total Manufacturing Output and Gross Domestic Product Per Capita for Selected DeveIOped Countries (1964) . . . . . . . . Total Manufacturing Output and Gross Domestic Product Per Capita for Selected Less Developed Economies (1964) vii Page 95 96 ’98 100 .'101 110 111 113 114 116 118 126 127 No. 5-9 5-13 The Coefficients of Scale Economies of U.S. Exports and Imports of Manufactured Commodities. The Coefficients of Scale Economies of U.K. Exports and Imports of Manufactured Commodities. The Coefficients of Scale Economies of South Korean Exports and Imports of Manufactured Commodities The Coefficients of Consumer Goods Ratio of U.S. Exports and Imports of Manufactured Commodities The Coefficients of Consumer Goods Ratio of U.K. Exports and Imports of Manufactured Commodities. The Coefficients of Consumer Goods Ratio of South Korean Exports and Imports of Manufactured Commodities. The Product Age Coefficients of U.S. Exports and Imports of Manufactured Commodities. . . . . The Product Age Coefficients of U.K. Exports and Imports of Manufactured Commodities. The Product Age Coefficients of South Korean Exports and Imports of Manufactured Commodities. . . . . The Product Differentiation Coefficients of U.S. Exports and Imports of Manufactured Commodities. . The Product Differentiation Coefficients of U.K. Exports and Imports of Manufactured Commodities. . . . . . . . . The Product Differentiation Coefficients of South Korean Exports and Imports of Manufactured Commodities. Summary of Performance of Neotechnology Theories of Trade: Trade Flows Inconsistent With the Predicted Pattern. viii Page 129 131 .132 .138 139 141 143 144 145 148 149 150 154 National Characteristics. Commodity Composition of Trade in Manufactured Goods of Twenty-Three Countries According to the Heckscher- Ohlin Theory and the Alternative Theories of Trade. The Measures of Commodity Composition of Trade and the Determining National Characteristic According to the H-0 Theory and the Alternative Theories of Trade. Simple Correlations Between the Coefficients of Commodity Composition of Trade for Twenty-Three Countries. Simple Correlations Between National Characteristics of Twenty-Three Countries. Summary of Performance of Theories of Trade: Bilateral Trade Flows. List of Non-Manufactured Commodities. Capital and Labor Coefficients Per Million Dollars of Value Added for U.S. Industries (1963). . . . . . . . . . . . . . . . . . ; Capital and Labor Coefficients Per Million Dollars of Value Added for U.K. Industries (1960). . . . . . . . . . . . . Capital and Labor Coefficients Per Million Dollars of Value Added for South Korean Industries (1966). . . . . . . Skill Requirements Per Million Dollars of Value Added for U.S. Industries (In Man Years). . . . . . . . . . . . Skill Requirements Per Million Dollars of U.S. Manufacturing Exports and Competitive Import Replacements. . . Skill Requirements Per Million Dollars of Value Added for U. K. Industries (In Man Years). . . . . . . . . . . . ix Page 159 163 165 170 171 180 190 191 194 195 197 199 202 No. Page A-8 Skill Requirements Per Million Dollars of U.K. Manufacturing Exports and Competitive Import Replacements. . . . . . . . . . . . . 203 A-9 Skill Requirements Per Million Dollars of Value Added for South Korean Industries (In Man Years). . . . . . . . . . . . . . . . 206 A-10 Skill Requirements Per Million Dollars of South Korean Manufacturing Exports and Competitive Import Replacements. . . . . . . 208 A-Il The Technology Coefficients of Commodities. . 210 A-12 Estimates of Optimum Plant Size in U.S. Industries (1954). . . . . . . . . . . . . . 216 A-l3 Skill Requirements Per Million Dollars of Manufacturing Exports and Imports of Twenty-Three Countries. . . . . . . . . . . 218 A-14 Capital and Labor Embodied in One Million Dollars of Manufacturing Exports and Imports of Twenty-Three Countries. . . . . . 222 A-IS Scale Intensity of Manufacturing Trade of Twenty-Three Countries. . . . . . . . . . . . 223 A-I6 Consumer Goods Ratio, Coefficient of Product Age, and Product Differentiation Index Embodied in One Million Dollars of Experts and Imports of Manufactured Goods for Twenty-Three Countries. . . . . . . . . . 224 No. 1-1 2-1 LIST OF FIGURES The Relative Price Definition of Factor Abundance. 'I‘he Relative Supply Definition of Factor Abundance. . . . . . . . . Factor Intensity Reversal. Different Production Functions Between Countries. Imitation Lag and the Technological Gap Theory. xi Page 12 33 38 65 INTRODUCTION This study is an attempt to test empirically the Heckscher-Ohlin theory and alternative theories of inter- national trade. International trade is brought about by differ- ences in the relative prices of commodities. "It is the inequality as to the relative prices in isolation," stated Ohlin, "that is a necessary condition for establish- ment of trade” [5, p. 7]. But the main body of the "pure" theory of international trade seeks to determine the predominant factor or set of factors which are responsible for the international differences in relative prices. There are many variables in a country's economic structure which have a bearing on that country's comparative advantage. As Kuznets pointed out: "Foreign trade flows . are affected by many complex factors in which techno- logical changes, social inventions, economic advantages, political revolUtions and diversities in the structure and endowment of nations all play their part" [2, p. 106]. Yet a theory of trade must be capable of determining the exact relationship of these or any other variables with the comparative advantage of a country within a systematic and logically consistent framework. 1 2 The Ricardian theory of trade singled out rela— tive labor productivity among countries in the production of different commodities as the predominant force in determining comparative advantage. According to this theory, pre-trade commodity price ratios in each country are determined by the relative average productivity of the factor of production (which to Ricardo was labor but could be any other factor as well) in producing different commodities. However, the Ricardian theory provides no explantion of what accounts for the differences in average labor productivity. The Heckscher-Ohlin (H-O) theory of trade attempts to explain the commodity composition of international trade throngh differences between the relative endowment of fac- tors of production among countries and the relative factor intensity of traded commodities. As introduced by Heckscher [1] and later elaborated by Ohlin [2], the theory considers other forces as secondary in nature and rarely strong enough to change the direction of the predominant force of the relative factor endowment. Thus, on theoretical grounds the H-0 theory assumes no variations in production functions among countries. In this respect the H-0 theory is diametrically Opposed to the Ricardian explanation of trade. For almost two decades the H-0 theory enjoyed the full respect of economists not only because of its fresh insight into the cause of comparative advantage but 3 also because of the simplicity and clarity of its logic which could lend itself to the neoclassical analytical framework. However, several empirical studies of the theory, beginning with Leotief's well-known tests of the structure of United States trade [3], indicated the limitation of the H-0 theory's explanation of the com- modity composition of world trade. The failure of the H-0 model to survive empiri- cal tests has led during the past several years to the development of new alternative theories for explaining the commodity composition of international trade. These new theories have undertaken two distinctly different paths. One group, the neofactor theories, attempts to reformulate the traditional version of the H-0 theory by modifying the concept of factors of production through the inclusion of factor qualities. The second group, the neotechnology theories, criticizes the assumption of similarity of production functions among countries in the H-0 theory and seeks to explain the commodity com- position of manufactured trade through differences in the technological capabilities of countries. However, despite their plausible assumptions, the alternative theories, especially those of the second group, lack a rigorous theoretical framework similar to the H-0 model. Through a reconsideration of the H-0 theory this study will attempt to show that the factor preportion explanation of trade may be acceptable under certain 4 restricted conditions. It is hypothesized that trade structure will tend to conform with the ”-0 theory more strongly between countries that are not widely different in their level of economic development. By further restricting the theory to an explanation only of the pattern of manufacturing trade, an even better performance of the theory is expected. This hypothesis is based on the presumption that the H-0 theory's critical assump- tions of identical demand pattern, similar production functions, and unique factor intensity ranking of com- modities will tend to hold more strongly under the above- state conditions. The above hypothesis, along with alternative theories of trade, will be tested in a comprehensive study of trade patterns of the United States (1970), the United Kingdom (1969), and South Korea (1969). The second part of the study is an analysis of the trade pattern of twenty-three countries according to the H-0 theory and alternative explanations of compara- tive advantage. The existence of a significant and systematic relationship between the commodity composition of trade and the national characteristics of each country is tested in order to determine the explanatory power of each theory. [1] [Z] [3] [4] [5] REFERENCES Heckscher, E. ”The Effect of Foreign Trade on the Distribution of Income." Economisk Tidskirft. 1919, pp. 497-512. Kuznets, S. Six Lectures in Economic Growth. Glencoe, IllinEis: Free Press, 1959. Leontief, W. ”Domestic Production and Foreign Trade, The American Capital Position Re-examined.” Economia Internazionale. February, 1954, pp. 9-150 "Factor Proportion and the Structure of American Trade: Further Theoretical and Empirical Analysis." Review of Economics and Statistics. November, 1956, pp. 387-407. Ohlin, B. Interregional and International Trade. Revised edition. CamBridge,TMass.: Harvard University Press, 1967. Chapter 1 HECKSCHER-OHLIN THEORY OF INTERNATIONAL TRADE 1.1 The Theory and its Assumptions Postulating a simple world of two countries which produce two commodities with two factors of pro- duction, the H-0 theory states that each country will have a comparative advantage in producing the commodity that uses the country's relatively abundant factor relatively intensively. The model is based on the following assumptions:* 1. Perfect competition exists in the factor and commodity markets. 2. Capital and labor (factors of production) are qualitatively identical in both countries. 3. There is perfect mobility of factors within the country but no inter-country factor movement. 4. Supply of factors are given and are fully .employed (full wage-price flexibility). 5. Production functions have the following characteristics: i. They are homogeneous of the first degree (constant return to scale); *These assumptions were first explicitly stated by Samuelson [28]. 7 ii. They are identical for the same commodity in both countries; iii. They exhibit different and unchanging factor intensities for different com- modities for all possible relative factor prices (no factor reversal); and iv. The law of diminishing marginal pro- ductivity holds. 6. No trade barriers or transportation cost exit. Under the above assumptions each country will export the commodity which uses the country's relatively abundant factor more intensively. It can be shown that the results follow from the above assumptions under the alternative definitions of "factor abundance": They are defined firstly in terms of relative factor prices (in the Ohlin sense [24, p. 7]), and secondly in terms of relative factor supplies in each country (as Jones suggests [13].) Subsequently, in section 1.2.3. the logical bases for choosing one definition over another will be introduced along with the problems that may arise by accepting the more logical definition. 1.2.1 The Relative Price Definition of Factor Abundance Given two countries, I and II, with country I relatively abundant in capital and country II in labor, and two commodities, X and Y, being capital and labor intensive, reSpectively. Defining relative factor abun- dance in terms of relative factor prices, country I is capital abundant if (Pk/Pl)2 >(Pk/P1)1, where Pk and P1 are capital and labor prices respectively, while the 8 subscripts 1 and 2 refer to the two countries. It must be shown that country I has a comparative advantage in the production of X, and country 11 in the production of Y, which would be manifested in a lower relative price of X to Y in country I and the opposite for country II. In Figure 1-1 let XX and YY represent the isoquants for commodities X and Y. By the homogeneity assumption any one of the isoquants for either commodity can represent the family of isoquants for that commodity, and because of the assumption of identical production functions in both countries, Figure 1-1 represents the relevant production conditions in both countries. The only matter of distinction between the two countries will be inequality of factor price ratios. As the figure indicates PO and CC' (which is parallel to DD') are the relevant factor price ratios in countries I and II, respectively. It is postulated that capital in country I and labor in country II are the relatively cheaper factor. From this it should follow that country I enjoys a comparative advantage in pro- duction of X, the capital intensive commodity, and coun- try II in the production of Y, the labor intensive com- modity. To show this a common measure of production costs for both commodities in the two countries can be deve10ped by converting the cost of production of each commodity expressed in terms of labor and capital inputs into its equivalent in terms of only one factor, say capital. \‘\ I .\\ \ K \ ‘~1|\\\ I , I kr- ---_._-_~_\__+___ \_ v (E ‘ \NK : _X 1.} __ . , k —~ .......... | ‘ I I I] I _________ _I__-_-1|__ _ k ' H I Y i I I I I I II I D I I: | i ll I t SI. V1. “1 Q 1 Figure 1—1 The Relative Price Definition of Factor Abundance 10 In country I, the production of a given quantity of X requires Otk of capital and Otl of labor. Ot1 of labor is equivalent to Ptk of capital. Therefore, the total cost of producing X in country I, with capital as the unit of account, is Ptk + Otk = OP. Similarly, for country II production of the same quantity of X requires ka of capital, and Ov1 of labor, Cvk + ka = OC units of capital. As a result the cost ratio of X between countries 11 and I is %%%%. In the same manner the relative cost of Y between the two countries will be (OD)/(OP). It can be seen from Figure 1-1 that OC/OPI>OD/OP, indicating that country 11 has a comparative advantage in production of Y and country I in production of X, the commodities which use the coun- tries' relatively cheaper factor more intensively. 1-2-2 The Relative Supply Definition of Factor Abundance Alternatively, factor abundance may be defined in terms of the relative supply of factors of production. By this definition country I is capital abundant and country 11 is labor abundant if Kl/L1:>Kz/L2, where K and L denote the physical supply of capital and labor in countries I and II, as indicated by the subscripts. Our task is to show that this difference in relative factor endowment will result in the comparative advantage of country I in the production of X, the capital intensive good, and country 11 in the production of Y, the labor ll intensive good. The Edgeworth-Bowely box diagram will be utilized in this demonstration. Consider once again countries I and II, each producing commodities X and Y, with factors K and L. In Figure 1-2 boxes I and II indicate the total supply of K and L in countries I and II. The origin of X iso- quants is at point Ox for both countries and of the Y isoquants is at Qy for country I and at Q; for country 11. Along any ray from the origin such as QXTS the ratio of marginal physical product of capital and labor in the production of X must be equal in both countries, if the production functions are homogeneous of the first degree and identical in both countries. Furthermore, the ratio of marginal physical products of both factors in the production of X and Y must be equal at points similar to S and T since these points are on the contract curves of countries I and II. It follows that the ratios of mar- ginal physical product for both commodities are the same at points 8 and T along any ray such as OXTS. As a result, rays QyS and Q§T must be parallel, remembering that the Y production functions are also homogeneous of the first degree as in the case for X. To prove that points similar to S and T along the rays from the origin are possible post-trade pro- duction equilibrium points, it must be shown that at these levels of production the relative commodity prices are equal in both countries [16]. 12 /’F /// / // / // / /"/ / /// / II Q; ,/’ // ’ / / / / / / / / / /Q y x Y s x Y I x Y K Figure 1-2 The Relative Supply Definition of Factor Abundance 13 Let L'XI and L'XII denote marginal productivity of labor in the production of X in countries I and II. Similarly, let L'YI and L'YII stand for marginal pro- ductivity of labor in the production of Y in the two countries. Recalling that under the assumption of per— fect competition the value of the marginal product of labor is equal in both industries in each country, the following relationships are obtained: PX ° L'X = PY ‘ L'Y (Ia) _or (PXI/PYI) = (L'YI/L'XI) (1b) where PXI and PYI are the prices for commodities X and Y in country 1. Similarly for country II: PX ' L'X = PY ' L'Y 23 II II II II ( ) 01" (PXII/PYII) = (L'YII/L'XII). (2b) However, because of the constant return to scale assumption the marginal productivity of any one factor is constant at all production points along any ray from the origin, such as OXTS. Hence at points T and S the fol- lowing equalities hold: L'X = L'X 3 I II ( ) 14 and L'Y = L'Y - (4) From relations (1), (2), (3), and (4) it can be concluded that; PX PY = PX PY 5 I/ II II/ 11 ( ) In other words, points of intersection of any ray from the origin OX with the two contract curves (such as S and T) are proved to be possible post-trade production equilibrium points for both countries. Furthermore, at any equilibrium point such as S and T, country I produces relatively more of commodity X and country II more of commodity Y. To show this in Figure 1-2 draw Q'yG parallel to OXS to intersect the extension of SQy at point G. Given that SQy is parallel to Q'yT, Q'yT is equal to 86. However, because OXS>>OXT, it follows that Q'yT/OXT,>GS/OXS (6) yet QyS is clearly smaller than GS; therefore Q yT/OXT >QyS/OXS (7) The above analysis shows that when the relative endowment of countries is defined in terms of the physical supply of factors of production, comparative advantage will be revealed in terms of the physical supply of com- modities. That is, each country will produce more of the 15 commodity that uses the abundant factor more intensively. In these terms country I has comparative advantage in production of X and country II in production of Y in the above example. 1.2.3 Definition of Factor Abundance and the Demand_Assumption It is possible for the comparative advantage in production of a commodity (as demonstrated above) to be nullified by a strong internal demand preference for that commodity. For example, although a higher ratio exists in the production of capital intensive to labor intensive goods in the capital rich country, a strong demand for the capital intensive commodity may make the relative price of this commodity higher in the capital abundant country. In other words, it is possible for the comparative advantage indicated by the factor endow- ment ratio to be reversed by a strong demand bias in the opposite direction. Therefore this version of the model requires the additional assumption that the demand patterns are similar in both countries. Although Ohlin admitted [24, p. 10] the existence of dissimilar demand patterns among countries, he maintained that such differences are small enough not to reverse the logical outcome of the model. However, even equality of demand functions is not sufficient for the logical consistency of the H-0 theory. The demand functions must also be such that differences in income 16 levels would not affect the demand patterns. This condi— tion is satisfied if it is further assumed that demand functions are homogeneous of the first degree [25], that is, the income elasticity of demand is equal to unity for each commodity in both countries. But such assump- tion may or may not be consistent with reality. If it is not, then the pattern of trade predicted by the factor endowment ratios may not be realized. Alternatively, the relative price definition of factor abundance implies that the commodity which uses more of the relatively cheaper factor in a country will have a lower relative price in that country as compared to the second country. Although this interpretation of factor abundance does not require an explicit assumption as to the demand condition, it falls short of a meaningful explanation of the trade pattern. What the theory states is nothing more than the unique Samuelsonian relation between factor price and commodity price rations [27,28]. Thus even though the relative price definition of factor abundance is logically true, it is "trivial," as Jones put it, in the sense that all it can explain is that "each country experts that commodity in the production of which it enjoys a comparative advantage" [13, p. 4], without really showing that the factor endowment ratio alone is responsible for the comparative advantage. Therefore the physical definition of factor abundance is to be considered the satisfactory explanation 17 of the pattern of trade, even if it may be invalidated by perverse demand conditions. It is this definition of factor abundance that is relied upon on all empirical verification of the theory. 1.3.1 Empirical Verification of the H-0 Theory: Leontief's Tests The first* comprehensive test of the factor proportion theory was carried out by Leontief [18] through the application of interindustry input-output relations. This method consisted of measuring the total (direct and indirect) capital and labor requirements of one million dollars of exports and of competitive import replacements (those imports that are domestically produced) in the United States. According to the H-0 model, it is expected that the United States, the most capital abundant country in the world, would engage in international trade to compensate for its scarce factor, labor, and would enjoy a comparative advantage in the production of capital intensive goods. However, Leontief found that a represen- tative one million dollar basket of U.S. exports requires a smaller capital/labor ratio than its equivalent amount of competitive import replacements, Table 1-1. Clearly, * Although MacDougal [21] and Kravis [14] attempted to test the H-0 theory their results cannot be considered conclusive because of the inadequacy of the methods which they applied. 18 Table 1-1 Domestic Capital and Labor Requirements Per Million Dollars of U.S. Exports and Competitive Import Replacements (of Average 1947 Composition) Capital 00 Dollars, 1947 prices Exports ....... 2,550,780 Imports ....... 3,091,339 Labor (1) Man years 182,313 170.004 k/l 13,992 18,184 (k/l) 1mp. 1.300 (k/l) exp. * I O 3'13 refered to as the coeff1c1ent of factor intensity of trade. Source: Columns 1 and 2 from W. Leontief, "Domestic Production and Foreign Trade, the American Capital Position Re-examined," Econom1a Internazionale. February, 1954. #- 19 this result is contrary to the H-0 prediction of the trade pattern, for it indicates that the U.S. specializes in the export of labor intensive rather than capital intensive commodities. As Leontief puts it: "This country [U.S.] resorts to foreign trade in order to economize its capital and dispose of its surplus labor, rather than vice versa" [18, p. 25]. This is the so-called paradox by which is meant a seeming contradiction between given facts and expected results. The Leontief test generated a strong wave of criticisms based on statistical and methodological grounds as well as an intense reevaluation of the theory and its assumptions. On statistical grounds the major reservation about the test was the method of estimating the input coefficients, especially with regard to capital coeffi- cients. Diab [8] considered Leontief's estimates of the capital coefficients for agriculture too high. But after reducing them to one half Leontief's estimates he obtained the same results. Leontief's capital coefficients are Open to another, more substantial criticism. As Buchanan [6] pointed out, instead of employing the estimates of the service of capital required per unit of output, Leontief used the estimates of the amount of investment in capital goods per unit of output. However, the investment coeffi- cients can be regarded as a proxy for capital (stock) coeff- icients only if the durability of capital in all industries are identical, Which is an implausible assumptiOn. On this 20 basis the capital coefficients employed by Leontief suffer from a strong deficiency. Furthermore, some reservations were expressed by Swerling [31] as to the selection of the year upon which the study is based. He points out that 1947 trade was ”highly atypical of world trading relations in general and of United States agriculture trade in particular." By that year, ”postwar disorganization of production overseas had not yet been corrected . . ., and close to half of United States exports were financed by grants and credits under various foreign aid programs.” In his second study, Leontief [19] considered most of the many criticisms that were directed toward his earlier test. He applied the 1951 trade flow of the United States to the 1947 input-output table of the industries. However, the result was not significantly different from what he obtained in the first study. On methodological grounds the main thread of criticism asks two questions: First, is the input-output method apprOpriate for the analysis of international problems, and second, is the extension of U.S. conditions of production to the estimation of factor requirements of U.S. imports a correct method of analysis of trade pat- terns? The difficulty with the application of input- output analysis, critics maintain, is in the fixed coeffi- cient characteristics of the input-output models which 21 imply that the input ratio for the production of each good is fixed and invariant with respect to the relative prices of inputs. Thus input-output models, Valavanis-Vail [37] argues, "except for rare luck are logically incompatible with international trade.” That is so because "we cannot be sure that every factor will be completely used up, ei- ther with trade or in its absence.” If each country em- ploys all of its factors fully before trade, there will be no assurance that the same will be true after a change in the world's composition of outputs, as the result of trade. Although the question raised by Valavanis-Vail is valid, it is irrelevant to Leontief's study. Leontief utilized the input-output tables for estimating the in- direct factor requirements of each product under the existing equilibrium condition [5, p. 177]. The other criticism of the fixed coefficient characteristics of the model is the view expressed by Ford [10, p. 67] that in the input-output models the input coefficients are given regardless of the output level of the industry. This implies that factor require- ment remain the same for a preportionate decrease or in- crease of one million dollars of exports and import com- peting goods. However, Ford maintains that ”an industry could be labor intensive for most increases in output and yet when it has to increase by its contribution to the one million dollars of trade it could be easily capital intensive." In other words, factor intensity reversal 22 may occur with scale changes. This, however, is incon- sistent with the homogeneity assumption of the H-0 theory. For factor intensity reversal to occur the isoquants must change their relative position with respect to the two factors of production as the level of production changes. But the characteristic of the homogeneous production functions is that they expand along the rays from the origin, that is, the marginal rate of technical substitu- tion is equal along such rays for each family of isoquants which cannot be true if isoquants change their relative positions. This criticism, therefore, is not valid on theoretical grounds. Besides, Leontief does not change the level of output but only estimates the factor require- ments for an average bundle of exports and import com- peting goods [5, p. 177]. Extension of the American condition of produc- tion to U.S. imports constitutes the second and potentially the most important methodological criticism of Leontief's tests. He employed factor input coefficients along with the interindustry input-output relations of the United States to estimate the factor requirements of a one million dollar bundle of that country's imports. This in Leontief's terminology is called the factor require- ment for import-competing industries for supplying, do- mestically, a one million dollar reduction of imports. It is argued by Ford [10,pp. 58-61], Swerling [31], and Elsworth [9] that in comparing exports and imports one 23 should take the actual factor requirements in the country where the product is being produced, instead of applying the same condition of production for both exports and imports. In an attempt to reconcile the H-0 theory with Leontief's results, Elsworth asserts that it is possible that the United States exports capital intensive goods (relative to its imports) and, at the same time in pro- ducing import replacements, also uses capital intensive methods. This type of argument may be interpreted to mean that the conditions of production in the United States and the rest of the world are such that factor intensity reversal occurs. This factor reversal may be caused either by widely different factor price ratios between two trading countries or by variations in pro- duction functions between two countries. But both these arguments are in direct violation of the assumptions of the H-0 theory, and if Leontief's test is a test of the “-0 theory then such methodological criticisms are in- valid. However, it can be argued that factor reversal, for either of the above reasons, is an empirical reality, and if so, we are outside the framework of the H-0 theory. These possibilities will be discussed in sections 1.4.2 and 1.4.3. In an attempt to salvage the theory, Leontief reconciled the paradox by redefining the factor endow- ment of the United States. He asserted [18,pp. 25-29] 24 that although it may appear that the capital/labor endowment ratio of the United States is higher than that of any other country in the world, the high efficiency of American labor makes this country a relatively labor abundant one. To be exact, he claims that this effi— ciency factor is about three. That is, using the same amount of capital, one man year of American labor is equivalent to three man years of foreign labor. This relative effectiveness of American labor is attributed to "entrepreneurship, superior organization, and favorable environment" [18, p. 29]. ILeontief, however, does not explain how he arrived at the efficiency factor of three or why the efficiency factor is attributed to labor only and not to capital. But this claim cannot be supported empirically. A study by Kreinin [15] shows that this efficiency factor is about 1 1/5 or at most 1 1/4, which fails to support Leontief's claim or his attempt in reconciling the paradox. Furthermore, Arrow, Chenery, Minhas, and Solow [1], in their inter-country study of production functions, have found significant differences only in over-all efficiency, and therefore even if Ameri- can techniques of production are more efficient, this situation cannot be attributed only to labor. Most impor- tantly, however, Leontief's explanation is theoretically unsatisfactory. As Robinson pointed out: ”A comparative advantage theory based on relative factor endowment cannot seek explanation in productivity concept by redefining 25 the units of measurement of factors of production, when it fails in what it had been trying to explain" [25]. 1.3.2 Other Tests of the H—0 Theory Tests similar to that of Leontief pertaining to the trade of Japan [33], Canada [40], West Germany [26], East Germany [30], India [6], and the United Kingdom [11] yielded mixed results. Bharadwaj's study of India indi- cated that India's exports to the United States are more capital intensive than their competitive import replace- ments from that country. However, he explained this nega- tive result by the atypical nature of Indian trade in 1951, which is characterized by unusually high imports of agri- cultural goods from the United States. Tatemoto and Ichimura [33] attributed the high capital/labor ratio embodied in the exports of Japan as compared to its import replacements to the high concentration of that country's trade with underdeveIOped Asian countries which are pre- sumed to have a lower capital/labor endowment ratio than Japan. They concluded that the Japanese trade pattern is consistent with the H-0 theory, especially since the capital/labor ratio embodied in Japanese exports to the United States estimated separately, is lower than that for Japan's total exports. Similarly, the higher capital/ labor ratio required for East Germany's experts as com- pared to its competitive import replacements is explained by Stolper and Roskamp [30] in relatively low capital labor endowment ratio of other Communist countries which 26 are East Germany's major trade partners. However, Wahl's [40] study of Canada's trade (1949) revealed that Canadian exports to the United States and to the United Kingdom are, on the average, more capital intensive than Canadian import replacements from these countries. This result along with the pre- sumption that both the United Kingdom and the United States were capital rich relative to Canada around 1949 appears to be inconsistent with the H-0 pattern of trade. The structure of West German [26] trade, and the fact that its exports are more capital intensive than its import competing replacements, cannot support or reject the H-0 theory because the relative factor endowment position of West Germany vis-a-vis its trading partners is not clear. Only a bilateral study of West German trade with individual groups of countries may offer a more definite statement of the comparative advantage position of that country. Recently R. Baldwin [2] updated Leontief's test of U.S. trade pattern by using 1958 trade data and 1958 factor requirement coefficients and input- output table. The result indicated that the Leontief paradox still holds. However, in a bilateral test, the study confirms the H-0 pattern in U.S.-Western European, and U.S.-Japanese trade. Based on the 1963 input coefficients of U.S. industries, the direct capital and labor requirements of 27 U.S. exports and import replacements for 1958-60 and 1968 are estimated in a study by the United States Tariff Com- mission [35]. The result indicated that while the para- dox prevail in both periods the 1968 imports are less capital intensive relative to experts than they were for U.S. trade during 1958—60. However, these estimates are deficient since they exclude the indirect capital and labor requirements of U.S. trade. The change in the rela- tive factor intensity of U.S. trade between the two periods may not be all attributed to a change in the com- modity composition of trade but also to variations in the structure of U.S. industries and a change in the degree of their vertical integration. 1.4.1 Reexamination of the H-0 Theory and its Critical Assumptions The inconclusive nature of the empirical tests of the H—0 theory is based by the critics on the five following factors: 1. Factor reversal: uncertainty about the existence of a unique factor intensity ordering of commodities for all relevant factor price ratios in different countries. 2. Different production functions: variations in production function for the same commodity in different countries. Factor quality differences may be considered as a case of variation in production functions. 3. Demand: substantial differences in the demand pattern of countries which may reverse the pattern of trade as indicated by comparative advantage in production of commodities. pee 28 4. Natural resources: special characteristics of natural resource base commodities. 5. Barriers to trade: tariffs, quotas, or other instruments of commercial policies which disturb the free flow of inter- national trade. The following sections examine the above violations of the H—0 assumptions, attempting in each case to evaluate the empirical validity of the assump- tion and to point out the conditions under which the H-0 assumptions may be considered realistic. 1.4.2 Factor Intensity Reversal A necessary condition for realization of the H-0 pattern of trade is the assumption that the production functions exhibit different (among commodities) and un- changing (among countries) factor intensities for all possible factor price ratios. This means that a commodity which is capital intensive relative to another commodity remains capital intensive irrespective of the relative factor prices, and so also will a labor intensive good. This is called the strong Samuelsonian factor intensity assumption. If production functions are of the Cobb-Douglas form, bi L (3)0. b>01 (8) where Qi is output of industry, 1, K and L are capital and labor inputs, and Ai’ ai, and bi are the parameters, 29 it is shown below that the relative factor intensities are independent of factor price ratios, and thus the strong factor intensity assumption holds. From (8) the marginal productivity of labor and capital in industry 1 can be written as: (9) II > H' U H- 7Q H' L“ bQi/aL bQi/OK = A a.K L (10) and the marginal rate of substitution between capital and labor: (BQi/BL)=(3Qi/3K) = (b,/a,)(K/L) ° (11) Recalling that under perfect competition the marginal rate of substitution is equal to the factor price ratio, the optimal capital/labor ration in indus- tries i and j will be: >< H' II (K/L), (ai/bi)(w/r) (12) >< l (K/L), (aj/bj)(w/r) . (13) With equality of factor price ratios among industries in equilibrium, the relative factor intensities can be writ- ten in the form: 30 (Xi/Xj) = (ai/bi):(aj/bj) (14) which is independent of the relative factor prices. However, production functions may be best fitted in the form: Qi = (A.K i + a.L ) (15) as suggested by Arrow, Chenery, Minhas, and Solow [l], in which case we can show that relative factor intensity is not necessarily independent of factor price ratios. The marginal productivity of capital and labor in industry i can be written as: bi + 1 a,(Q,/L) (16) bQi/aL bi + 1 Ai(Qi/K) (17) DQi/bK and the marginal rate of substitution between capital and labor will be: bi+l (bQi/DL):(3Q173K) = (ai/Ai)(K/L) ° (18) Equality of the marginal rate of substitution with factor price ratios, under the perfect competition assumption, yields the following optimum factor use ratios in industries i and j. 31 1/(bi + 1) (Ai/ai)(w/r) (19) X N (K/L)i 1/(b. -+ l) , (K/L), = (Aj/aj)(w/r) J (20) X II where l/(b + 1) = Z is the elasticity of substitution for each industry. With market equilibrium, the relative factor intensity can be written as Z ° Z 1 (Xi/Xj) = [(Ai/ai) :(Aj/aj) . z.-z. J1(w/r) 1 3° (21) Therefore, factor intensity can be independent of factor price ratios only if 21 = Zj' That is, factor intensity reversal is ruled out only when the elasticities of sub‘ stitution for industries 1 and j are equal. With factor intensity reversal the neat and clear conclusions of the H-0 theory do not necessarily hold. Consider the relation for the optimal factor use ratio for industry 1 which can be rewritten from equation (19) in the logarithmic form: Log(K/L)i = Log(Ai/ai) + l/(bi + l)Log(w/r)(-22 This relation between the factor use ratio and relative factor prices can be drawn as a straight line for any 1 B;_T—T ' industries the elasticities of substitution are different, industry with a slope of Zi = If for any two the two lines will definitely intersect at some critical 32 factor price ratio, at which point the relative factor intensities will be reversed. In Figure 1-3 the straight lines of XX and YY represent the optimal factor use ratio for industries X and Y at different factor price ratios. At the relative factor price of (w/r)2 commodity X is capital intensive and commodity Y is labor intensive. However, when the relative price of labor increases to (w/r)1 the Optimum factor input ratios indicate that Y is more capital in- tensive than X, and at the point (w/r)O both commodities have the same factor intensity. The far-reaching impli- cation of this possibility is that if the relative factor prices in the two trading countries are such that they are as widely different as (w/r)1 and (w/r)2, or more precisely if they are such that they fall on different sides of (w/r)o, then nothing definite can be said about factor intensity of commodities, and the H-0 pattern of trade will not hold for at least one of the trading coun- tries. Let E1 and E2 denote the factor endowment ratio of countries I and II for which the range of relative factor prices in country I will be between (w/r)'1 and (w/r)"1 and in country 11 will be between (w/r)'2 and (w/r)"2. As shown in Figure 1-3 factor reversal occurs in the relevant range of relative factor prices. Country II, which is richly endowed with labor, will have a com- parative advantage in Y, the labor intensive commodity. 33 Figure 1-3 Factor Intensity Reversal 34 Country I, which is capital abundant, will have a comparative advantage in the production of X, the labor intensive commodity in that country. Thus country II's production and trade conform with the H-0 pattern while those of country I do not. The important question is how widespread is this phenomenon, if it exists at all. In Samuelson's view [29], "factor intensity reversal is much less important empirically than it is interesting theoreti- cally.” It was Minhas [22] who suggested that factor reversal is in fact an empirical possibility. He main- tained that, based on the study with Arrow, Chenery, and Solow [1], production functions are not internationally characterized by the Cobb-Douglas form but are best fitted in the form of CES production functions represented above by equation (8). In a test of twenty-four industries in nineteen countries Minhas found [22,pp. 35-39] that in fact factor reversal occurs between five pairs of industries (from twenty-five pairs) in the relevant range of factor price ratios. However, the result of these estimates depends upon the method of estimating the value of elasticities of substitution. Leontief [20], using Minhas' data but with an alternative method of estimating elasticities, found that only seventeen of 210 possible factor reversals occur in the relevant range of factor prices between the U.S. and India. He therefore concluded that the strong 35 factor intensity assumption is supported by Minhas' data. Alternatively, the strong factor intensity assumption can be tested by ranking the factor intensities of industries among countries. Then, if the assumption that for any ratio of factor prices "the Optimal ratio Of capital to labor in any given industry 1 is always greater or less than any other industry j were true, then the ranking of industries between the two countries with widely different factor price ratios must match" [22, p. 39]. In fact Minhas performed the same ranking for the U.S. and Japan and found the Spearman rank cor- relation coefficient of only 0.328, "which is nowhere near unity, the value implied by the strong factor in- tensity assumption.” However Ball [3] showed that when Minhas' ranking is applied to only non-agricultural goods, the rank correlation coefficient for both direct and total input requirement of U.S. and Japanese industries in- creases. If both agricultural and food products are eliminated on the basis of greater dissimilarities in the composition Of output and methods Of production among countries, the rank correlation coefficient increases further. Ball maintains, therefore, that factor intensity assumption cannot be rejected as was done by Minhas. Nor can it be accepted. Similar results were introduced by Lary [17] who used non-wage value added per employee as an index 36 of capital intensity in his bilateral comparison of manufacturing industries between the U.S. and the U.K., the U.S. and Japan, and the U.S. and India. Lary takes the direct input coefficients to measure factor intensity. But for a two-factor H-O model only the total requirement coefficients are the proper measure of factor intensity. And if only direct factor input is used, ”we no longer remain within the bounds of a two factor world and all sorts of extraneous things (like differences in the degree of vertical integration among industries) can effect the nature of the results” [22, p. 41]. Although there is no strong basis for rejecting the uniqueness of factor intensity ranking of commodities among countries, the issue is not settled. However, with reference to Figure 1-3, it may be concluded that the wider the factor price ratio differential between two countries, the higher the probability of factor reversal for a larger number of commodities. 1.4.3 Production Functions and Factor Qualities Two necessary assumptions of the H-0 model are the equality of production functions and factor qualities in the trading countries. However, differ- ences in factor qualities can be transformed into dif- ferences in production functions, and thus their equality is the only significant assumption with respect to the inter-country condition Of production. 37 Thus the empirical validity of the H-0 theory is not insured if variations of significant magnitude are observed in the production functions in different countries.* Consider Figure 1-4, which depicts a situa- tion where production functions Of X and Y are distinctly different in countries I and II. It can be Observed that relative to Y, X is capital intensive in country I and labor intensive in country 11. Recalling that country I is abundant in capital and country II in labor, according to the theory both countries can produce X cheaper than Y. Therefore the direction of trade cannot be determined unequivocally but will depend on the relative intensity of each product in each country and the relative factor endowment of the two countries. Suppose the relative factor intensities Of commodities and the relative factor endowment of country I are such that the relative price of X there is less than that in country II. Then country I will export X and country II will export Y. For coun- try I the pattern is consistent with the H-0 theory, but for country II it is not. Therefore the pattern of trade is not necessarily consistent with the H-0 pattern if wide variations exist in the production functions as demon- strated above. *This problem is different from factor intensity reversal, which may be caused by differences in the elasti- cities Of substitution of production functions. Here, how- ever, while the elasticities of substitution may be equal, the problem arises from differences in efficiency and scale coefficients of the production function of a commodity among the countries. 38 BI“ 00 mOHMDCDOU cmm3umm chfluOcsm Cowuoswoum ucmumwwwa qua ooowflm m.woq MIA woq HH huucsoo H %uuasoo idly-J 39 Although there is no justification for assuming similar production functions for all types of commodities in all countries, it is reasonable to assume a high de- gree of similarity in production functions of manufactured goods, mainly because of the widespread diffusion of technology among countries, and the standardized methods required for production of these goods. Therefore on an a priori basis it appears that this adverse possibility can be reduced if the H-0 theory is applied only to the manufactured goods. 1.4.4 Identical Demand The analysis in section 1.2.2 indicated that the logical consistency of the H-0 theory, with its relative supply definition of factor endowment, requires an explicit assumption about demand patterns in trading countries. According to this assumption, demand functions must be identical in both countries and should be homogeneous of the first degree to insure independence of demand pattern from income. However, neither parts of this assumption may be considered realistic on an a priori basis, specially when we consider countries with significantly different income levels. In general the wealthier nations are the more developed economies, and economic deveIOpment in most cases implies a higher degree of industrialization, auto- mation, and urbanization, all Of which will inevitably affect the consumer's conception of utility and thus his (lemand pattern. Although the equalization force of the 40 international demonstration effect may tend to influence the demand patterns of the less developed countries (LDC's), it cannot be assumed that the patterns are identical in such widely different countries as India and the U.S. Furthermore, a unitary income elasticity of demand for all goods is hardly an empirical reality; different goods have different income elasticities. The ”-0 theory faces Special difficulty if a systematic bias exists in demand for capital intensive commodities in the wealthier countries, which are with rare exceptions the capital rich countries. It can be anticipated that in these countries there will be a demand bias toward manufactured consumer durable goods, higher quality housing, and for non- traded services. The first and second items are def- initely capital intensive commodities. For services, however, relative factor intensity is not as easily de- termined, since there are significant variations in quality and definition of services between the very high and very low income countries. Certain aspects Of manufacturing in an industrialized society may be considered as labor intensive services in the less developed economies.* At the same time changes in the quality of services in the development process often embodies high degrees of * . . . For example, the teleV1s1on set 1n a modern economy replaces the old Coffee House in a village of a traditional economy for the evening entertainment. 41 mechanization and automation which tend to increase their capital intensity.* Therefore a distortion in the H-0 pattern of trade may occur as a result of a systematic consumption bias toward capital intensive commodities in the capital rich countries. A capital rich country (U.S.) which pro- duces relatively more Of the capital intensive commodities than a poor, labor abundant country (India), may demand so much more of the capital intensive commodities that the pattern of trade will be reversed. This possibility is likely to be limited if we apply the theory to coun- tries with not too dissimilar income levels. 1.4.5 Natural Resources The tests of the H-0 theory consider capital and labor as the only significant factors in determining the commodity composition of trade. However, Diab [8] and Vanek [38,39] have pointed out that exclusion of natural resources may distort the indicated pattern Of trade for capital abundant countries that are not well endowed with natural resources. The distortion is caused they argue, by the high proportion of natural resource primary goods, which have a relatively high capital in- tensity, in the imports of these countries. *The most notable change of this type may be observed in communication, information, and transporta- tion services. 42 However, the difficulty in the treatment Of natural resources in the H-0 theory may be considered from a different perspective. A considerable amount of capital and labor must be expended to bring a resource into a form which is usable for production purposes, namely primary goods. As long as the conditions of the production Of primary goods are similar in all countries we can treat them as intermediate inputs within the two- factor model of the H-0 theory, with no problem. But difficulty arises when we apply factor requirements for the extraction and utilization of natural resources of one country to another one. In a country with an unfavor- able endowment Of natural resources, their supply can be expanded only at a relatively high cost and with the employment Of highly capital intensive methods. Hence the actual production functions for natural resources can be different for two countries because of the rela— tive accessibility and quality of resources. And since there is no reason tO believe that natural resources are distributed evenly among the countries of the world, they cannot be treated very easily in the context of the H-0 theory, which requires identical production func- tions for each commodity in all countries. It should be emphasized here that the difficulty with natural re- sources comes from their characteristic as a produced commodity and not because of their exclusion as a factor of production. The only solution is to consider natural 43 resources as non-competing imports when they are not easily available in a country. But since we have no Operational criterion for "easily available" resources, and because any natural resource is scarce in one coun: try or another, the H-0 theory appears to be incapable of including natural resource oriented commodities. In reality the problem is more complicated as a result of the mobility of capital across national boundaries to the countries in which natural resources are located. However, the important consideration then is the location Of capital, not its ownership. That is, a capital poor but natural resource rich country may become abundant with capital as a result of the inflow of foreign capital for the extraction and export of its natural resources. One can no longer then consider the recipient country capital scarce if its relative endowment of capital changes vis-a-vis the other country. 1.4.6 Trade Barriers A theory of comparative advantage can predict the pattern of trade only if no artificial barriers exist to the free flow Of international trade. Otherwise, al- though the theory may be capable Of explaining comparative advantage, that is the forces that determine the relative prices of commodities in isolation in each country, it can fail to predict the commodity composition Of their trade. This, Travis [34] points out, is the most important reason 44 for the failure of the H-0 theory in predicting the pattern of world trade. This view is supported in a study by Tarshis [32], which examines the wholesale prices of selected commodities in the U.S., the U.K., the Soviet Union, and Japan. He found that the comparative advantage of each country, as reflected in the relative commodity prices in that country, is consistent with the factor propor- tion theory. In many cases, however, the direction of trade is clearly contrary to the pattern indicated by the relative prices of commodities. A study by Vaccara [36] revealed significant and positive correlations between nominal tariff rates and the direct labor cost and labor use coefficients in 311 American manufacturing industries. But a sub- sequent study by Basevi [4] showed that once nominal rates are replaced by effective tariff rates the above relationships are no longer significant. Iqbal [12] applied a more accurate measure of factor intensity along with effective tariff rates to a limited number of less advanced industries and demon- strated that a positive relation exists between the labor/ capital requirement (direct plus indirect) of these in- dustries in five less deve10ped countries and the effective rate of protection in the deve10ped countries. Therefore, while there are no conclusive studies Of the relationship between comparative advantage and the 45 tariff distortions in the total U.S. trade and the trade among developed economies, it is reasonable to expect, based on Iqbal's study, some systematic distortions in the trade pattern between the developed and the less deve10ped economies. (I II) [1] [2] [3] [4] [5] [6] [7] [8] [9] REFERENCES Arrow, K.J., Chenery, H.B., Minhas, B.S., and Solow, R.M. "Capital--Labor Substitution and Economic Efficiency." Review of Economics and Statistics. August, 1961, pp. 225-50. Baldwin, R. ”Determination of the Commodity Structure of the U.S. Trade." American Economic Review. March, 1971, pp. 126-46. Ball, D.S. ”Factor Intensity Reversal in International Comparison Of Factor Cost and Factor Use." Journal of Political Economy. February, 1966, pp. 77-80. Basevi, G. "The United States Tariff Structure: Estimate of Effective Rates of Protection of United States Industries and Industrial Labor." Review of Economics and Statistics. May, 1966, pp. 147-160. Bhagwati, J. "The Pure Theory of International Trade: A Survey." Survey of Economic Theory, Growth and Development prepared fer the American Economic Association and the Royal Economic Society. New York: St. Martin's Press, 1965. Bharadwaj, R. Structural Basis of India's Foreign Trade. Bombay: University of Bombay,'1962. Buchanan, N.S. ”Lines on the Leontief Paradox.” Review of Economics and Statistics. August, 1954, pp. 286-289. Diab, M.A. The United States Capital Position and the Structure of its FOreign Trade. Amsterdam: NOrth-Holland Pfiblishing, 1956. Ellsworth, P.T. "The Structure of American Foreign Trade: A New View Examined." Review of Eco- nomics and Statistics. August, 19S4, pp. 279—85. 46 [NJ [11] [12] [13] [14] [15] [16] [17] [18] [19] [20] [21] 47 Ford, J.L. The Ohlin-Heckscher Theogy of the Basis and Effects of Commodity Trade. New York: Asia Publighifig House, 1965. Hodd, M. "An Empirical Investigation Of the H-0 Theory." Economica. February, 1967, pp. 20- 29. Iqbal, Z. The Comparative Advantage of Developing Countries in the Manufacturing Industries and fhe Effects of Generalized Tariff Preferences. Ph.D. Thesis, Michigan State University, 1971. Jones, R. "Factor Proportions and the Heckscher- Ohlin Theorem.” Review Of Economic Studies. 1956-57, pp. 1-10. Kravis, I.B. "Availability and Other Influences on the Commodity Composition of Trade." Jour- nal of Political Economy. April, 1956, pp. 143-44. Kreinin, M. "Comparative Labor Effectiveness and the Leontief Scarce Factor Paradox." American Economic Review. March, 1965, pp. 131-40. Lancaster, K. ”The Heckscher-Ohlin Model: A Geometric Treatment." Economica. February, 1957, pp. 19-39. Lary, H. Import of Manufactures From Less Developed CountriEs. New YOrk: NBER, COIumEia University Press, 1968. Leontief, W. "Domestic Production and Foreign Trade, The American Capital Position Re-examined.” Economia Internazionale. February, 1954, pp. 9-45. . "Factor Proportion and the Structure of American Trade: Further Theoretical and Empirical Analysis.” Review Of Economics and Statistics. November, 1956, pp. 387-407. ________ ”International Factor Cost and Factor Use." American Economic Review. June, 1964, pp. 335-45. MacDougal, G.D.A. ”British and American Exports: A Study Suggested by the Theory Of Comparative Advantage Part 1." Economic Journal. December, 1951, pp. 697-724. [22] [231 [24] [25] [26] [27] [28] [29] [30] [31] [32] [33] 48 An International Comparison of Factor Minhas, B.S. Amsterdam: North-Holland Cost and Factor Use. Publishing, 1963. Mookerjee, S. Factor Endowment and International Trade. Bombay, I968. Ohlin, B. Revised Edition. University Press, Interregional and International Trade. Cambridge, Mass.: Harvard 1967. "Factor Proportions and the Compara- Part I.” Quarterly Journal May, 1956, pp. 169-92. Robinson, R. tive Advantage: of Economics. Roskamp, K. "Factor PrOportions and Foreign Trade: The Case of West Germany." Weltwirtschaftilches Archives. 1963, pp. 319-26. "International Trade and Equali- Economic Journal. Samuelson, P.A. zation of Factor Prices.“ June 1948, pp. 163-84. . ”International Factor Price Equal- ization; Once Again.” Economic Journal. June 1949, pp. 181-97. ”A Comment on Factor Price Equal- Eization."’ Review of Economic Studies. 1951- 1952, p. 121. Stolper, W., and Roskamp, K. ”Input-Output Table for East Germany With Application to Foreign Trade." Bulletin of Oxford University Insti- tute of Economics and Statistics. November, 1961, pp. 379-92. "Capital Shortage and Labor Sur- Review Of Economics 286-89. Swerling, B.C. plus in the United States." and Statistics. August, 1954, pp. "Factor Inputs and International Price Comparisons." In M. Abramovitz and others (eds.), The Allocation of Economic Resources. Stanford, CaIifOrnia: ’Stanford University Press, 1954, pp. 236-44. Tarshis, L. and Ichimura, S. ”Factor Proportion The Case of Japan." Review November, 1959, Tatemoto, M., and Foreign Trade: of Economics and Statistics. pp. 442-46. [341 [35] [36] [37] [38] [39] [40] 49 Travis, W.P. The Theory of Trade and Protection. Cambridge, Mass.: Harvard University Press. 1964. United States Tariff Commission. Competitiveness of U.S. Industries. TC Publication 473, WESh- ington, D.C., April, 1972. Vaccara, B. Employment and Output in Protected ManufacturingyIndustries. Washington, D.C.: BroOking Institution, 1960. Valavanis-Vail, S. "Leontief's Scarce Factor Paradox." Journal of Political Economy. 1954, pp. 523-28. Vanek, J. The Natural Resource Content of United States Foreign Trade, 1870-1955} Cambfidge, Mass.: M.I.T. Press, 1963. "Factor Proportion Theory: The N-Factor “‘ 'C'a—se' Kyklos. October, 1968, pp. 749-56. Whal, D.F. "Capital and Labor Requirements for Canada's Foreign Trade." The Canadian Journal of Economics and Political Science. August, 1961, pp. 349-58. Chapter 2 ALTERNATIVE THEORIES OF TRADE 2.1 Introduction The apparent failure of the H-0 model to explain the commodity composition of trade prompted the development of alternative theories. While the H-0 theory assumes uniformity of factors of production and equality Of production functions, alternative theo- ries seek to explain the flow of trade through dif- ferences in either factor qualities or in the produc- tion functions. The factor quality approach, or the so-called neofactor theories,* may be considered an extension of the H-0 theory. They attempt to replace the simple concept of capital and labor by a more realistic and somewhat more complex notion Of factor endowment which incorporates factor qualities as well as quantities. On the other hand, the second approach, or the so-called neotechnology theories,* in sharp con- trast with the H-0 theory, sets out to explain the flow of international trade through differences in produc- tion functions. *As referred to by Hufbauer [11, p. 195]- 50 51 2.2 Neofactor Theories: The Human Capital Approach Proponents of neofactor theories maintain that while the logical framework Of the H-0 theory is valid, its treatment of factors of production is hopelessly crude. The central concern in this approach is the development of a more meaningful measure of labor in- stead of what the H-0 theory has provided for, namely simple and undifferentiated labor. It is argued that by considering the stock of physical capital and the number of workers at work as the relevant measures of society's endowments of factors Of production, one ig- nores a substantial part of society's resources which have been channeled into the stock of ”waiting” in the form of human capital, [15]. Similar to the stock of physical capital, the human capital endowment of an economy, manifested in the skill composition of the workers, will affect the productive capacity and the structure Of relative costs and outputs of the economy [4,23]. Proponents of the human capital approach therefore maintain that the H—0 theory ignores a significant part of society's resources embodied in the form of human capital, and in their view there should be little surprise if the theory cannot stand the test of empirical verification. The recognition of skills as a determinant of the commodity composition of trade dates back to 1956 Ivhen Kravis [18] found that American workers receive 52 higher wages in export industries than in import competing industries. He attributed the relatively high wages in export industries to the higher skill mix of these in- dustries. This hypothesis was later tested and confirmed by Waehrer [27] through a statistical analysis of the re- lationship between wages and skills in Kravis' industries. Both Kravis and Waehrer concluded that the U.S. has a com- parative advantage in the production and export Of skill intensive commodities. While the above studies present some relationship between the skill mix and comparative advantage, the effort was not put within a systematic model for explaining com- modity composition of trade. The general contention Of the human capital approach has been put forward in two different forms. Human capital may be considered along with physical capital as one composite factor of produc- tion, capital, with unskilled labor (the number of man years) being the second factor. This approach was sug- gested by Kenen and Yudin [16], Kenen [15], Bharadwaj and Bhagwati [S], and Roskamp and McMeekin [22]. They es- timated the wage differentials between unskilled labor and various skilled categories, The excess of earnings of a skilled over an unskilled worker was considered as a return to the human capital invested in him for educa- tion and training. The stock of human capital is then estimated by applying a ”realistic” rate of return to his earnings. 53 Kenen [15], through the application of the above method and assuming 12.7 and 9 percent rates of return to human capital, measured the physical and the human capital embodied in one million dollars of Ameri- can imports and exports for 1947 trade. However, the results were only a slight improvement over those of Leontief (in terms Of the H-0 model) and thus the para- dox prevails. With a similar approach, Roskamp and McMeekin [22] undertook a study of West German trade and found that one million marks of Germany's exports contain a higher ratio of human capital to labor and human capital to physical capital than it's one million marks of im- port competing commodities. They concluded that Germany experiences a comparative advantage in the production of skill intensive relative to capital and labor inten- sive commodities. Bharadwaj and Bhagwati [5] applied Kenen's method to Indian trade during 1953-54. Although their results indicated that Indian exports are, on the ave- rage, more labor intensive relative to imports, an unexpected situation existed that the inclusion of human capital reduced the degree of relative labor intensity of Indian exports. Using wage differentials as a proxy for human capital has several shortcomings. First, one must choose a "standard" wage rate for "standard" unskilled labor, 54 a selection which is Often arbitrary and therefore may bias the results rather considerably. Once the "stan- dard" wage rate is established, a more difficult task is to select a "realistic" rate of return to human capi- tal. Once again the choice is relatively arbitrary and can affect the results significantly. However, the other shortcoming of Kenen's method lies in attributing the total wage differentials to skill levels, and thus to human capital. What is assumed away is the effect of the monOpoly power of labor unions in the labor market on the wages of the skilled workers. Within the context of Kenen's approach, the stock of human capital may be measured directly by estimating the cost of education for each skill level. In this way Baldwin [2] estimated the total capital embodied in imports and experts as the sum of the physi- cal capital and the educational cost for required skills. He found that for U.S. trade in 1962, after excluding natural resource intensive commodities, the coefficient of factor intensity of trade* becomes 0.97, and thus the paradox is reversed. *The ratio of factor intensity of imports to exports, or II = (K/L)m (K/L)x 55 An alternative method for assessing the human capital requirement of trade was introduced by Keesing [12, 14]. He attempted to find the relationship between the commodity composition of trade and interindustry skill differentials by looking into the actual skill mix Of industries. "The availability of labor skills deter- mines the pattern Of international location and trade of manufactured products, those not closely tied to natural resources," said Keesing [12, p. 28]. This hypothesis is based on the observation that labor is much less mobile internationally than capital. While in the ab- sence of labor mobility skills may be acquired only through a long process of industrial experience and edu- cational training, the mobility of capital and the trans- formation of new technology embodied in capital goods are rather prevalent among economies. Therefore, within the framework of the H-0 theory, it is only the relative endowment of skilled to unskilled labor that determines the pattern of trade. Kessing classified labor skills into five categories: 1. Professional, technical, and managerial. II. Craftsmen and foremen. III. Clerical, sales, and services. IV. Operative (semi-skilled). V. Laborers (unskilled). 56 Applying the immediate skill requirements of U.S. industries* in 1950 to 1957 imports and exports of the U.S. and eight other countries, he found that U.S. exports are more skill intensive than imports by using three different definitions of skill intensity (Table 2-1). Moreover, for the other countries he discovers that the relative skill intensity ratio Of exports to imports is larger than one for West Germany, Sweden, and the U.K. and smaller than one for the Netherlands, Bel- gium, Italy, France, and Japan. Keesing's results cannot be considered conclu- sive and affirmative verification of the human skill theory without properly identifying the relative skill endowment of the countries under study. If one accepts that the U.S. is the most skill abundant country in the world, then the U.S. result is a positive confirmation of the theory. However, very little can be concluded from the trade structure of the other eight countries without specific knowledge of their relative skill en- dowment in terms of trading partners. Hufbauer [ll] attempted to test for the gener- ality of Keesing's hypothesis by correlating the skill intensity of experts with the skill endowment of a ~__ *The application of U.S. skill coefficients to the=‘trade structure of other countries may be justified on the basis of the tests performed by Keesing [12], and fhxfbauer [11], which show significant similarity of rank- ting of skill intensities of industries among countries. In Table 2-1 Keesing's Estimate of Skill Intensity of U.S. Trade (1957) Definition of skill Group (I + II) Gropp_I Group II intensity ratio Group (IV + V) Group (IV + V) Group (V + V) EXport skill ragig 9 WV Import skill ratio 1'7“4 1:582 l-BLJ Note: Groups refered to in this table correspond to the skill caLegorIes used by Keesing. Source: [12, p. 291, Table 3]. 58 selected group of countries. He found 0.695 and 0.822 for the Spearman and weighted rank correlation coefficients between the two variables. I In criticism it should be noted that both Keesing and Hufbauer apply the immediate (i.e. the final stage) instead of the total input requirements of each industry, thus ignoring the indirect input requirements for the industry. Their estimates will thus be biased to the extent that the degree Of vertical integration and the indirect factor requirements may vary among industries. 2.3.1 Neotechnology Theories The H-O theory assumes that the technological conditions of production are similar in all countries. However, in actual fact a country may enjoy a comparative advantage as a result Of favorable technological conditions. A country with large domestic markets can have a comparative advantage in the production of a decreasing cost commodity. More significantly, the comparative advantage may be the result of the deveIOpment of a more efficient process for producing the Old commodities or from the monopoly position resulting from new innovations in producing new products. The proponents of neotechnology theories of trade maiJTtain that "a complete explanation of trade must take account of differences in technological capabilities and 59 the relative efficiency among countries and industries [1,p. 241]. All neotechnology theories assume demand as given and similar factor endowment ratios in all countries, and they set out to explain the commodity composition of trade through comparative technological advantage. In this section several neotechnology explanations of trade will be presented under the four alternative theories of scale economies, stage of pro- duction, technological gap, and the product cycle. 2.3.2 Scale Economies According to the scale economies explanation of trade, a country with a large home market will tend to enjoy a comparative advantage in the production of those commodities which exhibit the greatest scale economies. Hufbauer, the first proponent of this thesis in 1966, pointed out that the above preposition carries with it a "hidden" assumption, that "internal commodity price ratios are established in all countries before trade Opens; that is to say that all goods are produced every- where before trade commences" [10, p. 23]. The assump- tion implies that small nations could not, with reliance oninternational markets, develop a scale economy industry. HoWever, Hufbauer argued that while contrary to this assumption production and trade are Often simultaneous, 60 geographical and psychological barriers to trade, along with the protective measures of tariffs and quotas, prevent a small country from taking advantage of international markets [10, p. 3; 11, p. 176]. In a detailed study of the trade and deveIOpment of synthetic materials Hufbauer found that the size Of home markets, as measured by gross domestic product, ”exercises considerable influence over the direction of synthetic materials trade" [10, p. 72]. Baldwin [2] tested the scale economies hypothesis by applying Weiss' estimates of scale index [29] for each industry (the percentage of employees in establishments with 250 or more employees) to an average one million dollar bundle of U.S. imports and exports for 1962. The results supported the theory in the cases of U.S. total trade and trade with Canada and the less developed countries but contradicted it in trade with Western Europe and Japan. In a recent study, Hufbauer [11] introduced a more sophisticated method for testing the scale economies theory. As a measure of the scale coefficient he estimated the percentage change in value added per worker for a per- centage change in the plant size for each one of the three- . . * dlgit SITC group Of U.S. manufacturing industries. This _ . * A detailed explanation Of the method of estima- thII of this coefficient and the coefficients of the {killowing three neotechnology theories will be presented in (Hiapter 5, where the bilateral trade structure of the IJ-S-, the U.K., and South Korea will be tested. 61 coefficient is then weighted by the relative size of exports of each commodity group in the expert baskets of the U.S. and twenty-three other countries for 1963 to Obtain the co- efficient of scale economies embodied in exports. Hufbauer tested the theory by estimating the Spearman and the weigh- ted correlation coefficients between the rank ordering of scale coefficient of exports and the size of manufacturing output of the economies as a measure of economic size. He found the two coefecients to be 0.627 and 0.788, respect- ively and concluded that the test confirms the theory. 2.3.3 Stage of Production The stage of production hypothesis attempts to explain comparative advantage in terms of the level of economic development and the technological sephistication of production in the economy. The essential logic of the thesis grows out of economic development studies which are concerned with structural transformation and growth through import substitution. The theory states that the less advan- ced economies tend to specialize in the production of fini- Shed consumer goods while the more advanced econOmies spec- ialize in production of consumer goods. The reason for this Pattern of specialization is that the required technology fOI' the production of consumer goods is less complicated and is 31655 capital intensive. In addition there already exists a SIHDStantial domestic market in the less developed economies f0!“these commodities which were imported prior to their 62. domestic production [24, p. 5; 9, pp. 6.71- The only test of this theory was performed by Hufbauer [11] by estimating the proportion of consumer goods in the exports of twenty-four countries (as explained in section 5.3), he found the Spearman and rank correl- ation coefficients between consumer goods ratio in exports and the GDP per capital as a measure of the level of economic development to be coefficients of 0.818 and 0.801 respectively. This, in Hufbauer's view, confirms the theory's hypothesis. 2.3.4 Technolggical Ggp The dynamics of industrial development manifested in the introduction of innovatiOns and imitations is considered by the technological gap theory as the significant factor which shapes and affects the structure of manufacturing trade among countries. The effect of technological change on the structure of trade became the focus of studies during the 1950's when the dollar shortage problem became apparent. The works of Hicks [7], Balogh [3] and Williams [28], among others, concentrated on the con- sequences of technological progress, particularly of the U.S., on the relative prices of commodities in the international market. However, Kravis [19] pointed OUt that the effect of technological change on the structure of trade is not only through the forces of 63 comparative cost and therefore through relative price adjustments but, more significantly, through "the advantages which result from the possession of the newest product and of the most recent improvement on the old kind" [19, p. 151]. This is the so-called availability thesis. Since the innovation of new products and processes is the result of systematic efforts of pro- ducers in research and development, Keesing [13] and Gruber, Mehta, and Vernon [6] attempted to test Kravis' hypothesis indirectly by considering the relationship between the level of activities in research and deve10p- ment and the export performance of a selected group of U.S. industries. They found a positive and significant correlation between the proportion of expenditure on, and the proportion of scientist and engineers in research and development and the level of export performance of the industry. The simplistic version of the technological gap theory introduced by Kravis in the above prOposition was formulated into a dynamic model by Posner [21]. Posner assumes no international differences in factor PrOportions, no factor mobility, and no transportation cost among countries in this model. If a country introduces a new product or an improvement of an old product, the country will enjoy a comparative advantage In the production of that product in the international 64 market until other countries become capable of producing this new product. The lag between the time that the product was introduced and the time that another country can imitate it is called "imitation lag" (totz in Figure 2-1). The size of the lag may depend on the availability of technological requirements for the imitating country and on the extent of legal barriers to patent rights. However, once the other countries are able to set up similar production processes, the leading country loses its favorable position and the export of the new commodity declines. Of course if new innovations take place in the advanced country, by the time comparative advantage is lost in production of one commodity innovations will provide comparative advantage in production of other commodities. The important point, however, is the extent of the aggregate imitation lag. Hufbauer [11] tested this theory by correlating the average age of exports of twenty-four countries (explained in section 5.4) and the GDP per capita as the measure of industrial sophistication. He obtained Significant Spearman and weighted rank correlation coefficients of 0.698 and 0.864,respectively,and considered the result a positive test of the theory. 0 U1 Country I Production Exports Amount of Pro- A///////’~\\\\\\\\_ T' duction & Export 1me o 1 2 tN Production Country II (Rest of the World) Country 1 introduces the new commodity. Export starts as a response to the newly created international demand. Country 11 starts producing the commodity. Country I loses its comparative advantage in the product and export declines to zero. Source: Hufbauer [40, P~ 24]' Figure 2-1 Imitation Lag and the Technological Gap Theory 66 2.3.5 Product Cycle To the extent that the product cycle approach deals with the timing of innovations and the level of technological SOphistication of economies, the theory is little different from the technological gap theory. However, the product cycle theory attempts to consider the elements which bring about the imitation gap and in this respect it is one step beyond the technological gap theory. Hirsch [8] and Vernon [26] maintain that the more sophisticated economies can produce the new and differentiated product. The high income of the economy justifies the expectation of large profits through developing the new product. In the early stages of pro- duction the producer is confronted with the uncertainty concerning the dimension of the market, the specifications of the inputs, and the exact method of production. But once the product is introduced and its production expanded, a certain degree of standardization in the method of production and the product itself takes place. Before standardization, the innovator is the sole producer, and he therefore enjoys a comparative advantage on the international market. But with standardization the product can be imitated, which is referred to as the maturing period for the product, and the first producer may lose his 67 comparative advantage if the cost of production is less for other economies due to more favorable factor endow- ment positions. Hirsch's study of the U.S. electronic industry [8] showed that between 1960 and 1964 the competitive position of the matured electronic products deteriorated while the industry maintained its lead in the growth (not yet matured) products. . Hufbauer [ll] attempted to test the product [cycle theory in a cross sectional study of exports of twenty-four nations by correlating the index of product differentiation (explained in section 5.5) embodied in exports and the GDP per capita. He obtained Spearman and weighted rank correlation coefficients of 0.724 and 0.763, respectively, and concluded that the result is a positive confirmation of the product cycle theory. * 2.4 An Evaluation of Alternative Theories of Trade The neofactor and neotechnology theories of trade, as presented in this chapter, choose distinctly ; 3% By applying Hufbauer's measures, a recent study by the United States Tariff Commission [25] tested the scale economies, the technological gap, and the product CYCle theories for U.S. trade during 1958/60 and for 1968. The results confirmed the technological gap and the Product cycle theories for both periods. However, in the case Of the scale economies hypothesis while the 1958/60 trade is consistent with the predicted pattern, incon- s1stency is observed in the 1968 trade. 68 different variables as the predominant forces in the determination of comparative advantage. For neofactor theories, the determining element is the relative endowment of factors of production in each country. Added to the traditional version of the H-0 theory is the admission of differences in factor qualities through consideration of relative endowment of human capital. Thus, neofactor theories preserve the logical framework of the H-0 model and its static characteristic. In contrast neotechnology theories depart from the relative endowment of factors of production ex- planation of trade and seek to determine comparative advantage through relative differences in production functions. In this case, since the predominant variable is the relative technological advantages of economies, the theories are molded in a dynamic form which attempts to explain the effect of technological change and its un- even incidence on countries and commodities on the pattern of trade. The only exception to this generalization is the scale economies theory which is primarily concerned with the relative efficiency of production achieved as a result of the economic size of the country. While stage of production, technological gap, and product cycle theories each give a different hypothesis for the commodity composition of trade, all three theories are essentially saying the same things: The more advanced economies will have comparative advantage in the production 69 of the more sephisticated commodities. The level of sophistication of commodities is measured by its newness in the technological gap theory and by the degree of nonstandardization in the product cycle theory. Howeyer, since new products are also nonstandardized products, there is little difference between the two theories. While the stage of production theory differs in its explanation of comparative advantage from the other two theories, the underlying elements remain rather similar. Consumer goof: are relatively more standardized both in commodity form and in the production process and require less SOphisticated methods of production which are available in the less advanced economies. Given the degree of similarity between these three theories one should expect that they would perform similarly in a test of international trade flows. In fact, Hufbauer's study, as pointed out above, confirms all three theories, along with the scale economies and the human capital approach. 3': The correlation coefficient between the coefficient of product age and the product differentiation 1ndex of trade of twenty-three countries was found to be -0.710 in Table 6-4 Chapter 6 below. *3! The correlation coefficient between the consumer good ratio and the index of product differention 0f trade of twenty-three countries was found to be ‘0-796. in Table 6-4, Chapter 6 below. 70 In addition to what was mentioned in section 2.2 about the shortcoming of Hufbauer's test of the human capital theory, his tests are open to another criticism. If one wants to verify the consistency of trade flows with the comparative advantage of countries, attention must be directed toward trade performances as reflect in import relative to export characteristics of each country and the country's national characteristics. Hufbauer instead based his conclusions on a series of correlations between export characteristics and national characteristics, as hypothesized by each theory. Thus he may conclude, for example, that country I has a comparative advantage over country II in the trade of skill intensive commodities if country I's exports contain a higher skill ratio than the exports of country II. However, this result amy be misleading if by inclusion of imports we find that country I is a net importer of skill intensive commodities while country II is a net exported of them. In conclusion, despite several tests of alternative theories of trade presented in this chapter, there is no conclusive result with respect to their empirical validity. REFERENCES [1] Arrow, K.J., Chenery, H.B., Minhas, B.S., and Solow, R.M. ”Capital-Labor Sustitution and Economic Efficiency." Review of Economics and Statistics. August, 1961, pp. 2291597 [2] Baldwin, R.E. "Determinant of Commodity Structure Of U.S. Trade." American Economic Review. March, 1971, pp. 126-46. [3] Balogh, T. The Dollar Crisis: Cause and Cure. Oxford: B1aCkweIl, 1950. [4] Becker, G. "Investment in Human Capital: A Theorectical Analysis.” Journal of Political Egonomy. Supplement, October 1962} pp. 9-49. [5] Bharadwaj, R., and Bhagwati, J. "Human Capital and the Pattern of Foreign Trade: The Indian Case." The Indian Economic Review. October 1967, pp. 117-42. [:6] Gruber, W., Mehta, D., and Vernon, R. "The R&D Factor in International Trade and Investment of United States Industries." Journal of Political Economy. 1967, pp. 20-37. [f7] IHcks, J.R. ”An Inaugural Lecture." Oxford Economic Papers. June, 1953, pp. 117-35. I:8] Hirsch, S. "The United States Electronic Industry in International Trade." National Institute Economic Review. November, 1965, pp. 92-97. [9] Hirschman, A.O. "The Political Economy of Import Substitution in Latin America.” Quarterly Jgurnal of Economics. February, 1968, pp. 1-32. I10] Hufbauer, G.C. Synthetic Materials and the Theory of InternationaI’Trade. Cambridge, Mass.: HarvardiUniversitngress, 1966. 71 [11] F [12] [13] .1141 [15] [16] [117] [418] [159] [20] [21] 72 . ”The Impact of National Characteristics and Technology on the Commodity Composition of Trade in Manufactured Goods.” In R. Vernon (ed), The Technology Factor in International Trade. New YOrk: NBER, Columbia University Press, 1970. Keesing, D.B. ”Labor Skills and International Trade: Evaluating Many Trade Flows with a Single Measuring Device." Review of Economics and Statistics. August, 1965, pp. 287-94. "The Impact of Research and Development on United States Trade." Journal of Political Economy. February, 1967, pp. 38-48. "Labor skills and the Structure Of Trade in Manufactures." In P.B. Kenen and R. Lawrence (eds.), The Open Economy, Essays on International Trade and Finance. New YOrk: Columbia Ufiiversity Press, 1968. Kenen, P.B. ”Nature, Capital, and Trade." Journal of Political Economy. October, 1965, pp. 437-60. Kenen, P.B., and Yudin, E. ”Skills, Human Capital and U.S. Foreign Trade." Columbia University International Economic Workshop, December, 1968, Mimeographed. Kindleberger, C.P. Foreign Trade and National Economy, New HaveniiYale University Press, 1962. Kravis, I.B. "Wages and Foreign Trade." Review of Economics and Statistics. February, 1956, pp. 14-30. ”Availability and Other Influences on the Commodity Composition of Trade.” Journal of Political Economy. April, 1956, pp. 143-55. Ohlin, B. Interregional and International Trade. Revised Edition. Cambridge, Mass.: Harvard University Press, 1967. Posner, M.V. "International Trade and Technical Change." Oxford Economic Papers. October, 1961, pp. 323-41. [22] [23] [24] [25] [26] [27] [28] [29] 73 Roskamp, K., and McMeekin, G. ”Factor Proportion, Human Capital and Foreign Trade: The Case of West Germany Reconsidered." Omarterly Journal of Economics. February, 1968, pp.152-60. Schultz, T.W. "Reflection on Investment in Man." Journal of Political Economy. Supplement, OctOber, 1962, pp. 1-8. United Nations Economic Commission for Latin America. Economic Bulletin For Latin America. March, 1964, pp. 1-49. United States Tariff Commission. Competitiveness of U.S. Industries. TC Publication 473, Washing- ton, DIC., April, 1972. Vernon, R. ”International Investment and Inter- national Trade in Product Cycle." Quarterly Journal of Economics. May, 1966, pp. i90-207. Waehrer, H. "Wage Rates, Labor Skills and United States Foreign Trade." In P.B. Kenen and R. Lawrence (eds.), The Open Economy, Essays in International Trade and’Finance. New York? Coiumbia UniverEity Press, 1968. Williams, J.H. Trade Not Aid: A Program For World Stabilimy. Cambridge Mass.: Harvard Univer- sity Press, 1953. Weiss, L. "Concentration and Labor Earnings.” Paper 6405, Social System Research Institute. Madison: University of Wisconsin, 1964. Chapter 3 EMPIRICAL TEST OF THE HECKSCHER-OHLIN THEORY: A BILATERAL TRADE STUDY 3.1 Introduction This part of the study attempts to test the H-0 theory by considering the commodity composition Of trade of the United States, the United Kingdom and South Korea. In the case of the United States, the novelty of the test lies in the employment of a more recent, hitherto lJnUSCd in this context, set of input-output tables for 15)63 [24]. It provides a basis for a new look into the s tructure of American foreign trade. In order to examine the generality of the H-0 tliecnyg it was considered desirable to test the hypothesis fc>r a developed Western European economy and for a less clexneloped economy. At the time Of this study the United KiJigdom was the only major Western European country for Mdiich the required data was available. The only other staudy of the U.K. pattern of trade was carried out by Hodd [11] for British trade during 1947-48. However, given the atypical nature of world trade in the immediate 74 75 post war years, a study of the commodity composition of U.K. trade for recent years may well be in order. In selecting a less deve10ped economy for a test of the H-0 theory, one encounters a dual problem. The detailed sectoral data required for an input-output study are seldom available for less developed countries. On the other hand, most less developed economies specialize in the export of primary commodities. However, the intent of this study is to consider the commodity composition Of trade in manufactured goods as well as the total trade of each country. It was therefore necessary to select among the less developed economies one that had a considerable amount of manufactured exports. With these requirements ‘the choice of a country becomes very limited, and South lmHm w: m a ..IMMNWW A\M . . o w to .ch m .wcm we . am we .mummp. . u a H e a H om sfiu\ev s a eea.a Ha.eaa mm.mwem ee.maa oo.aaam .a.m.m.p mom.a mm.nmm em.Hme mm.ooa ma.nmom meEocoom voccmHm maamuucoo Hoe.o no.6ma mm.~m~m oo.maw mm.es- moses nooom ome.o mm.aaa Hm.NH0m mm.wHN Aa.meam eaoe ae6.H mw.ama om.moA~ em.meH ~m.mamm eoaooee eaoea mme.a me.ama mo.mom~ ew.mma mo.ee~m eoaoo< emH.H ew.~ma em.aam~ om.maa ee.m~am ooeeoeoom eoooaosoo mood mmH.H am.mo~ mm.mmm~ ae.maa me.Hao~ eeeoeo mmm.a Hm.HmH nm.-m~ w~.mwa ma.mmmm eeeaeow soz eee eaaeooooe omo.a me.mo~ mm.eoam ma.wma He.mem~ eases ham.o ma.ea~ Ho.emm~ mm.mo~ mo.mam~ Noooeem sonoo mem.o eo.~aa m~.ewa~ ne.mma mo.eae~ .m.= smm.o mm.oaa ae.eoe~ mm.ema em.oae~ «you mem.o ha.mma me.aoa~ wa.oma He.mae~ emu AHO.H e~.mo~ mo.oom~ mm.oaa Ao.aoe~ seasoeoom eoooao>oa aeo.a ma.ao~ Hm.oomm mo.ama mowmma~_ name: a\e» Aoeo» eezv Aoooaav Aoeow ooze Aoooamv a x a e > sa Haw we a . meeomxm meeomzd nous enema .m.: cams Aeoeoaoeu mooaeH HH< .ooaoaeoaaou Haas mucoamomaaom uuoaaH O>Huwuoaaoo pom muuoaxm .m.: we mumaaoa cowaafiz Mom mucmsoufiavom momma use Hmufiamu mum OHQMH 84 Indeed the results (Table 3-4) indicate that the paradox no longer exists in the case of U.S.-world trade. Moreover, U.S. trade in manufactures with Japan and the centrally planned economies conforms with the theory. But the structure of trade with Africa, Latin America, Australia and New Zealand, and the U.S.S.R., still does not reveal the expected pattern. However, by confining the H-0 theory only to trade in manufactured commodities, a more accurate measure of factor intensity of trade will be Obtained when the nonmanufactured inputs are excluded from the estimate of factor intensities. Once this is done the relative factor abundance of U.S. vis-e-vis the rest of the world becomes more pronounced. The coefficients fall in all cases in Table 3-5. But the paradox ,prevails in trade with the same countries and regions indicated in the previous case. The U.S. results lend substantial support ‘t<3 a less generalized H-O theory which attempts to ex- IDQLain.the pattern of trade among countries with similar ]_ervels Of economic development. Whether the trade of 611.1. or only of manufactured commodities are included, Transatlantic and Canadian trade follow the H-0 pat tern. When only manufactures trade is considered, t}‘€3 (:omparative advantage of the U.S. in the production 81151 (export of the relatively more capital intensive 85 .NIm OHpoH mom .voeaaoxo mum .xwpcoam< N14 manna .Hnlmo muouoom mo muamcH "mouo:uoow umnuo cam moousom an eAH.H mm.~ma mm.mmam em.ema mm.o~m~ .m.m.m.: eHm.H ee.em~ mm.eamm me.~aa oe.awe~ moaeoeoom eoeeeao sfiaeeoeoo oem.o mm.mma om.eamN om.em~ am.mmaa moses sosom emo.o m~.mea em.amm~ no.ema mm.~ma~ mama NHA.H Ho.mea mm.mHH~ 6H.HNH ~N.~oe~ eoaooee eaoea mmm.a He.mma om.ma- o~.eHH mo.aamN eoaou< mma.a Ha.mma mo.~a- H~.~ma No.NmeN moeeoeoom eoooaosoo mood NH~.H o~.ewa -.emo~ aa.ama oe.moa~ eeeeeo see.a Hm.mma oh.HmmH em.HeH oo.emm~ eeoaeou 362 one eaaeeoooa aeo.a mm.me~ oa.ome~ mm.eaa No.~ema oeoee Nmo.a a~.eofi me.momm Nw.mma om.ma- Nooooem oesoo Nem.o Ho.oma mo.~aa~ oo.~aa em.amo~ .e.: eem.o me.ema HH.MaH~ oH.m~H em.~oa~ «emu Ham.o mm.eaa mm.ma- Na.maa am.oeoa emu amo.H NN.emH ee.oma~ oo.mea ao.meo~ moaeoeoom eoooaoson eeo.a mw.maa AH.HmH~ om.eeH na.~ea~ names a\x Amoeos ooze Aoooaav Amoeo» ooze . Aoooamv H s a a a x memooxm memomzo use: enema .m.s ohms «Apovsauxm mum muouoom ooa>~om we muaaaHv.mucoaoOmHaom uuoaaH O>Huauoaaoo cam muuomxm .m.D mo mumaaoo cowaaflz uom mucoaoufisvom Honda pom Hmufimmo mum magma 86 H mwo.H mqm. How. was. omq.a mHH.H HNO.H c~o.a mqm.a Hmw. ans. mum. moo. com. cma. 0mm. A\M> Hm.owH. Hm.MNH no.awa nw.NwH om.mma om.owH HN.~mH wo.mwa no.HmH Nw.mmH oo.omH mn.ona oo.owa om.~ma wH.owH om.omH Amummw cmzv A mHmomxm oo.mqom wq.mm¢N ca.qm¢~ mm.amm~ nm.mm<~ Hm.~nmN mn.¢am~ mm.owm~ mo.mqq~ H~.mamm mm.m¢oN q¢.wwq~ mm.~mqm Ha.mHmN wo.~o¢~ mo.mme~ Aoooaav e .NIm OHLMH mom .xfivcmam< .HI< oaan mom moflufipoaaoo emusuummsomalcoc mo umflH ecu pom o~.oma om.wmom mN.HoN mo.manm OH.NNN mm.wq- mH.qu Ho.mw¢N o~.~na om.qONm n¢.mma NN.mmom om.mo~ Ho.qmw~ no.¢wa Hq.womm wo.¢oa o~.~on~ ow.mma om.~mmm mm.m- Hm.OHmN ma.oma cm.oqo~ «c.0aa w~.omm~ mm.oma no.¢~¢N nn.mmH mo.o~c~ mN.mmH mm.oqm~ Aoumo» emzv Aoooamv A x mHmomZH "mMUOGuOOW Hmfiuo USN mGUHSOm a. .m.m.m.D moHEocoum woscmam haamuucoo mouom Sodom mfim< mOHuoa< sauna mowum< moasocoom vomoao>oa mmoq mwmcmu wcmHmoN Boz vow mwamuums< comma omousm nonuo N .M.: ¢Hmm 0mm mowaocoom emaoao>mn DAMO3 EHHZ mQ¢MH .m.D Omma «Avovaaoxm ow ovouh wcfiuauomwnomzlcozv mucoaoomammm uuoaaH o>HuHuomaoo pom muuoaxm .m.D mo mumaaoa :owaafiz pom muaoamudeom uonmq com Hmufiamo «In manna 87 commodities becomes more pronounced. In these cases, (Tables 3-4 and 3-5) the trade with Japan and the centrally planned economies also follows the H-0 pattern. It may be interesting to note that the so- called East-West trade, which is included in the category of the centrally planned economies, conforms to the forces of comparative advantage as explained by the H-0 theory. Among the socialist countries, the U.S. trades only with the U.S.S.R. and Eastern European countries. Trade with the latter constitutes two-' thirds of this category. Although the less generalized H-O theory allows for divergence of trade pattern of the U.S. with the less developed countries (the cases of U.S.-Latin /\merican and U.S.-African trade), the U.S.-Asian trade and i n particular the U.S.-South Korean trade consistently (:canform to the theory. l3..4 The Commodity Composition Of U.K. Trade The estimates of capital and labor (1960) and tilt) input-output table (1960) published by the Department (31‘ Applied Economics of Cambridge University [4,5] 1717C>vide the data for estimating the total capital and labor requirements of British industries. The trade ‘Jéltki as reported by the United Nations [21] were grouped t(3 Hiatch the industry groupings of the input-output .mlm manna mom "mouocOoom pom mooeaom .xfivcoaa<..aa< OHan mom moaufiwoaaou covsaoxo mo umfla ago you .eoesaoxo mum .xweooaa< ~14 macaw .anlmo paw .OHIH muouoom mo musacH 88 um emo.a wa.eea mm.~omm He.aaa oo.mmwm .e.m.m.: was. em.oea Na.eamm ma.mma em.mem~ moeeoeoom eoeeeao saaeooeoo mmm. no.0aa Aw.NmN~ mm.ee~ mm.~mma eoooe nosom owe. mm.oAH mm.mem~ am.em~ ea.maoN eao< ame.a Ne.NAH aa.mem~ em.~ea ma.aaa~ eoaooee edema meo.a an.eaa ma.eoe~ Nm.~mH mo.aae~ moaode mes. oh.oaa m~.eem~ mm.ama em.eee~ moaeoeoom eoaoaosoo mood Hao.a m~.maa nm.mm- ea.-a ea.moe~ eeeeoo oaa.a Ha.aaa os.-m~ aa.eea Na.omm~ eoeaeou 3oz eee oaaeoooee NNm. m~.aea H~.emm~ oe.nma ma.eaam eeoee Hoe. NN.meH om.mms~ Hm.mHN am.eao~ Noaooom ooeoo saw. oo.mea Hm.mmm~ aa.eaa ma.ee- .e.= sew. ea.oaa me.oem~ mo.ama mm.wm- aeem mam. oo.aea am.eam~ ea.mea mo.am- own was. ea.oAH ma.e~m~ ~m.oma ee.ea- moaeoeoom eoooaosoo ems. No.0aa mm.wmm~ Hm.HmH Hm.mom~ oases a\e isomos ease Aoooaav Amoeos eezv Aoooaac a > A x A x memomxm Smog PS: may 6:: ES Avopaaoxm mum ovmua cam musncH mcausuomMDGNZIcozv mucoaoomamom uuomaH w>auwuoaaoo pom mouoaxm .m.: mo mumaaon cowaafiz pom muooaouwsvom Honda pom Houflmmo mlm manna 89 table. In doing so, some difficulty was encountered in assigning commodities to industries since the detailed composition of each input-output sector was not reported. Thus the grouping of the commodities was based on the sectoral description of the input-output table which may be somewhat inexact. Before considering the relative factor intensity of U.K. trade, the relative factor endowment of that country with respect to its several trading partners must be examined. It is reasOnable to presume that the U.K. is capital abundant relative to the less developed, centrally planned, and "other EurOpean" economies. The relative factor endowment position of the U.K. with respect to the deve10ped economies is shown in-Table 3-6. The endowment position of the U.K. is not substantially different from the EEC (Italy excluded*), the EFTA (Portugal excluded**), and the Southern DominiOn countries. However, the U.K. is definitely capital abundant relative to Japan and capital scarce relative to the U.S. and Canada. The factor requirement of U.K. total trade is shown in Table 3-7. While the relative endowment position of the U.K. is somewhere in the middle of the [spectrum with respect to the world as a whole, the U.K.‘s :hnports are, on the average, more capital intensive than * Italy is excluded because its K/L endowment ' position of $2,600 makes it definitely capital scarce relative to the U.K. it Similar to Italy. 90 its exports. However, the results in Table 3-7 indicate that U.K. exports to the relatively capital scarce economies of Japan and "other Europe" are more capital intensive than the country's imports from them. Con- versely, its imports from the U.S. (contrary to Hodd's findings [11]), and Canada are more capital intensive than the U.K.‘s exports to them. Both situations confirm the H-0 theory.' While the trade pattern with the centrally planned economies is consistent with the H-0 theory, the relative factor intensity of trade with the less developed coun- tries, with the exception of South Korea, is contrary ‘to the prediction of the model. As in the U.S. case, little change in the pattern of trade is observed once the inputs of service sectors are excluded (Table 3-8). Exclusion of non-manufactured inputs and commo- dities, Table 3-10, results in almost identical coefficients of relative factor intensity when compared with the case that only manufactured inputs are excluded, Table 3-9. In comparison with Tables 3-7 and 3-8 some minor changes ier noted in the coefficients. However, the pattern 'remains unchanged with the only exception being that of ‘trade with the U.S.S.R., which no longer conforms to the ll—O theory. Summing up, the inclusion or exclusion of manu- facrtured inputs and trade does not alter the relative 91 Table 3-6 Capital/Labor Endowment Ratio of Selected Developed Economies (1964) Fixed Capital per Manu- facturing Employee Country or Economic Region (U.S. $) U.K. 4,000 Canada 8,850 U.S. 7,950 Japan 3,100 EEC (Excluding Italy) , 4,440 EFTA(Excluding Portugal and the U.K.) 4,590 Southern Dominions 4 4,150 Australia 5,3001 2 New Zealand 10,0001’ South Africa 3,000 lFixed capital per employee in all sectors. 2The estimate is unreasonably high and is not included in the ratio for the region. Source: See Table 3—1. 92 intensity pattern Of U.K. trade significantly. U.K. trade with the U.S., Canada, ”other Europe,” and Japan, all deveIOped economies, are consistent with the H-0 pattern. On the other hand, U.K. trade with the less developed economies of Latin America, Africa, and Asia with the exception of South Korea, cannot be explained by the model. Interestingly, the U.K. trade with Eastern Europe consistently conforms with the predicted pattern. Little may be said, however, about the trade with the EEC, EFTA, and Southern Dominion countries, given their similar relative factor endowment positions. 3.5 The Commodity Composition of South Korean Trade Capital and labor input coefficients for South Korean mining and manufacturing industries (1966) were obtained from national statistical sources [6,19]. Unfortunately, similar data for agriculture and the service sectors were not available. The Japanese input coefficients for 1955 were substituted for the missing data [7,8,9,10,25]. Although this substitution may result in some errors, it is believed that the input coefficients of South Korean argiculture and service in- dustries are substantially similar to that of Japan in 1955. The Japanese data were the closest substitute for the missing South Korean data. South Korea is unquestionably less capital 93 ._HNH aoum meme memuy .ng scum magma unauDOIusaaH .xflwceaa< .ml< manme CH wwucmmmua mucmflofimmmou uonmq wam Hmuflmmu "mmuusom .mofiuw< Lusom mam vcmammm Bmz .mfiamuum:mHmom:V vcm cfimmm .vcwamuH .maHmoH .mommuoq K ‘ . Ag\xv - a\x . .s m u » Hawaauom wagesfiuxm m HmuH caeaauxm N z Aq\mv H «mm.aq mo.o~w am.womm H¢.NH¢ Hm.omqm . maouam cumummm Nmm.o «H.maw ma.m-m m~.aom m~.-am .m.m.m.= omm.o ma.~mm ma.aomm mH.aHm a~.oq~m mmaaocoum emacmflm sHHmuuamu mfla.o “a.oom am.oomm qw.o¢oa mo.oaam mmuox nusom HHo.H ma.aom mm.¢~¢m mm.HmOH cm.mq~q mam< omo.H am.o~m m~.owcm ~m.mam Ha.~omm moaum< HRH.H mm.m¢m Hm.mamm mm.~ow Nm.m-m muauma< sauna HHH.H m~.omm «N.mmcm o~.mom mm.oaam amasoaoum umaoampon wwwa oma.o mo.mmw ~q.~mqm oo.¢om ma.amqm mmcoacaaoa anonusom Hwo.H o~.oam m~.mamm mH.mow m¢.~m~m manage Hma.o -.~mw oa.~mmn am.~ma qm.m~¢n gamma ahm.o oo.~aw n~.momm mm.cmm mm.ammm Hawsuuom mafia «maouam Hague mam.o ~m.aam q~.om¢q am.m~m Nq.Homm mmn omo.H “H.mmw mo.m~om ma.w~m Hm.wmom name: gxx Amumm» cmzv Am.oooflm mav Amumm» amzv Am.oooam may F . H mamomxm . mamomzH . meHs amaze .m.: mood Avmwsaocu wuaaaH HH< .mmfiufivoaaou HHwufiuwaaou van muuoaxm .M.D mo mumaaoa coaaawz umm mucwawufisvwm uonmg van Hmuflamo mum waAMH 94 ohlm OHfiflH. 00m "WOUOfiUOOM Hwfiuo USN meHfiom .vmv5Huxm mum Navamam< .ml< wanes aw.HmIm~ muouomm mo musmaH * mum.o cam.o mum.o cam.o Nmo.H moa.H ~¢H.H N¢H.H omm.o noo.H «Ho.o ¢ow.o mw~.o ~5N.H Noo.H oom.o Nmo.H H\M» H a. Hm.omo mm.nmn Ha.mah cm.eon o~.mmn m~.mmn NH.mo~ an.an wo.a¢h Hm.nnh Hm.aoh aH.~NN w¢.~u~ mw.0¢n Hm.¢¢h ow.am~ mo.n¢m a mm.a¢¢~ om.nmm~ mm.¢HmN ow.mo¢~ mm.uam~ nn.HmoN mm.mmn~ co.nmo~ NH.mom~ on.~om~ oa.eam~ ¢~.0HmN oo.mwmm mm.wmom mn.mnm~ om.omm~ aw.w-~ a Nw.m¢n m¢.aeh mN.Nhn m¢.¢mw mo.¢aw ma.woo mm.qu. em.¢mn 5N.amn .mc.omc «H.mmn Hm.oms nn.mun. Hm.mus mn.¢on mm.n- oa.om~ A mum «Hana me.omn mama mcoaawaon unusuaom m avenue aunmn Hmwsuuom mafia mnouam umnuo ¢ mmn QAMOB mBHB mnuwm mo uuaach wucmamomaamm uuoan m>wuwummabu van muuonxm 2M.= mo mumaaoa cowaawx uwm mucmamuaavmm gonna vnm kufiumu 95 .wlm manna mom "mmuOCuoow penuo vcm mmuusom .xfiwcmma< .HI4 maan mmm mmwuwvoaaoo vmuSuummscmancoc mo umHH ecu new « «mm.o mm.me Hm.ammm Na.wmm mm.o~qm maousm :uoummm mHo.H ~m.~am ma.HN~m mm.oom oo.~omm .m.m.m.: aqa.o om.omm oo.~mmm om.OHo HH.moqm mmHEOcoum emccmHm sHHmuucmo mom.o mm.mom Hm.Homm «R.MQHH Hm.qum mmuox zusom wqo.H No.omm oo.o~qm mm.mQOH wo.mmq¢ mHm< mOH.H aw.¢nm «n.0mqm Hm.mm~ «H.maom mUHuma mmm.H mm.mqm om.Hamm Hm.oao ~H.©¢He «UHuma4 aHumH qu.H aw.wnm mH.oaqm -.mom mH.mq~¢ mmHaocoom emQOHm>ma mmmH mmm.o oo.mmm mw.omqm 0H.mam 5H.m~mm mmnOHcHaoa aumnusom HHH.H Hm.mHm mm.qamm qw.wm~ om.m~Hm memcmu Nmm.o Hm.mqm mm.o~mm qa.~mm mm.~mqm gamma oqm.o ww.aom Ha.~mwm Hm.oqm cm.HNmm Hawsuuom maHa «oaousm Hague Hmm.o me.on m~.omqq Hm.wom Hm.m~mm m «Hum Nm~.H mH.omm NH.mmmm mm.mqm qm.~mmq N 0mm Noo.H «c.5mm mo.mwmm em.qu Hw.qHHm .m.= coo.H No.m~m mm.mqwm om.mqm m¢.wmnm mmHaocoom emQOHm>mo Nmo.H mo.omm mn.qmom 00.5mm “c.0qmm aHmoz Hmumm» amzv Hm.OOOHm mav Hmumm» cmzv Am.OOOHw mav H\z» a a memomxm A M memomZH zHHS MQfiufiummaoo cam muuoaxm .M.D mo mumaaom cowaaflz umm mucmamuflscmm uonmq use HmuHmmu m-m memH 96 umHH use you .Nlm UHQNH 00m ”mMHOn—UOOM HGSUO USN mmuhnflom .xfivamaa< .HI< manna mom wmfiuavoasoo wmvaaoxm mo .vmvsauxm mum xfivcmmm< .mn< manna ca HmImN vcm MIH muouomm mo munmcH * mma.o No.mom os.-mm oq.mam H¢.oamm maousm cumummm MHo.H ma.mom mm.oaHm No.Ham HH.mmmm _ .m.m.m.p mam.o a~.oqm om.o~mm aa.mmw o~.oomm mmHeanom emccmHa sHHmuuamu qow.o Ho.mam -.m-m oo.~mHH mo.¢mmm mmuox nuaom mqo.H ma.omm aH.ommm ma.amOH BH.quq «Hma qhH.H aw.mom_ mn.mqqm Hm.~an mm.mHem moHuua mmn.H am.qmm Hm.oqmm mo.oeo ha.MOHq «UHumaa cHumH omH.H om.aom cm.amqm «a.Hmm mm.mo~c meaoaoom emaon>ma mmmH oam.o -.¢aw mH.ooqm NH.~om Hm.mHmm mmaoHaHaoo cumsuaom mHH.H Ho.oom oo.~omm wH.mq~ mo.ahom «emcee HNa.o a~.mmw “H.mmmm wa.HNa ~m.ommm gamma an.o MH.oom mH.mmam m¢.¢mm Hm.¢wmm Hamsuuom msHa «maousm “mayo mam.o “H.500 mm.cha mm.mnw ma.wmqm m «Hem mm~.H mm.eam om.omnm 00.0mm mo.w¢n¢ N 0mm Hoo.H o~.maw ms.mmmm ~m.oow om.amom .m.= coo.H Ha.mow mH.mHmm mo.mmm -.mmam mmHaocoom emaOHm>oa moo.H Hm.oam mm.~oom mo.aqm qo.qomm aHmoz Hmummw amzv Hm.OOOHw mav Hmumm» gaze Hm.OOOHm mav H\x» H . x H x . memomxm mamomzH maHs mnema .x.a moaH «Avmvsauxm mum mvmuH van wusacH mcfiuauomwzcmzl cozv muamamomammm uuoaaH m>auflummaou vcm muuoaxm .M.D mo mumaaon cowaafiz umm muamawufisvmm honed new Hmufimmo oaum magma 97 abundant than any of the developed economies. The factor endowment position of South Korea relative to its less deveIOped trading partners is shown in Table 3-11. Because the required capital and labor estimates were not available for many less developed economies in Africa and Latin America, the capital/labor endowment ratios of the major trading countries with South Korea are con- sidered. In the Middle East, among the three major trading countries with South Korea, the capital/labor endowment ratio of Iran could be estimated. However, it is reasonable to assume that Kuwait and Saudi Arabia are more capital abundant than Iran which in turn is relatively more capital abundant than South Korea. Similarly, although data are not available for Burma, India, Pakistan, and Indonesia, it is safe to assume that each of these countries, and four of them together, are more labor abundant than South Korea. The estimates of relative factor intensity of fSouth Korean trade (1969) indicates that the country, in its total trade, imports capital intensive and exports .lahor intensive commodities (Table 3-12). In trade with developed economies the coefficients ()f South Korea-U.S. and South Korea-Australia and New Zealand indicate inconsistency with the predicted pattern. It is interesting to note that the relative factor 98 Table 3-11 Capital/Labor Endowment Ratio of Selected Less DevelOped Countries (1964) Country or Economic Region Fixed Capital per Employee (U.S. $) South Korea 315 Latin America 950 Chile Mexico Panama Peru * Africa 1950 Ghana Moracco Nigeria ** Middle East 1100 Iran Kuwait Saudi Arabia Burma, India, Pakistan, and Indonesia Not Available Taiwan and Thailand 400 Malaysia, Singapore, and Hong Kong 1200 * The estimate is unreasonably high. ** The given estimate is only for Iran. Sources and method of estimation explained in the footnote to Table 3-1 except here the KIL ratios include capital and labor of all sectors since similar data for the manufacturing sectors were not available for the above countries. 99 intensity of trade between South Korea and U.S., estimated with the American coefficients and input-output table, was consistent with “-0 pattern. However, using South Korean production conditions the results become inconsistent with the theory. It was shown in sections 1.4.2 and 1.4.3 that this form of results may be obtained if either the Samuelsonian factor intensity assumption or the assumption of equality of production functions among countries are violated. Moreover, the relative factor intensity co- efficients of South Korean trade with the less developed' countries, each group considered separately, confirm the “-0 pattern. 9 When non-manufactured commodities and inputs are excluded (Table 3-13), the relative factor intensity estimates of trade with all regions become consistent with the predicted pattern of the H-0 theory. 3.6 Conclusion The result of the study, as presented in this chapter, reveals that the H-0 theory provides a rather powerful explanation for the bilateral trade of nations. 111C theory faces no serious empirical contradictions as 1C)ng as the pattern of trade among either developed or lczss developed countries or regions are considered. The ressults improve even more if the analysis is confined to thc: trade of manufactured commodities. 100 manme udQudOIusmcH .HHNH Bonn muwumaumum ovmuy .Hma souw .xauammm< .qi< maamH cw omuamwmum muamwoamwmoo Hosea one Haufimmo “mmousom H x shawls. AH\MV H auo.H aa.mem mo.OmNm Nm.maa~ am.Hmm~ mace mac: mam .muommwcqm .mfimhmamz aaH.H MH.mmmm Ha.mHa~ H~.H~mm oa.~mHm aaaHaaae aaa aaaaaa mam.o wa.mmmm o~.amam oa.H~a~ mo.a~a~ aaaaaoaaH aaa .oaumwxam .afivoH .mauam maa.H o~.mam~ ma.m~Hm Na.HeaH Hn.aHmN uaam aHaaaz mm~.H mm.aowm oH.omHa o~.amoH am.amaH aaauma omm.o o~.moom mo.moom aa.mmam aa.mHm~ aoauaaa cauaa NHH.H mm.a~om ma.a~oa oa.omm~ oa.mma~ aaaaoaoam aamoHa>am aaaa mam.o mm.a~m~ ma.mmw~ oa.mo~m mm.moo~ aaaHaaN aaz a aaHauuaaa mam.H Na.aaam wH.moa~ mo.HmmH Hm.mmm~ caman maa.H mm.mmm~ ~a.mma~ Na.amaH mm.~om~ .x.= acm.H ma.mam~ aH.aom~ ~o.mmaH mo.mmm~ amounm auauaa: Haa.o am.mmm~ ~a.ammu mH.mmH~ m~.mamH .m.= mom.H mm.amHm om.w~m~ mm.Hmo~ mm.oa- aaaaoaoum aamoHa>am maN.H ma.maHm mm.nmm~ ~a.mmo~ oa.mom~ mHmoz maaaa» aazv AcooHa .m.=v Aaaaa» aazv mooon .m.=v H\x> H M mHmomxm A .mHfiOMZH M mHHZ MQ¢MH z¢umox mhbom moma AomvsaocH muaaaH HH< .mmuuanoaaoo HHauwuoqaou can muuoaxm Nalm UHAMH cmmqu nusom mo muamaon cowaawz pom mudmfimuwaumm Hanna can Hauwmmo 101 H mo umwa m pom mNH.H wmo.H o~w.o NH®.H NmH.H mmN.H omm.o HNH.H me.H. omo.H MHH.H HN©.H eon.H NHq.H m\a> mm.wana qq.o~ma mm.mmma mo.mma om.wqmm «n.5HnH Ho.NooH «0.00ma mH.wo~H Hm.HwMH mo.HNoH w~.momH mo.omwa Hm.mmwa Awumm» cmzv A mHmomxm mm.ama~ o~.ommm 0H.HHHM HH.O¢©N mn.oqom qo.mmo~ mq.co~m mm.oH- No.Ho~N cm.owm~ mm.mmn~ Hw.quN Ho.mqmm oo.woq~ ma.OOOHm mmv M .xfiuammm< .HI< wanna mam .aaaaHaxa aaa xHaaamm< .aua aHaaa am mmimm .NHim aHaaH mam Na.maNH mm.Hoa~ am.mamH Na.aHaN aa.waaH mm.aNHN ma.HOm am.mmaH mH.aaHH aa.~ao~ am.maOH aH.amHN H~.ma~H aa.mmo~ ma.mNmH ma.mmHm aa.aoNH cm.HmN~ mn.~mHH aH.n~HN m~.anHH mo.maH~ am.amHH ~m.mm- No.awHH mm.am- ma.mmHH aN.mm- Aaaaam aazv ma.OOOHm mam a a mamomzm "mmuocuoom umsuo use we uuaom mmfiufivoaaoo vmusuoamsamaiaoo com mid mucuomm mo musmoH « wooM waom a muommwaam .mwmmmHmz vdaawana a cmaama mammcoocH .cmumfixmm .mHocH .mausm aaam aHaaHz mofium¢ mafiuma< :Humq mafiaoaoum wmmoam>mn m ocmammN 3oz a mwamuums< comma .M.= amouam cumummz .m.: mag mafiaoooom vmmoam>mn mHHB ma¢MH ZHuHumaaoo vow muuonxm ammuox cusom mo muaaaon cofiHHHz mom wucmamuaavmm momma mam Hmufimmo MAIN maan DAMOZ mama 102 Moreover, these findings are contrary to Linder's assertion ||7| that countries with similar levels of economic development will tend to exchange manufactured commodities with similar characteristics among themselves. l1] l4] [5] 1"] l7] l8! [9] [10] [11] [12] REFERENCES Baldwin, R. "Determination of the Commodity Structure of U.S. Trade." American Economic Review. March, 1971, pp. 126-46. Bank of Korea. Economic Statistics Year Book 1968. Seoul, Korea, 1969. Creamer, D., Dolseovolsky, S.P., and Borenstein, I. Cgpital.in Manufacturing and Mining) its Formation and Financigg. Princeton, N.J}: Princeton University Press, 1960. Department of Applied Economics, Cambridge Univer- sity. A Program for Growth: Production, Capital and Labor. Cambridge,’Mass.: MIT Press, 1963. ____. A Program for Growth: Production, Capital and Labor. Cambridge,_Mass.: MIT Press, 1963. Economic Planning Board and Korean Reconstruction Bank, Report of Minigg and Manufacturing Census, Series I-- Basic TablesL 1966. Seoul, Korea, 1967. Economic Research Institute, Economic Planning Agency. National Income Accounts. Tokyo, Japan, 1957. National Wealth Survey. Tokyo, Japan, 1955. . Economic Bulletin No.1. Tokyo, Japan, February, 1939. . Capital Structure of Jgpanese Economy. ToKyo, Japan, 19567 Hodd, M. "An Empirical Investigation of the “-0 Theory.” Economica. February, 1967, pp. 20-29. Hufbauer, G.C. "The Impact of National Characteristics and Technology on the Commodity Composition of Trade in Manufactured Goods." In R. Vernon (ed.), The Technology Factor in International Trade. New York: NBER, Columbia University Press, 1970. 103 [lb] NH [18] [19] [20] Ill] [22] [23] [24] L251 Output Structure for 1963.” 104 Labor Office. Year Book of Labor Geneva, 1966. International Statstics. Kenderick, J.W. and Labor . Press, 1961. Productivity Trends: Capital New York: NBER, Princeton University Leontief, W. "Domestic Production and Foreign Trade, The American Capital Position Re- Examined." Economia Interazionale, February, 1954, pp. 9-45. ”Factor Proportion and the Structure of American Trade: Further Theoretical and Empirical Analysis." Review of Economics and Statistics. November, 1956, pp. 387-107. Linder, S.B. An Essay on Trade and Transformation. New York: Wiley, 1961. Miernyk, W.H. The Elements of Input-Output Analysis. New York: Random House, 1965. Input-Output: Seoul, Research Department, Bank of Korea. Interindustgy Relation Tables for 1963. Korea, 1966, Table 4. B.C. "Capital Shortage and Labor Surplus Review of Economics 286189. Swerling, in the United States.” and Statistics. August, 1954, pp. United Nations. Statistical Papers, Series D. 1970. Commodity Trade Statistics. New York, 1969, . Year Book of National Accounts Statistics. _« New York, 1966. United States Department of Commerce, Bureau of Census. U.S. Foreign Trade Statistics Classification and Cross CTassification 1970, Washington, D.C., 1970. Economics. "Input- Survey_of Current Business. November, 1969, pp. 16—47, Table 3. , Office of Business Watanabe, T. "Approaches to the Problem of Inter- Country Comparison of Input-Output Relations.” in United Nations. International Comparison of Interindustry Data. New York, 1969, pp. 187-210. Chapter 4 THE HUMAN CAPITAL THEORY OF INTERNATIONAL TRADE: A BILATERAL TRADE STUDY 4.1 Introduction and Methodology In this chapter the bilateral trade structure of the United State, the United Kingdom, and South Korea will be examined in an effort to test the human capital theory of international trade. Following Keesing [4,5], the labor input requirement of each industry is classified into six categories: 1. Professionals and technicians. 2. Administrators and managers. 3. Clerical workers. 4. Sales workers. 5. Manual workers. 6. Service workers. Let sti represent the number of workers of cxztegory t (t=1,2...6) required to produce one million ch>11ars of value added in industry i. Then the total Teanirement of group t workers by the jth industry 105 106 will be n E =Z—s r, (1) tj 1=l ti 13 where r is an element of the total requirement input- iJ' output table of the economy. Morevoer, let mj and xj represent the value of the jth commodity (industry j's output) in a representative one million dollar bundle of manufactured imports and exports respectively. Then the total requirement of the t category of workers per one million dollars of imports and experts will be (St)M and (St)X respectively. Where: n n (3 ) = 25: ( IE: 5 r ) n3 1: M J=l 1:]- ti ij (2) n n (St)X g 1.4:; ( i=lsti rij) "j Substituting (1) into (2) the following relations may be obtained: n (St)M = £Z: stj mj (3) . n , (St)X = Z S x 107 However to determine the relative skill intensity of trade, the six categories of workers must be classified into skilled and unskilled groups. Here the following three measures of relative skill intensity of trade appear plausible: (Sl)M / <35>M )5: (4a) (Sl)x / (SS)X XS - (31>M / [(34)M + (s5)M + (S6)M] (4b) 2 _ (51)x / [(34)x + (35)x + (86)}{1 i(si)M + (82%] /[(S4)M + (SS)M + (8654] ) (4c S = 3 [(51% + (52%] / [(54)x + (Ss)x + (S6)Xi Substituting (3) into the above equations, the Ct3efficients may be estimated by: 1'1 n 2 ~ Z — XS j=l Slj mj / j=l 85;] mj (53) l = n n Z E x. / Z 3 x 108 J=i t=4 3:]. K8 = 2 nu 26 n ‘_ (Sb) ‘2‘ E x. / - S X j=l 13 J t=4 J=1 t3 J 2 n 6 n Z 2: ti mi] / _. Z: " m t=1 J-l t—4 J-l t3 J . (5c) 4 J=l H J xi: 2 n 6 [Z Z End/[53: 2.1.] The problem involved in selecting one definition of skill intensity over the other two concerns the vari— ations in training and educaitonal requirement for different categories of workers in different industries and countries. Group 2 of skill categories (administrators and managers) contains the most notable variations in the level of training and education. The most consistent cartegories are the professionals and technicians (group 1) aiid manual workers (group 5). Therefore, ‘X: is considered tlie most reliable and. 635 the least reliable index of rc:1ative factor intensities. The skill coefficients of selected industries of tlue U.S. and the United Kingdom were obtained from a nuinpower study by Horowitz, Zymelman, and Hernstadt [1]. llowever, similar data was not available for South Korea, 109 so as an approximation the Japanese data for 1950 were employed. The Japanese coefficients were the closest available substitute for the missing data. The input- output tables and the trade data were obtained from sources explained in Chapter 3. 4.2 The Commodigy Composition of U.S. Trade The United States may be considered relatively skill abundant compared to the less developed and to the centrally planned economies. The relative skill endowment of the U.S. with respect to the developed economies is shown in Table 4-1. Estimates indicate that the U.S.is skill abundant relative to any one of the deve10ped countries except Sweden. However, taking the EFTA as one group, the U.S. is definately more skill abundant than any of its trading partners. The coefficients of relative skill intensity of U.S. manufacturing trade is presented in Table 4-2. The estimates of 315 and X25 give almost identical results. U.S. exports are relatively more skill inten- sive than its imports in its trade with the world as a whole and with each economic region considered, except the U.S.S.R. With 'XSS as the index of relative skill intensity, U.S. exports remain more skill intensive 110 Table 4-1 Skill Endowment Ratio of Selected Developed Economies (1961-1964) Skilled Employees as a Percentage of Country or Economic Region Total Labor Force United States , 0.108 Canada 0.106 Japan 0.049 EEC 0.081 * France 0.083 Netherlands 0.092 Belgium 0.080 Germany 0.100 Italy 0.046 EFTA 0.088 * United Kingdom 0.095 Norway 0.080 Sweden 0.129 Austria 0.068 Denmark 0.078 Portugal 0.027 Other Europe 0.047 Greece 0.034 Iceland 0.046 Ireland 0.071 Spain 0.041 Yugoslavia 0.056 Australia 5 New Zealand 0.101* Australia 0.103 New Zealand 0.093 *Estimated by Hufbauer [2]. Source: The ratios indicate the percentage of professional, technical, and related workers, category "0" of ILO classification, in the total active population. Data from the International Labor Office [3] except those with asterisks. 111 HON.H Hon.o mqm.o moo.o «oo.H waw.o aHmow:% mam .cfimmm .wamHmHH .vcmamoH .MUmmuo a qom.a wmo.o mqm.o wnq.o «mm.o man.o «No.0 nom.o oom.o 505.0 omc.o wmm.o own.o mmm.o How.o Hum.o H» m .m.m.m.: mafiaocoom veacwam maamuucmu menox nuoom mfim< mofiuma< cfiumq moauw< mafiaoooom Ummoam>mn mama moacmo vcmaamw 3oz w mHHmuums< semen enouam penuo .M.D «Hum 0mm mmwaocoom ommoam>ma an names mHHz madMH .m.D OmmH mo saHaaaaaH HHHxa a>HaaHaa mo aaaaHaHmmaoo Nlo oHAMH 112 than imports in its trade with the world as a whole and with most of the economic regions. However, in two cases, Australia and New Zealand and Latin America, reversal occurs. This, and the fact that X 35 coefficients are consistently larger in magnitude than ‘X15 and ‘X25, indicates that the U.S. comparative advantage is less in the commodities which require relatively more administra- tive and managerial workers than those that require more professional and technical workers. 4.3 The Commodity Composition of U.K. Trade The relative skill endowment of the U.K. is shown in Table 4-3. Relative to its trading partners the U.K. is less skill abundant compared to the United States and Canada, has an almost similar skill endowment ratio compared to the EFTA and the EEC, and is relatively more skill abundant vis-a-vis Japan, ”other Europe," and the Southern Dominion countries. It is also presumed that the U.K.‘s endowment ratio of skilled to unskilled workers is greater than any one of the less developed regions. An examination of the relative skill intensity of U.K. trade in Table 4-4 reveals that the U.K. is an exporter of relatively skill intensive commodities in its trade with the world as a whole with "other Europe,” Japan, the Southern Dominions, less deve10ped, and 113 Table 4-3 Skill Endowment Ratio of Selected Developed Countries and Economic Regions (1961-1964) Skilled Employees as a Country or Economic Region Percentgge of Labor Force a United Kingdom 0.095 United States 0.108 Canada 0.106 Japan 0.049 EEC (Excluding Italy) 0.092 EFTA (Excluding Portugal and the U.K.) 0.091 Other EurOpe, plus Portugal 0,044 Southern Dominions 0.061I Australia 0.103 New Zealand 0.093 South Africa 0.019 * Estimated by Hufbauer [2]. Source: The ratios indicate the percentage of professional,teehnical, and related workers, category "0" of ILO classification, in the total active population. Data from the International Labor Office [3] except those with asterisk. 114 .xfiecmamm .wI< maan ”meadow .mofiuw< auaom a .mcmammm Bez .mfiamuums< q .mw>oamowsw a .cfimmm .mcmHmHH .mamamoH .mummuu m .Hmm5uuom wowesfioxm N .mamuH wcHeDHoxm H mum.o owo.o omo.o mmousm cumummm wmw.o oN~.o omn.o .m.m.m.n «mm.o mom.o mom.o mmflaoooom emccmam maamuucmu qmq.o amm.o Hmm.o mmuox Lusom mwm.o eqq.o H¢¢.o mfimm 5mm.o mmn.o How.o magnum Head mmmé mom .o 3:92 523 mom.o Hmo.o omo.o mafiaocoom vmmoam>mn mmma cow.o omn.o 5mm.o mmcofiafiaoo aumnusom mmo.H 5mm.o ooo.a memomo wam.o mmn.o om~.o cmmmh amm.o aHa.o NHa.o Hawaaaom aaHm mamoaam aaaao oao.H mmm.o Nmm.o N mHmm oqo.a moo.a o~o.a H 0mm «QH.H «HN.H MAN.H .m.D amm.o mam.o mNm.o aaHaoaoam aamOHapam Nom.o omm.o wNm.o QAMQ3 m» m» w» meHz amaze .x.s mamH somwcflx vmufia: ecu mo mmmue wsfiusuomwscmz mo hummomucH HHfixm m>HumHmm mo mummeHmwmoo etc maan 115 centrally planned economies. On the other hand, the contrary is true for U.K. trade with the U.S. and Canada. Both situations confirm the human skill theory of trade. The above relations prevail.regard1ess of the coefficient of factor intensity that is used with the exception of Canada. Trade with Canada is consistent with the theory if the 'X15 and X35 definitions of relative skill intensity are used and is incensistent with the Xzscoefficient. However, given the similarity of skill endowment ratios of the U.K., the EEC, and the EFTA, no statement may be made with regard to the pattern of trade with these economic regions. 4.4 The Commodity Composition of South Korean Trade The relative skill endowment of South Korea compared to other less developed econOmies is presented in Table 4-5. South Korea is less endowed with skilled workers than are Latin America, Malaysia, Singapore, and Hong Kong. However, skilled workers are more abundant in South Korea than in the Middle East, Burma, India, Pakistan, Indonesia, Taiwan, and Thailand. And the relative endowment position of South Korea is similar to that of Africa. A comparison of South Korean skill endowment with that of the developed countries, shown in Table 4-1 supports the presumption that the 116 Table 4-5 Skill Endowment Ratio of Selected Less Developed Countries and Economic Regions (1961-1964) Country or Region South Korea Latin America Chile Mexico Panama Peru Africa Ghana Morocco* Nigeria Middle East Iran Kuwait Saudi Arabia * Burma, India, Pakistan, and Indonesia Taiwan and Thailand Malaysia, Singapore and Hong Kong Skilled Employees as 3 Percentage of Labor Force 0.022 0.037 0.022 0.015 0.017 0.016 0.035 * . Data not available. Source: See Table 4-3. 117 country is less skill abundant than any group of developed economies. The commodity composition of South Korean trade showed in Table 4-6 indicates that its trade with all of the developed countries conforms to the human capital theory of trade. However, inconsistencies betweerithe theory and empirical findings arise in the cases of trade with Africa, the Middle East and with Burma, India, Pakistan, and Indonesia. Among less developed countries, only the trade with Latin America and Taiwan and Thailand confirm the theory. The above situations exist with all three definitions of relative skill intensity of trade. 4.5 Conclusion The human capital approach to the factor pro- portion theory of international trade performs rather satisfactorily in explaining the commodity composition of manufacturing trade. The results of the empirical tests concerning the bilateral trade of the U.S., the U.K., and South Korea reveal a general consistency between the theory and the actual flow of commodities. The bilateral trade of the U.S. and the U.K. with developed and less developed economies conforms with the predicted pattern of trade. The same hold true for the trade of South Korea with developed economies. However, the trade of South Korea with other less developed economies 118 .xfimcma94 .oal< maeme "muuaom mnm.a mNm.o oHN.H mua.~ moo.~ om¢.H mmm.H moo.H mmm.H HNm.H NQH.N Hmm.H nmw.H omn.H mmw.a wmn.o mnm.~ Heo.q me.MH NNm.¢ nwm.~ HmH.a mn mmmq vcmammu 3oz m mfiamuum=< cmmmh .M.= maouam cumummz .m.= mafiaocoom momoam>ea n4mo3 :HHB MQHuaHam mo auaaHUHmmaoe mum mHan 119 does not conform to expectations based on the human skill hypothesis. The failure of the theory in cases of trade among less developed economies may be the result of significant variations in the definition of skill categories of workers in these countries. However, the problem may not be wholly a statistical one. Skill intensity reversal may be an empirical reality among less developed economies. Some skill intensive commodities may be produced by highly capital intensive methods in a country lacking the required skills to the extent that modern machinery may be substituted for skilled workers. These variations in skill requirements of industries may not be substan- tial enough to change the pattern of trade between developed and less deve10ped economies, yet they may be significant among less deve10ped countries which have rather small differentials in their skill endowment ratios. In the absence of any serious empirical test of skill intensity reversal among less developed countries (essentially due to a lack of data), the above proposition remains speculative in nature. Furthermore, this inconsistency between the prediction of the model and the actual flow of commodities may be, at least partially, a result of extensive trade restrictions in the less developed countries. [l] [2] [3] [4] [5] REFERENCES Horowitz, M.A., Zymelman, M., and Hernstadt, I.L. Manpower Reguirement for Planning: An Inter- national Comparison Approach, Volume II. Boston: Northeastern University Press, 1966} Hufbauer, G.C. ”The Impact of National Characteristics and Technology on the Commodity Composition of Trade in Manufactured Goods.” In R. Vernon (ed.), The Technology Factor in International Trade. New York: NBER, Columbia University Press, 1970. International Labor Office. Year Book of Labor Statistics. Geneva, 1966. Keesing, D.B. "Labor Skills and International Trade: Evaluating Many Trade Flows with a Single Measuring Device." Review of Economics and Statistics. August, 1965, pp. 287-94l "Labor Skills and the Structure of Trade In Manufactures." In P.B. Kenen and R. Lawrence (eds.), The Open Economy, Essays on Inter- national Trade and Finance. New Yofk: NBER, Columbia University Press, 1968. 120 Chapter 5 NEOTECHNOLOGY THEORIES OF INTERNATIONAL TRADE: A BILATERAL TRADE STUDY 5.1 Introduction The distinct charactersitic of neotechnology theories of trade is their consideration of technological requirements of production as the predominant force in determining the commodity composition of trade in manu- factured goods (see section 2.3). In this chapter the four neotechnology theories of trade will be tested through an examination of the trade structure of the U.S., the U.K., and South Korea. 5.2 Scale Economies According to the scale economies hypothesis the main determinant of the comparitive advantage of a country is the extent of the scale economies enjoyed by its manufacturing industries [2]. Producers located in economies with large markets are able to take advantage of economies of scale and can therefore produce their outputs more cheaply than can producers in smaller markets. Thus the theory maintains that large economies will have a 121 122 comparative advantage in the production of those commodi- ties where production is characterized by increasing return to scale. In testing the scale economies theory, however, economies of scale for various industries must be estimated. This may be done by measuring increasing return to scale as reflected in the production functions or by considering the cost functions directly [11]. Engineering data may be used to estimate the relationship between inputs and output. However, in this approach the nontechnical aspects of production are not considered. Thus, the plant or the process function, as Walters [11] calls this class of production functions, may not be relied upon in estimating economies of scale. Although it is possible to observe increasing return to scale in the plant, the producer may not realize economies of scale due to high cost in the nontechnical aspects of production. Alternatively, the extent of increasing return to scale may be found by estimating the production function of an industry as a relationship between inputs and the value of output in a cross-sectional study of the firms in the industry. In this method output is measured by the value added of a firm in the industry, with capital and labor being considered as inputs. In addition to the specification problem and other econometric constraints, 123 the most troublesome aspect in estimating production functions is the difficulty in considering various qualities of inputs in various firms within an industry. By taking the number of workers as a measure of labor input, labor quality as reflected in skill and age composition in different firms is not accounted for. A similar difficulty exists in estimating the capital input into the production process. The apprOpriate measure of capital input is the estimate of the flow of capital ser- vices. However, by taking the stock of capital of a firm as the capital input, not only is the degree of utilization of the factor not reflected in the measure, but the similarity of composition of stock of machinery, equipment, building, and land is also assumed. Consequently, to the extent that the larger plants may employ higher quality capital and labor than the smaller plants, the resulting efficiency observed may not be attributed to scale alone. The difficulty in estimating the economies of scale directly from the cost function lies in determining the cost of production for various firms under imperfect market structure [11]. The cost of production is defined as the difference between total revenue and the return to entrepreneurial capacity. In perfect competition, where the return to entrepreneurial capacity is the opportunity cost of that factor, the price of the output 124 may be considered as the long run average cost [11]. However, in a market characterized by monOpolistic elements, where specialized factors are controlled by the firm and are not marketable, the true nature of the cost function may not be estimated. In this study, however, an additional difficulty is encountered in estimating the scale economies. The test of the scale economies theory requires measures of economies of scale for over one hundred industries, a task beyond the limits of this study. Furthermore, no estimate exists of production functions or cost functions for such a large number of industries. Consequently, for the purpose of testing the scale economies theory of trade, two measures which approximate scale economies by methods other than those explained above will be utilized: scale elasticity and optimum plant size for a large number of U.S. industries as estimated by Hufbauer [3] and Saving [5], respectively. Hufbauer estimated scale elasticity of U.S. industries based on 1963 Census of Manufacturing [9] by the following method. Let Vi represent the ratio of value added per worker for a given size class of plant to the average value added per worker for all establishments in that industry, and let N represent the average number of workers employed per establishment in a given size class 125 of plants. Then: mi Vi = kN where k is a constant andixi is the coefficient of scale elasticity for industry 1. The coefficient'cK indicates the percentage change in output per worker as the result of a percentage change in the plant size. However, to the extent that larger plants may employ more skilled workers or a higher ratio of capital per worker than the smaller plants, the above estimate of scale economies is exaggerated. With this deficiency in mind, Hufbauer's estimates of scale economies is applied to the trade structure of the U.S., the U.K. and South Korea. Although it is plausible that the scale coefficients may vary among countries, in the absence of similar estimates for the other two countries, it is reasonable to assume that the ranking of these co- efficients is similar for manufacturing industries * among countries. The scale coefficients are directly applied to a representative one million dollar bundle of manufacturing imports and exports of the above three countries in their trade with the world as a whole and with several economic fl: The same problem is encountered in applying the U.S. coefficients to the other two countries in tests of technological gap and product cycle theories, and using the Japanese coefficients for all three countries in the test of the stage of production theory. 126 Table 5-1 Total Manufacturing Output and Gross Domestic Product Per Capita for Selected Developed Economies (1964) Country or Economic Region Total Manu- GDP/Capita 2 facturing Output U.S. $ in Billions of 1 U.S. $ U.S. 173.04 3000 Canada 10.55 2110 Japan 21.56 720 EEC 19.04 (95.20) 1464 EEC(Exc1uding Italy) 19.45 (77.80) 1640 France 27.53 1580 Netherlands 5.55 1430 Belgium 4.11 1460 Germany 40.61 1770 Italy 17.40 1030 EFTA 7.75 (46.49) 1557 EFTA (Excluding Portugal & U.K.) 3.18 (12.71) 1650 U.K. 32.22 1710 Norway 1.81 1880 Sweden 5.62 2100 Austria 2.90 1030 Denmark 2.38 1680 Portugal 1.56 420 Southern Dominions 3.72 1454 Australia 5.63 1810 New Zealand N.A. 2046 South Africa 1.88 507 1The figure for each region is the simple average of the manu- facturing output of the principal trading countries in that region. EFTA and the EEC may be considered as unified markets, the figures in parentheses which are the total market size represent this alternative measure. 2The figure represents the GDP per capita of the principal trading countries in the region. Sources: U.N. [8]. 127 Table 5—2 Total Manufacturing Output and Gross Domestic Product Per Capita for Selected Less Developed Economies (1964) Country or Economic Total Manu- GDP/Capita2 Region facturing Output inl In U.S. $ Billions of U.S. $ South Korea 0.51 140 Latin America 0.91 409 Chile Mexico Panama Peru Africa 0.14 160 Ghana Morocco Nigeria Middle East3 1.54 224 Iran Kuwait Saudi Arabia Burma, India, Pakistan & Indonesia 2.65 75 Thailand & Taiwan 0.36 117 Malaysia, Singapore, and Hong Kong 1.055 301 l’ZSee Table 5-1. 3The estimates are only for Iran. Sources: U.N. [8]. 128 regions. These results are presented in Part A of Tables 5-3, 5-4, and 5-5. However, since the a< coefficient for industries may be either positive or negative, the measures of scale economies emboided in imports and exports may also be either positive or negative. In order to obtain a consistent estimator of the relative scale intensity of trade, the coefficients of scale economies of imports and exports were transformed in the following manner: X SC =(c+o / where XEC is the coefficient of relative scale intensity of trade, and 0mamowmw paw .Hmwnuuom .cfimam .wcwamoH .pcmHmHH .mommuo H Hum .%.+fi8:f+~ssu x Nm~.o omo.mm mas.o seqo.o somo.o nawH.o- .m.m.m.: mmq.o HN¢.HN ns~.m ono.o somo.o mmqo.o- mausocoum emccmam saamuuamo omo.o Noa.mfi mmw.o amm~.o mnqo.o o~ao.o- «whoa nusom MNH.o Hmn.afi mmH.~ mmao.o oooo.o wafio.ou mam< omq.o mas.ma wom.m maqo.o ammo.o msmo.ou monuma< shame m-.o omo.H~ amm.q Hams.o o-o.o Hamo.o- muauwa Na~.o mm~.mH o~o.m mmmo.o coco.o mmmo.o- mmuaocoum emaoam>mn mama ~mo.a o-.¢~ qm¢.o~ maao.a ommo.0 mmmo.o memcmo mNH.H mom.- sam.m~ smmo.o mmoo.o Nomo.oa wcmfimmw 3oz use «Hamuum=< oa~.o mmm.o~ mos.oa Hmom.o oweo.o Hmso.o magma mmH.o mmo.am mwm.m omflm.o mmmo.o mmqo.o Hmaousm “mayo moa.o Raq.na omm.ma NNmm.o omno.o mamo.o .x.: «00.0 mwo.o~ Nom.~H mamm.o onco.o mamo.o «Ham Nom.o oom.- moo.o~ Homm.o coco.o Naqo.o 0mm mom.o oNo.mN HHo.o~ comm.o quo.o maso.o mmaaocoum emQOHm>ma amn.o omn.- omN.NH Howm.o mmoo.o «mmo.o name: Aaaeuiuv Anus lac x5 2. Now» memomxm memomzH Hum» memomxm mamomzH . . oomemz m.ozH>mam00:» 0cm .cwmam .0cmamuH avcmamoH .mommu00 Axd + 0.00\Azd + 0.00 uw0m> x» 000.0 000.0H 000.5 0000.0 0000.0 0000.0 mucusm aumummm 005.0 000.0 000.0 0005.0 0000.0 H000.0I .m.m.0.0 000.0 000.00 000.0 0000.0 HH~0.0 0000.01 mmancoom nuccmam 5HHmpuam0 050.0 0H0.50 H00.H 0000.0 5500.0 5000.0: mmuox Lusom 00H.0 500.H0 000.0 0000.0 0050.0 0000.0: mwm< 000.0 050.5H 000.0 0000.0 H000.0 0000.0! mowum< 000.0 000.0H 000.5 0000.0 0000.0 0000.0: mofiuma< cfiumq m5~.o 000.0N mmh.m wmom.o Ncoo.o quo.o- m0050coum cmaoam>mo mama 000.0 005.0H 000.0 0000.0 0000.0 00H0.0 0mGOHGHBoQ dumSuzom 550.0 000.0H 000.0 0000.0 0000.0 0000.0 mcmumo 000.H 000.0 000.0H 0000.H 0000.0 0500.0 cmamh 0H0.0 050.0H 000.0 0000.0 0000.0 0H00.0 Hmwsuuom mafia 0oaousm awnuo 000.0 0H0.0H 0H5.0H 5000.H 0000.0 0000.0 0m0 000.0 000.0H 000.0H 0000.0 0000.0 0000.0 00003 sax: :0 3 3., x5 :5 00m» « > 4 mHH3 00 000.0 005.00 000.0 000.0 000.0 000.0 050.0 000.0 000.0 000.0 000.0 000.0 000.0 000.0 A000 mHmomxm Domemz m.UZH> .050 0000.0: 0000.0 0000.0: 0000.0 0000.0: 0000.0: 0000.0 0000.0: 5000.0: 0500.0: 5000.0: 0000.0: 5000.0: 0000.0: 0 0 mHmomxm oomHmE 0.0m34mmmmld 0000.0 0000.0: 0000.0 0000.0: 0550.0 0000.0: 0000.0: 0000.0 0000.0 5000.0 0000.0 0500.0 0000.0 0000.0 2 d mHmomfi0 .z.0 0cm .x00amaa< .00:< 0cm .00:< mm0an “mmousom .Ax5 + 0.00\025 + 0.00 u0om>¥ 0com weer 0cm .muommwc0m .m0mzm0mz 0:000m00 0cm c030m0 m0mmdo0c0 0cm .cmum0xmm .m00C0 .mEpsm ummm 000002 000004 mo0pma< a0um0 mm0Eocoom 0m000m>wo mmm0 0sm0mm0 3oz 0cm 000muumz< cmmmh .M.0 maousm cumummz .m.0 mm0Eocoom 0m000m>mo Q0003 $903 mom0 mmm0 «cacao 0cm0w00 3oz w m00muum=< amamw onousm umnuo .x.0 «Hum 000 mm0soaoom 0o000m>m0 0 00003 mHHS MQ mmma.o m0m~.o ooH~.o 0000.0 0000.0 0050.0 5000.0 5000.0 0000.0 0000.0 5500.0 0000.0 0000.0 0000.0 0000.0 0050.0 moo~.o H“ mhmomxm 0500.0 0000.0 0050.0 5000.0 0050.0 0000.0 0000.0 0000.0 0000.0 0000.0 0500.0 0000.0 0000.0 0000.0 0000.0 0000.0 0500.0 0 0900020 maouam cumummm .m.m.0.0 mm0Eocoum 0mGGM0m 000muucmu mmuox nusom M0m< mu0uu< mu0uma< a0um0 mm0Eocoum 0w000m>m0 mmm0 m500c0800 cumnusom 0 mvmcmo ammuw 0m03uuom ms0a muonsm umnuo 0 00900 0 000 .m.0 mm0eocoom 0mmo0m>w0 00002 $903 MQ<0H .M.D 0000 mm0u0voEEou mmusuumwssmz mo muuanH 05m mupoaxm .M.0 00 00umm mwoou “manmcou 00 mucm0u0wmmoo mzH 510 m0an 140 In the South Korean case, Table 5-8, the pat- terns of trade with all economies, except with the Burma, India, Pakistan, and Indonesia group, are consistent with the stage of production theory of comparative advantage. 5.4 Technological Gap The commodity composition of trade in manufac- tured goods is determined, according to the technological gap theory, by the relative technological advantage of a country vis3a-vis other economies. The theory asserts that the more technologically advanced country will enjoy a comparative advantage in the production and export of newer products, while the less advanced country will have comparative advantage in the export of older products. To the extent that technological advances are the result of a systematic and costly process of research and development, the level of economic development, as measured by Gross Domestic Product per capita, may be con- sidered as the index of the relative technological sephis- tication of economies. The relevant commodity characteristic is its age, and it is measured by the date of its first appear- ance in the export market [4]. Hufbauer [3] estimated the date of the first appearance of commodities in the export schedule of the U.S. These estimates, as reported in Table A-ll, Appendix, are applied to the trade struc- ture of the U.S., the U.K. and South Korea. 141 .mlm $0”an 0mm "mmUHDOm qum.o qum.o ncmm.o waox mac: 0am .muoamwsfim .mammmamz NNH¢.~ m00m.o m¢mo.o wamafimne eam sesame ~m-.o nmmm.o mmmm.o mfimmaowaH 0am .amumfixmm .maecH .mausm «umo.o wwm0.o oumm.o “mam chqu mmmo.o mmwm.o moao.o muaum< mm00.o “man.o mmmm.o mafiuma< sauna Hoom.o ommq.o wm00.o mafiaoaoom umaoam>ma mama unmo.o Noam.o memo.o wcmfimmw 3mz use «00muum50 momm.o ommo.o 5mm~.o gamma aoqa.o Homo.o 0000.0 .M.: cmm0.o 00-.o m0m0.o «gonna :uoummz 005m.o mom0.o ~mm~.o .m.: mcoq.o “mam.o 050N.o mmflaocoum coa00o>ma wm00.o NmHm.o mmNN.o aqmoz m; a . a > memomxm memomzH mHHz mmma wmmq mvmcmo vcmammw 3oz a wfiHmuums< comma mmousm umsuo .M.D 49mm 0mm mmfiaocoom vmaoaw>oo H nqmoz mHHB mQmn mmmq WK . ON NN o 0N @COHCfi—HOQ aHUr—USOW 00 . 2 0.0.: e 328 oe.m~ mm.a~ :mamh -.HN Nw.¢~ Hewsuuom mafia oaousm nonuo 3.2 00.2.. 0 «4:0 0N.~N cm.- H can mm.q~ mw.o~ .m.= 0m.- o~.- moaaocoom wmaoam>mn 0N.NN Nc.m~ aqmoz HH 0 mamomxm mamomzH meHz mQmo mmmg mm.HN mn.mH wcmammu 3oz pom mfiamuum=< mm.m~ Na.o~ Gamma oq.mm co.NN .M.D 0m.- om.a~ mmounm cueummz mm.m~ mo.mm .w.D Hm.mm mm.H~ manoeoom vmmoaw>ma NN.mm oo.am oqmoz 00 0 memomxm memomzH mHHz mamn mmmq mcmcmu vcmammm 3oz w mflamuums¢ comma mmouam umsuo H a; «emu 0mm wmwaocoom umaoam>mn QAMOS mHHB madme .m.D ohmH mmfiuawoaaou kuzuomwscmz mo muuoaaH cam muuonxm .m.: we muamfiuwmmmou coaumwucmumwmwn uesvoum one 149 .qlm manme mom "wwuocuoom cam mmousom 0000.0 0000.0 0000.0 maouam 0000000 0000.0 0000.0 0000.0 .0.0.0.0 0000.0 0000.0 0000.0 000000000 0000000 000000000 0000.0 0000.0 0000.0 00000 00000 0000.0 0000.0 0000.0 0000 0000.0 0000.0 0000.0 000000 0000.0 0000.0 0000.0 0000000 00000 0000.0 0000.0 0000.0 000000000 000000>00 0000 0000.0 0000.0 0000.0 0000000000 00000000 0000.0 0000.0 0000.0 000000 0000.0 0000.0 0000.0 00000 0000.0 0000.0 0000.0 00000000 0000 0000000 00000 0000.0 0000.0 0000.0 0 0000 0000.0 0000.0 0000.0 0 000 0000.0 0000.0 0000.0 .0.0 0000.0 0000.0 0000.0 000000000 000000000 0000.0 0000.0 0000.0 0000: 0000 00 0 00» 0000000 0000000 0003 00000 .0.0 0000 00000008500 00050000300: NO 0000080 cam muuoaxm .m.: we 000000000000 cowum0ucmummwwa uuswoum mnH main maan 150 .mlm mHQMH mom “mmuusom 0000.0 0000.0 0000.0 0000 0000 0 000000000 .00000002 0000.0 0000.0 0000.0 00000000 0 003000 0000.0 0000.0 0000.0 000000000 0 00000000 .00000 .00000 0000.0 0000.0 0000.0 0000 00000: 0000.0 0000.0 0000.0 000000 0000.0 0000.0 0000.0 0000000 00000 0000.0 0000.0 0000.0 000000000 000000>00 0000 0000.0 0000.0 0000.0 0000000 300 0 000000000 0000.0 0000.0 0000.0 00000 0000.0 0000.0 0000.0 .0.3 ammo.N noom.o mmmH.H waouam 8000003 0000.0 0000.0 0000.0 .0.0 0000.0 0000.0 0000.0 000000000 000000>00 0000.0 0000.0 0000.0 00003 0000 00 0 00» 0000000 0000020 0003 00000 200000 00000 0000 00000008800 vmuauomwacmz mo manomaH was muuomxm cmwuox nusom mo mucmwofimwmoo ao0umHuamumwmwn uosuoum use calm mHQmH 151 The scale economies theory is concerned with the technological advantage derived from the domestic market size. However, two alternative approaches to testing the theory resulted in equally unsatisfactory results. The outcome of the test is troublesome in the cases of the U.S. and U.K. trade with other developed economies. Simi- larly, several contradictions are encountered between the actual commodity composition of trade and the predicted pattern in the trade of South Korea with other less devel— oped countries. The other three theories--stage of production, technological gap, and product cycle-~consider the rela- tive level of technological development of economies as the determinant of the commodity composition of world trade. The variation in these theories is the commodity characteristic which can differentiate a technologically advanced and sophisticated commodity from a less advanced and less sophisticated one. Variation in the performances of these theories is, therefore, the outcome of the discriminating quality of the index of the commodity characteristic. Furthermore, since the commodity coefficients of a single country are applied to all three countries in these tests the other determining element in the performance of each theory is the universality of the index of the commodity character- istic among countries. 152 An overview of the results of these three models in Table 5-15 indicate that all three theories are equally suited for CXplaining the trade of manufactured goods 3- mong developed countries. However, while the stage of productions theory fails to explain the flow of commod- ities between deve10ped and less developed countries, the other two theories reveal no serious deficiency in this respect. Nevertheless none of the three theories could CXplain the commodity composition of trade among the LDC's. In conclusion, only the technological gap theory and the product cycle theory are capable of providing a framework for explaining the bilateral commodity compo- sition of trade of manufactures. Even these theories fail to explain the pattern of trade among less developed econo- mies. The poor performance of the neotechnology theories to CXplaln the commodity composition of the trade of less deve10ped countries may be attributed, at least partly, to the effect of foreign investment in these countries. Al- though a country's level of technological achievement is closely related to its level of economic development, the availability of new technology through the flow of direct foreign investment may disturb this relationship. Thus a country at a lower level of economic development may have a comparative advantage in producing technically SOphisticated products as the result of the availability 153 of technology acquired through foreign investment. While this effect may not be substantial enough to disturb the pattern of trade between developed and less developed countries, it may yet be significant among less devel- Oped countries which have small differentials in their level of economic development as reflected in their GDP per capita. 154 Table 5—15 Summary of Performance of Neotechnology Theories of Trade: Trade Flows Inconsistent with the Predicted Pattern Scale Economies Theory A--Hufbauer's Method Trade of U.S. U.K. South Korea With Canada EEC Latin America EFTA Africa Japan Middle East Scale Economies Theory B--Savigg's Method Trade of U.S. U.K. South Korea With Canada EEC Latin America Australia Japan Africa New Zealand Stage of Production Theory Trade of U.S. U.K. South Korea With Canada Latin America Burma, India, Pakistan, & Indonesia U.S.S.R. Africa U.S.S.R. Technological Gap Theory Trade of U.S. U.K. South Korea With Japan Africa Middle East Product Cycle Theory Trade of U.S. U.K. South Korea With Canada Australia, New Zealand Middle East, Burma, India, Pakistan, Indonesia REFERENCES [l] Hirsch, S. ”The United States Electronic Industry in International Trade." National Institute Economic Review. November, 1965, pp. 92-97. [2] Hufbauer, G.C. Synthetic Materials and the Theory of International Trade. Combridge, Mass.: Harvard University'Press, 1966. [3] . "The Impact of National Characteristics and Technology on the Commodity Composition of Trade in Manufactured Goods." In R. Vernon (ed.), The Technology Factor in International Trade. New York: NBER, Columbia UniVersity Press, 1970. [4] Posner, M.V. "International Trade and Technical Change.” Oxford Economic Papers. October, 1961, pp. SZS-RT} [5] Saving, T.R. "Estimation of Optimum Size of Plant by Survivor Technique." Quarterly Journal of Economics. November, 1961, pp. 569—607. [6] Stigler, G.J. ”The Economies of Scale," Journal 9f Law and Economics. October, 1958. [7] United Nations. Commodity Trade Statistics. Statistical Papers, Series D. New York, 1969, and 1970. [8] . Year Book of National Accounts Statistics. New York, 1966. [9] United States Bureau of Census. U.S. Census of Manufactures. Washington, D.C., 1963. [10] Vernon, R. "International Investment and International Trade in Product Cycle." Quarterly Journal of Economics. May, 1966, pp. 190-707. 155 156 Ill] Walters, A.A. "Production and Cost Functions: An Econometric Survey." hconometrica. January- April, 1963, pp. 1-66. Chapter 6 THEORIES OF INTERNATIONAL TRADE: A STUDY OF TRADE PATTERNS OF TWENTY-THREE COUNTRIES 6.1 Introduction The H—O theory and alternative theories of trade have been empirically tested for the bilateral trade re- lations of the U.S., the U.K., and South Korea. The gen- erality of these theories will now be tested further for the existence of a systematic relationship between the national characteristics and the commodity composition of trade in manufactured goods only, as hypothesized by each theory, for a significantly large number of countries. 6.2 Selection of Countries In selecting countries for the following tests, the aim has been to include the major exporters of manu- ikustured goods. On the other hand, to test for applicabil- it), of the theories in explaining the commodity composition oiT‘trade of countries with varying degrees of economic de— velcnnnent, inclusion of a number of less deve10ped countries iii the sample was considered necessary. But since the fol- .lowiru; tests are confined to the flow of manufactured com- nuxlities, many less developed economies had to be excluded. 157 158 Only those countries with a significant size of manufactured export relative to their total exports were selected. To provide a basis for comparison with a similar study by Hufbauer [2], his criteria for selecting countries was adOpted, and the twenty—three countries presented in Table 6-1 are considered for this study. The only depar- ture from Hufbauer's list of countries is the exclusion of Israel. The significant inflow of capital and highly skilled workers into Israel during the last two decades, along with the very rapid rate of economic transformation during this period, required special attention for the study of its trade structure outside the scope of this study,(for example see [1D. 6.3 Methodology The relative factor intensity of trade of each country is estimated by applying the apprOpriate measure of the commodity characteristic, as deve10ped and ex- plained in the previous chapters, to the country's repre- sentative bundle of exports and imports of manufactured goods. However, application of coefficients of commodity characteristics estimated for one country to the trade of all countries implies the assumption of non-reversal of the proper characteristic among countries. The ideal method would have been to estimate the coefficients based on each country's experience. However, given the limita- tion of this study, only minor refinements may be under- taken in this context to improve the testing procedure. 159 .mim paw Him moanme "mmousom om qw.o mHo.o com mflwcH ow wn.o «Ho.o com cmumfixmm omH mm.o Hmo.o omaa cmsame oqa Hm.o mmo.o omw mwuox nusom oom mm.o eqo.o coma mcoM wcoz omm ma.a omo.o comm mw>mamowsw ems om.a “No.0 coma Hmmsuuom oma mm.q omo.o ooom oofixmz omm ss.s Hso.o ceaa camam HHH asouu own om.HN mqo.o ooam cmamh omoa os.mH eso.o ooom samaH omoa om.m woo.o oooq mfiuuw=< omaa mm.m «mo.o omma mpcmauwzuoz ooqa Ha.q owo.o coca adawamm owma mm.m~ mwo.o come museum owoa mm.m w~o.o ommm xumacoo oaha NN.~m mao.o oooq aoemcax sagas: caaa Ho.os ooa.o omma semapmu oawa mo.m noa.o comm mwamuums¢ omma Hm.a omo.o ooao smsuoz OOHN No.m mma.o oocn :mvam HH asouu oHHN mm.oa ooa.o omwm mvmamu ooom qo.mna moa.o ommn mmumum kuH63 H asouo Am .m.= Am .m.: CHV mo macadaflm mouom xuoz educaaem mafia Am.m.p :Hv ch usauso Hmuoe mo N m mm Isuommscmz pom muwmmo you mQU wcfiuauomwacmx mmmmoaaam voaafixm Hmufiamo wmxfim muucsoo mowuwwumuumumxu HmaOflumz Hue magma 160 In the cases of the H-0 theory and the human capital approach, the above estimation problem may be re- duced by dividing the countries in the sample into three groups based on similarities in their level of economic deveIOpment. Taking Gross Domestic Product per capita as an index of the level of economic development, coun- tries are classified, arbitrarily, into the three follow- ing groups: The U.S. and Canada constitute Group I; the major industrialized economies of EurOpe plus Japan form Group II; and the other economies in the sample (charac- terized by a Gross Domestic Product per capita of $500 or less) are placed into a third group (Table 6-1). For each group of countries, a representative country was selected. For Group I the U.S., for Group II the U.K., and for Group III South Korea. The total capi- tal, labor, and skill requirements of the industries of each representative country, estimated from its industry data and input-output table, are applied to a representa- tive one million dollars of imports and exports of each country in the group. Assuming that there are less vari- ations in each group than among the countries in the whole sample, this method is one step toward using each one of the twenty-three countries' own national statistics. For neotechnology theories this refinement in es- timating the factor intensity of trade of each country was not possible, since only one set of production coefficients are available. Therefore, for these tests the same set of 161 * coefficients are applied to the exports and imports of each country. Although the assumption of similarity of technical conditions of production among countries may not appear realistic, it may be considered satisfactory for manufactured goods, especially for the coefficients of scale economies and product age. The scale economies theory assumes similarity of capital and labor input requirements in the produc- tion process of goods among countries. Therefore, it is plausible to assume that the coefficients of scale economies, which are essentially a technological char- acteristic of production and are estimated by assuming that the capital/labor ratio and skill compositions are constant for different industries, remain invariant among countries, given the widespread diffusion of technology and the standardized methods of production of manufactured goods. The product age coefficients may not be the same in all countries. In fact there is little reason to believe that they are. However, what is relevant in the technological gap theory is the age of the product on the international market from the date of its first appearance. On this basis the age of each product * U.S. coefficients for scale economies, techno- logical gap, and product cycle theories, and Japanese coefficients for the stage of production theory were applied. 162 is best estimated by the date of the product's first appearance in the U.S. export schedule. This will be true on the assumption that the U.S. has been the innovator in production of all manufactured commodities, an assumption which is not all too unrealistic. The assumption of similarity of the estimates of consumer goods ratios and the product differentiation index is, however, less realistic. But in the absence of alternative estimates, the same set of coefficients will be used to find the commodity composition of each country's trade. The test for the existence of a systematic relationship between the commodity composition of trade and the national characteristics for each theory will be carried out by the method of least square regression analysis. This will be discussed in detail in the following section. 6.4 Results and Conclusion The characteristics of the commodity composition 3% of trade in manufactured products (1969) of twenty-three countries are shown in Table 6-2. Each set of X, coeffi— cients reflects the relative factor intensity of imports to that of exports according to the corresponding theory. ““** x’ Except U.S. (1970), Mexico (1968), Pakistan (1968), 1ndia(l966). 163 .xaeamaaa .oan< ecm .ean< .mHua mmanme "mmuuaom quo.a waa.o Nmm~.o wmm.¢ maem.a mmo.m wmm.m mqm.m oam.a maccH Naom.a mmm.o Nmnm.o omn.o q-~.a mqo.~ noo.n «Ho.n OHN.H cmumflxmm NHO¢.H «mm.o mmoq.o wmm.~ ewHH.H moo.H wwu.~ emm.m wm~.H cmSHmH ommm.a omm.o wmq¢.o mqu.m mqu.H o~n.a Nmm.m oqa.m NH¢.H «whom nusom nooH.H mwm.o qaao.o Hem.q mOHN.H mH~.H muo.H awo.a wnH.H wcox wcom oqno.a oam.o ammo.o mwm.H mwNH.H om~.H omq.H omq.H «mo.a mH>mamownw mmm~.H qmw.o HHNq.o qu.q mamo.a mqo.a mom.m qam.m NHN.H HmwSuuom qqu.a qom.o mmaq.o moo.m omo~.a Noq.a omo.m omo.~ mmm.o oonoz mmo~.H mmw.o ~m~m.o mmn.a quo.a qu.H omc.a omq.a N¢H.H Gammm HH msouu mqom.o mmH.H mmnm.o oqw.o cumm.o moH.H ma~.H NNN.H wmm.H amamn Nmoo.H qoo.H mmao.a on.o NNHO.H on.H Hmw.a nmN.H Hmo.a hamuH momm.o oNo.H mwoo.H mom.a mooo.o Hq~.H moH.H moa.H OHH.H mwuums< Hmao.a ocm.o mmqm.o HHo.H wmmo.H oqm.o Nmm.o mum.o man.o mccmaumnumz owwo.a Mmm.o quH.H Hum.o moe¢.o Ono.H ONH.H nma.H omm.o adfiwamm o¢HO.H Nam.o wmnw.o mmm.o momm.o mmo.H mmo.H cmo.a wom.o museum mmmm.o omm.o Nwa.o qmo.a momm.o mom.o mmm.o nmm.o om~.H xumsama qum.o Hmo.H ammo.a New.o mwNm.o Hoo.o omm.o nmm.o mmo.a aovwcfim wouwcs o~cm.o omo.a Hmnm.a mmn.o mwoa.a cam.o 005.0 Hon.o CNH.H mamaumo «Hmm.a new.o momw.o H¢N.H wmc~.a Hmo.H HOH.H mmo.a oam.o mfiHmuum=< Humm.o omm.o Hmwm.a mHH.N mnmo.H me.o mom.o oam.o qmm.0 mashoz mumm.o amo.a mmom.a neo.a oncm.o mmm.o qom.o nom.o mm~.H amvmzm HH macho «HQH.H mmm.0 Hoqq.a Nom.o Nmmm.o mmo.H qu.H mqo.a mwm.o mvmcmo mnow.o NoH.H cmmq.a nmn.o Howwuo mmm.o mun.o Hum.o omm.o mmumum vmufica H msouu var mwmr me» Now» How» m» N» H» A\M> m m m mvmue mo mmauownw o>wumcumua< man can cHH£OIMm£omxomm any on wcfivuoou< mmfiuucsou moununhuamze we mvooo vmusuemmsamz cw evmuw mo cowuwmoaaoo huwvoaaoo Niw oHQmH 164 'To tesst Idle (zxrriaruit()ry' ptnver“ oi‘ crush ‘tb(n)r)r, the designated X coefficients of the twenty-three countries are regressed on their national character— istics (Table 6-1), as postulated by the theory. The independent variable is the national characteristic and the dependent variable is the coefficient of commodity composition of trade. Then the regression equation for each test will be yj = a + b Xi + e where Vi is the coefficient of commodity composition of trade and Xi is the national characteristic for country i. The stochastic error term is e, and a and b are the parameters. The least square regression method is applied to each set of variables indicated in Table 6-3 for each theory. The following results are obtained: (1) Heckscher-Ohlin: XK/I = 1.2520 - 0.000045 (E/f) R2 = 0.3342 J (0.000014) (2) Human Capital: 5: = 4.2144 - 35.7273 (SK) R2 = 0.5592 (6.9201) s i 2 _ . K = 4.2140 - 35.7162 (5K) R — 0.5004 2 (0.9020) 2 K: = 1.8550 - 9.3854 (SK) R = 0.7042 ‘ (1.1375) (3) Scale Economies: Hufbauer's Scale Coefficients , K = 1.1235 — 0.00171 (MF) R“ = 0.1092 5C1 (0.00107)* 165 A3 3 535m mmw mugged you man up popsmmma .%uuaaoo mnu mo coaumofiumasaom Hmofiwoaoanomu mo Ho>mH one ppm muammo you mac mp no m: usauso mGwHSuumwsama mo muam mnu %n cmusmmoa unwaaoam>mp afiaoaooo mo Ho>oa may mz usmuso wcfiusuommscma mo oufim enu >2 panammma .uoxuma oHumeov mo muwm Mm huuadoo one we ucmESOpam Hafixm m>HumHom A\w. huucnoo ecu N. we ucoEBOEcw gonna paw Hmufimmo m>HumHmm mowumfiuouomumnu HmCOfiumz ea» muuoaxm mo umnu ou muuoaaa mo aoau ImHuqmumwwfiv uosvoua mo owumu 6:9 «mm > muuoawm mo umnu ou muuomafi mo mmm.mwmum>m any we» muuoaxm mo umnu ou muuoaaw aw powwonam mvoow “panacea mo oaumu way Now» .Hom> muuoaxm cu m>wumHmu muuoaaw :« weaponEm moHEocoom mamum m . N . H m r m > m r mcmuu mo zuwmamucw Haaxm m>wumamm A\M> .A\M> mpmuu mo >ufimamuafi uonma paw Hmufinmu m>flumaom mvmue mo coauHmOQEoo muwpoefioo mo musmmwz mzH macho u05poum new HmonoaocnomH bowuuspoum we mmwum mafiaocoum mamum somouma< Hawxm amasm caanouumEUmxomm NaomLH mpmua mo mmfluomzH m>flumcuoua< msu paw muomzH 01m mnu ou wafivuooo< mowumwumuomumso HmGOHumz waflcfiEHmuma one paw mwmufi mo cowumeQEoo xufiwofiaoo mo mounmmmz 63H mum magma 166 Saving's Scale Coefficients Xscz = 2.4755 - 0.01639 (MP) R2 = 0.1041 (0.01049)* ~ (4) Stage of Production: , Xcg = 0.8146 + 0.00444 (MF) R“ = 0.1504 (0.00230)* reg = 0.4061 + 0.000432 (GDP) R2 = 0.7611 (0.000053) (5) Technological Cap: Kage = 0.9072 + 0.000058 (009) R2 = 0.2856 (0.000020) (6) Product Cycle: ’7 Ypd = 1.3642 - 0.000203 (009) R = 0.5018 - (0.000044) where R2 is the coefficient of determination of the regres- sion equation, and the numbers in parentheses are the standard errors of estimated coefficients. Those designated by an asterisk (*) are not significant at the 99 percent probability level. Among the regression equations the scale econo- mics and the stage of production (with the size of manufacturing output as the explanatory variable) do not perform satisfactorily. The level of significance is below the 99 percent probability level, and the coeffi- cients of determination are very low compared to others. The poor performance of the scale economies hypothesis as tested by both methods of measuring econo- mies of scale may be attributed to the unrealistic assumption inherent in the theory that small nations cannot rely on international markets in producing 167 commodities with economies of scale. Considering that production and trade often take place simultaneously it appears more realistic to assume that deveIOpment of large scale production depends more on the technical capability of the economy than on its absolute size. To the extent that large scale production often involves more advanced methods of production and management and a higher degree of industrial integration, it is more persuasive to consider comparative advantage in production of scale intensive commodities as a function of the level of economic deveIOpment. To test this alternative hypothesis to the scale economies theory of trade, the following two regressions were tested: (7) xscl = 1.2308 - 0.000099(MF) - 0.000120(GBR) R2=0.2976 (0.001197) (0.000052) 3662 = 4.0351 - 0.00712(MF) - 0.00174(Efifi) R2=0,5198 (0.00969) (0.00042) where the dependent variables, Kscl and. ‘XSCZ’ are the relative scale intensity of trade of the twenty-three countries under this study, as measured by Hufbauer's and Saving's methods. Manufacturing size (MF) and Gross Domestic Product per capita (GDP) of these countries are considered as the independent variables. The results indicate that while manufacturing size is not statistically significant in either regressions, CUP is significant at 95 and 99 percent probability levels 168 in the two equations, respectively. Although both rela- tions appear satisfactory with GDP as the explanatory variable, the scale economies estimated by Saving's method indicates a stronger performance. Therefore, it may be pointed out that although the scale economies theory, as hypothesized by Hufbauer, does not explain the commodity composition of trade, it can provide the basis for an alternative explanation of the commodity composition of trade, as suggested above. In all other cases, however, the regression equations indicate a systematic relationship between the dependent and the independent variables as each theory hypothesizes. In each case the relationship is highly significant and the high coefficients of determination indicate the ability of the national characteristic to explain the commodity composition of trade. These results may seem to point to a difficulty of interpre- tation of the explanatory power of each theory. The results, however,may be explained by con- sidering the interrelation between the proposed alterna- tive theories. As an attempt in this direction we may separate the six alternative hypotheses into two groups of factor proportion theories (Hypotheses l and 2) and neotechnology theories (Hypotheses 3 through 6). The first hypothesis in the factor proportion group selects relative intensity of physical capital in production and its endowment in the economy as the 169 relevant measure. The second hypothesis, while assuming that physical capital is highly mobile, considers human capital as the predominant force in determining compara~ tive advantage. However, examination of Tables 6-4 and 6—5 reveals that the capital and skill endowment of countries on the one hand, and capital and skill inten- sities of traded commodities on the other hand, are highly correlated. Therefore, although the two theories look at comparative advantage from different angles, signifi- cant similarities between their selected variables give rise to almost equally powerful theories. However, if the size of R2, that is the closeness of the sample regression line to the sample observation points, is chosen as a criterion for distinguishing the "better” theory from the other, the human capital approach has an edge over the H-0 hypothesis. Similar situations exist for the three remain- ing hypotheses in the second group. The stage of pro- duction, technological gap, and product cycle theories all perform satisfactorily in the test. However, this is not surprising. All three theories consider the level of technological development of the economy as the relevant national characteristic, measured by GDP per capita. On the commodity side the comparative advantage is reflected, according to these theories, in the level of technical SOphistication of the products. 170 .N10 mHQmH "mousom ooo.a ea» cOHuMfiucmumwwfiQ uospoum oa~.o- ooo.a 6mm» mw< uosvoum 005.0- Nae.o ooo.H was Ofluwm mwoou Headmcou mm~.o HN¢.QI mom.01 ooo.a New» “65.0 ome.o- mam.ou nom.o ooo.a Hum» mmHBOGOUm mHmUm mww.o Nam.o- mmm.ou oqm.o mam.o ooo.a m» mmm.o Nom.on mma.o- nom.o sam.o ~sm.o ooo.a m» mmm.o mom.ou Hma.ou mom.o mom.o Hem.o aam.o M» moaamm Haaxm amm.o oma.o wom.on omm.o «mo.o “Nm.o “as.o maq.o ooo.a A\M> oaumm nonma\amaaamo var 6mm» we» New» How> m» N» A\M» uom mwmua mo coauamonaoo huflvoaaou mo mucmHUfimwmoo mnu amo3umm mGOflumamuuoo maaafim w m mmfluucsoo mounHlmuamze «no manna 171 . Hl© ”Ha—QB. ”MONA—0m coo.H mwm.o maa.o mam.o mac ooo.H mum.o cmq.o ooo.H mmw.o Mm mmwuuaaoo omusklmucwae mo moflumaumuumumnu HanoHumz comaumm mGOfiumHouuoo manfiwm mic manna mam muwmmo umm mac mcwuauuumaamz mo anew caumm ucwaaovcm Hawxm oaumm Honmg\amufiamo 172 The difference between the three theories is the various ways of measuring the concept of product sophistication. One would expect, a priori, that the three measures-- namely the ratio of consumer goods, the age of product, and the index of product differentiation would be very closely related. Indeed Table 6-4 shows a high degree of correlation between the measures. Even if the coeffi- cient R2 is selected as the criterion for discrimination, the stage of production and the product cycle theories perform equally well. However, the more interesting aspect of the results is that both groups of theories can explain the commodity composition of trade despite significant differences in their approach. Once again this outcome can be explained in light of the interrelationship between the selected variables. The level of technological deveIOpment of a country is closely related to its level of economic develOpment, to the extent that technological achieve- ments can be obtained through costly processes of research and development. On the other hand, physical and human capital are both forms of society's savings. The higher the level of economic deveIOpment (measured by GDP per capita) the higher is the ability of the country to channel its productive resources into physi- cal and human stock of "waiting"--that is, investment. As Table 6-5 shows there are very high correlations 173 between GDP per capita and capital/labor endowment ratio (0.9153) and relative skill endowment ratio (0.9194) of countries in the sample. Similarly, among the commodity characteristics, the coefficients of skill intensity are highly corre- lated with the coefficients of consumer goods ratio and with the product differentiation index. This implies that the producer goods and the less standardized products are those that require highly skilled workers in their production process, and a country having comparative advantage in its trade according to one theory will also have comparative advantage according to the other two theories. To conclude, it may be pointed out that all the alternative theories of trade, except the scale economies, can explain the commodity composition of trade of manu- factured goods on a global scale. The fact that they are all powerful in explaining trade should cause no difficulties since the variables selected by each theory are a close approximation of those selected by the other theories. REFERENCES [1] Hirsch, S. "Technological Factors in the Composition and Direction of Israel's Industrial Exports? In R. Vernon (ed.), The Technolpgy Factor in International Trade. New York: NBER, Columbia University Press, 1970. [2] Hufbauer, G.C. "The Impact of National Characteristics and Technology on the Commodity Composition of Trade in Manufactured Goods."In R. Vernon (ed.), The Technology Factor in International Trade. New York: NBER, Columbia University Press, 1970. 174 Chapter 7 SUMMARY AND CONCLUSIONS The theory of international trade attempts to determine the relationship between the national economic structure and the commodity characteristics of the country's trade. According to the orthodox version of the Heckscher—Ohlin theory this relationship is between the relative endowment of capital and labor of the economy and the relative intensity of these factors in the production of different commodities. Thus the hypothesis of the theory is that a country will have a comparative advantage in the production of the commodity which uses the country's more abundant factor more intensively. The Leontief studies of the structure of U.S. trade for the years 1947 and 1951 presented results contrary to the prediction of the H—O hypothesis. These and the subsequent empirical studies of the theory casted serious doubt over the validity of the H‘O model. Dissatisfaction with the performance of the H—0 theory has led to the development of several alternative theories of international trade. 175 176 Within the logical framework of the H-0 theory, the human capital approach seeks to explain the commodity composition of trade through differential skill require- ment of industries and the relative abundance of a coun- try's skilled to unskilled workers. However, a new breed of theories, the so-called neotechnology theories, have set aside the relative factor abundance and factor intensity concepts. These theories maintain instead that the commodity composition of trade in manufactured goods is determined primarily by the relative technological advantage of a country in producing various commodities. For the scale economies theory this technological advantage is derived from the economies of large-scale production which may be enjoyed by producers who are located in economies with a large domestic market. On the other hand, for the other neotechnology theories--stage of production, technological gap and product cycle--the predominant national characteristic is the level of the country's technOIOgical achievement. That is, the more technologically advanced economies will have a comparative advantage in the production of technologically more sophisticated products. The difference among these three theories lies in variations in defining the technological sophistication of commodities. This study has tested the H-0 theory and alterna- tive theories of international trade. Through a 177 reconsideration of the H-0 theory it was shown that the failure of the theory to explain the commodity composition of trade was caused by the theory's restric- tive assumptions in the face of the generality of its sc0pe. It was pointed out that although the theory may not hold at the level of generality presented by Ohlin, it is still capable of explaining the commodity composition of trade under certain restricted conditions. It was thus hypothesized that trade structure will tend to conform to the H-O pattern between countries that are not very disproportionate in their level of economic deveIOpment. Furthermore, it was suggested that the theory would perform even more satisfactorily if it was restricted to the flow of manufactured goods. The basis for this hypothesis was the pre- sumption that the critical assumptions of identical demand patterns, similar production functions, and unique factor intensity ranking of commodities would hold more strongly under the above stated conditions. Applying the most recent input-output tables and industry input requirements of the U.S., the U.K., and South Korea, the test of the H-O theory revealed that the less generalized version of the theory provides a powerful explanation for the bilateral trade among either developed or less developed countries. The results were even more pronounced when trade only in manufactured goods was considerd. However, the 178 theory was not verified in the case of trade between developed and less developed economies. The results were contrary to Linder's hypothesis that the commodity composition of trade in manufactured goods tends to be similar among countries with similar levels of economic development. The Keesing method was applied to test the human capital approach to the theory of international trade. The total requirement of workers of various skill categories was estimated for a representative one million dollar bundle of imports and exports of the U.S., the U.K., and South Korea based on their national statistics. The result indicated the ability of the theory to explain trade of developed countries among themselves and with less developed countries. However, in several cases the theory was inconsistent with the existing pattern of trade among less developed countries. In an attempt to test neotechnology theories, Hufbauer's estimates of commodity characteristics, and Saving's estimates of optimum plant size were applied to the bilateral trade structure of the same three countries. Although capable of explaining the composition of trade among developed countries, all four theories failed seriously in their attempt to explain trade among less developed countries. 179 Generalizing the result of the bilateral study of these three countries, it may be concluded that various existing theories are useful in explaining different segments of the international flow of commodities. The orthodox version of the H-O theory, with physical capital and homogeneous labor as explanatory variables, is well suited for the trade within the two groups of developed and less deve10ped economies. In fact it is the only one. Although the human capital approach and neotechnology theories fail to explain trade among less developed countries, they provide a satisfactory explanation of trade of manufactured goods among the developed countries (a summary of performance of the theories is presented in Table 7-1). While the above tests were concerned with the consistency of the pattern of trade flows between economies according to various theories, it is also relevant to verify the existence of a systematic relationship between national characteristics and the commodity composition of trade for a large number of countries. This aspect of the study improves upon Hufbauer's similar work by introducing two methodological changes. The commodity composition of trade for the H-0 theory and the human capital approach were estimated for twenty-three countries from the total factor requirements of the representative country of each three 180 ++ ++++ «Homo uospoum ++ ++ ++++ ++++ Gowu new Hmoww loseoum Ioaocsome mo ammum ++ ++ ++ ++++ ++++ ++++ + + + macaw: eonamz mcw>mm umzmnwsm AM» v 1 1 mowuma¢ :Huma I I mowuw< + I mmfisocoom page 1>ma mmmq + + mpmcmo I 1 vcmammw 3mz w awamuums< I cmamh mmousm umsuo .M.D ¢Hhm 0mm mafiaocoom cmaoam>mn ++ + ++++ +++ + I QAMOB umHHZ mQma o o o o o o o o QAMOB mHHz mn~m umsmnmdm Aw» v wvmue w mmfiufivoaaoo eHo%o emu Hmuww 1osvoum m < Hmuwamo u:QCH vmusuomw w mudaaH uoswoum 10HoanoeH we mmMum mmflaoaoom mamom cmadm 13amx 01m Ham cum mucosa Ap.ucoov H1m manme 182 + + + + + + + + .M.: + + + + + + + + vacuum cueummz + + + + + + + 1 .m.: + + + + + + + + mmaaocoom vmaoam>mo o o o o o o o 0 quoz mHHz mafimH owmue cmwuox Lusom1lu + + + + + + + + maousm cuoummm + + 1 o o + 1 + .m.m.m.b + + + + + + + + mmaaocoom pwccmam zaamuucmu + 1 + + + + + + mmuom nuaom + + + + + + 1 1 mam< + + 1 + + + 1 1 mowuw¢ + + 1 + + + 1 1 moauma< sauna + + + + + + 1 1 mmaaoaoom vmao 1>ma mmoq vosumz cosuwz defiu wafi>mm umsmnwsm AM» v mvmua w moHquoaaoo eacho Qmo Hmoaw Iospoum m < Hmufimmo usmcH mCHHSuomw a muaacH uoavoum Ioaoanooa mo ammum wmaaoaoum mamom amadm 13cmz 01m Ham 01m zuomse Ae.acouc a-“ magma 183 + + + + + + + + meow mac: w muommwch mfimxmamz + + + + + + + + UCQHHmfiH a chHmH 1 + I + + 1. + + mHmm 1cocaH a cmumfixmm .meGH .mausm 1 1 + 1+ I 1 + + ummm mavcfiz + 1 + 1 1 1 + + mofluw< + + + 1 1 + + 1 moauma< sauna o o o o o o o o mmflaoaoom vmmoam>mo mmoA 1 + + + + + + 1 ccmHmmN 302 a mwamuum=< + + + + + + + + amamw wonumz vocumz defiu wafi>mm umsmnmsm AH» v mvmuH w mequanou mauxu mmo Hmuuw Iosvoum m < HmWfimmo usmcH mcwHOuomm w muamcH nonvoum 1oaocnumH mo mmmum moHBocoom eamom cmadm 13cm: 01m Ham 01m xuomza he.aeouv a1a magma 184 groups of countries. The estimates of the relative factor intensity of trade obtained by this method are more apprOpriate for a study of this nature than are those found by applying the direct input requirements of industries based on one country's experience to the trade of all countries-~the method used by Hufbauer. Moreover, the relationship sought here is between the commodity composition of trade (the coefficient of commodity characteristic of imports relative to experts), instead of export characteristics alone (as done by Hufbauer) and the national characteristic for each theory. The results of the regression analysis indicated that, with the exception of the scale economies theory all theories perform satisfactorily. However, among them, the human capital approach, the stage of produc- tion theory, and the product cycle theory make the strongest showing. Bilateral tests of the trade of the three countries and the regression tests of the world trade of twenty-three countries provide measures of the explanatory power of the existing theories from differ- ent view points. In the test of the world trade of twenty-three countries, the sample included countries with a sig- nificant share of manufacturing exports in their total eXport basket. Consequently only a small number of less developed countries pass this requirement. Even 185 among the countries included in the third groups, over half are the so-called semi-developed countries: Spain, Portugal, Yugoslavia, and Hong Kong. However, trade in manufactured goods among these twenty-three countries constitutes the significant portion of the total world trade. Therefore, the latter test covers most of the world trade. If this criterion is selected for choosing one theory over another, then the above test is the relevant one. However, if the criterion for the selection of a "good" theory is the theory's ability to explain the network of trade flows for any group of countries, then the results of the bilateral tests are the measure against which the theories must be evaluated. Of course an ideal theory would perform well according to both criteria. But as pointed out before, the result of the bilateral tests of the trade structure of the U.S., the U.K., and South Korea showed that no single theory is capable of explaining the flow of trade among and between the two groups of developed and less- developed economies. While the H-0 theory was shown applicable to trade among less developed countries, it fails most seriously in explaining trade flows between deve10ped and less developed countries. It is in this direction that the alternative theories reveal their own comparative advantage. Therefore it should not be surprising to see in the regression tests of the trade 186 of twenty-three countries, which capture most of the trade among developed countries and trade between these and less deveIOped countries, that the H-0 theory does not perform as well as the human capital approach, the stage of production, and the product cycle theories. However, the varying performances of theories in different spheres raises the relevant question of the level of generality expected from a theory of inter- national trade and the adequacy of the existing theories. A theory of trade is expected to be capable of explaining the structure of commodity flows across all national boundaries. One clear conclusion of this study is that none of the existing theories fulfill this expectation, and in this respect none can be regarded as an adequate theory of international trade. This deficiency may be attributed to the simplistic nature of the explanation that these theories provide. Each theory singles out one economic factor as the predominant force in determining the commodity composition of world trade. The factor proportion theories consider the endowment of resources as the determining variable by assuming the same level of technological development in all countries. On the other hand, the neotechnology theories assume the contrary. And both groups assume identical demand in all countries. However, while there is some truth in each theory's proposition, and each of the preposed variables can I87 explain a part of reality, none can provide a complete explanation. These variables, namely capital and skill endowment of the country, the size of the market, national income, and the level of technological achievements, all play a significant role in determining the pattern of trade. Yet despite their deep interrelations, none may be considered as proxy for the rest. The most challenging task in developing a general theory of trade is the consideration of the interrelation 'between these variables in an effort to formulate an operationally meaningful theory. However, as Professor Keesing pointed out [1], the most striking characteristic of these variables is that “they are intimately related to the growth and deveIOpment process.” Consideration of the problems of economic development and technological change along with the endowment of resources within a dynamic framework, thus, appears to be the inevitable path for the development of a general theory of trade. In addition, the inclusion of the structure of trade pre- ferences, trade barriers, and transportation costs within the context of the theory may prove fruitful. Alternatively, an effort may be made to sharpen the focus of the existing theories in explaining different segments of international trade. This may be done by incorporating the various theories into two or three models for different commodity flows distinguished according to the commodity types (e.g.,homogeneous, 188 non-homogeneous) or the economic structure of the trading countries (e.g., developed, less developed). A possible method will be to construct composite measures of commodity characteristics by incorporating various measures employed in this study. REFERENCE [l] Keesing, D. Comments on G. Hufbauer's paper "The Impact of National Characteristics and Technology on the Commodity Composition of Trade in Manufactured Goods." In R. Vernon (ed.), The Technology Factor in International Trade. New York: NBER, Columbia University Press, 1970, p. 275. 189 APPENDIX 190 Table A—1 A List of Non-Manufactured CommOdities SITC Description of the Commodity, 00 Live animals 01 Meat and preparations 02 Dairy products and eggs 03 Fish and preparations 041 Wheat, unmilled 042 Rice 043 Barley, unmilled O44 Maize, unmilled 045 Cereal, nes., unmilled 051 Fruits, fresh, and nuts, fresh or dry 052 Dried fruits 054 Vegetables, fresh, frozen, or simply preserved 0611 Raw sugar, beet and cane 0616 Natural honey 07 Coffee, tea, cocoa, spices 08 Feeding stuff for animals 09 Miscellaneous food preparations 121 Tbbacco, unmanufactured 21 Hides, skins, and fur skins 22 011 seeds, oil nuts,m ad 011 kernels 23 (except 2312) 24 25 26 (except 2662) 27 28 29 321 331 341 Crude rubber (except synthetic rubber) Wood, lumber, and cork Pulp and waste paper Textile fibers (not manufactured into yarn, thread or fabrics) and their waste (except synthetic fibers) Crude fertilizers and crude minerals (excluding coal petroleum, and precious stones) Metalliferrous ores and metal scrap Crude animal and vegetable materials, nes. Coal, coak, and briquettes Petroleum, crude and partly refined for further refining Gas, natural and manufactured Animal and vegetable oils and fats 191 Table A—2 Capital and Labor Coefficients Per Million Dollars of Value Added for U.S. Industries (1963) NO. Description of Industries Capital Co- Labor Co- efficients efficients ($1000's) (Man Years) 1 Livestock & livestock products 2552.57 76.41 2 Other agricultural products 2729.98 48.70 3 Forestry and fishery 2424.65 89.35 4 Agricultural, forestry & ‘fishery services 1070.00 118.36 5 Iron & ferroalloy ore mining 1587.89 45.47 6 Non-ferrous metal ores mining 1573.68 67.22 7 Coal mining 2087.50 1035.62 8 Crude petroleum & natural gas 1796.15 31.70 9 Stone & clay mining & quarry 1583.87 64.72 10 Chemical & fertilizer mineral mining 1603.83 41.54 11 Food & kindered products 792.11 82.14 12 Tobacco manufactures 1647.53 52.94 13 Broad & narrow fabrics, yarn & thread mills 955.69 153.28 14 Misc. textile good 8 floor coverings 968.53 103.44 15 Apparel 398.98 167.35 16 Misc. fabricated textile products 812.49 140.98 17 Lumber & wood products, excluding containers 939.74 167.14 18 Wooden containers 1119.74 229.88 19 Household furniture 647.44 143.62 20 Other furniture & fixtures 671.70 116.66 21 Paper & allied products, excluding containers 1252.89 78.84 22 Paper board containers & boxes 1320.45 97.88 23 Printing & publishing 505.05 96.40 24 Chemicals 8 selected chemical products 922.20 51.11 25 Plastics & synthetic materials 934.93 61.08 26 Drugs, cleaning & toilet preparations 899.70 38.95 27 Paints & allied products 1110.89 56.65 28 Petroleum refining & related industries 2048.48 50.90 29 Rubber & misc. plastic products 917.88 90.24 30 Leather tanning & industrial leather 686.44 111.11 31 Footwear & other leather products 466.73 179.35 32 Glass & glass products 1032.46 91.21 33 Stone & clay products 990.29 86.54 192 Table A-2 (cont'd) NO. Description of Industries Capital Co— Labor Co- efficients efficients ($1000's) (Man Years) 34 Primary iron 8 steel manufacturing 1478.46 75.42 35 Primary nonferrous metal manu- facturing 1561.86 82.27 36 Metal containers 984.89 73.60 87 Heating, plumbing 8 structure metal products 1004.05 106.23 38 Stamping, screw machine products and belts 859.42 118.44 39 Other fabricated metal products 931.61 100.18 40 Engines 8 turbines 852.98 76.30 41 Farm machinary 8 equipment 948.12 92.62 42 Construction, mining 8 oil field machinery 944.55 79.41 43 Material handling machinery 8 equipment 758.45 71.97 44 Metal working machinery 8 equipment 622.55 92.54 45 Special industry machinery 8 equipment 869.43 91.28 46 General industrial machinery 8 equipment 1126.10 85.31 47 Machine shop products 625.06 139.82 48 Office, computing 8 accounting machinery 982.27 99.75 49 Service industry machinery 889.52 74.50 50 Electrical industrial equipment 8 apparatus 687.90 96.19 51 Household appliances 712.27 74.35 52 Electrical lighting 8 wiring equipment 641.41 98.07 53 Radio, television 8 communication equipment 694.66 86.07 54 Electronic components 8 accessories 634.48 104.46 55 Misc. electrical machinery, equipment 8 supplies 619.81 95.00 56 Motor vehicles 8 equipments 567.21 58.05 57 Aircrafts 8 parts 790.95 81.35 58 Other transportation equipments 979.41 109.11 59 Scientific 8 controling instruments 1084.51 99.80 60 Optical, ophthalmic 8 photography equipment 1036.07 72.78 61 Miscellaneous manufacturing 793.64 116.36 62 Transportation 8 warehousing 1900.86 123.01 193 Table A-2 (cont'd.) No. Description of Industries Capital Co- Labor Co- efficients efficients ($1000's) (Man Years) 63 Communication, excluding radio 8 T.V. 3762.77 101.13 64 Radio 8 T.V. broadcasting 4560.40 98.89 65 Electricity, gas, water 8 sanitation services 1566.93 65.16 66 Wholesale 8 retail trade 1181.88 51.45 67 Finance and insurance 37.88 83.67 68 Real estate 8 rental 4625.00 21.83 69 Hotels, personal 8 repair services 899.00 185.68 70 Business services 1855.60 121.65 71 Medical 8 educational services 2798.57 123.02 Sources: Bureau of Labor Statistics, Patterns of U.S. Economic Growth. Bulletin 1672, U.S. Department of Labor, Washington, 1970. J.W. Kendrick, Industrial Composition of Income and Product. Brooking 1968, pp. 151-75. Capital stock of Industries estimated by Cart and Goddy and used to find the Capital Coefficient of U.S. industries by Z. Iqbal in The Comparative Advantage of Developing Countries in the Manufacturing Industries and the Effect of Generalized Tariff Preferences. Ph.D. Thesis, Michigan State University, 1970. pp. 186-188. 194 Table A-3 Capital and Labor Coefficients per Million Dollars of Value Added for U.K. Industries (1960) NO. Description of Industry Capital Co— Labor Co- efficients efficients (US $1000's) (Man Years) 1 Agriculture, forestry 8 fishery 1347.30 527.74 2 Coal mining 703.45 213.03 3 Mining 8 quarry, nes. 1487.18 370.88 4 Food processing 1367.06 444.30 5 Drink 8 tobacco 1142.86 286.21 6 Coke ovens and coke 4294.18 199.58 7 Mineral oil refining 11884.62 563.19 8 Chemicals 1822.73 261.36 9 Iron 8 steel (melting, rolling 8 casting) 2071.11 335.71 10 Iron 8 steel (tin plates 8 tubes) 1195.88 690.72 11 Non—ferrous metals 1155.69 314.37 12 Engineering 8 electrical goods 835.30 415.96 13 Ship building 8 marine engineering 832.40 566.64 14 Motors 8 cycles 1135.39 408.89 15 Air crafts 975.00 380.10 16 Railway locomotives 8 rolling 1328.13 820.31 stocks 17 Metal goods, nes. 822.03 414.65 18 Textiles 1847.25 562.04 19 Leather, clothing 8 footwear 551.72 612.68 20 Building materials 1175.82 429.75 21 Pottery 8 Classes 860.87 440.99 22 Timber 8 furniture 751.27 536.62 23 Paper 8 printing 1192.79 391.25 24 Other manufactures 1181.00 491.27 25 Construction 399.22 447.54 26 Gas 6239.58 502.23 27 Electricity 13412.56 363.55 28 Water 20543.48 - 295.03 29 Transport 8 Commerce 4233.64 429.60 30 Distributive trades 878.50 430.94 31 Servicesi nes. 937.37 565.00 Source: Department of Applied Economics, Cambridge University, A Program For Growth: Production, Capital 8 Labor. Cambridge, Mass.: MIT Press, 1966. 195 Table A-4 Capital and Labor Coefficients Per Million Dollars of Value Added for South Korean Industries (1966) Capital Co- Labor Co- efficients efficients (US $1000's) (Man Years) NO. Description of Industry \OCDNO‘UbWNi-d Rice, barley 8 wheat* Other agriculture* Forestry* Fishery* Coal Other minerals Processed food Beverages 8 tobacco Fiber spining Textile fabrics Finished textile products Saw mills and phywoods Wood products 8 furniture Paper 8 paper products Printing 8 publishing Leather 8 leather products Rubber products Basic chemicals Intermediate Chemicals Finished chemical products Fertilizers Petroleum 8 coal products Cement Ceramic, clay, 8 stone products Iron 8 steel Steel products Non-ferrous metal products Finished metal products Machinery excepts electrical Electrical machinery Transport equipment Misc. manufacturing Electricity* Banking, insurance, 8 realestate* Water 8 sanitation* Communication* Transportation 8 storage* Trade* Other services* 1430.00 1430.00 433.00 691.00 744.75 989.11 2058.19 933.87 2103.40 2681.43 1433.70 2191.89 549.55 1428.20 2037.67 1885.67 2079.29 2512.81 2512.81 1474.30 2751.13 1597.81 2546.53 1493.50 1632.52 1632.52 2822.16 2271.54 2646.40 1671.37 1689.26 1357.13 3304.00 774.00 3304.00 2610.00 2610.00 556.00 1736.00 5600.00 5600.00 2100.00 2400.00 903.61 942.01 1140.36 310.81 997.02 1847.30 923.22 880.43 1996.60 630.11 1096.82 1673.62 1912.77 719.37 719.37 650.14 323.64 539.56 182.87 1323.89 665.10 665.10 944.13 1690.15 1381.81 870.30 980.87 1771.23 438.00 699.00 438.00 714.00 714.00 1390.00 2760.00 196 Table A-4 (cont'd.) Sources: Economic Planning Board and Korean Reconstruction Bank, Report on Mining and ManufacturingLCensus, Series I— Basic Tables, 1966. Seoul, Korea, 1967. Bank of Korea, Economic Statistics Year Book 1968. Seoul, Korea, 1969. * Data obtained from 1955 Japanese experience from the following sources: T. Watanabe, "Approaches to the Problem of Intercountry Comparison of Input-Output Relationsf' In U.N., A Survgy: International Comparison of Interindustry Data, New York, 1969, PP- 187‘210° Economic Research Institute and Economic Planning Agency, National Income Accounts.1957, Economic Bulletin No.1, February,1959, Capital Structure of Japanese Economy, Tokyo, Japan. 197 Table A-5 Skill Requirements Per Million Dollars of Value Added For U.S. Industries (In Man Years) No.* Professionals Administrators Clerical Sales Manual Service 8 Technicians 8 Managers Workers Workers Workers Workers 1 0.54 45.39 0.46 0.15 29.35 0.15 2 0.34 28.93 0.29 0.10 18.70 0.10 3 5.36 3.75 1.52 0.54 70.85 0.71 4 7.10 4.97 2.01 0.71 93.86 0.95 5 2.82 1.27 2.36 0.05 36.92 0.91 6 0.13 1.88 3.50 0.07 54.58 1.34 7 12.43 29.00 31.07 3.10 931.02 5.18 8 4.34 2.82 3.90 0.16 19.59 0.22 9 2.07 4.40 3.62 0.45 52.23 0.91 10 3.07 2.37 3.03 0.17 31.61 0.37 11 1.72 6.08 8.46 5.59 56.10 1.72 12 0.95 2.12 3.97 2.17 39.92 1.54 13 2.61 4.29 11.19 1.84 125.69 2.76 14 1.03 4.14 7.45 2.48 85.34 1.14 15 1.67 6.69 12.05 4.02 138.06 1.84 16 1.41 1 5.64 10.15 3.38 116.31 1.55 ‘17 1.84 10.20 9.36 2.01 137.39 2.17 18 2.53 14.02 12.87 2.76 188.96 2.99 19 2.73 7.76 13.21 4.16 110.16 1.87 20 2.22 6.30 10.73 3.38 89.49 1.52 21 3.86 3.47 8.59 2.21 57.40 1.34 22 4.80 4.31 10.67 2.74 71.26 1.66 23 8.68 7.23 17.74 19.67 39.91 1.06 24 7.82 3.48 7.82 2.86 26.37 1.07 25 9.35 4.15 9.35 3.42 31.52 1.28 26 5.96 2.65 5.96 2.18 20.10 0.82 27 8.76 3.85 8.67 3.17 29.23 1.19 28 7.53 2.95 9.21 1.27 27.64 0.71 29 5.23 4.06 11.10 2.25 62.72 1.53 30 1.00 6.56 9.00 2.22 86.78 1.11 31 1.08 4.82 16.14 3.05 147.61 1.79 32 4.29 5.47 8.76 2.10 66.40 1.00 33 3.89 5.97 8.22 2.25 62.74 0.87 34 3.62 1.73 7.42 0.60 58.68 1.43 35 6.01 3.46 9.46 1.56 58.25 1.40 36 6.99 4.12 9.79 1.47 48.06 0.96 37 10.09 5.95 14.13 2.12 69.37 1.38 38 11.25 6.63 15.75 2.37 77.34 1.54 39 9.52 5.61 13.32 2.00 65.42 1.30 40 7.02 4.27 10.22 1.67 49.98 0.99 41 8.52 5.19 12.41 2.04 60.67 1.20 198 Table A-5 (cont'd.) No.* Professionals Administrators Clerical Sales Manual Service 8 Technicians 8 Managers Workers Workers Workers Workers 42 7.31 4.45 10.64 1.75 52.01 1.03 43 6.62 4.03 9.64 1.58 47.14 0.94 44 8.51 5.18 12.40 2.04 60.61 1.20 45 8.40 5.11 12.23 2.01 59.79 1.19 46 7.85 _ 4.78 11.43 1.88 55.89 1.11 47 12.86 7.83 18.74 3.08 91.58 1.82 48 9.18 5.59 13.37 2.19 65.34 1.30 49 6.85 4.17 9.98 1.63 48.80 0.97 50 14.33 4.14 14.04 1.44 58.10 1.25 51 11.08 3.20 10.86 1.12 44.91 0.97 52 14.61 4.22 14.32 1.47 59.23 1.27 53 12.82 3.70 12.57 1.29 51.99 1.12 54 15.56 4.49 15.25 1.57 63.09 1.36 55 14.16 4.09 13.87 1.43 57.38 1.23 56 7.89 1.56 7.95 0.35 37.84 0.99 57 9.68 2.28 10.58 0.49 54.83 1.46 58 12.98 3.06 14.18 0.65 73.54 1.96 59 15.87 5.19 16.97 2.20 65.47 1.10 60 11.57 3.78 12.37 1.60 47.74 0.80 61 9.89 7.45 16.64 3.49 74.00 1.28 62 4.06 7.13 21.16 0.86 83.65 4.31 63 10.82 7.79 51.27 1.62 26.09 1.62 64 10.58 7.61 50.14 1.58 25.51 1.58 65 5.15 3.52 12.90 0.65 40.14 1.04 66 1.08 9.78 7.10 15.28 10.24 7.05 67 2.59 14.56 38.74 18.66 2.93 4.77 68 0.68 3.80 10.11 4.87 0.76 1.24 70 47.08 10.10 36.98 3.65 13.02 7.30 71 71.23 2.46 15.75 0.25 5.04 25.71 * The numbers in this table correspond to description of industries in Table A-2. Sources: Data is calculated from the number of employees in each category per 1000 employment in each industry for 1960, from M.A. Horowitz, M.Zymelman, I.L. Hernstadt, Manpower Requirement for PlanningiyAn International Comparison Approach, Vol.11. Boston: Northeastern University Press, 1966, and employment per industry for 1963 from : U.S. Department of Labor, BLS, Pattern of U.S. Economic Growth. Washington, D.C., 1970, p. 98. 199 . . . mm.mHH Hm.©~ om.m mm.oos sm.m ma.s~ ma.m ~m.sa x ecmammN 3mz oso H «ma 0 com o wq.om mo.- sq.m «n.0w Nm.o ao.a~ Ho.a wo.m~ 2 sum «Hamuum:< . . . No.44H sm.om oq.m mo.~oa mm.m mw.¢~ Nm.m am.sa x mas o own 0 Nos o mn.omH qH.mN so.m CH.NNH ma.s ss.qm Ho.m ms.oa z cmams mmn.o mmo.o omo.o Nm RAH mm mm mm m mm woa co m cm mm mm m mo.ma x ma.mma oo.~m No.q mm.~¢a wa.n mm.mm mw.oa mn.HH .z .«mmoudm umzuo mmm.o Nmm.o mma.o ms.sHH ma.m~ ms.m Nm.mOH ms.m sm.s~ . “H.a os.oa x sm.sNH wa.o~ ma.m mm.mafi ss.o an.s~ Ho.OH sa.ca z .e.: . . . Ho.qafl om.o~ mq.m ma.m0a .Hq.m mw.sm. aa.o H©.~H x New C mas o ow“ o Hm.o~H ma.m~ Na.m 44.0HH wm.o No.4N am.m o~.ma z. wn . . . no.4HH os.o~ Hm.m mm.moa ~m.m so.s~ so.m Ns.sa x mmw o mas o Has 0 wo.m~a oe.q~ ae.m mm.mHH om.m sfl.s~ mm.a mo.ma 2 names TlMunu TIMucu Mucs Mun" Mun. Mung V. 11 .d a... o... o... 3...... 13 m. mm mm. mm. m m. m.” .. no : cs : co + 2.x. +.x.T. X.M X.n x.a N.1 e T. u. I. as n8 c. a I. a T. a I. a e a s a I. 13 e u u e a (x r\ (x C.1 T. 7s; a 1 o 1 T. 1 1 o a u T. I.u 9 SI (SP 88 S S 38 10:8 DPS + a T. S 3 I. I. p 1 a m. mu w s e O T... 1 S as E as 5 3 5 as... E mucwamomaamm uuoaBH m>wuflumaeoo cam muuoaxm wowunuommscmz .m.p wo mumaaoa coHHsz umm mucmawufiscmm Hafixm o-<-mapms 200 Hom.a Hom.o mqm.o woo.o coo.a maw.o Es omm.o \.I IA LC.» 00 S OHM.H owo.o mmm.o mw¢.o oom.o Han.o mNo.o I).—4 21 (8: s qom.a wno.o mqm.o wmq.o mmm.o mam.o mmo.o nom.o )L IT. .0 S mm.oma N¢.oo mmmq mvmcwo 201 .mqloa .ao .m manmh .moma umnEm>oz ..mmm¢wmsm uamuuso mo mo>usm.cfi :.moma pom monuosuum osmosonusacH: .mmo .mouanou mo ucmEoummmn .m.D mam .m|< manme "mmounom .mH>mHmomsw cam .cfimam .vcmHmMH .wcmamoH .mommuu waHvSHocH .1 A.v.ucoov ol< mHan 202 Table A-7 Skill Requirements Per Million Dollars of Value Added for U.K. Industries (in Man Years) Professionals Administrators * and and Clerical Sales Manual Service NO- Technicians Managers Workers Workers Workers Workers 1 2.11 220.60 7.39 0.53 292.90 0.0 2 4.05 2.13 9.16 0.0 193.86 3.20 3 5.19 5.19 17.43 0.37 333.05 5.19 4 10.66 20.88 55.09 30.21 293.24 13.32 5 4.29 6.30 58.67 9.73 192.91 10.59 6 17.16 8.18 32.93 2.99 126.13 7.78 7 48.43 23.09 92.93 8.45 355.94 21.96 8 27.70 12.55 42.86 10.19 148.45 10.98 9 17.12 10.41 32.56 2.35 253.15 8.39 10 34.54 16.58 63.55 3.45 535.76 17.27 11 16.35 11.32 37.10 2.83 228.55 7.86 12 47.00 12.89 81.94 9.15 262.47 8.32 13 52.70 11.33 66.86 2.27 390.98 11.90 14 40.48 9.00 50.70 1.63 276.41 9.00 15 37.63 8.36 47.13 1.52 256.95 8.36 16 76.29 16.41 96.80 3.28 566.01 17.23 17 23.22 14.93 39.81 2.90 310.57 10.34 18 5.62 20.80 37.09 6.18 464.81 31.47 19 3.06 22.06 43.50 5.51 520.78 6.74 20 23.64 21.49 66.18 11.60 292.66 10.31 21 13.23 16.76 40.13 5.73 347.06 8.38 22 6.44 25.76 41.86 7.51 426.61 5.37 23 21.91 19.56 60.25 10.56 268.83 9.39 24 25.55 26.53 65.34 11.79 338.49 10.32 25 15.22 37.59 24.17 1.34 359.37 2.69 26 33.65 18.58 95.93 16.57 313.89 16.57 27 24.36 13.45 69.44 12.00 227.19 12.00 28 19.77 10.92 56.35 9.74 184.39 9.74 29 26.21 10.74 67.45 3.44 281.82 34.80 31 139.56 13.00 87.58 7.91 89.27 86.45 * The numbers correspond to description of industries in Table A-3. Sources: Data is calculated from the number of employees in each category per 1000 employment in each industry for 1960, from M.A. Horowitz, M. Zymelman, I.L. Hernstadt, Manpower Reguirement for Planning: An International Comparison Approach, Vol. 11. Boston: Northeastern University Press, 1966, and employment per industry for 1960 from Department of Applied Economics, Cambridge University, A Program for Growth: Production, Capital and Labor. Cambridge, Mass; MIT Press, 1966. 203 mmo.a ooo.a w~.aqo 0N.HOH no.nm nN.Hmn mo.¢H om.5HH $0.0m mm.om x . oo.mam om.mm. oo.am mc.moq mm.NH mm.om om.m~ cm.en z mwmcwu m¢m.o own.o w~.o~m mo.HOH Nm.mn «c.9Nn mm.oH no.mHH no.0m oo.Hn x I qw.mmo no.mm. mn.mm mm.qoo mH.nH no.0HH mm.qm mm.¢o z comma «Nu.o «Ho.o om.Hmm mm.NOH wo.¢n ma.~¢m ¢¢.¢H mm.hHH «m.mN mH.mn N Hmwauuom msaa nq.mmo ~¢.~m ma.mm Hm.mmo No.0H ac.ooa nn.¢n mo.~m z moousm nonuo oao.a Nmm.o Nu.qqo mo.ooa Ho.nm ow-mmm Hm.ma o~.mHH oo.am me.mo x AammSuuom om.ooo no.mm mq.mm no.0mn H¢.¢H no.0HH mm.Hm mo.mo z .xmv ¢Hmm oso.a o~o.4 so.NHe mH.HoH ma.¢m ms.~om no.44 ms.oHH ~m.am om.ao x AsHmUH o~.onn co.mm om.cm no.o~n mo.¢a m¢.nHH oo.m~ oo.ou 2 .Nov 0mm scara mam.a aN.oHo Ho.na n<.em nn.hom ¢~.ca a~.caa He.om o~.uo x mH.Hmn om.am ~o.on madman hm.NH mn.qaa Hm.m~ ma.- z .m.: omo.o “mm.o mm.~mm oo.moa em. m mn.wem o¢.¢a n~.oaa ew.m~ N~.ns N moasoaoom . - mq.mmn om.om Ho.¢m ow.omm No.ua Hoa.o HNm.o om.aao mm.aoa mm.mn ou.~on o~.cH «n.5HH -.om mo.au N . \; oo.nan ww.¢m hm.cm wn.ocm Hm.eH an.oaa an.on mo.e@ z name: 0...... arr wmm Um“ Mum. mm mm mm m m m. m : on C. as 1 8 J I. 1 M 1 u 1 T. 1 a u o o a. (\ +.x.x. + x.T. x. x.n x.a x.1 a I. u. I. I\ a I. .u.L a I. a e a s a I. .8 s u u a a $11 318 13 .1I 1 13 an? Ius 8T (SP 88 8 s 88 1P8 abs +. a T. S 3 .1 .1 D. 1 s o o n w a o an Ass ARV Ass Ans Ase Ame ANV u Ass mucoaoomammm uuomaH m>fiufiummaou was monomxm wcfiuauomMscmz .M.D wo mumaaoa cOfiHHflz pom mandamufiavmm Hafixm ml¢ magma 7,“ mmw.o www.o wmm.o mow.o Hem.o mow.o oow.o L) (8 s qqq.o mmu.o mmw.o Hm©.o 0mm.o (83I) 31 s owo.o ©mm.0 mom.o Hmm.o Hqc.o How.o wow.o omo.o 43.4sm ms.mms mq.mmm om.omo mo.owm No.mqo oq.qoo Hm.mmm mq.awm Hm.mmw ms.qmm Nq.omm m¢.m©m mm.mq¢ mm.moo oq.mqo mm.oom om.mmm (9 + S + 7) 313x10“ paIIIxsun Amv (z + T) mm.mm ~©.mm NN.mOH ma.qm Ho.©m om.mm oo.mHH mm.mm qw.moa mm.mm Rm.moa mo.¢m HH.moa cm.mn om.NOH qa.om «m.noa qw.mm M 8 813x10 PBIIIX Anv qq.am mn.mm om.mm om.¢m mm.qm oq.om Ho.mm om.om Hm.mm oq.mq mq.mm mo.mm mm.Nm ma.mm mq.qm mm.mm qo.mm No.0m MUS 513x10 BOIAla Asv mo.maq om.mnm qm.omm mo.wwm mm.oqm oa.omm mo.mmm mq.oow om.mmm HH.N©N Hm.mqm mm.qu 00.0Nm wq.qoq Hq.qmm me.mmm om.mqm q~.Hmm w 318x10“ {anus Amv A.e.ucouv mu< mamas wm.ma qm.oa mH.mH Nu.mH mo.ma mo.oH om.ma mq.qa om.ma om.¢H Ho.ma oq.qH NN.¢H Ho.m qo.qa Ho.qa Hm.qH on.mH 813x10“ sates Aqv OH.mHH wm.m0H mm.NNH oa.woa N¢.HHH m¢.©oa mm.omH mm.qoa mm.waa mm.ooa oo.waa Nq.mm co.o~a mm.ow mw.maa mo.¢oa Nw.NNH qq.m0a M Slaxlo IBDIIGID AmV mm.mm mn.mm wo.mm mm.nm NN.mN ma.om mm.mm co.mq ma.mm NN.mm mc.mm Ha.om mq.nm No.m~ ma.m~ mm.qm qq.mm mm.mm pus sxaBsusw sxoneinsIuImpV Amv mq.mm cw.©m «m.mm «n.0m on.oo mm.om mm.nw NN.mq mq.nm oo.mq wN.mm om.mm No.m~ qn.Nm ma.mn mm.qm om.mm 0\ ON sueIquqoal pus SIBUOISSBJOJJIR Adv muonsm cumummm .m.m.m.: .mmHEocoom was IcmHm xaamuucmo mmuox Lusom mfim< moauwa mowuma< afiumq mmHaocoom vmaoam>mn mama waofiafiaon cumnusom 205 .moma .mmmum HHS “.mmmz .mmvfiunamo .oomanqmma mcofiumamm usmusouusde nmfiufium "cusouu you amuwmum ¢ .muwmum>fis: mmowunamo .mmwaocoom vaqu¢ wo ucmauumama vow .514 manma ”mmouoom .mofiuwd susom new .ccmammN 3oz .mwamuuma<*« .m«>mamow:w vcm .cwmam .vcmamuH .vcmamoH .mommuu mmwsaocfi maousm umzuo i A.v.ucoov w-< magma 206 Table A—9 Skill Requirements Per Million Dollars of Value Added for South Korean Industries (In Man Years) * Professionals Administrators Clerical Sales Manual Service No. & Technicians & Managers Workers Workers Workers Workers 1 5.60 1848.00 33.60 5.60 3701.60 0.0 2 5.60 1848.00 33.60 5.60 3701.60 0.0 3 10.50 18.90 100.80 4.20 1938.30 12.60 4 45.60 12.00 24.00 2.40 2304.00 2.40 5 18.07 9.04 94.88 6.33 747.28 20.78 6 34.86 24.49 123.41 5.65 730.07 29.20 7 3.42 38.77 77.54 100.35 907.73 7.98 8 16.47 10.88 83.30 5.28 177.78 18.03 9 3.99 29.91 66.80 10.97 879.36 14.96 10 1.85 62.81 112.68 60.96 1594.22 9.24 11 0.92 31.39 56.32 30.47 796.73 4.62 12 0.0 33.45 61.63 16.73 762.45 2.64 13 0.0 63.89 115.80 35.94 1766.99 5.99 14 7.56 23.94 71.83 13.23 500.93 10.71 15 122.84 75.68 205.11 57.03 622.99 6.58 16 3.35 56.90 100.42 90.38 1410.86 10.04 17 28.69 86.08 267.79 43.99 1444.14 51.64 18 33.09 35.97 135.96 20.14 471.91 19.42 19 33.09 35.97 135.96 20.14 471.91 19.42 20 29.90 32.51 122.87 18.20 426.48 17.55 21 14.84 16.18 61.17 9.06 . 212.32 8.74 22 15.10 30.76 104.13 14.03 353:94 16.72 23 0.73 6.04 12.80 2.20 157.82 1.28 24 0.53 43.69 92.67 15.89 1142.52 9.27 25 11.97 24.61 75.82 7.98 533.39 11.30 26 11.97 24.61 75.82 7.98 533.39 11.30 27 16.99 34.93 107.63 11.33 757.20 16.05 28 8.45 70.99 138.59 30.42 1418.03 11.83 29 48.36 77.38 196.21 22.11 1017.01 23.49 30 50.48 44.38 176.67 16.54 554.37 14.79 31 39.24 28.44 107.28 6.87 743.51 20.60 32 21.25 61.99 125.76 56.68 1486.06 7.08 34 3.50 70.60 470.43 125.82 18.87 7.69 36 189.92 29.27 94.96 2.14 404.84 19.28 37 29.27 34.99 168.50 2.14 456.96 18.56 38 5.56 382.25 115.37 600.48 193.21 86.18 39 894.24 88.32 601.68 24.84 477.48 662.40 207 Table A-9 (cont'd.) * The numbers correspond to description of industries in Table A-4. Sources: Data is calculated from the number of employees in each category per 1000 employment in each industry based on Japanese Data for 1950, from M.A. Horowitz, M. Zymelman, I.L. Hernstadt, Manpower Reqpirement for Planningp:An International Comparison Approach,_Vol. 11. Boston: Northeastern University Press, 1966, and employment per industry from Economic Planning Board, and Korean Reconstruction Bank, Report on Mining and Manufactupipngensus, Series I, Basic Tables. 1966. Seoul, Korea, 1967. 208 moo.a mmn.H Hmw.H qu.N Hmm.H mmw.a Hwa.a qm~.q mom.m moq.m mwm.q mmH.H mom.¢ mum.m 0mm.m mm.mama mN.¢m©H oo.¢HmH Hq.nwm om.NmHH qq.mmw om.NmQH on.aom qw.~moa Hm.nmm no.wmoa mH.oma «N.maoa m¢.Hnm (9 + s + v) 913x10” PaIIInsun Amv ~©.Ho qw.qw mH.ow mm.mn mo.mc mo.mm mm.oo m~.0¢ om.mm no.5m «5.55 Hm.mm mm.on mm.am (z + I) 513110“ paIIIxs ARV mucmamomaamm uuoaaH o>HuHumano cam muuoaxm mm.©H mq.na om.qa cm.wH mN.wH mm.oN N¢.©H cm.ma mw.NH Nq.ma oo.mH mm.wH «H.¢H ma.mH Slanlom aoIAJas on om.meH mm.qwoa mn.~mqa mq.mq¢ mo.NMHH oa.mmw mm.woma mm.o~m mm.mHoH Hm.qom m~.mmma mm.mam o~.HmmH mn.oma Slanlon {snusw Amv oo.oq om.Nm ma.oq oq.mm qo.mq mo.oH mm.mq sq.ma qo.oq mn.mm wN.Hm Nm.qm om.mq MH.NN szaxion sates Aqv cm.mma H~.mmH OH.0NH mm.nma cc.nma mm.qqa nm.mma Hc.HmH Hm.aMH 05.nca om.mma om.HqH om.oma no.0qa Siaaiom Ievrlato Amv om.om mm.wo q~.mm mq.om qo.mq No.mq N~.mm wH.qm HN.co Hw.mm ma.mo mq.Hm oo.mo 9N~o «- $18890 pus SJOJBJJSIUIUIPV R Amv Nw.oa oq.©H Hq.oa ww.w~ Ho.ma mo.nm Hm.oa mo.om mm.ma ow.mm do.NH No.am H¢.NH O M pus susIquqoal a SIBUOISSBJOJJ AHV 2x EN EN vcmammm 3oz .mwamuums< comma .M.D maousm cumumwz .m.D mmHEocoom wmaoam>ma namoz wcwusuommscmz ammuox susom mo mumHHoa SOfiHHfiz pom wucmEmufisvmm Hawxm OHI< maan 209 .s manna .000H wow mmanma mcofiumamm zMumswcHuwucH «uamusonusmaH .mmqu 00 xamm may .ucmauummmn noumwmmm 06m .mHI< mfinmy “mmousom . waox 0Com N50.H 500.0 5mm.H 00.000H 00.05 0H.0~ -.500H 00.00 00.00H 00.00 0N.0H x 0 .muoamwcfim. 00.000H 00.05 00.5H H0.000 00.H0 «0.00H 00.H0 00.00 x ..mfim%mamz. 000.0 000.0 000.0 00.000H 00.00 00.0H 00.05NH 00.00 HN.5NH 00.00 ~0.NH x 0smaamnh 00.05HH 00.00 05.0H 0H.N50H 00.~0H 50.00 0H.50 H0.0 z 0:0 cm3HmH mfimwaovaH 0HN.H. 550.0 000.N NH.005H 05.00 0H.5N 00.000H 00.00 00.05H 00.05 00.00 x 0cm .cmumfixmm «5.0NOH 0H.00 0~.0H 00.000H 00.00 05.00H 0H.00 50.5w z..mH0GH .maunm 05H.N H00.0 00H.0 00.050H 00.00 00.N~ 05.0HOH 00.50 50.00H 00.00 00.NH x . 00.000 00.00 50.00 H5.050 50.5H 00.HNH 00.50 00.00 x .ummm mavwfiz . . . 50.50HN 00.00 ~H.0H 00.000N H0.00 00.000 00.05 00.0 x «so N wnu ma mNN ma sm.s~m am.s~ om.- ow.~mm H~.~H mo.a- s~.mn mo.mn z augum< 000.H NN0.0 000.0 05.0000 «0.00 00.~H 00.050H 00.00 00.0HH 00.00 00.5 x H0.0m0 00.00 0N.0H 5H.000 00.00 00.000 00.00 50.QH wauuauua< awuwn 500.H 500.0 050.0 00.000H 00.05 05.00 5N.NO0H 00.00 5H.00H 00.00 «0.0a x moaaoaoum 00.000H 50.05 50.00 05.000H 50.00 0H.00H 00.00 05.00 00 coaoao>mn 0004 \I .5 \1 1n \1 A“ XIMunu xlmucu Mucu Mu Mucs Mung "w "v .1 .4 SE 12 II Von 10% 03 0 os 01 e P a 1 .. S .. S .. S 18 1.... J In 1.... 18 u m D O .8 9. rs + H.X. +ax.T. X.M x.n 3.8 x.1 a I. u. I. (x (x ,\ a I. a I. a I. a s a s a I. .3 e u u e a SIT. 7u13 13 1T. 1 1.3 ant. TrUS 31 (SP 39 s a. se 1P8 ops + a T. s 3 I. I. D. 1 m o o, e u I.\ 10 S 8 m “a A00 A50 A00 A00 A00 Amv AN0 9 AHV A.e.ucouV OHI< sagas 210 0050.0 0.0000 000.0 000.0 muoswoua . 0mo0usmom9pmza 0cm 0mc0o0vwz 000 0000.0 0.0000 000.0 500.0 mwcm0aum> .muc0m0 .mucm500m 000 mnw0.o 0.0000 wqo.o 000.0 muumuuxm mafiacmu 0cm waflxo 0mm 0000.0 5.0000 000.0 000.0 .oum .mwwsummzw 00cmwuo o0um£uc00 000 0000.0 0.0000 000.0 500.0 wmw 0mu2umc .Eno0ouum0 .0moo Eoum m0mo0Emno mwsuo 0cm umu 0muoc0z 000 0000.0 0.0000 000.0 000.0 m0m0umume m>0uomo00mm 000 0000.0 0.0000 000.0 000.0: m0mo0awzo 00cmwuoa0 H930 000 0055.0 0.0000 000.0 050.0: .oum .muamEm0m .m0mo0emno 00cm0u060 000 0500.0 0.5000 000.0 000.0 m0mo0am£o 00amwu0 000 0000.0 0.0000 000.0 000.0: muosvoum Bam0ouumm 000 wm~0.o 0.0mm0 Nm~.o ooo.o umpnsu afiumauaxm «.0mm 0050.0 0.0000 000.0 000.0 mmpsuommncma ooomnoa 000 0000.0 0.0000 000.0 000.0 mmwmuw>w0 000osoo0< 000 0050.0 0.0000 000.0 000.0 o00o£oo0maoz 000 0000.0 0.0000 000.0 000.0 mumao0uommcoo umwsm 000 ~000.o 0.0mo0 000.0 mma.o- mmaon 0am umwsm 00o 0500.0 0.0000 000.0 000.0 wm00muwwm> 0m>ummmum 000 5000.0 0.0000 000.0 000.0 u0=um vm>umwmum 000 00m0.o 0.0mm0 000.0 mma.o muowumumamua ammumo mqo 0000.0 0.0000 050.0 000.0 .mmc .m0m0umo mo .u3000 .0mmz 500 mmHN.o 0.0m00 cum.o 000.o aflfimma .ummcs mo .uaoaw .Hmmz 00¢ co0um0u wuma 00um0 mmfiaoc co0umwuomm0 mu0voEEou 0000 Icmumww0n mvmue mwoow Ioom uonvoum omu0m umasmaou m0mom mm0u00oEEo0 mo muam0o000mo0 0000ocnomh map 00l¢ m0an 211 5000.0 5000.0 0000.0 0550.0 0000.0 0000.0 0500.0 0000.0 0000.0 0000.0 0005.0 0050.0 0000.0 0000.0 0005.0 0000.0 0050.0 0050.0 0005.0 0000.0 0005.0 co0um0u -cmumMMHa uoauoum 0.0000 000.0 0.0000 550.0 0.0000 550.0 0.0000 550.0 0.0000 550.0 0.0000 050.0 0.0000 000.0 0.0000 000.0 0.0000 000.0 0.5000 000.0 0.5000 000.0 0.0000 000.0 0.0000 000.0 0.0000 505.0 0.0000 000.0 0.0000 500.0 0.0000 050.0 0.0000 000.0 0.0000 000.0 0.0000 000.0 0.0000 000.0 mama O0um0 mwmue m0000 umu0m umssmcou 000.0 000.01 000.01 000.0I 000.0 000.0 000.0 000.0 000.0 000.0 000.0 000.0 000.0 000.01 000.0 000.0 050.0: 050.0 000.0 000.0 000.0 wm0Eoc loom m0mom 1.0..coov mo0u0m0 m00uxmu 0m0om00 000 .ouw .zum00ou05m .mom0 .000DH 000 aw>oB .wo0u0mM m00uxmu umnuo 000 am>o3 .mo0u0m0 couuoo 000 wmmusu 0am Guam m00ume 000 uumop umama no umamm .0050 pwama mo mvme mm0o0uu< 000 wumop nmmma 0am 000mm 000 mmHSuommaamE xuoo 000 .mma .mmHSuomMscma 00oz 000 .oum .mwumon 0003000 .mummam> 000 “wand“ 00 mw0o0uu< 000 umnnsu 0o m0~0umumz 000 umsumm0 mo mmunuommdamz 000 uwnummA 000 .mmc .muoswoua 0cm m0m0uwuma 0mo0amnu 000 wa0mwu 0m0o0w0uum 0am mm00300mo 0mumuwao0mu .m0m0umuma u0umm0m 000 muoavoua o0cnowuouza vam mm>0mo00xm 050 mum~000uumm 0mu=uommaamz 000 mco0umumamuq 0a0cm0000 .0:0mcmm0o .mmmom 000 mo0umamoo 0cm 0umsamumm 000 m0w0umuma uo>m0w .mssmuma .m00o 0m0ucmmmm 000 c00umfipommo mu00ossoo 000m 001< m0£mH 212 0005.0 0.0000 000.0 000.0: oC00 000 0000.0 5.0000 000.0 000.0: 0mm0 000 0005.0 0.5000 000.0 000.0: 83:08:0< 000 0050.0 0.0000 000.0 000.01 0mxo0z 000 0000.0 0.0000 000.0 500.0! ummmoo 000 00mm.o o.wmm0 000.0 mom.ou .e:c0um0q .um>HHm 0mg oomm.0 0.0mm0 000.0 000.0- Umxuozc: .mwcfiwHOW can wcfiummo 000 m000.0 0.0000 000.0 000.0 mmcfiuufiw cam .mmafia .mmnse 000 memo.o c.00m0 “No.0 mao.o wow my“: wafluaauxm .mp0: 000 0000.0 0.0000 500.0 000.0 m0m0umums co0uosuumcoo xomuu 0cm m00mm 050 0000.0 0.0000 500.0 000.0 00uum 0cm 0000 . 050 0000.0 0.0000 500.0 000.0 mummSm 05m .mmum0a .m0mmum>0:0 050 0000.0 0.0000 500.0 000.0 mmamnm .mm00cm .mvou .mumm 050 0500.0 0.0000 500.0 000.0 mauom >um80ua umnuo 0cm muomc0 050 5000.0 0.0000 000.0 000.0 .ouw .mho00mouumm .cou0 mwcoam .aou0 00m 050 0000.0 5.0000 000.0 000.0 humuuom 000 0500.0 0.0000 000.0 000.0 mumsmmm00 000 0000.0 0.0000 000.0 000.0 mmm00 000 0005.0 0.0000 000.0 000.0 .mma .mmusuomwdcma 0muma02 000 0005.0 0.0000 000.0 000.0 m0m0umume c00uo=uumcoo >M0o 000 0050.0 0.0000 000.0 000.0: m0m0uwuma 0:000050 .ucmEmo .ma00 000 0000.0 0.0000 000.0 000.0 .oum .mm0uummmMu .m0c0um>oo ~000m 500 0500.0 0.0000 000.0 000.0: mm0o0uum 000uxmu malmvmz 000 c00um0u muma. o0umm mmmmmm c00umwuomma 0u0woaaou 0000 Icmum0000 mvmue m0000 Ioom uoauoum umu0m umssmcoo m0mom 0.0.uc000 HH-< magma 213 0000.0 0000.0 0000.0 0005.0 0500.0 0000.0 0000.0 0000.0 0000.0 0000.0 0000.0 0000.0 0500.0 0000.0 0000.0 0000.0 0000.0 5000.0 0000.0 0000.0 0000.0 0000.0 .Ilmmwwmmw lamum0000 uoanoum 0.0000 500.0 5.0000 000.0 0.0000 000.0 0.0000 000.0 0.0000 000.0 0.0000 000.0 0.0000 000.0 0.0000 000.0 0.0000 000.0 5.0000 000.0 0.0000 000.0 0.0000 000.0 0.0000 000.0 0.0000 000.0 0.0000 000.0 0.0000 000.0 0.0000 000.0 0.0000 500.0 0.0000 500.0 0.0000 000.0 0.0000 000.0 0.5000 000.0 muwa 00umm mwmue m0o00 umu0m umeamco0 000.0 mucm50020w 0mo0uuom0m o0umman 005 000.0 mauwumamm :00umo0::EEoom0wH 005 Hmo.o >u0u0uuum0m 0500:00uum00 wow unm800200 005 000.0 ummwnou03m .zuwc050me Mason o0uuom0m 005 000.0 .mma .mmocm00mam 0:0 0nmd0numz 005 000.0 mm0uum20a0 0m0omam How mmc0nomz 005 000.0 hpr0nomE umnumm0 0cm m00ume 505 000.0 mumc0coms 0a0xuo30mumz 005 000.0 mmGHnome mo0000 005 ~0o.o mumcfisums 0mu200900uw< N00 000.0 wma0soma 0c0umumcmw umzom 005 000.0 .mwc .mmsuuommscme 0mumz 000 000.0 mucm50020m 0003mm=om 500 0~H.o uumfluao 0mg 050.0 mc0£om8 no 0am: ~00 m0ooH 000 mmo.o- muflpfiu .mu0on .munc .mamwuom .m00mz 000 moo.o- mHHHuw 0G0ocmw 0cm muosvoum wu03 000 000.0 uuommcmuu 0cm mwmuoum MOM muwc0muaoo 00umz 000 000.0 mmHSuosuum 0am muuma 0musuo=uum vmnm0c0m 000 000.0: m0muwa mzouumwcoc msomam00mom0z 000 000.0: as0uo£u 0am 8:0amu0 000 000.0! :09 500 mmmmmm co0um0uomm01N000oaaoo 0900 1000 m0mom 0.0.ucoov 00I< «0009 214 0000.0 0500.0 0000.0 5000.0 0000.0 0000.0 0000.0 0000.0 5050.0 5000.0 0000.0 0000.0 0000.0 0000.0 0000.0 0000.0 0000.0 0500.0 0000.0 0000.0 110000000. ucmummw0o nonvoum 0.0000 500.0 0.5000 000.0 0.0000 500.0 0.0000 000.0 0.0000 000.0 0.0000 000.0 5.0000 000.0 5.5000 000.0 0.0000 000.0 5.0000 050.0 0.0000 000.0 0.5000 000.0 0.0000 000.0 0.0000 000.0 5.0000 000.0 0.0000 000.0 000000 550.0 0.0000 000.0 0.0000 050.0 5.0000 050.0 mumn OHHQM mvmufi mvooc umufim umadmcoo 050.0 000.0 000.0 000.0: 000.0 000.0 000.0 000.0 000.0 500.0! 000.0 000.0 000.0 000.0 000.0 000.0 000.0 000.0 000.0 000.0 mm0aoc loom m0mum mw0u0uum o0umm0m 000 muwuuma amuC0um 000 wumwuouwu 0::0m .mucmasuumSH 0mo0m52 000 mx000o 0cm mwnoumz 000 6000 mamc0u 0w000w>mm 000 mm00aasm o0camuwouocm 000 mucwesuumc0 Hmofiuao .0m00wwa .UHWHucmfiom 000 ummauoom 000 0:0nu00u mam 000 ppm unmoxm 0:0Su00u 000 mm000uum ~000B0m 05¢ .m0m00:m£ .m0000 0m>mue 000 mu5u0au=m 000 mmHSux0w 0c0unw00 0cm wa0ummn .wc0nas00 .mumu0amm 000 mumon 0am 000m 005 mummuou0< 005 acuoacoc .mm0u0£m> 0mom 005 mM0u0nm> HOuoa 0mom 005 mm0o0nm> 0m300m0 005 msumumamm mumC0£oma 0mo0uuum0m umnuo 005 mmmoansm 0mu00m8 How maumuwaam 0mu0uuom0m 005 c00um0uummn 0u0voaaoo 090m 0.0.ucoov 00I< Q0008 215 .0500 .mmmum 0u0mum>0cp 00052000 .0002 "3000 3mz ammMHw 0ma00umcumuc0 :0 neuomm 0000ocnome may .0.0mv .0 .coaum> a0 01¢ w00me Lmvooo amusuummsamz :0 mvmue mo c00u0moaaoo 0u0voaaoo wnu co 0wo0oanuma 0am mo0um0umuomumno 0mc00uwz mo uomQBH may: ..o.u .umsmnmam "muusom 0005.o 0.0000 5N0.o 000.0 .mm: .mm000uum nmusuummscmz 000 0~qm.o 0.0000 “Nm.o ooo.o mmHHQmsm hymGOHumum 0cm mafiwwo 000 5005.0 0.5000 500.0 000.0 m0000 0c0uuomm .mzou .mu0u00snamumm 000 c00um0u wumm 00umm mm0an co0um0uumm0 mu0uoasoo 000m Icmummw0a mvmph m0000 loom uunvoum umu0m um62mco0 m0mom 0.0.ucouV 00I< m0nmh 216 Table A-12 Estimates of Optimum Plant Size in U.S. Industries (1954) Commodity Optimum Commodity Optimum Classification Plant Size Classification Plant Size (SITC) (Millions (SITC) (Millions of $'s)fi of_§'s) 332 9.392 665 9.861 512 69.806 666 1.705 513 69.806 671 4-587 514 69.806 672 1- 714 531 1.536 673 26.534 532 1.411 674 26.534 533 1.536 675 26.534 541 48.874 676 26.534 551 0.105 677 26.534 553 0.105 678 26.534 554 0.105 679 26.534 581 69.806 681 4.067 682 40.67 611 1.282 683 4.067 612 1.282 684 4.067 629 1.282 685 4.067 631 0.230 686 4.067 632 0.160 687 4.067 641 0.497 689 6.750 642 0.497 691 2.183 651 0.512 692 2.183 652 1.162 693 2.183 653 1.162 694 0.564 654 0.311 695 0.433 655 0.311 696 0.433 656 0.143 697 0.564 657 5.141 698 0.564 661 0.329 662 0.815 711 2.183 663 0.704 712 5.320 664 26.333 714 1.900 217 Table A-12 (cont'd) Commodity Optimum Commodity Optimum Classification Plant Size Classification Plant Size (SITC) (Millions (SITC) (Millions of §fs) of §'s) 715 8.084 734 93.900* 717 0.533 718 0.895 812 6.166 719 4.204 821 0.160 722 10.867 831 1.283 723 1.006 841 0.683 724 10.867 851 1.283 725 1.006 864 0.387 726 2.263 891 6.064 729 2.273 892 2.329 731 41.600* 894 1.474 733 1.140 ! 895 0.560 * Estimates not provided by Saving. Source: Saving, T.R. "Estimation of Optimum Size of Plant by the Survival Technique." _anrterly Journal of Economics. November, 1961, pp. 598-602. 218 nom.o Hoo.o q¢w.o omo.H me.o wmm.o. cam.o oo5.o HOH.H mmo.o 5mm.o 5N¢.o H05.o wmo.H 0mm.o 5cm.o m¢0.H 5a.omm wo.5on ow.HHo 5o.mmm mm.amm N¢.¢cn om.mmc wo.¢mn oa.o~e mn.amm ~5.¢Nn 50.non 6m.o~a om.-a (9 + s + v) szaxlom PaIIInsun V Am mm.N5 mm.M5 mm.HOH ww.¢a mm.w5 mm.wo no.¢e no.55 ao.on «v.05 55.n5 nc.e5 ~m.q~ MH.@N o¢.mm O¢.¢~ (z + I) 313x10“ Pattrxs m¢.aH 5H.NN wN.mn 5m.¢n no.0N HQ.NN oe.wH m~.NN on.nH mn.o~ om.mH Hw.HN 5o.m mo.m 81331.10“ 3 5," «30; aarazss A \O V 00.non 5c.~mm cN.Nem mm.mem mo.man mo.¢mm 5m.Nne ou.¢~m H5.mmm ne.5on mm.¢m¢ 5m.mmm om.HHH nn.nHH m.noa N n .QHH @ sleaze“ funny“ A In V oc.ma < ZN ovaxmz Gammm HHH msouu cmmmh mamuH mauums< madmauwnuwz asfimamm magnum 220 000.0 000.0 000.0 055.0 500.0 000.0 000.0 000.0 500.5 055.0 000.0 050.0 000.0 (8:1):3 S :0 000.0 000.5 005.0 000.0 000.0 000.0 000.0 :I) L (9 00.0050 00.005 00.0000 05.050 00.0000 00.000 00.0000 00.050 00.0000 00.0000 00.0000 00.0000 00.0000 00.0mm )Mn ,‘(9 + S + 9 813x10 paIIInsu 00 v \T 00 000 00.0000 00.005 00.0000 50.000 00.0000 00.000 00.0000 05.000 00.0500 00.5000 05.0000 00.000 05.0500 00.000 M” 9N Slanio Isnu 000 00.00 05.00 00.00 00.00 00.00 50.00 00.00 00.50 00.00 00.00 00.00 00.00 00.50 00.00 Mucu 313x10 $319 000 0.0.00000 00:4 00000 00.000 00.000 00.000 00.000 00.500 50.000 00.000 00.000 05.000 00.000 00.000 00.000 00.000 00.000 M 0 313x10 1931131 000 sxaBsusn 00.50 00.00 05.05 00.00 00.00 00.00 00.00 00.00 00.00 05.00 00.00 00.00 05.00 50.00 A N v pus Sloaexnsrurmpv 55.00 00.50 00.0 00.00 00.00 00.00 00.00 00.00 00.50 00.00 00.00 00.50 00.00 Ch \0 O\ N susroruqoal pus steuorssagoxd 000 M0wc0 cmuw0xmm amB0mH mwqu suaom 0Com 0:00 00>00m0030 0mmsuuom 221 .0 00000 .0000 .0000 060 060000 .m8600000m 0000300000000 .000050I00000 .mmuox 00 0000 00B .0008000000 00000000 000 .0I< 0000B 8000 000 0:000 000 .0000 .00000 902 ".0002 .000000800 .0000I0000 000000000 000000300000 .003000 000 8000000 < .0u0mum>0c= 000000800 .000800000 000000< 00 0008000000 000 .5I< 0000a 8000.00 0:000 000 .0 00009.50100 .00 .0000 00080>0z .00000000 usmuuao 00.00>u:0 00 :.0000 000 manuoauum usmusouuSQSH: .000 .00008800 00 0008000000 .0.D 000 .0|< 00000 8000 0 0:000 000 000080000 00 00000000 0000 000 000080003000 00030 00000 ”0000000 A.0.ucoov 00:4 00000. 222 Table A014 Capital and Labor Embodied in One Million Dollars of Manufacturing Exports and Imports of Twenty-three Countries IMPORTS EXPORTS K L K L YK/L Country (U.S. $1000) (Man (U.S. $1000)‘ (Man Group I Years) Years) United States 2309.31 181.91 2338.85 170.52 0.926 Canada 2267.26 178.41 2240.93 174.44 0.989 Group II Sweden 3403.28 745.58 2550.38 703.55 1.259 Norway 2956.81 712.14 2337.28 559.35 0.994 Australia 2730.53 745.55 2899.99 646.40 0.816 Germany 3061.06 739.57 2742.72 741.89 1.120 United Kingdom 3804.04 874.03 3602.82 876.81 1.093 Denmark 3278.65 747.32 2533.27 713.57 1.236 France 2750.01 723.15 2923.81 744.23 0.968 Belgium 2857.89 738.62 3250.86 736.15 0.876 Netherlands 2876.59 759.99 3788.83 717.51 0.717 Austria 3029.60 765.38 2752.29 772.17 1.110 Italy 2794.75 701.08 3124.05 807.62 1.031 Japan 3237.78 660.68 2735.68 756.40 1.354 Group III . Spain 2185.21 1151.84 2354.66 1417.08 1.142 Mexico 2137.29 1162.64 2389.82 1202.18 0.925 Portugal 2181.00 1159.43 2554.22 1651.94 1.217 Yugoslavia 2263.84 1235.23 2342.60 1372.23 1.074 Hong Kong 2402.10 1439.52 2652.99 1873.60 1.178 South Korea 2237.24 1199.47 2408.06 1829.71 1.417 Taiwan 2149.25 1126.60 2430.21 1577.54 1.238 Pakistan 2089.01 -1091.10 3331.71 2105.71 1.210 India 1983.88 936.13 3144.74 1944.02 1.310 a (K/L)M K/L (K/L)x Sources:Total capital and labor requirement of industries are estimated for Group I from Table A92 and U.S. "Input-Output Structure for 1963? . 0p. cit., for Group II from Table A-3 and U.K. Input-Output Relations BET—311., for Group III from Table Ar4 and South Korea Input-Output, Inferifidustry Relation Tables for 1963, op. cit. Trade data from Commodity Trade Statistics. 223 NUflwerOU .z.: wcm xfiwcmma< .NH-<.asm HHI< mfinme .muaumflumum ocmua Um "mmuunom .Axd + N.ov\Azd + N.oV uH >« wmm.q mmo.~ H¢©.NH waqm.a q~ao.o: omqo.o mHUcH omn.o «mm.H omo.ma cunN.H mmoo.OI omqo.o amumflxmm wom.~ omm.q wmm.ma quH.H oqao.o ooqo.o cmafimfi qu.m qo¢.~ mcH.qH quN.H HNHO.OI maqo.o mwuox nuaom Hem.q mam.a. Hm¢.m moHN.H ooao.ou oomo.o waox waom mwm.H ch.n Now.qH mwNH.H NHHo.o ammo.o mH>mHmowaw qu.q mmw.m mom.ma mHmO.H Homo.o aoqo.o Hmwzuuom nmo.~ mew.o mmo.o~ omo~.a qu0.01 quo.o ouwxmz mwm.H mm¢.w nom.ma quo.a ommo.o Hoqo.o Gammm HH nacho oqm.o «mo.qa qmw.HH ommw.o o~¢o.o ooao.o Gamma eHw.o mam.oa ewq.ma NNHO.H ommo.o quo.o hamuH oom.H mn~.oa Hom.mH mooa.o oaqo.o ommo.o mauum=< HHO.H Hmo.mH HmH.mH mmmo.a camo.o ammo.o mvcmaumnumz Hum.o omo.na NmH.NH moqm.o quo.o mmmo.o Edfiwamm mum.o mqo.na mNm.qH momm.o oflqo.o m¢mo.o magnum cmshmu H¢~.H HoH.mH maw.ma wm¢~.H omoo.o mmmo.o mHHmuums< NHH.~ mom.n Noo.oa Nnao.a mNHo.o ammo.o hmsuoz nqo.H ~om.~H mom.mH omq¢.o quo.c mHmo.o amvosm HH asouu Nom.o Ncm.<~ H¢H.NN «moa.o mmqo.o quo.o mvmcmu nmn.o www.mw om~.ua Howw.o mmoo.o ammo.o mmumum wmufiaa om Hum H azouu N > memomxm magnum .. > 929:8 among 5558 QOmHmz m.OzH>mHmomnw mmnm.a mowo.o meow.o qmw.o qo.om m¢.HN HHNq.o mommwo qum.o Hawauuom qqu.a qmmo.o omnm.o qow.o om.m~ mm.o~ mmmq.o mmwm.o wme.o coaxmz mmo~.a qmo~.o comm.o mmm.o mm.o~ NH.- NmNm.o owcq.o mmmm.o awmam HHH msouu mcom.o omow.o nmmm.o NMH.H ~H.HN oo.q~ mmm¢.o mnn~.o ~o-.o Gamma mmoo.H «mow.o mmmm.o «oo.H mm.- H¢.N~ mmao.a owqm.o cmm~.o KnHmuH momm.o w~mm.o nmom.o omo.a Nw.H~ mm.- mmmo.a wo-.o nmm~.o mauum5< Hmao.a Homw.o meom.o oom.o wm.- Hm.HN mmqm.o cmam.o Nmm~.o mwamaumnumz omwo.H menu.o maow.o mum.o om.- qm.- cmoa.a Heem.o wo-.o avwmamm quO.H mmmm.o maqw.o Nam.o mo.- n¢.- wmsm.o omom.o mmc~.o muamum mm¢w.o momm.o mwmw.o o¢a.o Hm.- -.- -mm.o onm.o onN.o Mumaama ~q¢m.o ownw.o Hoam.o Hmo.H mN.NN Nq.m~ ammo.a moom.o cnmm.o Eocwafix vmuwan o~om.o ~com.o mmwn.o omo.H eN.HN mm.m~ Hmnm.a om-.o woom.o hcmauwo qamm.~ mwme.o seam.o new.o nm.¢~ oo.a~ moww.o Num~.o wao~.o mwamuums< Humm.o co~m.o o~mw.o amm.o «N.~N «H.NN mnwm.H moca.o oqom.o mmauoz mumm.o nwa.o mno.H om.0~ No.~N mmom.H wNmH.o maom.o amvmzm HH macho «HQH.H wom~.o oawm.o mam.o mo.HN mm.o~ Hoqe.a wmma.o Nm-.o mvmamo mnow.o ommm.o ~mm~.o NOH.H mm.om mo.- «mmc.a HmoN.o Hom~.o mmumum wouaa: . . H maouu an» mmm mu muuomxm wuuoaaH > muuomxm muuomaH r muuomxm muuomaH huucsoo macho uuavoum mam Hmowwoaocnome coauuzvoum mo meum mumps mo huomzh wmwuucsou mmunulhucmae now mvooo cannuUMwscmz mo mvmue mo muuoaaH cam muuomxm mo mumaaon COfiHHHZ mao ca vmfivonam xmvaH coaumfiucwuwwwfio uosvoum vcm .mw< uusnoum we ucmwuwmwmoo .Ofiumm muoou umadmcoo oHI< wHQMH BIBLIOGRAPHY 225 BIBLIOGRAPHY Arrow, K.J., Chenery, H.B., Minhas, B.S., and Solow, R.M. ”Capital-Labor Substitution and Economic Efficiency." Review of Economics and Statistics. August, 1961, pp. 225-50. Baldwin, R.E. "Determinant of Commodity Structure of U.S. Trade." American Economic Review. March, 1971, pp. 126-46. Ball, D.S. ”Factor Intensity Reversal in International Comparison of Factor Cost and Factor Use." Journal of Political Economy. February, 1966, pp. 77-80. Balogh, T. The Dollar Crisis: Cause and Cure. Oxford: Blackwell, 1950. Basevi, G. ”The United States Tariff Structure: Estimate of Effective Rates and Protection of United States Industries and Industrial Labor.” Review of Economics and Statistics. May, 1966, pp. 147-160. Becker, G. ”Investment in Human Capital: A Theoretical Analysis." Journal of Political Economy. Supplement, October, 1962, pp. 9-49. Bhagwati, J. ”The Pure Theory of International Trade: A Survey.” Survey of Economic Theory, Growth and Development. Prepared’for the American Economic Association and the Royal Economic Society. New York: St. Martin's Press, 1965. Bharadwaj, R. Structural Basis of India's Foreign Trade. Bombay: University of Bombay, 1962. Bharadwaj, R., and Bhagwati, J. "Human Capital and the Pattern of Foreign Trade: The Indian Case." The Indian Economic Review. October, 1967, pp. 117-42. 226 Buchanan, N.S. ”Lines on the Leontief Paradox.” Review of Economics and Statistics. August, 1954, pp. 286-289. Creamer, D., Dolseovolsky, S.P., and Borenstein, 1. Capital in Manufacturing and Mining, Its Formation and Financing. Princeton, N.J.: Princeton University Press, 1960. Department of Applied Economics, Cambridge University. A Program for Growth: Production, Capital and Labor. Cambridge, Mass.: MIT Press, 1963. A Program For Growth: Production, Capital and Iabor; Cambridge, Mass.: MIT Press, 1963. Diab, M.A. The United States Capital Position and the Structure of its Foreign—Trade. Amsterdam: North-Holland Publishing, 1956. Economic Planning Board and Korean Reconstruction Bank. Report on Miningyand Manufacturing Census, Series I--Basic Tables, 1966. Seoul, Korea, I967. Economic Research Institute, Economic Planning Agency. National Income Accounts. Tokyo, Japan, 1957. National Wealth Survey. Tokyo, Japan, 1955. Economic Bulletin No.1. Tokyo, Japan, Feb- ruary, 1959. Capital Structure of Japanese Economy. Tokyo, Japan, 1956. Ellsworth, P.T. "The Structure of American Foreign Trade: A New View Examined." Review of Economics and Statistics. August, 1954, pp. 279-85. Ford, J.L. The Ohlin-Heckscher Theory of the Basis and Effects of Commodity Trade. New York: Asia Publishing House, 1965. Gruber, W., Mehta, D., and Vernon, R. ”The R a D Factor in International Trade and Investment of United States Industries.” Journal of Political Economy. 1967, pp. 20-371 Heckscher, E. "The Effect of Foreign Trade on the Distribution of Income." Economisk Tidskirft. 1919, pp. 497-512. 227 Hicks, J.R. ”An Inaugural Lecture." Oxford Economic Papers. June, 1953, pp. 117-35. Hirsch, S. "The United States Electronic Industry in International Trade." National Institute Economic Review. November, 1963, pp. 92-_7. "Technological Factors in the Composition and Direction of Israel's Industrial Exports" in R. Vernon (ed.), The Technology Factor in International Trade. Neinork: NBER, Columbia University Press, 1970. Hirschman, A.O. "The Political Economy of Import Substitution in Latin America." Quarterly Journal of Economics. February, 1968, pp. l-32. Hodd, M. ”An Empirical Investigation of the H-O Theory.” Economica. February, 1967, pp. 20-29. Horowitz, M.A., Zymelman, M., and Hernstadt, I.L. Manpower Reguirement for Planning: An Inter- national Comparison Approach, Volume IIi Boston: Northeastern University Press, 1966. Hufbauer, G.C. Synthetic Materials and the Theory of International Trade. Cambridge, Mass.: Harvard University Press, 1966. ”The Impact of National Characteristics and Technology on the Commodity Composition of Trade in Manufactured Goods? In R. Vernon (ed.), The Technology Factor in International Trade. New York: NBER, Columbia University Press, 1970. International Labor Office. Year Book of Labor Statistics. Geneva, 1966. Iqbal, Z. The Comparative Advantage of Developing Countries in the Manudeturing Industries and the Effects of’Generalized'Tariff Preferences. PHLD. Thesis, Michigan State University, 1971. Jones, R. "Factor Proportions and the Heckscher-Ohlin Theorem." Review of Economic Studies. 1956- 57, pp. 1-10. Keesing, D.B. "Labor Skills and International Trade: Evaluating Many Trade Flows with a Single Measuring Device." Review of Economics and Statistics. August, 1963, pp. 287-94. 228 ”The Impact of Research and Development on United States Trade.” Journal of Political Economy. February, 1967, pp. 38-48. ”Labor Skills and the Structure of Trade in Manufactures." in P.B. Kenen and R. Lawrence (eds.), The Open Economy, Essays on International Trade and-Finance. New York: Columbia University Press, 1968. Kenderick, J.W. Productivity Trends: Capital and Labor. New York: NBER, Princeton University Press, 1961. Kenen, P.B. "Nature, Capital, and Trade." Journal of Political Economy. October, 1965, pp. 437-60. Kenen, P.B., and Yudin, E. "Skills, Human Capital and U.S. Foreign Trade.” Columbia University International Economic WorkshOp, December, 1968, Mimeographed. Kindleberger, C.P. Foreign Trade and National Economy, New Haven: Yale University Press, 1962. Kravis, I.B. "Wages and Foreign Trade." Review of Economics and Statistics. February, 1956, pp.5I4-30. ‘ ' "Availability and Other Influences on the Commodity Composition of Trade.” qurnal of Political Economy. April, 1956, pp. 143-55. Kreinin, M., ”Comparative Labor Effectiveness and the Leontief Scarce Factor Paradox." American Economic Review. March, 1965, pp. 131-10. Kuznets, S. Six Lectures in Economic Growth. Glencoe, Illinois: Free Press, 1959. Lancaster, K. “The Heckscher-Ohlin Model: A Geometric Treatment." Economica. February, 1957, pp. 19-39. Lary, H. Import of Manufactures From Less Developed Countries. New York: NBER, Columbia University Press, T968. Leontief, W. "Domestic Production and Foreign Trade, The American Capital Position Re-Examined.” Economia Internazionale. February, 1954, pp. 9-15i‘ . 229 ”Factor Proportion and the Structure of American Trade: Further Theoretical and Empirical Analysis.” Review of Economics and Statistics. November, 1956, pp. 387-407. ”International Factor Cost and Factor Use.” American Economic Review. June, 1964, pp. 335—45. Linder, S.B. An Essay on Trade and Transformation. New York: Wiley, 1961. MacDougal, G.D.A. "British and American Exports: A Study Suggested by the Theory of Comparative Advantage Part I." Economic Journal. December, 1951, pp. 697-721} Miernyk, W.H. The Elements of Input-Output Analysis.' New York: Random House, 1965. Minhas, B.S. An International Comparison of Factor Cost and Factor Use. Amsterdam: North-Holland PubliShing, T963. Mookerjee, S. Factor Endowment and International Trade. Bombay, 1968. Ohlin, B. Interregional and International Trade. Revised Edition. Cambridge, Mass.: Harvard University Press, 1967. Posner, M.V. "International Trade and Technical Change." Oxford Economic Papers. October, 1961, pp. 323-41. Research Department, Bank of Korea. Input-Output: Interindustry Relation Tables for 1963. Seoul, Korea, 1966. Table 4. Economic Statistics Year Book 1968. Seoul, Korea, 1969. Robinson, R., Factor Proportions and the Comparative Advantage: Part I." Quarterly Journal of Economics. May, 1956, pp. 169-92. Roskamp, K. "Factor Proportions and Foreign Trade: The Case of West Germany." Weltwirtschaftilches Archives. 1963, pp. 319-26. Roskamp, 230 K., and McMeekin, G. “Factor PrOportion, Human Capital and Foreign Trade: The Case of West Germany Reconsidered.” Omarterly Journal of Economics. February, 1968, pp. 152960. SamueIson, P.A. "International Trade and Equalization of Saving, T Schultz, Stigler, Stolper, Swerling, Tarshis, Tatemoto, Travis, W. Factor Prices.” Economic Journal. June 1948, pp. 163-84. ”International Factor Price Equalization; Once Again.” Economic Journal. June 1949, pp. 181-97. "A Comment on Factor Price Equalization." Review of Economic Studies. 1951-1952, p. 121. .R. "Estimation of Optimum Size of Plant by Survivor Technique." Quarterly Journal of Economics. November, 1961, pp. 569-607. T.W. ”Reflection on Investment in Man.” Journal of Political Economy. Supplement, October, 1962, pp. l-8. G.J. ”The Economies of Scale,” Journal of Law and Economics. October, 1958. W., and Roskamp, K. "Input-Output Table for East Germany with Application to Foreign Trade." Bulletin of Oxford University Institute of Economics and Statistics. November, 1961, pp. 379-92. B.C. "Capital Shortage and Labor Surplus in the United States." Review of Economies and Statistics. August, 1954, pp.'286-89. L. "Factor Inputs and International Price Comparisons." In Id.Abramovitz and Others (eds.), The Allocation of Economic Resources. Stanford, California: Stanford University Press, 1959, pp. 236-44. M., and Ichimura, S. "Factor Proportion and Foreign Trade: The Case of Japan.” Review of Economics and Statistics. November, 1959, pp. 442-46. P. The Theory of Trade and Protection. Cambridge, Mass.: Harvafd University Press, 1964. 231 United Nations. Commodity Trade Statistics. Statistical Papers, Series D. New York, 1969, 1970. Year Book of National Accounts Statistics. New York, 1966. , Economic Commission for Latin America. Economic Bulletin For Latin America. March, 1964, pp. 1-49. United States Department of Commerce, Bureau of Census. U.S. Foreign Trade Statistics Classification and Cross Classication 1970.iwashington, D.C., 1970. U.S. Census of Manufactures, Washington D.C., 1963. , Office of Business Economics. ”Input-Output Structure for 1963.” Survey of Current Business. November, 1969, pp. 16-47, Table 3. United States Tariff Commission. Competitiveness of U.S. Industries. TC Publication 473, Washington, D.C}, Apiil, 1972. Vaccara, B. Employment and Output in Protected Manufact- uring Industries. Washington, D.C.: Brooking Institution, 1960. Valavanis-Vail, S. ”Leontief's Scarce Factor Paradox." Journal of Political Economy. 1954, pp. $23-28. Vanek, J. The Natural Resource Content of United States Foreign Trade, 1870-1955. Cambridge, Mass.: MIT Press, 1963. "Factor PrOportion Theory: The N-Factor Case.” Kyklos. October, 1968, pp. 749-56. Vernon, R. ”International Investment and International Trade in Product Cycle." Quarterly Journal of Economics. May, 1966, pp. 190-207. Waehrer, H. “Wage Rates, Labor Skills and United States Foreign Trade." In P.B. Kenen and R. Lawrence (eds.), The Open Economy, Essays in International Trade and Finance. New York: Columbia University Press, 1968. Walters, A.A. "Production and Cost Functions: An Eco- nometric Survey." Econometrica. January-April, 1963, pp. 1-66. 232 Watanabe, T. "Approaches to the Problem of Inter-Country Weiss, L. Comparison of Input-Output Relations." In United Nations. International Comparison of Interindustry Data. NewFYOrk, 1969, pp. 187- 210. "Concentration and Labor Earnings." Paper 6405, Social System Research Institute. Madison: University of Wisconsin, 1964. ”Capital and Labor Requirements for Canada's Foreign Trade." The Canadian Journal of Economics and Politicdl Science. August, 1961, pp. 349-58. Williams, J.H. Trade Not Aid: A Program For World Stability. Cambridge Mass.: Harvard University Press, 1953.