__ - s RFSTRICTED REETURN TO | CIRCULMTING COPt Report No. WH-146a RlEPC)RTS DESKI TO E RETURNED TO REPORTS D = WITHIN J I G-EN,QP%L IVLL This report was prepared for use within the Bank and its affiliated organizations. I They do not accept responsibility for its accuracy or completeness. The report may not be published nor may it be quoted as representing their views. TNTERNATIONAL BANK FOR RECONSTRUCTION AND DEVELOPMENT TNTERNATIONAL DEVELOPMENT ASSOCIATION CURRENT ECONOMIC POSITION AND PROSPECTS OF BRAZIL Volume V Manufacturing Industry May 11, 1965 Western Hemisphere Department CURRENCY EQUIVALENT Currency Unit - Cruzeiro (svmbol Cr$ r% ~A rnrTr. 1- A r LAJtL i IIN%LI ni-I JL Quotation: June 1964 US $1 = 1, 200 Cruzeiros April 1, 1965 US $1 = 1; 840 Cruzeiros ?E3 MISSION brend A. tLe Vries .............. Chief of Mission Irving A. Sirke *..*....*..........**.*** Chief conist So Sbahid Eusain Oa*.*.... ,Gneral Economist Otto maiss ......... . . . . . eoneral Economist Heins Vergin O.. ......... ......-. General Economist Richard Goode ........... ..-..e. Adviser on Fiscal Problsms Sbigeharu Takahashi ................... Agriculture Economist Lains E. Silva *... . * . . . Agriculture Adviasr Arnauld R. de Soucy *........ a....... Education Adviser Bertil Walstedt d............ , nustrial Economist Uoln EcioUUdT, -- ------.;.--.-@ S.e . bC oi Robert ifaJ.5h-1*~EUnB1e*t1EW" AA44WM Rhna A. Adluqr ......... -?., . . Transrort Doonimist Frans A.Lo 1Inning o.o................. Tfmnsvort Economist Eric Schaefer .e** ...o.e.....e.. Hlghway Adviser E(Ward V. Breitenbach *..e............... Railroad Adviser Dietrich L. Regling ...s,..*. ... a..... Rail d Adviser Jobn A. Jaman P*** **** ort & Shipp g Adviser Thomas B. Winston *......0o.0z..e....... Editor Mis8 Var Elias *....*..*.**.e.*.... Mission Secretary Volume V Manufacturing Tndustry TABLE OF CONTENTS Page, Recent Growth and Dsevalopment Tong=rn Pro'ble.,. L1VS U]d1IlD U a4LI=1IIt=lQO 1 Conc'usions A - Approximate Grow-th Patterns for Brazilian Manufacturing Industries: 1959 (Census) - 1963 (Est.) - 1967 (Proj.) B - Investment Requirements in Manufacturing 1965-1967 C - Main Areas of Potential Import Substitution in Manufacturing ;YM1 7lnrFA C '7.17 .t -!.-TG -I'D'UOSl1t Recent Growth and Development 1. During the last two decades, Brazilian industry has grown at a vigorous rate estimated at about 7.5 percent per annum. The interruption of this growth in 1963 and 1964 was due primarily to political factors which affected business confidence and also to increasing difficulties in generating or obtaining adequate financing for investment. 2. Not only did industry grow at a high rate,; it also diversified greatly, particularly in the 1950's. airing this period the country became virtually self-sufficient in the production of motor vehicles (about $500 million were invested in that sector). Several shipyards and tractor plants were erected as well as plants producing heavy electrical and mechanical equipment. E.y 1964., two large new steel mills for the prodlction of flat products had gone into production. Changes in industrial structure between 1949 and 1959 (industrial census years) and crude estimates of changes between 1959 and 1963 are shown below. Percenta2e of value added. by industries 19L9 1959 1963 Consumer goods 52 v Intermediate oroducts 2q.1 28.0 30.8 Capital goods 22.4 33.3 34,6 100.0 100.0 100.0 Consumer goods were responsible for over one-half of the value added in 1 Q)i9; today they accounnt for only a little over one-+hird. 3r ?>Tĥir ex+vern.,L factors ,.r1-, ch mAad such g,.row.+vh possibl uer ia, large and rapidly expanding market, substantial possibilities for import substitlution, stroncg incentivres for establishing nrew inrdustries (cuastom^s prot+ect+ion, low taxes, and in some cases, financing on subsidy terms), and some very valuable foreign nve s m t (in e gJ automobile .arufacturing and hay electrlca:L equipment). Yet, the driving force was the Brazilian entrepreneur who thought big -Ar was wll.ng 4to 4-t be risks Tn terms ofP inn At first sight, it may seem inconsistent that the demand for capital and durable consumer goods should grow by about ten percent per year, i.e., considerably faster than overall investment activity. Closer study reveals that the high growth rate is due largely to assumed import substitution and (to a much lesser extent) to increased exports. Excluding these influences the growth rate would have been rather less than seven percent. Among the major growth points would be steel, chemicals, machinery and (of lesser overall impact) pulp and paper and oil refiningl/. g The prospects for steel are analyzed in Volume VI. A brief comment on the Brazilian oil industry appears in Volume I, Annex 4, page 52. 12~. The prospects of actually achieving growtch rates of the magnitude. indicated are perhaps particularly favorable in the machinery field. There are now 25-30 farly la-ge enter-rises in thi s sector covering? a range o-f equipment such as machine tools, railroad equipment, heavy electrical equipmen ts ndustr ial equipmUent, ec.Nt Mos+ of +1>hse are -b; ies of foreign firms. Their major problems are finance (particularly for work in progress and er-14s - ħ-- IN ~ ~ UL 4.L J - i1 ~-J-I' -;- progress-L ;dce st their buyers6)9 cor,peUti:on -wlthL forei,gn -ples and scmetimes design and engineering. In the latter area, purely Brazilian firm.s are han-capped both by the shortage of specialized en,ginee-ring firrmas and by severe exchange restrictions on the use of foreign consultants. ħeverthelħessg5S, fiancing remainis shne main problem in a ielu -wiiere "Ahe peF-.oU of gestation between the purchase of the necessary raw materials and the final cash payments for tne machines delivered is necessarily long. Recommendations with respect to equipment financing are made in Volume I, Annex 7. 13. Expectations of vigorous growth in the chemicals sector might be entertained on the grounds of (a) the degree of development already achielred by this industry; (b) the maturing in 1965-1967 of several irmportant petro- chemical projects now undertaken by Petrobras, and (c) investments urndcr way or planned by private industry in e.g., nitrogen and phosphate fertilizers, artificial and synthetic fibers,. plastics, carbon black, etc. Generally Epeaking, such growth would appear to be economically justified. The major problems of execution are in the petrochemicals sector (see para. 22 below). 14. Vigorous growth in pulp and paper is favored by a large and rapidly expanding market and apparently adequate raw material availabilities. It is hampered by the existing structure of the industry (mostly small producers), the lack of overall planning, and by price control for newsprint even thougi Brazilian prices are below the world market level. Long-run Problemsl/ 15. To gain perspective on Brazil's industrial future, it is useful to compare the recent past with a period stretching somewhat beyond 1970. The mission feels that Brazil will soon be reaching a crossroads of its industrial development. In the past, the dynamic of growth was provided by import substitution. with private entrepreneurs creating new industries in a climate of boom and inflation and with the Government investing heavily in oil and steel. Over the next decade there will be a marked decline in the role of import substitution. And while there is substantial scope for industrial exports it is difficult to quantifv the potential. Essentially, Brazilian industry must grow through the development of low-cost internal mass markets. 1/ The problem of industrial finance is discussed in Annex 7 of Volume I. -6- 16. From a policy point of view, import protection A tout prix must give way to studied imnort protection and deliberate export nromotion. Lower costs. higher quality, and improved industry structure will become more important than the crreation of new indu.st.rie Thi daring Pnt_n+rsnneiir must. jin forces with the cost-conscious professional manager. Industrial firms will nrofit from a- rime of' price stabili+y and the support of a capital mnrket. In a process of growth-with-efficiency, consolidation of small firms seems likalv- nod mnn-r r.ainnal fity.s will falI by the rTntrQiel ItAe d e of Sao Paulo will somehow have to be transmitted to other regions since better regio,nal balance wlll erae 4- L-e Mraet4- for- .-;A-,4-r --1 Aeed r conoA ~~ ~ ~ ULJ~ L~*O.,L Z\V U .LA. JLA%ALA V).L C"AA UJ~J growth. 17. The Government is aware of these needs. It has instituted measures t O pJr or .ot1e epUort, sto fVħacilit ate selff La.LJAħ11g, UU UgUt,eLJ.L. 441L Z)UJ)J.Ly f.L long-term capital to industry, and to improve the treatment of foreign investors. ItU is) 11-Vng16 tuwaħ-ud a iiLUL_LJU exch.aL1.ge raUre. UJUħLII oVeV-tufiiL1tl.[L,-LinuuIstry committees have been organized to recommend programs and policies for various nldustrial sectors. This is a good beginning but, in some respects, even ,nore vigorous action will be required. IMport substitution versus exports. For a country the size of Brazil, import substiLution is the normal starting point for industrial growth. Never- theless, the policies of import substitution where indefinitely prolonged will produce negative results. By the end of the decade import substitution will already have proceeded so far that it will no longer provide a dynamic element in industrial growth. -P;oreover, Brazil would be pushing into fields where its comprehensive advantages are slight or non-existent. Lastly but most important, in a protected market, comretition would decline and there would be a danger of deterioration in the industrial fabric. 19. Until recently, Brazilian industry has been protected from foreign. competition by too formidable defense lines: differential exchange rates and a protective tariff. Towards the end of December 196h, the exchange rate system was revised, and a substantial move made towards a unified rate. The free rate was raised from 1650 to 1850 which benefited exports, and the im:port surcharge was reduced from 30 to ten percent. Under the Government policies, the compulsory deposit on imports will be progressively eliminated. Never- theless, even a unified exchange rate will involve substantial discrimination against exports in view of heavy protective duties, ranging typically from 40 to 120 percent. This duty structure, especially the relationship between different duties, often runs contrary to the country's long term development interests. Thus, duties on textiles are prohibitive while duties on machinery and equipment are generally low. The latter industry is a dynamic growth point where "infant industry" support is desirable; the former industry would no doubt benefit from the pressure of foreign competition. The main instrument of protection is the "law of similars" which prohibits imports financed with external credits if similar items are produced in Brazil, and if the imports are financed without credits prohibits reduction in the existing tariff for such items. Thus, any Brazilian manufacturer who can demonstrate plans to produce items similar to those previously imported is more or less automatically provided duty protection needed to cover his costs, -7- however high Apart from its overlv nrotective intent; the law of similars puts too much faith in administrative discretion. 20. In the past, there has been little Government support of manufactured exprnts_, - d ht.q > duan echonme n ra sys-tem F -n in u n yll ,0nfr,t'rnble to exporters. Steep fluctuations in exchange rates have also discouraged exports si mn 4 + is A- C4 II+ +n 4 -r 0+ 4e ln,n+ --. org ,a +4 en - r nv.n+i n ishcrn sales one month may be extremely profitable and the next month a losing proponnn4+4 n h ".. mvee.nt- toward s a un4 -IA -4--ng a -t 'I An---d a V~ Wk' UJS o~W a L AL1 USl V ~u~ U U' u0J 'L LL.LħħU I. uV U'.L._.L. S~ great stimulus to exports. In addition, the Government has alnounced a series o1f7 m.easures desi.gned. t lo inc.ease exports, .L.L1L;J_Lu.dr s.Lf1icatL.ħUIon.LIV41 of- e Uor licensing (essentially a problem of preventing camouflaged capital exports) ad Ldraw-Jbacdk procedures, bettu t cie rLc;ħi L.L r ezpuv It ħiiUan ng. etc. Additional assistance for market study and promotion may be desirable. 21. Industry structure. The Brazilian textile industry is not in good shape. There are some efficient manufacturers and a vast group oI mailuaC ub- urers working with outmoded equipment. The introduction of synthetic fibers is lagging. Foreign competition is cut off by a tarifI waii. Saies taxes favor integration even where specialized dyeing and printing establishments would be economically superior. Small factory establishments, by avoiding taxes and by paying low wages subsist in spite of extremely low productivity. The Economic Commission for Latin America a few years ago made a pioneering study of the industry but there has been hardly any follow-up on this work., 22. Another case in point is petrochemicals. The large Petrobras refineries already existing or under construction will be able to supply all the needed petrochemical feedstocks. The Brazilian market is also large enough to pernmit reasonable economies of scale for most products. Yet, the growth of this sector thus far has been disappointing, and prices are excessive. On one side is Petrobras which will soon have a monopcly not only legally for feedstocks but also de facto for many intermediates, and eventually will be under strong pressure to move into final products. Will it make an economic and financial success in a field where world market prices for standard items are low, and profits customarily earned as a fruit of heavy investments in new products and new markets? - On the other side are a large number of private companies, many foreign-owned, jockeying for a position in the Brazilian market, hesitating to make major commitments in an area where there would be no competitive supply of major raw materials, and where repatriation of earnings is subject to various doubts and limitations. The question then is how best to achieve a partnership, or perhaps several partnerships, between Brazilian control of essential raw materials and foreign know-how. 23. Plan-ing for industrial grjith. While the central problem of growth is one of mobilizing and guidding the private initiative; sich growth twillI put a very heavy burden on Government creative thinking and Government adrmLn:s- trati nn . 2)1 In +te first place, a vision of fte Ti-hte most economnc lines of development of Brazilian industry will need to be supported by many studie's an.d translatedw into speci-'c ++~-,-rO New policies *.411 meon. a h,.zear Tj+th the past; they fill upset some vested interests, and may necessitate some nlnnaes inr Adeeply r ^+-A --1 r -oted1 p-al 4-zAe,s. 'Th a .d.Ie . cbLo sen ins-ven+ for getting the industrial program on the way is a series of joint Government- s urJ LAWLIU~U ULW0 # LJL LU W-L ULI someL no dbLU tIL 'AAAL.Fj I.-LVI.I03 ULJ4V$e ULIUI II u indu t y co---',4tees. ut, 4 it som notabl Cxep-4ns thtco ,4-,ees are not getting their work done. 25. Apart from the launching of industrial advisory committees, the Government' s tasks include (a) tne revarping oI import protection asid exparu incentives, and (b) delicate decisions regarding foreign participation in petrochemicals development, etc Miajor projects in steel and petrochemicals are also under execution. The recruitment of personnel capable of analyzing and dealing with these problems is far behind schedule. 26, At the plant level, efficient industrialization will call for a greatly increased supply of skilled administrators, accountants, and auditors. Industrial engineers are also likely to prove a bottleneck. Educational facilities in all these areas will need to be greatly strengthened (cf. Volume VIII). 27. Regional industrialization. Lagging industrialization in Brazil is a problem chiefly for the Northeast. Industrial employment opportunities are also poor in the North but in that area low population densities create serious barriers to industrialization. In the extreme South (Parana-Santa Catarina- Rio Grande do Sul), the solid agricultural and forest base provides both agricultural employment opportunities and a better market for 'Local industries. Next to the Northeast, the need for induced industrialization 'is probably the greatest in Mlinas Gerais and Goias. To some extent, however, M4inas would appear to grow through natural extension of the Sao Paulo-Rio industrial region. 28. Industrialization in Brazil really started in the Northeast (sugar and textiles) but until recently the relative position of that region stead:Lly deteriorated. In 1959, industrial employment as a percentage of the total population was about 1.0 as compared with a national average of 2.5 and figures of 5.0-8.0 for the most highly industrialized states. Moreover, the textile industry which was outworn and obsolete accounted for about 38 percent of the total employment whereas e.g., mechanical industries were unimportant. 29. Counting about 25 million people, the Northeast region represents a large and rapidly growing market. It is protected from competition by high transport costs, arising mainly from geographic conditions -9- but also from the poor condition of the railroad and coast-wise shipping systems. This nrotection is accentuaat d by thg rate structure, Rraziaini rail rates are not very regressive by distance; moreover rates for shipmen's frnm tfli Jrl;heart ar ahc-t. lin1f tfhp raf+ fnor tehp msnm kinrl of traffic moving into the Northeast0 30. The t-k, basic difficulties hampering industrial growth are a slim agricultural base iTth a poor irrigation potential n the one hanhd anr d ann undernourished, undereducated population on the other hand. Since climatic cnditio.nsa n,.a , A- a4r more -. ndA in, ds-t-l I--lnm,-4- opporunlt 4 ies more limited, it is difficult to attract skilled staff and management to the region. Tn P4 -,- --- fact, - -hs ,,- of -k.l co,-A - lo pecr .t -- -.; e rer.,neration4in U . -V, UL1-~Q~ L.YjJkUO u.. 0~M L.LJ0 .UiJAIWkCZIIU CL J jJħL J. LJ.L-LiJ.L.L L UIIU.LULħ V. UL _I.JI 1 the Northeast than in Sao Paulo. Industrial integration within the region is Calso Lhd',ullyul- Dy ui,h ine 4i, the V UIihua.e U ,.qU U IjUU IiU n,L'L-JM.L, especially the poor condi,tion of the roads north of Bahia and inadequate coasta bL or1ħpjIu 31. Tne Government is promoting tthe industrialization of the Northeast through tax benefits, preferential financing, and sector and project studies. The former are most important, including exemption from taxes and duties 1/ on imported equipment, from income tax and, in some states, from sales taxes- The existence of the Banco do Nordeste makes it easier to obtain long-term industrial loans in the Northeast than in any other part of Brazil and on very favorable terms./. 32, This type of program is bound to produce effects, and competent observers estimate that industrial prodiction has recently expanded more rapidly in the Northeast than in other parts of Brazil. 2Iuch of the growth has occurred in the State of Bahia where the oil boom has created a new climate propitious to investments. Projects already decided or under discussion for Bahia include carbon black, caustic soda, polyvinyl chloride, 1/ In a recent textile mill project, the customs duty exemptions reduced the cost of the investment by about 13 percent. In addition, firms invest- ing in approved industrial projects in the Northeast receive tax credits up to 50 percent of the income taxes due. Exemption from state sales tax might mean a reduction by nearly six percent in the cost of production. Finally, in a country wlhere profits on the order of 25 percent on the investment before taxes are not uncommon, a ten-year exemption from 25 percent Federal income tax adds greatly to the competitiveness and financial attractiveness of the operation. The usual conditions of lending are that the borrower accepts the exchange risks on the foreign exchange portion of the loan which may carry interest at seven percent. The effective interest on the domestic portion is m,fent1y about 21 percent which. at the recent rate of inflation.. involves a substantial subsidy element. The term of the loan is geared to the repayment capacity of the bhorrower hiut has never exc-eeded ton vers. -10- steel, ferro-manganese, oil field supplies, tractors, and nylon. Over the longer term, one might foresee a gradual shift of the textile industry towards the Northeast (similar to the migration of the industry to the South in the United States) and, on a much smaller scale, of leather and shoe manufacturing to take advantage of local raw materials. (Local hides are only partially salvaged and often poorly treated; substantial quantities of goatskin aie exported to the South). The major difficulties arise from the size cf the problem (ideally 50,000 new industrial jobs should be created each year at an estimated investment of $2,500 per job) and difficulties in acquiring complementary administrators, engineers, and foremen at reasonable remuner- ations. Investment Renuirements 33. A good estimntn nf invPqteme-nPt reniuirements in manuifactuiring 1965-1967 would require (a) information on existing capacity reserves in different. sectors; (b) on the extent to whrch growt.Xh can be obtained thrrough expnnsion of existing plants as compared with the more expensive alternative of building entirely newT plants - -d (c) reliable investment cost estimates for a certain number of major projects. Such information is generally lacking except for very n prelim-J n-J - stud,h nes of cher -Aca' oil ning, eel .d non=ferr,sc metals. 34. Based upon estimates furnished by the Government and the BNDE, we have pieced together the following picture of invest ent requirements in raiħuaufaciu-L.Lig 1964i.-J4-7 I fT* _ detal - -ee A__x 0/.- Gross fl'xed' Jnvestmnent rvlru"i shu-re -of eachJ_se-;V-,- J_ ;1- (Annual average in Fixed Increase MIJ.ħIion uS$ eouiva.lentuj investment net ouIupul Steel i80 22.5 9.7 Chemicals 140 17.5 13.1 Food, beverages and tobacco (90) (11.3) 9.9 Machinery and equipment 80 10 19.3 Non-metallic metals 50 6.2 7.1 Textiles 45 5.6 6.5 Oil refining 40 D 4,0 Pulp and paper 35 4.4 4.5 Non-ferrous metals 15 1.9 1.1 Other 125 15.6 2h.8 Totals 800 100.0 100.0 35. There is, of course, a considerable margin of error in the above estimates--particularly in the estimates for individual industries but also in the mamufacturing sector total. Nievertheless, our independent cross-checks are not consistent wiith the above total (see Annex B). 1/ The totals do not include increased working capital requirements for which w.1e P ave bhen unahl e +.t nrovrid a n iusqt:ry hbAkdiwn - Tho agppreppate re- quirements of the Brazilian economy for increased inventories are estimated in ol. I, pp. 8 anrd )3a. 36. The Table suggests the heavy concentration of investment effort needed in the steel ad c cals indusurie N o'y e e investments in these sectors great; they also include some very large projects. Mlost of the i,nvestmenlts in steel woUld Li Uco ,Ufor by three large mill expansions. Petrobras alone would propose to invest up to $60 1.1-llion equivZeULL ir. Li pe d_rUCa o ħcaħlU sector5 particularly in interreuħae_s (benzene, ethylene, butadiene and styrene) and nitrogeneous fertilizers. %JXLUi' IrI JSor pruQuctbj several of which are alreaay In an aavanced state Of planning or even execution, wiculd be adipic acid and adiponitrile (raw materiais for nylon), polyethylene. sulphuric acid and phiospnate fertilizers, caustic soda, aoda ash, calcium carbide and titanium dioxide. Except for soda ash and titanium dioxide,these are private projects and the bul-c of the investments will come from half a dozen rmiajor producers. 37. Ihe execution of such major investments within a four-year period (of which one year has already gone by) will call for a tremendous effort. The most difficult task, in terms of planning and execution, will fall upon the two new steel companies (Cosipa and Usiminas) and upon the Petrochemicals Department of Petrobras. But the real bottleneck in industrial expansion may well turn out to be a financial one. Industrial finance is discussed in Volume V, Annex 7. Conclusions 38. Our conclusions with respect to the industrial sector may be summarized as follows: 1. Great growth potential. The combination of a large market, ample raw materials, vigorous entrepreneurship, and low-co3t, excellent labor provides a unique basis for growth. 2. Brazil at the cross-roads. In the past, the dynamic of growth was provided by import substitution, with private entrepreneurs creating new industries in a climate of inflation while the Government was investing heavily in oil and steel. In the future, private entrepre- neurs must turn increasingly towards the creation of mass consumer markets. Exports must be given as much attention as import substi- tution. Lower costs, higher quality and improved industry structure will become more important than the creation of new enterprises. 3. ImDortance of planning. This change in direction will renuire a major reorientation of psychological attitudes, both in Government and in industrvy and in the institutional framework. A close Government- industry partnership will be needed to draw up long-term programs for major industrial sectors= Prinninnl tets-ioni n wniil, of - nf' bn e unp to indlustrv hut the Government should improve its work on both a staff and executive level to provide assistance and direction., -12- 4. Large investment requirements. There are strong indications that the investim.ents required for futurBe growth may have been ulnder- estimated, The mission estimates that manufacturing investments on the order of $8nn r.illon p-ii4rlTa ernt ,vr yer s 1 1n Abe di"ablez,. as a minimurm for the period 1964-1967 (including oil refining but nXcl,,A-n" --;' pouco -ar. 'D-1ort) 4.bi --r--t,,nts 4n steel and oil refining would account for about one-quarter of this 4total. 5 * ReħU;onal l L,.'Liħi zLa-- Regionon. a iuuLOUl.E1J1biti is Ix*Js;ħly important to a rapid expansion of markets. Improvements in infra- str-cutur-e Ouwbdstant - as T ncanIaves, and preferenTial faindllUlIlg will draw industries to the Northeast, Kinas Gerais, and the Extreme South, Telecommuications between the Center (Rio-Sao Paulo) and the periphery are poor; and few investments would have higher ecornomic priority and do more to further regional growth than an improved telecommunications system. The main obstacles to an industrial take- off in the Northeast, however, is probably the acute scarcity and high cost of skilled staff and labor. Ways must therefore be found to increase and retain the supply of regional talent.. AIiXNX A APKPROX:aB22: CaMITH PATTE i_ F MAN1&A&T12JUNG INDU 3TRIE,S jj§US) -196 (EST . ((Iross outpLt values in biLlion U.S. $ equivalent 9/ Sectsr - ]9MA} _ _ _ ~~~~~~~~~~~1963_ 96___ Sector.~ ~~~ FL C, I EX 7 1 c CP I Ex C Cons,er Goods IxLdus rie Food, beverages and tobaccov2 2.18 0.02 0.10 2.10 2.55 0.03 0.10 2.48 3.08 0.04 0.15 2.97 Text:iles (.'°° 0.00 0.0( 0.983 1.19 0.00 0.00 1.19 1.45 0.00 0.00 1.45 Clothing and shoes 0.27 0.00 0.(0 0.27 0.34 0.00 0.00 0.34 0.44 0.00 0.00 0.44 Furniture 0.15 0.00 0.)0 0.15 0.19 0.00 0.00 0.19 0.25 0.00 0.00 0.25 Phanraceuticals 0.15 0.01 0.(0 0.16 0.204W 0.01 0.00 0.20 0.25 0.01. 0.01 0.25 Soaps and perfumes 0.12 0.00 0.00 O.1L 0.1.5 0D.00 0.00 0.15 0.18 0.00 0.00 0.18 Prin-ting and publishing Qtm. _j1 0.Q( 0 2 Q02 0. 01 Q0.Q Q.?. 0.28 2aO °-°° 5022 Sub-total n nQ Q9 _Sa9j LI QrL1 Q. 5ai' 9AL 06 q=elLda"e Proaucts, Timber and wood products 0).21 0.00 0(.4 0.17 0.25 0.00 0.04 0.21 0.30 0.00 0 052i 0.2521 Pulp and p per 0.24 0.05 0.(00 .29 0.35 0.04 0.00 0.39 0.53 0.02 / 0.00 C) . 5f; Leather aad products C0.0 9 0.00 0.00 0.09 0.10 0.00 0.00 0.10 0.12 O.0o 0.00 0.12 Rubber products 0.17 0.01 O.00 0.18 0.26 0.00 0.00 0.26 0.35 0.00 0.00 u.35 Plastic products 0.05 N.A. N.A. 0.05 n.o8 0.00 0.00 0.08 0.1. 0.00C 0.00 0 .11w Oil Refining (0.18) 0.14 0.00 (0.32) 0 33 0.07 0.00 0.40 0.49 0.05; 0.00 C)-5h CheinLLoals (0.53) 0.10 O.C)1 (0.62) 0.80' 0.17 0,01 0.96 1.32 02Ci 0.02 3 5CtZ Sundry products C292 _S00 9D _" QQQQ _0 41Z O 2 0.00 ,)1 SSub-total I26 Qa1.82 2.2 O QQL 2AZ 339 0.27. 0a.0 .652 Non-meta1Lic minerals 0.3, 0.01 0.00 0.37i 0.49 (.01 0.00 0.50 0.68 0.02 0.00 O.7C;i Stee]L (0.22 ) 0.09 0.00 (0.31.) 0.32 (.09 0.00 0.41 0.56 0.04 0.04 O.56>J Non-,ferrouas metals (0.03) 0.03 0.O0 (0.06) 0.05 0(.07 0.00 0.12 0.08 0.0s; 0.00 0'.16 V Metal products (0.58) 0.05 0.00 (0.63) 0.91 0.04 0.00 0.95 1.28 0.04. 0000 1.32=1 NichnLicaL eqiLpment 0.22 0.23 0.00 0.45 0.36 0).23 0.00 0.59 0.48 0.37 O.0 1!. Electrical & Communications equipment 0.30 0.10 0.00 o.40 O.'54 0.08 o.oo 0.62 0.79 0.09 0.00 088O Transport equipment "12 (. Q00 PA21 *()12 0.00 1A05 ;A8 P.08 5Qa Sub-total Z24 Q OQ 0.C) ; p.d O2& -.2A 5.27, 0.06 ALU Grand Total 0 L .29 -, L _L__ difi Ice 3 of l?our-Kear Grow tU %i 106 1 106 12 DOTNO TES / The foLLowirug comments explain the msin procedures used in constructing the t,shle: a) The starting point forz the lE'aziian production figures is the 1960 census, giving gross outputs and net outputs (vaLues added) for the year 1959, by industries. b) In order to construct supply (dometstic production plus imports) and demend (exports plrus consumption) balance sheets, dozestic production figures for 1959 hadl to be converted into U.S. dollar equivalents. (:Ebcport and imnport values are actually shcom in IJ.S. dollars in officisJL trade statistics). The choice of an appropriate exchange rate is obviously important to this procedure. According to INF stsatistics, the average free rate in ]L959 vwas 204 Cr. per U.S. dollar while the average exporlt rate was 76. The EconDmic Commission for Latin AmeriLca in its study of the "Chenical Industry in latin America" (New York, 1963) used a rate of 165 for 1959 (for chemicals on=L). ECILA added the coament that at this rate, the average price for chemicals preduced in Brazil were beLow the U.S. average. The average rate used in official trade statistics varied according to the commedity group (171 for chemicals, 100 for zachinery ad equitLFent which is by far the largest group, 118 for sundry manufacturee). The 1959 average rate, for imported manmfactures ass about 115 but rough allowance for import duties actually pEid would bring this rate above 130. There is clearly no sin,g3Le iceally suitable rate for purposes of adding tradLe and production statistics. In selecting a rate of 150, we had in mind the fact that charges other than import duties probably added significantly to an import price which would be diractly comparable with prices f'or dormestic nanufiactures f.o.b. plant. By applying this rate unif'ormly to aiLL items, we theoretically overstated the relative Talue of machinery imports. Yet, for all .wE know, the average real value of the equipnent imported may have reflected a rate in excess of 150. c) 1963 totals were extrapolated fromt the :1959 figures baEed mainly upon relatively fragmentary production Indices published by thei Varg3s Fouindation and apon physical production figures available for some secuors (e.g., steel, cement, automobile production, etc.). d) A possible diffeirence in the avereLge priLce level for manufactured imports in 19633 as compared with 1959 was ignored. e) As a iurther check on the figures used above, we (i) compsred industrial census totals with national inoome statistics and (ii) spelled out the imnplication of' the exchange rate used in terms of per capita gross national product. This gave the followring results: Gross3 90mestlc Produe , in tsurcisr= Total Mfg. 1959 Qross Domestic Incame (Cr. $ trillon) 1.42 0.36 1959 Gross Domestic Product ( tt ) 1.79 (C0.45) The gross domestic product figure for manufacturing was derived on the assumaption that manufacturing would have the same share in GDP as in GDI (25.3 per cent). The ZNLtional Accounts total cif Cr. $ 0.45 tri]LLion compares with a ceLsus total of Cr. t 0.54 trillion. The diifference, it is believed, is due mainly to the inclusion of indirect taxes in the census tables. Per CapRta oss Dome t,c P )duct Dividing the above total of Cr. $ :1.79 trillion by the estimated population of 69 million gives a per capiLa gross nationaL product o' Cr. rK 26,000. At a rate of 150, this corresponds to EL per iapita gross domestic product in 1959 equivalent to a'bout U.S. i175. / In this context, items like saugar and meat are regarded as products of industry, if they eLre processed in wanufacturinc establishments. ltke ai-r i .i kmat ;_c'i'g nlant_. In some ptlhnr -rti-E cl tt t,z ? port, notably biL 'we discussion of Ioreign trade, meat and sugar are treated as products of agriculture. ,/ Changes in nroriuction 1959-1923 were estiniabed based nainly upon figures pubJ.ished by the Vargas Foa.ndation. Our ovr extrapola1.ions are shown wiLthin brackets jiicep of Phyrsical Out--,. hnwil ner cent Incre,e Total-per cnt increas_ 1960 1961, ]9 9 19592 Food 6 5 (4) -.8 i Beverages -.3 16 2.0 ) 17 Tobacco 3 15 0.5 ) Textiles 9 6.5 (7) -2.7 21 CLothing and shoes 0.8 26) Fairniture 2) Pharmtceuticals 9 12,5 12.5 (-3.5) 33/ Scaps and perfumes 4 13 2 4 25 Printing and publishing 9.5 6 (2.5) (2.5) 22 Timber and wood products 2.5 I.82 Palp and paper 7.8 5;0Q/ L3ather and products -7.2 17.4 Rubber products 0.9 7id/ ILastic prodtucts 44 0:1 refining 91 Chelmicals 10 13 14 3.6 a 33ndry products 21 Non-metallic minerals 1.6 6 7.5 -0.1 ' 4 Steel and ferro-alloys 11 9.5 7 3.4 Non-ferrous mretals bi5 A)tal products hachanical equipmert (14) (14) (20) 2.6 0 Electrical and commmunications equipment; 23 24 1(24) -3.9 831 Transport equipment 36 11 (30) -10.7 7W LI Figures for 1959-1962 from Economc Co nipsio _gr Iatin Amer gas "IDevelopment of Chemical Industries in Latin Alerica, 1959-1962 (ST/EClA/C0KF.15/L.4, November 12, 1964.). 2/ AccordLing to a recent FAO study, "Latin American Timber Trends and Prospects" (1963), consumption of imajor wood products grew by 2.7 per cent per year between 1948-1951 and 1956-1959; or say bt 12 per cent over a four-year period. It may be noted, however, that in. 1963, a year of depressed business conditions, production grew at about that same rate. We have assumad an income ealasticity of the demand for wood. products in relation to GN.P of 0.8. _/ Cia T. JanAr shows an iLncrease in Brazilian production of paper and board from 440,00 tons in 1959 t'o 560,000 tons in 1962, or an increase 1: 27 per cent over three years. Over the same period., according to estimates by J.C. Leone & Associates, domestic production of cellulose rose! from 79,CCO tons to 170,000 tons, or by 11'5 per cent. As a very rough estimate, we assumed a total increase in the production of pulpand paper by 40 per cent bettween 1959 and 1962. Adding the Vargas Foundation estimate of 7.8 per cent growth between 1962 and 1963, gives a four-year growth total of about 50 per cent. dJ Production of auto tires increased, as follows between 1959 and 1962 (million tires) 1SS 1962 Truck and bus tirea 0.80 1-1_ Motor car tires 1.11 1.72 QthL r vehicles 022S3 02.1 Index 152 -3- e/ Based upon the increase in the volume of crude processed (from g.6 m.-illion cu.m. in ]L959 to 16.3 million cu.m. in 1963), without a8llowanice for chanLges in the product composition. fj Based upon the increase in cement production. The Vargas Foundation shows a 33 per cent increase for the whole group of industries processing non-metallic minerals. g/ Based uDon a preliminary estimate of the increase in steel prodi)ntion 195'-19563 Published census and production index data give no breakdown for the "Metallurgical Sector," as between steel non-ferrous metals, and metal products. According to a rough (and very uncertain) calculation, the breakdown of the 1959 gross value mayr have been as fo:Llows (figures in million U.S. S equivalent): Steel .216 Non-ferrous metals 31 Metal products 579 j/ The increase in the gross value of production of major metals between 1959 and 1963 was computed as follows (based upon BNDE data): ~aantities itnp) ( 8 ;iin eqmiln-a a. ) 1L9519 1961 Oka > Copper 1.B 2.0 1.24 1.38 Aluminum 15.2 22.0 7 .5 10.9 Lead 5.5 17.0 1.04 3.2 Tin 1.25 2.4 3.1 6.0 Nickel 0.09 0.43 0.1 0.7 Gold(021 3.7 4 '.0 4.3 Silver (02) 7.9 7 _) 3 * Sub-total Lin2l &. Index l The prices used in the above calculations are the approximate 1963 import values. The difference between the 1959 figure of '17.3 million shown above and the f:Lgure of 3 :.illion equivalent used for the mEain tab3Le in Annex A is intended as an approximate reflection of (a) quality and processinLg difftrena1a as compared with the crude metals on which the above values are based and (b) the difference between Brazilian anad world market prices., j/ It is difficult to estimate the oultput of this sector. Since foreigu trade is relatively unimportant, procluctioin will grow roughly with consumption. Consumption, in turn, is inflluenced in part 1y the activity in the mnechanical and equipmeLent industries which, aczording to our estimates, wolaldheve grown by about 73 per cent between 1959 and 1963. Moreover, the change! from assembly-type manufacture to more and more integrated p-oduction would uldoubtedly have increased the demand for metals. On the other hanc, metal prcoducts also includie a nunber of Items for the household, agriculture, and building trsdies, the demand' for which presumably grew at a far slower rELte. Our estimate of a 50 per cent increase in consumpt.ion is probs.bly conservative. Because of some imporl substitution in 1959-1963, this corresponds to a 56 per cent increase in domestic production. 4/ The Vargas Foundation shows the following chasnges: Indices (Base Year = 100) 195292/ 196l/162 12L.&6L 19S9/12 Mcc; 1ni al -ndsLuotritls 1 . 3 . (.2LJ) 1.03 i. Electrical ancd cossnications 1.53 (l.2L) 0.96 1.83 equipment Tz-anLport equJsment 1.51 1.30 0.89 1.75 Figures witnin bnacVets arc nmi:S;ilg links roug-ly estimazed b9y ourszlves. -4- &/ See Development oi' Chemical Industries in Latin America 1959c-62 ST/ECLA/CONF.;L5/6.4,, Table 19. The 1962 total shown in that source has been inflated by the increase 1962-63 according to the Vargas Foundationi. / There are essentially two ways of projecting the demand for consumaer -oods 1967. One is the simnple extrapolation of past trenads. The other wouLld make certain assumptions regardin; the income elasticities of dei:and, based e.g., upon international comparisons. The general starting point for our projectLons is tLt the gross nationa2L prodsct would grow by 6 per cent per year non-:umlatively, i.e., by 24 per cent between 1963 and 1967. This comipares with eLn e stimated growth 1959-1963, accordling tc0 the Vargas Foundetion of 22 per cent. Since, the projected growrth in the gross national product 1963-1967 is so close to the estimated actual growth 1959-1963, one imight envisage that the growth rates for individual corLsumer goods industries might atlso be very similar. Ihe inc,ome elaiticities imLplied in this as3umption would be as folrLows: Food 0.59 Text'Lles 0.94 Clothing 1.41L Furniture 1.45 Pharmaceut:Lcals L.30 Soaps and perfum3s L.30 Prinlting and publishing Qi loal maiufactired con3umer goodul L.26 These elasticities are not unreasonable when compared with similar figures for other countries. In our projections for 1967, however, we have rade a few modifications. The figure fcor text;4es seems high. The DAono4mic CoGomiLssion for LIAtin America, in its study of "The Textile Industry of Brazil" (1963), suggests an inc0o7n elasticity cf only C.65. For reasons&mentioned elsewhere we freel that there is substantial scope for a reduction in the real price for textiles in Brazil which should have some eff'ect upon demand. We have assumed a combined iLncome and price effect yield:Lng an elasticity of 0.8:5. The figure for pricnting tad publishing seemes abnormally low, End w(e have adopted an income elasticity of 1.1. W EZxports 1967 wouldl be essentiEL1ly in meat (about $85 milliorn equivalent) and ujgar ($51 mLLlion equivalent). 72/ The supplies of pine timber appear to be dwindling. Although substitutLon of hardwood timber for pine timber will occur, prices for timber over the next f'our-year period are likely to rise.. In the absence of a thorough market stiudy it is diFficult to telL what effect this will have upon demand. Our estirate of' a 20 per cent increase in the consumption of' wood products as compatred with a 24 per cent increase in the gross national product may be on the high side. 'W'ith respect to exports, we have assumed a very slight increase in the suipplies of Parana pine available for export (mainly to Argentina) say from .94 million cu.,m. in 1963 to 1.0 millicn cu.m. in 1967. Because of the shortage, prices for pine :Lumber, both for domestic consumption and for exports, are assumed to increasei from $41.50 per cu.m. to $50. Allowing also for exports of other wood manufactures (Stat. No.. 224), we estimate that total lumber exports will grow fromz; $40 milliorL in 1963 to $51.5 million equiv- alent. j/ Based upon an assumed income elasticity of dexand of L.65, we arr:Lve at a 410 per cent increase in the demand for paper by 1967. Tbhs estimate coincides with a similar estimate oade by an independent paper expert,. Imports oF specLal quaLLities wiLL continue but at a reduced rate. There is a surplus of short-fibre pulp but export prospects are not promising. 2/ Based upon Fast trends aLnd the assuaed growth in gross national product, domestic consumption of refined oil products is expected to grow by 33 per cent. The scheduled addition of two new refineries will makie Brazil virtuallr self-sufficient in refined oil products except for liquified petroleum gas and lubricating oils. IO/ According to a recent ECIA study-/', production of chemicals (excluding 5oaps, detergents and pharmaceutical products) grew by acout 43 per cent between 1959 and 1.962. Even though there was only minor growth in 1963 (according to estimates made by the Vargas Foundation), production in 1963 was prcbably asout 50 per cent higher than in 1959. J/ DeveloFm,ent of' the Chemical Incu stries in Iatin AmsŽrica, 1959-1962. According to aniother ECLA s tudy, the demand for chemicals in Brazil (on the assumption of a 6 per cent per year growth in the fross national product) was exoected to grow by 15 per cent per year between 3L959 and 1965, fal'Ling off to 10.6 per cent per year between 1965 aind 1970. Brazilian consumption of certain mEajor chemicals (e.g., raw materials for plastics ancd synthetic fibers, fertilizers, etc. ) is in fact; abnormally low, and rapid growth in consumption over the next fire years would seemn possible. On this basis, we have assumed that; consumptioni of chemicaLs will grow ty 14 per cent per annuma, or say by 56 per cent over tUe next four years. An ites 'by iten appraisal suggests that production could grow even more rapidly, say by 65 per cent. TLhis reflects the maturinig in 1965-67 of several important petrochemical projects we being built by Bstrobras as well as investments uncier war or planned by private :industry in e.g., ztitrogen and phospate fertilizers, artificial and synttLetic fibers, carbon black, otc. 1/ After conversation with an industry expert, we haire accepted his estimate that consumption of coment will grow froma 5.2 million tons in 1963 to 7.2 million tons by 1967, an inerease by 38 per cent. Deanci for other construction materials is expected to grow in the same proportion. / It is assumed that the output of' roiled steel will grow from 2.15 adllion tons to 3.75 millLon tonis, that intports will fall from 0.4 imillion tons to 0.2 million tons, and that exForts on the order of 0.25 million tons will be developed. Domestic consulmption would gronr by 9.5 per cent per annum between 1963 and 1967. ThB follDWing price assumptions weire useid: Domestic prices $154 per ton; import prices $200 per ton; export prices $160 per ton. ]/ Genera-lly one would expeec; the demand for metal to vary closely with the output of metal products (see i]ootnote 14). Allovwing for some effort to out down the consumption of metals which are expensive in Birazil, this would yield an inelease in consumption, sayr, 35 per cent between 1963 and 1967. BMDE, however, expects demand to grow only by 15 per cent (extrapolating past trends on the implicit assunptLon that the high apparent consumption in 1963 was due, in great part, to inve:atory accumulation). Pending a more intensive studly of what actually happexned in 1963, we have retained the assumption of a 35 per cent increase in demand. Estimates of domestic production of metals are based uporn BNDE projections and known pLans. Improved mimming incentives introduced by the Goverrment have been taken into account. Prices for imported maetals are bused upon 1963 average imiport ,ralucs, Prices for domestically produced imetals are considerably higher. This reflects existing higb protection and (we hiave assumed) a higher average degree of fabrication for domestic imetal shapes. 2/ We have assumead that the demand for metal products woul3d increxase at; aboult the sane rate as activity in the machinery and equipment industries, allowing for some reduction in the use of metals pEir unit of oiutput. Specifically, we bave assaecd a 40 per cent increase in the demsand for metal products as compared with a 45 per cent increase in the output of equipment. aJ A predorinant portion of mechanical equipme:t is used by the manufacturing sector. Hence, one wouLd expect that the demand for mechanical equipaent would be rather closely connected withL industrial investlents. Admzitting that there wiLL be sharip year-to-year fluctuations, one wouild therefore expect indastrial investments to grev Iri Da wiLth mniufacturing output, thougkL not necessarily at exacily tbo same rate. Our estimtate for the growth of manuifacturing ositput 1963-1967 is of the order of 34 per cent. Since 1963 was a year cf low industrial investment (although detaiLed investment statistiCs are not availalAe in Brazil), on,s might expect industrial investnents to grow rather mtore than by 34. per cent. If we assume that industrial investments were at 90 per cent of their normuL level in 1963 (they were alnost certainly lower than that), the growth in investmkents corresponding to the growth in output would be 1.34/0.9 = about 50 per cent. Nevertheless, there may be somie factors causing investments to grow less rapidly than output. Thus, additional production ir e.g., automobile maxLufacturiLng, heavy electrical equipment, and steel should be obtainable at a lower investment per unit of output than the original investmients. For purposes of our projections, we have assumed tlht industrial investestments wiL1 grow by 43 per cent. If anything, we feel this estimate May be on the low sidet. In projecting imports, we have assumed a continuation of the 1959-1963 trend (GRF.PH A). Because 1963 was a year of abnorsally low equipment imports, thLis gives an increELse of about. 60 per cent; as ccmpared with a basic trend (i.e., if 1963 had been a normal year) of onl;y about 20 per cent. Dosostic production of mechanical equipment is obtaiined aLS the difference betweeer conauLption and imports. The proportion of consumption supplied, by d-stic equliLat. would decline slightly, from 61 per cent in 1963 to 57 jper centL in 1967. Even iso, doiestic produ:tion would need to grow by an average of 9 per cent ;eor yeEar. .J We have no detailed figures of the present structure of the demaind for electrical equipannt but the following represents our best guess. Prospective growth ratea 1963-1967 are also supplied: P_rjrtn of mnarkeLt AAsMd erowth 15163-67 Power indeistry 25 44 Industrial equipment 33 43 Consumer durables 1_ 100M 42 / The Chemical Inclustry in Latin tmerica (l964). -6- The growth rate for power indiutstry equipment corresponds to the planned rate of growth iZL geaLerating capacity between December 1962 and December 1966. The growth rate for industrial equipment is the same as the assimed rate oi growth in industrial investments. The increase for consumer durables might be slightly on tke highl side since there is reported to be some saturation of the mEarket in Sab Paulo and Rio. (Because of inflation, people are abnormally interested in inveslting in consumer durables). With respect to imports, we haLve assumed no chaage in the proportion of power equipment imported froms a broad. Fiwrthar import substitution is probahly both feasible anCd economically justified. The major obstacles are (a) tie up of foreign equipment purchases with foreign financing and (b) high prices fbr certain domesti.c materials and components (e.g., electric sheets, copper and aluminiLm semi-manufactures). On the other band, we hoLve assumed that imports of industrial, auto-, Eand cormmmications equipment will remn constant, and thb.t imports oi consumer items wiLll fall b one-third. Accorxling to best available information, the structure of cdemand and eupply for transport equipment in 1963 may bave been somewhat as follows: (million U.S. $ equivalent, assuming anecchange rate of 550): Motor vehicles 700 18 1 717 Other 230 102 - 312 Total 9 Q JJ_12D 1 1 2 Based uapon conversationEI with manufacturers, we have projected a 40 per cent increase in the demand for motor vebicles and tractors. We have assumed that the deaand for airplanes anad parts will be slightly lower in :1967 than in 1963. For the balance (mainly railway equiipment and ships) we have assumed an increase by one--third. These assumptions are consistent with thei mission's conclusions rLegarding the development of the transport sector. With respect to foreign trade, we haLve assumed that br 1967 imports of motor vehicles and parts would have fallen to negligible proportions, and would in fact be offset by an equal amount of exports. Imports of ships wouLld virtually disappear and be replaced by a slight eiport sarplus. On the other hands there would be no major changes in imports of railway equipment (e.g., diesel locomotives) or airplanes. GRAPH A RPA711 * (r-OWTH TRFKNDS IN IMPORTS OF (a) Mf1ECHANICAL EQUIPMENT AND (b) EQUIPMENT FOR MANUFACTURING IFAINJUU~ I 1(T MCU1'A CIIC I!At Ii~e mr ii | fe ' Io As ^ tn HcIA CM^1 TC'\ %IYIL.L.0I JIvd.S vJr lu. S. Dv OL.LIA L.w' J V . _L_ I %.P 400 [ - I 1 1 1 1 I I 400 0 TOTAL MECHANICAL EQUIPMENT - j | EQUIPMENT FOR MANUFACTURING INDUSTRY 350 ( | | ^ Z}350 _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ r _ _ 3 00 I 300f-- 1 -1300 25011 - > 19 - -250 200- -1200 150 j 150 100- J >40 DIFFERENCE: MECHANICAL EQUIPMENT MINUS MANUFACTURING EQUIPMENT 50t 1-- -150 (. 1 lI l l I 10 '58 '60 '61 '62 '63 '64 '65 '66 '67 iBRD- Economic Deportment 2528 A'1RIZX B INVIESTNT REQUIREMENTS IN MANUFACTTJRING 1965-1967 First 1.jprximation A. reasonably solid projection of investment requirements for manuf'acturirig industries can only be obtained through separate appraisals of grrQwth prospects and expansion and modernizauion req,)-reLThents in each major indus-rial sector. The fi=gures shown in the attached Table are a rough attempt at such a comparison0 MTeir validity is reduced, however, oy shorteo.min 9- ivn the aveilqh1ph infnormntAon arnei supnnnr-Hni nnalyses. Secondt Anryrt-,- m+ioncw For t.his re--son imabeoinrst+vocmr~ l ul-l total with a figure derived from more general considerations regarding the -elationso.iS,-p betvween heprojected ircrease 'n"',se^v outpuIt and thecn corresponding investment requirements. In the period 1955-1960, gross fixed domestic investments in BJDraz-l IlLUC UUatbvU DI DW 6I I cent) t,L0 IA) iU U.L of t ħDD grs atJ.ionaLJ. pr%ouzt. Let us assume that the average ratio 1964-1967 will be 16 per cent. Un- e_ r1. Ril;.L~ .LIE _r 110 E'U,1 1.U.L~ lo x __________r vli _[U.' __ _ilUU a UB_LL__ __e___. -. e_ _ fortunatl y, t,here are uuc recent figure;s for the unai-~ 0ħ mld.ħufactur-ng investments in the investment total. But if we make the hypothetical assuiription that, at the present stage of Brazil's econo--ic eve'lopi-menrt and in a normal year, manufacturing investmcnts would bear the same relationship to totai investments as manufacturing output to total output, we arrive at the folloiing capital-output ratio. 0.16 GDP x 0.28 = about 1.8 0.09 GDP x o.28 It is assumed that manufacturing has a 28 oer cent share both in gross domestic product and in gross fixed domestic investments. The numerator in the above statement expresses the investments in manufacturing in a normal year (0.16 being the relationship of gross fixe,d domestic investments to the gross domestic product). The denominator expresses the corresponding programmed increase in manufacturing output w1aich is very similar to the recent historical rate of growth. The above capital output ratio of 1.8 is similar to one which mnight be derived from the Table in Annex B (3.20/1.76 = 1.82). How realistic is our nreliminarv working hvyothesis that the shzare of manufacturing in gross fixed investment would be the same as its share in the gross national produet9? ecAuse of different nonditions, a compari- son with other countries can give no precise clue. GeneralLy, in industrially dpvep1nnedr1 rintrips- the shnre of mnnnf-t'iti-rinn in invest-ments i smal aler than its share in output. Nevertheless, there would seem to be tendency for the investment share to equal, or even to exceed, the output share in countries ulnderaorg ---itr' phase of industrialization, e.g., Israel or Taiwan. Brazil is clearly in a similar phase. Third Approx maticn According to the growth trend shown in Graph A, Anrnex A, a normal ~ve J. IJLJAjJ L U v UJ' L~Ii .SL_ -L.Ld.I. LI-ryLA L4L.U~ U UJ nY J I .L6'J levelI ofP equ-ipment im.ports by 4the B._azJilian.manufactuuing inC.1+ustr in 19673 would have been on the order of $250 million equivalent. The average import component of indust-ial investm,lents at the present moment would seem to beB on the order of 38 percent (judgment based mainly on data shown in the Industry Chapudr of tlllan, p. X1X-4)e U- t'nis basis, a normnal level oi ' nanufactiring investments for 1963 would have been on the ordler of $650 million eq-uiV-alent w'hich compares with our projected average annual invrestmenLts 1964-67 of $800 million equivalent. The difference is not easy to explain. Two possibilities come to mrinad In the first place, the above percentage import component of industrial investments was derived from "already quantified investments"; it seems likely that the import component would be lower in some of the traditional industries (food, leather. lumber, etc.) for which no data were supplied. But even in the ;"quantified investments" the percentage import component might turn out to be lower than estimated in the Plan. This would arise from the well-known tendency in most developing ccuntries to underestimate the local currency component of manufacturing investments. Conclusion Further study is needed of the probable investment requirements in the Brazilian manufacturing sector, 1965-1967. We have used three dlifferent approaches to arrive at a guess of these requirements, and obtained similar results by all three methods. This by no means proves the reasonable- ress of our figures. Thus, there is little empirical verification of our working hypothesis that manufacturing investments would represent 38 percent of total investments in Brazil in a normal year. Nevertheless, we have stated our assumptions. As more refined data become available, these can be used to modify our conclusions. ANNEX B^ TABLE FIIXED INVESTIENT REQUIREIENTS IN MANUFACTURING, -L964.-67 (c3illion U S. fequ-ivaieniT 17a :Investment Requirements Gross CL,:Lt Increase in Ou-tput 1963-67 1964--967 Ra tio Four-year Annual 96 19 Gross_ Net Net/Gross Total Average Consumer goods F'ood, beverages & tobacco 2.'55 3.(8 0,53 0.175 .33 .36 *09 2/ Textiles 1.,L9 1.45 0O26, 0.115 4.18 .045 / Clotbing and shoes 0.34 O044 O10 0.05 .148 0o4 .01 L,/ Purni. ture O.-L9 0.25 0.06 0o03 .54 .02 .005 h./ Pharmaceulticals 0.20 0.25 O.05 0.03 .58 (.o4 (.ol 5/ S;oap and Perfumes o.-L5 0.18 0.(3 0.015 .42 ( Printing & publishing 0.22 0.28 0,06 0.035 .60 .08 .02 6/ 1Wj34 1.02D9 °L .1i o72 .18 Intermedia1e 2roducts Timber & wood products 0.25 0.30 0.05 0.03 ,56 .02 .0 C05 7/ Plulp and paper 0.35 0o53 0.18 C).08 .46 .14 .035 13/ Leather and products 0.10 0,12 0.02 0.01 .46 .02 .005 '7/ Rubber products 0.26 0,35 0 . 09, c,045, .47 .08 .0C2 9/ Plastic products 0.08 0.11 0-.03 C).015 .58 .02 .005 7/ Oil refining 0.33 0.49 0o.16 0.07 .45 ,.16 .04 'LO/ Chemicals 0.80 1.32 0.52 0).23 .,4 ,,56 .L4 'Ll/ ,Sundry products 0.12 0.17 0.05 0.03 .,60 _ 4 .01 7/ 2,29 3,3 1,10 (0.51 .46 1.04 .26 Capi,taL_goods Non-meta nic minerals 0.49 o.68 0.19 0.125 .66 ,20 .05 12/ Steel 0,12 0.56 0.24 0 17 .70 _ 18 13/ Non-ferrous metals 0,05 0.08 0.03 0.02 ,67 0.6 .015 T1/ Metal products 0.91 1,28 0.37 0.145 ,,39 ,.14 035 CL7/ Mechanical equipment 0.36 0.48 0.12 c).o65 .,55 .10 .025 16/ Elect.& Comm. equipment 0,54 0,79 0.25 0).11, c,46 .10 .025 L7/ Transport equipment 0,93 1038 0.45 o)16 J35 12 c03 Y/ 3,6o 5.25 1.65 C).8o .-48 1,,44 .,36 GRAND TOTAL 10.j lj,5 38h4 1.76 ,46 ).20 .80 1/ Ideally, investment requirements in manufacturing 1964-1967 should be based upon the increase in output 1964-1968, Tnis is on the assumption of a minimum one-year average time lag between a given investment and the increase in annual output resulting from that investment. We have re- lated them to the growth in output 1963-1967 on the assumption that this would not differ significantly from the 1964-1968 total. 2 Based upon scattered data for a few other developing countries, we have adopted a capital-output ratio of 2.0, This may be appropriate for e.g., raw sugar factories and grain milling; it is probably higher than the rates to be expected for meat packing. 3/ An intensive ECLA study (New York, 1963) suggested a five-year re- equipment program for the Brazilian textile industrv at an estimate(d cost of U.S. $226 million eauivalent, of which $57.5 million equivalent for the North-East. This includes only machinery and equipment, an(i makes no all.owance for new buildings, In terms of total physical area, the existing factory space is no doubt adequate. Assuming a four-year total of $180 million equivalent and relating this to the expected increase in net output 1964-1967 of $115 million equivalent, we derive a capital-output ratio of 1;52. some fii,'res for inv-,estment renuirements for a new mill are available for a recent project. This involves a total investment; of Cr. $ 4 billion for an ann.ual gross output value (2 shifts at full canacitv) of Cr. $ 1.7 billion and a net output of Cr.$1.2 billion, yielding a capital-output ratio of 3.3. Th4.e ratio of net output to grons nitput would be abollt 0.7 which compares with an average (not necesaarily fuUly comparable) according to Vth 1960 census of 0. '38 Th_.e c is clin-d in nqrt. by the fact that the above estimate was for a mill producing high quality fabrics as cor.pared' -Witl thle rMtior"'1 a-Verage re ectir.g a releatively- nnnrnp. grade of cloth. A third indication of investment requirements may be had from BrLazir imports 4of txtile 4aX. rery ninnhrati p' a o n milfI ion- equivalent per year in 1961-63 (textile machinery proper, excluding - A. . J... .IUL41 ~LLJVL 1,,. U.Y 4....,4 ~ . -.A .,+.r -.4 lr 0_,cv1 -__ o"er eI-2 by ALh e textLle i Lj UsUry Aike% boilers, LJntpi-nal transport equipment, etc.). Assuming the impor'ts reprosented about 45 per ceIit of' tov;a4.i UkV sUIWU, U.L0WU would corepnd . a recent investment rate in the textile industry of about $45 million equiv- alent per year, i.e., the saiRe figure as Lel anUlual re-equpm.ent requirements according to the ECLA study. On balance, it is our feeling that the investment requirements shown in the ECLA study are a minimu V corresponUding to a per-fect reorganization and modernization program, and that the actual require- ments will prove higher. Since we are in no position to rake an independent estimate, we have retained the ECLA estimates, subject to our stated reservations. 4/ ICA "Industry Fact Sheets" (based upon United States conditions and c,ne sLift oVLY) show a capia L -output raiVo for ne-w slmall clothiLLng fact-ori.es of about 0.5. In Brazil, there would probably be shift work. On the other hand, the h.igher cost of the machinery ana the lower eiielency in the utilization of the equipment would tend to increase the capital.-output ratio. We have assumed a capital-output ratio of 0.75. 5/ A capital-output ratio of 0.75 was assumed. 6/ A capital-output ratio of 2.0 was assumed. 7/ The investment requirements for this sector are small in relation to the total requirements for manufacturing induistry. Yoreover, because of the desirability of rounding to the nearest $10 million equivalent., the annual average investment would correspond to a range of capital- output ratios. Hence, only a rough guess has been made to fill out the table. 8/ Based upon some very general ideas about the type of expansion required (one new newsprint mill, one integrated kraft paper mill, two rayon pulp mills, etc.) we have arrived at estimated investments on the order of $138 million for 300,000 t. additional paper and 245,O00 t..; additional pulp, .e. about $445 per ton paper integrated with pulp. 9/ This is a capital-intensive industry, and the capital-output ratio for a new medium-sized plant might be atleast 2.5. (See ICA Industry Fact Sheet). In fact, actual data from another plant suggest a very much higher figure. There are three major producers of tires in Brazil, and we suspect that they have enough napacitv to meet a 30 ner cent increase in demand without major investments. Nevertheless, in 1967, some investments may be requiired to meet the demand i n the late 1960ts. 'Je have assumed a capital-output ratio of 1.75. 10/ We have projected an increase in refinery output 1963-1967 by about 7.5 million cu.m. (about li5 ner c.ent.) representing a value at imDort parity of about $210 million equivalent. This is on the assumption of an avernae prioe of-r19 n$2 er cu*. m= fnr refJined produ_ts. (The 1967 market value shown in the table corresponds to a price of only a little above $~,20 perP CU ln T,h- S iS proh1-1y btcause +tis Z 1iiT ni projetedr f-rom 1959 when petroleum products in Brazil were valued substantially below Thek mair ; poto of +)i 7. miion"ii nn1;4+ - I .m. fnitiia oupu Y.r; 1 come from two large new refineries with capacities of about 2.5 million cu.. p e r ye a r. Data fJO a ISor -theU cost of -Su. hAe sS rSe f ir.Me JeA_ S i. S not av o1taib le., A typical figure for recently built refineries of that size is $20 per _. .m n_ 4 - A - _ 4 4- _ _ : A A L Je, 4, 1- 7 .) -4 .L.4J-_ CU .M. W. ,4 necessitate investments on the order of $150 million equivalent. In t us earlier estl,matles (no longr up-tUUUd J rfuteu)Io P - a uw LIctt.l investments in refining 1964-66 of approximately $100 mi'llion equivalent (cor-responding to $133 LIcħiħħULħ on uq-uiv on a fou--year basis. U cting more precise studies, we have adopted the higher benchmark figure of $150 million equivalent. 12/ Total projected invettments 1964-1970, according to the economic prcgr.-m exchange equivalent of about $87 million. This would include about $ I-. L4llion for. _L1 iJ. iI ch ca1lU.;J IILL.L.LħUon for ħLJ ertsLgLiħV1 0, CUIU. $113 million for various petrochemicals. Govet'ment investments pre- domr,nate in pe tEocherMij'als (say $75 mUllion equivalenlt) and 1ertilizers ($50 million equivalent). In inorganic chemicals, they represent only auou one-fifth of tne totai (ksoda ash). Petrobras alone,, according to preliminary estimates, would invest altogetner $35 million equivalent in 1965 and 1966 in petrochemical and fertilizer plants. According to an estimate prepared by the Sa6 Paulo Chemical Products Industry Association, expenditures of $199 million equivalent would yield an estilnated gross output value of $193 million. Assuming an average ratio of net output to gross output of 0.45., this would correspond to a capital-output ratio of 2.3. This checks roughly with an estimate based upon a selection of new projects (alkalis, fertilizers, polyethylene, PVC., artificial fibers, carbon black, and calcium carbide) which shor an average capital-output ratio (assuming 85 per cent capacity utillzation) of about 2.4. We have adopted the latter ratio for our projections. 12/ The bulk of the cement output in Brazil comes from cement plants with outputs exceeding 500 t/d. An industry expert has supplied the following approximate cost estimate for such plants: New plants: $45 per t/y Expansion of existing plants: $30 per t/y A recent project for a new plant indicates a substantially lower cost of $33 per t/y, exc'luding working capital. Apart from additional capacity. investments are also needed for conversions from the wet to the dry process. All but two plants are said to use the Wet process which has the drawback of 50 per cent higher fuel consumption. The cost of such conversion is said to be about $10 per tI/v. According to our rough estimates. Brazil is likely to need 2 million additional tons of cement by 1967 as compared with about 5 mill- ion tons arctunllv usedi in 196A nn th.is hasis. investment requirmem~nts may be estlmatd.2s follow~s: .M ,mi l 1 lion Exnansoin : m I0 rW1 tons + at $30: 45 New. : 0.5 million tons at 45: 22.5 0 nverS ' On: 1. r,ILl . ion tons at 10: MO t_ _,I rf A. - We have no details for other branches of the non-metallic minerals group. * Aother important L ite, ho-wever-, is fLat glass. For a neRw 100 t/d flat glass plant operating at 85 per cent of capacity, one igtgn figure with a capitaL output ratio oI 1.6. In translating tnese indications into an overall investment cost estimate for the non-metallic minerals group, one realizes that the 1963 gross value of output for this group is inexplicably high. Assuming an average price for cement of $15GO0 equivalent f.o.b. pl.nt, the total value of the 1963 cement production would be abouit $80 million equivalent. Assuming that cement represents 40 per cent; of the gross output of the non-rretallic mineral products industries, the total gross outout of this group would be $200 million equivalent. Yet, our table shows gross output value of no less than $490 mil'lon ecr valeint. Lacking the opportunity to examirne this difference, we have made the fol'Lowing investment cost estimate for the group: Cement 80 Other products (l.6x120x0.67) 128 208 L/ According to Voliime VI, the total investment requirements for steel 1964- 1967 would be about $700 million equivalent. The approximate investmlent cost per ton additional capacity created in 1965-1969 would be $169S corresponding to a capital--output ratio of say, 1.6. The bulk ct7 the now capacity, however, waLld not actually be commissioned until 1968-4970 which explains why the rela+tonship between invest,nents and increase ir output is much more unfavorable in 1964-1967. lV/ We have accepted the Plan estimate of $56.5 million equivalent for 196L-67. J$/ There are a large number of producers and variety of products in this industry, and it is difficult to know what would be a representative capital-output ratio. Cur best guess is a ratio of 1.0 . 16/ Same remark as under 1. In this case. we have assumed a canital-outDut ratio of 1.5. It seems that the 1963 output of this sector may be undervalued biit we have aAd no nnnnortui ty to nmrtsue this question 17/ r1nn-ayl Elec-t-ricr has n 7s comnleted +.he largest hpnir e1etri l equipment plant in South America. Brown Boveri has a large plant in the same field. T.i.e is q i H Ysoc4p rstra cap4 +.w here wThinh may 'h absorbed when the Electrobras program goes into high gear. Major new addition- A wi hardly be needed. There is also substanni+A excess capacity in the electrical appliances sector and in the plants making *nl r-ro.-.m.r~ .,n+awion s, ,-ui.nrm.ent V.r 4-his basis the projectedn4.a eavvncn n in outplt may be achieved at a relatively low investment cost. Never- 4theless, by 'I 67) Jir.est- n4ts may .U.a- to- Ibe --e for addtina +;---I ULitut ioL7n in t Cle ) 01960's Uand LIperhaAp f Loe J.p As JTo ... re V substitution in the late 1960's and perhaps for exports. To reflect the favorable unused capacity potential. we have assumed a marginal capital-output ratio of only 0.85. 18/ There is a substantial capacity reserve in the Transport equipment industries as indicated by the followin2 fieures: Motor vehicles: 196i output: 200000 vehicles anacri M PS! Assepm.hlv 2 O-onn vehi; l es Industries 375O000 vehicles but mode-rrizzatior. r.eeded va+^-s: ~~~19 o(. qtput:if~ 12,000f sl,-t ===LL: 2~~~~~~'-4,0v "'1. +4..Ut. QubCajr,d41L,ħ..1l .Increase iLL UU U sho LLUULU beV oJbtaUiIeCdL atd a re LaJtie.,y modest investment, and we have assumed a capital-output ratio of 0.7';. ANNTY G. Main Areas of Potential Import Substitution in Manufacturing P-nnndvn+4 nf ;4 or w,nn,. vi flea C-os vrlIofln caluet consumption (%) consumption 1967 import substitution +w^ ~~~~~~~~'I 90.2 , L'3 ff .A (EjS5 )14 ( -, 1 , - llionb - - >, f- 11 4 )- TC 4.7 / 1 LU.LL k\1LLJW_L_L11 U 'AP4 U~~L 4UI.LV-.VIVL1 J I Th UYCISI1%JC u ,AO Drr -V .L lp nd papsrJ~ 1.2~. 3D6 6L 50 3 Oil refining 17.5 9.2 8.3 540 45 m -1 ~~~,* ', ,.* I- L It, r'nn L ^Le-er,ca'ls 17 I.7f 1. L4. 4 ,Ouv UU Non-ferrous metals 58.5 50.0 8.5 160 13.5 riechalical euq-uip 391p >..04.o -5. 040 -42 Electrical equip. 12.9 10.2 2.7 880 24 Transport equip, 11.4 5.5) 15,9 1,45 86 l/ Although the degree of import substitution would be less in 1967 than i:n 1963, there would still be healthy growth in domestic production (about eight percent per year).