INDUSTRY AND ENERGY DEPARTMENT WORKING PAPER INDUSTRY SERIES PAPER No. 32 National Systems Supporting Technical Advance in, Industry: The BrazHi'an Experience June 199b . ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ . .. .. 44 4 ~ . ~ ~ .. _ ,.~~~~~~~~. I~~~~~ 3un 1990 - = = ___ ,_ __,_ _ _____ 7 , _ _ _ -___ .. _,__ . .._ L. _ A,._. ,.___-, -Ce- Th Wol Ban Inusr and Enrg Departmnt,PR {ir~~~~~~~~~~~~~~~~~~~~~~~~~~~4 - 4 4 - 4= 4 4 -- s .- 4 . 44W ~. o 4. G fr ' j W = - ~ --.- 4t _ '-' =:j-4 . ,_ *;..g * i °. \ -- ''- -44 -44l ° 4, f44 . c c " 440~ =4~ S.4~. * 444 * - =-& _44 .4*4- 444 44=' 4. .5-4444 4 W.X44 4.44~~4 -~~. ~~-.4444~~ . X~~~~~~~~~~~4-.4' ,,4.~ 4 44..44,, __ ~ ~~~~~~~~~~~~~~~~~~~~~~~~~~~ 444444 4i,4444 =4~44 ~ 44~4. -- _ 4 ' The 444.. World44~4 Pank Inusy ndEn4g 4pane,PR NATIONAL SYSTEMS SUPPORTING TECHNICAL ADVANCE IN INDUSTRY: THE BRAZILIAN EXPERIENCE Carl J. Dahlman- and alaudio R. Frihtak June 1990 Industry Developmeni Division Industry and EnXrg Dpartmet Polico, Research and External Affairs TABLE OF CONTENTS Page No. EXECUTVE SUMaMARY ............. ,...................................... i -ii L INTRODUCFION AND OVERVIEW .......................... * I IL BRAZIL'S COMPETITIVE POSIMTON AND THE TECHNOLOGICAL EFFORTS OF BRAZLIN FIRMS ........................... 3 A BrazisLaggng nternational Cnpett.ven.ss. 3 B. he Scope and Intensity of R&D Activities AnmongIndustrialIFs ............ 4 m. TECHNOLOGY POLICIES .... 9 A. Anms-Lenth TechnologyTransfer. 9 B. Imports of Capital Goods .....12......... 12 C. Foreign Iestment. .13 D. Fbin a Incentives for Lod Tehnology Devlopment .14 TV. INSTITUTIONAL NETWORK SUPPORTING TECHNOLOGY DEVELA(*MENT- 15 A. Historic Badground and Cufrent Structure. 15 B. PublcR&D lnttutes. .16 a Public and Agtao R&D 3Ependitures .18 D. Human Capita Formation .20 V. CONCLUSIONS .u ANNEX BlBIBOGRAPHY The World Bank does not acsept resposibility for the vles ewpressed herein which are those of the authors and should not be attributed to the World Bank and its affiliate organizations. The finding, interpretations and conclusions do not represent offical polcy of the Bank. The authors would like to thank Basanta Chaudhuri for research assitnce, Stephanie Genrd for editorial assistance, Anna Marie Maranon and Wilson Peiris for word prxcing suppor, and Valerie Chlsholm for secretarial services. EXECUTIVE SUMMARY J. During the late 1960s and most of the 197(s, Brazi was one of the fastet gowing ecmnomies in the world. Yet in the past decade, output growth of Brazilian manufacturing industy has slowed sificantly. In contrast with East Asian and sevral other newly Industrializing countries, Brazil had low average GDP and manufcXturng growth rates In the 198(-2.9% and 22% p.a, respectively. In addition, Brazil's share of manufctured exports in manufactured output-11.8% in 1987-Is relatively small, even though manufactured export growth has outstripped manufactured output growth since 1965. IL Many reasons explain Brazls inability to sustain growth. This paper focuses on technological factors behind Brazil's faltering competitie position-and on the role of Brazil's innovation system he M m _ nat sstem Is here defined as the network of agents and set of policies and institutions that affect the Introduction of new technology to the ecnomy, whether or not It Is new by world standards. Since in the vast majority of developing countries, technology is imported, the innovation system is defined vey broadly to include policies toward arms-length technology transfer, intellectual property rghts, imponation of capital goods, and direct foreign lnwestmenL lbe innovation system also comprises the network of public and private institutions and agents supporting or undertaking scientific and technological actvities, including research and development, diffusion of technology, and creation of tehcl human capitaL ;l; This paper's major contention is that the gap between the average Brazilian firms performance and the price-performance frontier has widened due to a combination of e limited technological Involvement by domestic producers; - regulatory and poUcy restrictions on both embodied and disembodied forms of technology imports and * weak institutiona suppor tO industrial firms. In addition, major gaps in the educational system, particularly low enrollment levels in secondawy school and in science and engineering, undermine the quality and quantity of the technical labor force and adversely affect the acquisition, use, and development of technology. iv. This paper suggests that much of the underlying motivation for the Brazilian Government's technology policies was not the improvement of the countrYs technical base but the more short-sighted consideration of saing or earning foreign xc hange. At the same time, the government's technology polices did appear to Improve the bargaining position of local fims in negotiating arms-length technology transfer deals. In addition, they promoted entry of domestic fims in areas that normally would be precluded through patent protection or the exercise of overwhelming market power (pharmaceuticals and electronics), while stimulating the creation of a fairly sophistcated capital goods sector. Moreover, such policies effecti* stered multnational dfms that traditionally were domestic market-oriented to become exporters (as In the highly suessful case of vehicles)+ This thereby forced the techological upgrading of their domestic operations. Fnaly the poLcies created for the first tme a financing mechanism for the technological needs of Brazilian industrial and engineering consulting finns. v. What this paper implicitly argues, however, is that these govement policies were caried too far and became outdated. Restrictive armhsenh technology transfer poliies, for example, assumed the presence of relatively unsophisticated domestic producers and the considerable eagerness of foreign suppLiers to market their latest technological wars Nearly two decades later, such assumptions have much less justification. Simarly, capacity creation in sophisticated industrial segments, such as capital goods and eletronics, was regarded as providing double benefits for economic developmenIt iaLlwed for import substitution in areas of beay foreign exchange outflow while creatng capabilities in strategic segments for the formation of skills and diffusion of knowledge. But acrs-the-board import substitution led to exmsive ' I ,j o I' t ;V - s- : 11 iil E - ii- ,} 'I *~1' ifi ii;s ' t ~ ~ ~~~~~~~~~~~~~~~~~I 0 I' . - I ._ . [!, , , 1S _ - . . g~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~-. L INThODUCTION AND OVERVIEW 1.01 Brai has the ninth largest GDP in the world, the stha largt population, and the 0fh largest area. Its per capita Income In 1987 was USS2,241, slightly above Mexim but below Argentia, South Korea, and Taiwa. The county is well endowed witb natural resources, has a productive agriculture sector, and a mature and diversifed industrial base. Manufacturing value-dded in 1987 was US$79 billion, th ninth largest in the world, about 15% larger than that of Canada, two times that of Mxico or Korea, and almost three times that of Austalia 1.02 During the late 196( and most of the 1970s Brazl was one of the bstest growing economies In the world. Yet in the past decade, output growth of Brazilian manufacturing industy has slowed significantly ( le 1.1). In contrast with the East Asian and several other newly industizing countries, Braz's average GDP and manufauring growth were low, 2.9% and 2% pa., respectively. In additin, Brazis share of mufactured exports in manufactured output, 11.8% In 1987 is small, even though manufacd export growth has outstipped manufactured output growth since 1965. T,ble 1.1: SELECTED ECONOMiC INDiCATORS -3bare of Grotb of Growl lf MeOnf GDP GDP Mn ouwt Manuf Epodus Esps in PerCapita Out 1911 196S.80 19g"o lC5s-0 198 1o.0 1980-88 1987 (USS) (%) (%) (%) (%) (%) (S) (%) Brat 2,241.07 8a8 9 98 2.2 22.1 6.2 118 Indonesia 476.09 ao 5.1 120 13.1 13.3 35A 44.8 Japt 23,195.02 65 329 6.7 11.9 5.3 32.4 Kam 4,081 9A 9.9 187 115 31.2 14.7 103.1 Mano 2,111.11 6;5 OS 7.4 02 9.4 21.1 27.1 Thaiand 1,06330 7.2 6.0 112 6.8 213 193 53.3 Thkew 1,19628 63 53 7.5 7.9 26.4 23.3 42.6 Sources: World Dcn:bp Report, 1990 (in prm) World Tab1s 19-90 EdI lutenonal FIabmal SatistIs, De. 1989. 103 Seval caues explain the countiy's inabilty to sustain growth Brazil faced a 38% decine in its terms-of-trade In 1977-8L nteational interest rates rose rapidly. Moreover, the flow of capital was reversed, from plus to minus 4% to 5% between the mld-1970s to the mid-198(L These shocks led to severe macroetjnomkc Instability, two major recess, and a strong loss of confidence in the longer-tenm prospects of the economy. Ultimately, the shocks brought about a significant contraction in the rate of investment. The ratio of investment to GDP-after reamcng nearly 25% in 1975-fell to 15 to 16% in 1988. 1.04 Brazil's falng pfmane relative to other industralizing counties has been due not only to 8 OntrtOn In the rate of investment but also to inefficient resource aloction and use. BegIning in the 195(s, producers were attracted to invest I new industrial segments by a combination of trade barriers, entry egulations, and fiscal, credit and other incentivs Such policies and.intitun arrangements also shielded producers from domestic and ntenational oDmpetition. As the industrial sector matured, these protective barres solidified, and firms becme ineag secure in thek market positn Protection from competition made fims less resilient and management less responsive to the rapid shifts in the international economy: an acceleratd rate of innovation, Intense temhnological and commercial -2- ialr, accompaied by Incre prtetonism in developed couatrl~&. At the same time that Brazirs tchnob poliies geally Inibited ntenadon telog transer, the technological capabilities of most local producers were enhanced oy marginall by the support Infraructwe 1.0 1hs paper focuses on technological factors behind Brls falteg compedtive poddon and the role of Bil's nnovation system he innoa yste Is here defined as the network of agents, and set of policies, and Wntutions that affet the introducdon of technology that is new to the economy. Since in the vast majority of develDoping countres technology imported, the novation system is defned very broadly to include polices towad arms-length tenology transfer, intellectual propery rights, impora- tion of capital goods, and direct fogn im ent. Te innovation system also compries the network of public and pdvate institutions and agents supporting or underting scientific and technoylgical actities, including research and development, dffusion and creation of technical human capitaL 1.06 Section U brefly d _cuss Bzil's competite status In world marwks and suggest that its faltering position is related to the limhed domestic technological effr of industrial produces Section m shows that the modest aure of these efforts bave not been offset by subsntial technolg imports, whkb have been resicted by regulatoiy acostraidts on arms-length technolo tansfer, direct foreigp investment, and capital goods Imports. Although some of these restrictions, in combinatio with govep ment financa support Induced frms to eae in technolgcal activities, relatively few producers were able to approach the international best-prctice ¶rontier in the 18L 1.07 In view of the modest technologia actwities unetaken by local private producers and the Ihmited inkows of foreign techokl, deveomnt efforts fall on the shouldets of the public sector. Seton IV notes that 80% to 90% of R&D pentures have been undertaken by the Government. Unforunately, dispersion of these public ctr R&D effort, and weak linkages between public R&D institutes and the producdve sector, have meant that the fow of useful techologl serces has not been commensurate with the volume of public setor reoroes alocated to technological activities. Pnaly, in human resoure formation-an area in which the public sector role is critical-weak investment dedsions have undermined the creation of technologil capabilties for the ftu seection V presets the papees conclusions. -3- - II. BRA'S COMPETITIVE POSION AND THE TECHNOLOGICAL EFFORTS OF BRAZILIAN FIRMS A. Brazil' LIg g luteratlonl Competitvenes 2.01 Brazil's competite position in the 1970h Improved siganficantly. Between 1970 anid 1980 Braz expanded Its world mar*et share in 83 of 93 IndustrI segments (3-dgt SJTC categories) with positie manufactured exports (Annex Table 1)Y Moreover, Brazil had an impressv record of markt divesification and product diffrentiation in the 197I Between 1971 and 1980, in most machinery and transport equipment segment-and in shoes, textles and fbers-the number of countries importng from Brazil inceased, and the number of producus eported in each goods categoy expanded. In many categories- such as office machey, boats and ships, aircraft and textiles-the eport cmposition shifted to higher vdue-added productsY 2.02 Slune 1980, the competitive dymi of the B_azlan economy have dhmged substtially, with constant or dedining market shaes In 46 of the 93 relekat SfTC segments, including technologiay sophisticated areas such as eic machinery and Instruments (Annex Table 1). The only significant exeptions were aircraft, telecommunications equipment, and electric distbution equipmenL Moreover, In cntrast with the 197(b, when 20 of 26 segments in which Bra had achievd a rvl comparative advantage-Le., nomalid market shares greater than one-increased their shares, only nine segments did so in the ed 198-7F Virually al non-trational and non-natul resource-baed manufictured products had their gsas rever Surprisingl, this revera occured in nontaditional products sucit as texles, garments, travel goods and handba, in whki Brail ould have expected to maintin its competitive position. 2.03 Brazi is now saddled with many fntemationally unompitie segments, their number having grown signifiantly since the eary 1980s. A study of 67 manufcturIg subsectors reeas that from 1983 to 1985, 45% were nternonally ompitnve (that Is, the ratio of domestic to international prices was lss than one) and 7% were margly competitiv (domestic-inteational price ratios between 1 and 1I). A similr comparison shows that In 1988 only 15% of the subsectors were ceary compettive, while In 12% Market share is defined as X where x is the counts exports of sood I and Xw is the wrld's exports of good i. See J. Tavares de Araujo, 'Mudanca Tecnogsca e Competitividade das Exportacies Brassle de Manufatr " IEWEAIIUFRJ, Texto paM Disao na 1982, Table Il mThe norized market share index is defined as (XIUX)(X1wIXw), where Xi is the countrys exports of good i, X is total ontry exports, Xiw is the world exports of good and Xw is total world exports; hence, the share of a pardcular good in the countrys exmrts is nornalized by the shae of world exports of that good in total world expors. The Index is quite sensitive to the competitive position of individual subsects in the national economy, and hages in the Index are indicative of shifts with respect to the world pries. 4 The only major exception is explosive o cnial products, wbich continued to increase thdr normalizd shares in the 19W( and showed in 1987 the highest "revealed" compuatie advantage. This category is belved to contain many arms expors -4 - of the casesmpetitiven was marginal. bes studies sgest that the poportion of fuly or marginly competitie segments fell by narly half (from 52% to 27%) between 1983/85 ad 1988. 2.04 Additional evidence of a deteriorating wompetitie position comes frm a 1989 survey of the views of 550 major Industrial produes Altho most Industrialists judged the Brails Industrial sector to be more efficient than at the beining of the decade, nearly 65% of those Interviewed also pereived the extent of modertion as "Insufficient' to keep up with the Interational price performance fontier, wherea only 3.8% thought that It was -quite adequate.# B. The Scope and Intensity of R&D Activities Among Industria Finmn 2.05 The lbility of Brazilian fims to keep up wit the international frontier is a reflection of a highl heognus industial structure, relatively few competite producers and limited technological effol Bra's lae and diversified industrial base-tbe result of a long process of Import subsUtution-is insufficiently 'deep." Relatively few firms are in a position to consolidate and expand intenational market share on the basis of substantial gains in productivity, continuous improvemets in product quality and reliability, or devlopment of new designs Most produces have based their market position more on extensve exploitation of natural resources and reliance on low-wae labor than on the qualit and productivity of labor itself and the Introduction of new or better producs. 2.06 Evidence strongl suggest that R&D, as a formal actvity, is undertk by relatiely few firms. In the perod 1975-79, only 1,500 industral fim or 07% of a universe of 210,000 legally established producers, conducted formal R&DF Even within this universe of 1,450 priate fims and 50 in the public sector, R&D expenditures are highly concentrated. Data from a 1983 subsmple of 1,118 enterprises show that public sector fims accounted for 62.6% of budgete R&D outss, with edt producers resposible for See J. Tavares de Araujo, L Haguenauer and J. Bosco Maciado, 'Proteao, Competitvdade e Desempenho Exportor da Ewnomia Brasikira as Anos S0W' Revista Brasileira de Comerclo Eaiori! ano V, no26, Novembro/Dezembro 1989, and Honorto Kume, "A Poltlca Tarifda Brasi- elra no Periodo 1980-85: Avalia;o e Reforma,' mjjeo 1988. IL Kume's 1988 price comparisons were for 88 subsecton On the basis of a representative sample of 715 products (8"igit BTN cas- sification) of the IPEA-FUNCEX data base. Intemational price comparisons tend to present well knwn problems of choice of an appropriate exchange rate, and comparability of products differen- tiated by quality and performane. Yet such price data can be suggestive of broad trends in a countrys competitive position in world markets. Al See Confedera9no Nacdonal da Industria, Competitividade e Estrateaga Industial: a Visso de Uderes Industriais Brmileitos Rio, 1990. 7J In 1985 the top 254 eporxtrs of maufcured goods were esponsible for 70% to 80% of total manufcued expors Tbese producers comprise approximately 6% of all medium and WVge firms with published balance sheets and less than 3% of al estbimn with over 100 employees. In addition, a number of smaller fims export through tradig houses; these fims are concentrated in shoes, cast iron, steel, and autopa Overall, however, exports are undertak by that nawrow set of large producers in meta-mecbanlcs, chemcals and petroemcals, and traditional segments. A( See J. Nunes, 'As Forcas niboas de P&D na Industria Brasleira, Revista Brasileira de T efl~IQgI~ Vol. 16, nZ 1985 Thds information coma fom fltms' Income tax returns. or half of the total expenditures Only about 25 private industrial groups were responsble for an additonal 17.4% of ependitures (Fable 2.1). Tale 1: R&D OUILAYS IN BRAUAN INDUSTRY, 193 W Nature of Ftirm P of Total Outlays (iS) Pul Seeo Entrpes 626 Pivate Groups 17.4 Prae Flm Faeud on Indidual Maret Source: I 11aully, Emprs Nacoalud-Panmam do Setow Elnvearlal em 1983, * Bmsilelra do ensai. vol.15, n3, 1984. W( 1,118 firms, of whkh 43 wer pWbc e _tm and 1,075 prvate producers. Firm I the .ectdc machinay and madeto-odr capital goods ubsectm were not included In Ib study. As a result, the percentage for public enterprise I biased upwards. 2.07 Me smallne of ndusty-related R&D In Brazil also is reflected In the distrution of researchers (Table 2.2). In 1986 there were 52,863 re sarcher out of 3.5 milLon college graduates, a relatvely low propoinu the ratio of researchers to population was 4 to 10,O-whereft in developed countries the number is about 40 to 10,000 Inhabitants. It is striking that 91.6% of researchers worked in gverm nt, the majority in publc universities (62.0%). followed by specialized S&T Insttutions (20%), state enterpries (3.4%) and other agendes (6.1%6). abk22 DISMIRMON OF R ARWHRS ACCORD TOD PLACE OF ACtiVITY 1986 I'-titution --TOW Doctor'ate No % No. % 1 Univerties 36112 683 9,952 6.6 (Public U ities) 3275 62 ; a. a Speaize ST Instito 10,856 205 1,38 11.8 Go _enmet Agencies 3,203 6.1 82 0.7 StateEntrpise 1,811 3.4 38 0.3 Prvate situos 727 14 60 0.5 Prvate Ffrms 295 0.6 15 0.1 Other 111 02 - ToWa 52,863 100.0 11,492 100.0 Sw=ron G. Martns and R. Quefro, -0 Perta do Peaqulsador ner," Revist Bmsilefa dsiqj~a. VoL 18, No. 6 Septmber 1987. y2 ' These are Petrobras (oi4 and petochemicals), Telebw (telecommunications), Electrobras (electricity), Nuclebras (nuclear energy), Slderbras (steel), CVRD (minerals), Embraer (airplanes), and Cobra (ltfrmatics). -6- 2.08 OJ8.4% of s4rer s wem I u*pnjobs. Mo of themk to private unvesidea (635%). whIle rtiely few whel in p t rea InstudtIonm (0O%) or other agenis (0%) PAr ly not w s the mInut shar of resc In private firms (O.6%) Even lnluding state enterpses, the pentwage of R&D pedn emplyed by the productive set s moest, 4 n for less than 4% of the to 2.09 Information compled (a the mid-19o ultec ta R&D continues to be underatk by a narro st of firm; a core of 366 fims can be rega as -R&D acdWe (able 23). As exted fom Intemationa ueience_ coo rlnteusv segmen sch as electroi, vehices and the chtWik t a heutl chaster ontain the largt share of ftms active in R&D. Converly, segments sch as food and beverages te , grments and ootwea, leather anid wood product have rehtve few R&Dlengagod flrmswith capital goods and meWt Wication In bei weu T8lt SCIORAL I)STRODTION OF WORT.Or4 NE11D AND R&D ENOAGED BRALIAN FRMS R&D eave %at IndntSW &caa R&D ad[" R&D # Zt aetwe ap itm s o a Tota bf am I s5td nbo ma. 24 17 7 S17 46 U Metal FabdoW 9 3 6 m Caa Goods 9 7 42 419 SI iIsI IV VeNd.' 1 12 5 70 24.3 V Auopas 12 8 4 30 8.6 VI RuWbberoPm 3 3 0 Oa, a vI 'Ektroal 90 4 86 258 34 ViE Petiocbec.b 34 12 22 IX amkds 47 8 39 X Par- 8 0 8 )ai Pba,ue aeu - 13 0 13 635 IV I" XII - PUlPad-Fa" . i9 --4-- - S 31 XI N0*4Vetalfle-wne 12 0 12 21 S.2 MV T_i, Gum., Fotw 8 4 4 Xv LdberGoods,Wood PsI 4 2 2 7i f S7f XVI FP4ad and wuBeFs 24 S 19 712 3.4 VII O0m 3 0 3 ci t .V 11 47 64 1145 9.7 C2 EIwiataisVI 90 4 86 258 349 C3 Vilm-X 102 -- 2D 62 43 16.1 Ca ThdhammtXIVXV 36 1i 25 1469 ',25 TOTAL 366 8 278 3869 9.5 Sowm Own c_mp ad. of R&D enga hosed on bfmmadm km ANM (Natona Amadatw of Idatrial R&D CaM= ft IsdUte of ldMEm ulof do Fedwu Uuwly of Rio de iaadm ankd acommnt of R&D atvIry by hadMifd m%l ln ie Dranm.fan. de Cacs data of fim cqxa1lngam $6 muliaIn 1965. y Tbe nlm of =t_ to aW maufuetuaM rw_fb pbW bbAl sbdet. and fa Gaa Gu eethntiD hp..1987. gdot U ndL 9i Indudna VIIJXL g/ Inudesa-XIV.X I lbe limited pardcipatin of R&D pnoned In te productive sectr Is ev more anuataed in tms of eearche with docora degresL e sbae for the poductve secr Is 0,46%, Mon in pubic eateprise R&D labs. -7- 240 Among R&D active fins, 88 are major eXpot, most concentrated in steel vehce, petemicals, chemials, and autopatts Some of these ane world class producers In sophqbited sens (such as aaf, misles, hig-peoma pistons, ad specia allos). Substanti producto design and marketing experience, combined with purposefl R&D effors, result In thei compedtite strength. A number of R&D acive produces are multiational firms, mostly In the metal-mechanics industies (paticdarly vehides), with some in rubber products and petrochemals Their R&D efforts were not necessarilY undetnaken in response to expot requirements (althougb many wre drawn to export markets in response to strong govemet inducements-such as the BEFIEX export program), but to local market demands (eg, for ethanol-based engines or lge hyroeltk turbines). However, the majority of R&D- engaged exportes are those that combine Bra's traditional sources of compratie advntage (abundant natural resoures and ixpensive labor) with a measure of product development and proems impemnts to stay competitive In world markets 1y are producers of steel, autoparls, pulp and paper, chemicals, teles, and food and beverages 2.11 Most fim that carry out R&D but do o export are in electronics and capital goods Thee are segments with taditionay high R&D intanslty, as suggested by Inerational exper,nce, and that since the mid-197 hawve expanded rapidly, behled high prtconist barrier Although govenment policies and situtional support mehnism have induced these producers to undertake substantial R&D, such acivity has been Insufficdent to close in the techological foler. Moreover, some efforts probably have been midiected and wastul Continuouy hih levels of protection have allowed producers to faU behnd best practie without ben penalized by competlo Acrostheboard import subsitution also led to excessive product tatio, with the dispersion of techoo al resoures. As a result, low vels of intuaindustry and intaflrm specalztion precluded these producers from reaping economies of scale in pwduction and developmt of now products 2.12 R&D Is just one dimension of firms' technologil actities, however, and posil not the most important Producdo planng and oraon, troubhootg resca aw matias adation, and qualty contml ae oth teologial tasks that have a direct bearig on firms' cmpeti- tvynss. But the fact that so kw firms were eaged In R&D is sigficnt, as fims ted to fwmalie R&D actis once compedtie pressures require new designs and major improvements in proon poceses 2.13 Moreover, despite the fact that many indust ms are enWed in minor technogic activities-rangIng from the desi of tools to improvemts in plant layout and production promesses-most of thse activties have had only lmited impact in brngg fim closer to best practicaM A detailed study of these frms' technologkal behavor has shown that mateal inventory was monitored by over 92% of sampled produces, but more than 81% did so manually. Similay, nearly 93% of fims were epgaged In quality control, while just 28.8% employed modem metods (such as stadstical process control).Y At the same time, only 23.5% were commitd to deveoping new products in a systematic way. Imitaion- through mmse engineering or hiring away competitoS-was peceived to be the enal rule in industry. lTese fndings suggest that most technology a in Brazilan idusty consists of mior innovations and adaptadons- uet to cater to a rtWively ptected domestic market but inadequate to sustain Its competitive position in word mares I/ See Helson sxap and Virn Matsco, e TenolUgbo da lndustral Brasieir un Aniise Exploratoria,' Textos Par Dcmo latemna no. 162, WPESIIPEA, February 1989. The study reports the answers to a detale questionaie on the tchnologi acdvitis of the Brazilian indusuty from a 1980181 sample of 4309 industral fms opeating 7156 plants. jgI IL Bragp and V. Matesco, gpL pp. 12-14. js/ H. Bra and V. Matesco, opg, report that over 67% of respondents pointed out that copying of product Ines from oDmpetitors is a common prce in their sector. -8 - 2.14 In sum, the evidence assembled In this section points to vety Umited commitment by Brazilian industri frms to R&D and to technological development In general. As shown in the next section, such lmited efforts are not offset by a regime to facflitate technology imports. The effectiveness of Imported technology is hampered, in any case, by the weak absoptive capacty of Brazilian industry. Box 2.1: METAL LEVE AND rTB LONG-TERM COMMITMENT TO TECI INOLOGICAL EXCELLENCE Metal Leve is a case that illustrates the trajectoty of a very dynamic domestic private firm committed since its Inception to technology development. In 1950 Metal Leve began 1950 producing pistons for local subsidiaries of multinational vehicle manufactures in Brazil (Ford, Volkswagen, GM, and Mercedes Benz). Its initial focus was on quality control, combined with well structued technolo transfer programs, involving engineering of piston manufacturing processes. Metal Leve entered the international market in 1965 as a way to use up its excess capacity, and induced by fiscal and credit incentives. Export-output ratios rose to 8% in 1970, 17%-20% in the 1980s, and 35% In 1988 The driving force for technical change during the early period was foreign buyers' demand for high quality. Efforts at quality Improvement at Metal Leve were part of a managerial culture that continuously emphasized the importance of best practice techniques and methods. Effective absorption of imported technology was the result of prior and detailed studies of the processes that were to be transferred, intense training in-house and at the premises of the technology supplier, and systematic adaptation of the technology to local conditions. In the late 1970s the firm began developing new products to oDmpete in markets that required Metal Leve to supply designs as welL! In 1979 it set up its own research center with financial assistance from the Agengy for Financing Studies and Projects-FINEP (US$2.2 mitlion), spurred largely by the challnge In export marketl By 1988 the staff of Metal Love's R&D center had grown to 230, and the company spent 2.7% of its sales on R&D. In addition, it contracted out research to local R&D centers and universities. In 1988 Metal Leve set up an Advanced Technology Center in Ann Arbor, Michigan, as an extcnsion of its Sao Paulo R&D center, motivated by the need to be better informed about new developments and the future plans of its American customers.! In September 1989 Metal Love opened a manufaturing facility in South Carolina to produce articulated pistons for diesel engines (a Metal Love innovation) to equip Caterpillars new diesel engines (to be introduced in 1990) and also cater to demand from Volo and Cummins. The case of Metal Lve (like that of Embraer-see Section IV), Mustrates the systematic, step-by-step technology efforts needed to become an internationaUly competitive producer. It also emphasizes the importance of export markets as a source not only of technical Information (supplied by buyers or gleaned from competitors) but of continuous pressure to improve performance. Like Embraer, Metal Leve also was quite successful in using Brazil's institutional system for supporting innovation. It maintains close interaction with FINEP, as well as its colaborative ventures with universities, technical and research institutions. These arrangements suggest that a motivated management with long-term commitment to technological excellence and with a record of technological and commercial accomplishments, is able to tap existing technology-oriented institutnal, human, and financial resources effectively. I/ Between 1970 and 1980, 31 new types of pistons and S9 types of a_ebeaiV wee int_dcd, hInuding hay dill? pist for ar dIesl engines and a pecal line for alool vehides yt Te tepo4ted cest of the new center ws USS3 plllion, 90% of whkb was finced by FINEP. See ceta MeJt Fedbny I, ' - -9- m. TECHNOLOGY POuLCIES 3.01 Explicit policies for idustril technology dvelopmet in Brazil have followed the broader ndustial policies of the Govement. lhe acquisition of fotreg technolog through ans-length technology trnsf, capitl goods imports, and direct fegn investment bas been strony Influenced by the gverment objective of minimizing the outflow of forign exchange and promotig import substitution. Similarly, flancal Incentives to foster R&D and boost the engneeng consultng capablities of national enterprises were foused on Import-subsdtuting acdvides or nadonal producers that were cmpeting In areas dominated by muldnatonal producers. A. Arms Length Technolok Trausfer 3.02 Brazils curret technology tra regime is designed to impme the brgEining position of national firms and mmie fore echange outflow. Sine 1958, the Cental Bank has controlled royalty payments and has established that the maXiDmum royalty should be 5% of net sales. The Goveran- meat's main concen was that foreig fms were using technolo taner paymemns as a way of remitting profits Subsequently, In 1962 the law govening foreign Inmestments in Brazill established stict controls on tehnolW tansfer, whkh are sti in effet. Th law requird that foreign payments for technology tansfer be registered with the Central Bank It also prohibited tenolkgy transfer payments between subsidiaries and their paent companies and between joint ventures and freign partners who held more than a 50% stake in the venture. Me rationale was that the parent firms already earned a return on the technology, thtough profts, and therefore should not be allowed to deduct technology Uesing payments. The law maintained the 5% maxImum royalty rate, and also rsicted deducting technology transfer payments for fiscal purposes to a maimum pedod of five yeams. 3M3 In 1970 the Natonal Institute of Industrial Propert (INPI) was created. Among other ponsibities, it was to regulate technology traf (taking over from the Central Bank). Techology tansfer agreements were divided into five categores Each agreement must be registered with INPI as to terms and payments authoriad, period of valdity, required Brazilin pardidpadon, and specil proviss, depeding on type of cntract. Royaltes for pent Ueesm and tradeakrk lienses may be paid only If the lieses are registered in Brazil Moreover, the intelletual property reghme in Brazil does not offr significant patent protection in the areas of chemis, ph, metal alloys and miures, and alloys in geeraL One of the principal rationale of the Brazlian property rights system is to protect technologiclly less capable local firms against potentially predatory behavior by foreign produces and to reduce the Qost of appropriating elevant foreig technology. However, a frail intelecual propery rights regime cluding weak trade secret proecdon possibly dets foreig firms fmm transfering or using lca their most advanced or up-to-date technology. 3.04 Induist & technology liense arements are ess onerous since such licesing does not depend on INPrs regitering the patent or trademarkif However, INPI treas such an agreement as a sale: after expkadon of the contract, INPI rerds the liensee as the ower of the technology. In addition the partes must satsfy INPI that the technology lI/ Law 4131. 1 f AC. Cardozo, 'A Implantaio de Leis e Regulmuentos Sobre Transferencia de Tecnologa: A Eixpeenciia do Brasi, mW mex 198& JW For more information on the regulaions covering the different types of agreements, see IL Rosenn, Regultin of Foreign Ivesment in Bral," 19& * 10. * Is rot avalable domestically ad that Its transfer to a Brazilan lcensee is In line with natona deveopment objectives; * brIngs real advantages for the dereomt of the dustral sector, * improves product quality and allows or subsitution of Brazilian products for imports; and * allows the Ucemee to absorb and muter the technolog within the letime of the contract Registratio of _tecnil and lndslrlmal nape _a e depends on coanincg INPI that the services are not avaiable locW and that there wil be short-term benefits to the sector. Ths might mean producing a quality product with expot and/r impot substution poIbItIes. For techal serice a _eememta INPI must be satisfied that the seves are not available domestially. It also must approve the detailed schedule of payments for foreign technicians. 35 Mmum appoved royalty rates range from 2% for plastic and rubber artiles to 5% fow electical equipment* Those roealties are based on net saes, wich can be consideably wer than total saleLff Royalties for trademarks have a maimum of 1%. In addition, royalty payments are taxd at a rate of 25%, and that tax is counted ithin the pecfied maximum royalty limits The maximum duration is five yeurs, with a possibl extension for another fie years, at the dicretion of INPL Also, as noted above, royWalty payments are not permitted between subsiarie and ther parents or between joint venture fms and foreign partners holding more than a 50% stake in the vture. 3.06 Technogy contracts prohibit restrie clauses lnvolviln limits on exports, pridng guideline use of tied inputs, purchase of other technolok; secrecy about the technology after expiation of the contract; and obligatosy transfer to the seler of improm ts made in the technology by the buyer. MTe seller, however, is obliged to pass on improvements made to the tecnolgy after it was sold and is lable for any lega action that originates because of defec in the tedcnoloy or intelectual property right infringements, even by third partes. In additon, fui disclosure of al technica data is requied, and engineedag drawings must be provided, as well as al informatin no essay to update the know-how. ITe latter requirement is of extreme cncern to many technoloVg suppliers who fear that teoloy information may be appopriated by third pates. 3.07 The impact of INPrs attempts to control technolgy import bas not been evaluated systematically but may be quite sgniflcanL Between 1972 and 1987 more than 23,500 contacts were submitted to INPI for approvaL Of those, about 18,000, or roughly 75%, were approved. INPrs apprv rate has followed a generlly downward trend, the result of its striter control, especall regarding tecnical services (Table 3.1). fL/ In 19S3 INPI issued Normative Act 64 whose pncipal objectve is to strengthen natonal technologcal capability. Under this Ac, INPrs approval for a technology transfer agreement is conditioned on whether the recevn firm has an adequate program to assimilate the techology and amy out R&D to grain great technological autonomy. The amount of the investment required depends on the fiancial situation of the Braiian ntractor and the amount to be paid for the Imported tecology. wI However, it appeas that since the New lodustil Polc was iniated In 198 royalties as high as 10% are being permitted for some higt-technolgy sectors such as softwar See K Rosenn, o .12/ Net sales is calulatd as total sales minus duties, taxes, Impored inputs and cmponents, commssions, transport, insuane and other deduction. 3,1 mN APIROVALS OF ¶BCHNt)oOY IRANM COMRAC1 \ ~~1978-19617 V ~~~~~VBAR CONTRACIS CONIIRACIS, PCENrAGES 'RESBNTED APPROVED APPROVED FOR APPROVAL 1978 1473 1451 9S5 1979 1456 1416 9773 1980 1576 L332 84.5 1961 1426 1178 826 1982 1438 1135 789 1963 196 969 74. 1984 907 786 86.7 1591 1043 769 73.7 1996 1185 865 74.7 1967 1S15 1213 65. TOTAL 13615 11134 8.6 SOURCE: Carow (1938) ted an DIN da 3. Fiy perct of all teology contracts appoved betwee 1978 and 1987 e conentrated In four of the 70 economic sectos that INPI uses to clasfy contact (mechanical, metallurgIcal, chemic and minig). Whie the share of most of the ladg secwts hae been lng, eleonc and delca materil a expandig rapil, with the propo of appvd cons i om 4% I978 to more than 15% in 1987. lbIs party reflecs the ased impoan of o technolor purases by the electon ind , in direct fort has bes severel rsted i umbe f its segments. 3.09 Data on technolo paymts are avmsble for 1979.lM7 (Ilb. 3.1), showing paymets of US$2.1 blion during that period. Rogty 81% of the totl was for eialid tehi services evec thoug thei shar fl parke ower the period. About 15% was for unpatented Industrial tecbnology (ilnduding 7.1% for npaed indastrial techn for maet40rder-cptal goods). lhe smlest share (Just 3%) went for patn and trademk likeses Thales _undstate the actual Ins sine payme ts by subsidia to their paret compani are not allowed. An almost coDuus decine in the values recorded is notable and this fll is nOt e aid br the ecnomic reoosio alonedoI 3.10 It is not dear to what extent INPI has imprvd the bargaining posto of domestic frms and helped them obta dol ore cheaply, bowe There are various reasons tosurmise that INPrs cotols may hlve been ddelterio to ntiona Int in two repects. Itr the regulations ate against national firms becase these firns must pm ther conts thrwp a bureaucatic process for apoval In contrast, sudiaie of o n mpaies hae direct amm their parent compane' tenoy and an tnsfer the cology egm It the amot deduct the cst Ibr tax puposes Semo, fDi fDreign teology supplie, including investo resicte transfr nuols as well as inadequacies in the direct foreigd inestmet regime (e.g, not being able to captliz technoloy contributions or to deduct tehnology fees) ma It not wortwhile to liceme or even use ther best technogy In BraL is situaton is awaated by the lack of adeqa pmotet from the itelecu prert regm The net rest Is that Brai may be receving obsoete or outmde technolog. -12- Tab32 PAYMEN73 FOR-TBCHNOLOOY hANSFD CONTRACT 1979-1987 OIn im of wta doom) Yaw Paat & 11 Indb Ta!nl Total baew CO _PII" T _d" 8ervlem 1979 9 6 11 287 313 1980 12 11 14 284 321 1981 12 18 13 234 276 1 5 17 10 218 240 - l§ 12 a10 14 182 218 1984 9 8 8 177 202 1985 S 21t 41 18 175 1986 2 20 43 11 184' 1987 3 39 26 106 174 TO`TAL 69 1SO 179 1705 2103 Somz AC. Caw am e bdL d on Anusl Repots of the Cal Dank L nImports of Capital Goods 3.11 Aaess to the most poductve capita goods Is citcal i an ecom Is to benfit frm embodied advances in Wchnology. B1a1 has relied much less st growing East Asian ecoomi on imponed capital goods for acquring technolog (k ble 34 )Moreover, the rado of capital goods hlmports to gross domestic invesbimn (01)) ha decased Mac 1g90 Thi rested fom the polc of promoting the local capital goods indlUStI through trade asiction aon imports and special financial incentives for lal production. Emesive protectio may be hampeig the International mp of loca iustry. Table 32 UPOR73 OF CAPITAL GOODS AS RAno OF ODI 1965 1970 197 191i ism 1986 1987 Hass Kan 1.949 3.76 2962 2.24 4.073 326 3848 Soth Korea 0991 0902 I 1.119 1114 1.0M4 1.046 mabysa 1.768 1.488 1.435 1.441 1AS4 1.541 1A36 Sngapore 578 3357 3.592 4.42 3A489 3801 4.140 Thaiand 0889 0.714 0.824 1.112 1129 0.944 L041 Brz 02S1 0327 048 0.455 0376 0299 0.259 Medeao 0352 03312 0312 0.370 0.400 0390 USA 0.181 Q270 Q442 075 0Q559 0.574 0.91 Sono: Own _ bi anl UN trade data and DESD ypam Note Cal goods wae defed as te followng Sr RevIin 2 epegom:, 711, 2, 714, 75, 716, 717,718 719 (m 7194n) 72, 72Z, 72 7249, 72i, 729,7 96,727,729, 731, 732 7323, 734, 7325, 7327, 733, 734, 73, AND 861. be r vadoen the ta a gwe tr tha am fr cme of the Aam conneAs bme tany of the imptd, capit p we _daWane as pal of am pred -13- 3.12 7h restrci capital goods Iimpot regime bas had significant Impact on as wel So many manustues in many speciazd product areas (ven more from in the U.S. or Japan) has meaut an overly diversied capital goods ndusy lacing the economies of scale or specalztion Iecessary to reduce cOs, keep up with new technologi, and stay compettive. Exessive dependence on more expsve and often outmoded locally-produced capital goods, espealy in electronics and process eontrol, bas placed the dowtream user industry at a dadvantage with respect to Brazis international compettos C Foreg Investent 3.13 Brazil's foreign ivemt poly has rflected the gvrmets objtve of attracting capital to specific industrial segments and (after 1972) of obtaing eWpot commitments from Individual fims. TechnoW development has been of secdady Impoace to import substitution and balance-of- payments consierations Although the Inflow of det fore investment (DFI) bas been coniderable, Brazil may not have been receiving the most advanced technlogy. he orall effect of restricon has made foreign investment in Braz, parularly instment with a high technology content, relaively less attractive than In other locations* 3.14 Profit remitances are limited to 12% of the original Investment plus reinvested eamings./ In addition, the original foreign currency vwalue of the inmestment is not correted for inflation. which means that the origi base for remittances decrases in real terms over time. Fthermor parent companies may not rie royalty payments for patents and trademak from their subsidaies and may not capitaze their tecnology as equity contributds to the investments. 3.15 For minority joint venture partnes, there are also arbitry and strict limits on maximum payments for technology trandsr (as noted above). In addition, the foreign Investor who brngs technolo to the joint venture has to pay a 25% tax on thi part of his capital contribution to make it paut of his remittance base. 3.16 Still, the foreign investment regime Is quite open. In genal, up to 100% foreign ownership has been allowed in most areas. Until recently, the only sectors closed to DFI were petroleum exraction and refining, orain segments of the inofmatics industy, communiations media, and most domestic aspon, although there were also some rmttins on mining, banking, insurance, and other finandal acitieS. The new Constitution adopted in 1988 has taken a more nationalistic posture toward DFL Its main restrictions were Ihmiting foreign control in miinng, mheal exploration, and production of electric power, and prohibiting foreign oil companies frotn engaging in oil exploration through dsk cacts However, the new Govnment that took power in March 1990 has announced a more open poliy toward fore investment, and a posible reduction in the scope of resrcto in the informatics sector. 3.17 Bua has the largest stock of foreip capital of any developing country. Registred foreign Inestment was US$27.9 biUlion in 1987. Almost 75% of all foreign ivestment is In manufctu 20% in servioes, 3% In the extactive mineral industy, less than 1% in agriculture, and 2% in other actties ZI See IC. Rosean, pu it. D/ Profit remittances above 12% of registered capital are subjected to a 40.60% supplmental income tax ;22/ Thi causes tax problems for US. companies becaue Sections 367 and 382 of the US Internal Renue Code force companies to include impute roylies on tehology transfens in their US. income for tax purposes See IL Rosenn, op._cit - 14- Within uctuig, the subsecors acounting for the: larst shares of foregn investment ge: automobiles (11%), basic Obeicals (9%), mechanical (9%), and electrical and communiCations (8%). Thl particu concentration occurred because most of thi investment came to lrail under specific sectoral programs for automobiles, shipbuilding, capital goods, and chemials. Such programs provided fiscal and financial incentives (as well as protecion in the local market) In exchange for perfomance requirements peraining to ivestment volume, local content, and exports.9 At the same time, foreign Investment in professional and industrial electronics, and data processing and telecommunications equipment, has been limited with podsible adverse impact on industrial competitiveness D. Financial Incentives for Local Technology Development 3.18 Financial incentimes bave been the main instrument for encouraging the development of technologc capabflities at the firm level Since 1973, FINEP (Agency for Finandng Studies and Projects) has used subsidized loans, risk-sharing instruments and, to a lesser extent, equity partcipation, to foster national firms' tenological activities. In the perid 197349, FINEP contracted 1,761 technology development support operations for approximately USS810 milloi The focus has been on the devlopment of import substitutes (as in chemicals and pharmaceuticals) and producs that would allow national frms to compete with foreign-owned producers in the domestic market (as in capital goods, electonis, vehicles and autoparts). here has been a growing emphasis on establishing R&D and quality ontrol labs and developing and improving products, process and tools. 3.19 Although there is not enougb information to evaluate the impact of FINEPs programs, It is likely that most national firms with significant R&D activity have benefitted from FNMEs assistance. Nonetheless, the efetvens of its actions has been limited not only by relatively narrow lending criteria, but by intenal bureaucratic obstades to timely operation, and the difficulties small, technically-based frms encounter in quahling for its progm Larger frms, on the other hand, incraingly find FPNEPs financial capabilities too limited for their innovation finance requirements* 3.20 In condusion, the technology policy regime in Brazl has been charaderizdw by objeces other than the aqWsition of technologic capabilidies that would allow firms to become inernationaly competitie As a result, the countly has failed to attra best-practice technology via direct foreign investment or through arms-length transactions A combination of weak domestic technological efforts and estrictive acoess to the most valuable foreign technoogy appears to have hampered the moderizaton fforts of Baziian fims As argued in the nest section, substantial public sector inolvement in technological activities has been a weak substitute for greater invlvement by the productive sector and a more flexible technology policy regime The main progm was BEFEX, established in 1972. This program, initially developed for the auto Industry, gave firms incentives in the forms of reduction of imports duties and other fiscal exemptions as well as financial incentives in return for specific export commitments. The program was expanded to other sectors and also to national firms. In 1989 it was estimated that SD% of all Brazilian manutu exports were covered under this program. SI Note in this respect that the value of contacted technology development opertions in i989 was less than US$18 million, little over one-tenth of what was contracted two years earlier in 1987. - 15 - IV. INNSTIErWRONAL N RK SUPPORTING TECHNOLOGY DEVELOPMENT A. B ackgnd and Current Strutre 4.01 The fist three decades of the 20th century saw the beginning of dusttion in Brazil, spurred by growing agricultwal income, the epaion of domestic trading actities, and profitable opportunities in the production of importables. Exept for foundy and metlwoldng for the ralroad network and the use of steam power, there was not much demand for engineering manpower or local research activities, howver. Technologies generaly were obtained from abroad by copying. The difficulty of Importing during the Great Depressio and World Wu 1 senred to deepen the proces of Import substitution and led to was growing demand for more sophitcated technical and support services. Thi brought about the establishment of some of the At major iust reseawh institutes (induding IPI in Sao Paulo) and the first unieresity with a stong emphasis on sdence and technology. 4.02 The most significant developments after WWII were driv by the militaiy, which had readily understood the strategic impornce of science and technology to the outcome of that conflict. The military set up an aerospace engineering taining institute (ITA) in 1946 and the National Science Couanc (CNPq) in 1951, with the initial purpose of understanding how to hamess atomic energy. Neverheess, Bris S&T effors until the mid-1960s were still isolated and dispersed, with lmited Impact, except for the development of technical human capitaL 4.03 Ftom 1964 to 1985 Brazil was under mlitwy rule. During this peiod, there was strong emphasis on formalizing the economic planning prooess This included dfting science and technology plans, establishing the National System of Scientific and Tehnological Development (SNDCr), and setting Up various speciaed Institutions to direct and finance scientific and technological development. The whole S&T system was expanded through a process of quantitative tareting, mirrodng the overall approach of plannig for growth. TLe system was centralized, and the main efforts were in expanding research capability in public sector institutes and enterprise This was largely to strengthen national technological capability in military and strategic areas 4.04 Since 1989 the Special Setariat for Sience and Technolo has been the apex S&T institution in BraziL In addition, 17 major federal implementing agencies for S&T policy receive at east 1% of the S&T budget. Combined they acount for over 90% of total allocations for S&T (Table 41). Approdmately 40% of the allocations finance research related activities (metrology, norms and standards, normative coordination, for example), and the remainder goes to postgraduate training and actual researcL - 16- Tae4 KEY S&T AGENCIS ACCORDING TO B)DGHTARY ALLOCATIONS 1 Agn Fbatlo Fhl 198 Budgetay A9oe. Amot _W pCWCt CNPq - The Natl CocO for S&T Dev. T,R 291,639 17.0 EMBRPA - BS AgicuL Rmswd. Corp, T,R 200,542 11.69 Secrutarlat Of Sec. of S&T g/ NA 183,903 10.72 CAPES - Ow. Ag. or Paugrd. Tann TJ 139,108 &11 CNEN - NaIl Counci for Nucdea Emu NAT 129,128 7.53 B#lRATfR -1r Rutal Extaion Cap 1 98016 5.77 IFES - Fedal Unv and oe TeadL lst. T.R 93,842 5.47 C;N -Natonl Sa CD N 89,431 5.22 BMFA - Joa of Staff R 72,760 4.24 Sacrutat iu, ol Mie and Energy i N 65,617 3.83 INWE - hstuWe of Spac Rarh TiR 48,421 2.8 PIOCRUZ - ounai InsL Owado Cn T,R 33,137 1.93 Mi. of Asniwautic R 31,931 1.86 DIMEIRO - Natl IstL of Meu, Stad. I, 27,178 1.58 Seerazarhi ot Mm. at Jus - 25,199 1A7 CEDATE - Caute for ThCh Sapw of Ed. W - 21,083 1.23 Cl - Center for anaties Tedolo R 19,91 1.16 INPA - Na InstL for Rae of Amazo Rag R 18,981 1.11 Subtotal 1,590,644 9.75 Total S&T Budge 1,71485 100. Sourcn CNPq, 'Recoua de Tewouo da Unbo Docao lalial, Dotso Fn, e Dqpm Reallrada 198049. may, 1989. it T stamn for Tnlag. R for eut% D for d h of inaon, I for Intitutlal devlopment an N for normaiv cOoILad hf Amounts in 1987 dolg Inludes the FNDCT ad the PADCr eeh fundh W 80% s I aftd r equity intment In NUCaEBRAS. %90% for dsao sdfc equpme and rurments. B. Public R&D Instituts 4.05 A complex and differentiated ovrnment institutinal newk carries out public sector R&D activities In BaziL In major areas, such as health and biolil scences, as wel1 as in experimtal physlS and chemistry, mot resourms are in the federal and State o Suo Paulo universities. Public enterprises' in-house R&D faciitis cover enea genraion and conservation oil explorto teule sialos, and aircaft develpmst. Covement institutions not attached diretly either to fderal or stt universities, or to public sectOr enterprises, play an important role in agriculture, food tchnology, Amazon eclog, biotechnology, tropicl disease researdc, physics and nuclear technology, aerospac*, and comput sciences. In additon, devlopmen efforts in various areas of engineerng and mining technlogy are undertaken by a few staeWvel institutions (the most important of which is IPT of & Paulo). 4.06 Goverment-undetaken industrial R&D has not been very effectie due to the geraly weak lnkes with the productve sector. A survey of sources of technolog in Brazilikn industry notes that less than 2.5%* 5.1% and 3.9% of product designs tool designs, and maufauring proesse, respectivl, has originated in rerch itstiutesL 4.07 Although industy ties are strong in some cases, the dispedon of resoure through an exes number of research projecls and competing aactides hampers effbrs An example of the dispersion of a research program in the publc sector is the work program of CPqD, the research unit of Telebras, BmIs public telephone hoing company. In 1988 CPDwas the largest and most sophistcated m5 IL Brap and V. Matesco, op Table 2.1. - 17 - applicationslabotory in Latin America. It employed 400 professionals directly engaged in R&D work, in addition to sponsoring personnel from ind stty and universities. CPqD's budget of US$60 million was allocated to a very broad research agenda, ap rodmately 80 R&D projecs in seven priority areas: electronic switching, digital transmission, optical communications, data and text communications, satellite com- munications equipment, tools, and materials. Even adjusted for skilled labor ost differences, a budget of US$60 million is minute for the scope of CPqD's research agenda. Budgetary allocation would average less than US$8 million per program or US$1 million per project, whereas each program could in itself justify the whole budget. The absence of research focus and specialization has led to systematic delays in the CPqD research chronogram and, therefore, to delays in the market launching of its applications. These delays have had a high cost for the economy because in many cases, such as the digital exchange program, government policy prevented the use of existing, less costly, foreign technology while waiting for the local technology to be developed.09 Bo 4.1: THE INCUBATION OF A TOP PERFORMING ENTERPRISE Despite certain fundamental weaknesses in the institutional support network for technological advanement, there are many instances of successful cooperation with industry. In these cases strong and effective links have been forged between the public support infrastructure and industrial firms, and producers have achieved international levels of competitiveness. An impressive case of a top public enterprise with strong roots in the public research infrastructue is that of Embraer. Embraer progmssed from a protected and subsidized public enterprise to a dynamic world-class competitor. In 1946 the Aeronautics Technology Institute (ITA) was established to generate high-level human technical resources. Drawing heavily on ITA graduates, the Aeronautics Technology Center (CIA) was created in 1947.Y Among other tasks, CTA was given the mission of designing an airpla- ne suitable for Btazilian cnditions. The Banderante turboprop had its first sucoessful flight in 196& The original team then trassfered from CTA to set up Embraer in 1969 and produce the Bandeirante on a commercial scaleY Five hundred Bandeirantes were produced. Most were sold internationally, and the plane found a niche in the commuter airplane market. Embraer followed the Bandeirante with seral other models, again focused strong on the international market} Embraer has been proficient at using public research institutions and universities as well as tapping finance from FINP. Its strong focus on the export market has ban crucial in offsetting development costs by perwtting lager production scale, in bringing in new Ideas for further technical change, and in demanding exacting performance standards. /I By 1988 ffA had tained more than 3,000 eaglees, 800 of wCh wae la ghe aernaut BKd The ITACrA gaduate _agkneesin& educadona/earcb e led to the spontaneo t of 8nel's (hit and largest industrl bigh to pk around San Jose dos Campos. Alon with Embaer it has lead to the establishm avy high tecdnobg companie includW Avibs (mbm). Orbia (msl6s), Eagesa (muTitaly equient), Tecaa (elecurnic comauncaktion equipmat), Cmposte Tecooga (compose maials), and Quantum (oftware). As of Februay, 1989 it had an accumD ated producion of 3,983 aIcrt. 2§/ C'qD however has been fairly -sccessful in transferring many of the products it has developed to industrial firms for lirge-scale production. By end-1987, 75 different products developed singly by CPqD -or in association with univesities (which were in charge of carnying out most applied r.search) and industria~ firms (generally focused on the later stages of development, such as prototyping) were being manufcurd by 25 producers. See C Frischtak, 'Specialization, Tcial Cbange and Competitiveness in the Braii Eectronics Industry," ndustry Series Pap. No. 15, The World Bank, Industry and Energy Department, 1989. - 18 - 4.08 in two areas howev, govenmnt R&D effrb have been quite effective agrotechnology and aerospace In agrotechnology, the goverment agenc EMRAPA (Empresa Braslera de Pesqusa Agropecuarla) has coordinated and party executed a complex and fa ranging natioal agricultural reearch program. lhis has resulted In, among other things, the incorporation of previously Inferdle land through a better understanding of sofl biology and the suYcessful introduction of new plant varieties. 4.09 In aerospace, Brazi (with China and India) probably has the most advanced research program among industrializing countries. CrA (Centro Tecnologbco da Aeronautca) and INPE (Instituto Naclonal de Pesquisas Espacis) have been the core R&D institutes in this subsector. CMA, in particular, has generated and efecidvely transferred aircraft, rocket, fuel and other aerospace-related technology to Industry. In addition, It has had an important role in the development of gasohol-based engines 4.10 Among industryorlented institutes, IPT (Instituto de Pesqulsas Tecaologicas) of the State of Sao Paulo is posibly the one that has forged the closest lUn with the producdive sector. Other than IPT (and two more state-level R&D institutes-those seing Minas Geras and Baia), government-led, industry-diected efforts have been smal and not vety effectiv C. Public and Agregate R&D Expenditures 4.11 Public sector R&D has been dominant not only in scope but in resourc allocated. Although there are no firm estimates of total R&D expenditures in Brazil-private sector outlays are not known wih ertaity-most sourca estimate that public sector expenditures amunt for betwe 70% to 90% of total R&D outlays Based on actual public expditures, a series of Imputed private and total R&D expenditures can be construed (Table 4.2). Tabk R&D EX?BtDlllRBU IN BRAI, 19818 ( consat "97 doom) yewr ?uhic apend alf GDP PRODM TEGD? bk Tn D? i981 966,133 253 O 0.38 0.54 0.42 12 1,184,970 M Z U6 0.6 0.51 1983 938*7 247W500,345 038 .54 042 1984 907,447 2538,14 036 0.51 A40 1985 1,242,271 262,347,174 0.44 0.63 0.49 1966 1,455,Z39 314,47,537 0.46 OA6 0.51 1987 1,485,363 323 0.46 66 0.51 1988 1A68,59 322,597.503 46 Q65 0.51 So= CNPq, Rcuros do Te da Uobe Damo lokial, Doco Fnal, e Dapa Raua 190.89" May 1989. gl For 198147 these are actual ezpeadltu for 1988, it h the final budget allocaota I IBI Is total Imputed ae qtum on the asmumpdan that publc pedtur muak up 70% of R&D outlas In SaL Sf TE2 is total hiputed expdtume on the a th t public xpedtm mak up 90% of R&D outlays in the coury. -19- 4.12 Although In absolute tms R&D outlays in Brazil ae not modest compaed to other industrializing countris, they are small in compaison to developed economies Crable 43). Moreover, after a signlflcant Inease In the 1 , R&D expenditures as a proportion of ONP remained basially flat thmouht the 1918, wherea they expanded substantially In the fast rowing East Asian economies As a result, absolute 6evels of R&D expenditues In South Korea, for example, are at lamst twice as large as in Brazil, although the Brazilian eonomy Is more than twie the sie of the Korean ecnomy. Tsble R&D BEDfrlVRE IN SELE)CTD COUNTRIS (R&D Bup_ntuz In absolute tnm s nda a w p utos or GN?) County Abslute Expend. it R&D Intenlty Amoh Amoumt Year 1970 1977 1912 lsti Yew BRAZaL 1448 1982 0.24 0.70 059 059 (1987) bf Agentina 1087 1981 na. 10 0.20 020 (1982) M o gal881 1984 020 030 020 060 (1984) left 1482 1984 .s. 0Q50 076 L0 (1985) S. KOreM 1307 1983 039 0.40 090 1i8 (16) Tab= SO 1985 0a us 090 1.06 (1985) jap 39117 1987 1.90 200 240 19 (1987) US 100823 1987 260 2.10 250 26 (1967) Soaur UNESO, Stdsl Yeatook 1988 Par, 1989. nd endit_ to mWmEn of 1982 dabom b Avge betwee the low and hih atmas of tabe 6.1. 4.13 Is Brazil devoting an insufflcient amount of es to R&D? It should be emphaszd that even if R&D outlas are sagnt, t certainly does not imply that Brazil is spending too little on scientific and tedhnological devlopmenL In the absence of bkde about rats of return in R&D compared to other ecoomic activities, ot much can be said about the optimal level of R&D expenditures. On the oe band, a number of cowtres that are aocating bWger amounts of resources to R&D are not neoessari raping greater benefits. For e.mple, between 1977 and 1985, India doubled its reative outlays to R&D witout major gains In terms of tecologiall capablties in the producte scto. On the other hand, counies that hae Inppved el compeie position have also been inresing ther R&D efforts on a systematic basis More Important, most of these efforis hve been underken by the productive sector. 4.14 In contas, not only the level and intensity of Brazil's R&D expenditures have been flat since the begifng of the deqde, lOut more Important, most R&D is both financed and undertken by 8erment (Table 4.4) At the beinning of the decade, n t was responsible for 67% of R&D fIing and ndusty for only 20%. Esumas for 1198 Idicate that with an additional 10% transferd from government and other sources, industys share of Invati financing may have reached 30% Tabk R&D FUNDING AND IN S0dC COUNIE if ) J - .. .. _ _~~~~~ -20 - Table 4.4: R&D FUNDING AND FEBNDnURES IN SELECTED COUNTRIES S Comtfy Year Sore of Fnding e of Eqendlture G I F 0 1 HE GA BRAZIL 1982 67 20 S 8 30 17 53 Axgentla 1981 95 - I 4 41 22 37 Meder 1984 IS I 1 83 30 S1 19 India 1984 87 13 - - 26 - 74 S.Korea 1986 19 81 - - 67 11 22 Japan 1985 21 79 - - 67 20 13 USA 1S86 47 59 - 3 73 12 15 Source: Unels, Statistct Y Padis 1988 aI0 stands for gomanment I for Inuat, F for focelgo 0 for other, HE for higher edacation, and GA for governmet agenc 4.15 Thb pattern of govenment domnace of R&D efpenditures is also apparent In other industrializing economies-such as Argentina, Mexdeo and India-that hae not been able to sustain their competitve position in world markets for more sophisticated goods In South Korea and Japan, in contrast, industqy both finances and caries out most R&D. In the U.S., financing is shred equaly by govenment (a good propordtion of R&D being defense related) and industty. although industry ultimatel undertakes most R&D. D. Human Capital Formation 4.16 Brazirs education system is one of the main obstacles to the countys moderizaton and tectbological upgrading Although major defidencies characterize all components of the system, more fundamental weaknesses aie observed in primary education and at the top end in science and engineering. 4.17 Pin=a and Secnday Schooling. Technical progress in industy is increasingly dependent on an educated labor forc Low levels of basic education and skills are not compatible with the compleaity, precsion, and consistenq of modern industry. Yet in 1980, 73% of the Brazilian labor force either had no education or had not completed primar school, a figure among the highest for middle-income countries. In 1985 Brazil's total secondary enrollment represented only 35% of secodry school-age popation, well, below the averag for middle income Cwuntries (Table 44 4.18 Brazls heavy investment in vodational and technical training has compensated to a limited extent for the weakness of the formal school systenL Although the proportion of students enrolled in vocational training as a proportion of the working age population is less than in South Korea or Taiwan, it is comparable to that of Mexico and considerably above other comparator oountries (fable 4.5). Yet the trainabiity of students at vocational and technical schools increasirgly depends on the quality of their basic school education More generally, the tcnological content of industrial and other economic activities requires growing levels of formal education for thl labo force , which the Brazin education system ias failed to provide. 4.19 EIgher Educado Between 1960 anid 1985, the total number of terdary students went fom 93,202 to 1,437,232, a rapid expansion. As a peroentage of their age group, students enrolled in higher education comprised 11%, up from 2% two decades earler (Table 4.5). Altho .gh that perentage does not compare poorly with other industrialWiqg countries (with the exception of South Korea), certain featurs -21 - of the higher education sstem td to undrine its dfctves as a breeding grund for tecnl labor and Innovation * First, the three-fold expansion of un uate enrollment during the 1970s, comblned with a decline n fMl-time fculty, has had an advwse effbt on the quality of univesity education across the board. Man postgraduate courses have been edged to remedy the insufficiendes of under- graduate trainlg. The low status and poOr qualcain of undergraduate teaches are the core of the problem. About 45,000 full-time teachs hird intialy on a provisional basis, without formal procedures or evaluation, are now tenured. Most lac academic training and bave not gone beyond a BA degree. They safve some 450,000 students fe, public untstie4ZY In addition, around 60,000 teaches, also not vwell qualifd and many with a large teadhing load in several Institutions, serve 850,000 students private schols * ,Second the postgraduate system also faces severe quality problems. Although in 1970 there were 57 doctoral programs in Brazilian uniersities, in 1985 the were more than 300, with another 800 providing trang at the MA. level Combined, they were graduating 5,000 students at both levels each year* Yet there is a wide quality vaane among postgraduate progms, and accordig to CAPES' (Coordenaclo de Aperfeicoamento do Ensino Superior) evaluations, only about one-fourth of new graduate programs are academially satisfactory. Most progams also fc uncertain finanial support from FINEP and CAPES, and very few have been able to dimersiy their funig by likng up with or supporing Inwvative effots in the productive sector. * bhd, the proporto of students in sdence, mteatles and engineering is relatively small compared to other industialzing ntries, being cnsiderably below fast growing East Asian economies and Mdico (Table 4.5). In engheering alone the dispropor is even greater. there ae more than four times as many enginern students in as a percentage of the popuation in South Kora, Hong Kong, Taian and Singapore as tbere are in Brazil; and there are three times as many in Mexico (Table 45. 4.20 The cmbination of low-quality udergraduate teaching few strong gduate programs, a relativ small pool of students in science, matbematics and engineering, and underuse of edsting educational capabilities outside fonal institutions of higher laring has contributed to Brazi's lagging R&D manpower. Brazil has fewer scientists and enges engaged in R&D in rdatio.n to population (256 per millon) than lwan (1,426 per million), Singapore (960 per million), South Korea (804 per miSlion), ad Argentia (360 per milion). 4.21 Moreover, the distribution of human resours aloUted to R&D is co trated in biologicl and health sciences, as well as applid social scienoes, human scences, and the atsditeaure gable 4.6). In two key fields, namely eni and agian scienes, the propoon of resehers is reatively smalL ThMeir seconday position in the ditibution of R&D resarche suggests that much R&D undertaken in Brazil is not closely related to productive acivitieL 27I See Simon Sarlzman, 'Brazwl Opportunity and Ctss in Her Education, lgh E't" 17, 1988, pp.99-119. 2f -S. Schwaptna. c p.104. -22- T1bW94 DDICAWO1N OF D4VWUfdWfl IN HMAN CAPiTAL 11N SBLECIUD NI(O AND WAAN &.ENo= Ibla R. Kqq Slajcr Ita Mude. Indi Idniani Jap Peren AP OtW p-Vd Il: (190) 101 97 103 105 laS 92 74 72 100 (198S) 96 100 105 115 104 t1 92 118 102 - EmdrnUe (195) 35 38 29 45 16 17 27 12 82 (198) 94 91 69 71 35 55 35 39 96 6 7 5 to 2 4 5 1 13 33 13 12 31 16 9 7 35 No. ci To"tltuy S --de.-I per 1W.0 Popahllam Po) past 3606 2060 1410 1406 114 158 776a 600 2006 No. of Tleid,y SWdeib i CSE Bb (0) 585 207 36 22 Su 563 1443 23S 707 000 (tv" (1 (tw 0 (19 0 (twM (M) AS of Popubdmlr Todal t30 06 067 0.89 40 0.70 021 0.14 058 ueba 22 136 4n72 0s 058 1.2. O7 03 07 No. of StWdnt In SM. g (MO 320. 151 27 162 3233 336.9 1269 1373 4869 AsS of FPd_ Teal 0.76 O7 051 7 0.24 02 0.9 oa9 00 Usban 110 L0 0m 73 034 0M59 0.86 03 q3 No. of StudesaB In (i0) 227A 127 21.1 154 1646 2818 397.0 1093 4189 As % of Pap b Tosal 054 04- 0A41 Q61 0.13 035 0.06 0*7 034 Udbn 0.78 085 0.42 06 0.17 0450 27 0.27 A45 NM of Students EntIh -b V0calca ThIbnl (00 814.5 406 31.7 94 148S1 853g6 397.7 10613 141S4 (Yew (6 (198) (1984 (19 1968) 61) (198O) (1986) A% E'Poputma ot Wettin Apg 3.06 324 086 0.54 13 20 0.1 1.14 L71 SaMIe Ia "Eaiiln Jadomzra Suca in Demlapin Cc.nldw ID S LAUl VA J( bmj (aft.) Cetj i D edom L=_ 1AID*O illan fb Od8S datB lkm Wod4.Doagen RaRo 1988; 1.1eC, "a Path 1989 and Gout d Rofqwb aor Stat Yaltbook of Remblo c aa 19C T_bLe oesume of Rqtawbic fCan,iity of eAEduebnof of, _ f bi Gmemir udse aSd aiglsug Udac safdslua iaue .ntbeadw and aisute eemen mdldn e*agined aidtlzceut lzde, av i scl n onmwokhv oy d.y Yf Phtw odmr ote_ ; ad_ 'qw nne ( * I sg2s f l Sttt , it I IX8 J f el-s§fiiP 8 1 -24- V. CONCLUSIONS 5.01 Thi papers major contention is that a combination of limited tedhological involvement by domestic producets, regulatoq and policy rtriton on both embodied and dismbodied forus of technology Imports, and weak nstitutional support to industrial frms, has increased Brazlian frms' distance from the price-performance frtier. in addition, major gaps in the educational system, paticularly Iow enrollment levels in secondary school and in science and engineering, compromise the supply of technical labor force and the acquisition of technological capabilities in the future. 5.02 As in most discussions of this knd, thee propositions are subject to a number of caveats. Firs, the suggestion that relatively few frms are actively engaBed In technology dvelopment omits the fact that not long ago, say in the ealy 1970s, rfewer tan thepresent core of 350 R&D-active flrms (with a contour of 1,200 producers) were technologically activ. In many cases, imprve technological perfomance came as a result of competitve pessures fkom the intntiona rwet or national firms, particularty those in capital goods and eecronl, suppt o genme tecology polcs and isttudons was criticaL 5.03 Tis paper suggests that much of the underlying motvatioa fot the Brazin Govnment's technology poLies was not the imprvement of the coutWs technical base but the more short-sighted consideration of saving or eauning foreig excage. A seond caveat is now in order. Although that geal statement is on the whole correct, the Govment's technology policies also improved the barining position of local fims in negotiating arms-length technology transfr deals In addition, they promoted eatry of domestic firms in areas that normally would be precluded through patent protecion or the eerise of overwhelming market power (pharmactica and electonics), while stimulating the creation of a fairly sophiscated capital goods sector. Moreover, such policies effectively steered multinaonal frms that tradtionally had been domestic market-oriented towad exporti thereby forcing the tecnolgical upgrading of their domestic operations. Finally, the poLies aeated for the first time a fnanng mechanism for the technoloSg needs of Brazilian industrial and engineering consulting fims 5.04 What the paper is implicitly arguig, however, is that these govenmnent polides were carried too far and became outdated. Restricdve ams-length technology transfer policies, for example, asmmed the prence of relativel unsophistcated domestic producs and the considerable eagerness of foreig supplies to market their latest technological wares. Nearly two decades after these policies were flrst implemeted, te assumptins are much less justifiable Similarly, capacity creation in sophisticated industial segments, such as capital goods and electronics, was regarded as double benefits for eomic development it alowed for import substitution in areas of beavy foreign exchange outflow wble creating capabilities in strategic segments for the formation of skills and diffusion of knowledge. But aoss-the-board import substitution led to ewesver diversification, fagmentation of efforts within fims, and substantial waste in gvrnmet- supported technological pursuits. 5.05 A thrd and finacseat conems the institutional support system for technical advance in industry. his paper argued that despite a number of important eptdons (such as EMBRAPA, CIA and IT), the public R&D network is not vy responsive to the producdve needs of the economy. Moreover, oveapping and fagmentay efforts dissipate sare resourCeS. What was left unsaid, howevr, is that the faults of the institutional network are not unique to Brazil nor to industrializing countries, for that matter. A more demanding industril sector, chllenged by competitive markts and foused on fewer product areas, would be a far beter user of existng Instutional resources (as the experenc of Metal Leve and Embraer show) and an efctive force for reform. -25- AMt TABLt 1: MMSITS TV MUW;T SHAMRES FOR LRAZI'8 M CU REPORTS, 1970-1287 MARET SHARES RATIOS SITC COMMDITIES 1070 1980 1887 80t70 87/80 512 ORANIC CHUIICALS 0.003 0.009 0.012 3.0 1.3 513 1NMt0ElEnZTS,OXIOZS,ETC 0.000 0.004 0.008 9.3 1.8 514 0TII INORGAIC CHEMICALS 0.001 0.002 0.004 1.6 1.8 515 RAIOACTI ETC MATERIAL 0.000 0.000 0.000 0.0 37.1 521 CAL,PETROLEUM ETC CUM 0.000 0.000 0.004 5.1 88.1 531 SYNT DYE.ART rNDGOSLAXES 0.000 0.001 0.001 4.2 0.7 532 DYES NESTAINO PR ODS 0.040 0.041 0.052 1.0 1.3 533 P'IGENTS.PAINTS.ETC 0.000 0.002 0.002 11.1 0.9 541 MEDICINAL ESC PRODUCTS 0.002 0.003 0.003 1.6 1.1 551 ES8ENTL OIL,PERFUME,ETC 0.023 0.018 0.014 0.8 0.8 553 PERFUMR.COSoETICS.: 0.001 0.005 0.003 4.8 0.8 554 SOAtCLEANG ETC PEEP 0.000 0.006 0.002 18.1 0.4 561 PFTILIRS AMANWACTURED 0.000 0.000 0.001 31.9 4.2 571 PLOsmiES,PYRogcs ID 0.002 0.041 0.168 23.8 4.1 581 PLNSTC MA.TMLIALS ESC 0.000 0.003 0.007 20.4 a.4 599 CBDIICALS 3es 0.002 0.008 0.008 3.4 1.1 611 LEAMTER 0.021 0.029 0.032 1.4 1.1 812 LTATM IETC MAIUPACTSs 0.005 0.025 0.044 5.1 1.8 613 FMF1 SKINS TmIUEI).SHESS3 0.000 0.001 0.002 55.4 2.4 621 MATERIALS OF RUBHER 0.000 0.003 0.003 18.9 0.9 629 RUBBIR ARTICLES DE8 0.003 0.011 0.016 4.1 1.4 631 VENEJS=lr* DhoDwlTC 0.025 0.023 0.022 0.9 1.0 632 WMO MANUFACt.RES 3ES 0.003 0.012 0.010 4.7 0.8 633 CMRK MANUFACTURES 0.000 0.001 0.001 18.4 0.7 641 PAPER AND PAPVRBOARD 0.000 0.006 0.009 78.5 1.5 642 ARTICLES OF PAPER ETC 0.000 0.004 0.006 16.1 1.4 851 T8XTILE YARN AND THREAD 0.004 0.020 0.018 5.1 0.9 652 COTTON FAHRICS,IWVER 0.008 0.016 0.015 2.6 1.0 653 WCWE9 TEXTILES NORCOTTON 0.001 0.004 0.002 5.3 0.8 654 LACE.r ZBO.TULLEt.ETC 0.004 0.003 0.003 0.7 0.9 655 SPECIAL TKTILE ETC PROD 0.003 0.020 0.008 6.0 0.4 856 TXTILE ETC PRODUCTS 3ES 0.003 0.028 0.034 8.1 1.2 657 FLOOR COVR,TAPESTRY ETC 0.000 0.001 0.001 5.2 1.2 661 CMENT ETC BUILDING PROD 0.000 0.006 0.004 17.3 0.7 662 CLAY,RZRACTCRT BLDG PED 0.002 0.012 0.014 6.8 1.2 663 OTH DOhMETAL MINERAL MFS 0.001 0.007 0.005 7.5 0.7 664 GLASS 0.010 0.005 0.006 0.5 1.0 665 GLASSWASE 0.000 0.004 0.006 14.6 1.6 s66 POTTERY 0.001 0.007 0.009 7.0 1.2 667 PEARL.PBEC-.541-P STONE 0.004 0.002 0.005 0.6 2.1 671 PIG IRON ETC 0.026 0.073 0.145 2.6 2.0 672 IRK.STL PRIMRT F5RHB 0.015 0.009 0.064 0.6 7.5 873 IRoN AD STEEL SHAPES 0.009 0.007 0.023 0.7 3.4 674 IRNS1TL UNlV,PLAT5,SRZZT 0.004 0.011 0.022 2.9 1.9 675 IRON,8TREL EOOP.STR1P 0.000 0.004 0.004 34.8 1.0 676 RAInM! RAILS ETC IRN,STL 0.000 0.005 0.000 25.0 0.0 677 IRR,STL WIE UCL W ROD 0.002 0.005 0.007 2.8 1.5 678 IRON,ss TUESs,PIPEs,ETC 0.001 0.010 0.007 8.3 0.7 679 IR.STL CASTINGS UNlRD . 0.002 0.005 . 2.8 691 STRUCTURES AND PARTS NMM 0.000 0.002 0.004 14.9 1.5 692 METAL TA8 S,BOXES,ETC 0.001 0.009 0.007 10.0 0.8 693 WMRE PRODUCTS N ELECTS 0.000 0.009 0.012 19.4 1.3 694 STL,COPPR MAILS.NUTS8ETC 0.000 0.003 0.003 10.9 1.1 695 TOOLS 0.004 0.007 0.005 1.6 0.7 696 CUTLERY 0.006 0.020 0.018 3.6 0.9 697 BASE MTL BDtEEOLD EQUIP 0.001 0.012 0.007 15.2 0.6 698 HIEAL MANUFACTURES RES 0.000 0.003 0.003 7.8 1.2 711 PO4ER MACHINERY NON-ELEC 0.000 0.013 0.014 34.5 1.1 712 AMICULTURAL MACDINERY 0.001 0.016 0.016 23.2 1.0 714 OFFICE MACKINES 0.007 0.011 0.002 1.6 0.2 715 HETALRKII FMACSINERY 0.002 0.006 0.003 3.1 0.5 717 TEXTILELEATHER MACUHRY 0.002 0.005 0.005 3.2 1.0 718 MACHS FOR SFCL INDUSTRYS 0.002 0.007 0.005 2.7 0.8 719 MACHINES DsE ONELECTRIC 0.Ol 0.005 0.005 5.7 0.9 722 ELEC PAR MACH,SWITCHGEAR 0.00 0.005 0.003 5.1 0.7 723 ELECTR DISTRIBUSTI GMACM 0.001 0.002 0.006 2.2 2.3 724 TLWICATIONSEQUIP 0.001 0.005 0.010 4.7 2.0 725 DQD3STIC ELECTRIC EQUIP 0.001 0.004 0.005 2.7 1.5 726 8LCR-MKDCL,ZRAT EQUIP 0.001 0.001 0.000 1.7 0.3 729 ELECTRICAL AC ES 0.001 0.005 0.003 3.7 0.5 731 RAILWAY VEIIICLES 0.000 0.016 0.004 47.4 0.3 732 ROAD MDTOR VEHICLES 0.000 0.008 0.009 18.3 1.1 733 ROAD VEHICLES NM-NOTOR 0.000 0.007 0.001 48.0 0.2 734 AIRCRAFT 0.000 0.003 0.011 40.5 3.3 735 SHIPS AIND DCTS 0.001 0.007 0.010 8.2 1.4 -26- ,me Tabl 1: (Cont4.) MARKE SIEARS RATIOS 8iTC CXW2linE 1970 1980 1987 80/70 87/80 812 LMMI6,NKATNG,LGBEM EQU 0.000 0.004 0.003 9.0 0.7 821 FM TItME 0.001 0.003 0.002 1.9 0.9 831 TRAVEL 00OU6,IAXnBDIi 0.000 0.013 0.012 41.9 1.0 641 CLOTIZEG NOT f FMUR 0.001 0.003 0.003 6.0 1.0 842 FUR ESC CLOTHES,IOD 0.000 0.017 0.004 299.8 0.2 851 POOtYXnNR 0.005 0.033 0.057 6.7 1.7 861 INE M*APPARATS 0.000 0.002 0.002 11.5 0.8 662 PWTO,CIUDI2 SUPPLZRS 0.001 0.007 0.010 6.7 1.4 863 DEVLOXPED CM rFIum 0.000 0.001 0.000 2.3 0.1 884 N$1U38 AIID ClCES 0.000 0.001 0.001 15.1 1.5 86 2 OUNf EVMR8.PMR:S 0.001 0.007 0.000 1.8 0.1 892 MIRTMnn MATTR 0.001 0.003 0.002 2.1 0.6 893 ARTICLES OF PLASTIC 15S 0.000 0.004 0.002 22.4 0.4 694 oTr8,s'spToo3 3G0S.ETC 0.001 .00G 0.003 5.7 0.s 895 OCFICR SUPPLiES mm 0.000 0.e02 0.005 13.0 2.1 $86 NORKS OF ARTETC 0.000 0.000D 0.000 0.7 0.1 897 am,U ,8UR WuA,JWB5LRY 0.004 0.002 0.003 0.6 1.3 899 OR MWFMACTURED ODODS 0.001 0.005 0.005 7.9 0.$ 0.002 0.007 0.008 3.5 1.1 1, -27- IBlBIOGRAPHY Braga, IL and Mats, V, D mpeuho Tecdoglco da Industral Bslra uma Anjise Exploratorla,' Tatos Nra Di ntna no. 162, JN%PESJPEA, Februay 1989. Cardozo, A.C, "A Implantago de Les e Regulamentos Sobre Thsferacla de Teologla: A Experileda do Bri t m 198& CNPq, PRecursos do Tesouro da Uniso: Dotacao Indal, Dotacso Final, e Despesa Realzada 1989 May, 1989. Coakdasu Nadonal da Industrais RIO, 1990. FPIsctk C IS WIi*tloa, Techic Che and Compeldvo- to the b Br a Ectas bndusuy,' kUM §WAj!jWNo- 15, Mm. World flAak W"usr and Enexg Department, 1969 Go nmet of Republic of dhina, M1ty of Edcation, St s o ippb ot 1984 oenumet of Republc of Chdna, Statistical Yerook of Republic of Cina 19 Tawa Kume, IL 'A Poitca Taria Brasleia no Perodo 1985: Avalaco o Reforma,' mIm, 198& Lul. S. plaing Industial Sucem in Deeloping CoDUni, i S Lal VN. N.(e) Crrent Issues in Deelomt Eonomi London, Macmailln (ftoming). Martins, 0. and Queiroz , '0 Perfl do Pesquad Brasilei Revista lelm de TecnoJji VoL 18, No. 6, September 1987. Nunes, 1, 'As Formas Ilidoas do P&D na Industrh Brasilea, Revisa Brasilefra do Tecnoloeba VoL 16, No. 2, 198S. PaulUny, E, 'Empresa Nacdonal-Panorama do Setor Enmprsl em 1983,' Revista BRasiera de Tcnmlqgi' VoL 15, n.3, 1984. Roseau, K, Regulation of Foig Investment In Brazi, , 1988 Schwaan, S, Brazik Oppotnity and Cris in Higher Education,' Ihe Educaton 17, 1988, pp. 99-119. Tavar de Araujo, 1., Haguenawur, L and Mabad o, J.L, 'otesao, Compdtltdvade e Dsempeh E 1xwro da Economia Baseleim nos Anos 80,' Revsta Braei do Comedo ri ano V, No.26, NovembroiDeembro 1989. Tavares do Araujo, J. 'Mudanca Tecnoga e Competitdade das E 1prace B ieras de Manufatuados, IEIFEAUPJ, Teo para DI o 1982, Table m UNESOO, Stasical Yerbook 1988 Pars, 1989. World De pmt Ren 1990. -28- RMUSTRY SERIE PAPERS No. 1 Japanese Direct Foreign Investment: Patterns and Implications for Developing Counties, February 1989. No. 2 Emerging Patterns of International Competition in Selected Industrial Product Groups, February 1989. No. 3 Cbanging Firm Boundaries: Analysis of Technoloy-Sharing Alliances, Febray -1989. No, 4 Technological Advance and Organizational Invation in the Engineering Industry, March 1989. Nlo. 5 Eprt Catalyst in Low-Income Countries, November 1989. No. 6 Overview of Japanese Industrial Technology Development, March 1989. No. 7 Reform of Ownership and Control Mechanisms in Hungaiy and China, April 1989. No. 8 The Computer Industry in Industrialized Economies: Lessons for the Newly Industiializing, February 1989. No. 9 Institutions and Dynamic Comparative Advantage Electronics Industry in South Korea and Taiwan, June 1989. No. 10 New Environments for Intellectual Property, June 1989. No. 11 Managing Entry Into International Markets: Lessons From the East Asian Experience, June 1989. No. 12 Impact of Technologiqal Change on Industrial Prospects for the LDCs, June 1989. No. 13 The Protection of Intellectual Property Rights and Industrial Technology Development in Brazil, Septembe 1989. No. 14 Regional Integration and Economic Development, November 1989. No. 15 Specialization, Technical Change and Competitiveness in the Brazilian Electronics Industry, November 1989. -29- RNDusrYX sI PArERS c.at'd No. 16 Small Trading Companies and a Successful Export Response: Lessons From Hong Kong, December 1989. No.i 17 Flowers: Global Subsector Study, December 1989. No. 18 The Shrimp Industy: Global Subsector Study, December 1989. No. 19 Garments: Global Subsector Study,, December 1989. No. 20 World Bank Lnding for Small and Medium Enterprises Fifteen Years of Experience, December 1989. No. 21 Reputation in Mam*ctured'Ooods Trade, December 1989. No. 22 Foreign Direct Ivestment From the Newly Industialized Economies, December 1989. No. 23 Buyer-Seller Links for Export Development, March 1990. No. 24 Technology Strategy & Policy for Industrial Competitiveness: A Cae Study of Thaila4d, February 1990. No. 25 Investment, Productivity and Comparative Advantage, April 1990. No. 26 Cost Reduction, Product Development and the Real Exchange Rate, April 1990. No. 27 Overcoming Policy Endogeneity: Strategic Role for Domestic Competition in Industrial Poliey Reform, Aprfl 1990. No. 28 Conditionality in Adjustment Lending FY8O-89. The ALCID Database, May 1990. No. 29 International Competitiveness: Determinants and Indicators, March 1990. No. 30 FY89 Sector Review Industry, Trade and Fmance, November 1989. No. 31 The Design of Adjustment Lending for Industry: Review of Cufrent Pracice, June 1990. -30- IDUSIRY SERIES PAPRS eAn'd No. 32 National Systems Supporting Technical Advance in Industryr. The Brazilan Experience, June 26, 1990. No. 33 Ghana's SnmUll Enterprise Sector: Survey of Adjustment Response and Constraints, June 199Q. No. 34 Footwear. Global Subsector Study, June 1990.- No.- 35 Tightening the Soft Budget Consraint in Rdming Socilist Economies, May 1990. No. 36 FEee TiAe Zones in Export Strategies, June 1990. No. 37 - Electronics Development Strategy: The Role of Government, June 1990 No. 38 Export Finance m the Philippines: Opportunities and Constraints -for Developing Country Suppliers, June 1990. No. 39 The US. Automotive Aftermarket: Opportunities and Constraints for Developing Country Suppliers, June 1990 - For extra copies of these papers please contact Miss Wendy Young on xtension 33618, Room S-4101 ~~~~~~~~~~~~~~~~~_I ,,- iii ENIERGY SERtIES PAPERS No. 1 Energy Issues in the Developing World, Februay 1988. No. 2 Review of World Bank Lnding for Electric Power, March 1988. No. 3 Some Considerations in Collecting Data on Household Energy Consumption, March 1988. No. 4 Improving Power System Efficiency in the Developing Countries through Performace Contracting, May 198$. No. 5 Impact of LowerOil Prices on Renewable Energy Technologies, May 1988. No. 6 A Compareson of Lamps for Domestic Lighting in Developing Countries, June 198& No. 7 Recent World Bank Activities in Energy (Revised October 1989). No. 8 A Visual Overview of the World Oil Markets, July 1988. No. 9 Current International Gas Trades and Pnces, November 1988. No. 10 Promoting Investment for Natural Gas Exploration and Production in Developing Countries, January 1988.- No. 11 Technology Survey Report on Electric Power Systems, February 1989. No. 12 Recent Developments in the US. Power Sector and Their Relevance for the lDeveloping Countnes, February 1989. No. 13 -Domestic Energy Pricing Policies, April 1989. No. 14 Financing of the Energy Sector in Developing Countries, April 1989. No. 15 The Future Role of Hydropower in-Developing Countries, April 1989. No. 16 Fuelwood Stumpage: Considerations for Developing Country Energy Planning, June 1989. . No. 17 Incorporating Risk and Uncertainty in Power System Planning, June 1989. No. 18 Review and Evaluation of Histotic Electricity Fecasting Experience, (1960. 1985), June 1989. ) .~ . -32- ENERGY SERIES PAPERS cont'd No. 19 Woodfuel Supply and Environmental Management, July 1989. No. 20 The Malawi Charcoal Project - Experience and Lessons, January 1990. No. 21 Capital Expenditures for Electric Power in the Developing Countries in the 1990s, February, 1990. No. 22 A Review of Regulation of the Power Sectors in Developing Countries, February 1990. No. 23 Summary Data Sheets of 1987 Power and Commercial Energy Statistics for 100 Developing Countries, March 1990. No. 24 A Review of the Treatment of Environmental Aspects of Bank Energy Projects, March 1990. No. 25 The Status of Liquified Natural Gas Worldwide, March 1990 No. 26 Population Growth, Wood Fuels, and Resource Problems in Sub-Saharan Africa, March 1990. No. 27 The Status of Nuclear Power Technology - An Update, April 1990. No. 28 Decommissioning of Nuclear Power Facilities, April 1990. Note: For extra copies of these papers please call Ms. Mary Fernandez on extension 33637.