Latin America

Latin America

World Development, Vol 12, No 5/6, pp 645-660, 1984 Printed m Great Britain 0305-750X/84 $3.00 + 0.00 Pergamon Press Ltd. Latin America* SIMON TEIT...

1MB Sizes 3 Downloads 102 Views

World Development, Vol 12, No 5/6, pp 645-660, 1984 Printed m Great Britain

0305-750X/84 $3.00 + 0.00 Pergamon Press Ltd.

Latin America* SIMON


Inter-American Development Bank and the Catholic University of America, Washington, D. C. and FRANCISCO



Consultant, INTAL -- Inter-American Development Bank, Buenos Aires Summary. - - Developing countries are expected to ~mport but not to export technology However, the 1970s have witnessed the beginnings of a small but growing flow of technology exports from some of these countnes, particularly the larger and relatwely more mdustrmhzed Latin American countries, as well as others m Asm On the basts of avadable mformat~on from country studies for Argentina, Brazil and Mexico, and prehmmary reports on the Andean sub-region and on technology purchasers, the paper discusses 0) what is being exported, i e types of goods and services as well as export modahtles or export channels; (n) where these exports go, i e distinguishing between mdustrmhzed and developing country markets, (m) who the exporting agents are, i e. dlstmgmshmg between pubhc, private, multinational and local exporters, 0v) restitutions and pOllOes, i e promotional functions of the government, pohcles directly and indirectly related to technology exports, (v) prehmmary evaluatmn and prospects, including considerations on technological development, protection, and comparative advantage The paper fimshes with some prehmmary conclusmns

1. A NEW LATIN A M E R I C A N T R A D E PHENOMENON According to conventional wisdom, developmg countries are expected to import technology, not to export it In the 1970s, however, an incipient but growing flow of technology exports began to emerge from several semi-industrialized countries In Latin America, the group of technology exporting countries consists primarily of those with the largest and relatively more industrialized economies.1 These technology exports included disembodied technology, supplied through hcences and technical assistance contracts, as well as industrial and civil construction projects, various consulting and engineering services (with or without the supply of equipment), sales of turnkey plants and direct foreign investments. An examination of the experience of the countries for which information is avadable2 seems to show an evolutionary sequence, going from import-substituting industrialization to an lncreasmg flow of manufacturing exports, recipe-

ent direct foreign investments m nelghbounng countnes, and, finally, sales abroad of technical information and technology services. Sectoral circumstances related to particular emphases m policy and differences m natural resource endowment introduce a rich range of variations around that main thrust. Notwithstanding sharp contrasts m terms of policy, organizational environ-


*This paper is based on field work and other results of a larger, coordinated research project on technology exports from seml-mdustnahzed countries jointly sponsored by the Inter-American Development Bank and The World Bank A previous version of this paper was presented at the Thxrd Latin American Regxonal Meeting of the Econometric Sooety, Mexico, 19-22 July 1982, Mexico City A somewhat different presentation of stmdar materml was included m InterAmerican Development Bank, Economic and Soclal Progress Report m Latm Amerwa (Washington, D C 1982), Chap 6. The wewpomts presented m this paper are those of the authors and do not purport to represent the officml position of the Inter-American Development Bank



ment, and resource endowment, semiindustrialized countries in Asia appear to have followed a broadly similar pattern This paper has five sections In Section 2, the nature of the technology exports is analysed by country, in Section 3, the destination of the exports is rewewed; Section 4 discusses the agents involved, Section 5 presents preliminary conclusions and an evaluation.

2 N A T U R E OF T H E T E C H N O L O G Y EXPORTS (a) Argentma ~ Although


involvement in trade [ able ]


lel ]lll(,qOl,q'

Argl_ nllnd


and investment in foreign markets dates from an earlier period, 4 that related to technology exports is largely recent in origin It is also relatively spontaneous, it is not, in other words, mainly the result of government promotional policies s The p h e n o m e n o n is difficult to quantify Neither m o v e m e n t s ot capital nor contractual arrangements are always registered, because there are no requirements to that effect (except when tax incentives have been granted) In addition, technology exports from Argentina are still hmlted in volume and have been largely concentrated in Latin America Table 1 presents data on foreign sales of tnfra,structure projects and turnkey manufacturmg plants and other industrial facdmes amounting to approximately us $696 million and US $106 mllhon, respectively, for the Illdlt alo# ~ Of HIIIOlI,'IIS CtlJ+.lcd


[nlra~tructurc projects

(197g 82) No 32 + A m o t l n l 696 At inllhon Ater
( 196¢~-81) No 147~ A n l o u l l l $42S~ q illllllt)ll Avcr,tgc dlllOUil[ $73 t) million Riillgc $2 4 $12li0 llulllon

( I 9"~4-S2) No gg A i l l o u n t $9S4 nulhoil A%clilgc dllllitlnl $2() (I Illllhon Rdngc n d

]ndu~lilal proJCcl~ prolculs

( 1973 ~2) No Illg * A l n o t l n [ $1D6 4 million Avcr
(OllMJiilllg ,llld technical dkM%|dllLl_

(196t)-bi2) No 79' A m o u n t $22 inilhoil A~er
( Iq7~ 81 ) No 49" A l n o t l n l $2S4 2 nlllhon A,,er,igc
( 1979 Sl ) No h ~Xmotini $]t') 7 mllhol/ A'.Cl,igc
l)lrcLl |orcigil III~CI~IlIICIli

(1965 mid 81) loldl No 122"" A m o u n l $87 1 rmlhon M,inufaciurmg No 61 A l l l o t i n i $4q 4 lllilholl

(197f~-81) 1 old] No 210 II' A m o u n l SAt(R) mflhon MdrltlIdl_lUl'lng No ¢,9 Aniotlnt $211 mlHlOll

(1977 gO)

] oi
",c,u#~e~ Age entma %ntcr (19S ~;) Kalz and ko~,tcoll (198~), I'ba=tl %-ILovldl (1982a), ,%h't;~o I) d~

period 1973-82. They include 32 cwtl engineering works and 108 industrial projects. The average value of the cwil construction works was substantially higher than that of the industrial plants (a phenomenon which, as we shall see, wdl be repeated in the cases of Brazd and Mexico). Most of the industrial plants exported belong to the food and chemical industries. Direct foreign mvestrnent has been rather limited and &rected for the most part towards nelghbourmg Latin American countries The high proportion, - - more than 50% - - geared to manufacturing industries Is noteworthy It is hkely that a substantial share of the 61 investments m manufacturing industries involves the deployment abroad of skdls, reformation and experience prewously acquired by the investing company at home. Such is clearly the case of compames in the textile, food, agricultural machinery, pharmaceuticals, automobile parts and steel industries. Direct foreign investments m the oil sector, accounting for about 30% of the total amount through four projects can also be regarded as revolving technology transfers The remainder includes investment abroad in construction (8%, 20 projects); agriculture (3 8%, 5 projects); trade (3.2%, 20 projects), transportation (1.2%, 6 projects) and serwces (1.4%, 4 projects).


Foremost among the exports of mfrastructure prolects (Table 2) are turnkey telecommumcatlons, hospital and airport facilities, pumping stations, and oil and gas terminals Among the mdustrtal prolects, the large number of food and pharmaceuttcal plants, as well as projects such as an antibiotics plant and a radioisotope plant (exported by the National Atomic Energy Commission to Peru), warrant specml mention. 6 In the field of techmcal asststance, engmeermg and consultancy contracts, the acttwty appears to be rather more hm~ted. A total amount of US $22 mflhon through 79 projects is reported (m this case the amount refers to remuneration for services rendered) If eight bmatlonal projects between Argentina and ne~ghbourmg countries were to be included the figure would amount to US $60 mflhon and 87 projects, respectively; 24 of the contracts were &rectly related to mdustrml projects.

(b) Brazil 7 Brazil's presence abroad grew substantmlly during the 1970s In addition to the growth of its commo&ties and manufacturing exports, this increase took the form of more &rect foreign

Table 2 Examples of te¢hnologtes and prolects exported ltcm Infrastructure


Industrial projects


Argentina Telephone exchange Turnkey airport Off p~pehne and filhng stations Llqmd gas terminals Turnkey hospital Port warehouses

Meat packmg plant Bakery Instant soluble coffee plant Plants for the processsmg of soybeans, otrus frmt, milk Sulphunc aod, cement blocks, pharmaceutical products plants Antibiotics and pestlodes plants Planning ot transport system Urban and regional development feastbdJty studlcs Water supply systems Planning of cattle production

Brazd ttydroeleetnc plant and transmission hncs Hotcl. watcr supply systcm and roads

Residential centre Off plpehne and storage facilities Port repairs Airport expansion Sugar mills and ethanol (from bn)mass) distilleries Charcoal and coke-based stccl Pulp and paper, soft drinks construction Flldterldls Bicycle, manioc, household appliance plants Planning and supervision of telecommunication systcms Pre-mveslmenI studies Mineral resource development Managemcnt of transport systems

Mexico Hydroelectric plant Water supply project Off pipelines, ports (plcrs) lrngatLon system Roads, bridges, ,urport

Carbon-black, and bagassc paper plants Direct-reduction steel plants Glass and bottle-making plant Non-wovcn fibre process Furfural plant Design of lead oxldc and fertdlzer plants Irrigation and drainage studies

Land registration surveys appraisals, aerhd photography Admlnlstrdtlve and dccounlmg proccdurcs Aqueduct and ~satcr supply, studies and designs

Geological prospecting Source Prepared by the authors based on data from sources given m Table 1



investment, exports of industrial projects, and participation in international tenders for largescale civil construction works Direct foretgn mvestment appears to be concentrated not in the manufacturing but In the oil sector and, to a lesser extent, in commerce, finance, and so on. Investment in manufacturing industries represented only about 5% of the total In this respect, it differs strikingly from Argentina Furthermore, it is interesting to note that, as in the cases of Argentina and Mexico, technology exports contribute much larger earnrags than those derived from direct foreign investment (although strong caveats should be borne in mind regarding this comparison) ~ According to Central Bank of Brazil data, in the period 1971-78 the cumulative value of earnings from foreign sales of technology was US $926 million, whereas income from direct foreign investment amounted to only about $125 mllhon In this period, foreign investment to the amount of $400 million had been authorized. 9 In number, the bulk of the technology export contracts belongs to the period 1976-81 (77%) and the remainder (23%) represents the cumulative total up to 1975 The concentration in the recent period is even greater in the case of exports of industrial technologies (88%) and of construction projects (86%) In contrast, it appears that a fairly large proportion of the consultancy contracts were concluded in or before 1976 10 As to the amount of the contracts, data are not complete The sample for which data on amounts are available includes 148 out of 308 contracts (48% of the total) The proportions are as follows: (l) industrial technology, 21 out of 49 (43%), In) construction, 58 out of 147 (39%), and (ni) consultancy and technical assistance, 69 out of 112 (62%). Thus, the sample is dominated by consultancy and technical assistance projects, whereas construction projects and industrial technologies are relat,vely under-represented

As can be seen In Table l, the average value of construction contracts is very much larger than that of industrial technology or consultancy and technical assistance contracts The mean values are infrastructure contracts, $73 9 million, manufacturing plants, $13.6 milhon, and consultancy and technical assistance, $4.5 million The value of a civil engineering contract averages approximately five times that of an industrial technology contract and 16 times that of a consultancy and technical assistance contract However, it should be borne m mind that the proportion of value added in the infrastructure projects will usually be less than in the other exports If we take the total amount of $4926 million for the 148 contracts for which data are available, and, since most are for the period 1977-79, we compare them with the Central Bank data on the foreign exchange earnings from these exports (which for those years totalled $672 million), we can see that approximately 14% of the amount of the contracts returned to Brazil as foreign exchange earnings Obviously, these data are greatly influenced by the construction project contracts which, while only accounting for 39% of the total number of contracts, represented 87% of their total value The 145 cases of reported exports of mdustrtal technology include (Table 2) plants and processes for the production of ethanol, steel, food products, machinery and equipment, chemical products, paper, consumer durables, etc Numerically the most important are machinery, equipment and components, 21%, sugar and alcohol, 19%, steel, 14% The proportions for the various types of mdustrtal technology exports, are shown in Table 3 Engineering, consulting and technical assistance contracts predominate with 52% and are followed by sales of turnkey plants with 34% of the total In approximately 18% of the 49 cases

Table 3 Braztl -- type~ of mdu3trtal technology exports

Engineering, consulting and techmcal assistance* Turnkey plants L~cences Total

Number of cases

Proportion (%)

76 49 20 145

52 34 14 100

Source See Table 1 *Includes special cases hke contracts revolving reciprocal transfers or international technology marketing agreements

LATIN AMERICA of sales of turnkey plants, technology exports are coupled with direct foreign investment. The 147 cases of mfrastructure pro]ects included show that hydroelectric projects and electricity transmission lines represent approximately 27% of the cases; roads and related works, 16%, urban construction, 14%; water systems, 10%, ports, 6%. The remainder includes communication, railways, airport and eight other projects. With reference to contract values, the concentration of construction projects is very great since one single contract accounts for 29%; three for 55%; and five for 63% of the total. The largest construction contracts are related to projects located in countries that export oil to Brazd. I1

(c) Mextco 12 Because of the lack of centralization of Mexican information, data by type of activity are not readily available. One of the reasons IS that, unlike other countries that supervise technology contracts, tenders or direct foreign investments, Mexico, with no tradition of exchange controls, has not deemed it necessary to establish central registries for this information. I3 The information available Is thus incomplete and is essentially based on: 0) data from two consortia devoted to the export of technology and technical services; (ii) the records of a tax incentive for these exports; and (in) secondary sources, such as newspapers and specialized }ournals. The range of Mexican industrial technology exports is extensive (see Table 2), and includes technologies for the manufacture of steel, glass, paper, non-woven fabrics, miscellaneous chemical products, etc. Companies in the following industries are known to have foreign


investments abroad: mining, pharmaceuticals, cereal processing, engineering, automobiles, fertdizer, glass and oll equipment. Unfortunately, data indicating the amounts of these investments are not available By taking into account the average reimbursement of taxes it is possible to obtain the lowest estimated value of exports according to fiscal incentive records, This is shown in Table 4 for 1977-81. The values turn out to be well below those for Brazil, but are roughly comparable in value with the data for Argentina. The total includes approximately $778 million in construction contracts exported by Construmexico, an export consortium, and $43 million in industrial technical services exported through Tecnlmexico, another export consortium. However, the amount shown in Table 1 is much higher: around $1052 million, including $984 million in construction contracts, $17 million m industrial projects and $51.5 million in consulting and technical assistance services The difference still has to be accounted for Infrastructure projects represent approximately 94% of the total amount exported

(d) Andean Pact countrtes Preliminary evidence from the Andean Pact countries 14 indicates that there are a good number of manufacturing and engineering companies, particularly In Colombia, Peru and Venezuela, with an experience of exporting technology resembling that of companies in Argentina, Brazil and Mexico The preliminary data 15 shown in Table 5 suggest that the group of Andean countries reviewed there together constitute the fourth largest source of supply of technology and technical services in the Latin American region.

Table 4 Mextco -- value of technology export~ esttmated according to value of tax retrnbursement~, 1977-8l (rmlhons of US $)

Construction firms Manufacturing firms Consultancy, engineering and research organlzatxons Total






139 7 07

134 7 36

120 4 33

97 7 72

98 2 630 7 15 1 30 0

12 141 6

36 181 9

24 126 1

13 106 2

Source See Table 1 *Includes $6 4 mdhon not allocated on a yearly bas~s

62 119 5


21 1 681 8*



Technology export operattom by ('olombta, Peru and Venezuela (functional and ~ectoral breakdown in number o] operations) Manufacturing industry Construction Licence agreements, techmcal assistance and engineering Complete serwces plants

Colombia Peril Venezuela Total

21 7 12 40

-2 5 7

Electric projects Others 7 -11 18

13 4 8 25

Total 41 13 36 9(t

Source Sercovlch (1982b)

A total ot 90 technology export operations arc reported in Table 5 with about equal shares accounted for by industrial technology and construcnon works The breakdown according to type of operanons Is as follows. 40 (or 44%) sales of engineering services, hcenslng agreements and technical assistance contracts; 25 (or 28%) ovll construction projects, excluding electric power, 19 (or 20%) electric power projects, and 7 (or 8%) turnkey industrial plants. The number of operanons by companies m Colombia and Venezuela is considerably higher than in Peru The export of turnkey plants, although accounting for a small share of the total, shows a broad pattern of accumulation of knowledge and experience across sectors, e.g Peru ~s present with seafood processing plants, and Venezuela with b~omass ethanol plants The sale of engineering services for industrial projects appears to be the most frequent kind of operanon

degree of Involvement in foreign tenders for civil works and infrastructure and a relatively h~gher degree of Involvement in direct foreign investment in manufacturing, 0v) technology exports from Mexico appear to be smaller in amount than those from Brazil but larger than those from Argentina, (v) there is a larger share of turnkey plants m Argentine industrial exports, which is possibly due to the incentive provided by promononal decrees for these exports (see Section 5, below), as well as to its headstart in capital goods manufacturing; (Vl) Brazilian contracts have an average value much higher than that of Argentine and Mexican contracts in all categories of transactions This is due to the weight of a few very large contracts in the Brazdlan sample


(e) Comparison

(a) Argentma

A comparison of the three major exporters of technology in the region shows that (1) the bulk of the amount of the exports of the three countries consists of ClVll engineering services for lnfrastructural construction, (n) In general, Argentina and Mexico have smaller amounts of technology exports and lower levels of participation in foreign engineering and construction contracts than does Brazil This is probably due, at least in part, to lower promotional actlwty and financial support provided by the governments of these two countries, (lU) Argentine companies display a lower

The data available (see Table 6) show that almost all exports of mdustrtal projects went to Latin American countries, and that 65% of them went to nelghbourlng countries Several of the industrial projects exported went to Cuba under special bilateral trade arrangements A very large proportion of mfrastructure prolects also went to Latin America, although the share of the region is lower if measured m value terms Although most technical assistance and consultancy prolects were also destined for Latin America, and pamcularly to nelghbourlng countries, a not negligible proportion - - 17% - - went to other developing regions

Source See Table 1 n a not available

Direct foreign investment

Technical assistance and consultancy

Industrial projects

Civil infrastructure projects


85 7 8

Latin America USA Europe

Latin America Africa, Asia and Middle East developing countries Industrialized countries

82 17


Latin America Africa, Asia and Middle East developing countries

98 2

88 10 2

n a

n a n a

99 1

65 35

Latin America Africa, Asia and Middle East developing countries

89 11

Amount (%)



71 9


80 10


79 20




Latin America Africa, Asia and Middle East developing countries Industrialized countries

Latin America Africa, Asia and Middle East developing countries Industrialized countries Latin America Africa, Asia and Middle East developing countries Industrialized countries



n a

n a n a


75 8


48 47



Destmatton of technology exports (percentages)




Table 6



45 35


73 9





Latin America Africa, Asia and Middle East developing countries lndustrlahzcd countries Unknown

Industrialized countries

Latin America Middle East developing countries

Latin America



n a

n a

n a n a

n a

n a n a




r" >



proprietary process for the production of spongeiron by direct reduction with natural gas

Out of the total number of mdustnal technology exports, 80% went to Latin America and the remainder to other developing regions and to the industrialized countries by equal shares However, the latter have a higher unit contract value which brings their share of the total value to 1 7 ° , against 8% for other developing regions and 75% for Latin American countries In the case of Clvll infrastructure prolects the situation is very different According to the number of contracts, the destination of the bulk of them was Latin America (79%), only 1% went to industrialized countries and the remaining 20% to other developing regions However, in value terms,the share of these regions is much higher due to the heavy weight of a few very large contracts in Middle East oil-exporting countries The destination of 71% of the consultancy and technical asststance contracts was Latin America, 20% went to industrialized countries and the remainder went to countries in other developing regions The itemized data available for 1979 and the first half of 1980 show that: (l) as was to be expected, Brazil has an adverse position in its foreign technology accounts with the lndustriahz,~" countries and a slight surplus with the eloping countries, including Latin America; (il) the bulk of foreign technology proceeds (93% in 1979 and 82% in the first half of 1980) came from industrialized countries, ~6 (ni) despite the deficit mentioned in item (1), Brazil has a surplus for hcences for the use of trademarks with the lndustriahzed countries, and - - although this is an item of minor importance - - a deficit with the developing countries

(c) Mexico The destination of Mexican technology exports was almost exclusively Latin America. The concentration was 100% for the export of

(d) Andean Pact countries Central America and the Caribbean are the main markets for technology exports from Andean Pact countries (Colombia, Peru and Venezuela) They account for more than 40% of the total number of operations Intra-Andean Pact exports constitute the second most ~mportant flow, accounting for 33% of the total The relatively less developed countries within the Andean sub-region, that is, Bolivia and Ecuador, are the main recipients of these exports. The remainder goes to other developing countries, in Latin America, and elsewhere, and to industrial countries

(e) Comparison Most of the trade in technology ~s lntraregional However, In the case of industrial projects, Brazil and Mexico account for substantial exports to the industrialized countries, and Argentina and Brazil for a considerable number of civil projects to extraregional developing countries It is worth noting that in the Central America and Caribbean areas, technology supphes from Andean Pact countries are often an alternative to those from other relatively more developed countries in the region As may be seen in Table 7, the destinations of trademark registrations from Argentina, Brazil and Mexico also reflect the predominance of Latin America as the market for this type of export, followed by the industrialized countries and then by other developing countries The proportion decreases from Argentina to Mexico, while the proportion of the other countries, both lndustrlahzed and developing, increases.

infrastructure projects Of the industrial project exports, the Latin American share was 73% whereas the lndustnahzed countries accounted for 18% and other developing countries, 9%. This holds true in pamcular for production technologies or the sale of industrial processes In the steel sector, technology exports went to Latin America, Iran, Indonesia and lraq The extra-regional exports were all from one Iron and steel company (Hojalata y Lamina), and involved its

4 AGENTS FOR TECHNOLOGY EXPORTS (a) Argentina Most of the recorded industrial technology export~ from Argentina are made by national compames A notable exception Is the export of telecommumcatlons eqmpment by Standard Electric, a local subsidiary of IT] ~ Sales of



Table 7 Destmatton of trade mark regtstrattons Argentma, Brazd and Mextco (196%78 averages, m percentages) Argentina Latin American countries Industrmhzed market-economy countries Other developing countries Total






15 7 100

29 8 100

33 23 100

Source Umted Nations Conference on Trade and Development (UNCTAD), 'The role of trade marks m the promotion of exports from developing countries', TD/B 6/AC 5/2 (1 October 1981), p 6

industrial plants are made for the most part by smalll or medmm-scale national companies. Many were the result of international tenders. The proportion of the value of the contracts accounted for by transnatlonal companies (TNCs) was 28% , and the number, only 3%. Among mfrastructure prolects, those exported by Techmt, a TNC affihate, were substantial. Thus, exports of TNCs represented approximately 68% of the total value of these contracts, but a rather smaller proportion (40%) of the number of contracts The agents responsible for these exports are not only manufacturing companies, some are engineering companies or engineering consortia. One private trading company is very active Sometimes more than one company is involved m an operation and subcontracting takes place. Likewise, there are two private compames (Bndas and Perez Companc) actwe xn od development work abroad, which usually part~opate m joint ventures with other international compames. Also. worthy of mention is the sale of one plant by the National Atomic Energy Commission, a government agency, which accounted for 12% of the total value of the industrial contracts

(b) Brazil Two public restitutions have played a predominant role in the recent outward thrust of the Brazilian economy: P E T R O B R A S and the Bank of Brazil. P E T R O B R A S is a pubhc enterprise active in od exploration, development and production, both within the country and abroad It has a subsidiary, I N T E R B R A S , which is devoted to international trade, including the sale of Brazdian technologies I N T E R B R A S does not hmlt Its business to oil, but is also active

in other industrial, commercial and direct investment fields. During the 1978--80 period, according to official data, it invested $148 million, most of it abroad. Its net annual income from foreign subsidiaries was $13 million in 1978; $18.1 million in 1979; and $21.3 million in 1980 I N T E R B R A S has taken part in technology exports amounting to approximately $1500 million through 17 contracts in 11 countries (It should be noted that all this activity dates from 1976, when the institution was founded, to 1980 ) I N T E R B R A S offers a package of services to potential exporters of technology including the identification of commercial opportumtles, preparation of project analyses, the furnishing of bid and performance bonds, procurement services in Brazil and abroad, and the formation of joint ventures with foreign companies to exploit markets abroad. For ~ts part, the Bank of Brazil is actively involved In many operations abroad through numerous branches and subsidiaries PETROBRAS, through its subsidiaries I N T E R B R A S and B R A S P E T R O (the latter being devoted to oil exploration and development, with total investment and income from services rendered abroad in 1982 worth $69 milhon and $14 million, respectively), and the Bank of Brazil, are Brazil's most important direct forogn investors In recent years these state enterprises - - also including Companhia de Pesqmsa de Recursos MInerais (CPRM), in the mining field - - have tried to obtain contracts in od-exporting countries, as a qutd pro quo to offset Brazil's great dependence on foreign oil. As a result of these efforts, major projects were obtained m Iraq and Venezuela, and smaller ones m Algeria, Kuwait, Nigeria, Saudl Arabia and other African countries In the private sector, large engineering and construction firms, such as Mendes Jumor and Camargo Correa, play an important role in the



exports of infrastructure projects These firms gained substantial experience m the domestic market through the ambitious pubhc investment p r o g r a m m e carried out m Brazil durmg the 1950s and 1960s When they go abroad they also pave the way for a host of much smaller, specialized Brazilian enterprises, which participate m foreign activities by subcontracting While there is some evidence of direct foreign investments in the manufacturing sector including the export of technological know-how, there have also been cases in which the sale of technology was incidental, and merely reflected the expansion abroad of an Industrial and commercial conglomerate (like the case of Brazlhan agricultural and industrial investments m Nigeria)

G o v e r n m e n t pohcles may also have indirect effects For example, it supports the training ot a large number of professional personnel employed m state enterprises (who many times set themselves up subsequently as bus,nessmen on their own account), or, more important, Jt facilitates the development of local engineering capablhttes by means of public sector procurement and tenders Such examples clearly apply to Brazil and Mexico, but the lower public investment during the 1950s and 1960s m Argentina probably accounts for the more limited impact of government actwlty m this field

(a) Argentma (c) Mexuo In the g o v e r n m e n t sector, the d~rect acttwtms of autonomous companies such as the Mexican Petroleum Institute, which ~s part of Petr61eos Mexlcanos ( P E M E X ) , w the state petroleum enterprise, merit attentmn Large private engineering companies with hundreds of engineers like Bufete Industrial and Ingenlerfa P a n a m e n c a n a of Mexico, are rarely found in other developing countries The Mexican consortia for the export of construction services ( C O N S T R U M E X I C O ) , and for the export of engineering and techmcal services ( T E C N I M E X I C O ) , referred to above, are also unusual With respect to affiliates of foreign c o m p a m e s m Mexmo, easy generalizations cannot be drawn On the one hand, they have stimulated local demand for techmcal servmes, providing experience which subsequently enabled Mexican firms to obtain contracts to work in projects in which TNCs participated On the other hand, they have not themselves been very ,icnve m technology exports, which were generally m,lde by national companies



G o v e r n m e n t s can play a direct promotional role In this field by means of such instruments as tax incentives, financing, providing risk coverage and various supporting services. Table 8 shows how this ~s done, to varying degrees, in the three major countries studied

The incentives offered in Argentina tot technology exports take the form of subsidized credits and the reimbursement of taxes on exports The inclusion ot services together with the goods to which these benefits are granted has been questioned in some cases A m o n g the promotional factors that appear to have played an ~mportant role in Argentina are Decrees Nos 2785 and 2786 of 1975, which encouraged exports of complete industrial plants and civil engineering works, and the special credit granted to Cuba m 1973 for the same purpose. The promotional measures consisted. initially, m a reimbursement eqmvalent to 40% of the value of exports of industrial plants which was subsequently reduced to 25% Jn 1976 and to 10% in 1981 Pre- and post-export | n l a n c x n g w e r c ,ll~,o p l o vlded Prc-financmg was subsl&zed with ,l 1% enterest rate Post-fll3anclng was ehmnlated 111 1976 The financing ot Argentina's technology exports was the responsibility of the Banco de la Naclon Argentina, a state lnStltutmn simdar to the Bank of Brazil H o w e v e r , a difference from Brazil has been the lack of funds for the pre-financmg of international tenders; this has become a source of problems for potential Argentine exporters. To compete in these tenders it is necessary to open letters of credit with a prior guarantee deposit and to furmsh performance bonds Specml commercial and other risks are involved The government could help stimulate exports by bearing part of these risks or creating insurance schemes to cover them As will be seen below, this is done in Brazil, where the government furnishes the necessary bid and performance bonds It should also be noted that general economic



Prefinanclng (1% subsidy) Financing (eliminated in 1976) Special credit to Cuba for purchase of complete plants


Decrees 2785 and 2786 to promote exports of complete mdusmal plants and complete civil engineering works Reimbursement of 40% of value exported, subsequently reduced to 25% (1976) and to 10% (1981)

Source See Table 1 n a Not available

Information and support services


Risk coverage

Tax incentives


INTERBRAS provides exporter, with services in identifying commercial opportunities, project analysis and selection, securing guarantees, procurement, sponsoring joint vcntures, etc

Bid bond (2% of contract) Performance bond (10% of contract) Prefinanclng and financing by Bank of Brazil

Exemption from direct and indirect taxes for exporters of turnkey plants

Exemption from profits tax


Table 8 Technology exports promotional mstrument~


Construction services export credit through BANOBRAS Manufacturing export credit by FOMEX Covers pre-exportatlon Encourages formation of consortia CONSTRUMEXICO for civil projects and TECNIMEXICO for other tcchmcal

Since 1973, reimbursement of 11% if Mexican component is 50% or more Since 1975 an addmonal 4% if export Is made by consortium Exemption from taxes on imports ot equipment Pre- and post-shipment credit guarantees








conditions and government pohcles may have had an ~mportant, though unintended, effect in this field The Argentine recession, coupled, in recent years, with a pronounced over-valuation of the peso as part of its antHnflationary policy, resulted m depressed foreign sales hand-in-hand with a contraction of domestic demand This not only reduced the posslblhtles of contributing to the adjustment process through an mcrease in the exports of goods and services, but also served to erode the good-will estabhshed in external markets

(b) Brazil Tax incentives are part of a package of benefits granted for the export of non-traditional products in general, Some of the most important incentives are (l) exemption from taxes on export earnings, (n) exemption from taxes on insurance commissions related to services rendered abroad; (Ul) exemption from direct and indirect taxes for national companies that export technology and goods, including turnkey plants: (iv) permits for the temporary export of capital goods (thus permitting their eventual sale and/or leasing) and for the free import of capital goods for temporary use These incentives are included in relatively recent decrees, most of which have only been enacted since 1975 The Bank of Brazil offers bonds for covering the risks of failure to wm a tender and to fulfill the operation spec~ficatmns of exported projects. The bid bond is usually 2% of the value of the contract and the performance bond, 10% Although no detailed statistics are maintained, data are available on technology export operations for which insurance has been granted since 1976 The guarantees furnished represented a contract amount of $3484 milhon This amount should be compared with the $4936 mllhon for the 148 contracts for which information is available and, for the most part, corresponding to the period since 1977. is Compames exporting through I N T E R B R A S automatically have access to these guarantees The Bank of Brazil also grants pre-export financing, although this facility has only been used marginally The data on export financing shows that in the period 1977-79 the Bank of Brazil granted financing of $475 million for 15 projects Of this total, 80% was for infrastructure works, in particular hydroelectric prolects In industrial technology, the Bank financed only two projects; a steel plant in Peru for $15 million, and a ceramics plant in Ghana for $17 5 million

Only a hmlted amount of resources appears to have been available for this financing (c) Mextco In 1973, the reimbursement of taxes granted to exports of manufactured products (which dates from 1971) was extended to technology exports This reimbursement is 11% of the total value exported If the Mexican component is 50% or more, if not, It is pro-rated according to the lower value Exemption from the Import tax is also granted to equipment when it IS used in a foreign project (as in Brazil). In addition,since 1975, an additional 4% reimbursement has been added and is paid whenever the export is made through an export consortium This results in a possible maximum reimbursement of 15% of the value of the project. The consortium is committed to increase the value of its technology exports by 15% a year This incentive led to the establishment of the two consortia mentioned above C O N S T R U M E X I C O , for exports of construction services, and T E C N I M E X I C O , for exports of technology and non-construction techmcal services. B A N O B R A S (National Public Works Bank) grants credits for the export of construction services F O M E X (Bank of Mexico) extends credits for manufacturing exports through a redlscountmg operation that allows Mexican banks to provide Mexican companies competitive credit terms on their exports. The credits granted cover' (i) expenditure prior to export (preparation of studies and bids, transfer of technologies and provision of services, etc ), (n) financing of the export itself, and (ni) guarantees for credits covering both pre- and post-shipment risks

6 PRELIMINARY CONCLUSIONS, EVALUATION AND OUTLOOK (a) Prehmmary conclu~gtons Examination of the data presented in this paper plus the results of interviews conducted In a number of firms in the three major exporting countries permits us to advance some preliminary conclusions (I) The data show that technology exports annually amounting to hundreds of millions of dollars have been made in recent years from these countries These exports include contracts for construction and civil works as well ab sales of industrial technology The

LATIN AMERICA value of the exports of construction contracts is much higher than that of industrial technology exports (il) The exports already represent a substanual part of the foreign technology accounts of the countries stud~ed, to which they make a positwe contribution. (m) Most of the exports, and m particular those for industrial projects, go to other countries in the region. This ~s an area m which a future increase in rater-regional trade may be expected. (iv) The government policies of openness to international trade and of providing support In finding new markets play an ~mportant role m promoting these exports, as do subsidies, financing, and insurance The existence of a network of banks abroad may also be a substantial aid (v) Also important, although Indirectly, are pohcles for promoting industrial development, the estabhshment of local consulting firms and, m general, the training of high-level techmcal personnel. (vi) As a rule, exports of construction services appear to be closely (inversely) linked to the domestic investment cycle m which the state plays an important role. These exports then appear to be a supplement to the use of the capacity of the sector to supply the domestic market (although, m some cases, the extent of their foreign involvement has allowed some companies to become largely independent of the domestic market) (vii) At the sectoral level, the promotion gwen to a specific achvlty or industry, combined with natural resources and the eventual development of adaptive knowledge that can be exported, may play an important role A sIgnffi(ant example ts Brazilian exports of plants to produce alcohol from sugar-cane (wu) At the company level, some businessmen favour a technological response to the demands of the market; others do not. Inter-firm differences in behawour are sometimes as strong or even stronger than rater-industry differences m this respect The attitude of businessmen with respect to the promotion of research and development can also play an important role, as can their attitude toward risk, since venturing into foreign markets is always more risky than merely supplying the domestic market (b) Evaluation and outlook

(I) Defmmon and nomenclature The technology exports investigated include:


the export of turnkey plants, civil and industrial construction plants, consulting services, hcences and direct foreign lnves:ment. In addition to the problem posed by the availability of data, we may note that the studies on the subject carried out m different counmes use different nomenclatures for similar exports. For example, 'mdustrml projects' may be understood as the sale of industrial plants, including the equipment and all or part of a package consisting of' design, construcuon, installation, start-up and training of the personnel of the chent The definition used In Latin American studies is usually narrower and corresponds to the concept of 'turnkey plants', which usually means that a single contractor is responsible. In India, the sale of part of the package would also be included under the headmg of 'Industrial plants'. In Korea, any sale of equipment exceeding $100,000 is called a 'plant export' These definitions are somewhat broader, and of course, include all cases of sales of turnkey plants. 19 In pnnciple, mdustrml technology exports reflect the existence of capabllmes for design, techmcal adaptation, and creation of products and processes, as well as for meeting specifications for quahty and operation, market orgamzauon, financing, etc It is usually assumed that these skdls do not exist or are only just emerging m developing countries Thus, the evidence for the existence of sophisticated technological and orgamzauonal capacity m the companies and institutions parUclpatmg in technology exports is to be noted The nature of the knowledge revolved and exported in the case of large construcUon works appears to be rather d~fferent from that generally assumed for industrial technologies. It would seem to include: (1) capacity to plan, orgamze and supervise the movement of large masses of materials and manpower, (n) abdity to maintain good industrial relations with unskilled workers m temporary jobs (and usually under difficult enwronmental condmons), and (in) capacity to organize extensive procurement and subcontracting serwces.

(u) Demand and ,~upply considerations From the point of view of demand, it was expected (and was confirmed by the data) that other developing countries would be the principal purchasers of the technologies exported, since the projects sold would be better adapted to their needs. This is not only because of the cultural and political ties that may exist between buying and selhng countries, but also because of the similar needs of the market and



economic environment, and the particular technical solutions that have been developed to meet these needs in the exporting countries From the point of view of the technologies involved, it is not easy to explain this outward flow from developing countries by such theories as the product-cycle theory, in accordance with which only the sale of technologies for the manufacture of 'mature' or standardized products that can be easily produced on a small scale and with labour-intensive methods should be expected Yet, the technologies exported from the Latin American countries include, for example, those related to the manufacture of such products as synthetic fibres, pulp and paper, carbon black, various chemical products, steel, pesticides and fertlhsers, radioactive isotopes, pharmaceutical products, and complete telecommunications faclhtles, etc., which hardly fit the postulates of this theory It should be borne in mind that, in order to export these manufacturing techniques, certain technical quahficatlons are necessary, even though the technologies involved may be wellestablished. Even in the case of more traditional products, such as textiles and domestic electrical appliances, the knowledge required for manufacturing must be distinguished from that necessary for exporting a plant or setting up and operating manufacturing facihtles abroad A question may also be raised about the method chosen for the export of techmcal knowhow Why is it more advantageous to the companies in question to export plants or projects than to export the products themselves or to grant hcences for their local manufacture 9 Exports may be restricted because local production in overseas markets is protected Llcences may not be justified by the difficulty of patenting knowledge essentially consisting of operational 'know-how' On the other hand, there is a flow, albeit an incipient one, of hcenslng and patents that is already providing foreign earnings for these countries. It may also be asked: why do they not invest abroad 9 This would require furnishing more capital, assuming greater responsibility in the day-to-day operation of the plant, and so on, than contractual sales of technology Some technology exporters have done so, but the comparative advantage of others could lie more than anything else in what they sell: technology adaptation, plant operating know-how, engineering services, and, in some cases, the construction of the necessary capital goods 2o Sometimes, as pointed out above, the technology content of the operation may be incidental, When this is so, it is worth asking where the strategic interest of the suppher lies, it

may refer to a continuous supply of intermediate inputs or it may consist of the sale of capital goods, whereby the appropriation of benefits will take place through other channels than revenues from direct sales All the operations analysed were market transactions, atlhough in some cases governments Intervened or the transactions had some political basis. Technology sales made in competitive markets could give us some indication about future directions in the comparative advantage of the countries under study 2~ The competitiveness of many developing country technology suppliers may arise simply from the fact that they serve markets for the same type of products and services as demanded by their home market Such is the case of many engineering, construction and capital goods producing firms 22 Among markets with a reasonable degree of competition, capital goods and oil prospecting services could be mentioned In the first case, an advantage in terms of relatively low-wage skilled manpower is often noticed In the second, technical capability seems to be up to international standards though there is a disadvantage in terms of lack of capital resources to be committed in cases of high contractual risks.

(ui) O u t l o o k The emergence of a technological capacity in the Latin American countries analysed has occurred in the context of import substitution under tariff protection as the basis for industrialization Although some argue that these countries can acquire self-reliance in technological matters without acquiring it in production for the domestic market, this argument appears to ignore the fact that a large part of technological learning is acquired by overcoming obstacles in such production Furthermore, the importation of technology has been a necessary first step In the technological development of many companies Technology exports appear at a stage in the evolution of economies that have already achieved a certain level of industrialization and development of their human resources If the phenomenon has been correctly diagnosed as one that is evolving rather than merely a cyclical or temporary occurrence, then a future expansion of the flow of these exports by these countries is to be expected Furthermore, it is expected to grow throughout the region as other countries reach similar stages of development :~

LATIN AMERICA A c o m p a r i s o n with o t h e r seml-lndustriahzed countries, especially those m the Far East, a p p e a r s to indicate that t h o s e countries that have c o n c e n t r a t e d their industrial and technological d e v e l o p m e n t on a few sectors may appear remiss or tardy m achieving technology exports, smce they lack the mdustrtal links, and the necessary technical quahficattons, to s u p p o r t a b r o a d flow of these exports. T h e c o m p a r i s o n , although p r e h m i n a r y and tentative, with Asian countries, such as India and K o r e a , on the one hand, and Tatwan, on the o t h e r , a p p e a r s to support this thesis If it ts c o n f i r m e d , widely based mdustrlahzat~on, w~th p r o t e c t i o n e x t e n d e d to a n u m b e r of mdustrles and for p r o l o n g e d periods of t~me,whtch was a dlstingutshmg feature of the Latin A m e r i c a n e c o n o m i e s , may be one way to achieve technological d e v e l o p m e n t leadmg to the eventual export of acquired knowledge


In m a n y respects, the experience of India m this field is similar to that of the L a t m A m e r i c a n countries. Thts similarity a p p e a r s to stem from their having followed similar p a t t e r n s of industriahzat~on with high levels of p r o t e c t i o n , e m p h a s i z e d the a d v a n c e d t r a m m g of h u m a n resources, and u n d e r t a k e n a substanttal degree of g o v e r n m e n t participation m the p r o m o t i o n of technologtcal d e v e l o p m e n t . O f course, we are not yet in a position to evaluate the e c o n o m i c vtabxhty of these exports: that Is to say, to ascertam to what extent they are m a d e at competitive or subsidized costs, and ff the latter, to what extent their p r o m o t i o n ts justified as part of a process of lndustrxahzation and technological d e v e l o p m e n t that may eventually benefit the p e o p l e of the regton

NOTES 1 We shall mainly refer to the experience of Ar~;entma, Brazil, and Mexico, countries for which prehmmary studies exist While the Andean group countries also show some technology exports, thts experience will be only briefly discussed here since it has not yet been adequately surveyed 2 In some countries, foreign sales of technology are not always duly documented because of the lack of offlcml mcentwes for them and because of the nonexistence of custom records In other cases, registrations exist only for the export of tangible assets incorporating technological know-how, such as the sale of mdustrml plants, documentauon for which is available because they have been offlcmlly promoted 3 Data on technology exports and d~rect foreign investment from Argentina are from Bolsa de Comerclo de Buenos Aires (1980), Katz and Abhn (1978), UNIDO (1983), Softer (1983), CEPAL (1982), Katz and Kosacoff (1983)

8 In the first place, total foreign investment income and not just the part of it that is repatriated should be considered Secondly, returns from both technology exports and direct foreign investment may not be duly captured by the available statistics, since tied-sales, capltahzatlon of amortized assets, etc may be mvolved 9 This amount appears to be substantially larger than in the case of Argentina, but it should be noted that a large proportion of it (around 80%) is accounted for by two major public enterprises, l e PETROBRAS and Banco do Brasll, whereas heavy foreign involvement by state-owned enterprises is not observed in Argentina Excluding these public investments, the remaining 20% (or $80 million) comes fairly close to the corresponding figure for Argentina 10 This is possibly related to the earlier acquisition of skills involved in the appraisal of projects, and the carrying out of pre-mvestment studies, prior to the implementation of investment projects

4 In addmon to known exports of tradmonal products, mentmn should be made of the relatively early direct investment abroad of such companies as Alpargatas, Bunge y Born and Smm Di Tella

11. The special bilateral nature of these contracts requires that these exports be viewed in the appropriate perspective

5 The special facdltles granted to facilitate exports of industrial plants to Cuba probably consutute the mare excepuon

12 The data on technology exports from Mexico are from Dahlman and Cort6s (1982) We wish to express our thanks for permission to use this information

6 Contracts for techmcal assistance provided by Yaclmlentos Petroliferos Fxscales (YPF) (as for example to Ecuador), were possibly not registered because they represent government to government agreements and did not include the export of equipment or capital goods

13 The data available is much better and the coverage more complete, for example, in countries like India Protection tends to favour statistical and economic analysis of these matters by requiring preparation of the data base necessary for control purposes

7 The data on technology exports from Brazil are from Sercovtch (1982a)


Sercovlch (1982b)



15 Direct foreign investments arc not included in Table 5 16 It ~s likely that the explanation ol this curious lact has to do with the way in which the lnanagcment and accounting of torelgn income stemming from technological transact|ons is earned out by Brazilian companies with royalties from other areas probably being accounted tot m industrialized countries It ~s worth noting that a good number of these companies (including those state-owned) havc set up offices abroad for this purpose 17 YPF In Argentina and P E T R O B R A S In Brazil also sell techmcal assistance and participate in thc development ol technological processes having accumulated a good deal ot engineering expertise ,it all levels through their own investment programmcs It is worth noting that only recently have these companies reahzed that their techmcal capabilities bear market value They hitherto had provided a grcat deal of technical assistance and training services to similar, although less experienced, compames in the region on a non-colnmerclal basis through A R P E L (thmr regional technical association) Similar considerations are vahd fl)r state-owned enterprises in other fields like railways, energy, etc 18 ( , i r e m u m be t a k e n v, flh t h e s e c o n l p a r l s o u s s]ncc d a t a o n the , u n i l u n t s i n s u r e d i n c l u d e b o n d s luM l o r part~clp,ltnlg m t e n d e r s w h i c h d o e s not n c c c s s , l r f l y m e a n t h a t t h c w o u l d - h e c x p o r t s l u s u r e d WClC a c t u a l l y made

19 The risk associated with usmg these broader definitions ~s that they can come to include data on exports ot capital goods wnh little or no provision of technical services, which have usually not been included in the other countries The data would thus not be comparable

2(1 Demand-side studies, which have also been undertaken in the context ol the Inter-American Development Bank's project on technology exports, suggest that somenmes the key to the advantage of the developing country supplier may he not so much in whether the knowledge offered is readily adapted to recipient country conditions but m how well their organization and working methods can be adapted to the ahcn environment The stgnlflcancc o1 this can be better appreciated if ~t is recalled that the cultural distance between Argentina or Brazil and some Middle East fir African countries (to which they have exported technology) may be bigger than that between the former and Indusmahzed countries 21 It is necessary to note, however, bclorc drawing conclusions about comparative advantage that countries like those studied are under the lnluence ot important changes in their economic structure due precisely to the development process q his process is characterized by the gener,mon ot externahtles and learning effects The more rapid the process of cconom|c development, the more comphcated the phenomenon may be leading to dllterences in the speed of factor accumulaUon which may s~gnlficantly mod,jy revealed comparative advantage 22 Another interesting insight I r o n the demand-,,lde studms referred to above is that sometimes succes,fful technology suppliers engage in aCtlVmCS in the foreign market that they have never perlormed at home Tl-ms the foreign market becomes a learnmg held protected by goodwill gained through previous contracts business contacts developed and, sometimes polmcal backing 23 See Tel/el (19831 hir a statement about technological development seen as an evolutionary process

REFERENCES Bolsa de Comerclo de Buenos Aires, Proyec¢t6n Externa de la Ernpre~a Argentina (Buenos Aires October 1980) CEPAL, 'l.os scrvlclOS de consultona cn la Argentina La olerta local y la expencncm cxportadora (Buenos Aires 1982) Dahlman, C and M Cort6s, "Technology exports I r o n Mexico', (The World Bank, February 1982), pubhshed as Mexmo study m this issue Inter-American Development Bank, Economu and Social Progress Report m Latm America (Washington, D C 19821, Chap 6 Katz, J and E Abhn, "From infant mdustry to technology exports the Argentine experience in the international sale of industrial plants and engineerlng works', I D B / E C L A Research Program on Science and Technology Buenos Aires, Working Paper No 14 (Buenos Aires 1978) Katz, J and B Kosacoff, "Direct foreign investment of Argentine Industrial enterprises', in S Lall et al,

Hie New Multinationals The spread oJ Tturd Worht Enterprises (Chlchester J Wiley, 1983) Sercovlch, F C , 'Brazd as a technology exporter (Inter-American Development Bank August 1982), published as Brazil study in this issue Sercovlch, F C , 'Report from ,t short term mission to Andean Pact countries' (Inter-American Development Bank March 19821 Solfer, R , 'Exportaclones de tecnologia desde Argentina', B I D - I N T A L , (Buenos Aires 1983), published as Argentma study m this issue T e n d , S 'Toward coneeptuahzatlon of technological development as an evolutionary process', 7th World Congress of the Internatmnal Econonuc Association (Madrid 5-9 September 1983) U N I D O , 'Technology exports from developing countries, the cases of Argentma and Portugal', Developrnent and Tram[er of Tethnology ,Series No 17 (1983)