STEEL AND ECONOMIC swemmmt: ctapsz-ouwm RATEOS IN THREE LATIN AMERECAN STEEL MAN-rs ' T513535 for the Dames a? N1. 75D; V MECE’flGAN STATE UNIVERSITY" Dwifi G. 53.an V 1963- THESIS This is to certify that the thesis entitled STEEL AND ECONOMIC DEVELOPMENT: CAPITAL-OUTPUT RATIOS IN THREE LATIN AMERICAN STEEL PLANTS presented by David G. Greene has been accepted towards fulfillment of the requirements for f/q ./ / 1 _£/L/N;//h [ZiA//V/ /(/L l) J Major p{o/feésorU U7 ,./ 7 0-169 LIBRARY Michigan State University ABSTRACT STEEL.AND ECONOMIC DEVELOPMENT: CAPITAL-OUTPUT RATIOS IN THREE LATIN AMERICAN STEEL PLANTS by David G. Greene The years since‘world'War II have been marked by great rises in the aspirations of the peoples of the under- develOped nations manifesting itself in a drive for national growth and develOpment. This desire for economic growth.has led to policies designed to promote industrialization. The steel plant has become to these aspiring nations a symbol of economic develOpment and national prestige. However, a steel plant represents a considerable investment and should be undertaken only after thorough analysis of the existing alternatives. Unfortunately, this is not always done. In this paper the investment of capital resources in three Latin.American steel plants is appraised. The plants involved in the study are Companhia Siderurgica Nacional of Brazil, Chile's Compania de Acero del Pacifico, S.A., and Altos Hornos de Mexico, S.A. The principal instrument used for the appraisal of these investments is the capital-output ratio (COR). In the first chapter motives generally cited for establishment of a steel industry are discussed and appraised. Most of these arguments are seen to have little economic foundation. 2 The second chapter includes a description of the development of the three steel plants and a discussion of the economic and political conditions leading to their establishment. It is shown that the establishment of each plant was a product of economic nationalism. In the third chapter three general types of invest- ment criteria are reviewed: marginal productivity, pro- graming, and factor intensity. Because of theoretical objections and empirical difficulties, the first two alter- natives are rejected in favor of the third. As used in this paper, the factor-intensity criterion involves comparison of the COR of the steel plant with the incremental capital- output ratio (IGOR) for the economy and the manufacturing sector. In the next chapter the theoretical and empirical problems involved in the calculation and interpretation of 003s are discussed. The following chapter shows the results of calcu- lation and caparison of the com. The 003s of the firms are compared over time and the reasons for and implications of the downward trend of firm GORs are discussed. The roles of price changes, depreciation policy, and changes in product mix are given special emphasis. Interfirm collapari- sons of 003s are made in order to determine the nature of the relationship between size of firm, degree of development of the economy, and the firm's COR. The COR of each firm is compared with hypothetical alternatives based on the 1003s for the economy and 3 manufacturing sectors. In the case of Brazil and Chile the firm CORs were substantially above the Icons for the economy and manufacturing sector. The difference was less sub- stantial in the case of Mexico. The evidence indicates that investment in.the steel plant was considerably more I'capital absorbing" than the average investment being made in the economy at the same time. Finally, a comparison is made of the cost of domes- tic steel and imported steel (without tariffs). In all cases domestic steel was considerably mere expensive. This indicates that the steel industry must be assisted by tariffs or subsidies. Investigation shows that the steel plants have been generously protected and subsidized. The findings of this paper indicate strongly that investment in.a steel plant is strongly capital absorbing and must be appraised carefully; iFurthermore, the study indicates that the predisposition of international lending agencies toward approval of steel plants may lead under- developed countries to direct their planning toward these projects rather than to consider logical alternatives. STEEL AND ECONOMIC DEVELOPMENT: CAPITAL-OUTPUT RATIOS IN THREE LATIN AMERICAN STEEL PLANTS By. \‘0 (3' David Gg'Greene A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Economics 1963 323?83 A 72-44 ACKNOWLEDGMENTS The author wishes to thank Dr. Paul Strassmann of Michigan State University for his guidance and encouragement in the completion of this dissertation. The bulk of the research was done at the Export-Import Bank of Washington, D.C. , with the permission and cooperation of Walter Sauer, Executive Vice-President. The patient and expert assistance of many of the staff members of the Bank is greatly appreci- ated. I especially want to thank my wife, Roberta, for her moral and physical support and for putting up with me during the writing of this paper. 11 TABLE OF CONTENTS AcnowIEDMNTS O O O O O O O O O O O O O C O O O 0 LIST OF TABIES O O O O O O C O O O O O O O O O O 0 L18 T OF A PPENDIX TABIE S O O O O O O O O O O O O 0 0 Chapter I. II. III. ImODU CTIO n O O O O O O O O O O O O O O O STEEL AND ECONOMIC DEVELOPMENT . . . . . . Steel and the Developmental Process . . . Motives for Establishing Steel Plants . . Balance-of-Payments Difficulties . . . The Symbolic Role of Steel . . . . . . Shortages of Steel . . . . Iron Ore Deposits and their Develop- mental Effects . . . . . . . . . . . THREE LATIN AMERICAN STEEL PLANTS: ESTABLISHMENT AND DEVELOEMENT . . . . . . Development of the Companhia Siderurgica nac1°m1 O O O O O O O O O O O O O 0 Establishment of Compania de Acero del Pacifico . . . . . . . . . . . . . Establishment of Altos Hornos de Mexico, 8 ”A O O 0 O O C O O O O O O O O O O 0 Conclusion . . . . . . . . . . . . . . . THE CAPITAL-OUTPUT RATIO AS AN INVESTMENT CR1 ERICK e e e e e e e e e e e e e e e e Marginal Productivity Criteria Programming Criteria . . . . . Capital Intensity Criteria . . Conclusion . . . . . . . . . . iii Page ii vi viii 5: ¥;FSo> cnc- er Ia Chapter IV. V. VI. VII. TABLE OF conmws (continued) EMPIRICAL PROBLEMS . . . . . . . . The Incremental Capital-Output Ratio ICOR of the Mbnufacturing Sector . Capital-Output Ratios of the Firm . Capital Stock of the Enterprise . value of Output . . . . . . . . . Adjustment of the Capital Stock for Price Changes . . . . . . . . Re-evaluation of Depreciation . . . Estimation of Value Added . . 'USING THE CAPITAL-OUTPUT RATIO TO APPRAISE THE STEEL INVESTMENT: THEORETICAL AND Adjustment of Value of Output foi- éhénéeé in the Price Level . . . . . . . . . . smry . . . . . . O C O C . . . O . C . PIANTS O O O O O O O O O O O O O O The Capital-Output Ratio Over Time Expected Findings . . . . . . . . Actual Fmdmgs O O O O O O O O 0 Implications . . . . . . . . . . Interfirm.Comparisons of CORs . . . Expected Findings . . . . . . . . Actual Findmga O O O O O O O O O cone lu‘ ion 0 O O O C O O O O O O O EVALUATING THE STEEL INVESTMENT BY COMPARISON WITH ICORS FOR THE ECONOMY AND MANUFACTURING SECTOR . . . . . Technical Considerations . . Comparison with the Incremental Capital- Output Ratios for the Economy . . Comparison with the Incremental Capital- Output Ratios of the Manufacturing sector 0 O O O O O O O 0 O O O 0 COST OF DOMESTIC AND IMPORTED STEEL . iv CAPITAL-OUTPUT RATIOS OF THE THREE STEEL Page 77 77 82 a: 88 9O 92 9h 96 98 98 98 106 11h 117 117 127 13h 137 137 1110 153 TABLE OF CONTENTS (continued) Chapter Page Expected Findings . . . . . . . . . . . . . 153 Actual Findings . . . . . . . . 15h Protection and Subsidization of the Steel Industry . . . . . . . . . . . . . 163 VIII 0 CONCLUSIONS 0 O O O O I O O O O O O O O O O O 171 Evaluation Using CORs and ICORs . . . . . . 171 Costs of Domestic and Imported Steel . . . 172 Implications . . . . . . . . . . . . . . . 173 BBLIOGRAPEY O O O O O O O O O O O O O O O 0 O O O O 176 APmme O O O O O O O O 0 O O O O O O O O O O O O O 186 Table 1. 10. 11. LIST OF TABLES Brazilian.Rationa1 Production of Pig Iron, Steel Ingot, and Rolled Products and Per Cent Accounted for by Companhia Siderurgica N8010nfl1, 19:46-58 0 e e e e e e e e e o e 0 Comparison of Ingot Capacity and Ingot Pro- duction: Altos Hornos de Mexico, S.A., 1914-9“ 60 e e e e e e e e e e e e e e e e a Mexican National Production of Pig Iron, Steel Ingot, and Rolled Products and Per Cent Accounted for by.A1tos Hornos, 19“.“.‘58 e e e e e e e e e e e e e e e e e Annual Depreciation as a Percentage of Book Value of Fixed.Assets: Altos Hornos, Companhia Siderurgica Nacional, and Compania de Acero del Pacifico . . . . . Capital-Output Ratio, various Formulations, Companhia Siderurgica Nacional, 19h7-59 . Capital-Output Ratio, Various Formulations, Compania de Acero del Pacifico, 1951-59 . Capital-Output Ratio, Various Formulations, Altos Rornos de Mexico, S.A., l9h5-59 . . . Average Annual Change in Capital-Output Ratios, Post Break-in Period, as Indicated by the Slape of a Straight Line Fitted by the Method of Least Squares, CSN, AHMSA, and CAP . . . Comparison of Costs Per Ton of Production of Pig Iron by Different Processes . . . . . . . Estimated Plant Capital Investment Cost Per Ton Required to Produce Various Tonnages of Ingots Per Year by Different Processes . Output Range of Different Types of Mills Rolling Flat Steel . . . . . . . . . . . vi Page 33 50 53 91 107 108 109 111 123 12h 126 Table 12. 13. 16. 17. 18. 19. 20. 21. 22. 23. LIST OF TABLES (continued) Selected Measures of Economic Deve10pment: Brazil, Chile, and Mexico . . . . . . . . . . Comparison of Average, Post Break-in Capital- Output Ratios of CSN, CAP, and AHMSA, Values and Ranks 0 O O O O I O O O O O O O O O O O 0 Comparison of the Gross COR of the Steel Plants, Post Break-in Period, with the Appropriate ICOR for the Economy, l939-h9 . . Comparison of the Net COR of the Steel Plants, Post Break-in Period, with the Appropriate IGOR for the Economy, 1939-“.9 e e e e e e e 0 Chile: Capital Stock and Gross Product, Mifl.5 , Manufacturing, and Construction, 19 -2e 0 e e e e e e e e e e e e e e e e 0 Mexico: Investment and Value Added in the Manufacturing Sector, 1939-50, in.Market and 1950 Price, 0 O O O 0 O O O O O O O O O O O 0 Comparison of the Adjusted Capital-Output Ratio for AHMSA, 19h7-58, with the Incremental Capital-Output Ratio of the Mexican.Manu- £8015me SOO'COI‘, 19h0'50 e e e e e e e e e 0 Comparison of Estimated Transportation.Require- .ments for Coal, Iron Ore, and Finished Steel to Deliver One Ton of Finished Steel to Domestic Market, Domestic Steel Plant Com- pared with North.American Plants . . . . . . Estimated Transportation.Requirements for the Production of One Ton of Pig Iron at Selected Locations in Latin.America and the United States 0 O O O O O O O O O O 0 O O O O O O 0 Estimated Cost of Finished Steel Products in Three Latin American Steel Plants and Sparrows Point in 19h8 Dollars . . . . . . . Comparison of Estimated Delivered Cost of Imported Steel (c.i.f. ) with Estimated Domestic Prices, CSN, CAP, and AHMSA . . . . Cogparison of Prices of Imported and Domestic eel at Monterrey, Mexico, with and without Tariff and Tariff Per Ton, 1955 . . . . . . . vii Page 128 130 lhl 1&2 1H6 1&8 150 155 156 157 159 168 Table 1. 10. 11. LIST OF APPENDIX TABLES Wholesale Price Indexes: Brazil, Chile, and HOXiOO, 1939-59 0 e e e e e e e e e e e e e e Companhia Siderurgica Nacional: Fixed Asset Stock: Gross and Net: Book Value and 1950 hiOOS , 19u2'59 e e e e e e e e e e e e e e e Compania de Acero del Pacifico: Fixed Asset Stock: Gross and Net: Book Value and 1950 ”1608 , 19h7-59 e e e e e e e e e e e e e e e Altos Hornos de Mexico: Fixed Asset Stock: Gross and Net: Book Value and 1950 Prices, 19'42'59 e e e e e e e e e e e e e e e e e e 0 Brazil: National Product and Capital Formation: ‘Market and 1950 Prices, 1939-58 . Chile: National Product and Capital Formation: Market and 1950 Prices, l9hO-S8 . . . . . . . Mexico: National Product and Capital Formation: Market and 1950 Prices, 1939-58 . Companhia Siderurgica Nacional: Tonnage Output by Product, Index of Total Output, and Index of Output‘Weighted by value of Product, 1950:100, 19h6'59 e e o e e e e e e e e e e e Altos Hornos de Mexico: Tonnage Output by Product, Index of Total Output, and Index of Output We ted by Value of Product, 1950= 100, 19%. 9 e e e e e e e e e e e e e e e e Chilean Exchange Rates, Conversion of Data for Compania de Acero del Pacifico from.Pesos to Dallars O O O O O O O O O O O O O O O 0 O O 0 Compania de.Acero del Pacifico: Tonnage Sales, Finished and Semifinished Products, Domestic and Export, Tonnage Sales Index, viii Page 186 187 188 189 190 191 192 193 19h 195 LIST OF APPENDIX TABLES (continued) Table Page and Sales Index'weighted by Value of Product, 1950:100, 1950-58 a e e e e e e e e e e e e e 196 12. Compania de Acero del Pacifico: Total Ship- ments by Pr0du0t, 1951-57 0 e e e e e e e e e 198 13. Compania de Acero del Pacifico: Estimation of Exports of Smmifinished Products, 1951-55 0 e e a e e e e e a e e e e e e e e e 199 1b. Altos Hornos de Mexico: Estimation of value Added and value Added as a Per Cent of 88168 ’ 191111-59 e e e e e e e e e e e e e e e 200 15. Companhia Siderurgica Racional: Estimation of Value Added and.va1ue Added as a Per Cent Of $8108, 19h7-59 e e e e e e e e e e e e e e 201 16. Compania de Acero del Pacifico: Estimation or value Added, 1950-1957/58 e e e e e e e e 202 17. Compania de Acero del Pacifico: Estimate of Value Added, 1958-59 and 1959-60 . . . . . . 203 18. Deflation of Sales and Value Added: CAP, CSN, and AW 0 O O O C C O O O O O O O O O C O 0 20“ 19. Capital-Output Ratios for CAP, CSN, and AMSA O O O O O O O O O O O O O O O O O O O O 2 05 INTRODUCTION The post4Wor1d War II era has seen a great rise in the aspirations of the peeples of the underdeveIOped world. Accompanying this has been a drive for economic growth.and deveIOpment. In.many of the less developed countries, this has resulted in economic policies designed to promote industrialization.and among these nations the steel industry has become the symbol of economic deve10pment and national prestige. A steel plant represents a considerable invest- ment for an underdeveloped country and, therefore, should be undertaken only after thorough analysis of the costs and benefits of this investment and of the existing alternatives. Unfortunately, this is not always done. In this paper the investment of capital resources in three Latin.American steel plants has been appraised. These plants are: Brazil's Companhia Siderurgica Nacional, Chile's Compania de Acero del Pacifico, S.A., and Mexico's Altos Rornos. The principal instrument used for the appraisal of these investments is the capital-output ratio. The first chapter will discuss and appraise the motives generally cited when underdeveloped countries con- sider the establishment of a steel industry. It will be seen that many of these arguments have little economic 1 2 foundation, and that the steel industry should not be appraised on any ”special" basis. The next chapter will consist of an.historical sketch of the deve10pment of the three steel plants as well as the economic and political factors leading to their deveIOpment. Relevance to other underdeveloped countries will be discussed. The third chapter will discuss the various invest- ment criteria that have been suggested in the literature. The use of capital-output ratios to examine and appraise the three steel plants will be defended. The fourth chapter will discuss the technical and empirical problems involved in the application of the tech- niques used in analyzing the available data. The remaining chapters will show the results of the calculations and comparisons of these capital-output ratios. The capital-output ratios will be compared over time, and the implications of the trends of various formulations of the capital-output ratio will be discussed. Interfirm comparisons of the capital-output ratio will be made in order to determine the nature of the relationship between size of firm and degree of development of the economy with the capital-output ratio. The capital-output ratio of each firm will then be compared with hypothetical alternatives based on the incremental capital-output ratios of the economy and its manufacturing sector. This will be done in an attempt to judge the "extravagance" of the use of capital 3 in the steel industry. Finally, the price of steel produced in the domestic plant will be compared with the price of imported steel less duties. Along with this there will be a discussion of the protective devices used to insulate the Operations of the steel plant from international compe- tition. It is hoped that through these comparisons a better picture can be gained of the costs of establishing a domes- tic steel plant. CHAPTER I STEEL AND ECONOMIC DEVELOPMENT Steel and the Developmental Process The infatuation of underdeveloped nations with the idea of a steel industry is not difficult to understand. For, after all, steel is almost indispensable to the devel- Opment of modern agriculture and industry. Consequently, there is a positive correlation between the level of devel- Opment of an economy and its consumption of steel. In fact, a study made for the Economic Commission for Europe indicates that in the early stages of a country's economic development large increases in the per capita steel con- sumption correspond to very small changes in gross national product.1 Furthermore, the general pattern of steel utili- zation is such that during the early stages of development almost 90 per cent of steel consumption takes place in what might roughly be described as the investment sector: con- struction, production equipment, and transportation. Later 1However, as development proceeds, the average percentage increase in steel consumption associated with a change in gross national product declines fairly rapidly. United Nations, Department of Economic Affairs, Economic Commission for EurOpe, Lon Term.Trends and Problems of the EuroEean Steel Industry (ST7ECE7STEEL7I, IprII, I960) eneya, , p. . 5 in the deve10pmental process, as absolute steel consumption increases, the share of steel used in investment declines and that in consumption increases.l Thus, the eagerness of underdeveIOped nations to assure themselves a steel supply is not difficult to under- stand. The possibility that steel will not be available can be a serious developmental bottleneck. In peacetime, the key factor in determining whether a country can import sufficient supplies of steel is the availability of foreign exchange. Thus, the following alternatives would be available for obtaining steel: A. Foreign exchange earnings may be increased by beginning production of a new export commodity or by expanding output of present exports . B. The foreign exchange may be increased by cre- ation of industries competing with imports or increasing production in such industries that already exist. C. Domestic steel production may be undertaken. The domestic production of steel may be started at several levels: 1. The manufacture of pig iron from ore-- establishment of a fully integrated steel industry. 2. The manufacture of steel from imported pig iron and scrap--a semi-integrated plant. 3. The manufacture of finished steel products llbid. 6 from imported semifinished products (bars, billets, and slabs) or ingots--nonintegrated plant. A fourth alternative, increasing foreign exchange availabilities by general restriction of imports, is also possible. However, discussion of this alternative will be restricted. First, this policy by itself does not directly involve investment, and therefore, cannot be appraised on the same basis as the other alternatives. In actual prac- tice, however, import restriction is used in combination with a policy of import substitution (domestic production of the commodity whose importation is being limited). In this case, the analytical emphasis should be placed on the selection of the commodity whose production is being encouraged by the import restriction; the restriction of imports being viewed as a form of subsidy. The primary interest of this paper is in the investment decision itself, not the techniques used in supporting the enterprise. Second, import restriction does not seem to offer a promising solution to the problem. As Higgins points out: In most underdevelOped countries, finding foreign exchange for an ambitious development program by cutting other imports would require inroads into the semi- necessity or necessity category, with accounpanying sacrifices of the masses of the people.1 This generalization would also seem to hold if "substan- tially increasing imports of any group of goods" was substi- tuted for "an ambitious deveIOpment program." And, to the lBenjamin Higgins, Economic Develo ent (New York: W. W. Norton and Company, 1959), p. 575. 7 extent that it does not hold, the possibility of increasing steel imports in this way is limited. A policy of import restriction as an inducement to domestic production carries with it the danger of ignoring what may be an important function of imports. As Hirschman has put it: The advocates of protection and industrialization . . . were probably far too intent on blaming imports for the economic backwardness of their countries to recognize that meorts fulfill the very important function of demand formation and demand reconnaissance for the country's entrepreneurs.1 Import restriction may thus lead to import substi- tute production "too early" which may be relatively costly and inefficient. As Nurkse points out: The market for an imported commodity, small as it is to start with, becomes even smaller in real volume as the price to the consumer increases. The initial effect on real income is bad andeay very well lead to a fall in domestic savings. . . . The 2import restrictions should therefore be used sparingly.2 Furthermore, a policy of import restriction is inflationary in that it reduces the flow of goods while releasing funds for spending at the same time.3 1A1bert o. Hirschman, The Strate of Economic De- velopment (New Haven: Yale UnIversIty Press, I958), p. I23. 2Ragnar Nurkse, ”Trade Theory and Development Policy,” Economic Develo nt for Latin America, ed. Howard S. Ellis and Henry C. Wa IIIcH (New YorE: St. Martins Press, Inc., 1961), p. 25h 3See Ragnar Nurkse, "Some International Aspects of the Problem.of Economic Development, The Economics of Underdevelo ent, ed. A. N..Agarwa1a ana S. P. SIEEEI' (LoEEon: Oggora‘University Press, 1958), pp. 267- 68. 8 Judging by the deveIOpment of Latin American steel industries and those of several other underdeveIOped countries, it is obvious that the first two alternatives have been combined with.the third--establishment of a domestic steel industry. And this has generally involved establishment of a fully integrated plant. While the solution of the steel problem.does not necessarily require the domestic production of steel, certain factors have acted to bias the decisionemaking process in favor of establishing a domestic steel industry. The more important of these factors are: 1. Balance-of-payments difficulties. 2. The symbolic role of steel. 3. Recurrent steel shortages. h. Dissatisfaction with the develOpmental stimulus of domestic iron ore deposits. Motives for Establishing Steel Plants Balance-of—Payments Difficulties Balance-of-payments difficulties have been an important factor in orienting investment toward import- substitution.1 ‘Many of the arguments used against the development of imports in order to earn.the foreign exchange necessary for steel imports are based on past lflmitations in the capacity to import and on anticipated deterioration of 1Hollis B. Chenery, "Comparative Advantage and DeveIOpment Policy ' American Economic Review, LI, No. 1 (March, 1961), its-Em the terms of trade. In their study of the iron and steel industry of Latin America, the Economic Commission for Latin.America reported: Capacity to import . . . seems to establish specific limits in each country above which iron and steel imports have risen only occasionally. No matter how strong the pressure of internal demand for greater quantities of steel products, there have been no available means to purchase them abroad. Concern over the alleged deterioration of the terms of trade has pervaded the works of the Economic Commission for Latin.America and the writings of its influential Executive Secretary, Dr. Raul Prebisch.2 The view of the Commission is that because of the long-run tendency of the terms of trade to run against countries exporting raw materials and foodstuffs, the possibility of promoting development by continued expansion of export of primary materials is ltmited. Their analysis attributes this deterioration in the terms of trade to the fact that the gains of technological progress in developed countries 1United Nations, Department of Economic Affairs, A Stud of the Iron and Steel Industr in Latin.America, I . ev. , ovem er , vo s.; ew ork, 195h), 83. Hereafter cited as A Stud of the Iron and Steel Industr . This work is actually a report of a coHTerence e Bogota, Colombia, sponsored by the Economic Com- mission for Latin.America and the Technical Assistance Administration for the purpose of promoting the development of iron.and steel industries in Latin.America and, there- fore, can hardly be considered an unbiased source. 23cc Raul Prebisch, "Commercial Policy in the Under- develOped Countries," American Economic Review Papgrs and WOOOOng, nu, NOe y, g "’ e 10 remained in the hands of producers, whereas in primary pro- ducing countries the gains of technological improvement took the form of lower prices.1 The Prebisch-Singer thesis, as this argument is usually called, has come under heavy attack on both empiri- cal and theoretical grounds.2 However, regardless of the merits of this thesis, it does not constitute a justifi- cation for the establishment of a steel industry. In analyzing the reasoning behind the decision to establish a domestic steel plant, it is important to differ- entiate between the arguments for industrialization in general and those for establishing a steel industry in particular. While acceptance of the Prebisch-Singer thesis would result in industrialization as a general policy goal, it might not necessarily support the establishment of a steel industry. A policy of industrialization may include production of any range of manufactured import-substitutes. A policy of industrialization could even be based on the development of industrial exports . In the circumstances in which the less developed countries find themselves today, the question may be raised why, instead of promoting output of imports- substitutes, they do not promote manufacturing for 1Hans Singer, "The Distribution of Gains Between Investing and Borrowing Countries,” American Economic Review Papers and Proceedings, XL, No. 2 (May, , - . 2See, for instance, Gottfried Haberler, "Terms of Trade and Economic DeveIOpment," in Ellis and Wallich, Economic Develoflent for Latin America, pp. 275-307. 11 export so as to benefit from expansion of markets and trade among the more advanced countries. This is indeed possible.1 As far as export of primary goods is concerned, even if the terms of trade for the exporting nations are declin- ing, such a deveIOpment does not necessarily result in a dimunition of economic welfare. Using Haberler's example, if productivity in the production of an export good has risen by 10 per cent, so that the same factor input makes a 10 per cent larger output, or the same output can be produced by 10 per cent smaller input, and if the prices of exports in terms of imports have fallen by less than 10 per cent, the country is still better off.2 However, it is clear that wide fluctuations in the demand for exports of primary commodities, such as have been experienced in some Latin American countries, can have a disrupting effect on a development program by causing shortages in essential imports. This problem may be more prOperly solved by international commodity stabilization. Thus, it can only be concluded that balance-of-payments difficulties cannot be used as the sole justification for establishing a steel industry when the supply of foreign exchange earned, and, therefore, the supply of steel may be increased by any of a wide variety of investments. 256 1Nurkse, "Trade Theory and Development Policy," p. I n ZHabzrler, "Terms of Trade and Economic DevelOp- ment, p. 27 . 12 Investment in a steel plant must be viewed as one of an array of alternative investments, and judged only in.the light of alternative investments. The Symbolic Role of Steel Steel seems to have an almost mystical attraction for the countries of the underdevelOped world. The basis for this attraction can be easily explained. According to some writers of the "goo-political" school, steel is the cornerstone of national power. In his World Resources and Industries, Erich W. Zimmermann.writes that the core of mechanization and national power is a "coal-iron complex" which can produce large quantities of steel. He goes so far as to say that "without steel all else is meaningless, 1 ineffective, and inarticulate." Along this same line, Walter Voskuil has written: Iron and its alloys hold the central position in our mbdern economy. Iron is the key with.which man multiplfies his power to produce and increase his wealth. It is almost certain that writings linking the deve10pment of iron.and steel industries to the progress of 3 civilization itself have created an image of steel as being lErich‘W. Zimmermann, World Resources and Industries (rev. ed.; New York: Harper an o are, , p. . 2Wa1ter H. Voskuil, Minerals in World Industr (New York: McGrawbHill Book Company, nc., , p. . 3Charles W; Merillfi "The Significance of Mineral Industries in the Economy, Economics of Mineral Industries, ed. Edward H. Robie (New‘YOrE: KierIcan Institute of MInIfig, Metallurgical and Petroleum.Engineers, Inc., 1959), p. l. 13 the primary requisite of economic deveIOpment. The impor- tance of this industry has thus become magnified to the point where William A. Haven, vice-president of Arthur McKee and Company, a firm.that designs steel plants, could write that the production of iron and steel within any country is the most requisite of all so-ca1133_HE§ic industries as a solid and permanent foundation for the deve10pment of manufacturing, transportation facilities and agriculture and can be justified at almost any cost.1 An iron and steel industry is thought of as con- ferring political power and prestige on a country. "It gives a sense of independence from, almost equality with, other more developed countries."2 Thus, the steel industry plays a symbolic role in the deve10pment process of under- deveIOped countries. .Anderson and White wrote the following in reference to Colombia's recently completed Paz del Rio steel plant: This enterprise is expected to serve more than the usual economic ends. The basic function of the ordinary steel plant is to make a profit. Paz del Rio must lead a nation toward self-sufficiency and incarnate a vision of progress. . . . It is not fair to evaluate Paz del Rio exclusively on its economic performance. The national iron and steel industry also has a symbolic role to play in Colombian life.3 1WilliamA. Haven, "Selection of Steelmaking Processes and of Locations for Integrated Iron and Steel Works," A Stud of the Iron and Steel Industry, 11. 358. (Italics—53337) 2Norman J. G. Pounds, The Geo ra of Iron and Steel (London: Hutchinson'UniversIty LIErary, I959), p. 62. 3C. Langdon White and Donald J..Anderson, "Steel and Symbolism at Paz del Rio," Inter-American Economic Affairs, v, No. 9 (Spring, 1956), 8A and 93. IL This symbolic role of a steel industry has undoubt- edly been a factor in enhancing its attractiveness to under- develOped countries. It is, however, rather extreme to announce that a steel plant can be justified "at almost any cost." Just what are the costs to the economy of this symbol? An inefficient plant producing expensive tariff protected steel can turn out to be a symbol of futility and a deterrent to domestic steel using investments. The attraction of steel may also have been.enhanced by the Soviet experience in economic development. The success of Soviet planning methods in inducing rapid eco- nomic growth.and their stress on self-sufficiency in basic industrial commodities have been an influence on India's five-year plans. These plans omit any criteria of efficiency or comparative advantage for the goal of cre- ating heavy industry.1 Shortages of Steel Another frequently mentioned reason for the neces- sity for domestic steel production is that shortages of steel have been encountered at critical phases of economic growth. For example, the Economic Commission for Latin America reports that analysis of steel consumption in.Argentina, Brazil, Chile, Colombia, and.Mexico shows that during most of the past twenty-five years aLmost all of these countries h 1Chenery,American.Economic Review, LI, No. 1, h2- 3. 15 were unable to obtain the steel they required.1 The evidence usually cited as proof of steel shortages involves the allegation that steel consumption has not "kept pace with some of the factors which should influence its demand," e.g., real income, investment, or manufacturing output . 2 The use of the "steel shortage" argument may be criticized on two grounds. First, failure of steel con- sumption to keep up with other indicators of economic growth does not necessarily indicate a steel shortage. It may mean that the deveIOpment of the economy is following a pattern which.does not require as much steel as in the case of some other developmental patterns. DevelOpment of the food, textile, chemical, or electronics industries, for instance, would not require as much steel as the develOpment of machinery or automotive industries. The former pattern of industrial deveIOpment may be a natural adaptation to the country's resource base (and comparative advantage) rather than a distortion caused by lack of steel. Second, this argument again ignores the fact that steel shortages are only a symptom of a shortage of foreign exchange. And, as such, the arguments given above regarding the balance-of—payments difficulties apply. Establishment 1United Nations, Department of Economic Affairs, Economic Commission for Latin America, Iron and Steel Trans- form. Industries in Selected Latin.American Countr es (E7CN.I27377, August, I955) (Bogota, I955), p. 2. 2A Study of the Iron and Steel Industry, I, 59. 16 of a domestic steel industry cannot be justified by the fact that there are chronic or periodic shortages of foreign exchange. There is the possibility, however, that the short- ages of steel may at times be unrelated to the availability of foreign exchange. During wartime or in other periods when steel is in tight supply in the world markets, it might be difficult to obtain steel even if foreign exchange were available. The shortages of steel which were experienced during World War II were an important factor in the decision of some countries to promote domestic production of steel.1 If domestic industry requires a definite supply of steel, then greater weight must be given to greater domestic pro- duction in the investment decision calculus (if it can be shown that domestic production would ultimately result in greater regularity of supply). There is some question, however, as to how mmch.weight this factor should be given. What is the likelihood of occurrences like World War II or the Korean War resulting again in an international steel market so tight that underdeveloped nations cannot obtain critical supplies? Iron Ore Deposits and their DeveIOpmental Effects The presence of substantial deposits of iron ore in some underdeveIOped countries has undoubtedly influenced the 1United Nations, Department of Economic Affairs, Economic Commission for Latin America, Economic Surve of Latin.America, 1948 (E/CN.12/82) (New YorE, I9H9), p. IO. l7 develOpment of domestic steel industries. The underde- veloped areas of the world are richly endowed with iron ore deposits, possessing over half the world's ore. Unfortu- nately, the abundance of ore is not matched by comparable resources of fuel.1 The presence of a large mineral deposit in an under- developed country results in pressure on the government to insure maximization of its developmental potential. If the deposit is not being mined, the pressure is toward finding a market for the ore, domestic or foreign, and getting production started. If the ore is being mined, the pressure is toward controlling the use of the ore for whatever is thought to be the greatest gain for the country. In either case, the pressure is great for the establishment of a domestic ore-utilizing industry. The major source of this pressure is dissatisfaction with the deve10pmenta1 effects of an exporting.mining industry. This dissatisfaction centers around the magnitude and direction of the linkage effects of mining, especially when comparing processing and finishing industries. As Hirschman explains: The grudge against what has become known as the enclave type of deveIOpment is due to this ability of primary products from mines, wells and plantations to slip out of a country without leaving much of a trace on the rest of the economy.2 1Pounds, Geography of Iron and Steel, p. 168. ZHirschman, Strategy of Economic Development, p. 110. 18 The linkage effects of mining, beside being very small, are very largely forward. That is, the relationship of mining to the other sectors of the economy results from sales to these sectors rather than from purchases from them.1 Beside the fact that forward linkages are less of a stimulus to further deveIOpment than backward linkages, in the case of enclave mining industries the forward linkages are largely to the economy of the developed (importing) country.2 Marvin Bernstein has described an example of this type of situation in Venezuela: Venezuelan iron ore . . . has become part of the United States steel industrial complex which extends from hiring geologists to selling fabricated steel to con- sumers. Such raw material sources become ”economic enclaves"--portions of the local economy in reality attached to and integrated.with a foreign industry. In essence they are extensions of the economy of the advanced nation; embedded in the territory of the less advanced nation. They are resources located in.but not part of the economy of the underdeveloped nation.3 Because of the vertical integration of the iron mines into metal producing complexes, the rate of exploi- tation of the deposits is tied to the rate of production in.the advanced country. It is strongly believed, however, that for the greatest possible long-run gain to the under- deveIOped nation, the rate of exploitation should be 11tid., p. 107. 2Ihid., p. 116. 3Marvin D. Bernstein, "Foreign Investments and Mineral Resources: A.Few Observations," Inter-American Economic.Affairs, XIII, No. 3 (September, I959), HE. This type of argument is more often associated with H. W. Singer fiAmerican.Economic Review Papers and Proceedipgs, XL, No. 2, l9 coordinated with the country's general economic develOpment. There is the expressed fear that . . . irrespective of its potential ore resources a feeder mine may be developed no more at a given stage than is warranted by the capacity of the plant which it has been designed to serve. Conversely, a very close link between the intake of a particular smelter or refinery and the output of a particular mine may result in a rate of extraction which differs significantly from the Optimum. 1 This is the type of fear which leads to a policy of close control over mineral exports, or even to government ownership. In the case of iron ore, such a policy almost invariably means that if the deposit is to be exploited at all, it will have to be exploited by domestic processors. Integrated steel producers are reluctant to deveIOp ore deposits which they cannot fully control. In countries where ore deposits are nationally owned or nationalization seems imminent, exploitation has largely depended on domes- tic industry, as, for example, in Brazil andMexico.2 There is also the possibility that deposits of ore may be relatively small and scattered, as in the case of Mexico. The deposits of ore may also be of such.quality as to make them.unsuitable for develOpment.3 DevelOpment of 1United Nations, Economic and Social Councilfl Non- , ferrous Metals in'Underdevelo d Countries (ST/ECA/3 OctOBer, I955) (New YOPE, 195E): P- II;- 2U.S., Tariff Commission, Iron Ore: Re ort on Investigation No. 2; Under Section I95) Tarzzg got 0: 1930 as on, D. .: e ommission, , p. . 3U.S., Department of Interior; Bureau of Mines and Geological Survey, Materials Surve , Iron Ore" (Washington, D.C.: U.S. Government Printing 0* ice, 1956), p. V-83. Hereafter cited as Materials Survey. 20 iron ore deposits requires Operations on a larger scale than.most minerals. Large steel producers usually attempt to assure themselves of a long-term supply of uniform ore of specific characteristics for their furnaces. Uniformity is Obtained by procuring shipments of ore from certain districts and, therefore, small deposits of ore are not of any interest to them.1 There is naturally a great temp- tation to exploit these deposits domestically. The presence of iron ore deposits whether mined by foreign or domestic interests, or deposits not being mined at all, may lead to an underestimation of the benefits of mining as compared to the potential benefits of a domestic steel industry. This may be due, in part, to the tendency to lump all the various products of mines, wells, forests, and farms into the category of "primary products" and then ignore individual differences. Surely, the prospects for export of salt.mackerel or coffee or mahogany are not comparable with.the prospects for iron ore.2 It takes vast quantities of coking coal and fluxes as well as iron ore to produce steel. But the economic production of steel requires more than a combination of lUnited Nations, Department of Economic Affairs, World Iron Ore Resources and their Utilization (New YOrk, , p. O Zln fact, the projected export market for this commodity in.the'United States, at least, seems rather favorable. See Lewis 0. Delwart, The Future of Latin .Mmerican Exoorts to the United Sta es: ’- and 1' 0 ”as: a; on, i. .: ‘a one ‘ ann ng ‘ssoc a on, '60), pp. 20-21. 21 those elements which result in the steel-producing chemical reactions. It requires a potential steel market large enough to permit the building of a plant of sufficient size to overcome the diseconomies of small-scale production. Even if all these elements are present, establishment of a steel industry may not represent the best possible utili- zation of scarce capital resources. (Although admittedly in this case, the chances are far better than if any of the essentials are missing.) This can be determined only by a systematic evaluation of investment alternatives. In this chapter some of the factors which have led underdeveIOped nations to establish steel industries have been discussed and analyzed. It has been demonstrated that the usual arguments used to support this decision are not by themselves convincing, and that they may lead to a serious misallocation of resources. Given the confusion between shortages of steel and shortages of foreign exchange, dis- satisfaction with deveIOpmental effects of mining ore, and the exaggeration of the importance of steel, thereis the distinct possibility that various import substitutes and exports will not be compared. The result may be investment in a project with considerably lower productivity than an available alternative. In the next chapter an analysis will be made of the various methods or criteria suggested in the literature for the allocation of investment. There will be a discussion of each general approach to the allocation process and an 22 attempt to appraise the applicability and practicality of each when trying to evaluate the advisability of investment in a steel industry. CHAPTER II THREE LATIN AMERICAN STEEL PLANTS: ESTABLISHMENT AND DEVELOPMENT The three steel plants with which this study is concerned share, to some extent, a common background. All three are products of what might be called "economic nation- alism"--the desire to achieve or preserve independence as against foreign economic dominance, whether real or imagined. There is also the hOpe of achieving status or prestige in the world economic community. The building and the planning of each of these plants was an early manifestation of the turning toward industrialization as a way out of the vicious cycle of economic underdeveIOpment. The financing of each of these plants was carried on with the assistance of a develOped country, in this, case the United States. And in each case, the motives of the devel- Oped country can be said to be that of serving its own interests in a critical period of tension and world strife. Each of the countries involved in the study was at approxi- mately the same stage of develOpment at the time of the building of the steel plant. Although the economic charac- teristics of the three nations differed considerably, each was concerned with "steel shortages"--this was characterized 23 21.1 by low steel consumption per capita, dependence on imports for most of their steel supplies, and chronic shortages of foreign exchange needed to pay for steel imports. These characteristics are common not only to the three countries discussed in this study, but they prevail throughout the underdeveIOped world where steel has become a symbol of economic progress. In order to understand the political and economic conditions influencing the decision to establish a domestic steel industry, it is helpful to consider in greater detail the Brazilian, Chilean, and Mexican cases . Moment of the Companhia Siderurgica Nacional The modern era of Brazilian industrial develOpment began about 1890 with the abolishment of slavery, making possible the develOpment of an internal market, and the beginning of prosperous times for the coffee trade. These develOpments were accompanied by an increase in immigration of European workers, bringing their skills and capital. By 19111 Brazilian industry was producing about 300 million dollars per year. World War I saw a sharp increase in Brazilian industrial production to an estimated 690 million dollars by 1918.1 Between World Wars I and II, the upswing continued, spurred on by an unprecedented prosperity in the coffee industry. Between 1920 and 1937, the number of 1Re ort of the Joint Brazilian-United States Tech- nical ConfissIon (RIO 3e Janefio, I9E9), p. 89. 25 establishments increased by over 35 per cent, and by 1938 industry passed agriculture as a source Of national income.1 ' It should not be assumed that this indicates that Brazil giggle 1938 was anything resembling a highly developed economy. Actually, in 1939 only 10 per cent of the pOpu- lation was engaged in manufacturing. Textiles, foodstuffs, and clothing accounted for nearly 60 per cent of the value of production.2 It would seem that until 1930 the industrial sector grew in spite of rather than because of government activity. Economic policy was marked by an official disregard of industry, transport and power and by support of the coffee trade.3 In 1930, the dictator Getulio Vargas came to power in the wake of world-wide depression. This marked the beginning of an economic transformation as the traditional policy of coffee protection at the expense of industry began to be replaced.“ The Vargas regime stressed national unity, and this had its corollary in a vigorous program of economic nationalism. The establishment of Companhia Siderurgica lCorwin Edwards, ”Brazil's Economy in the War and After," Economic Problems of Latin America, ed. Seymour E. Harris (New YorE: McCraw-HIII BOOE Company, Inc., 191414), p. 278. 23c ort of the Joint Brazilian-United States Tech- nical ComiIssIon, p. 33. 3John J. Johnson, Political Cha e in Latin America (Stanford: Stanford UniversIt? Press, I958), pp. IKE-63. hJose Jobim, Brazil in the Mak_il_1_g (New York: Mac- millan and Company, 1 a PP- - . 26 Nacional was undoubtedly a manifestation of that national- ism.1 The founding of Companhia Siderurgica Nacional, to be referred to as CSN, was preceded by many years of devel- Opment of the Brazilian iron and steel industry. In fact, Brazil has a long metallurgical tradition. The rich ores of Minas Gerais have been exploited for.many years. Pig iron has been produced using charcoal fuel on a fairly large scale since the 1920's, and small furnaces have been Operated since colonial times. In 1860, a Belgian.named.Monlevade started the first steel mill and rolling mill in.Minas Gerais. It was the beginning of the Companhia Siderurgica Belgo Mineria, which until CSN was the largest steel producer in Brazil. In 1925, the Companhia Brasileira de Usinas Meta- lurgicas was organized. By 1930, Brazil had eleven blast furnaces with a total capacity of 100,000 tons of pig iron per year. Actual output, however, was only 36,000 tons. Between 1930 and 1939, eight more furnaces were built, increasing annual capacity of pig iron to 200,000 tons, and actual output to 160,000 tons. By 1939, consumption of steel in Brazil was hh2,000 tons per year, with domestic production accounting for llh,000 tons.2 At that time the lHenry'W. Spiegel, The Brazilian Econom (Philadel- phia: The Blakiston Company, I9E5), p. II. 2United Nations, Department of Economic Affairs, Economic Commission for Latin America, Economic Surve of Latin America119g9 (E/CN.12/16h/Rev. 1) (New YorE, I95I), p. 255. 27 entire iron output was produced in charcoal furnaces. In early 1939, the Brazilian government took the first steps toward expanding the steel industry. At that time the government had planned for the establishment of a mill to be jointly owned by the government and the United States Steel Corporation. The company gave tentative approval but later withdrew in the light of the world situation.1 In March of 19110, Vargas appointed a commission to submit a plan for the establishment of a mill to produce rails, structural steel, and plates. The plan submitted by the commission was approved by Decree No. 3002 of January 30, l9hl, thus creating the National Steel Company. In July, 19110, Oswaldo Aranha, the Brazilian foreign minister, wrote a letter to Warren Lee Pierson, President of the Export-Import Bank. Its purpose was to present Brazil's final plans for the establishment of its national steel- making industry. Some weeks later the Export-Import Bank decided to aid Brazil in the establishment of the steel plant. This was the first instance of the financing of an inportant industrial project abroad by the Bank.2 The amount of dollar exchange necessary for pur- chases in the United States was estimated at 25 million dollars during the early stages of planning. The cruzeiro 1United Nations, Department of Economic Affairs, Economic Develognent in Selected Countries (New York, 19117), p. O 2Export-Import Bank, Case History: Volta Redonda. (Mimeographed . ) 28 equivalent for internal costs was estimated at 35 million dollars. The government of Brazil, through the Banco do Brasil, decided to guarantee the Export-Import Bank loan, and to assure full sale of domestically issued stock by purchase, if necessary. CSN issued 35 million dollars of preferred and common stock. Seventy per cent Of this was ultimately purchased by the government.1 Subsequent study indicated that the initial foreign exchange requirements were underestimated, and that the purchase of steel mill equipment in the United States would cost AS million dollars rather than 25 million dollars as originally expected. Thus, on June 19, l9h0, the Export-Import Bank authorized a loan, NO. 269, of NS million dollars to the CSN. The loan was granted at h per cent interest to be repaid in 36 equal semiannual payments beginning on October 1, l9h7, and termi- nating April 1, 1965.2 The Operation of the preposed steel mill would have been impossible without first assuring a supply of high quality iron ore. To this end, the Companhia Vale do Rio Doce was organized in l9u2 to Operate the Itabira iron ore deposits and.the Vitoria-Minas railway. In.March.of that year, an agreement was signed between the British government, lErnest H. Vaughn, "A Study of the DevelOpment of National Steel Plants in Selected Latin American Countries" (unpublished Pth. dissertation, University of Texas, 1955), p. 25. This is a study in economic geography more directly concerned with locational problems. ZIbid. 29 the Export-Import Bank, and the Metals Reserve Company. Under this agreement, the British government acquired control of the Itabira Iron Ore Company, Ltd., which had extensive claims in the area, then coded the prOperties to the Brazilian government. The Export-Import Bank extended a credit of 1h million dollars for construction of a railway from.the ore deposits to the Port of Vitoria. The Companhia Vale do Rio Doce was organized to run the mines and railway. Its initial capital was 10 million dollars, 55 per cent of 1 Between which was subscribed by the Brazilian government. l9h2 and 1950, the Export-Import Bank lent the Companhia Vale do Rio Doce 26.5 million dollars for improvement of facilities and expansion of mining Operations. The site chosen for the plant was the small plan- tation town of Volta Redonda about 65 miles northwest of Rio de Janeiro. The steel plant has since borne the name of that town. TranSportation seemm to have been the major factor in locating the plant. The coal deposits of Brazil are in the State of Santa Catarina, more than 1,000 miles to the southwest of the ore fields of Minas Gerais. It would, therefore, be necessary either to locate the plant in Santa Catarina and haul the ore to it, haul coal to Minas Gerais, or select a logical intermediate location. The attempt to settle on a location occasioned considerable bickering. The final site being inland and 65 rail miles from the principal 1GeorgeWythe, Industr in Latin America (New YOrk: Columbia University Press, , p. . 30 market at Rio de Janeiro seems less than ideal. It requires a 250-mile rail haul from.Lafaiete to the north for ore and over 500 miles by a rail-water-rail route from.the Barro Branco coal deposits in Santa Catarina.1 Construction of the plant began in late l9hl. It was expected that production would begin in l9hh. The tech- nical and economic planning of the plant and its construction were carried out with the assistance of the American firms Hoppers Company, Inc., and Arthur G. McKee and Company.2 Completion of the plant took longer than had been antici- pated and the rolling facilities were not completed until late 19(17.3 The installation at Volta Redonda was built with a view toward future expansion. However, during the first years of Operation the plant Operated at considerably less than capacity. Inability to obtain.maximum.output during the first few years of production.was attributed to supply problems including shortages of raw materials, caused by insufficient rolling stock and electricity, and to break- downs in the blast furnace. On the demand side a ”lack of consumer familiarity with domestically produced products" 1WilliamA. Haven, ”Selection of Steelmaking Processes and Locations for Integrated Iron and Steel Works," A Study of the Iron and Steel Industry, II, 35h. ZExport-Import Bank, Case History: Volta Redonda, p. l. 3Memorandumto the Board of Directors of the Ex-ort- Immort Ban u 1 = ' ' 'e: 'pp, ca on 30. ' o ‘upp emen re- ’0. .', p. . 31 is cited.1 In 1950, the CSN began the first step in a two- phase expansion program to increase production of Volta Redonda to 1 million tons of rolled products yearly by 1960. Also during this year the fourth Open-hearth furnace was installed. The first phase of the expansion included: installation of two new steel furnaces bringing the total to six, installation of a second blast furnace with a capacity of 1,300 tons per day, addition of 21 new coke ovens, new hot and cold rolling mills, a new finishing mill for tin plate, and electrolytic tinning equipment. This was to increase production of pig iron to 700,000 tons by 19511.2 In order to finance this expansion, a second loan (No. h81) for 25 million dollars was Obtained from.the Export-Import Bank on July 20, 1950. The loan was to be repaid in 36 semiannual payments, between.August, 1955, and February, 1973. The equivalent of 25 million dollars in cruzeiros was raised in Brazil to pay for all domestic costs of production.3 In.February, 19Sh, the second blast furnace at Volta Redonda was blown in, completing the first stage of the Plan 10.8., Department of Interior, Mineral Trade Notes, 2Walter Godfrey, Economic and Commercial Conditions in Brazil (London: Her Ma es y s a onery ce, , 57W— 3Vaughn, "A Study of the DevelOpment of National Steel Plants in Selected Latin American Countries," p. 25. 32 de Milhao. To begin the second stage of expansion, the government increased the capital of the Companhia Siderur- gica Nacional by 500 million cruzeiros, and began negoti- ations with the Export-Import Bank.for a loan of 35 million dollars.1 This loan was granted to the company on February 1, 1956 (Loan No. 770). This phase of the expansion program would bring ingot capacity up to the goal of 1.25 million tons per year, and rolled product capacity to 1 million tons. Completion of this phase of expansion was expected in 1961. The total cost of the expansion was estimated at 37 million dollars plus 2 billion cruzeiros.2 In l9h7, CSN's first full year of production, the company produced 37.5 of the nation's pig iron, 37.9 per cent of its steel ingot, and 31.8 per cent of its rolled products. During the period of 1955 through 1958, the company produced, on the average, h8.2 per cent of the pig iron, 58.2 per cent of the steel ingot, and 5h.8 per cent of the rolled products made in Brazil (see Table 1). While production of rolled products at Volta Redonda increased by 2&6 per cent, production of rolled steel by other producers in Brazil increased by 110 per cent. Increases in the production of steel were, therefore, not at all limited to 1United Nations, Department of Economic Affairs, Economic Commission for Latin America, Economic Surve of Latin America, 195A (E/CN.12/362/Rev. 1) (New YorE, I955), p. 136. 2U.S., Department of Interior, Mineral Trade Notes, In NO. 6 (June, 1960), IN. 33 Table 1.--Brazilian National Production of Pig Iron, Steel Ingot, and Rolled Products and Per Cent Accounted for by Companhia Siderurgica Nacional, 19h6-58 (thousands of tons) L _—=m— Pig Iron Steel Ingot Rolled Products National National National Year pro- Per cent pro- Per cent pro- Per cent duction CSN duction CSN duction CSN 19u6 370.7 25.8 3 2.6 2h.9 230.2 5. 19h? h80. 37.5 3 7.0 37.9 296.7 31. 19u8 551.8 no.1 t83.1 50.5 u03.5 A9.1 19h9 511.7 37.7 615.1 50.1 505.5 tho9 1950 729.0 h6.5 778.6 5h.0 623.3 h6.1 1951 776.2 hh-O 8h3.0 55.2 696.6 h9.2 1952 811.5 uu.2 893.3 53.3 gl9.h 50.1 1953 880.1 u2.1 1016.3 h7.5 1h.5 uu.6 195A 1088.9 M9.% 1188.3 51.2 9g0.8 h3.2 1955 1068.5 u6. 1162.5 57.3 9 2.1 52.2 1956 1152.9 118.0 13751; 53 .8 11111.8 50.7 1957 1251.7 50.7 1299.2 61.9 972.8 61.2 1958 1356.1 h7.8 1359.2 59.0 1125.3 55.3 Sources: Brazilian production--Instituto Brasileiro de Geo rafia Economia, 0 Brasil Em Numeros (Rio de Janeiro, 1960 , p. uh. Firmfis production--Companhia Siderurgica Nacional, Rela- torio do Directoria (Volta Redonda, l9h7-59). 3h those at Volta Redonda, although production there did increase at a faster rate. Production of pig iron, steel ingot, and finished products in the other firms comprising the Brazilian steel industry more than doubled as well (Table 1). Establishment of Compania de Acerogggl_Pacifico The primary instrument of economic nationalism.and the state mochanism.for the planning and promotion of eco- nomic develOpment in Chile is the Corporacion de Fomento de la Produccion (Chilean DevelOpment Corporation), more commonly referred to as Fomento or CORFO. The Compania de Acero del Pacifico, here to be called CAP, is clearly the handiwork of CORFO. CORFO was born in 1939 amidst the devastation of the catastrophic earthquake of that year. It was also during the great international depression. The combination of these two events seems to have resulted in the crystallization of the political force known as the Pepular Front. It was this party and its successor center- left coalitions that guided Chile's economic nationalism in the critical period of l9hO-h6. It was during these years that CAP was conceived.1 1Alvin Cohen, "Economic Change in Chile,” Latin American.Eono a , School of Inter-American Studies, University of filorida, II (Gainesville: University of Florida Press, 1960), 7-13, and Federico G. Gil, "Genesis and Modernization of Political Parties in Chile," Latin American.Mono ra hs, School of Inter-American Studies, University of PIorIda, XVIII (Gainesville: University of Florida Press, 1962), u3-h8. 35 Chile is a country rich in mineral deposits and particularly well endowed with iron ore. In addition to mineral deposits, Chile already had a rudimentary steel industry on.which to build. The E1 Tofo deposit, in the province of Coquimbo, was develOped for large-scale pro- duction in 1913 by Bethlehem.Chile Iron Ore Mines Corpo- rations; however, a small tonnage had been mined in 1910.1 As early as 1889, the Chilean government began investigating the possibilities for domestic iron and steel production. The initial project began in 1905, when the government granted the French firm, Hauts Fourneaux.Forges et Aciereries du Chili, a concession to exploit 5,000 hectares of national forest lands near Valdivia. The company planned to use the Prud'homme process, charging logs directly into the blast furnace. Plant construction was begun in 1908 and operations commenced in l9lh. The plant Operated occasion- ally, mostly on.an experimental basis, until 1919 without achieving commercial success. The plant was located at Corral Bay at the mouth of the Valdivia River.2 In 192k, the furnace was purchased by a private company, Compania Electra-Siderurgica e Industrial de Valdivia (ESVAL), and was set up with government assistance. In 1933, the furnace was blown in again, using charcoal 10.8., Department of Interior, Materials Survey (May, 1956), p. 1'78. ZDanilo Vucetich, "Operation of the Charcoal Blast Furnace at Corral Using Mixtures of Metallurgical Coke and Charcoal," A Study of the Iron and Steel Industry, II, 158. 36 instead of wood.1 The next year, the government bought the company, and a Siemens-Martin furnace and a rolling mill were installed. Over a period of twelve years (l933-hh), the blast furnace produced 83,000 tons of pig iron. In l9hh, it was replaced with a new one designed to use char- coal.2 Up until the inauguration of the Huachipato works of the CAP, the Corral plant was the only producer of pig iron in Chile. Prior to the Opening of the Huachipato plant of CAP, production of steel was limited to four firms: ESVAL (pre- viously mentioned), Laminacion y Fundicion, S.A. (Lamifun), Establecimientos Metalurgicas "Indac," S.A. (INDAC), and Fabrics y Materiale de Guera de Ejercito (FAMAE), now called Fabrics y Maestranzas de EJercito. During the period l9h0-hh, consumption averaged 3 about 125,000 tons per year. Per capita steel consumption averaged about 25 tons per capita. By l9u0, Chile had a nascent manufacturing industry employing almost 300,000 peeple or 17 per cent of the working force. This percentage was surpassed only by Argentina among the Latin.American nations.’4 1W‘ythe, Industry in Latin America, p. 159. 2Vucetich, "Operation of the Charcoal Blast Furnace at Corral . . ., p. 158. 3A Study of the Iron and Steel Industry, p. 85. “United Nations, Economic Survey of Latin America, 194 , Tables I and II. 37 Production averaged approximately 35,500 tons per year from.l9kS-h9, and consisted almost entirely Of rounds and bars. The total annual consumption during that period was 152,h00 tons per year.1 Within.this background, CORFO and the Chilean government undertook to develOp a steel industry. The Huachipato plant Of CAP represents the fruition of these efforts. The steel plant was one of the first projects undertaken by CORFO. According to a United Nations report, it was felt that the plant would rest on a sound economic basis because of the convenient location of ore, coal, and limestone. In fact, it was expected "that the average production cost of iron and steel products will be approximately ten per cent lower than the corresponding import price, excluding present customs duties," so that "no increase in customs duties is contemplated."2 As will be shown later, these predictions proved to be rather Optimistic. Plans for building an integrated steel plant at Huachipato were announced in 19MB, following an investi- gation and report by United States experts. CAP was founded for this purpose with a capital of 50 million pesos. One-half Of this amount was subscribed by 1U.S., Department Of Commerce, Investment in Chile (Washington, D.C.: U.S. Government Printing CITIce, 1930), pp 0 161‘ 62 e 2United Nations, Economic Development in Selected Countries, p. 112. (Ital cs m. e. 38 CORFO.1 The site selected for its plant was on San Vicente Bay, near Concepcion, two miles south of the port of Tales- huano. This site had the following advantages:2 1. Because Of the tidewater location, iron ore could be delivered 500 miles from El TOfO by ship. 2. The coal mines of Lots and Schwager were 25 miles south Of San Vicente, with rail and water transpor- tation available. 3. Hydroelectric power was obtainable from the Abanico hydroelectric plant on the Rio LOJa. h. Fresh water was provided by the near-by Rio Bio-Bio. 5. Satisfactory highway and rail transportation was available. 6. There was ample supply of labor from.Concepcion, Lots, and Talcahuano. In order to finance foreign exchange requirements, CORFO submitted its prOposals to the Export-Import Bank with a request for a loan. In September, 1985, the bank approved a loan of 28 million dollars (Loan No. 378, September 11, l9h9, to be repaid in no semiannual payments beginning June, 1951, and ending December, 1970).3 1Wythe, Industry in Latin.America, p. 159. 2c. Langdon White and R. H. Chilcote, "Chile's New Iron and Steel Industry," Economic Geography, XXXVII, NO. 3 3Vaughn, "A Study of the DevelOpment of National Steel Plants in Selected Latin.American Countries," p. lho. 39 It seems clear that the decision to make a loan to Chile, and, in fact, the entire relationship between the Export-Import Bank and CORFO was influenced by the tremen- dous pressure the United States was bringing to bear on Chile to declare war on the Axis. In fact, Chile received no aid from.the Export-Import Bank between 1939 and l9h3 when she finally broke relations with the Axis.1 A loan of 10 million dollars was secured from.the Bank of Chile, and capital Of 15 million dollars was sub- scribed (38 per'cent by CORFO). In accordance with one of the credit restrictions imposed by the Export-Import Bank, the company engaged the KOppers Company of Pittsburgh to supervise construction Operations. Construction.was started in.mid-19h7, and all sections were in Operation before the end of 1950. Facilities included a blast furnace, 57 coke ovens, a Bessemer converter, a merchant mill, a blooming mill, and a plate, tin, and sheet mill.2 By agreement with CORFO, CAP took over the adminis- tration of the ESVAL plant in 1950.3 The total capacity of the CAP Operation at the completion.of the first stage was 203,000 tons of pig iron and 236,h00 tons of ingot.h 1Cohen, "Economic Change in Chile," p. 15. 2White and Chilcote, Economic Geogggphy,.XXXVII, NO. 3, 262. 6 3U.S., Department of Commerce, Investment in Chile, p. l 2. #Export-Import Bank, Press Release NO. 8;. ho Shortly before the completion of the plant, blast furnace capacity was found to be greater than anticipated owing to high quality ore; it soon became evident that steel production facilities were inadequate to handle all the excess production Of the blast furnace. Therefore, an expansion program was planned which would permit the con- version of the entire blast furnace production into finished products. The total cost Of this program was estimated at 10 million dollars plus 271,000,000 peace.1 0n.August 8, 1951, the Export-Import Bank granted the company a loan Of 10 mdllion dollars at h per cent repayable over twenty years (No. 502). The domestic costs of expansion were met by CORFO.2 The program.was carried out gradually over the years 1951 to 195h. Completion of the second Bessemer con- verter and addition Of the new Open-hearth furnace raised ingot capacity tO 375,000 tons per year. ‘With expansion Of the steelmaking shOp, rolled product capacity had reached 227,000 tons per year by the end of 1955.3 In the first four years of Operations, 1951-Sh, production of pig iron increased 25.8 per cent, ingot prO- duction increased 86.3 per cent, and output Of finished 1Compania de Acero del Pacifico, Memoria Anual, 1951 (Santiago, 1952), p. 8. 2Vaughn, "A Study of the DevelOpment of National Steel Plants in Selected Latin American Countries,” p. 55. 3United Nations, Department of Economic Affairs, Economic Commission for Latin America, Economic Surve Of Latin.America, 1956 (E/CN.12/h27/Rev. 1’ (New YorE, 1557’, p. go. hl products increased 71.9 per cent. In 1955, however, prO- duction was curtailed. This was due to: (l) the closing down Of the blast furnace for relining; and (2) a 61-day shutdown of Operations caused by two successive strikes, first by miners and then by Huachipato workers.1 A study carried out by the CAP in 1955 indicated that by 1959 Chilean steel consumption would reach 3h5,000 tons Of rolled products per year. It was estimated this would require an ingot production Of about h30,000 tons.2 0n the basis of these estimates, the company prO- Jected a two-stage expansion plan for 1956-59. The first stage provided for improvement Of the plant's hot rolling Operation.which would result in an increase Of capacity from 256,000 tons per year to 291,000 tons per year. Although the increase in productibn was not substantial, this improvement was supposed to result in reduced costs and improved product quality.3 The cost or this phase of expansion was 10,163,000 dollars, 3,550,000 dollars Of this was obtained through an.Export-Import Bank loan in June Of 1956. The total cost of the second was estimated at 29 million dollars. The dollar requirements Of 16 million dollars were Obtained from.the Export-Import Bank in a loan 1Compania de Acero del Pacifico, Memoria Anual, 1956 (Santiago, 1957), p. 9. 2United Nations, Economic Survey of Latin America, 1956, p. 90. 3Compania de Acero del Pacifico, Memoria Anual, 1956-1951 (Santiago, 1957). p. 11. 82 granted in.February, 1957. The peso costs of the program were financed through a capital increase by the company.1 Phase I Of the expansion program.was completed in 1957; phase II was substantially completed in 1960. At that time, annual capacity had been raised to meet the goals set for the 1956-59 program-430,000 tons of ingot and 315,000 tone of rolled products. The company has already formulated an expansion program.for 1956-6h. This program.would result in an increased output Of rolled products to h35,h60 tons by l96h. The program.inc1udes installation Of a second blast furnace, 39 new coking ovens, and other miscellaneous alterations. The transition to the higher outputs would not involve any fundamental expansion Of steelmaking and recently installed rolling equipment. The latter are thought to be adequate to absorb the additional production from the second blast furnace.2 After the beginning Of Operations of the CAP plant in 1951, Chilean consumption Of steel increased very rapidly from a base Of 100,000 to 125,000 tons per year before 1951 to about 225,000 tons in 1955. However, since 1955, Chilean demand has stagnated and the company has had to turn to external markets to maintain output at a high 11bid. 2"Rolled Iron and Steel Products in Latin.America," Economic Bulletin for Latin America, IV, NO. 2 (October, ’ O 143 level.1 Today Chile is producing substantially all of its domestic consumption. Insofar as self-sufficiency was a goal Of the Chilean government, CAP is a success (except for the fact that the production of steel depends on foreign sources for coal). The pertinent question is then, what is the cost Of this self-sufficiency? Establishment of Altos Hornos de Mexico, S.A. The year 19h0 marked a change in the course of eco- nomic develOpment Of Mexico. The administration of Presi- dent Manuel Avila Camacho, which took Office that year, made industrialization the central feature of Mexican economic policy.2 The preceding Cardenas administration had encour- aged industrialization indirectly at best, Cardenas himself finding little place in his personal philosOphy for the building Of large industry. He was far more interested in agricultural reform and the building of rural agricultural cOOperative communities. The policies Of his administration reflected these interests. But beginning with the Camacho administration, a change began to take place; land reform was slowed down and industrialization became an immediate national goal.3 The formation Of Altos Hornos de Mexico was lnide, ppe 18‘190 2Sanford Mosk, Industrial Revolution in Mexico (Berkeley: University of CalifornIa Press, 1950), p. 31. 3Oscar Lewis, "Mexico Since Cardenas," Social C e (New YorE: in Latin America Toda , Richard R. Adams et a1. Vifitage Books, I931), p. 286. uh both a product and a reflection Of that change. Altos Hornos was not the first steel producing establishment in Mexico. In fact, it was not even the first fully integrated steel plant in Mexico. The first fully integrated steel firm, Compania Fundidora de Fierro y Acero, commonly referred to as the Fundidora, was founded in 1900 by a group of Mexico City and Monterrey businessmen. This firm was established during the first great period Of Mexi- can economic growth during the rule Of the dictator Diaz. Diaz' policy included programs to introduce new industries and to foster and protect domestic industry. The steel producing facilities of the Fundidora, constructed during his term of Office, was the largest iron and steel producer south of Birmingham, Alabama, until 1941.1 During this period, all of the pig iron produced in Mexico came from the furnaces of the Fundidora. Between 1910 and l9h0, production Of pig iron doubled but still remained at an extremely modest 90,000 tons per year. During the period 1925 to l9h0, Mexican steel consumption rose from 165,900 to 221,200 tons per year. In 1925, 37 per cent was produced domestically: by l9h0 this had risen to 57 per cent.2 The year l9h1 marked the beginning of a new era for 1Carlos Prieto, "La Industria Siderurgica," Mexico: Cincuenta Anos de Revolucion, Fondo de Cultura Econom.ca ex co, 0 O, , p. O 2A Study of the Iron and Steel Industgy, 1, 5h. AS the Mexican steel industry. In 19h2, the Fundidora began construction Of a second blast furnace and also began to expand its steelmaking department. In l9h3, Altos Hornos de Mexico, S.A. started construction.of a blast furnace in Monclova, Coahuila. And in 19h6, HoJalata y Lamina, S.A. began production Of flat steel and tin plate in.Monterrey.1 The.motivating forces leading to the establishment Of Altos Hornos may have been somewhat different from.those prevailing in some other underdeveIOped countries. Mexico, after all, already possesses a modern, fully integrated steel plant. However, a study of the Mexican steel market made in 19hl indicated that Of an estimated consumption Of 380,000 tons annually, approximately 250,000 tons were imported. Analysis of the composition Of steel imports during the years 1937 through 1981 revealed average yearly imports Of flat rolled products Of h9,500 tons.2 Up until l9hl there was virtually no production of flat rolled steel products in Mexico. Production in.the Monterrey plant consisted almost entirely Of structurals and railway supplies. It was originally decided to establish a rolling mill using imported hot-strip coils. However, fear of being at the mercy Of foreign.suppliers led to the 1mid. 2Harold R. Paps, "Five Years Of Achievement at Altos Hornos Steel Company," Basic Industries in Texas and lorthern.Mexico, Institute of Latin American Studies, Uni- versity of Texas (Austin: University of Texas Press, 1950). p. 51. It will be noted that these consumption figures are considerably higher than.the estimates cited on page uh. h6 decision to build a fully integrated plant. The site chosen was the town of Monclova. It was at the junction Of two railroads, one from the Sabinas coal basin to the north and the other from Chihuahua to the west. Substantial iron ore deposits are found in both Sabinas and Chihuahua. The site also provided an adequate supply of fresh water. In December of 19111, an agreement was reached between Mexican businessmen and bankers, the Nacional Financiera, and Armco International Corporation for formation Of the company and construction of the mill.1 The Nacional Financiera, a government formed develOpment bank, purchased 50 per cent of the stock and all Of the bonds Of the company.2 In return for its technical direction, patent rights and management services, Armco received roughly one-fourth Of the common and preferred stock of the company, plus a fee of one dollar for each ton shipped for the first fifteen years.3 Mexican bankers and industrialists purchased the remaining fourth Of the stock. The need for dollar exchange was met in 19112 when, through the Nacional Financiera, and with the guarantee Of the Mexican government, a loan of 8 million dollars was Obtained from the Export-Import Bank (Loan No. 323 at 11.5 per cent interest, repayable in 22 11bid. , pp. 51-53 . 2Vaughn, ”A Study Of the DevelOpment Of National Steel Plants in Selected Latin American Countries ," p. 1110. 3"Altos Hornos de Mexico, S.A.,” Revista de Econo- mia, V, HO. 8 (August, 19112), 8. h? semiannual payments from August, l9h6, through March, 1957) .1 The diplomatic implications of this loan and others made to Mexico by United States agencies has an.interesting background. The State Department of the United States was aware that Mexico was being drawn into cOOperation with the Axis powers. Germany, Italy, and Japan needed oil and Mexico's other Oil markets were cut Off because Of an inter- national embargo resulting from.the Mexican exprOpriation of foreign Oil properties in 1938. The United States attempted to bring about a settlement Of claims against Mexico and a reconciliation with her. In 19hl, a MexicansAmerican General Agreement was signed settling the payments for exprOpriated prOperty. In return for Mexican concessions, the United States pledged that long-term, low-interest loans would be made to rehabilitate the Mexican economy.2 The loan.for the construction of Altos Hornos was part Of this program, The basic units Of the plant consisted Of secondhand equipment. A used blast furnace was located near St. Louis, a scrap universal plate mill in Indiana, and a ladle was found in Pennsylvania. These were purchased, dismantled, and shipped by rail to Monolova. The first material for 1Vaughn, "A Study Of the DevelOpment of National Steel Plants in Selected Latin.American Countries," p. 1h3. 2Howard F. Cline, The United States and Mexico (Cambridge: Harvard University Press, I953), pp. ZEV-h9. h8 construction Of the furnace arrived in April, l9h2; con- struction began in October Of that year. Twenty months later, in June, 19hh, the blast furnace went into prO- duction; and, in October of that year, the first basic Openrhearth furnace and the universal plate mdll began production.1 The original plant consisted of a blast furnace of 200 metric tons per day output, two 65-ton Open-hearth furnaces, a universal hot mill, two cold rolling mills, three centrifugal cast iron pipe machines, three tinning lines, and the various necessary maintenance shOps.2 (The cold rolling mill was purchased new and did not begin prO- duction until February, 19116.)3 Expansion and improvement Of the plant started almost before production. Expansion necessitated two refinancing programs:h l. In.February, 19hh, the number Of preferred stock was increased, raising total capitalization from 22,310,000 pesos to h0,000,000 pesos. The bond issue was increased by 10,000,000 pesos. 2. In.May, 19h7, the common.shares were increased 1Pablo Sada, ”Some Notes on the Organization of Men- clova Steelworks," A Study Of the Iron and Steel Industry, II, 338. 2Pape, "Five Years of Achievement at Altos Hornos Steel Company," p. 58. 3Sada, ”Some Notes on the Organization Of Monolova Steelworke," p. 339. “Pepe, "Five Years Of Achievement at Altos Hornos Steel Company," p. 59. 119 to 60,000,000 pesos, while the preferred shares were reduced by 10,000,000 pesos by converting some preferred to common, on a one-for-One basis. The expansion thus financed made possible an increase Of capacity from 30,000 to h0,000 tons per month. It enabled the improvement of blast furnace Operations to 350 tons per day and allowed the addition of a third Open- hearth furnace with a capacity of 105 tons. It also enabled the expansion of the two original furnaces to that level.1 From.l9hh to 1950, annual production of iron and steel gradually increased until it reached 116,136 tons of finished product in 1950. In the same year, ingot pro- duction reached 123,hh2 tons or 82.3 per cent Of the 150,000-ton total capacity (Table 2). In 1951, another expansion program.was planned; by then, the company had achieved an income record which enabled it tO accumulate funds for expansion. However, the company found it could not finance the necessary dollar exchange for purchases in the United States. The Nacional Financiera therefore applied to the Export-Import Bank for a 5 million dollar loan. The loan was granted (Loan NO. h87-D, 5,000,000 dollars at h per cent interest repayable in 20 semiannual payments from.November, 1953, through.May, 1963).2 1Ibid. 2Vaughn, "A Study of the DevelOpment Of National Steel Plants in Selected Latin.American.Countries," p. 1&3. 50 Table 2. --Comparison of Ingot Capacity and Ingot Production. Altos Hornos de Mexico, S .A., 19h9- 60 Production as a Ingot Ingot Yb‘r capacity production p°iagzgit3r l9h9 150,000 118,156 78.8 1950 150, 000 123.hh3 82.3 1951 180, 000 1%3, ,167 79.5 1952 180, 000 6 750 92.6 1953 220, 000 182, 306 82.9 195 280, 000 219,173 78.3 195 280, 000 256, 075 91.5 1956 360,000 31h,172 87.3 195 A20, 000 358,635 8h.h 195 h30, 000 353.533 82.2 1959 500,000 h15.368 83.1 1960 600, 000 5&6, 222 91.0 Source: Letter from.Luis Barranco, Comptroller, AHMSA, to A. Carl Cass, Chief, Engineering Department, Export- Import Bank of Washington,D DC. ., August 8, 1961. 51 The expansion program embarked upon included the addition of a fourth Open-hearth furnace and the con- struction Of a new blast furnace with an 800-ton per day capacity. There was also an increase Of rolling capacity by 60,000 tons per year. Along with this, capacity of the Old Open-hearth furnaces was to be increased to 115 tons, and two new tinning lines were to be added.1 These inprove- ments were substantially completed by 1955, bringing ingot capacity to 280,000 tons per year, rolled product capacity to 1811,1100 tons per year, and pig iron capacity to a theo- retical £120,000 tons (based on a 350-day Operating year). By 1955, production of ingot had passed 250,000 tons per year and that of finished products 200,000 tons per year. In 1950, the old 350-ton blast furnace was taken out Of production and rehabilitated. Production also began on the fifth Open-hearth furnace. In December Of 1956, AHMSA applied to the Export- Import Bank for a credit of 19,259,809 dollars for further expansion Of steel mill facilities and the develOpment of iron ore deposits.2 A loan of 15 million dollars was granted in August Of 1957 to enable this expansion program. As Of June 30, 1960, AHMSA had completed the expansion lExport-Import Bank Altos Hornos de Mexico S.A. Memorandum Januar 1 19 . Re: zareai’Es NO. 323 m E8745, Facionai Fig i no era and AHMSA request for revision of maturities. 2Memorandum to the Board of Directors of the Export- Import Ba L ec er J . 52 program started in 1956, involving an expenditure of about 25 million dollars. The purpose of this program.was to increase production of ingot from 300,000 to 560,000 tons per year, and to increase production of finished products from 235,000 to u79,000 tons per year. Altos Hornos has, since 1950, produced about one- half of Mexico's pig iron and about one-third Of her steel ingot and rolled products. Between 1945-h7 and 1956-58, production Of rolled products by AHMSA (as Altos Hornos is usually called) increased by 267.5 per cent. During the same period, production by all other Mexican steel producers increased by 200.7 per cent (Table 3). There can be nO doubt that there has been a tremen- dous change since the inception of AHMSA in the quantity Of steel consumed in.Mexico and in the prOportion Of that steel produced domestically. In l9h5, imports constituted almost 60 per cent Of a consumption of rolled products of h3h,OOO tons. In 1958, imports constituted only about 30 per cent Of a national consumption of 1,320,000 tons.1 But this does not.mean that the establishment Of AHMSA was the cause Of this change. Nor can it be implied from.this that some other investment at the time might not have made a greater contribution to economic growth. Furthermore, this does not indicate what the consumption of steel in Mexico would have been if it were available at an unprotected price. 1"Rolled Iron and Steel Products in Latin America," Economic Bulletin for Latin America, IV, NO. 2, 23. 53 Table 3.--Mexican National Production of Pig Iron, Steel Ingot, and Rolled Products and Per Cent Accounted for by Altos Hornos, l9hh-58 (thousands Of tons) Pig Iron Steel Ingot Rolled Products Year National National National pro- Per cent pro- Per cent pro- Per cent duction AHMSA duction AHMSA duction AHMSA 19hh 159 13.3 175 3.8 1%h 0.3 19h5 210 22.6 230 19.2 1 6 22.0 l9h6 2A0 30.3 258 28.5 207 31.u 19h7 236 26.7 291 32.2 232 39-8 19h8 175 3 .2 291 33.2 232 37.0 1989 200 3 .6 371 31.9 297 35.9 1950 227 u7.7 390 31.6 312 37.3 1951 251 hl-S #67 30.7 373 38.9 1952 30h 39.0 533 31.3 #28 33.1 1953 212 h7.8 525 3u.7 h20 36.h 1958 252 5 .u 609 36.0 ABA 38.0 1955 328 5 .0 725 35.3 577 3h.9 1956 h08 50.h 880 35.7 707 32.0 1957 A29 h7.2 10h9 33.8 838 30.8 1958 A96 51.u 1115 31.7 890 27.2 Source: Mexican.Production--Fondo de Cultura Economica, Mexico: génouenta Anos de Revolucion (Mexico, D.F., 1961), pp. 218- 514 Conclusion The founding of each of these steel plants came at a similar time and for similar reasons. Each was clearly a manifestation Of a wave Of economic nationalism. In each case the establishment Of the steel plant was influenced by the external political strife and economic dislocation Of the times. Each was built with considerable assistance from a more develOped nation, the United States. In all three cases there was a history of steel production in the country before establishment of the mill and there was a raw material base more or less well suited to the production of iron and steel. As will be seen later, the steel plants were care- fully protected in a hothouse environment of subsidies and tariffs. It is Obvious from.the discussion of these plants that once established there was a tendency for them.to expand almost continuously. Obviously, attitudes and con- ditions similar to the ones that led to the establishment of these plants contributed to the decision to expand them. Added to this, however, is the fact that once the plant is firmly established it gains the political and economic power to work in behalf Of its own expansion. The expansion of the plants cannot be regarded as necessarily an indicator of their "success," especially in view Of the sheltered environ- ment in which they Operate. The combination of social, political, and economic 55 factors involved in the establishment of a national steel plant makes the decision difficult to appraise. In fact, it even makes the decision to attempt an appraisal a difficult one. No attempt will be made in this paper to consider or weigh the political and/Or sociological gains accruing from the establishment Of such a plant. A steel plant is an economic entity and, therefore, the decision to build one should be based on some economic criteria. In the following chapter some of the criteria for selecting projects and allocating investment funds will be discussed. An attempt will be made to select and justify the selection Of one Of these criteria. CHAPTER III THE CAPITAL-OUTPUT RATIO AS AN INVESTMENT CRITERION In the previous chapter it was strongly emphasized that a decision concerning the advisability of any particu- lar investment should be made only after appraising the available alternatives. This principle is just as true in the case of investment in a steel plant as it is with any other investment. The need to appraise theavailable alternatives is acute. The examination given to most projects rarely goes beyond determining whether there is a need for the products to be produced and whether it is well planned from an engi- neering standpoint. Hirschman has written with regard to his experience: In spite of all the insistence on "over-all planning," I have yet to see a project that is well-conceived rejected by national or international agencies disposing Of investment funds on the ground that the investment required is too high for ”balanced” develOpment.1 The primary purpose of this paper is to determine 1Albert 0. Hirschman, "Economics and Investment Planning: Reflections Based on Experiences in Colombia ," Investment Criteria and Economic Growth, Center for Inter- nationaI Studies, Hassac'fiusetts Institute of Technology (Bombay: Asia Publishing House, 1961) , pp. 1111-115. 56 57 whether it has been the case in the investment of capital in the national steel plants of Brazil, Chile, and Mexico. The purpose Of the present chapter is to describe various cri- teria that have been suggested for the appraisal of invest- ments and to select one Of them, In this case the selection of criteria is somewhat different from the normal case in.which a prOposed project is being considered in that the steel plants already exist. Ag_hgg investment decision making.must differ somewhat from the post hoc appraisal Of an existing enterprise. Theoreti- cally, however, the same criteria may be applied in either case. In fact, for purposes Of evaluation, planned invest- ments are treated as if they were actually in Operation by use Of various projective techniques. The actual appraisal of the investment decision involves three problems. First, there is the problem Of selecting and justifying the theoretical basis for apprais- ing the investment. Second, there is the problem of develOping a technique for manipulating the available empirical data in the application Of the selected criteria. Unfortunately, these two problems cannot be resolved sepa- rately since the availability (or inavailability) of data must influence the selection Of criteria. The first problem will be dealt with in this chapter; the second, insofar as it is really possible to split them, will be examined in.the following chapter. There is still another difficult problem that must 58 be resolved. This concerns the definition of "available alternatives." Hirschman has warned that unless an econo- mist evaluating a given project . . . can say in detail'what the alternatives are, unless he can produce projects in a state of readiness- tO-be-undertaken similar to the one he attacks, Ee wiII not andishouId not 5e Iistened‘to. *The reason for this position is tfiat there exist not only conceivable better uses of funds, but many worse ones as well, and these are the ones that are surely going to be undertaken if no ready alternative is produced by our objective economist.1 Unfortunately, it is not possible to compare the investment in a steel industry with other projects that were "in a state of readiness-tO-be-undertaken" at that time. This is because it is not possible tO develOp prOposals for projects which could have been.undertaken ten to twenty years ago. This does not mean, however, that an evaluation Of the investment in steel plants is of no use. It is still important tO determine whether capital has been expended in these projects "in a particularly lavish and unnecessary way,”2 even if nO specific alternatives are indicated. Insofar as the criterion chosen can indicate generally the suitability or misallocation of resources, the analysis can still be of value, especially for those countries contemp plating investment in steel industries. Hollis Chenery has provided an excellent starting point for a discussion of criteria for evaluation Of 11hid., p. ul. (Italics mine.) 21bid., p. 1.8. 59 investments. He divides his discussion of investment cri- teria into three categories: factor intensity criteria, marginal productivity criteria, and programming criteria.1 Each type of investment criteria has as its goal solution Of the allocation problemp-picking out project(s) which.wou1d make the greatest contribution relative to its cost. Marginal Productivity Criteria Neoclassical economic theory has as its answer to the allocation problem the rule that, given the distribution of income and perfect competition, Optimum.factor allocation results when in every use the value of the marginal product is equal to the price of the factor. Since factor prices are assumed to be equal everywhere, Optimum division is achieved when the value Of the marginal product of a factor is equal in every use. If there is perfect competition and profit maximization, this state will be achieved automati- cally. However, it has been argued that because of external economies and market imperfections in underdeveloped coun- tries an allocation policy based on marginal productivity will not lead to Optimum.resource allocation.2 These 1Chenery, American Economic Review, LI, No. l, ll-h9. 2Paul N. Rosenstein-Rodan, "Problems Of Industriali- zation in Eastern and Southeastern.EurOpe," Economic JOurnal, LIII, NO. 3 (June-September, l9h3), 202-11; Tifior Scitosty, "Two Concepts of External Economies," Journal Of Political Economy, LXII, NO. 2 (April, 1958), 1h3-52; Ragnar NurEse, O ems of Ca ital Formation in Underdevelo ed Countries Trondon: Ciford University Press, I953), Cfigpter V. 6O arguments stress the influence that the action of one prO- ducer may have on the costs of another. It is argued that because of external economies there may be differences between equilibrium.price and social value or, in other words, between the private marginal product of an investment and the social marginal product. Because of this difference it has been suggested that social marginal productivity (SMP) should be used as the measure of a project's contri- bution. As Kahn explains it, account should be taken of ”. . . the total contribution Of the marginal unit tO the national product, and not merely that portion of the con- tribution or its costs which may accrue to the private investor."1 The decisionemaking process would then involve ranking projects according to their SMP and then going down the list until the available funds have been exhausted. Unfortunately, the concept Of SMP and the "net con- tribution to national product" is not completely defined. And the vagueness Of the word ”social".makes it difficult to appraise the usefulness of this criterion as an evaluative device. In an effort to improve this situation, Chenery has attempted to reduce the SMP criterion to a formula.2 At the same time he Offers modifications tO allow for artificial 1Alfred E. Hahn, "Investment Criteria in DevelOpment Programs,” rterl JOurna; of Economics, LXV, NO. 1 (February, 3351), 35 aHOllis B. Chenery, ”The Application of Investment Criteria,” rterl Journal Of Economics, LXVII, NO. 1 (February, , - . 61 elements in the price system and to provide for an evalu- ation of labor and doreign exchange considerations. As Belshaw incisively notes, Chenery's formula is actually a sophisticated capital-output ratio in which a ”social value" replaces the usual output measures. This output is deter- mined by the effects of the project on the national income, balance-Of-payments, employment, etc.1 Chenery's approach does not, however, include a measure of external economies (nor do the other SMP criteria).2 An attempt has been.made to apply Chenery's formula to establish a priority system for develOpmental investment in the Philippines.3 However, difficulties in precisely measuring employment effects, balance-Of-payments effects, and not benefit derived from the use Of domestic raw materials (which.are included in the priority formula) for the steel industry, as compared with.even an hypothetical alternative makes the use of this formula impossible. Whatever criterion is decided upon, it is clear that these factors will bear considerable dis- cussion. Galenson and Leibenstein have suggested that the crucial factor in investment decisions ought to be what they lMichael Belshaw, "Operational Capital Allocation Criteria for Development Planning," Economic Develo ent and Cultural Change, VI (April, 1958 , - . 2Higgins, Economic DevelOpment, p. 660. 3Philippines National Economic Council, The Five- ‘Year Economic and Social DevelOpment Proggam.fcr Years - an a : O Dunc , e 62 call the “marginal reinvestment quotient"--a measure of the savings generated by a project.1 They argue, on the basis of their assumption, that the aim of a project ought to be to maximize per capita income at some time in the future and that the most productive project is the one that yields the greatest savings. Therefore, the most profitable projects are the most desirable and, they argue, the most profitable projects are the ones with the greatest capital intensity.2 The "Galenson-Leibenstein thesis" has been seriously attacked in a series Of articles in the Quarterlprournal_g£ Economics by Neisser,3 Bator,” Eckstein,S and Hirschman.6 Bator makes a strong argument against Galenson and Leiben- stein's use Of a social welfare function in which lower income would be preferable to higher income as long as the former resulted in a higher prOportion being saved, so that IWalter Galenson and Harvey Leibenstein, "Investment Criteria, Productivity, and Economic DevelOpment," rterl JOurnal Of Economics, LXIX, NO. 3 (August, 1955), 3E3-E7. 21hid., pp. 357-58. 3Hans Neisser, "Investment Criteria, Productivity, and Economic DevelOpment," Quarterl JOurnal Of Economics, LIX, No. h (November, 1956), - . “Francis Bator, "On Capital Productivity, Input Allocations, and Growth," Quarterl Journal of Economics, LXXI, NO. 1 (February, 1957), 36-136. 5Otto Eckstein, "Investment Criteria for Economic DevelOpment and the Theory Of Intertemporal Welfare Eco- nomics," Quarterly Journal of Economics, LXXI, NO. 1 (February, , - . 6Albert 0. Hirschman, "Investment Criteria and Capital Intensity Once Again,” Quarterly Journal of Eco- DOMICS, WII, NO. 3 (Auguflt, 1 , " 10 63 income at some future period is maximized. In general, and in the context Of the traditional con- fines Of the economist's formula maximizing setups, instantaneous Pareto-efficiency is a necessary condition for a full-fledged dynamical, intertemporal efficiency. . . . To be on an efficient growth path . . . one must maximize current output no matter what the consequences for the marginal productivity partial derivatives.1 Bator also points out that the applicability of Galenson- Leibenstein's conclusion.with regard to capital intensity is seriously limited because of the assumptions which have to be made in order to reach these conclusions. .Among these assumptions are: 1. that a larger marginal productivity of labor and smaller employment figure will be associated with a lower wage bill than with a small marginal productivity Of labor and large employment; 2. that a smaller wage bill divided by a smaller output will give a smaller ratio Of wages to output, hence a smaller ratio of wage to profits than a large wage bill and a large output; and 3. that a large rate Of savings times a low income will necessarily result in a larger volume of saving than a small rate of savings times a highincome.2 Neisser has also attacked Galenson and Leibenstein's social welfare function: Superiority of the capital intensive process as to 5 1Bator, Quarterly JOurnal of Economics, LXXI, NO. 1, 10 . ZIbid. , pp. 1011-105 . 6h profit originates only in the lower wage bill, which is the consequence of the lower employment power Of the capital intensive process. Therefore, the social cost of unemployment should not have been neglected.1 In a later work, Leibenstein seems tO have softened somewhat on his earlier stand on capital intensity, suggest- ing that "expansion Of entrepreneurship," effect On future savings habits, effects on future consumption patterns, and effects on pOpulation growth ought to be considered in investment allocation (in addition to SMP).2 He admits, however, that such a criterion might be difficult to apply. As Hirschman correctly points out, Leibenstein's theorizing "seriously impairs the usefulness Of the SMP criterion.with- out replacing it with a manageable new instrument."3 Thus it appears that neither orthodox marginal theory nor its better known variation offers a criterion which is both theoretically acceptable and empirically manageable. Programming Criteria The ideal method for allocating investment funds in an.underdevelOped country would probably involve some use Of a form.of linear programming framework. Such an approach h uh 1Neisser, Quarterly JOurnal of Economics, LXX, NO. , 6 0 2Harvey Leibenstein, Economic Backwardness and Economic Growth (New'York: JOKE WiIey and Sons, Inc., ’ p. O 3Hirschman, The Strategyof Economicygevelopment, p' 77a 65 would have an important advantage in allowing for large, as Opposed to marginal, changes in the economic structure.1 In the programming approach the criterion.used to compare projects is the social profitability of each project measured from the shadow prices. The shadow price of each commodity (input) is that in which.the price of each is equal to its cost of production (or its Opportunity cost).2 A linear programming approach would also make pos- sible the reconciling Of any number of restrictions in the economic system. As Chenery points out: Both the factor-intensity ratios and the partial pro- ductivity measures assume that there is one principal restriction on the system, the scarcity Of capital. They do not allow for the fact that in allocating capital according to one Of these rules some other restriction on the system, such as the supply of foreign exchange, of skilled labor, or of a particular commodity may be exceeded.3 Despite the Obvious theoretical advantages such an approach might have in planning for future develOpment, the problems involved in setting up the program are critical. The develOpment Of such a program.presupposes the avail- ability of good statistical information which is unlikely in underdevelOped countries. These limitations may be such as to make the application Of this technique almost impossible. 1Chenery, American Economic Review, LI, No. l, 31. 21bid., pp. 32-33. 3Ibid. See also Hollis Chenery, "The Role of IndustriaIization in.DevelOpment Programs," American Eco- nomic Review Papers and Proceedings, XLV, No. 3 (May, 1955), 66 These problems are even more crucial if the analysis involves going back any number of years to the inception Of the project which is to be evaluated, as it does in this case. Because Of theoretical Objections and lhmitations in data availability, neither the marginal criteria nor the programming approach.can therefore provide a satisfactory tool for the appraisal of a decision made sometime in the past to invest in a steel industry. The availability Of data severely limits the range of investment criteria that can be used. In fact, in this case the choice is narrowed to the use Of some sort of factor intensity criterion. Logic suggests that this should be some type Of capital intensity formulation. Capital IntensityvCriteria The capital intensity approach is at once the most Obvious and the easiest to apply Of the types of investment criteria. In its simplest form this criterion suggests that since capital is generally the scarcest resource in an under- develOped country, projects should be selected which.use the least capital in relation to their output. The tool most Often used in the measurement Of this relationship is some form Of the capital-output ratio. The capital-output ratio is, in general terms, the relationship between investment in a given project, industry, or economy for a certain time period and the output Of that same project, industry, or 67 economy over a similar time period. As far as selection Of projects is concerned, capital intensity consideration would thus lead to choosing those projects with lower capital- Output ratios over those with higher ratios. The incremental capital-output ratio measures the average output associated with a given or average investment (ignoring causation). In a sense, any particular project can be considered incremental from the point Of view of the entire economy. The capital-output ratio for the project may, therefore, be compared with the average incremental capital-output ratio for the economy. The latter provides a standard against which the (investment) use of capital in any new project can be measured. In other words, if the capital-output ratio of a prOposed project exceeds the incremental capital-output ratio for the economy as a whole, then the output associated with the investment in.the project is less than that associated with some hypothetical "average" investment in the economy. Moreover, assuming that the project involves a manufacturing industry, as in this case, a comparison can.be made between the capital- output ratio Of this project and the incremental capital- Output ratio of the manufacturing sector or that Of any Of the subcategories (census divisions) Of manufacturing. This briefly is the approach used in this study. Only in this case, instead of comparing various prOposed projects on the basis Of their capital-output ratios, a project that is already instituted is compared 68 with an hypothetical "average" project undertaken in the same period. This approach.has two outstanding advantages. First, it is empirically manageable, relying on obtainable census and corporation data. Second, it circumvents the difficulties involved in establishing alternatives for purposes Of comparison. Unfortunately, the capital-output ratio, although seemingly a simple concept, is rather sticky, as the concept can.be interpreted in nineteen different ways, depending upon the situation.1 First, there are important theoretical differences between the capital-output ratio itself and the marginal or incremental capital-output ratio. As indicated earlier, the capital-output ratio relates existing total capital stock to the total output. It is, as Lerner has put it, "none other than.the average product Of capital standing on its head and it becomes quite harmless as soon as one gets used to the "2 The incremental capital-output ratio, unusual posture. however, is definitely not the marginal productivity of capital standing on its head. The incremental capital- output ratio relates changes (additions) in the stock of capital with increases in output with no attempt made to 1Friedrich.A. Lutz, "Introduction," The Theor Of Ca ital, ed. Friedrich A. Lutz and Douglas C. Hague (fondon: St. Rartin's Press, Inc., 1961), p. ix. 2Abba P. Lerner, Ca ital-Out t Ratios Production Functions and the Theor Oi Ca itaI IEvestment, p. 9. (HiieograpEed.) 69 hold other factors constant. In other words, while the marginal productivity is a partial derivative, the supply of all factors other than capital remaining constant, the incremental capital-output ratio is a full derivative, the supply of all other factors being variable.1 This differ- ence naturally has important implications as far as the potential usefulness of the ratio is concerned. Two general types of uses have been suggested for the capital-output ratio. The first of these involves the use Of the incremental capital-output ratio as a tool for estimating the volume of investment necessary to expand output. The second suggests the use Of the capital-output ratio as a criterion for the allocation Of investment funds among alternative projects. The use of the incremental capital-output ratio in a model for determining rate Of investment needed tO produce a target increase in per capita income is relatively simple. If we let Y represent national income, 4y represent the absolute increase in income, and I be investment, then I/Y equals the rate of investment and Iflhy is the incremental capital-output ratio. It follows that: I.AY=AY Y’ ‘I "Y or rate of growth of income is equal to the rate of invest- ment times the reciprocal of the incremental capital-output 9 lBator, Quarterly Journal Of Economics, LXXI, No. 1, 0. 70 l The rate of growth Of per capita income can then be ratio. determined by subtracting the rate Of growth Of population from.the rate of growth Of output. On the other hand, given a desired AY/Y and the incremental capital-output, the required investment is easily determined. This is an over- simplification.but it is approximately correct for short periods Of time and low rates of growth. Difficulties arise largely from develOping an.empirical concept of capital. A second problem.invo1ves the doubtful process Of imputing a given output to a given input. The definition Of capital involves such esoteric problems as distinguishing land from capital, treatment of inventory changes, handling of depreci- ation, and adjustment for changes in the price level. These problems are largely technical and definitional and will be left to the following chapter. The second problem arises from the fact that capital- output ratios are not determinants of the rate of growth. 2 Therefore, to argue that They are merely ex post ratios. because a given increase in output has been associated with a given change in the capital stock in the past, some future increase in capital will result in a prOportional increase in output is not correct. Capital-output ratios have little predictive ability if factors unrelated to additions in capital stock can cause great fluctuations in output. Four 1Leibenstein, Economic Backwardness and Economic Growth, pp. 176-77. 2Lerner, Capital-Output Ratios . . ., p. 11. 71 factors which.may have this effect are: (1) unutilized capacity prior to addition to the capital stock which is employed after investment takes place, (2) increases in the labor force simultaneously with capital additions, (3) changes in the nature or availability Of "free factors," and (h) changes in the quality and/or energy Of the labor force.1 There is the further possibility that both capital and output are determined by the same set Of factors. The capital-output ratio may be a relatively passive result of interaction between the prOpensity to save and the role Of technological progress.2 These considerations limit the usefulness of the incremental capital-output ratio as a predictive device. This is not, however, of great consequence in this study. Here we are not using the incremental capital-output ratio to predict future investment needs. It is used instead as a means of develOping a standard against which the invest- ment Of capital in a steel industry and the results of that investment can.be measured. Therefore, in this case, the criticisms Of capital-output technique which are more telling are those of its use in investment allocation. In the case Of this study, the investment in a steel industry 1Leibenstein, Economic Backwardness and Economic G‘I'OWth, ppe 171-780 2Evsey D. Domar, "The Capital-Output Ratio in the United States: Its Variation and Stability," in Lutz and Hague, The Theory Of Capital, p. 117. 72 will be evaluated by the comparison of its capital-output ratio with that of others in the economy. In this case we are dealing only with known ratios and our analysis does not depend on the future stability of the ratio. The major objection to the use of the capital-output ratio as an investment criterion is that it ignores the existence Of other factors Of production. It does not measure the marginal productivity of capital, nor does it consider the requirements of other resources. Therefore, the minimum capital-output ratio criterion applies only if either capital is the only scarce factor on the system.or other inputs are so abundant relative to capital that the latter is the dominant element. To the extent that this condition is typical of underdevelOped economies, the capital intensity criterion is applicable. Another objection to the use of a capital intensity criterion is that it is not applicable to social overhead capital. First, there is almost no way of measuring the output of social overhead such as dams and railroads;1 second, this investment is almost invariably characterized by a very high capital-output ratio. However, the problem Of evaluating social returns of a project exist whatever criterion is used. And since, in this case, we are not comparing the steel investment with other specific invest- ments, the problem of evaluating the output of an lGeorge Rosen, Industrial Cha e in India (Glencoe, Illinois: The Free Press, I958), pp. 5I-52. 73 alternative which is a social overhead project is avoided. The return to social overhead projects such as dams or railroads show up in increased output Of enterprises uti- lizing their services.1 Therefore, when investment and output are aggregated, as in average incremental capital— output ratio for the economy, the social output capital investment will not result in an abnormally high ratio. Thus, the use Of the incremental capital-output ratio as a standard for comparison minimizes the difficulties normally caused by the social overhead investment. Another Obvious difficulty encountered in the use of capital intensity as an investment allocator concerns the fact that length of service of capital equipment plays an important part in determining the capital-output ratio. This problem.will be discussed in detail in a later chap- ter. Perhaps the most serious theoretical criticism of the use of the capital intensity criteria (and indeed Of all the usually suggested investment criteria) has been suggested by Hirschman: The question of priority must be resolved on the basis of comparative appraisal of the strength with which progress in one of these areas will endure progress in the other. In these basic types of develOpment decisions, it is therefore not sufficient to supple- ment, qualify and otherwise refine the usual investment criteria. We.must evolve entirely new aids to thought 1The output of investment in health or educational facilities might show up in the form of increased labor productivity spread throughout the economy. 7’4 and action in this largely uncharted territory Of effi- cient sequences and Optimal develOpment strategies.l Hirschman hypothesizes that one growth sequence may be more effective than another because it economizes decision-making ability or provides incentives to political action or develOps administrative ability. However, as Chenery has pointed out, "the empirical significance of these psychological and sociological factors remain to be established."2 At any rate, they cannot yet be accounted for in economic terms. Conclusion The decision Of an underdevelOped country's eco- nomic planners to invest a large quantity of the nation's resources in a steel industry is clearly a decision which should receive the most careful consideration. Yet, as seen in the previous chapter, there are factors Operating which result in a strong bias in favor Of the establishment Of a steel industry. The bias may be so strong as to preclude any unprejudiced analysis (by policy makers) of the conse- quences of this investment. However, there can be no doubt that a systematic analysis of the investment alternatives is called for before capital is allocated. In this study the investments have already been made and the best that can be done is an ex post analysis. This has the advantage, 1Hirschman, The Strategy Of Economic DevelOpment, P. 79. ZChenery, American Economic Review, LI, No. 1, hO. 75 however, of providing us with perfect knowledge of the per- formance of an investment project. Analysis will not have to depend on predicted results. In this chapter some of the investment criteria suggested in the literature have been reviewed in order to determine which would be empirically and theoretically apprOpriate to the problem at hand: evaluation of three recently established steel plants in underdevelOped coun- tries. Three general types Of investment criteria have been reviewed: marginal productivity criteria, programming cri- teria, and factor intensity criteria. Because of theoreti- cal Objections and empirical difficulties, the first two have been rejected, as far as this study is concerned, in favor of the third, factor intensity. More specifically, this criterion involves a comparison Of the capital-output ratio of the steel plant with the average (incremental) capital—output ratio for the economy and the industrial sector. As has been pointed out and discussed, this tech- nique is not completely satisfactory. But it has been shown that the theoretical Objections to this approach are not so strong as to destroy its usefulness. Using this approach, it is possible to make a preliminary judgment of the feasi- bility of the investment of capital by an underdevelOped nation in a steel industry. This approach has one over- whelming advantage over the other suggested criteria. It is empirically manageable even given the limited data available. 76 In the next chapter the empirical problems involved in the calculation of the capital-output ratios and the assumptions upon.which these calculations are based will be discussed in detail. The results Of the specific calcu- lations will also be presented and discussed. CHAPTER IV USING THE CAPITAL-OUTPUT RATIO TO APPRAISE THE STEEL INVESTMENT: THEORETICAL AND EMPIRICAL PROBLEMS As has been indicated, the evaluation Of investment in the develOpment of a domestic steel industry will be based on capital intensity criteria. This will basically involve comparison of the capital-output ratio (COR) of the steel plant in which the investment has been made with the average incremental capital-output ratio (ICOR) for the economy as a whole and with that of the manufacturing sector Of the economy. Use Of the COR for an enterprise and the ICOR for an economy, or any subsector thereof, involves certain theoretical and empirical difficulties. The Incremental Capital-Output Ratio For purposes of this study, the ICOR for the economy is used as a standard with which the capital-output ratio of the steel plant may be compared. The ICORs provide at least a rough indication Of the capital required to expand output. They also can provide in a more general way a measure Of the productivity of capital in alternative uses. In using the ICOR, it is assumed that there is a relationship between increases in output and increases in the stock Of capital, 77 78 with the former being brought about by the latter. Other factors affect the ratios of additions to capital to additions to output: 1. Changes in other input resources. 2. Technological change. 3. Changes in.the social factors influencing pro- ductivity. h. Changing reliance on foreign markets; Nevertheless, the relationship is so strong as to warrant examination. In its most common usage, the term."capital“ refers to reproducible capital stock of a nation. This would nor- mally include: (1) all construction and improvement attached to land; (2) machinery in the hands Of producers, both public and private; (3) the inventories of government and business; and (A) the net balance Of claims against foreign countries. Capital formation would measure additions to the capital stock, either gross, if depreciation is not deducted, or net, if depreciation is deducted. The sum Of the first and second factors comprises domestic fixed capi- tal formation. .Adding net additions to inventories would give total domestic capital formation, and adding the net balance Of claims against foreign countries gives the total national capital. Thus, the ICOR could include in its numerator either: 1. gross or net domestic fixed capital formation, 2. gross or not total domestic capital formation, 79 or 3. gross or net national capital formation. For purposes Of this study, the definition Of capi- tal formation is limited to the first Of the above cate- gories--gross or net domestic fixed capital formation. There is some question as to whether it might have been more apprOpriate to include inventories and use total domestic capital formation. The lack of availability Of data in this case obviated the necessity Of making this decision. The exclusion Of inventories could be compensated for by also excluding inventories from the firm CORs. However formulated, measures Of capital formation probably understate the true value. For instance, they would normally exclude substantial investment in the agri- cultural sector Of the economy (improvements made by the farmer's own labor).1' Furthermore, domestic fixed capital formation excludes capital investments in assembling and training the labor force, and it excludes the use Of non- reproducible natural resources. This underestimation must be kept in.mind when evaluating the results of the calcu- lations presented in the following chapter. The reliability Of data is recognized as a very serious problem.when conducting empirical studies on the economies Of underdevelOped countries. There is always the 1Simonxiuznets, "International Differences in Capi- tal Formation and Financing,” Ca ital Formation and Economic Growth, National Bureau Of Economic ResearcE (PFiEceEon: 'FTIEESton University Press, 1955), p. 20. 80 uncomfortable feeling that the statistical margin of error present may make the conclusions reached doubtful, at best. However, using the ICOR probably results in a greater degree of accuracy than would the COR. This is because an estimate Of the total capital stock is not necessary to the calcu- lation. It would ordinarily be expected that annual invest- ment data might be more accurate than an evaluation of the entire capital stock, especially since annual investment is a more crucial statistic than the size Of the capital stock to a develOping country. Even if one did not consider the data altogether reliable, the ratios would still be of interest. After all, if the planning authorities of the countries in question were actually deciding on investments on the basis of ICORS, they would have substantially the same data. The con- clusions they reached on the basis Of this information would be of further interest. Incremental output could be considered either change in gross national product or change in net national product. The decision would be dictated by whether gross or net domestic fixed capital formation is used. If at all possible, the denominator (change in out- put) of the ICOR should reflect only changes in real product. It is really not sufficient to express both capital formation and changes in output in market prices because price rises affect both differently. Changes in product in market prices are a result of changes in additions to output and 81 in.the initial volume of output. Capital formation in market prices reflects price changes only in additions to capital stock. Because Of this, the resulting ICORs will be tOO low if prices are rising. In this study, both.output and change in capital stock are deflated using the wholesale price index. A complicating factor is that the ICOR has a tendency to fluctuate widely from.year to year, due both to changes in the growth of output and in the rate of capital formation. In order to reduce the cyclical and erratic fluctuations in the data, average figures over time periods are used. ICOR of the Manufacturing Sector One would expect the ICORs for the various industrial sectors of any economy to vary widely. The ICOR for the economy as a whole would depend on the differing sectoral ratios and the weight Of each.sector in the total product. The economy's ICOR used as a standard for comparison is such a composite. While it is not possible to present the ICORs of "concrete prOposals in a state of readiness to be under- taken," it is possible to narrow the comparison to the manufacturing sector of the economy. As would be expected, however, the further one attempts to disaggregate the data, the more evident gaps in the information become. This is especially true of the availability Of data concerning investment in manufacturing and also concerning prices in this sector. Therefore, the ICORs calculated for the 82 manufacturing sector are more crudely adjusted than the ICOR of the economy. In the case of Chile and Mexico, it was possible to approximate the procedure used for the economy as a whole--that is, estimating additions to manufacturing capital stock, adjusting for price change, and dividing by change in.manufacturing output. For Brazil, however, data on annual additions to capital stock were available for only part of the manufacturing establishments, and there were output data relating specifically tO these enterprises. Capital-Output Ratios Of the Firm Derivation Of the COR for any enterprise involves the dual problem.of definition and calculation of the value of output and the definition and calculation of the size of capital stock. The latter problem is somewhat more compli- cated so it will be tackled first. Capital Stock Of the Enterprise For purposes of this study the definition Of capital will be limited to what is commonly defined as fixed assets. This would normally include land, plant, equipment, and vehicles. Defining capital stock in this manner has several technical and theoretical advantages. First, it allows maximum comparability with data for the economy as a whole on domestic capital formation. This figure would include, with the exception of land, the same group of assets. Second, if the COR for any firm is to be economically meaningful, the implicit assumption.must be made that the 83 output is in some way "produced by" the capital of the entire capital stock. Fixed assets are those most inti- mately connected with and most obviously essential to the production process in a technical sense. Ideally, however, a definition Of the stock of capital should include all the capital used to produce a given output. The major short- coming Of the definition Of capital as fixed assets seems to be from.this point of view the exclusions Of inventories. The exclusions of inventories may be justified on the grounds that inventories, while necessary for production, do not contribute directly to it. Except perhaps for goods in process the stock of inventories may be more of a mana- gerial prerogative than a technological imperative. A third advantage Of restricting the definition Of capital is that it increases the meaningfulness of inter- firm comparisons. This is because the definition omits those assets which cannot be easily defended as being essential to the productive process, such as cash.and deposits in banks (at least in part), investments in govern- ment securities, accounts receivable, other investments, etc. Finally, by limiting the discussion to fixed assets, we simplify the problem of adjustment for changes in price as it is not necessary to evaluate the effect of price changes on intangible assets. Unfortunately, there are no statistics of capital corresponding to those Of, say, output or employment. Corporations generally report data on their capital stock 81+ in the form of book value Of assets. These data, while appearing accurate, are Often.misleading for two reasons. First, all assets are reported in terms Of their original cost. They are, therefore, composite values, aggregated at different prices. In order tO be of any use, they should be adjusted to a given level of prices. Second, assets are usually reported in depreciated or net form, The methods for determining depreciation are generally established by a combination of governmental decision and accounting proce- dure. At best, they only incidentally reflect the actual physical deterioration Of assets. Corporate depreciation policy is, in general, conservative. That is, assets are usually written Off too rapidly, and some assets are com- 1 In order for the pletely written Off but remain in use. accounting data to be Of any use, then, they must be adjusted for changes in price and, if possible, compen- sation.must be made for overdepreciation of assets. The task of adjusting for changes in price level is extremely difficult because it requires knowledge of the price index for each type of asset and a knowledge of the data of acquisition and life span Of each.ssset. However, these problems can be overcome by using some generally acceptable price index and applying it to the "lump" of assets acquired during each year. The problem.of handling depreciation is more complex 1See Tibor Barna, ”On.Measuring CaBital," in Lutz and Hague, The Theory Of Capital, pp. 75-7 . 85 than indicated at first glance. First, the question as to whether the capital-output ratio is made more meaningful by allowing depreciation should be answered. Deductions for depreciation would be warranted only tO the extent that the aging of a machine results in a decline in its productive efficiency. In his analysis of several Indian industries, Rosen has argued for the use Of undepreciated capital figures using the assumption that in underdevelOped coun- tries a piece Of machinery tends to be used to the same degree and efficiency until it is discarded (and generally for a period far beyond the accounting life).1 Domar has also argued that the assumption of declining efficiency is 2 However, he seemm reluctant to allow that use unreliable. Of completely undepreciated data is justified. ’This problem cannot be solved on an a pgiori basis. Perhaps the best thing to do is to calculate the capital- Output ratio using the depreciated as well as the undepreci- ated values and to see if it makes a significant difference in the appraisal Of the project. The use of depreciated values brings up two further problems. The first concerns the effect Of the re-evaluation of assets for price changes on depreciation. The second is the possibility of changes in corporate depreciation policy over time (perhaps because of rising prices, and because of the lack of uniform 1Rosen, Industrial Change in India, p. 55. aDomar, "The Capital-Output Ratio in the United States: Its Variation and Stability," p. 93. 86 depreciation policy on the part of the three firms con- sidered here). If the period of time over which the COR is to be calculated is one of substantial inflation, and if depreci- ation were on an historical cost basis, the depreciation reserve would not cover the replacement costs of depreciated assets. If the asset stock were adjusted for price change, then depreciation would have to be similarly adjusted. This would be relatively easy if the firm itself did not try to compensate for the rising price level by either accelerating depreciation or adding to the depreciation reserve an adjustment factor. If this were not true, an alternative technique would be to redepreciate the asset stock on the basis of the adjusted values. The simplest method would involve calculation of depreciation on a straight-line group depreciation basis, assuming all assets have uniform life _spans. This would involve the selection and justification of some rate Of depreciation. This technique would solve the problem.oaused by adjustments for price changes and have the added advantage Of assuring consistent depreciation policy over the entire period under study and identical policies on the part of the three companies in question. This, in fact, was done. Value of Output The relevant measure Of the value of the output Of an enterprise, if any comparison is to be possible between 87 firms, is value added. Obviously, sales value would vary widely with the stage of the productive process and the degree Of integration of the firm. Value added has the further advantage Of being an income concept, thus allowing comparison with the usual aggregate measures Of income. Calculation of value added involves excluding the value of intermediate products used up in the production Of the final products. This may be done either by deducting from the gross value of output (sales), purchases from other producers (in this case largely fuels and raw materials), or by summing the values of wages, profits, and other incomes payable for factor services. The technique used in each case in this paper depended on the availability of data. Any monetary measure of output will be affected by changes in the price Of the product. And even a physical measure of output will be subject to cyclical and random variations. Handling of the price problem for the output for the steel companies is much less difficult than that involved in the valuation of capital. As long as data on the physical volume and composition of output are available as well as information on the relative prices of the compo- nents of output, it is possible to construct an index of the value Of product in prices Of one year. This index may be used to estimate the value of output in constant prices. Rosen has indicated that the relevant measure of output should be value added adjusted to a capacity rate of Operation. This, it is argued, would eliminate the cyclical 88 and random fluctuations in output and give a "truer" measure of the COR.1 It is not clear, however, that such an adjust- ment is apprOpriate in this case. It is possible, for instance, that an investment project was built too large for the size of the market. The consequent under-utilization of capacity would result in a high capital-output ratio; such a finding would be hidden by adjustment of output to capacity. It may also be interesting to investigate the possibility Of chronic Operational difficulty which would also be mani- fested by fluctuations in the COR. These variations would also be hidden by such adjustments. Calculation or the COR for the firms then involves solution of the following prob- lems: 1. Adjustment Of the capital stock for price changes. 2. Re-evaluation Of depreciation. 3. Estimation Of value added. u. Adjustment Of value of output for changes in the price level. Adjustment of the Capital Stock or r ce hanges If the COR of a firm is to be in any way comparable to the ICOR for the economy, or for a sector Of that economy, it must be stated in the same terms as far as prices are concerned. As mentioned previously, the use of unadjusted 1Rosen, Industrial Change in India, p. 57. 89 data introduces distortions into the analysis. So in this study all data have been adjusted to prices of the year 1950. Another difficulty that had to be overcome was that data on the asset stock of CSN were not available for 19hh- A6. Data were not available for CAP for l9h7-u8. Data were not available for these years because this was part Of the construction period; there was no production and no balance sheets were prepared. Since the deflated value of the capi— tal stock is the aggregate Of the deflated annual increments, these data were necessary to Obtain a value of the capital stock in 1950 prices. Therefore, the increments for the missing years was estimated. This was done by assuming that the percentage Of the capital stock Of the year of completion of the plant acquired in the first year of construction was the same for CAP (in 19h7) and CSN (in 19hh) as it was for AHMSA (in 1982). This gave a figure for capital stock in the first year. Data for the remaining missing years were estimated on the assumption Of equal increments. This left one final problem, selection of a price index. It would be difficult to exaggerate the problem this presents. First, part of the fixed capital was purchased domestically, part was purchased with dollars in the United States. CSN and AHMSA keep their books in domestic cur- rency, so that the dollar expenditures were converted to domestic currency at a rate of exchange selected by the company. CAP keeps its books in dollars, sO the domestic 9O expenditures were converted to dollars at a rate of exchange selected by the company. In the case Of CAP the exchange rate used was known so that the data could be reconverted to domestic currency. Deflation of the data is complicated by the fact that the precise prOportions Of domestic and for- eign expenditures in each year are not known for any of the companies. It is further complicated by the fact that data on the domestic price levels in the countries concerned are extremely scarce. Even under the best empirical conditions, as in working with United States' data, construction of a price deflator for capital goods is an extremely tricky prOposition. The capital stock was deflated using the domestic wholesale price index. This choice was made more out of necessity than conviction, the wholesale price index being the one index available for all three countries for the length Of the study. Adjustment of the capital stock for price changes is essential if the COR is to be meaningful. Because of the different effect it has on capital, a stock, and output, a flow, inflation automatically tends to reduce the COR. Reeevaluation of Depreciation In the examination of depreciation policies of the three companies, it was found that none of the companies seemed to follow a strict historical cost depreciation policy (see Table A). The Compania de Acero del Pacifico 91 Table h.--Annual Depreciation as a Percentage of Book Value of Fixed Assets: Altos Hornoa, Companhia Siderurgica Nacional, and Compania de Acero del Pacifico Annual depreciation as a percentage of book value Of fixed assets Year AHMSA CSN CAP l9hh 1.0 19h5 1.u 19h? 3.6 0.7 19h8 3.5 1.3 19h9 3.h 3.6 1950 6.0 ---b 1951 5.0 8.8 5.5 1952 11.0 5.6 n.1, 1953 hoh %°9 hoO 1951; go .5 11.2 1955 .8 10.0 u.6 1956 10.6 13.2 11.6 1957 11.9 11.7 h.7 1958 10.2 12.2 h.2 1959 9.8 12.1 2.5 aData from.first year of depreciation. bNO depreciation. Source: Calculated from Appendix Tables 2, 3, and h. 92 annual depreciation averaged at 5 per cent of the value of the Huachipato plant annually. However, the company added to the depreciation reserve an allowance for change in exchange rates between Chile and the United States. This amounted to a de facto allowance for Chilean inflation. Therefore, depreciation for this company was not strictly on an historical cost basis. Altos Hornos depreciation averaged h per cent Of historical cost annually from.l9h6 through 1958. In 1955, depreciation was accelerated to a rate of approximately 10 per cent per year. For CSN the rate Of depreciation averaged 7 per cent from.l9h6 through 1953 and approximately 1h per cent from 195k on. In order to calculate the depreciated value of capi- tal stock in current prices and in 1950 prices, depreciation was recalculated by straight-line group method assuming all assets with a useful life Of 25 years. This is the life span suggested by the United States Internal Revenue Service. Estimation Of value Added For Altos Hornos value added was estimated at the difference between total sales and the cost Of raw materials and fuels.1 The value added by manufacture Of Companhia Siderur- gica Nacional for the years 1951-59 was estimated as sales produced at Volta Redonda less the cost of materials 1Altos Hornos de Mexico, S.A., Informs Anual.(Mon- clova, Mexico, 19hh-55). 93 purchased. However, the sales produced at Volta Redonda did not account for the total sales Of the company although this difference is relatively small. There is no way Of directly estimating the value added component Of this part of sales. Total value added was therefore simply estimated by applying the ratio Of value added to sales produced at Volta Redonda to the value of total sales.1 For earlier years it was assumed that the prOportion of value added to total sales was the same as the 1951-59 average. For the years 1950 through the fiscal year 1957-58 CAP's value added was estimated as the summation of profits, labor costs, general overhead, and depreciation.2 For the fiscal years 1958-59 and 1959-60 value added was estimated as cost of sales (raw materials and supplies, general manufacturing expense, selling expense, and staff participation) plus depreciation and profits less estimated 3 cost of raw materials and supplies. 1Price'Waterhouse and Peet, Com.anhia Siderur ica Nacional Financial Statement 19 l- o e ane O, I95I-60), and Companhia derurg ca ac onal, Relatorio do Directoria, 1986-1951 (Volta Redonda, Brazil, I9E2-52). aFor profits, overhead, and depreciation the source used was KOppers Company, Inc., and Compania de AcerO del Pacifico, Proggam.E l96l-196Q (Santiago; 1958), Exhibit H-l, "Historica rO . and as Statement. Labor costs esti- mated On the basis of averages for 1950-55 as reported in KOppers Company, Inc., Re ort on the Operations of Compania de Acero del Pacifico (unpuBIished). 3Cost Of sales, depreciation, and profits from Compania de Acero del Pacifico, Memoria Anual 19 8-19 9, and 1959-1960 (Santiago, 1959 an . s as 0 cost 0% raw ma er als from.Koppers Company, Inc., report cited a ove. 9h Agjgggment of value of Output for nges in the Price Love The simplest possible technique of deflating the value Of sales would be to divide the total tonnage sales into product categories (i.e., rails and accessories, bars and structural steel, thick plate, cold rolled sheet, hot rolled sheet, tin plate, etc.), multiply the tonnage in each category produced during any given year by the base year price and sum the values of each category's sales in base year prices. Unfortunately, it is the practice Of AHMSA and CSN to report production tonnage rather than tonnage sales. It was, however, possible to use the production data with base year prices (1950) to construct an index of the value Of output along the lines suggested above. This index was ethen used to estimate the value Of sales in 1950 prices.1 Of course, this technique will approximate the actual deflated value of output to the degree that output increased at the same rate as sales. The deflating of sales for CAP was complicated by the fact that the company exported a considerable prOportion of its total output. A large prOportion of these exports were semifinished products (these have a lower value per ton than finished products). The problem was further compli- cated by the fact that accurate data on export prices were not available. However, data on tonnage sales, domestic and 1Sales, year n, 1950 prices = sales 1950, current prices x index of value Of output, year n, 1950 prices. 95 export, and.money sales, domestic and export, were both available. A weighted index of the value Of sales in con- stant prices was constructed by dividing sales into the categories domestic, finished exports, and semifinished exports. The index was computed as follows: Sales, year n, base year prices 3 W omes c onnage, ase year x domestic tonnage, year n + (finished export tonnage x base year price per ton) + (semifinished export tonnage x base year price per ton). For this study it was necessary to use 1955 as a base rather than 1950 because the 1950 data only represented three months of production. Sales in 1955 prices were con- verted into an index and the index was shifted to a 1950 base then applied to the 1950 sales to get the value Of sales in 1950 prices. The 1955 price per ton Of exports was estimated on the basis Of the relative prices of semifinished and finished steel products in the United States in 1955. The f.o.b. mill composite price Of finished steel was used for the finished exports and the f.o.b. mill price for Bessemer billets was used for semifinished products. These were then adjusted to equal the weighted average price of CAP's exports in 1955 as follows: Let z be the adjustment factor z=(price, composite, 1955 I finished exports, 1955) + (price, Besgemer billet 19 x semifinished e arts 19 Total export sales, 19 96 and: price Of finished exports, 1955 8 (price, composite) (z), price Of semifinished exports, 1955 = (price, Bessemer billets, 1955) (2). For a complete description Of the technique used, see Appendix Tables 11 through 13. Summary It has been shown that it is possible to calculate various series representing the CORs of the individual firms and the ICORs Of the economies and their respective manu- facturing sectors. These ratios can be calculated in con- stant or current prices and for depreciated or undepreciated values. The differences in the techniques Of calculation have been noted and the theoretical implications and empiri- cal shortcomings Of each.have been indicated. The specific technical solutions to umpirical prOblems (insofar as the problems were soluble) have been discussed. Given the various COR formulations for the steel plants and the corresponding ICORs Of the economies and their manufacturing sectors, several series Of comparisons can be made in evaluating the investment in the steel plants. The plant CORs can be compared with the economy and manufacturing sector ICORs. This would give some idea of the relative ”extravagance” Of the use of capital in the steel plant. The use Of the COR technique also allows two other evaluative comparisons to be made without additional 97 computations. First, it is possible to examine the COR of each plant over time. From this series it is possible to evaluate the effect of Operational experience and the develOpment Of the domestic market on the plant COR. It also makes it possible to examine the effects Of plant expansion on the COR. Second, it is possible to make inter- firm comparisons Of CORs. From these same conclusions can be drawn as to the factors influencing the level of the COR in the same industry but in different economies. Along with these comparisons involving the COR, it would be interesting to compare prices of domestic steel in Brazil, Chile, and Mexico with prices of imports. The import price, as normally reported, would include import duties and would reflect the degree Of protection afforded the steel industry. A more meaningful relationship would be that between the domestic price of steel with the c.i.f. price Of imports. The difference between the domestic price and the cost of imports represents the inplicit cost to the economy of domestic steel production. In order to make this comparison, an estimate was made of the c.i.f. price Of imported steel. The results of this comparison and Of those mentioned above will be reported in the following chapters. CHAPTER V CAPITAL-OUTPUT RATIOS OF THE THREE STEEL PLANTS The Capital-Output Ratio Over Time Expected Findings One Of the most pervasive findings in empirical studies Of the capital-output ratio of industries or groups of firms is the phenomenon Of the decline Of these ratios over time. This phenomenon has several important impli- cations for this study. First, if the COR is to be used as a device for evaluating an investment on a post-hoc basis, then finding a declining COR might prove bothersome. If the COR for a firm has declined over time ,1 then there would be no one COR with which commrisons could be made. Second, the use of the COR for the appraisal of hypothetical invest- ment alternatives would be made more meaningful if it were possible to include in the analysis a discussion Of the possible behavior Of the COR. Finally, a discussion Of the movements or trend Of the COR Of the firm would be helpful in evaluating the effect of such Often discussed factors as y 1The expression of the decline of the COR as a function Of time is used here merely to avoid jumping directly into an analysis Of the causal factors. 98 99 economies and diseconomies Of scale and labor versus capital intensive projects in investment allocation. Most Of the research concerning the CORs of manu- facturing firms has been Of the American experience. These findings will serve as a basis for discussion. In an inves- tigation Of the COR in the American economy, Creamer found that the ratio for three-digit SIC manufacturing industries, almost without exception, increased from 1890 to a peak sometime between 1900 and 1929 and then declined there- after.1 The reasons for the declining 003 are not at all clear; however, certain factors are generally thought to be influential. First, Of course, is the fact that a rising price level will naturally result in a falling COR. This can.be corrected for, and, in fact, is in the study cited above.2 A second problem.involves depreciation. Use Of depreciated value of assets will result in a declining lDaniel Creamer, Sergei Dobrovolsky, and Israel Borenstein, Ca ital in Manufacturi and Min (Princeton: National Bureau 0 cono c esearc , r ce on University Press, 1961), Table A-lh. 2However, it has been argued also that the decline of the COR is due to defects in the capital deflators. While output deflators are constructed by'a market basket technique, measuring price changes Of a group of commodities fairly constant in nature over time, capital deflators are constructed by measuring the amount of factors used in the production of capital equipment. Because Of this capital deflators are much less successful in eliminating quality changes and technological improvements than.are commodity deflators. 'For this reason a capital price index so con- structed rises tOO rapidly and, therefore, the capital- 1output ratio appears to decline. George Anderson, "The 4Apparent Decline Of the Capital Output Ratio," ggarterly piournal of Economics, LXXV, NO. 6 (November, 19 , -3h. 100 capital-output ratio if the rate Of depreciation does not reflect the declining productivity of the equipment over time--the asset stock is overdepreciated. In.a more theoretical vein, Kuznets ascribes the pattern of the COR to what he calls the industry's "life cycle of capital."1 First, in the early stages Of an industry's life there tends to be an underutilization of the capital stocke For example, time is necessary to build up to capacity Operation. This would help to explain initial declines in the COR. Secondly, Kuznets points out that given a constant life period of depreciable capital, a high rate of growth Of capital means a low ratio Of depreci- ation to the total capital stock (and thus a rising COR). On the other hand, a slowing down of the rate of growth.af capital means a rising ratio Of depreciation to fixed capi- tal (and a declining COR). Third, Euznets postulates an "accelerator effect.” In the early stages of an industry's development, when the rate Of growth Of output is high, capital tends to grow at a higher rate and the COR thus rises. In the later stages, when the rate of growth.of output is smaller, the capital stock grows at an even smaller rate and the COR falls. Each Of these hypotheses might be applicable to individual firms. In an underdeveloped area, we would y 1Simonfiuznets, Ca ital in the American.Econom (Princeton: National Bureau of Econaiic Research, ceton University Press, 1961), pp. 86-90. 101 expect to find a sharp initial decline of the COR of a new firm in the first few years Of Operations, possibly due to a slow growth Of acceptability of a domestically produced import substitute or to the slow growth Of the market in response to availability Of domestic sources. The second hypothesis, regarding the relation.between growth Of capital and the ratio Of depreciation to asset stock, would result in an upward or downward movement Of the COR only if the rate Of growth Of capital were accelerating or decelerating, but not if it were constant. The effect on any one firm would depend on its particular capital growth rate. There is, therefore, no a priori reason to assume that this would lead the COR for an individual firm to increase or decrease. Furthermore, this hypothesis would not explain declines in a gross (undepreciated) COR. The "accelerator effect" is not really an hypothesis at all, but is merely a description of an actual condition. If output growth is accelerating and the COR is rising, capital must be increasing at a faster rate; if the COR is falling, capital must be increasing at a slower rate. It might be hypothesized that this acceleration is due to preparation for future output. But there is an Opposite effect which.might lead one to predict a reverse reaction on the part of the COR. When output is growing at a high rate, Operations tend to be near capacity, but as the rate of increase declines, Operations may fall to a lower per- centage Of capacity. This would indicate a higher COR as 102 the rate Of growth Of output falls. In fact, Creamer's findings (10 not support Kuznets' "hypothesis."1 The expla- nations presented by Kuznets, with the exception Of the first, do not explain the pattern of CORs found by Creamer nor do they Offer any insights into the expected pattern Of CORs for an individual firm. It can be argued that the decline in COR in Ameri- can manufacturing since 1929 has been due to technological improvements which economize in the use Of labor. This would be true if the capital-labor ratio were constant. If instead the capital-labor ratio rose, then a fall in the COR would depend on the increased productivity Of labor (g) out- weighing the decreasing returns to capital with substitution of capital for labor, as easily seen by manipulating the identity, 5 8% . %. There is no a priori reason to believe that this would be true, either for the manufacturing sector or an individual firm. Intimately related to the question Of the behavior of CORs over time, and Often confused with it, is the question of the relationship between the COR and size of the firm. Available evidence strongly suggests that there is a direct relationship between the size Of the firm and the COR. For instance, Stanley S. Schor found that in 19117 the highest COR was found in the group Of firms Of greatest size in eighteen out Of a total of twenty-two manufacturing lCreamer, Dobrovolsky, and Borenstein, Capital in Manufacturing and Mining, p. 911. 103 industries.1 This would mean that if the average size of firm.has grown over time, say since 1929, the COR has declined in spite of a "size effect" which would tend to move it in the Opposite direction. Numerous attempts have been made to explain.the size effect. Since Schor's data and earlier evidence2 were based on CORs using sales as a measure of output, an Obvious temp- tation exists to ascribe the higher COR in larger firms to vertical integration.3 This, however, can be corrected for by using value added as a measure of output rather than sales. This was, in fact, done by Marcus, but the direct relationship between size Of firm and COR was still evi- dent.h Schor has argued that the higher COR Of larger firms is due tO the fact that since they are larger they are not subject to the competitive pressures to economize.S This, 1Stanley 8. Schor, "The Capital-Product Ratio and Size of Establishment for Manufacturing Industries" (unpub- lighgd PheD. dissertation, University of Pennsylvania, 19 2 . 2Walter A. Chudson, The Pattern of Cor crate Financial Structure (New YOrE: NaEionaI Bureau of Economic esearc , . 3A vertically integrated firm would Obviously have a higher capital-sales ratio than the total of its component subun ts. #Matityahu.umrcus, "Capital-Output Ratios and their Variation" (unpublished PhaD. dissertation, Brown Univer- sity, 1963). 5Schor, "The Capital-Product Ratio and Size of Establishment for Manufacturing Industries," p. 87. 10h in effect, implies that they do not try to maximize profits. This, while not impossible, is not a generally acceptable explanation. Scitovsky has argued that ”large firms are likely to be in a better bargaining position vis-é-vis the producers of equipment and, therefore, Obtain them [capital goods] at more favorable prices than do the small firms."1 This is saying, basically, that factor price ratios are not the same to large as to small firms. This argument has been made also in.a more refined formbyMarcus,2 who has shown that the cost of borrowing (and therefore of capital) is lower for large than small firms, and that there is a posi- tive correlation between the cost of borrowing and the level of the COR. This argument could.have important implications for develOpment policy. First, the price Of capital to a firm.subsidized by foreign loans would not reflect the domestic price ratios of capital to labor. If the price of capital in terms of the rate of interest were lower, the firm would use a more capital-intensive technique then if the choice had to be made on the basis Of the domestic factor price ratios. Second, once a firm.became estab- lished, the factor price ratios (even domestic) could shift in favor Of capital leading to still higher CORs as the firm 1Tibor Scitovsky, "Economic Theory and the Measure- ment of Concentration," Business Concentration and Price Polic , Special Conference Series, NaEionaI Bureau of Eco- nomic Research (Princeton: Princeton University Press, 1955). p. 11. " ZMarcus, "Capital-Output Ratios and their Vari- ation, p. 112. 105 increased in size. Another possible cause of the "size effect" is the difference in product mix between large and small firms. If large firms use more economical methods Of production than small ones, then there would be better reason to suppose that large and small firms would be producing the same product using different techniques Of different capital intensity. The small firms would in this case not survive. The aggregate data probably conceal the fact that the larger firms are manufacturing a different product mix requiring a higher 0011 than the small firms.1 This would suggest that higher CORs would result from.the growth Of a firm.on1y if this growth were accompanied by a shift in the product mix in favor Of more capital-intensive goods. Again there are no a priori reasons to assume this will be true Of any individual firm or industry. Davis' study indicates that the reasons for the direct relationship between size and COR varies from industry to industry. In the agricultural machinery industry, for instance, the relationship seems to be due to product specialization. The larger firms manu- facture heavy implements and tractors, and the smaller firms produce simpler implements. In the men's shirt industry, on the other hand, the smaller firms have lower CORs because they tend to subcontract some Operations and also rent 1Evidence of this is presented by Hiram.Davis, "Relationship of the Capital-Output Ratio to Firm Size in American.Mhnufacturing, Review Of Economics and Statistics, XXVIII (August, 1956), 288-931 106 building space.1 In sum, there is no conclusive evidence concerning the probable trend of the COR for a firm over time. Some hypotheses would lead us to expect a downward movement, others, especially those related to size, would lead us to expect an upward movement. There is even the possibility that the "time effect" may be cancelled by the "size effect" resulting in a constant COR. Actual Findings In Tables 5 through 7 various formulations Of the COR have been presented for Altos Hornos (AHMSA), Compania de Acero del Pacifico (CAP), and Companhia Siderurgica Nacional (CSN). These ratios are presented in gross (unde- preciated) and net terms, using undeflated and deflated data. As would be expected, there was a sharp decline in the COR Of each company. (This decline was apparent regard- less Of the method of calculation of the COR.) Since the output of the first year of Operation generally represented production of less than a full year, the COR calculations shown begin with.the second year or the first full year of Operations. This was the year l9h7 for CSN, l9h5 for AHMSA, and 1951 for CAP. As might be expected, there was a tend- ency for the COR to drOp sharply for the first few years of Operations. This drOp was a result Of growth Of output ~ 1Ibid., pp. 288-89. 107 Table 5.--Capita1-Output Ratio, Various Formulations, Companhia Siderurgica Nacional, 19h7-59 Undeflated data Data in 1950 prices Year NFA adj/'NFA adj/ GPA/VA NFA/VA spa/ya urn/v1 VA VA adj l9u7 18.31 18.06 26.2h 25.98 2h.18 .02 19u8 6.u0 6. 11.22 11.07 9.90 .92 19h9 5.h8 5.15 9.63 8.91 7.85 7.82 1950 n.05 3.83 7.30 7.02 5.90 5.90 1951 3.19 2.73 5.h6 5.02 u.21 h.27 1952 3.32 2. 0 5.99 5.3% u.u2 h-37 1953 3.59 2. 5 6.19 5.3 u.uh h.%2 1958 3.68 3.01 6.9h 5.82 h.96 h. 9 1955 2.7a 1.95 5.75 h.57 h.00 3.87 1956 2.09 1.26 5.1u 3.77 3.33 3.%6 195 2.25 1.20 5.93 h-02 3. 2 3. 2 195 1.82 0.89 5.h3 3.39 3.39 3-h7 1959 1.28 0.63 5.30 3.07 3.26 3.39 Note: GFA = gross (undepreciated) fixed assets NFA - not fixed assets VA 8 value added NFA adj = net fixed assets, adjusted depreciation VA adj 8 value added, adjusted for changes in product mix Source: Calculated from.Appendix Tables 2 and 18. 108 Table 6.--Capita1-0utput Ratio, Various Formulations, Compania de Acero del Pacifico, 1951-59 Undeflated data Data in l950prices Y9“? NFA adj/INFA adj] GFA/VA. NFA/VA GPA/VA MFA/VA VA VA adj 1951 2.29 2.06 5.99 5.60 5.75 5.38 1952 l.h9 1.2h u.68 h-IZ h-33 8.02 1953 1.65 1.30 n.8u h.03 h.32 3.62 195 1.u0 1.0u h.80 3.77 h.13 3.72 195 0.92 0.62 n.8h 3.51 3.99 3.70 1956 0.63 0.87 5.29 3.65 h.27 3.78 1957 0.82 0.61 5.05 3.27 3.96 3.23 1958 1.18 0.96 7.35 8.72 5.67 n.7h 1959 1.75 1.51 8.60 5.56 6.65 Note: See Table 5. Source: Calculated from Appendix Tables 3 and 18. 109 Table 7.--Capital-0utput Ratio, Various Formulations, Altos Hornos de Mexico, S.A., 1985-59 Undeflated data Data in 1950 prices Ybar NFA adj/ NFA adj? GFA/VA NFA/VA GPA/VA NFA/VA VA VA adj 19’45 29e62 20e9u- 26039 37e17 2,4038 ----a 1986 8.72 8.52 5.56 6.56 8.93 8.13 1987 3.37 3.12 3.76 8.78 3.21 2. 1988 3.05 2.7 8. 1 8.35 3.87 2.50 1989 2.32 2.0 3. 6 3.32 3.10 2.81 1950 2.19 1.81 3.67 3.22 2.82 2.82 1951 2.23 1.83 3.36 3 .52 2.55 2.59 1952 3.03 2.53 8.10 8.28 3.18 3.12 1953 3.01 2.83 3.92 3.83 2.90 2.93 1958 2. 1.95 3.86 3.18 2.52 2.58 1955 3.02 2.35 8.21 3.70 3.15 3.18 1956 2.82 1.69 3.83 2.75 2.80 2.82 1957 2.38 1.87 3.78 2.22 2.69 2.78 1958 2.75 1.69 8.33 2.92 3.15 3.25 1959 2.61 1.83 8.16 2.23 2.95 3.08 aNot calculated. Note: See Table 5. Source: Calculated from.Appendix Tables 8 and 18. 110 rather than of changes in the capital stock. It would not be expected that a plant could start capacity Operations in the first year Of production largely because Of the techni- cal, Or engineering, and marketing problems. For AHMSA the period of sharp decline lasted until 1986 or about three years; for CAP about until 1952 (about two years); for CSN until 1989 or 1950 (about four years). This time might be called the "break-in" period. It is not possible to say precisely what factors influence the length of the "break- in” period, but among them might be: the length Of time necessary to complete facilities still under construction at the time production began, the scale of plant relative to size of domestic market, and government policy in limiting imports. At any rate, whatever the specific reasons, a break-in period during which the COR drOps Off sharply should be expected and is evident in the data presented. What is more interesting and important is what happens to the COR as the firm grows after the initial period. In order to investigate what actually happened to the post breaksin COR Of each plant, the trend Of the COR has been estimated by a straight line fitted by the method Of least squares for the years after the break-in period. As shown in Table 8, the COR appears to fall for all three firms if the common COR formulation ratio Of net fixed assets, book value to value added at market prices, is used. If the data are deflated, a decline should be apparent in the rate of fall Of the COR. However, line 2 Of Table 8 111 Table 8.--Average Annual Change in Capital-Output Ratios, Post Break-in Period, as Indicated by the SlOpe of a Straight Line Fitted by the Method of Least Squares, CSN, AHMSA, and CAP Formulation of capital-output ratio CSN Com an and ears AHMSA CAP 1950-59 1987-59 1952-57 NFA, book value/VA, market prices -.l66 NFA, 1950 prices/VA, 1950 prices -.181 NFA, adjusted depreciation, 1950 prices/VA, 1950 prices -.111 NFA, adjusted depreciation, 1950 prices/VA, adjusted for product mix, 1950 prices -.106 GFA, 1950 prices/VA, 1950 -.096 -.087 -.158 -.081 -.038 -.031 +.008 -.050 +0009 " eOhé Source: Calculated from Appendix Tables 8-6. 112 does not indicate this for either CSN or AHMSA. This is because, as noted in the previous chapter, deflating net fixed assets directly has the effect of vastly understating depreciation.1 Therefore, a more accurate picture requires that the gross fixed assets be calculated in terms of 1950 prices and then redepreciated. The results of this prO- cedure are shown in line 3. This shows that when fixed assets are deflated and redepreciated the COR does not decline nearly as rapidly. The rate of decline for CAP and AHMSA falls by two-thirds and that of CSN by one-third.2 At least part Of this decline can be shown to have been.the result Of changes in the product mix. ‘Vaiue added has been.recalculated under the assumption that the product mix as well as prices remain the same as they were in 1950. This is done by revaluing sales on the basis of the price per aggregate ton in1950,3 and assuming that all products have the same prOportion (or approximately the same) Of 1This is true because the value Of assets purchased before the base year will be inflated by the price adjust- ment procedure. 2Note that this still does not eliminate the possi- bility that the decline may be due to overdepreciation of assets. If the physical deterioration of the average assets is less than 8 per cent a year, the rate used for redepreci- ation, then the net asset stock has been undervalued in this calculation of the COR. 3 Mone sales 19 0 Sale: your X = O a onnage x Tonnage out- Output, 1950 put year X The accuracy of this technique rests partly on the assumption Of constancy Of the ratio of sales to production. 113 value added per money unit of output. Given these assumptions, the COR, using constant product mix and not fixed assets in 1950 prices, was calculated. Line 8 clearly shows that part of the decline of the COR again disappears. In fact, for AHMSA, the COR now appears tO rise slightly, while it falls somewhat less in the case of CSN. The ratio for CAP seems to decline more drastically when using this calculation. This is because in 1957 (and 1958) the product mix changed drastically in favor of exported semifinished products with a much lower value per ton. In order to con- vert this to 1950 product mix, it was first necessary to convert the semifinished products to finished tonnage equivalents,1 then assign a value per ton on the basis Of 1950 prices. This results in a higher value added and, therefore, a lower COR.2 In order to get some further insight into the effect of depreciation, the trend of a COR using group fixed assets and value added in 1950 prices has been calculated (line 3). For AHMSA and CAP, the COR now has an upward trend (as coma pared with the equivalent line 3); for CSN the trend is still downward although at about half the rate Of the trend of the net ratio. This indicates that at least for CSN the 11.33 ton semifinished I 1 ton finished. 2Note that for CAP, because Of the falling output and value added, the calculation was limited to the period 1952-57. Notice also that for this company the lepe of GFA/VA is greater than the lepe of NFA/VA. This is due to the fact that the COR undeflated COR rose sharply in 1957 and when the COR rises, MFA/VA rises faster than GPA/VA. 118 downward path Of the COR cannot be explained away as merely a result of the treatment of depreciation. implications It seems clear that the apparent decline of the COR can be accounted for, in part, by price changes, depreci- ation policy, and changes in the product mix. Unfortu- nately, the simple deflation process used in this paper cannot adequately adjust the capital stock for changes in the price level. The effect Of deflation on the trend of the COR is, however, clear. If accounting depreciation were matched precisely to the physical deterioration of assets, then depreciation would not contribute to the decline of the COR. A somewhat more intriguing question is that of the influence Of changes in product mix on the COR. However, since it is not possible to calculate the COR for each product a firm produces because all are joint products, it cannot be said which products are more capital-intensive than others. If a plant is built with a given capacity for several different products, and if the change in product mix involves the uneven growth of outputs Of these products, such that those of greater value per ton increase later, then the change in product mix will result in a declining COR (along, of course, with the increased rate of output). The pattern Of steel consumption in the course Of economic develOpment indicates that this may happen. Steel 115 consumption tends to shift with develOpment from lower value per ton rails and structural to the higher value flat rolled products.1 Of course, if facilities for the production Of higher value per ton goods are added to the plant, then the effect on the COR depends on the relationship between the increase in capital and the increase in output (value added). For any individual plant the change in product mix involves a combination of both effects--varying rates of growth to capacity output and increase in productive ability. For CSN the "product-mix-effect" appears to be mild and is a result largely of a shift away from rails and accessories to hot rolled sheet. This would follow the generally expected developmental pattern Of trends to rolled products. The AHMSA plant was designed specifically for the manufacture Of flat rolled products. The product-mix-effect in this case was due largely to a great increase in the share Of tin plate in the total output. This is a phenomenon usually accompanying the growth Of domestic agricultural output and the increased use Of canned and tinned fruits, vegetables, and.meats.2 Judgments concerning effects of the change in product mix for the Chilean plant are more difficult to make. As in the other cases, there has been a shift in production to flat products. But counteracting this has 1United Nations, Long Term Trends and Problems of the EurOpean Steel Industry, pp. IO-I2. 2Ibid., p. 17. 116 been an increase in the sale of semifinished products. Moreover, a large prOportion Of the tonnage sales of CAP are exports; the export price differing from.the domestic price. It seems that much, if not all, of the apparent decline in the CORs of the three steel plants are a result of price changes and depreciation policy. If the values Of output and asset stock are adjusted for price changes and reassessed depreciation, much of the decline disappears. Furthermore, if value of output is adjusted for changes in the product mix, the COR of AHMSA no longer appears to decline. If this were generally the case for steel plants, an evaluation Of the COR of a planned steel plant need not be complicated by anticipating the possibility Of a declin- ing COR, once the product mix was given. Unfortunately, the pattern Of behavior Of the COR of CAP is not clear because Of the extreme variations in the prOportion Of finished and semifinished products in tonnage sales and because Of changes in the prOportion Of export sales. In this case, the product mix changes have resulted in an upward movement Of the COR (since fiheir elimination increases the lepe of the trend line). None of the adjustments of capital and/Or output will result in the elimination of the downward trend of the COR of CSN. However, it seems reasonable to assume that the decline cannot continue indefinitely as the firm grows. 117 Interfirm.Comparisons of CORs Expected Findings If the COR is a measure of capital intensity, it could be hypothesized that industries in underdevelOped countries would have lower CORs than their counterparts in develOped countries. Using CORs as an indicator Of capital intensity, Bhatt found that empirical evidence does not support this hypothesis. He found that the capital coeffi- cients (COR) of at least one underdevelOped economy are higher than the corresponding capital coefficients Of at least one develOped economy. If the CORs in underdevelOped countries are lower than those in develOped countries, the differences between them.are not very great.1 It is not necessary to discuss in great detail the technical limitations of Bhatt's study,2 when the more important question is whether his findings should be as surprising as he seems to feel they are. If carefully considered, the conclusion.must be made that they are not surprising in the light Of general economic considerations. Besides these considerations, certain characteristics of the 1v. v. Bhatt, "Capital-Output Ratio of Certain Industries: A Comparative Study of Certain Countries ” Review of Economics and Statistics, XXVI (August, 1958), ZSuffice it to say that shatt uses the ratio of book value Of plant and equipment to gross value Of output for his COR. The use Of such a measure, given the effects Of price changes and depreciation on the COR, make such inter- national comparisons almost meaningless. 118 steel industry may lead us to expect an inverse relationship between the degree of develOpment of the economy and the COR of the steel industry. Getting back to Bhatt's findings, if the situation in an underdevelOped country were such that the ratio of the price of capital to the price Of labor were higher than the corresponding ratio in a develOped country, firms in the develOped country would be using a higher ratio of capital to labor than firms producing the same product in the develOped country. If it is accepted that the price ratios actually differ in this manner, the question still remains as to whether a higher COR actually indicates a higher capital-labor ratio. The answer to this question depends on the nature of the production function, and there is some reason to believe that it should be negative. If the industry were one Of economies of scale, even if it were assumed that the capital-labor ratios were identical, the CORs of the two firms need not be identical. Clearly, the larger firm would have a lower COR. If we assume economies of scale, different factor price ratios, and different size firms, the extent to which the COR is a measure Of the capital-labor ratio depends upon the substitutability of labor for capital; the closer the elasticity Of substitution is to one, the more closely the capital-labor ratio will be measure by the COR. The point Of all this is that the COR is not necessarily a measure Of capital intensity. Even if a high COR indicated a high capital-labor 119 ratio, such a finding need not be surprising. The argument for use of labor intensive techniques rests on the pre- sumption.that the relative domestic prices of capital and labor dictate the use of such techniques. However, the presumption may not be a useful guide. To the extent that investment projects are financed with borrowed capital and built with imported machinery and equipment, the factor pro- portions employed will depend on the cost Of capital in the lending country rather than on the domestic factor prices. Furthermore, what seems to be a low price Of domestic labor in relation to that of domestic capital may be deceiving because of the lack of experienced and technically skilled workers. According to one expert on the steel industry, this shortage of skills can best be Offset by the using of methods and processes in which the Operations requiring skill are entrusted to machinery. Skill in.maintenance, mechanical, and elec- tric work seems to be more easily found or develOped than that required for complicated manual Operations, 1 such for example, as in the sheet or tinplate rolling. Along similar lines, Hirschman has pointed out that production functions may not be the same in develOped as in underdevelOped countries: The marginal rate of substitution of labor for capital is larger in underdevelOped countries as with the loss in managerial and labor efficiency consequent upon the adOption of less capital-intensive methods, more labor 1Haven, "Selection of Steelmaking Processes and Locations for Integrated Iron and Steel Works," A Stud of the Iron and Steel Industr , II, 356. Other autEoriEies seem to hold that manual s ills are more easily develOped than maintenance skills. See especially Hirschman, 222 Strategy Of Economic Development, pp. 139-82. 120 is needed than in develOped countries to make up for a given decrease in capital. . . . With this situation . . identical, relative factor prices should result in the adOption of more capita1-intensive processes in the underdevelOped country than the industrial one.1 These factors would all gravitate against the use of more labor-intensive production processes in underdevelOped countries and would thus help to explain why findings such as Bhatt's might not be unexpected. If the analysis is shifted from the general to the particular, as in the case of steel, more possible reasons can be found for deviations from the "expected" direct relationship between degree Of "underdevelOpment" (as it supposedly reflects the relative availability of labor and capital) and the size Of the COR (as it supposedly measures the capital intensity of a particular industry or firm). If the steel industry were one of economies of scale, then an inverse relationship between size and the COR would be indicated. DevelOped economies, with larger markets would have larger plants. Therefore, we would find an inverse relationship between level of economic develOp- ment and COR (or capital intensity in Bhatt's terms). Several studies declare that the steel industry is one of "economies Of scale." However, the evidence Of this gener- ally involves a comparison of capacity of finished steel production.and cost Of production per ton of finished steel, 1Albert 0. Hirschman, "Investment Policies and Dual- ism.in UnderdevelOped Countries," American Economic Review, XLVII, No. 8 (September, 1957). 567- 68. 121 showing a decline in the latter as the scale of the plant is increased. These conclusions do not prove economies of scale because they do not assume constant factor prOportions. But the estimates presented in the report of the United Nations Expert Working Group for production costs in plants of different sizes do indicate an inverse relationship between COR and size Of firm.1 This decline is, at least in part, due to a substitution of capital for labor and a ris- ing productivity Of labor. This manifests itself in the estimates by declining capital and labor costs per ton as the size Of plant increases. .Another factor that may account for possible dis— parity between factor price ratios and the COR is the fact that the economics of steelmaking is influenced in large part by the quality and costs of raw materials--especially iron ore and coke. The capital-output ratio of a particular steel plant is determined, at least to some extent, by the steelmaking process used and unless there are several kinds of ore to choose from.there is only limited Opportunity to pick alternative processes. The integrated steelmaking plant involves three steps: manufacturing pig iron, refining pig iron into steel, and rolling the finished products. The process used for conversion Of ore into pig depends largely on the avail- ability Of coke or coking coal. Pig iron or an acceptable 1A study of the Iron and Steel Industry, 1, 113-16. 122 substitute can be manufactured using blast furnace tech- niques or any of several electric or "direct reduction processes." The latter are designed primarily for use in areas without coal suitable for coking. As can be seen from Table 9, the labor and capital costs per ton vary considerably according to the process used. The suitability of various techniques for the con- version Of pig iron into steel ingot is determined largely by the phorphorus content Of ore as phosphorus is not removed in the reduction of iron ore to pig iron.1 Charles Ramseyer prepared the following estimates Of investment costs per ton Of steel ingot for various steelmaking processes presented in Table 10.2 The data available do indicate considerable variation in the invest- ment cost per ton Of the various processes. The extent of mechanization in the final step of 1Haven, "Selection of Steelmaking Processes and Locations for Integrated Iron and Steel Works," A Study of the Iron and Steel Industry, II, 387. 2The acid Bessemer process can be used with ore whose phosphorus content does not exceed 0.08 per cent. The basic Bessemer process requires a minimum phosphorus content Of 1.75 to 2.00 per cent. The basic Open hearth.uses prac- tically any combination Of scrap and pig iron with a phos- phorus content up to l per cent. Basic electric furnace processes are generally used for scrap melting with charges of pig iron, possibly up to 80 to 50 per cent. Duplex processes are used in cases where the phosphorus content is too high for basic Open-hearth or electric furnaces, but not high enough for basic Bessemer. Here the making of two slags for prOper dephorphorizing is required. Charles F. Ramseyer, Comparative Investment Costs for Different Steel- making Processes," A Study of the Iron and Steel Industgy, II, 316-32e 123 Table 9.--Comparison of Costs Per Ton Of Production of Pig Iron.by Different Processes Technique It°m' A s C D E Assembly costs $29.57 $29.57 $27.91 $28.72 $19.85 Other 0.65 1.35 0.65 23.80 11.80 Wages and salaries 2.79 5.50 8.01 7.60 6.25 Depreciation 9.12 10.86 12.63 5.82 8.00 Total cost per ton 80.83 87.88 85.15 60.58 81.10 Installation cost (million dollars) 28.50 19.00 22.10 8.00 2.25 Note: A. B. C. D. E. Source: ' United Nations, Department of Economic.Affairs, World One blast furnace, American practice, 800 tons/day. One blast furnace, American.practice, 800 tons/day. Two blast furnaces, Swedish practice, 200 tons/day each. Three TyslandrHole Electric furnaces, 100 tons/day each. Two Wiberg Electric furnaces, 68 tons/day each, Iron Ore Resources and their Utilization (New YOrk, I987). 128 Table lO.--Estimated Plant Capital Investment Cost Per Ton Required to Produce Various Tonnages of Ingots Per Year by Different Processes Tonnage Technique For your A s C D E r 100, 000 $39.6 $59.18 $86.76 $----a $33. 03 $86.73 250, 000 23.7 36.82 37.71 19.93 86 .25 27.36 500,000 16.78 28.93 30.58 20. 01 38.09 22.82 1,000,000 ----a 19.81 28.02 17.63 31.16 18.50 aSize not apprOpriate for this technique. Note: A. Acid Bessemer process with auxiliary electric furnaces. B. Basic Bessemer process with.auxiliary electric furnaces. C. Standard stationary basic Open-hearth.process. D. Basic electric furnace process. E. Acid Bessemer--basic Openshearth, tilting furnace, duplex process. F. Acid Bessemer--basic electric furnace, duplex process. Source: Charles F. Ramseyer, ”Comparative Investment Costs for Different Steelmaking Processes,” A Stud of the Iron and Steel Industr in Latin America, UniEed Naiions, Department OinECOnomic 3/Rev. l/Add. 1, November, 1958) (New YOrk, 1958). p. 3 26 125 integrated steel production, rolling, is greatly influenced by the size of the market. Table 11 indicates a range Of types of mills for the rolling Of flat sheet and the output range in tons per year of each. The trend in rolling mills has been toward continuous, high-speed electrified units.1 The main advantages Of continuous and semicontinuous mills are "the superior quality Of their product, a better yield for higher productivity as a result Of almost complete mechanization Of all Operations, and reduced costs."2 A continuous mill is one in.which a slab or bloom is reduced in thickness by "ironing" continuously in the same direction through a long series Of rolls. The ratio of roller speeds must be maintained precisely, and all other Operations in the plant must be geared to the mill. It is not possible to build a small continuous mill.3 Thus, the COR found in a particular steel plant cannot be expected to reflect merely the relative factor prices in that country. It is also influenced strongly by the extent of the market, the size and nature of the resource base, and the nature of steel technology. This assumes a given state Of the arts. However, the nature Of steel technology is subject to change through 1United Nations, World Iron Ore Resources and their Utilization, p. 37. 2United Nations, Long Term.Trends and Probloms Of the European Steel Industry, p.7108. 3Pounds, Geoggaphy of Iron and Stee , p. 17. 126 Table ll.--0utput Range of Different Types of Mills Rolling Flat Steel Output range (thousands Type of tons of hot rolled coils per year) Mechanized sheet and tin plate mills 80- 60 Steckel hot strip mill and reversing cold mill 120- 350 Semicontinuous hot strip mill and reversing cold mill 300- 500 Semicontinuous hot strip mill and tandem cold mill 500- 800 Continuous hot strip mill and tandem.cold mill 1250-2500 Source: United Nations, Department of Economic Affairs, Economic Commission.for Europe, LO Term Trends and Problems of the Euro an Steel Industr (ST7ECE7STEEI7I, IpriI, I930) 611678 , g p e e 127 innovations which may make it more adaptable to the market and resource conditions Of underdevelOped countries. Steel technology represents an independent variable as far as current alternatives are concerned but may be a dependent variable in longer run considerations. Actual Findings The next step in the study should be to compare the CORs Of the firms with the degree of develOpment of the three economies involved. It is difficult to take three countries, all generally considered "underdevelOped,“ and rank them.according to their degree of develOpment. This is especially true in.the case of Brazil, Chile, and Mexico, all in the forefront Of Latin American economic growth. Using the traditional measures of economic develOpment-- per capita gross national product, caloric intake, etc. (Table 12), it seems that Chile is the most develOped of the three. (It has the highest gross national produce per capita, the highest caloric intake per day, the highest urban pOpulation, the highest energy and steel consumption per capita, and the smallest prOportion Of the pOpulation in agriculture.) If it is assumed that the ratio Of the price of labor to the price of capital is a function Of the degree of develOpment (the greater The degree of develOpment the higher the ratio), Chile's steel plant would be expected to have a higher COR than either Brazil or Mexico, assuming, of course, that the COR is a measure Of capital 128 Table 12.--Selected Measures Of Economic DevelOpment: Brazil, Chile, and.Mexico Measure Brazil Chile Mexico Gross national product er capita, 1952-58 (U.S. dollars Caloric intake per capita per day, 1958 Per cent Of pOpulation in agricul- ture, 1958 Per cent Of pOpulation in urban areas, 1955 Energy consumption per capita (megawatt hours) Per capita steel consumption (kilograms per year) 230 360 220 2585 2682 2260 58 30 58 20 81 28 5.7 7.7 6.8 25 58 31 Sources: GNP--Statistical Office Of the United Nations, Per 37 Ca ita National Product Of Fift -five Countries, 1952-1958 (3T7STIT7SERE7H) (New York", I957) Others--Norton Ginsberg, Atlas Of Economic Develo nt (Chicago: I96I). University of Chicago Press, : .r limufl. aJfls . .‘ .Ffi... .fi .. .K . (Gn‘ mum. F . - .31.... Us r. .I‘ 129 intensity.1 If the steel plants' CORs are compared using Bhatt's naive formulation (ratio Of net book value Of fixed assets to sales), the ratios found are: 0.35 for CSN, 0.58 for CAP, and 0.66 for AHMSA. This would not be as expected. However, the above finding is only Of curiosity value because Of the methodological limitations of Bhatt's technique. If the CORs are compared on the basis of values in prices of the same base year, using value added instead of sales and a common basis for depreciation, an altogether different picture emerges. Whether the COR is considered to be the mean value of the years after break-in, the lowest value Obtained in any one year, or the value Obtained for the last year of the study, CAP still shows the highest COR. AHMSA shows the lowest. This would be an agreeable finding2 if the OCR were a measure of capital intensity rather than a measure of the average productivity of capital. However, it is possible, as previously indicated, that a high capital-output ratio for a firm (low average productivity of capital) does not reflect high capital intensity. There are other factors which do a better job in explaining the relative levels of the CORs. 1Itmust be remembered also that the Chilean plant was planned and constructed about five years after the other two. Therefore, Chile would have been five years more develOped than the other countries. This would reinforce the expectations concerning the size Of the COR. 21: it is assumed that Brazil and Mexico are at approximately the same stage Of develOpment. 130 Table l3.--Comparison of the Average, Post Break-in Capital-Output Ratios Of CSN, CAP, and AHMSA, Values and Ranks M CSN CAP AHMSA Mean Values GFA, 1950/VA, 1950 5.98 5.26 3.93 Rank 1 2 3 NFA, adj, 1950/VA, 1950 8.18 8.38 2.99 Rank 2 1 3 Lowest Values GFA, 1950/VA, 1950 5.30 8.68 3.36 Rank 1 2 3 NFA, adj, 1950/VA, 1950 3.26 3.96 2.52 Rank 2 l 3 Values in Last Year GFA, 1950/VA, 1950 5.30 7.35 8.16 Rank 2 l 3 NFA, adj, 1950/VA, 1950 3.26 5.67 2.95 Rank 2 1 3 sourgziculated from.data in Tables 8, 5, and 6. 131 Several other factors would lead one to expect that CAP would have a higher COR than the other two firms. The first involves technical conditions of production. The ore supplied to the Huachipato plant by agreement with Bethlehem Chile Iron.Mines has a slightly higher phosphorus content than can.be handled in an Open-hearth furnace, but lower than that needed for basic Bessemer steelmaking.1 Because of this, the CAP steelmaking facilities were designed for an acid Bessemer basic Open-hearth duplexing Operation. As indicated in Table 10, this duplex process requires a higher investment per ton of output than any other steelmaking technique. Further, the highest investment cost per ton using this technique is in the approximate output range of CAP (250,000 tons per year). It is difficult tO draw any conclusions with respect to the relationship between size and the COR. ‘Using the depreciated fixed asset figures, CAP has the highest ratio. It is also clear that from the point of view of ingot capacity CAP is the smallest of the three plants (approxi- mately 830,000 tons per year as compared with 600,000 tons for AHMSA, and about 1,000,000 tons for CSN in 1960). CAP is also the smallest plant from the point Of view of prO- duction of finished products. This seemingly clear picture is complicated by some Of the following factors: First, lHaven, "Selection of Steelmaking Processes and Locations for Integrated Iron and Steel Works," A Study Of the Iron and Steel Industry, II, 387. 132 while CAP is the smallest firm and has the highest COR, CSN, the largest firm, does not have the lowest COR. Second, while CAP was the smallest firm in the later years of the study, during the period 1952-57 finished production at CAP exceeded that of AHMSA. The rolled ingot capacity of CAP was higher as well. Third, the picture is clouded by the fact that since 1955 Chilean demand for steel has stagnated, and the CAP plant has been Operating below capacity. This automatically makes for a higher COR for the firm. The smallness of the CORs for AHMSA may be partly a reflection Of the fact that the original capital equipment consisted of rehabilitated scrap.1 It could also be a reflection of the difference in product mix between AHMSA and the other two firms. The Mexican plant was designed to manufacture only flat rolled products. The other plants devote a considerable portion Of their capacity to the manu- facture of unrolled products, eSpecially rails and struc- tural shapes. For this to affect the COR in the manner suggested, it would be necessary to assume that flats have a lower COR than non-flats (they do have a higher value per ton). Since data on employment for the three firms are sketchy, at best, it is not possible to say precisely 1This by itself would not result in a lower COR if the lower price of this equipment reflected accurately its lower productivity. It is difficult, however, to evaluate the increase in the real value of scrap equipment that results from its rehabilitation. 133 whether the declining COR evidenced by CSN and, to some extent, by AHMSA is a result of "economies of scale" or is a result of an increase in size coupled with a change in factor prOportions. However, partial data do indicate that there has been an increase in the capital-labor ratio for both AHMSA and CAP.1 The declining COR Of CSN suggests the intriguing possibility that the ratio for this firm may eventually fall below that of AHMSA; the gap between them.has been.narrow- ing. This would leave an inverse relationship between the sizes of the firms and their CORs. It would also leave an approximately direct relationship between the COR and the degree of develOpment of the economy. This would leave us with the conclusion that the COR varies inversely with the size Of the firm.snd directly with the degree of develOpment of the economy. This is a somewhat paradoxical situation for we would normally think of the most develOped economy supporting the largest steel market and, therefore, the largest firm. In this case the reverse is true; the small- est plant is in the most highly develOped country--Chile. This clearly indicates that the "degree of develOpment" actually involves two distinct effects: first, it influ- ences the factor price ratios; and second, it determines the 1The capital-labor ratio for CAP increased from 0.579 to 1.120 million pesos per worker over the period of 1951 to 1959. The capital-labor ratio Of AHMSA increased from.%;.8 thousand pesos per worker in 1986 to 65.1 thousand in 19 . 13h size of the market. Although Brazil and.Mexico are less well developed than Chile, they clearly have larger demands for steel (in the same way that, say, India has a larger steel market than Canada). On the whole, then, it would seem.that the size of the market and size of the firm.have a greater predictive ability insofar as CORs are concerned than the largely intuitive reasoning about factor price ratios. Conclusion The CORs of all three firms when calculated on the basis of book value of fixed assets decline over time. This finding is common to almost all studies of industrial CORs. The reasons generally expressed for these declines have been shown not to be entirely satisfactory. Furthermore, it has been shown that these explanations seem to become even less satisfactory when the additional relationship between the COR and size of the firm is considered. On the whole, there does not seem to be any a priori reason to assume that the COR of any one firm.will increase or decrease over time, assuming that depreciation, price changes, and changes in the product mix are adjusted for. In the case of the three firms in this study, adjustment for depreciation, price changes, and product mix changes do, in fact, substantially reduce the rate of decline in the COR, after the sharp initial decline of the "break-in" period. The decline in the COR of AHMSA 135 disappears entirely. However, even using adjusted data, the COR of CSN declines over time. The fact of this decline must be taken into consideration in any further evaluation of the firm. 'Unfortunately, the operations of CAP have been quite variable, and the firm.has been Operating for a shorter time so that the results are not clear enough to draw any conclusions with regard to the trend of the COR. This study suggests several important qualifications to the often expressed hypothesis that we should expect to find lower CORs for any industry in an underdevelOped country than would be found for its counterpart in a more developed country. This hypothesis is based largely on the intuitive belief that the relative factor prices in underdeveloped areas would favor labor-intensive technology. This argument is then extended to the favoring of projects with low CORs on the presumption.that these necessarily have low capital- labor ratios. This line of reasoning has been shown to be fallacious on both counts. First, the domestic factor price ratios may not be as commonly assumed. Second, the domestic price ratios, even if as assumed, may not be appropriate in the case of borrowed capital. Third, labor-intensive projects may not necessarily have low CORs; this depends on the nature of the production function. In the case of the steel industry, the possibility of economies of scale might lead to lower CORs in countries with bigger steel markets and larger firms. Furthermore, the COR of a steel firm.depends on the particular production 136 process being used. This, in turn, is dependent upon the resource base and the size of the market. It was found that the Chilean firm.had the highest COR. This would seem to indicate a direct relationship between degree of development and COR. However, this relationship is complicated by several other factors. First, some technological considerations would lead us to expect a higher COR for this firm. Second, it is the smallest of the three firms. Third, this firm.has been operating below capacity. There is some evidence of a relationship between.the COR and the size of the firm. The COR of the Brazilian firm may, in the future, be lower than that of AHMSA. This would leave the largest firm.with the lowest COR and the smallest firm, CAP, with the highest ratio. This is, however, specu- lation because only data on future Operations of the firms will verify this. CHAPTER VI EVALUATING THE STEEL INVESTMENT BY COMPARISON WITH ICORS FOR THE ECONOMY'AND MANUFACTURING SECTOR Technical Considerations Since there is considerable doubt about the useful- ness of the COR as an indicator of factor intensity, one should not slip into the error of assuming that advocating investment in low COR projects necessarily implies labor- intensive rather than capital-intensive projects. Moreover, there is reason to believe that the common assumption con- cerning factor price ratios in underdeveloped areas may be misleading. Use of the COR as a criterion in investment allocation implies only that a given amount of funds ought to be invested in projects where its average productivity is high (COR is low) rather than low. This in no way implies that the COR need be the sole criterion used, but if the availability of capital is limited, it should at least be considered. We still have not completely come to grips with what is surely the most troublesome aspect of the use of the COR as an investment criterion. This is that the COR measures the relationship between input and output at a single moment in time, but most investments involve returns spread over 137 138 several years. This requires discounting of future benefits when investments of different life spans are being con- sidered. The present value of output should be calculated as: Y1 Y2 Yn -___+—-—-§ l + r (l + r) where V is the capitalized value of the stream of income Y at rate of interest r. If two alternative investment projects have equal life spans, however, then the COR is an acceptable substitute for a present value formulation. In this study we are not only comparing three different investments but are also measuring the COR of the enterprise against the ICORs for the economies as a whole. If the life span of the steelmaking facilities does not differ appreciably from the average life span of fixed capital in the economy, the time dimension offers no diffi- culty and the COR can be substituted for a present value formulation. It is still not clear exactly which formulation of the ICOR is apprOpriate for our purposes. The most common COR formulation is the ratio of net domestic capital for- mation to net national product. This shows the relationship between increases in the fixed capital stock (assuming depreciation shows the loss of productive ability of the capital stock) and the consequent increases in income net of depreciation. This is probably the most significant formulation of the ICOR if it is being used for estimating 139 capital requirements. However, if we wish to assume that the ICOR actually represents an hypothetical "average" investment, a new problem is introduced. The important factor in evaluating a new investment is its total capital requirement and not some hypothetical "net" figure.1 There- fore, the gross investment figure rather than the net investment seems to be appropriate in this case. Not only is this measure more appropriate, but it is also empirically more satisfactory. First of all, there is the distinct possibility that depreciation exceeds the withdrawal of capital from productive use,2 so that the figure for net domestic capital formation would be underestimated.3 Second, the use of gross capital formation figures would allow the use of the gross capital stock figure for the steel plants, thus avoiding any errors that may have been introduced into the analysis through the revaluation of depreciation. There is some question.whether output or value added should be considered in gross terms (including interest and 1As a matter of fact, if depreciation actually mmasures a deterioration of productive ability (and if we assume fixed prOportions), the gross and net capital-output ratios should be the same. If, say, a 5 per cent decline in the capital stock results in a 5 per cent decline in output, COR remains unchanged. aFor a discussion of the Philippine experience in this regard, see Higgins, Economic Development, pp. 651-53. 3This is why a calculation of net domestic capital formation divided by changes in net domestic product, 1950 prices, for Chile resulted in an ICOR of O.h6 for the period l9hO-56. This can be compared to a gross ratio of 2.50. 1&0 depreciation) or in net terms, excluding depreciation. The exclusion of depreciation from value added in the comparison of CORs compensates to some extent for the difference in productive lives of alternative projects. Projects with shorter life spans have greater annual rates of depreci- ation. Subtracting depreciation from.value added will, therefore, raise the COR. However, it is possible that depreciation allowances may be reinvested in new capital equipment (if the rate of depreciation exceeds the actual wearing out of equipment). In this case, it would be sufficient to deal with the gross figure, but, as indicated previously, the use of the net value added figures would introduce certain additional statistical difficulties. In any case, the COR of the firm was calculated using both.the net and gross value added figures, and these were compared with ICORS for the economy calculated using changes in net domestic product and gross domestic product, respectively. The results of these comparisons are presented in Tables 1h and 15. Com arison with the Incremental Ca ital- Output RatIo for the Economy The comparisons of the capital-output ratios of the companies with the ratios of domestic capital formation to changes in the domestic product have been.made on the basis of three values of the firm's COR: the mean value after "break-in," the lowest value in any year, and the value in the last year of the study. To facilitate comparison, each lhl Table lh.--Comparison of the Gross COR of the Steel Plants, Post Break-in Period, with the ApprOpriate ICOR for the Economy, 1939-N9 Brazil Chile Mexico Incremental capital-output ratio for the economy 2.56 2.50 2.u5 Capital-output ratio of the company CSN CAP .AHMSA Mean value 5.9h 5.26 3.93 Lowest value 5.30 h.63 3.36 Value last year 5.30 7.35 h.l6 Ratio of company COR to ICOR for the economy Mean value 2.28 2.10 1.60 Lowest value 2.07 1.87 1.3? Value last year 2.07 2.9h 1.70 Note: The gross COR for the steel plants is the ratio of gross fixed assets to value added in 1950 price. The gross ICOR for the economy is the ratio of gross domestic capital formation to the change in gross domestic product, in 1950 prices. Source: Capital-output ratios for the firmm--Table 12. Incremental capital-output ratios--calculated from data in Appendix Tables h, 5, and 6. 1&2 Table 15.--Comparison of the Net COR of the Steel Plants, Post Break-in Period, with the A prOpriate ICOR for the Economy, 1939- 9 Brazil Chile Mexico Incremental capital-output ratio for the economy 2.6h 2.72 2.66 Capital-output ratio of the company CSN CAP AHMSA Mean value 7.87 6.99 h.67 Lowest value 6.36 5.37 3.89 Value last year 6.88 10.30 h.99 Ratio of company COR to ICOR for the economy Mean value 3.02 2.57 1.72 Lowest value 2.ul 1.98 l. 6 Value last year 2.h9 3.80 l. Note: The gross COR for the steel plants is the ratio of gross fixed assets to value added minus depreciation, 1950 prices. The net ICOR of the economy is the ratio of gross domestic capital formation to change in the net domestic product. Source: See Table 1h. 1&3 COR of the three firms has been expressed as a percentage of the appropriated ICOR for the economy. The ICOR for the economy is based on the weighted average of the years 1939- 59 for Brazil and.Mexico and l9hO-56 for Chile. These periods include practically the entire time of planning and production for all three plants. Average figures were used because of the wide variation generally found in ICOR figures from.year to year. The gross ratio comparisons in Table 1h indicate that the lowest value of the COR of CSN was 107 per cent above the ICOR of the Brazilian economy. This could be broadly interpreted as indicating that the increment unit of output due to the investment in the steel plant used 107 per cent more capital than the average increment of output in the economy.1 Since the COR for CSN is declining, the last year's ratio was also the lowest. The mean COR of CAP was 110 per cent above the aver- age Chilean ICOR between 19h0 and 1956. It was only 87 per cent above the Chilean economy's COR if the lowest value of the COR is used. However, the lowest value in this case was achieved in the first year of Operation and never approached again. The spread between the COR of the steel plant and 1The average ICOR for the economy includes a AC and atLI from.the steel investment. To the extent that the ICOR of the steel plant exceeds that of the rest of the economy, the ICOR for the economy as a whole, including steel, will be greater than if capital and output increments for the steel plant are excluded. The ICOR for the economy has been calculated both ways but the resulting ratios are not appre- ciably different. um the ICOR of the economy was smallest in Mexico. The mean COR of AHMSA was only 60 per cent above the economy's ICOR. The spread was as low as 37 per cent in the year in which AHMSA's ICOR was lowest.1 As can be seen from a casual comparison of Tables 1h and 15, the investment in the various steel plants appears more favorable when the comparison is on a gross basis than when the net value added figures are used. This is true even though depreciation has been re-evaluated. Because CSN has the highest COR, it is most affected by the use of the net COR. The lowest value of this company's COR is now lul per cent above the economy's net ICOR where it was only 107 per cent above on a gross basis. Use of net rather than gross comparison has less effect on the appraisal of CAP and still less on that of AHMSA. Com.arison with the Incremental Ca ital-Output Ratios of tEe Hanfifacturing Sector It is well known that there is a wide variation range of ICORs from.aector of the economy to another.2 1It‘would probably be apprOpriate to discuss the factors determining the level of the ICOR for any economy, but not enough is known about this to permit a full-fledged analysis. It is more than interesting to note, however, the similarities in the ratios found for Brazil, Chile, and Mexico. 2For instance, Simon Kuznets found that during the period l922-h8 ICORs for the American economy ranged between 0.03 for mining and 1.31 for manufacturing (Capital in the American Economy, . 199). Benjamin Higgins in his Economic ‘fievelo ment (p. 6kg) quotes a United States Department of State Report, "Intelligence Report NO. 7607" (Washington, ILLS Therefore, it would be informative if the COR of the steel plant could be ICOR of an apprOpriate sector. The apprOpri- ate sector in this case would be manufacturing. Unfortu- nately, there are little data available covering annual capital formation in the manufacturing sectors of the economies with which.we are concerned. The data for the Chilean economy seem.more complete than that for the other two economies. The Economic Come mission.for Latin.America has published some data on capital stock and output in 1950 prices for the manufacturing sector of the Chilean economy between 1915 and 1952 (Table 16). Manufacturing was defined as including mining and con- struction. Using these data, the ICOR for this sector has been calculated for 19h5-52 as 1.9h. This would be the ratio of net capital formation to the increase in the gross product in this sector over the period. Since the ICOR for the manufacturing sector was calculated using net capital formation, an equivalent ratio has been calculated for the company. This is the ratio of net fixed assets,adjusted depreciation to value added. The lowest value of this COR was 3.96 which was 10h per cent higher than the sectoral D.C., 1958), as finding the following ICORs for various sectors of the American economy: Sector ICOR Trade and Services 0.1 or less Manufacturing 1.3 to 2.0 Housing and Public Utilities 10 or more lh6 Table 16.--Chile: Capital Stock and Gross Product, Mining, Manufacturing, and Construction, l9h5-52 (in 1950 prices) .__~._.. .. __ . ._ 'Y Gross rOduct Ca ital stock ear (EIIlions of pesos) 19h5 29.u uu.9 19u6 33.3 A7.3 19h? 33.8 50.5 19h8 3h.8 52.7 19h9 36.1 57.1 1950 36.1 61.6 1951 37.h 6%.9 Source: United Nations, Department of Economic Affairs, Economic ve of Latin Commission for Latin America, Economic Sur AmericaI 1951-1952 (E/CN.12/29I7Rev. 2) (N Y E I953), p. . ew or , 1&7 ratio.1 The mean value h.38 was 126 per cent above the ICOR for the manufacturing sector. In ranking the three countries according to avail- ability of data, Mexico would follow Chile. An estimate of the ICOR for Mexican manufacturing was derived from data on investment in machinery and equipment in the.manufaoturing sector and census value added estimates.2 The estimate of investment did not include an estimate of construction. Using the data on investment in.machinery and equipment, an estimate was made of total investment in fixed capital items. This estimate was based on the Brazilian experience as to the per cent of fixed capital formation of enterprises spent on construction during the period 19h7-56, an average of 39 per cent.3 This estimate was then deflated using the whole- sale price index along with the estimates of value added. The resulting sectoral ICOR was 2.92 (see Table 17). In order to check this estimate of capital, another estimate was made using census data on capital stock. 1The lowest COR for the company, not adjusting depreciation, was 3.51, or 81 per cent higher than the sectoral ICOR. 2Combined Mexican Working Group, The Economic Devel- O ent of Mexico (Baltimore: JOhns HOpkins University Pgess, I955), p. 171. For investment data value added by manufacture from Mexico, Direccion General de Estadistica, Censo Industrial de los Estados Unidos MexicanosI 19%0 ex co, D. ., , and Mex co, 8 ados n dos ex canos, Quinto Censo Industrial 1 Tercer Censo de Transportes, 1950 D. 0’ O THexico, BBrazil, Conselho Nacional de Estatistica, Boletina Estatistico, No. 61 (January-March, 1958), p. 27. 1&8 Table 17.-~Mexico: Investment and Value Added in the Manufacturing Sector, 1939-50, in Market and 1950 Prices m Investment in Estimated Total invest- A. Year machinery and total ment in 1950 equipment investment prices (millions of pesos) 1939 171 280 89& 19&0 160 262 796 19&1 186 305 86 19&2 120 197 50 19&3 208 3&1 729 19&h 338 55 967 1985 5&8 89 05 19&6 839 1375 1 71 19h? 1217 1995 2561 19&8 1279 2097 2511 19&9 116& 1908 2083 1950 1205 1975 1975 B. 'Year Value added Value added in 1950 prices 1939 1001.0 3793.8 1950 8793.8 8793.8 Sources: Investment data--Combined.Mexican'Working Group, The Economic Develo ment of Mexico (Baltimore: JOhns HopEIEs University ress, , p. 1. Value added--Mexico, Direccion General de Estadistica, Censo Industrial de los Estados Unidos Mexicanos, 19&0 (Mexico, D.F., 19521, andfMexico, Estados Unidos Mexicanos, Quinto Censo Industrial Tercer Censo de Transportes,gl9§0 (Mexico, 5.F., I957). 1&9 Census data indicated an increase in the capital stock from 2,287 million pesos in 1939 to 12,879 million pesos in 1950. However, these figures are net book values. In order to convert the change in book value to a change in 1950 prices, the following estimation technique was used: AC ._._ Total investment in manufacturi 19 O rices TUUEI investment in manufacturifig, maréet prIces (Fixed assets, census 1950 - Fixed assets, census 1939) "Total investment in manufacturing, market prices" is the sum of the annual investments in.manufacturing, 19&0-50, valued at their market prices as presented in Table 17. "Total investment in manufacturing, 1950 prices" is the sum of deflated value of these increments. In effect, what this gives us is the reciprocal of an average price index for the period, base 1950, which has been weighted by the volume of investment in each year. The increase in the census value of assets is then multiplied by this reciprocal to obtain the value of the increase in 1950 prices. Using this tech- nique, net investment in manufacturing was estimated at 1&,511 million pesos. This would yield a net ICOR for the sector of 2.60. Comparing the estimated ICOR for manufactur- ing with the apprOpriate COR for Altos Hornos, it seems reasonably clear that the capital associated with each unit Of output for Altos Hornos was probably not substantially higher than the average ICOR of the manufacturing sector of the Mexican economy. It is difficult to appraise the relative merits of 150 Table l8.--Comparison of the Adjusted Capital-Output Ratio for AHMSA, 19&7-58, with the Incremental Capital-Output Ratio of the Mexican Manufacturing Sector, 19&0-50 Ratio of COR for Estimated ICOR, Mean COR: AHMSA to ICOR manufacturing AHMSA for the economy Gross 2.92 3-93 1'35 Net 2.60 2.99 1-13 Sources: Manufacturing ICOR calculated from data in Table 16. Mean CORs for AHMSA from Tables 1& and 15. 151 the "gross" or "net" estimates of manufacturing ICOR. The net estimate is based on census data which aside from the usual shortcomings with respect to accuracy has the additional handicap of the lack of complete comparability between the 1939 and 1950 census. The 1950 census included all registered firms; the 1939 census included only those firms with sales greater than 10,000 pesos. Since annual data are not available, it would not be possible to deflate directly the census increases, even if an apprOpriate price index were handy. ‘Moreover, since the "deflation" Of the net index requires use of data from the gross index, the errors already there are compounded. Since the construction component of investment had to be estimated rather indi- rectly, the possibility for error is increased. 0n the whole, the latter (gross) approach is preferable. However, as indicated above, either estimate gives similar results insofar as the appraisal of the investment in a steel plant is concerned. There are no data available concerning annual capi- tal formation in the manufacturing sector of the Brazilian economy over the period in this study. However, as in the case of Mexico, there are census data available on the book value of fixed assets and value added for manufacturing establishments for the period 19&0-50.1 These data were 1The Brazilian census re orts capital invested (fixed capital) for the years 19 0 and 1950 and value added for 1939 and 19&9. Brazil, Instituto Brasileiro Geografia Economics, Recenseamento Geral do Brasil, 19&0, "Serie 152 used to estimate the ICOR for the manufacturing sector in a manner similar to the Mexican "net" estimate. It was assumed that annual capital formation in the manufacturing sector followed closely the annual pattern for the economy as a whole. The census reports showed an increase in book value of 31,966 million cruzeiros over the period 19&0-50. This was multiplied by the ratio of net domestic capital formmtion at 1950 prices to net domestic capital formation at market prices over the period. This was done to find the value of the increase in 1950 prices. The result was then divided by the difference between value added for the manu- facturing sector in 1939 and 19&& in 1950 prices for an ICOR of 2.37 for the manufacturing sector. The mean value of the net COR (NRA/VA) of CSN was about 77 per cent above this; the lowest value was about 37 per cent above the net sectoral ICOR. Nacional," Vol. III, Censos Economico (Rio de Janeiro, 1950), and Instituto Brasileiro Geografia Economics, Conselho Nacional de Estatistica, Resenseamento Geral do Brasil, Vol. IV, Censos Industrial (RIO do JaneIFo, I957). CHAPTER VII COST OF DOMESTIC AND IMPORTED STEEL Expected Findingg Given the international transferability of techs nology and of the skills required for production of steel, together with difficulty of adaptation to different capital- labor price ratios, the costs of domestic steel production ‘will ultimately be a function of the size of the market (economies of scale) and the location and quality of raw materials. The raw material base has not been a serious problem for the three countries of this study. All three are at least reasonably well endowed with iron ore. The Brazilian reserves are enormous, with.the deposits in Minas Gerais alone estimated at over a billion tons Of 55 per cent hematite.1 The Tofo, Algarrobo, and Romeral deposits in Chile have an estimated 200 million tons reserve. The Cerro de Mercado deposits in Durango, Mexico, contain.reserves estimated at more than 100 million tons.2 Unfortunately, none of the countries is well endowed with coal that can be converted into metallurgical coke. fir 1U.S., Tariff Commission, Iron Ore, p. 66. 2United Nations, world Iron Ore Resources and their Utilization, p. 302. 153 15h Lack of coking coal as in Norway and Sweden more often proves to be a greater handicap than lack of high-grade iron ore (as in Norway and Sweden).1 Chilean ore from.the Lots and Schwager deposits must be blended with imported coal (about 35 per cent) to produce a suitable coke. Brazilian coals from.Santa Catarina are also blended with.imported coals (about 50 per cent). Only the Mexican Sabinas deposits are of a quality such that no blending is required.2 Despite the coal handicap, it has been estimated that the raw material transportation require- ments for the three Latin.American plants are comparable with, if not superior to, North.American locations (Table 19). If, however, transportation requirements from the plant to the principal market are considered, then only the Chilean plant has a lower requirement than.Birmingham. Alabama, the closest American plant (Table 20). Actual Findings Estimates of the costs of production of finished steel in the three plants as compared with American pro- duction costs show higher totals for the three Latin American plants. These figures are, however, estimates based on 1Haven, "Selection of Steelmaking Processes and Locations for Integrated Iron.and Steel WOrks,” A Study of the Iron and Steel Industry, II, 352. 2George Bulle, "Steel Production in Latin.America,” A Study Of the Iron and Steel Industry, II, 302. 155 Table l9.--Comparison of Estimated Transportation Requirements for Coal, Iron Ore, and Finished Steel to Deliver One Ton of Finished Steel to Domestic Market, Domestic Steel Plant Compared with North American Plants (in railway ton-mile equivalents) Domestic markets Plant location Rio de Valparaiso, Mexico Janeiro Chile City Volta Redonda, Brazil 2009 n.a. n.a. Concepcion, Chile n.a. 1&2& n.a. Monterrey, Mexico n.a. n.a. 1839 Pittsburgh, Pa. 21193 2720 2386 Sparrows Point, Md. 2957 318& 2856 Birmingham, Ala. 16118 1711: 1308 Notes: Ton-mile equivalents for Birmingham are low because calculation of these equivalents involves the use of a factor, based on relative prices, to convert ocean ton-mile to railway ton-mile. The distance by rail from Birmingham to port is not sufficient to offset the considerably lower ocean mileage from port to port. n.a. indicates not applicable. Source: United Nations, Department of Economic Affairs, World Iron Ore Resources and their Utilization (New York, I955), T__—'—_—_"— P. O 156 Table 20.--Estimated Transportation.Requirements for the Production of One Ton of Pig Iron at Selected Locations in Latin.America and the United States (in railway tonrmile equivalents) Plant location Ore location Coal location Tonsmdles Volta Redonda Lafayette, Santa Catarina 1553 Brazil and West Virginia Huachipato El Tofo, Chile Concepcion and 1050 West Virginia Monclova Durango, Sabinas 963 Mexico Pittsburgh Mesabi Western 1116 Pennsylvania Sparrows Pt. La Serena, Pennsylvania 2209 Chile and West Virginia Note: These estimates allow for differences in the prOportions of ore and scrap uses in develOped and underdevelOped areas. They include loading and unloading costs and Panama Canal tolls. Allowance is also made for the difference in iron content of the ores. Transport by ocean is converted to railroad mdles at the rate of 1 sea mile equals 0.15 rail miles. Source: United Nations, Department of Economic Affairs, World Iron Ore Resources and their Utilization (New YOrk, I95U7, pp- - - 157 Table 21.--Estimated Cost of Finished Steel Products in Three Latin.American Steel Plants and Sparrows Point in 19&8 Dollars 22222212222292.2222" Volta Redonda 716 $85.&1 Huachipato 230 82.&& Monclova &30 83.10 Sparrows Point 1000 71.92 Source: United Nations, Department of Economic Affairs, A Stud of the Iron and Steel Indust in Latin America, I (E7CNTIE/ ev. , ovember, ew or , , 6. 158 hypothetical production costs. They do not include trans- portation costs for finished products either from the Latin American plants or Sparrows Point to the major market in each country. In order to get a comparison of domestic versus import costs, the cost of steel (per ton) bought in Pittsburgh and c.i.f. port of entry Talcahuano, Chile : Santos, Brazil; and Vera Cruz, Mexico, has been estimated. The estimate is based on a composite finished steel price f.o.b. mill, Pittsburgh, plus rail transport to Baltimore, Maryland, and ocean freight to port of entry. This has then been converted into domestic currencies to get an import price per ton and compared with the average price per ton of the domestic producer.1 The costs of transportation from the port or domestic plant to the major market has been omitted. This is probably not a serious distortion because Talcahuano and Concepcion are practically twin cities. Santos is the port of Sao Paulo, Brazil's major industrial center, and since it is only 200 miles by sea from Rio de Janeiro, the cost differential is small. In this case the omission favors the domestic price. In the case of Mexico, the port of Vera Cruz is approximately 200 miles from Mexico City while Monclova is over 500 miles away. However, Monclova is only about 100 miles from the industrial center at Monterrey. The exclusion biases the data in favor of the 1As previously mentioned, this average price is based on production rather than on sales tonnage. It is also inaccurate insofar as the domestic product mix deviates from the product mix upon which the composite price is based. 159 Table 22.--Comparison of Estimated Delivered Cost of Imported Steel (c.i.f.) with Estimated Domestic Prices, CSN, CAP, and AHMSA A. Companhia Siderurgica Nacional--— Domestic price as Im ort Domestic 2 a Cruzeiros/ton Per cent 19&7 1,800 2,683 1&9 19&8 2,182 3,765 132 19&9 2,258 &,066 1 0 1950 2,291 &,&08 192 1951 2,300 &,307 187 1952 2,&08 &,621 192 1953 2,719 %,950 182 195 2,761 ,3g3 231 195 6,689 8,2 8 125 1956 6,83& 10,788 158 1957 9,061 12,153 13& 1958 10.797 15,6 3 1&5 1959 18,&&0 23,8 0 129 B. Compania de Acero del Pacifico Domestic price as Import Domestic Year a percentage of price price Pesos/ton import price Per cent 1950 7,&22 8,775 118 1951 11,187 19,6&& 176 1952 15.252 26,767 175 1953 15.913 36,10& 227 195k 29.9h3 37,307 127 1955 6.692 63,736 137 1956 5,&96 169,&58 198 1957 12&,h76 228,103 183 1958 185.57? 321,111 171 (continued) ].v'fl. .lhnlnflflwlpl). ,I .. . . «L31. )l- 160 Table 22 (continued) C. Altos Hornos de Mexico, S.A. Domestic price as Import Domestic a percentage of Y9“? price price import price P°8°3/E6n Per cent 1905 3%3 378 110 19&6 3 9 561 1&& 1913 nu? 558 125 19& 731 831 11% 19&9 98& 1,033 10 1950 1,022 1,101 108 1951 1,037 1,155 111 1952 1,055 1,187 112 1953 1,211 1,202 100 195 1,102 1,%2& 129 195 1,269 1, 127 1956 1,891 1,91 101 195 2,156 2,083 97 195 2,2&3 2,272 101 1959 2,268 2,375 10& Sources: 1. Composite finished steel prices f.o.b. mill, Pittsburgh, l9&5-59--American.Meta1.Market, Metal Statistics (53rd ed.; New'YOrk, 1960), p. 215. Basic steel shipping rates from United States Atlantic ports to a. Talcahuano, Chile--Letter from.W. A. McGonagle, Atlantic and Gulf West Coast of South America Con- ference, May 1, 1962. b. Santos, Brazi1--Letter from.K. D. Thornton, Traffic Managzr, River Plate and Brazil Conferences, August ’ 1 2. c. Vera Cruz, Mexico-~Letter from C. V. Cutler, Smith and JOhnson (shipping), Inc., May 25, 1962. Railroad freight rates for steel, carloads from Pitts- burgh, Pennsylvania, to Baltimore, Maryland, 19&5-60-- Baltimore and Ohio Railroad Company, Tariff 0, "Local and Joint Export Freight Tariff Apply on ron and Steel Articles, Billets and Pig Iron and New Iron and 161 Table 22 (continued) Steel Rails and Iron and Steel Railroad Cross Ties"; Interstate Commerce Commission No. 23585. Effective September 18, 19&&, and successive issues thereof. Exchange rates--Internationa1 Monetary Fund, Inter- national Financial Statistics, various issues. a. Brazil, 19&7-52, rate applicable to essential imports 1953-5h, most favored import rate 1955-56, government import rate 1957-59, favored import rate b. Chile, 19&7-53, principal import rate 1953-55, Official rate 1956-58, free trade rate c. Mexico, 19&5-59, free market rate (selling). 162 domestic producer for delivery in Mexico City, biases the data in favor of imports for Monterrey. In any case, the choice of Pittsburgh as the origin of shipments favors the domestic producers. Birmingham steel would be even cheaper for Brazil and Mexico while imports from the West Coast of the United States would avoid Panama Canal tolls for Chilean purchases. At times steel could also be obtained in Mexico from the Sheffield Division of Armco in Houston, Texas, about &50 miles from Monterrey by rail. Even given the probable overestimation of the import price of steel, the import price has been almost without exception lower than the domestic price. The only exception to this has been the more recent experience of AHMSA. In 1951:, the devaluation of the Mexican peso sharply increased the import price in pesos. For CAP the domestic price of steel has ranged between 27 and 127 per cent above the import price. Moreover, there is no evidence Of a narrowing of the gap between the domestic and inport price. For CSN the domestic price has varied between 25 and 131 per cent above the import price. The difference between the domestic and import prices of steel indicates that the domestic steel industry is protected, and, therefore, subsidized in some manner. This protection and subsidization can take any of several forms (or it might involve several or even all Of them): prohibition or restriction of imports, import duties and taxes, unfavorable exchange rates for competing imports, 163 guaranteed returns on capital, exemption from taxes, etc. As will be seen, all Of these, and others, have been used in the protection of the steel plants in this study. This indicates rather clearly that profitability of the enter- prise cannot be used as a practical criterion for the evalu- ation of investments. This is so because, within very broad limits, any project can be profitable if the government wishes it to be so. Protection and Subsidization of the SteeI_Industry The link between the degree Of protection afforded the steel plant and the meaningfulness of its COR should be fairly clear. If, and to the extent that, the steel industry has been afforded a greater degree Of protection than other domestic industries, the difference between the COR of the steel plant and the ICOR for the rest Of the economy will be underestimated. Protection of an industry allows it to keep its domestic prices higher than they would be given compe- tition from.imports. And, assuming inelastic demand, this results in an exaggeration of "output" (actually sales) in money terms. If the only protective device being used was a tariff, the degree of distortion in CORS introduced by a relatively higher tariff on steel than on other products ‘would be indicated by the difference between the tariff on steel and the average level of tariffs on imported counter- parts of other domestic products. While this is a relatively 16& simple concept, it is extremely difficult to estimate. It involves determining: (l) the individual products produced by a domestic industry, (2) the tariff rates applying on the imported counterparts of these products, (3) the ad valorem burden of various duties and taxes, (&) the extent to which tariff rates may not represent actual protection, and (5) the degree of tariff protection that characterizes the industry as a whole.1 More important, of course, is the fact that pro- tective devices other than tariffs were commonly used-- licensing, exchange controls, advance deposits, etc. These would make comparisons of the type suggested above impos- sible. Besides this, these industries were not only pro- tected, but were also subsidized in many ways. The direction of the effect of subsidies on the COR is not clear. A subsidy may allow a firm to charge a lower price or it may support a higher price, depending on the exact nature of the subsidy and the elasticity of demand for the product. Empirically, subsidies are even more difficult to identify and evaluate than even tariffs. For these reasons no attempt has been made to evalu- ate the magnitude of the distortion introduced into the analysis of comparative CORs by the existence of tariffs and/Or subsidies. All we have attempted to do in this 1Beatrice Vaccara, Em lo ent and Out ut in Pro- tected Manufacturi Industries (Easfiifigton, E.C.: TEe IE tItfitgon I966) Brookifigs s , , p. 11. 165 section is indicate the magnitude of the differential in price between domestic steel and hypothetical duty-free imports. In the following section, the extent and tech, niques of protection and subsidization in each case will be discussed. The Companhia Siderurgica Nacional received sub- stantial assistance from the Brazilian government in its establishment. Aside from.assistance in financing, dis— cussed previously, which enabled the firm to Obtain capital at very favorable rates (this was true in all three cases), it provided exchange for foreign equipment and materials which were not available to other firms. Government depart- ments were ordered to purchase their steel from Volta Redonda even if they could‘buy at lower prices in foreign markets or in Brazil. The mill was given preferential tax treatment. Import restrictions through exchange controls cut off foreign supplies.1 Insofar as tariff protection of the new firm is concerned, Simon.Hanson claims that . . . while it was still in the process of construction government Officials discussed with other steel pro- ducers the desirability Of raising protective tariffs, which although they were adequate for existing facili- ties to be run profitably, might not be adequate for the new enterprise in which the government was interested.2 In more recent times, Brazilian protection of the steel industry has depended on a complicated exchange 1Simon G. Hanson, Economic Develo ent in Latin America (Washington, D.C.: Inter-Ifierican.ATTaIFs Press, I95I), p. 205. ZIbid. 166 auction system. This system.has involved the use of multi- ple selling rates, with steel products, despite their essentiality, not appearing among the favored categories Of imports.1 The extent of Chilean protection of the steel industry can be gauged from the fact that as late as 1959 CAP's prices to the domestic market were &1 per cent higher than those quoted by principal American manufacturers, but 2 This may not 27 per cent EEEEE the price including duties. completely measure the barriers to import of steel. Until 1956, importation required licenses which.were issued at the discretion of the government. Since 1956, importers have been required to make import guarantee deposits ranging between.5 and 1500 per cent of the c.i.f. value of imports.3 Of all Latin American countries, Mexico has probably had the longest history of protectionist.measures.u This protectionist hand was extended generously to Altos Hornos. This company and other steel producers received various tax exemptions and subsidies that other industries did not get lFor a discussion of the complications and dis- tortions introduced by the Brazilian multiple exchange rate system, see Alexandre Kafka, "The Brazilian.Exchange Auction System," The Review of Economics and Statistics, XXXVIII (August, 1956), 358-22. 2U.S., Department of Commerce, Investment in Chile, p. 163. 31bid. “Wythe, Industry in Latin America, p. 66. 167 l The industry and firm also benefited to the same extent. from Mexican tariff and import restriction. In December, 19&&, the tariff on iron and steel was increased by about 100 per cent of the former rate and ranged up to 900 per cent. The industry (and also that of Brazil) has been aided by a ban on the export of steel scrap.2 The Mexican tariff since 19&7 has consisted of a levy (per cent of weight plus a systom of ad valorem rates). Furthermore, the ad valorem tax is based on the invoice price or on an official value, depending upon.which is higher--the official price has been at times rather arbitrarily administered.3 Besides this, the Mexican government has at times refused to issue import permits for steel products. In 1955, the tariff on baser steel products averaged 35.2 per cent of value. This was after the tariff was raised sharply the year before. In June Of 195& the government began to restrict granting of import licenses for iron and steel products and refused to grant licenses for import of steel products manufactured in the country.h This was coupled with a devaluation of the 1S. G. Ross and JOhn B. Christensen, Tax Incentives for Industr in Mexico (Cambridge: Law SchooI of Harvard UnIversIty, I959), pp. 38, 75 et passim. 2U.S., Tariff Commission, Economic Controls and Commercial Polic in.Mexico (Washington, D.C.: U.S. Govern- ment FFIEtIng UTTIce, I9E6), p. 27. 3U.S., Department of Commerce, Investment in Mexico, p. 106. “U.S., Department of Interior, Mineral Trade Notes, XL, No. 5 (May, 1956), p. 19. 168 Table 23.--Comparison of Prices of Imported and Domestic Steel at Monterrey, Mexico, with and without Tariff and Tariff Per Ton, 1955 Price of PIPICB or U . S . hports Mexican M61168]: Item per ton Steel tariff Without With """7P Tariff; tariff tariff Pi6gé ton Blooms and slabs 106.29 128.5 n.a. 22.25 Structural shapes 135.00 2&&.1 151.55 89.18 Merchant bars 136.10 193.37 1&7.l7 57.27 Rails 127.08 1&9.37 181.97 22.29 Plate 132.80 190.05 115.&0 57.25 Hot rolled Strip 128.9& 186.21 115.&0 57.27 Cold rolled strip 171.37 228.6 1&3.22 57.27 Cold rolled Sheet 150.98 208.2 1&3.22 57.27 Tinplate, "coke" 2&0.&7 293.5& 213.56 53.07 Tinplate, electro- lytic 211.82 26&.89 n.a. 53.07 Wire rod 1&&.37 201.6& 15&.O8 57.23 Wire 171.37 222.17 163.28 50.80 Note: n.a. indicates not sold. Source: Altos Hornos de Mexico, S.A., Rolled Products in Mexico, unpublis e an o as g on, D.C., 1955, Tables 1& and 15. mpor repor Marketing of Steel Flat 0 e port- II III 169 peso. Because of the multiplicity of protective devices and subsidies, it is not possible to measure the real cost of domestically produced, as compared with imported steel.2 It is clear that the cost is greater than the difference between the domestic price and the c.i.f. price of imports-- if this is so, the cost is substantial. The attempts to justify the incurring Of these costs to the economy have been discussed in a previous chapter. They involve all the arguments and motives for establishing a domestic steel industry in the first place. Of course, if the steel industry were justifiable on the basis of compara- tive advantage, these arguments would be superfluous. This possibility is not ignored here. In fact, it is implicit in the infant-industry-type argument: A number of factors work against the initial efficiency of iron and steel industries in underdevelOped countries and prevent the economies which are found in.most devel- Oped countries. . . . Time and experience are required to perfect the necessary engineering, managerial and labor skills and to attain an efficient work force.3 1The effect of this move on steel consumption was immediate. Mexican output in 195& of rolled products rose by 72,000 tons or 16.6 per cent over the 1953 level; how- ever, steel consumption was lower than in 1953 because of a drastic decline in imports. United Nations, Economic Survey of Latin America, 19511. p. 180. '—"'"'""" 2Besides direct aid to the steel industry, these firms have benefited from government aid in the promotion of the domestic coal industry and the reduction or abandon- ment of duties on coal. 3United Nations, World Iron Ore Resources and their Utilization, p. 13. 170 There is, however, little evidence that the steel plants are now, or are becoming, competitive in an inter- national sense. The final rationalization seoms to be the foreign-exchange saving argument. The ECLA has estimated that the hypothetical plants would require 67 per cent less foreign exchange per ton as compared with the delivered price of imported steel.1 This argument represents the ultimate rationalization of industrial autarky. Acceptance of foreign exchange savings as the primary investment cri- terion makes possible the justification of almost any project. As long as it is physically possible to produce the product involved, using imported raw materials and intermediate goods with a total cost per unit of output below the existing import price of the product, some foreign exchange savings can be realized from domestic production. 1A Study of the Iron and Steel Industry, I, 16. CHAPTER VIII CONCLUSIONS Evaluation Using CORs and ICORs The evaluation of the steel plants involves compar- ing the COR of the firm with the ICORs for the economy and manufacturing sector over some apprOpriate time period. The evaluation is made assuming that the ICORs indicate the average increases in output associated with capital formation during the time the steel plant was being planned and built. This gives a rough idea of the alternatives foregone in the investment of capital in the steel industry. Comparison of the gross CORs of the steel plants and the ICORs for their respective economies indicates that the COR for the steel plant was substantially above the ICOR of the economy in the case of Brazil and Chile, but less so in the case Of Mexico. For Brazil and Chile the gross COR was approximately twice the ICOR for the economy. The differ- ence between the company COR and the ICOR for the economy was substantially smaller in.Mexico. The CORs and ICORs were also compared on a "net" basis--subtracting depreciation from.value added for the plants and using net national product as an output measure. The steel plants showed up even less favorably in this case. 171 172 Altos Hornos, however, still appeared more favorably than the other two plants. Finally, the CORs of the steel plants were compared with estimated ICORs for the manufacturing sectors, using data of lower quality. In the case of Chile, the COR of CAP was 126 per cent above the ICOR for the manufacturing sector. The net COR of CSN was only 37 per cent above the net manufacturing ICOR in Brazil. The ICOR for the Mexican manufacturing sector was estimated using two different techniques. The COR found for AHMSA was 35 per cent above one formulation of the sectoral ICOR (gross) and only 13 per cent above the other (net). In any case, the investment in a steel plant appears more favorable, using the ICOR criterion, when compared with investments in the manufactur- ing sector. Even in the case where the steel investment appeared most favorable, the company COR was definitely above the average ICOR for the sector. This indicates that, more often than not, there were investments being made in manufacturing which had greater output increments associated with them.than were found in the case of the investment in steel. Costs of Domestic and Imported Steel A comparison was made of the costs of domestic and imported steel. In all cases it was found that the cost of domestic steel was substantially higher than an estimated hypothetical average cost Of imported steel, exclusive of 173 tariffs. This indicated that the steel industry was the beneficiary of protective devices. These, and the various subsidies received by the steel plants, were investigated and were found to be numerous and substantial. It was also indicated that protective tariffs and subsidies, to the extent that they are greater for the steel plant than for other industries, distort the ICOR comparisons. The differ- ence between the domestic and import price was found to be the smallest in the Mexican case (although this may be a statistical, rather than a real, phenomenon). Implications It may not be possible to say, post hoc, that investment in a steel plant was, or was not, justified. However, this study does raise some important questions in cases which are presumably "successes." The evidence does indicate that investment in a steel plant was substantially more "capital absorbing" per unit of output than was the average investment being made in the economy at the same time. It also indicates that the cost of steel made domestically is substantially above the import price. Furthermore, each of these industries was heavily subsi- dized and protected. Comparison of the steel COR with the ICOR of the economy and the manufacturing sector does not in any sense imply that the old or pro-industrial industry structure is to be preserved if the steel plant is rejected because its 171+ COR is substantially above the ICOR. 0n the contrary, the ICOR, being incremental, shows the increases in output associated with new investment, it does not represent the average productivity of the capital stock now in existence. It must be remembered that the three countries involved in this study are among the more developed of the group generally referred to as "underdevelOped." Also, they are reasonably well endowed with the resources necessary for production of steel. Furthermore, they have all had previ- ous experience with the domestic production of steel. If the investment in steel plants in these countries were to be classified as "marginal," or if serious questions were raised as to the wisdom of such investments, how would investment in steel plants be classified for less develOped, less well- endowed countries? The Economic Commission for Latin America, which has long been in the forefront of those promoting domestic steel industries, has apparently had second thoughts about the develOpment of autarkic capital goods, and presumably this would include steel industries.1 These thoughts have seemed to develOp from a realization that the protectionistic measures necessary to develOp these industries represent barriers to economic integration. One of the first arguments made in this paper was 1A1bert 0. Hirschman, "Ideologies of Economic Development in Latin.America," Latin American Issues, ed. Aézert O. Hirschman (New YOrk: TEe Twentieth Century Fund, 1 1 , p. 20. 175 that the steel investment should be compared with available alternatives. This implies that capital would be available for these alternatives as well as for a steel plant. If it is not, then any arguments concerning alternatives are entirely fruitless--there are no alternative costs. The well-known predisposition of American lending agencies to favor the sO-called "big concrete and steel" projects may lead underdevelOped countries to direct their planning toward such projects rather than to consider the alter- natives. This is a particularly dangerous situation, given the already strong symbolic attraction of steel and the biases, ideological and pseudo-economic, which exist in.many underdevelOped areas. The implication here is that the lending policies of international develOpment agencies, particularly the World Bank and the Export-Import Bank, should be re-evaluated. BIBLIOGRAPHY BOOKS Adams, Richard N., et a1. Social Cha e in Latin America Today. New York: Vintage Books, 1961. Agarwala, A. N., and Singh, S. P. (ed.). The Economics of UnderdevelOpment. London: Oxford University Press, American.Metal Market. Metal Statistics. 53rd edition. New YOrk, 1960. Center for International Studies, Massachusetts Institute of Technology. Investment Criteria and Economic Growth. Bombay: As ism . Chudson,‘Wa1ter A. The Pattern of Cor orate Financial Structure. New YbrE: NationaI Bfigeau of Economic ResearcH, 19&5. Cline, Howard F. The United States and Mexico. Cambridge: Harvard University PFess, I953. Combined Mexican Working Group. The Economic DevelOpment Of Mexico. Baltimore: JOhns HOp n vers y ess, 1953. Creamer, Daniel, Dobrovolsky, Sergei, and Borenstein, Israel. Ca ital in Manufacturigg and Min . Princeton: a one Bureau 0 conomic Researc , Princeton Univer- sity Press, 1961. Delwart, Lewis 0. The Future Of Latin.American.Ex the United States: NationaI PIanning Assoc a on, orts to Ellis, Howard S., and Wallich, Henry 0. (ed.). Economic Develo%nt for Latin America. New York: am .a” c.’ . Fondo de Cultura Economics. Mexico: Cincuenta.Anos de Revolucion. Mexico, D.F., I93I. 176 177 Ginsburg, Norton. Atlas of Economic Development. Chicago: University of Chicago Press, 1961. Godfrey, Walter. Economic and Commercial Conditions Brazil. London: ‘Her MajestyTs Stationery Office, 1955 Hanson, Simon G. Economic DevelOpment in Latin.America. Washington, D. C.: IntereAmericanUAfTEirs Press, 1951. Harris, Seymour E. (ed.). Economic Problems of Latin America. New YOrk: McGraw-Hill Book Company,IHc., Higgins, Benjamin. Economic Development. New York: W. W. Norton and Company,‘I959} “"1 Hirschman,.A1bert 0. (ed.). Latin American Issues. New YOrk: The Twentieth Century Fund, 1961. . The Strategy of Economic Development. New Haven: YaIe University Press, 1958} M, Hurtado, Carlos Sanchez. La Industria Siderur ica Chilena y Ibero Americano. Santiago, CEiIe: EditoriaI Nasci- mento , 1952 . Instituto de Economics de la Universidad de Chile. Des- arrollo Economico de Chile 19 0-19 6. Santiago, Chile: or a n vers er a, . ., . JObhm, JOse. Brazil in the Makin . New YOrk: Macmillan and Company, I9E3. Johnson, John J; Political Chan e in Latin America. Stanford: Stafiiord University PFess, I958. Kuznets, Simon. Ca ital in the American Econom . Princeton: National Bureau 0? Econdfiic ResearcH, Prificeton Univer- sity Press, 1961. Leibenstein, Harvey. Economic Backwardness and Economic Growth. New YOrk: 3053 WiIey anH Sons, IEc., I957. Lutz, Friedrich A., and Hague, Douglas C. (ed.). The Theor Of Capital. London: St. Martin's Press, Inc., I96I. Mosk, Sanford. Industrial Revolution in.Mexico. Berkeley: University 0 a orn ress, . National Bureau of Economic Research. Business Concen- tration and Price Polic . Princeton: Ptinceton Uni- versity Press, I955. 178 National Bureau of Economic Research. Ca ital Formation and Economic Growth. Princeton: Princeton University ress, . Nurkse, Ragnar. Problems of Ca ital Formation in Under- develOped Countries. Lonaon: Uiiora University PFess, Philippines National Economic Council. The Five-Year Economic and Social Develo ent Programs for Ybars 1951- I96I. HaniIa: TEe CounciI, I957. Pounds, Norman J. G. The Geo a h of Iron and Steel. London: Hutchinson University Litrary, I959. Price waterhouse and Peet. Com ania Siderur ica Nacional, Financial Statement, 195::I959. Rio de ghneito, I95I-60. RObie, Edward H. (ed.). Economics of Mineral Industries. New York: American Instifite OT HinirT'g, HetaIIurgical and Petroleum Engineers, Inc., 1959. Rosen, George. Industrial Cha e in India. Glencoe, Illi- nois: The Free Ptess, I958. Ross, S. G., and Christensen, thn B. Tax Incentives for Industr in Mexico. Cambridge: Law SchoI OT HErvard University, I959. Spiegel, Henry W. The Brazilian.Econom . Philadelphia: The Blakiston Company, I9E5. University of Texas. Institute of Latin American Study. Basic Industries in Texas and Northern Mexico. Austin: University of Texas PFess, I950. Vaccara, Beatrice. Em lo ent and Out ut in Protected Manufactur§gg Industries. Wasfiifigton, 5.0.: TEe Brookings me u on, 960. Voskuil,'Wa1ter H. Minerals ianorld Industr . New YOrk: McGrawaHill Book Company, Inc., I955. Wythe, George. Industr in Latin America. New YOrk: Columbia University PFess, I9E5. Zimmermann, Erich W. 'World Resources and Industries. Revised edition. New Tort: Harper ana EEOtEers, 1951. 179 ARTICLES "Altos Hornos de Mexico, S.A.," Revista de Economia, V, NO. 8 (August, 19&2), 1-8. .Anderson, George. "The Apparent Decline of the Capital Output Ratio," Quarterly Journal of Economics, LXXV, NO. Bator, Francis. "On Capital Productivity, Input Allocations, and Growth," Quarterl Journal of Economics, LXXI, NO. 1 (February, 1957), 86-I06. Belshaw, Michael. "Operational Capital Allocation Criteria for DevelOpment Planning," Economic DevelOpment and Cultural Change, VI (April, I958), 1&6-63. Bernstein, Marvin D. "Foreign Investments and Mineral Resources: A Few Observations," Inter-American Economic Affairs, XIII, No. 3 (September, , - . Bhatt, V. V. "Capital-Output Ratios of Certain Industries: A Comparative Study of Certain Countries," Review of Economics and Statistics, XXVI (August, l95E), 309-20. Chenery, Hollis B. "The Application of Investment Cri- teria," Quarterly JOurnal of Economics, LXVII, No. l (Febmary, , "' e . "Comparative Advantage and DevelOpment Policy," Afierican Economic Review, LI, No. 1 (March, 1961), 11- H9. . "The Role of Industrialization in DevelOpment Ptograms," American Economic Review Papers and Proceed- M, HIV, no. 3 (may, , - 0 Davis, Hiram. "Relationship of the Capital-Output Ratio to Firm.Size in.American.Manufacturing," Review of Economics and Statistics, XXVIII (August, I956), 288- 93. Eckstein, Otto. "Investment Criteria for Economic DevelOp- ment and the Theory of Intertemporal Welfare Economics," Quarterl Journal of Economics, LXXI, NO. 1 (February, 1957): 58-85. Galenson, Walter, and Leibenstein, Harvey. "Investment Criteria, Productivity, and Economic DevelOpment," Quarterly JOurnal of Economics, LXIX, No. 3 (August, , "Oe 180 Hirschman, Albert 0. "Investment Criteria and Capital Intensity Once Again," Quarterl JOurnal of Economics, LXXVII, No. 3 (August, I958), E89-7I . "Investment Policies and Dualism.in'Under- Eavelo ed Countries," American.Economic Review, XLVII, Kafka, Alexandre. “The Brazilian Exchange Auction System," Review Of Economics and Statistics, XXXVIII (August, , 8-22. Hahn, Alfred E. "Investment Criteria in DevelOpment Pro- ams, Quarterly JOurnal Of Economics, LXV, NO. 1 February, , "' e Neisser, Hans. "Investment Criteria, Productivity, and Economic DevelOpment," Quarterl JOurnal of Economics, LXX, NO. & (November, 1956), SEE-E7 Prebisch, Raul. "Commercial Policy in the UnderdevelOped Countries," American Economic Review Pa ers and Pro- ceedings, XL , o. ay, , - 3. "Rolled Iron and Steel Products in Latin America," Economic gulletin for Latin.America, IV, No. 2 (October, I959), :32e Rosenstein-Rodan, Paul N. "Problems of Industrialization in Eastern and Southeastern EurOpe," Economic Journal, LIII, NO. 3 (June-September, 19&3), 202211? Scitovsky, Tibor. “Two Concepts of External Economies," JOurnal of Political Economy, LXII, No. 2 (April, 195&), Singer, Hans. "The Distribution of Gains Between Investing and Borrowing Countries," American.Economic Review Papers and Proceedings, XL, No. 2 (Kay, I950), E73-85. White, Langdon C., and Anderson, Donald J. "Steel and Symbolism at Paz del Rio," Inter-American Economic Affairs, V, No. 9 (Spring, I956), 82-93. ‘White, Langdon C., and Chilcote, R. H. "Chile's New Iron and Steel Industry," Economic Geography, XXXVII, NO. 3 (July, 1961), 258-66. 181 GOVERNMENT PUBLICATIONS Brazil Conselho Nacional de Estatistica. Boletin Estatistico, No. 61, Januarbearch, 1958. Conselho Nacional de Estatistica. Instituto Brasileiro Geografia Economia. .Anuario Estatistica. Various issues. Conselho Nacional de Estatistica. Recenseamento Geral do Brasil Volume IV Censos IndustriaI. Rio de JaneiFO. Conselho Nacional de Estatistica. Recenseamento Geral do Brasil, 19&0. Rio de Janeiro, I950. Instituto Brasileiro de Geografia Economia. 0 Brasil Em Numeros. Rio de Janeiro, 1960. Chile Servicio Nacional de Estadistica y Censos. Estadistica Chilena. Various issues. Mexico Direccion General de Estadistica. Censo Industrial de los Estados Unidos Mexicanos, 19&0. Mexico, 0.F., I952. Estados Unidos Mexicanos. Quinto Censo Industrial Tercer Censo de TransportesI 1950. Mexico, 5.F., I957. United States U.S. Department of Commerce. Investment in Chile. wash- ington, D.C.: U.S. Government PFintiEE 0TTice, 1960. U.S. Department of Commerce. Investment in Mexico. Wash- ington, D.C.: U.S. Government Prifitifig 0Tfice, 1958. 'U.S. Department of Interior, Bureau of Mines and Geological Survey. Materials Surve , "Iron Ore." Washington, D.C.: U.S. Government rinting Office, 1956. ‘U.S. Department of Interior. Mineral Trade Notes. Various issues. 182 U.S. Tariff Commission. Iron Ore: Re-ort on Investi-ation U.S. Tariff Commission. Economic Controls and Commercial Polic in Mexico. Washington, D.C.: 0.8. Government 5132137522163: 19&6. UNITED NATIONS U.N. Department of Economic Affairs, Economic Commission for EurOpe. LO Term Trends and Problems of the Euro ean Steel Industg. Geneva, I959. U.N. Department of Economic Affairs, Economic Commission for Latin America. Analysis and Projections of Economic DevelOpment. VO . . ew or , . U.N. Department of Economic Affairs, Economic Commission for Latin America. Economic Survey Of Latin America, 19&8. New York, 19149. U.N. Department of Economic Affairs, Economic Commission for Latin America. Economic Survey of Latin America, 1999. New York, 1951 . U.N. Department of Economic Affairs, Economic Commission for Latin America. Economic Survey of Latin America, 1951- 19 2. New York, I953. U.N. Department of Economic Affairs, Economic Commission for Latin America. Economic Survey of Latin America, 1953. New York, 195& . U.N. Department of Economic Affairs, Economic Commission for Latin America. Economic Survey of Latin America, 195&. New York, 1955 . U.N. Department of Economic Affairs, Economic Commission for Latin America. Economic Survey of Latin America, 1955. New York, 1956. U.N. Department of Economic Affairs, Economic Commission for Latin America. Economic Survey of Latin America, 1256. New York, 1957. U.N. Department of Economic Affairs, Economic Commission for Latin America. Iron and Steel Transformgg Industries in Selected Latin American oun r es. ogo a, o om a, 1955 . 183 U.N. Department of Economic Affairs. A Stud of the Iron and Steel Indust in Latin America. VoIs. I afia II. New YbrE, I95E. U.N. Department of Economic Affairs. Economic DevelOpment in Selected Countries. New YOrk, I987: U.N. Department of Economic Affairs. World Iron Ore Resources and their Utilization. New YorE, I950. U.N. Economic and Social Council. Nonferrous Metals in UnderdevelOped Countries. New YorE, 1958. Statistical Office of the United Nations. Per Ca its National Product of Fifty-five Countries, I952-I95g. New YorE , I957 . Statistical Office of the Unitethations. Statistical Year- book. New Ybrk. Various issues. Statistical Office of the United.Nations. Yearbook of National Income Accounts, 1252. New Yer , . MONOGRAPHS Cohen, Alvin. "Economic Change in Chile," Latin.American Monogra¥%§. Vol. II. School of Interwiifirican Studies, n vers y of Florida. Gainesville: ‘University of Florida Press, 1960. Gil, Federico G. ”Genesis and Modernization of Political Parties in Chile," Latin.American.Mono a he. Vol. XVIII. School of Inter-25erican Sttaies, University of Fézrida. Gainesville: University of Florida Press, 1 2. DISSERTATIONS Marcus, Matityahu. "Capital-Output Ratios and their Vari- ation." Unpublished Ph.D. dissertation, Brown‘Uni- versity, 1963. Schor, Stanley 8. "The Capital-Product Ratio and Size of Establishment for Manufacturing Industries." Unpub- léghed Ph.D. dissertation, University of Pennsylvania, 1 2. Vaughn, Ernest H. "A Study of the DevelOpment of National Steel Plants in Selected Latin American Countries." Ugggblished Ph.D. dissertation, University of Texas, 1 . 18& UNPUBLISHED MATERIALS AND REPORTS Altos Hornos de Mexico, S.A. Informs Anual. Monclova, Mexico. 19&5-60. Altos Hornos de Mexico, S.A. Marketi of Steel Flat Rolled Products in Mexico. Report to tEe Eiport-Ifiport BEEF ET‘Washington,7070., 1955. Cass, A. C. O eratin Re orts. Altos Hornos de Mexico, S.A., 1955. Companhia Siderurgica Nacional. Relatorio do Directoria. Volta Redonda, Brazil, 19&7-60. Compania de Acero del Pacifico. Memoria Anual. Santiago, Compania de Acero del Pacifico. Proggoma, 1958, Enero a Junie. Santiago, 1958. Export-Import Bank. .Altos Hornos de Mexico S.A. Memoran- dum. Januar' '1_. 'e: re. s ‘O. a-- 3 -3, ‘ac ona I o:nc era and AHMSA, request for revision of maturities. Export-Import Bank. Case History: Volta Redonda. (Mimeo- graphed.) International Monetary Fund. International Financial Sta- tistics. Various issues. KOppers Company, Inc. and Compania de Acero del Pacifico. Program, 1959-1961;. Santiago, 1958. Keppers Company, Inc. Re ort on the O erations Of Com.ania de Acero del Pacifico. Various issues. Lerner, Abba P. Ca ital-Out ut Ratios Production Functions ana tfie TEeo OT CapitaI Ifivestfient. (Mifieograpfied.) Letter from W. A. McGonagle, Atlantic and Gulf West Coast Of South America Conference, May 1, 1962. Letter from.Luis Barranco, Comptroller, Altos Hornos de Mexico to A. Carl Cass, Chief, Engineering Department, Export-Import Bank of Washington, D.C., August 18, 1961. Letter from Alvaro Flores, Altos Hornos de Mexico, S.A., April 25, 1962. 185 Letter from.K. D. Thornton, Traffic Manager, River Plate and Brazil Conferences, August 9, 1962. Memorandum to the Board of Directors of the Ex ort- Im-ort Memorandum to the Board of Directors of the Ex ort-Im ort Rant, December I9, I950. Memorandum.of July 27, 1959, Export-Import Bank. Price Waterhouse and Pest. Com anhia Siderur ica Nacional, Financial Statement. Report to tHe Export-Lfiport Bait. Various issues. Railroad Freight Rates for Steel, Carloads from.Pittsburgh, Pennsylvania, to Baltimore, Maryland, Baltimore and Ohio Railroad Company, Tariff g5- , Local and Joint Export Freight Tariff App y on ron and Steel Articles, Billets and Pig Iron and New Iron and Steel Rails and Iron and Steel Railroad Cross Ties." Interstate Com- merce Commission No. 23585. Effective September 18, 19&&, and successive issues thereof. Report of the Joint Brazilian4United States Technical Com- mission. Rio 3e Janeito, I959. 186 Appendix Table 1.--Wholesa1e Price Indexes: Brazil, Chile, and Mexico, 1939-59 Year Brazil Chile Mexico 1939 28.7 20.6 31.1 19h0 30.7 22.7 32.9 19&1 32.3 26.5 35.3 19&2 37.9 36.1 38.8 19u3 h7.h 39.8 &6.8 1918 50.9 u1.1 57.3 19&5 61.6 g3.7 63.9 19&6 71.1 0.5 73.5 198% 80.2 65.1 77.9 19& 88.2 7u.7 83.5 1950 100.0 100.0 100.0 1951 119. 130.8 12&.0 1952 131. 162.2 129.1 1953 151.5 199.h 126.5 195& 197.0 313.0 138.0 1955 222.7 552.0 157.0 1956 266.7 90 .8 16&.6 195 298.5 128 .8 172.1 195 3hh.8 1616.9 179.7 1959 &60.0 2100.2 181.9 Sources: Brazil, l939-&9--United Nations, Department of Economic Affairs, Economic Commission for Latin.America, Anal sis and Economic Pro actions of Economic Develo ent, Vol. II, eve Opment of BraziI' (E7CN.I55208/Rev. 1) (New YOrk, 1956)) p0 195. Brazil, l950-59--Conselho Nacional de Estatistica, Instituto Brasileira Geografia Economia, Anuario Estatis- tico, 1959 (Rio de Janeiro, 1960), p. &3&. Chile, l939-59--Servicio Nacional de Estadistica y Censos, Estadistica Chilena (various issues). Mexico, l939-50--Combined Mexican Working Group, The Economic Develo ent of Mexico (Baltimore: Johns Hopfiifis Uniiersity ess, , p. 0. Mexico, l951-59--Statistica1 Office of the United Nations, Statistical Yearbook (New York, various issues). 187 Appendix Table 2.--Companhia Siderurgica Nacional: Fixed Asset Stock: Gross and Net: Book Value and 1950 Prices, 19&2-59 (million cruzeiros) Fixed assets Net fixed assets Year Book 1950 Book 1950 value prices value prices Adjusted 19&2 u93.0 1,300.8 &93.0 1,300.8 1913 983.5 2,251.2 9&3.5 2,251.2 1926 1,393.9 3,136.1 1,393.9 3,136.1 19&5 1,8%&.§ 3,867.& 1,8&&.& 3,867.& 19&6 2.h 1. n.763.5 2.h6h.9 h.7 0.0 8.673.0 19&7 2,621.5 h.398.2 2,586.8 h,8 9.5 &,550.2 1988 2,703.1 5,103.6 2,633.8 &,938.1 u,&1%.u 19 9 2,810.6 5,1&1.7 2,639.2 h.935.7 &,3u .9 19 0 2,897.2 .228.3 2.739. 5.0 -h u.zzu.u 1951 2,972.7 5,291.& 2,551.9 &,85 .2 1,075.9 1952 3,232.6 5.h88.6 2,639.& &,887.8 &,053.6 1953 3.859.8 5.902.6 3,068.2 5,129.7 h.231.5 195k 5,586.1 6.7 8.9 h.315.3 5,686.5 h.8 6-h 1955 6.70h.9 7.2 1.3 h.765.7 5.78h.h 5.0 7.7 1956 7.373.h 7.532.h h.h57.1 5.51h.9 5.027.1 1957 8,391.3 3,873.0 &,&90.0 5,333.1 5,063.2 1958 10,058.0 .356.& &,92&.7 5,222.0 5,222.1 1959 13.390.9 9,080.9 6,639.9 5,268.2 5,597.3 Notes: Net fixed assets adjusted is the value of the asset stock redepreciated at & per cent per year. Depreciation is the difference between fixed assets and not fixed assets. Deflators for these calculations are from Appendix Table 1. Source: Companhia Siderurgica Nacional, Relatorio do Directoria (Volta Redonda, various issues). 188 Appendix Table 3.--Compania de Acero del Pacifico: Fixed Asset Stock: Gross and Net: Book Value and 1950 Prices, 1987-59 (millions of pesos) Fixed assets Net fixed assets Year Book 1950 Book 1950 value prices value prices Adjusted 19&3 &96 762 19& 972 1.399 19u9 2,12& 2.751 1950 2.3h3 2.970 1951 2.705 3.2h7 2.h29 .036 3.117 1952 3.560 3.77h 2.967 .327 3.893 1953 h.hh0 h.125 3.515 .hBH 3.679 195& 5.20h .859 3.866 3.501 3.835 1955 6.17 §.635 8.159 .359 3.826 1956 12.12 .293 9.082 .657 8.273 1957 13.368 5.777 13.668 .739 .527 1958 37.050 6.932 30.079 .851 5.h06 1959 60.2h7 8,037 .219 ,172 6,191 Note: See notes to Appendix Table 2. Source: Compania de Acero del Pacifico, Memoria Anus; (Santiago, Chile, various issues). Asset Stock: 189 Appendix Table &.--A1tos Hornos de Mexico: Gross and Net: and 1950 Prices, 19&2-59 (thousands of pesos) Fixed Book Value Fixed assets Net fixed assets Year Book 1950- Book 1950 value prices value prices Adjusted 19&2 11,883 30,626 19h3 32.708 75.115 19th 57.056 117.61h 56.h7l 116.593 112.909 19&5 6&,885 129,866 63,&02 127.33h 119.986 19&6 70,057 136,903 67,122 132,115 121,536 19kg 7&.98u 1A3,228 69.325 138.611 122,123 198 82.327 152.022 3.737 139.298 12h.836 1989 100.018 171.335 8.03 153.999 137.295 1950 105,606 176,923 87.3h 151,993 135.805 1951 181.755 206.075 116.339 173.003 156.713 1952 203,008 253.521 169.336 211, 22 19&,017 1953 2 ,101 270,195 180,659 217, 27 199,882 195& 26 .590 302,&33 2 ,268 238,992 220,021 1955 808. 8 391.515 31 .015 300.930 293.hh1 1956 h6h.2 8 825.880 3 .h67 299.6u5 310.275 1957 555.118 878.217 38 .975 307.372 3h3.922 1958 30.355 575.733 1:89.623 355.238 1118.807 1959 17,152 623.&50 &&6,196 360,207 &&1,18& Note: See notes to Appendix Table 2. Source: Altos Hornos de Mexico, S.A., Informs Anual (Monclova, 1959): PP- 22'23- Appendix Table 5.--Brazil: Formation: Market prices 190 National Product and Capital Market and 1950 Prices, 1939-58 (thousands of millions of cruzeiros) prisce 'Ybar GDCF NDCF GDP NDP GDCF NDCF GDP NDP 1939 6.2 2.9 39.3 36.0 21.6 10.1 136.9 125.& 19&0 6.3 2.9 82.3 38.9 20.5 9.8 137.8 126.7 19&1 7.1 3.& &8.3 &&.6 22.0 10.5 1h9.5 131.8 19&2 7.9 2.7 55.0 g9.8 20.8 g.1 1&5.1 131.8 19&3 11.2 &.0 69.5 2.3 23.6 .% 1&6.6 131.& 19 15.2 6.9 89.5 81.2 29.9 13. 175.8 159.5 19 13.9 &.6 10&.u 95.1 22.6 7.5 169.5 158.u 19&6 20.5 10.9 133.5 123.9 28.8 15.3 187.7 17&.3 198; 26.5 17.0 169.5 160.0 33.0 21.2 211.3 199.5 198 25.9 15.5 190.7 180.3 29.& 17.6 216.2 20&. 19&9 33.6 22.5 217.7 206.6 3&.7 23.3 225.1 213 1950 33.2 20.5 255.8 2&3.1 33.2 20.5 255.8 2&3.1 1951 59.0 &3.6 309.& 290. h9.3 36.& 258.5 2&5. 1952 65.1 17.8 358.3 336. u9.& 36.7 268.8 255.& 1953 50.1 28.6 u33.5 112.0 33.1 18.9 286.1 271.9 1958 82.& 5&.6 580.8 553.0 &1.8 27.7 28&.7 270.6 1955 89.9 55.3 698.3 663.7 &0.& 2&.8 313.6 289.0 1956 106.5 62.& 889.5 8h5.h 39.9 23.& 335.5 338.h 1957 12&.5 71.7 1131.1 1010.3 &1.7 21.0 356.1 338.3 1958 165.6 101.1 1299.3 123&.8 &9.5 30.2 388.1 368. Note: GDCF a gross domestic capital formation NDCF = net domestic capital formation GDP = gross domestic product NDP = net domestic product Sources: Economic Commission for Latin America, ent, Vol. ections of Economic Develo§m eve Opment o Braz l939-&9--United Nations, Department of Economic Affairs, 1956), p. 193. l950-59--Statistica1 Office of the United Nations, Year- book of National Income Accounts, 1952 (New York, 1960), p. . 26& Anal sis and Pro- m c onom c /Rev. 1) (New YOrk, 191 .Aoooa .saom sozv ommawooBSooee csooaH Heooaocz mo soontao»..anoacaz screen on» to coaoeo sweeteners -- .Hne ooaamp .xomoa ..<.m .eaaecaaeocaem Hoaaooeom nomaaosami m -o om edwno oo ooaeomoom OHHonaomon .OHaSO co redeemabwnb ma op moaaomoom op opspfipmnHuuomuomoH "meoasom .m canoe Kaomonom on snow oom . ”epoz m:.o~a on.m:a es.m se.ma m».mm~m mm.aumm :m.mo :o.oom mom ma.moa o~.:aa mm.o oa.o~ -.omom mo.mm~m no.4m no.6:m mos om.ooa em.mca m:.H pm.ma mm.mmma mo.omeH sm.ma sm.oma omma oe.maa mo.mma ma.o m~.mfl mo.mmo m:.e:oa Ho.H Hm.om mmoe mm.moH 5:.mma mo.e SH.mH :o.omm :H.H~m mm.m m .a smoa mo.ooa oe.osa mo.o ma.oe or.omm «n.amm om.H om.mm mmoa om.osa mo.oma os.a oo.ma om.cmm oo.mmm ao.m ~:.:m «mos oa.sme om.mma mo.m HS.:H oe.mme os.ooa :o.m :m.oa Amos ma.e:a mo.mma om.o mm.sa ma.esa mo.mmH m:.o mm.ea omoa ma.oma Hm.mma mfl.m mo.ma mm.oaa -.oma om.a mw.ma mama mm.mma sm.mda mo.H o~.ma om.oo om.moH om.o o:.o were mm.HHH mo.HmH ms.m ::.mH Hm.ms mm.ma :e.m ma.m some oa.mme o:.emfl oH.e mo.:e a:.mo Ho.so oa.m em.c mead m:.:ea mm.:~e mo.: om.:a oo.om mm.:m ma.m om. meme mm.moa om.:HH :m.m H:.HH om.me no.5: mo.o on. some :H.mo ma.ooa am.: om.mH sm.om mm.om os.a om.: more m~.om ou.~m mm.a mm.m sa.om mm.am ::.o mo.m were :a.mo om.aoa oo.e mo.me :m.:~ oo.~m om.e 6:.m Haas mm.mm Ho.mo oo.m mm.0H sa.om mo.Hm om.o mm.m ones .32 one .82 .88 .52 one .82 some are» meoaamomofi seamen nomad: mm-o:oa .ocoaam omoa one scream Hensoao one anemone Hoaoaaez “nosed moaaada oaoaoomuv «soapmanom ”cameo--.o canes smoother ..)l'lll.lll(l I‘ll" a't .omm .o .xoooa .meow.cczi mwme raccoooom eaoonH Hmuoapam no moonasew .uwoapez daemon on» mo eowumo Hosannapsumuuwmnamoa .nnH .o . mmofl .mnonm haunno>wnb.amamoom munch «ones landmmv ooawez no pfloflfloaopen oasomoom one aroma madame: amount: oosaoaoouuomnomoa 192 uncommon .m canoe waosooom o» epoo oom "epoz oam.mm o::.mo mse.m mao.oH oco.moe ooo.sHH om:.o osm.AH mmofi omo.mm omr.om cam.: omo.o oma.mo ooo.moH oom.m oem.mH Amos moo.m :aa.sm moa.: men. oom.~m oao.:o omm.o oas.ma omoe oma.o oom.mm Haa.m mmm. ooo.ms ooo.sm omo.m om .HH mmoa mua.er and.mm m::.m mmm.o oco.ao oc:.oo ems. on .o rmoe mmm.a: ram. omm.m ooo.m omn.mm oem.om o m.m osm.a mmoe mmo.se oma.me mma.m mm.o oom.sm oom.om o 6.: ome.m «mom mmm.om mmm.m: mom.~ m:.m oam.me oo:.mm oHH.m oom.o ammo Soo.o: mmo.m: moo.m amo.m uoo.o: mmo.me ooo.m mmo.m omoe omo.on mma.mm mm .m omm.m omo.mm omm.mm aeo.m AHH.m ores mmo.om om~.mm as .m nee.m ooH.om omm.mm omm.m mmm. roe mom.mm oom.mm osm.m mom.m So~.~m omo.om moo.m moa.e roe Ha:.:m moo.am mmm.a cmr.: Hmm.m~ mom.cm oem.a Hom.m more mm:.om mom.mm mmo.H mom.m m::.oa oao.o~ oeo.a o>~.m meme mam. m Hom.om mom.e oao.m mmm.ma aom.se Ho oma.a some mam. m oo:.mm Hmo.a om».~ mos.mH Hom.me a cam.e mead mmo.m~ one. m ore oom.~ mmw.o oos.oH on moo.H were and.mm amo.mm omm.H m:~.~ one.» ~::.m em: one Here omn.om omo.am mmo.e omm.m soo.o mud.» Arm mac ores moe.o~ mos.em mm» omo.m mmw.o amo.o mrm are ones .32 one moor .88 one one some some are» ueoaam omoa noeaaoxpemamm mm-omoH .oeeaam once one comes: Hooaoeo use someone saucepan Auoeeo no asoaaaflav ”soapssmom ”ooaweZTn.h edema Kauwomod 193 uponhmlhmmmmlwmmmmunwmmoam «owed .Nm were no asoaaaoaoz one .Aomuusoa .mowooom mpdo>v maaopoeeun oo oaaOpeHom .Hssouoez moamaoaeowm «Homeosoo ”mooasom mm~.m Hom.m moe.m no» a moadeusho ocean nae pecan penanm>dmm scone ooHHoa oeoo poems oeddoa pom cacao momma Hoops unemposhpm one whom meuaommooom one ofiaam scam "meowao omoa op mmaoaooos Nomad psavso no wnapowaez "epoz ms.m:m oo.mmm mo .aeo mmo.o moo.om ome.m:~ mom.mme mm:.oo oom.~oa raa.mm omoa Hm.amm Hm.oem a: .Hmo mmm.o mmo.ma Hom.oma m~.mmH m a.mo Hom.maa m:m.am mmoa mm.aom om.~om mm:.mmm o o.mo ooo. H moo.mme mm.mma com.mm mom.om m:~.oo amoa mm.mom mo.eom omo.o~m oom.oa mH~.mH mo~.mee mam.m- rmm.om osm.mo mmm.mme mmoa om.maa mm.mse the.mem omm.cm mco.me smH.oHH oem.mae soa.ma aaa.mm mom.om mmoa Ho.msa mm.msa omo.mar omm.es osm.me so:.oa aom.:c moo.sm mea.eoe oom.mm smoe mm.ome mm.omfl and.mem :H:.o: mom.:a mso.o~ mmm.mm mum.mm meo.:~ mso.mm mmoe m~.mma om.mma mmo.oom moa.m: meo.me mmo.aa oea.em oom.m omo.am oom.aa mmos No.6mH om.oea Hom.m:m m:m.m: oao.ma omo.oo mem.mm 6mm.m H 5.5» m:m.m: amoe oo.ooe 60.66H oa.mwm mma.cm mm:.oa mmo.om mmm.om omo.sm m o.o: omo.oo omoa os.ms Ho.os mm.omm ooe.o~ Amm.HH ooo.:m oao.~m moo.mm moo.o~ mam.om area Hm.mo mo.oo sam.moe oam.o smm.o m m.m: oam.am moa.o~ moo.~m Hao.eo were mm.om mm.mm msm.:o A m.~ Hme.ae mmr.ma oem.om ooa.Hm some oa.e mm.: aeo.ma Hao.ma ores Momma Hound poems 900nm poems Henna He mouhou pseudo poouso Hmpoa evade ocean oedaoa pOHHOA cacao Innocence nmooom anew moonmaoz emancoa use -eaaao mace new means one team a tease om-oeoa .ooauomoe .nosooam to csaa> an schemes: ashore no stoma one coarse Hence no NoosH .posooam ho psmuso owadaoa ”Henoaeoz seawaoaoodm mafifluQEOOnu.w manna Naoaoamd .mmo .ooawoz_m« muooooam meadow pmHm Heepm no :Hpexamz ..<.m .ooawoz em mosaom moua< .NooH .mN Hwaom ..<.m .ooawoz op mocaom no» Sosa heaven .mmoa ..<.m .ooawoz op momaom nova< .mpao om nausea .mmmo .o .d "mooadom cow. Oman mend ammo mom H opmHo 8H9 mmm.H ocean ooHHoa oHoo mmm...--_ coaHm so» he mouem _EepH “needed omoa on woaeaoooe Noose psnpso Ho mmapmmaoz "ouoz 19h om-eroH .ooHu Hopes to wooaH .aesooam an esearc emmsmoa oeawoz omoH .posOOAm no esHo> ho commune: unease Ho HoosH one .psnpso to mowaom woodmau.o edema Hdonoond o:.:mm o:.o:~ mow.mom mos.m mm:.~m mmm.om om~.:mH omoH om.:Hm am.mom omo.Hsm omm.m ozo.Hm moo.oo oam.mo mmoH mo.on mo.Hmm mmm.mmm mo~.m om.mm ocm.oo msm.ao amoH Ho.boH mo.eoH mom.omm moo.o mm.om ooo.:o oam.ma mmoH mo.mmH ms.maH mm:.Hom amm.mH SHm.:m mou.oo mom.mc mmoH Ho.moH om.mmH mom.mmH m:~.o :mm.:~ arH.om omo.mm smoH SN.MMH Ho.HMH Hew.mmH mHm.s omH.oH Ham. 6 mmb.Hm mmoH NH.HmH ma.HmH com.H:H mHH.0H Hmo.mH on. o oom.om mmoH oS.MHH ~m.mHH oma.omH oom.HH Ham.mH 6.66 omm.om HmoH oo.OOH oo.oOH omH.oHH ms:.mH mmS.HH Hm:.mm sm:.om omoH mo.mm m~.Ho mmm.ooH mmH.HH moo.o H m.me mom.oe oon om.oo oo.ma Heo.mm 6mm.o oom.m m m.om omH.H: mroH om.mo :~.ma m::.Ho omo.mH :Ho.H mo .mm amm.oe asoH ms.a: HH.om aoH.mo mmm.eH mam.m sm~.o oeH.mm meoH 2.5 oH.mm mm .3 HoH.m 32mm mmoH mm.m mm.: no .m :o:.H moo.: msoH Momma Momma Oman cacao poem» pompoo pounce Haves coma saa ooaaoa epmam anew mooanoz smashes peso oHoo 195 Appendix Table 10.--Chilean Exchange Rates, Conversion of Data for Compania de Acero del Pacifico from.Pesos to Dollars Exchange rate Weighted annual Period used by CAP average rate Jan., 1950-Dec., 1950 30:1 30:1 Jan., 1951-Dec., 1951 60:1 60:1 Jan., 1952-July, 1952 90:1 Aug., 1952-Dec., 1952 125:1 lO&:l Jan., 1953-Dec., 1953 110:1 110:1 Jan., l95&-0ct., 195& 110:1 Nov., l95&-Dec., 195& 200:1 125:1 Jan., 1955-June, 1955 200:1 July, 1955-Dec., 1955 300:1 250:1 Jan., 1956-Apr., 1956 300:1 May, 1956-June, 1956 500:1 367:1 July, 1956-June, 1957 600:1 600:1 JUly, 1957-JUne,'1958 770:1 770:1 July, 1958-June, 1960 1050:1 1050:1 Note: The weighted average annual exchange rate was used to convert CAP data from dollars to pesos. Company reports were issued annually, covering a January to December period through 1955. In 1956 a report was issued for the first six months, and since that time reports have been issued annu- ally covering a July through June fiscal year. After con- verting the data to pesos it was shifted to an annual, January through December, basis. Conversion to pesos for l9&7-&9 was made on the basis of the official exchange rate. Sources: Compania de Acero del Pacifico, Memoria Anual (Santiago, 1950-1960/61), and International Monetary Fund, Inter- national Financial Statistics (various issues). 3336.83 u unmask sedan oauuoson .m .aeeh moaned on» non seem ooaao no mood Ho oasmeoo omoa some hooves wda posooao mmoa so cameo eae3_wo«peee name eon nammfiez_ema .apmwaez. .m .umoa we as soaposooao on cease no cause earn our wadsdaeo mensadane oMoH ma seams emmssoa .N .e nonwe .ao hem compo ewuao>e owoa euaawowe aeaaou I omofl eweanp uaoowo ”aroaaou as Heath on» ma eaea ovem.eaez madam omofi Haev Hmoa ad on mmofl dd do» men eouan emeaebe afie- ena mndsflems oopsaaaee ee3.omod ad omeswoa vacuum .H 196 «Icahn m.oma H.~mm Hom.om mem.mm oom.moH moo.ooH mos.oom mmoH H.Ho~ o.moo mom.mo sm~.mm mHH.HMH Hoo.mmH o~:.omm smoH o.ooa ~.Hoo omm.oo mm~.: Hoc.:n oom.omH erm.:m~ omoH H.mma m.mo~ omm.wm mmm.: mpm.m: and.emH mm.-~ mmoH o.HHm o.oms on». m o.oH oom.:m :H:.m~H oH.om~ smoH m.oHo m.mm» mem.- H-.HH moa.mm a::.amH cam.MHm mmoH n.aom o.ooo mHH.om mHH.on som.msH mm~.mAH mmoH :.oo: m.H:: mos.oH mom.m smm.~m moo.:oH omm.o~H HmoH o.ooH o.ooH mom.m msH.om mom.mm omoH wcoaH eeHee Muwuw Ummnwmdu monmdwdm paomwm eduaoaon Hsuoa anew moaancz mromOH .ooauo oH evo§o0hm HO 05Hm> hp cepnwwo3.N0dnH madam one mucosH aeHem e onsoa one oeneamdm .aoaom ewesmos .aaoowm one oaaaeson coaudoem Hem ahead to .eocoooam oceaHsHuHsom eHseosoo--.HH eHnaa wHeoooom 197 .oeH .o .HoooH .maou.aon H.oo comma moapeapsam Hope: apexaex Heuoz.neo«haadrunooaan oeoawmdudaoa one ooomumau enamogaoo .NH canes Haoooooduaue an ommsmo amoem e.eeosooanm anon one aHeoam HcoHaoaeHme .m oHanam . coH-omoH ass .oeauaeem Heo oneod so summoaoo one ..onH ..oo eaooooMuneeHee paoowe one empooaon ego. .oasesdoasz HSHHOpHom "omsausemv osmouaosd onenH mwoaamo sow asheoam «HaansosH an .ooepamm uemosam moaneouuomofl .eeaeu oapuoson “aeoasom .aOpoeu useaoeshoa on» we N shook sense amount area» a N «momeasdudaea .meHse emsasOp w oemeaddudaen .eeana .m.b + cease ewednoa H mommacau .eoaao .m.Dv «hwsoaoo on» mo aeHeu eaoowo Hana» on» assoc oasos_odoa meannese on» needs nsoam nose Ho aeeaao on» no son on» can» on passes oopasnos ones_moe«am 03» enema mommadwoaoos not upesooao demodwau Ho sedan eon neusoeeamea as poms no: upeaaao Apnoea-memo no ocean ooh .eouan Heep» oomnflnuu enamogaoe one made: .mposooao oemmdqahdaon no e>au .neuoam pends: on» ma .Hdda .n.o.u .meoaan openedoa edema op wwaoaoooo doummwet one: upmomoosoo enema .aposooan oomadsaudaee use oommfimdu how haepoaeoee renames: one: sedan whoowm .n Hoeanaaoov HH cHoaa AHescaoe 198 .Aomofi .oweapsemv edema e oaomm emwmmldmmumummm..ooaudomm Hoe ahead so maoegaoo «common oae.omm :em.:m~ mmm.amm :oH.omm osm.nHm mma.mcH omm.omH coacsmHna Hones ma~.sm om~.o ooH.oH mHH.HH emo.~H coo.e moo.: oonoHnHeHaem Ho~.mrm mom.~:m moH.~Hm moo.mHm oom.moH omH.ooH mmm.m~H encroach ocneHaHe Hosea meo.om so:.:H mms.m csm.HH omn.m :HH.S mmm.m ooHo one adeasposhpm mo~.moH mom.aoH per.mo mo:.mm Hmo.mo amm.mo omo.se ooaHo eHc can .aoono .oach mom.omH m:~.mmH mem.mHH m:o.mmH oom.mo osm.mo mac.m~ eaoseoan adenoaea_aeaoa umoH omoH mmoH :mmH HmoH acetone mmofi Nmoa :83 Sm-HmoH .aesooam an aoaosmHnm Hosea "ooauaesm Heo outed em sasmgsoouu.ma edema Naomoomm 199 wwwdn «Hound c.7380: .oeanaomm Hoe hood .ON .Aomofi ahODovoov N .02 e>H emOHhoEd.:dpmA how mauedasm odsomoom a.moaaoa<.maamq ad uposooam Hooum one noaH ooaaomcunuaaoowo Hence em mawemao .m .o .xomoH .omaHocemv ounpeomm one anemones moans eapnoson .NH edema Mausooo.mom m::.omm omm.~oH ooo.~oH Hmm.mHH scene one pummeaea_aeuan Hence mmoa :moH mmoH NmoH Hmoa no aoHoasHoem mm-HmoH .ecesooam connHoHuHScm to outcome “coauaeem Hem oaeo< so easeosoouu.ma canes Hdowooom 200 Appendix Table l&.--A1tos Hornos de Mexico: Estimation of Value Added and Value Added as a Per Cent of Sales, 1988-59 (thousands of pesos) Value added Year Sales Ragfimggzzigls 32:2; assgifsof 1988 1,69 1,095 599 35.36 1985 15.1573 13.287 2,191 18.19 1986 36, 36 21,690 18,886 80.63 1987 50,979 28,760 22,219 83.58 1988 71.880 88.881 26.959 37.78 1989 110,037 66,999 83,038 39.11 1950 127,817 9,685 88,172 37.69 1951 150.952 7.358 63.598 82.13 1952 167,868 100,870 66,998 39.91 1953 183,692 109,272 78.820 80.36 1958 261,852 151,699 110.153 82.07 1955 325.082 189,896 135.193 81.59 1956 833.787 282.258 191.529 88.15 1957 538.680 297.098 237.586 .83 1958 589.738 283.788 265.950 8 .3 1959 679,803 366,589 312,898 86.05 Source: Altos Hornos de Mexico, S.A., Informe Anual (Monclova, 1959). pp- 22-23. 201 . noznua usoaAop .auaouom suao>v sanopoonan on ouhouaaom .Hunoaoaz madmanhovum manna aoouuhonuo . dam anonaH uphonxm on» on vacuum .unoaovupm Hedonanamn Hanoaoaz «0H ashouam manna Boo .poom can ousonhop¢3_ooahmunmmnamoa .naaahopaa_3ah no once one aunovom avfiob as coonvoan modem "noonoom asm.mas mo~.mom.a o 6H ”1...“... “may. gm m.~ . E annmhm ME. cocoa osHa> uoHsm meow unsoaaom as .mmuamod voahoa on» mom came on nodes Quad» Ho oupun omeno>u on» wnans oopuadvno as: cocoa osaub omnhdma munch on» ham .uunooom spao> pa woosuohn uaosoona an ocean awash no doapnononm dado on» nfiupzoo hon» pun» ooaduuu an: pa cocoa osHab Hana» wnwpaadoawo nH .omoH Hound hano undofludnwau one: aunovom apHo> pa coosvoan anon» can» nonno madam .om udmod now th0 can» one .Aoanom on» hon hHao cannaaspu ohoz_unxouxsonp puoo .nhonpo one «unavom aaHo> as coonconn «pegboan .uanonogsoo o3» no pudunoo nodes u.hnugaoo mama ”opoz cocoa odH¢> Hoo.m~:.oa omo.m~o.ma we.mm apa.mm~.o Hm».~oz.o wom.~ma.:a omoa «so.mmm.m mom.omo.oa no. m :o .eoo.m owm.amfl.: omp.oh~.m omoa :m:.wma.m omm.~mm.~ a#.om ma .m~:.m oam.oo:.m :oe.~mm.o hmoa mmm.mmm.m ooe.oo~.o Ho.om aom.om:.m o:o.:am.m 0mm.omo.o omoa em~.m::.m mmo.mam.: om.~m mmm.mom.~ o:w.aaa.a moo.owo.: mmoa oam.mm:.a oua.auo.m mp.mm oo .pam.fi mao.mma.a ~o~.Hm:.m :moa Hmm.:>o.a mo:.mmw.a oe.~m a» .moo.a omm.m:~ am:.ams.a mmoa and.mum moo.moo.a om.om aoo.mmm m»:. we oo:.:am.a mmoa a:a.mma Hmm.mp:.a m~.mo woo.omm 633. o: mmm.nmm.a Hmma coups modem avcovom 05Ha> Hmpop uonum no R a as cocoa unannnoo upao> as use» Hauos 05Hm> ndmflhoaaa_3am vooduonn madam eopmaapum AnahwoNSho Ho ndndmnonuv amusdwa .uoaum Ho undo hem a as pound osaub one pound osaa> no noupuadaam "asqowouz aoaMfifihoudm «unnamaoouu.md oanaa Nduaonn< 202 .HOflub 690.0 .HO 0085.00 n .owuapnamv .uunoo gonad can» =.unoaouapm anon can panonm Huoahoanam= .Hum pandnwm .Awmaa aduo ad .au cam .ooauaoam Hop onoo¢ cu «Haanaoo can ..ooH ..oo uaoomom ”cannon .Auozana usoanapv ooauaoam H06 ahead 06 canuoaoo ho anoauuaomo on» no phenom ..onH ..oo uaommou ad copaoaoa apnoo pooaao on apnoo gonad Ho awash 0:9.aoau voaoaupno coop open space aoan Haaonow .npnoo nonwa .uuduoaa no noapaaasa on» no hdpooaaund uopufiupno 60666 osaa> .noaawaoonmoo was oncomwo vaonno>o ”00oz m.~6 6.66 m.H6 0.6m 6.6m 6.66 a.me 5.5m 6.6m uoauu mo 6 6 on condo osaa> am~.~: no:.mm o:m.o~ «no.6: mmm.mm m: .o: 666.6m muo.:m m66.» madam mm.om mom.:m oao.ua mo~.6m m6~.om mo .zm omo.mm 666.6a -6.~ 66666 66Ha> ma.H cam.6 6H:.m 66:.m 6am.: :6m.: moo.m omm A66 .an panopn mnapaaomo mmm.6 who.m H66.m oom.~ uo~.6 65m.m 666.m moo.m an .H 6monno>o Hmaonam omp npnoo henna 666 wannaanm wmu.m NHO.H emonno>o pagan mom.m Hm6.m mo~.m moo.: a:m.m ~66.n ~66.: mao.m noapaaoonaoo m66.ma 6H6.ma oam.m Ham.ma oofl.ma aoo.~a 66m.m 666.6 mom.m puoo gonna 6666aaanm :za.om mum.am :om.¢fi 6mm.mm >6m.~m moa.- H66.mm mma.:m mHH.m p.66 soonan emoa om\umma ~m\omoa 6:66 mmma 6mma mmoa «mod Hmwa omaa op .nmh AahuHHov .m.D no uvnauoonvv mmxamma-ommfl .66666 onaa> uo noapmadpnm "oedudowm Hop onoo< op mauamaoonu.oa manna Kuvnonmd 203 Appendix Table 17.--Compania de Acero del Pacifico: Estimate of Value Added, 1958-59 and 1959-60 (thousands of U.S. dollars) 1958-59 1959-60 Cost of sales (raw materials and supplies, general manufacturing expense, selling ex ense, and staff participation 39,85h 10,107 plus: depreciation 6,001 6,960 plus: profits 1,076 8.552 less: raw materials and supplies (12,961) (11,200) equals: value added 36,967 37.819 Sales 56.287 57.“1 Value added as a percent of sales 65.7 65.6 Note: Cost of raW'materials and supplies estimated from partial data for the JanuaryiMarch, 1960, July-December, 1958, and JanuaryaMarch, 1959. Sources: Koppers Co., Inc., Be ort on O erations of Com nia de Acero del Pacifico (various Issuesi, and UompanIa de Icero HeI PacITico, Memoria AnualI 1958-1959, and 1959-1260 (Santiago, l9 and . 20h .66 666 .66 .66 666666 «6666666 “condom .Aomoa an an K68 9056069 0366 on» 6085666 6606066996 £66636 nowpon6onn ommnqoa no Ne6nn 6onnwnozns one ha uoauu OM66 on» wcahamfipflda 63 609656660 66 609966 0&9 .ownano Han 9056069 non 60965666 60666 Bonn 6096656660 .6 60965666 60666 oflaa> .6966 6opaHne6n: onp_q6 66 60666 on 60666 0566» no 06966 0866 on» manna 609636660 soap on 60666 0566> .omoa d6 adenoqaoo aos6onn_nouo no e§Hu> can hp 6opamnoz_uu unmade 66066666 no Ho6n6 on» hp omoa d6 60666 no 0566» on» munhannpana ha 6opaano6 066 60666 0666 none nH «0902 666.666 666.666 666.666 6.6666 6.6666 6.6666 6666 666.666 666.666 666.666 6.6666 6.m666 6.6666 6.6666 6.666 6.6666 6666 666.666 666.666 666.666 6.6666 6. 666 6.6666 6.6666 6.6666 6.6666 6666 666.666 666.666 666.666 6.6666 6.6666 6.6666 6.6666 6.6666 6.6666 6666 666.66 W66.66 666.666 6.6666 6.6666 6.6666 6.6666 6.666 6.6666 6666 666.66 66.66 666.666 6.666 6.666 m.6666 6.6666 6.666 6.66 6 6666 666.66 666.66 666.666 6.666 6.666 .6666 6.6666 .666 6.66 6666 666.66 666.66 66 .666 6.666 6.666 6.6666 6.666 .666 6.6666 6666 66.66 666.66 66 .666 6.666 6.666 6.6666 6.666 6.666 6.666 6666 66.66 666. 6 666.666 6.666 6.666 6.6666 6.66 6.66 6.666 6666 666.66 666. 666.666 6.66 6.6 6 6.666 6666 666.66 666.66 mM6.66 6.66 6.6 6.666 W666 666.66 666.mm 6.66 6.666 6.666 6.666 666 666.66 666. 666.66 6.66 6666 666.6 666.6 666.66 6666 666 66 66 66666 666 66 6> 66666 666 66 6> 66666 “women 6q6650nuv 6606606666 nonaadav “nonemnonfiafiav 666% «m2m4 <6:6<_6n6 .666 .666 ZmU Ado 666666 66666 666 66666 no 666666666--.66 66666 66666666 205 Appendix Table 19.--Capita1-Output Ratios for CAP, CSN, and.AHMSA A. Gross fixed aaeeta/ B. Gross fixed aaeete/ Year value added, book value value added, 1950 prices CAP CSN AHMSA CAP CSN AHMSA 1985 29.62 26.39 19&6 8.72 5.56 1987 18.31 3.37 26.2h 3.76 1988 6.h0 3.05 11.22 h.71 19h9 S.h8 2.32 9.63 3.86 1950 h-OS 2.19 7.30 3.67 1951 2.29 3.19 2.23 5.99 S.h6 3.36 1952 1.%9 3.32 3.03 h.68 5.99 h-lO 1953 1. 5 3.59 3.01 n.8u 6.19 3.92 195& 1.ho 3.68 Z-hh 1.80 6.9h 3.h6 1955 0.92 2.7h 3.02 u.eu 5.75 h.21 1956 0.63 2.90 2.u2 5.29 S-Ih 3.83 195 0.82 2.25 2.3a 5.05 5.93 3-7h 195 1.18 1.82 2.75 $.35 5.83 n.33 1959 1.75 1.28 2.61 .60 5.30 n.16 0. Net fixed assets/' D. Net fixed aaeeta/ Year value added, book value value added, 1950 prices CAP csn AHMSA CAP csn AHMSA 1985 28.98 37.17 19&6 8.52 6.56 1987 18.06 3.12 25.98 n.7u 1988 6.2h 2.7 11.07 n.35 1950 3.83 1.81 7.02 3.22 1951 2.06 2.73 1.83 5.60 5.02 3.52 1952 1.2h 2.70 2.53 h.12 5.3% n.2u 1953 1.30 2.85 2.u3 h-OB 5.3 3.83 195 l-Oh 3.01 1.95 3.77 5.82 3.1a 195 0.62 1.95 2.35 3.51 8.57 3.70 1956 0.37 1.26 1.69 3.65 3.77 2.75 195 o. 1 1.20 1.h7 3.27 u.02 2.22 195 0.96 0.89 1.69 n.72 3.39 2.92 1959 1.51 0.63 1.83 5.56 3.07 2.23 (continued) ‘2 ‘n. 206 Appendix Table 19 (continued) F. Net fixed assets, E. Net fixed assets, adjusted depreciation/ Year adjusted depreciation/ value added, constant value added, 1950 prices product mix, 1950 prices CAP CSN AHMSA CAP CSN AHMSA 1905 20.38 19&6 n.93 h.17 1987 28.18 3.21 .02 2.78 19h8 9.90 3.87 .92 3.50 l9h9 7.85 3.10 7.82 2.81 1950 5.90 2.82 5.90 2.82 1951 5.75 h.21 2.55 5.38 h-27 2.59 1952 8.33 n.82 3.1a h.02 8.37 3.12 1953 8.32 h-hh 2.90 3.62 u.%2 2.93 19511 11.13 11.96 2.52 3.72 n. 9 2.58 1955 3.99 8.00 3.15 3.70 3.87 3.18 1956 u.27 3.%3 2.80 3.78 3.%6 2.82 1957 3.96 3. 2 2.69 3.23 3. 2 2.2% 1958 5.67 3.39 3.15 h.7h 3.h7 3. 1959 6.65 3.26 2.95 3.39 3.0“ G. Net fixed assets/' H. Gross fixed assets/ value added.minus value added minus Year depreciation depreciation CAP CSN AHMSA CAP CSN AHMSA 1905 1986 19h? 3.78 h.38 1988 8.75 5.81 l9h9 3.67 h.S? 1950 8.3h 3.30 13.18 n.31 1951 5.39 2.95 6.99 3.39 1953 .36 5.90 3.h6 6.00 8.22 h.70 195& 5.1% 6.77 2.92 5.95 9.h7 8.02 1956 .hl h.29 3.30 6.70 6.h2 “-53 195 3.96 8.95 3.17 6.h3 7.30 h.h0 195 .03 n.3o 3.81 10.30 6. 8 5.28 1959 n.21 3.5h 6.89 h.99 _ni . Inn-4| . "'Tl'iifitiflfinflflifl'tfllfififllflfifllflififlfliflijflfilfi“