December 1991

December 1991







Brazil and the World Economy in the 1990s

Emerging trade and investment issues


Winston Fritsch & Gustavo H. B. Franco

Departamento de Economia, PUC-RJ

Rua Marques de São Vicente 225

Rio de Janeiro RJ 22453

tel (021) 274-2797 Fax (021) 294-2095



Part I: The changing framework of trade and FDI policies

2. Trade policy

2.1) The trade regime in the 1980s

  1. Protection through an administratively decentralized system of QRs
  2. Special import regimes and tariffs as a secondary line of defense
  3. The neutralization of protection desincentives through export incentives

2.2) International trade policies

2.3) The new challenges of the 1990s

2.4) The 1990 trade policy reform

3. Direct investment policy

Part II: The emerging issues

4. Trade policy

4.1) Prospects of the trade liberalization program

4.2) Trade policies: the regional dimension

4.3) Trade policies: the multilateral dimension

5. New directions in FDI and technology transfer policies

6. Conclusions

7. References

Prepared for the conference "Latin America's Integration into the World Economy: confronting the choices", organised by the Inter-American Dialogue, The Brookings Institution, Washington, 18th-19th December 1991.



This paper is a comprehensive synthesis of the key policy issues in current Brazilian trade and direct investment policies both in their domestic and international (regional and multilateral) dimensions. It is divided into two parts. Part I examines the lessons of the 1980s, summarizing the key dimensions of the trade regime in existence in 1990, the main features of its sweeping reform launched in March 1990, the recent challenges to the traditional framework of Brazilian international trade policy, and the regulatory practices affecting FDI in Brazil. Part II is forward looking, and examines the policy problems arising in the context of the ongoing attempts at changing the pattern of Brazil's integration into the world economy. It discusses the prospects of sustaining the trade liberalization experiment and its consistency with the Brazilian positions in regional issues - namely the MERCOSUL and the Bush Initiative - and the Uruguay Round. Finally, we discuss the new directions in the treatment of foreign capital likely to emerge in the current discussions now taking place in the Brazilian Congress. Finally, a section summarizing the main conclusions closes the paper.


Part I: The changing framework of trade and FDI policies

2. Trade policy


2.1) The trade regime in the 1980s

During the troubled 1980s, despite dismal growth and record high inflation rates, the evolution of Brazil's export growth was maintained well above world market rates and accompanied by an impressive diversification in their commodity composition towards manufactured goods - now accounting for around 60% of total exports. In parallel, contrary to this increasing "outwardness" on the export side, Brazil has radically reduced her import propensity, reaching near autarky levels in terms of manufactured imports. Both macroeconomic and structural factors played a role in determining such phenomena, but a crucial influence was the evolution of the trade regime. The key features of the latter can be summarized as follows:

(i) Protection through an administratively decentralized system of QRs: The coverage of quantitative and administrative restrictions on imports was 100% all through the 1980s and the agency responsible for import authorizations - CACEX - progressively developed a system of consultations with business associations which resulted in a marked diversification and decentralization of administrative obstacles to imports. Sectoral agencies and regulatory bodies were granted powers to restrict specific imports mostly through local content requirements imposed as conditionality to the access to fiscal incentives, credit from official institutions and public procurement. This decentralized and highly discretionary system of protection, given the great levarage enjoyed by sectoral agencies in project approval and in the definition of sectoral industrial policies, resulted in a de facto privatization of regulatory functions in trade and industrial policy making which turned regulation into an activity designed to protect established interests, and to enhance non-contestability of markets.

(ii) Special import regimes and tariffs as a secondary line of defense:

As a corollary of decentralization of import controls many "special import regimes", under which specific items were granted tariff and other border tax exemptions, appeared. Importing firms that happened to by-pass local content requirements invariably applied for tariff exemptions or reductions under such regimes, which were granted by specific legislations, or on an ad-hoc basis by the Tariff Commission. As a result, nearly 67% of Brazilian imports entered the country in 1985 with tariff reductions or exemptions. An interesting consequence of such "special regimes" was the large difference between legal tariffs and the ones effectively practiced. In 1984, for example, while the average tariff in manufacturing was 90.1% the "true", i e. effectively paid rate, measured by the import tax revenues as a percentage of imports' value, was 19.1%. In 1989, when legal tariffs have been brought down to 49.4%, the average "true" tariff stood at 8.3%.

(iii) The neutralization of protection desincentives through export incentives:

The high levels of protection created a generalized anti-export bias which, however, was offset in specific industries by export incentives placing such industries in a more or less neutral regime. In 1984, for instance, the aggregate value of all export incentives was quite significant and, interestingly, it exhibited a positive and significant (rank) correlation with effective rates of protection: 0.626 for 1973 and 0.723 for 1977 - suggesting that export incentives played a key role to neutralize the anti-export bias implicit in the structure of protection. In addition, exporters had access to imported inputs and capital goods at special conditions through a special program - the BEFIEX - in exchange for export performance commitments on a firm basis. The program's effectiveness can be assessed by the fact that exports under the BEFIEX increased more than tenfold from 1974 to 1981, from US$ 335 million to US$ 2,581 million, and another fourfold during 1982-89, from US$ 2,343 million to US$ 8,979 million. The program's share in manufactured exports increased from16% in 1975-79, to 23% in 1980-84, reaching slightly over 50% in 1987, where it remained since.

2.2) International trade policies

Following the return to democratic rule, but previous to the recent liberalization decision, Brazil had been active in reviving the idea of subregional integration with Argentina and Uruguay, but mostly prompted by the lure of saving scarce dollars in times of foreign exchange shortage by some kind of barter trade, or some system of reciprocal prefences. Yet, the protectionist bias of domestic policies of the leading partners, precluded the advance of such arrangements, but more recently, with the proliferation of unilateral liberalization initiatives, regional integration revived as an important issue. Landmarks in this process were the signing of the first bilateral protocols by Presidents Sarney and Alfonsin in 1986, launching cooperative ventures in several areas. These diplomatic moves had the fundamentally political motivation of restablishing an amiable climate in bilateral relations, which had reached a low ebb in the late seventies. However, as far as trade creation is concerned, results were little short of disappointing. Trade actually shrank in areas specifically contemplated by special protocols, such as wheat and capital goods, and special bilateral trade promotion instruments simply did not work due to their overlap with extremely unstable macroeconomic conditions in both countries. Nevertheless, this has not prevented the two presidents to define, by the end of 1988, a ten-year time frame for the construction of a common market, which was eventually ratified by the national parliaments in 1989.

In July 1990, the outlook for regional trade integration would again be altered by two important events. First, a change in presidential styles in both Brazil and Argentina plus Brazil's sweeping trade liberalization decision, facilitated the big-bang approach to regional integration adopted in the Buenos Aires Treaty, signed by presidents Collor and Menem, which went as far as to decide for a linear and across the board tariff elimination in bilateral trade by end-1994. Second, President Bush's celebrated Initiative for the Americas speech called, among other things, for the formation of a free trade area ranging from Anchorage to Tierra del Fuego.

The Bush Initiative squarely placed two crucial strategic questions for both Brazil and Argentina. The first related to the choice of the negotiating strategy as between a purely bilateral negotiation or an attempt to negotiate as a bloc of sub-regional integrated economies. The other related to the optimum timing and extent of free trade with the United States. The answer given to the first question was relatively simple. As bargaining power in bilateral reciprocal negotiations depend on the size of domestic market, there was the perception that it would be interesting to concentrate efforts to negotiate en bloc. Such efforts would, however, be eased by the harmonization of trade policies and this added to the momentum for the signing of the Assunci_n Treaty, on March 1991, between Brazil, Argentina, Uruguay and Paraguay, creating the MERCOSUL: a common market for goods, services and factors which should be operational by 1994, enlarging the already far reaching objectives of the Buenos Aires treaty. The second strategic question was not a pressing issue, as no concrete American diplomatic action followed the Bush speech, as in the second half of 1990 the agenda of trade policy makers was crowded by the expected completion of Uruguay round in the Brussels GATT ministerial.

In the multilateral front, Brazil's active participation in the GATT was traditionally marked by a leading role in discussions relating to special and differentiated (S&D) treatment granted to developing countries under Part IV of the General Agreement, as well as by concern with system-wide issues and multilateral disciplines - a feature stemming from Brazil's extremely diversified composition of exports in terms of both commodities and markets. This stand was, of course, consistent with domestic trade policies. On the one hand, high levels of import protection through foreign exchange controls required sustaining the rights to use such means of absolute protection allowed by S&D under Article XVIIIb. On the other, as market access problems became of paramount importance to sustain Brazil's increasing outwardness, she aggressively denounced both GATT's long-standing lack of concern with liberalizing primary product trade as well as the neo-protectionist breaches of GATT rules affecting market access for manufactures. For a long time this asymmetry - using protectionist (GATT legal) policies allowed by S&D herself, but denouncing its (GATT illegal) use by other developed trade partners - posed no problem to the very able Brazilian Foreign Office officials for, from a strictly legalistic standpoint, it was perfectly defensible under GATT rules. However, its long term use was a source of growing irritation in the OECD - and, especially, in the United States. Indeed, throughout the negotiations leading to the Punta del Este GATT ministerial, when the Uruguay Round was launched in 1986, Brazil followed a line consistent with such positions, defending an agenda for the Round that would concentrate on dealing first with the backlog of unfinished business rather than on the so-called "new themes" - TRIMS, TRIPS and trade in services - as proposed chiefly by the United States. This strategy did not succeed and, following Punta del Este, Brazil was caught againg in the awkward position of arguing for market access without promoting a freer trade regime at home. The recent trade liberalization will thus create the opportunity of reducing Brazil's isolation and of an important shift in her capacity to influence events in the final stages of the Round.

2.3) The new challenges of the 1990s

The highly interventionist character and the multiple distortions in the Brazilian trade regime were generally minimized by the claim that they generated unimportant static allocative costs as export performance was sound and, at least up to the early eighties, so was overall output growth performance. However, since the onset of the protracted growth crisis of the early part of the decade, four new factors emerged to change perceptions as to the usefulness of the trade regime in stimulating industrialization and trade performance.

The first of them is historical, as the perceived development challenges have definitely changed. As Brazilian industry grew and matured, the meaning of "industrialization" became quite different from what it was in the heyday of import substitution and so did the nature of the questions asked about strategic choices regarding industrialization. The policy problem today is no longer how to save foreign exchange by protecting the establishment of industries to produce for the domestic market at not too unreasonably large cost differentials relative to world prices in the not too distant future. The great challenge today is how to reform the policy framework built to answer the former problem - so as to overcome its obvious inadequacies to generate the managerial dynamism required to meet the challenges posed by the world's fast changing technological trends and the continuous need of export upgrading by middle-income manufactured exporters.

The second is related to the fact that the need to overcome the macroeconomic constraints on growth increased the concern with the attainment of efficient industrialization. There is a growing consensus that the task of restoring sustained growth after the painful contraction of investment levels in the 1980s requires lifting a potential foreign exchange constraint imposed by the debt burden, restoring domestic savings from its very depressed levels, and redressing government financial balance. However, this should be achieved in a context of rising demands for greater distributive equity and the provision of social overhead services given Brazil's dismal social indicators. Now, the only way to simultaneously lift the foreign exchange and savings constraints on growth, and allow a rapid and sustained rise in real wages and of government consumption expenditures on social services is to increase productivity growth, as it would prevent both the erosion of manufactured exports competitiveness as well as the shift in income distribution towards wages and a consequent fall in the savings ratio. These macroeconomic imperatives placed the concern with efficiency considerations at the centre of the stage, and led to a a growing perception that the trade and industrial policy framework has become detrimental to the attainment of dynamic efficiency and that this handicap is more serious now than ever before.

A third factor informing the will to reshape trade and industrial policies have to do with external influences. The slow move of the opinion of local elites towards deregulation - especially in the sphere of trade and industrial policy - was also undoubtedly reinforced by the almost permanent pressure from some OECD governments and multilateral organizations on Brazilian public authorities for trade liberalization and industrial policy deregulation. No doubt, trade policy reform in Brazil was implemented with a view to improve the climate of Brazil-US relations to pave the way to a reopening of debt negotiations following a long moratorium, and in a time in which far reaching agreements for sub-regional integration in the Southern Cone were taking shape; the United States put forward a proposal for an hemispheric FTA in the context of the Bush Initiative; and the Uruguay Round of multilateral negotiations neared its conclusive stage. In addition, the multiplication of Latin American examples of liberalization cum deregulation programs created the sensation that Brazil should not stand against international trends.

Fourth, the political motivation to advance trade reform and industrial deregulation in Brazil had much to do with popular disenchantment with the nature of state intervention in Brazil. The generalized perception was that the multiplication of fiscal and regulatory favours, widely seen as a process of privatization of the state, could not be justified by any rational criteria, and that it was a product of the disproportionate influence of special interest groups. Popular attitudes towards government intervention became similar to those reportedly seen in Eastern Europe. A growing perception, on the part of a substantial part of the local elites, of the political unsustainability of the rent seeking arrangements sustaining past trade and industrial policies, has become instrumental to strengthen the political drive towards a more market oriented economy.

2.4) The 1990 trade policy reform

The traditional trade policy framework was seriously shaken in March 1990 with the announcement, by the newly elected Collor government, of a sweeping trade liberalization programme launched as part of a broader reform in the traditional principles, objectives and instruments of industrial policy. The implementation of trade liberalization was conceived to take place in a phased way, following a seemingly conventional path. There would be an immediate rationalization of the import regime, whereby some duty exemptions granted on an ad hoc basis by the Tariff Commission under "special import regimes" would be abolished. Then, the actual liberalization process would begin with the abolition of some QRs, especially those enforced by CACEX, allowing a more extensive application of the tariff schedule, followed by a reduction of the height and sectoral dispersion of the tariffs over time

The first important step in the liberalization sequence was the abolition of a list of around 1,300 prohibited import items in May 1990, when new - and high - tariffs were created for the previously prohibited products. In July, the more important mechanisms of quantitative control of imports were relaxed. Several "buy-national" programs and local content requirements were, however, maintained, though with more moderate dispositions. Shortly after, it was decided that a new tariff would be put into force at the beginning of 1991, together with advance notification of the pattern of fall up to 1994, when tariffs would average 20% within a range from zero to a maximum of 40%. In the same breath, the government announced the revision of the existing "market reserve" policy for "informatics goods", which limited both imports and the activities of foreign firms locating in Brazil to service the domestic segments of the mini and microcomputers market, which has by now been completed.

The new tariff schedule, with the projected yearly variations until 1994, was announced on 1st January 1991 to be in force from mid-February. Generally speaking, the methodology followed consisted in classifying the 13,500 items in the Brazilian tariff according to the following seven brackets: (i) zero tariff, for products with natural comparative advantages (mainly primary or semi-processed traditional exports), with natural protection (due to high transport costs), with no competitive domestic production and commodities with low value added; (ii) a 5% rate, for products which already paid 5% in 1990; (iii) rates between 10% and 15%, for products using zero tariff products as their main input (such as the paper and pulp or cotton chains); (iv) a 20% rate, for the bulk of manufactured products; (v) a 30% rate, for fine chemicals, wheat, wheat biscuits, pasta, TV sets, record players, video cassettes and sound equipment; (vi) a 35% rate, for autos, trucks and motocycles; and finally (vii) a 40% rate, for informatics goods.

Table 1

The new Brazilian tariff, 1991-94 (in %)

1990 1991 1992 1993 1994

Average 32.2 25.3 21.2 17.1 14.2

Standard deviation 19.6 17.4 14.2 10.7 7.9

Source: CTT/MEFP.

Having classified the products according to this general rule, the pattern of fall was designed to follow a rule of concentrating the heavier reductions during the first two years on intermediate and capital goods, thus enhancing effective protection and competitiveness of consumer goods sectors. Moreover, besides the fall in the average tariff over time, protection was to become more homogeneous among goods, that is, the variance of tariff rates was to fall. The aggregate result of this exercise is shown in Table 1. If one takes into consideration that the tariff in force after the 1990 round of abolition of QRs had a maximum of 105%, the extent of the projected liberalization looks quite impressive.

3. Direct investment policy

Foreign direct investment has been a crucial actor in Brazil's industrial development. The weight of foreign firms in domestic value added in manufacturing is around a third on average, and their share in manufactured exports is greater than the 40% and, in both instances, their shares are substantially higher in the more dynamic and technologically sophisticated branches of manufacturing. Foreign capital has been important in Brazilian industrialization in different ways through the years and the key to this successful "mariage" has been, on the one hand, extensive investment opportunities and, on the other, the enforcement of regulation in line with international trends. In the 1980s and 1990s both premisses came to question: macroeconomic instability reflected, as alluded above, the difficulty in the redefinition of a new development model, i. e. a new pattern of investments and economic activities, and the precise definition of policies, general and more specifically related to foreign capital, to support the new model. This section discusses existing FDI policies, as a prelude to the broader issues to be addressed in section 5 on the reform of such policies.

The basic piece of Brazilian legislation regulating foreign investment (the Profit Remittance Law) dates back to the 1960s. The fact that it has remained untouched for the past few decades is commonly taken as a highly desirable aspect of foreign capital regulation in Brazil. Its principle is quite simple: it is based on the notion that direct foreign investments should be registered at FIRCE, a division of the Central Bank, in their currency of origin, and that dividend remittances, although free from quantitative limits, are taxed at a rate of 25%, and progressively if they exceed 12% of registered capital. There is no sectorial discrimination, except for a provision that is rarely recalled which sets an 8% limit on dividend remittances resulting from profits on activities connected with "superfluous" goods and services. Reinvested profits may be included under the value of the capital registered with FIRCE, and there are no limits on repatriations.

Over the last few years, however, many new forms of international investment have appeared. This which required the development of new practices, often discretionary and tentative, mostly to deal with investments involving intangible assets and technological links. In this area, Brazil's regulatory regime has not been duly established and is certainly not as well-matured as the Profit Remittance Law. The passing of time has aggravated some distortions that are acknowledged to exist in FIRCE administrative practices and formal regulations. An example is the criteria used for registering foreign capital. Given that the basic principle of the law is progressive taxation due on "excessive" rates of return (over 12%), it is necessary that the concept of registered capital be as close as possible to the concept of capital relevant to the economic decisions. Yet, FIRCE criteria underestimate amounts of foreign investment in Brazil first, because investments' book values (in foreign currency) do not undergo any monetary correction for inflation in their country of origin and second because because inter-company loans are not considered as direct investments, a procedure that contrasts with that adopted by, for example, American statistics, and which to a large extent explains the fact that amounts of US investment in Brazil according to FIRCE are substantially lower than US Department of Commerce figures.

The fact that direct investment stocks registered with FIRCE are underestimated suggests that rates of return - as measured by amounts remitted and reinvested as a proportion of registered capital - may be very high, which would mean that foreign companies become subject to marginal tax rates applicable to remittances of over 12% of registered capital. Surprisingly, however, as may be seen in Table 2, average "rates of return" according to Brazilian statistics do not even approach 12%.

Table 2

Dividends and Reinvestment 1978-1989 (in US$ million and average annual ratios as % of registered stock)

1978-82 1983-87 1987 1988 1989 1990

  1. Dividends

Million US$ 492 974 909 1,539 2,383 1,591

As % of stock 2.9 3.7 2.9 5.0 7.0 4.5

(2) Reinvestments

Million US$ 881 555 617 714 531 273

As % of stock 5.0 2.1 2.0 2.3 1.5 0.8

(1)+(2) as % of stock 7.9 5.8 4.9 7.3 8.5 5.3

Source: BACEN

Firstly, it may be noted that much caution is needed in taking the ratios in Table 2 as rates of return, in view of distortions caused by transfer prices, "black market" remittances, or debt-equity swaps, as well as difficulties in distinguishing economic from accounting profit. With these caveats in mind, it may be noted that this Table's "rates of return" are surprisingly low, contrary to what was suggested above. No doubt this should be seen as an indication that profits from Brazilian subsidiaries are being transferred through other channels. As royalty payments amount to a very limited sum, thus offerring little room for disguised transfers, the remaining possibility is the practice of transfer prices in trade between Brazilian subsidiaries and their head offices. Total exports for foreign companies in Brazil reached US$ 7.1 billion in 1985, with imports of US$ 2.3 billion. An appreciable amount of this trade is intra-firm, as is well known, and it is quite probable that transfer prices are practiced in such trade. Only an empirical investigation has been carried out on transfer prices practiced by foreign companies, which showed that, for a sample of some 1,006 imports made by Brazilian and foreign companies in Brazil in 1979, foreign companies paid prices between 21% and 38% higher than those paid by Brazilian companies for the same product. These proportions thus suggest overestimation of import prices (35%) and would imply transfers of around US$ 1.2 billion through over-invoicing of imports. If similar proportions for under-invoicing are applied to exports, amounts transferred would reach some US$ 1,3 billion, which would raise the amount transferred to around US$ 2,5 billion in 1979. It is probable that there is some exaggeration in these figures, but they would certainly justify some effort by the authorities aimed at checking the real extent of such practices.

Transfer pricing is certainly an old issue and the only motivation in bringing it to the fore is to note that the Brazilian regime worked quite well, dispite its trend towards overtaxation of FDI, provided that transfer pricing was allowed in significant magnitudes. One distortion cancels the other and both sides coexist quite peacefully in an unwritten pact that lasted for at least two decades. It is uncertain, however, whether, in view of the new realities of the 1990s, the same pattern of events could be maintained, as discussed in more detail in section 5.

Part II: The emerging issues

4. Trade policy

4.1) Prospects of the trade liberalization program

In the prevailing environment of persistent macroeconomic imbalances, the sustainability of Brazil's trade liberalization experiment is not to be taken for granted. There are threats to be faced both in the short and the medium run. In the short run, there is the classical sequencing problem between trade liberalization and stabilisation policy under conditions of macroeconomic instability, or the idea that trade liberalization pressuposes a previous successful stabilisation effort in order to: (i) avoid superimposing the fall in aggregate domestic demand caused by the stabilisation programme to the sectoral adjustment strains accompanying trade liberalization and (ii) to prevent the usual temptation to transitionally target the exchange rate to objectives of stabilisation policy, thus leading to its appreciation which increase structural adjustment costs as well as the danger of balance of payments problems during the implementation of the reforms.

The failure of the Collor administration to control inflation - now stabilized around 20% per month - in spite of an almost continuous fall in activity levels and rising unemployment, enhances these threats and increases the political odds against the liberalization experiment. Although, as mentioned above, current tariff levels are still providing a comfortable margin of protection to most domestic industries, a macroeconomic environment of "repressed hyperinflation" through very contractionary demand management and appreciating real exchange rates may increase the adjustment costs of trade liberalization beyond politically acceptable limits.

This is not to say that, given Brazil's very closed industrial system, the pressure of import penetration is likely to cause significant shifts in aggregate industrial employment levels. During 1989-90, the average share of domestic demand in total manufacturing output was around 95%, against 9.5% for exports and only 4.5% for imports. In an economy such as this, the impact of a 20% rise in real imports on domestic output and employment can be countervailed by an increase of less than 1% in domestic demand. Opposition to trade liberalization may, nevertheless, stem from sectorally or regionally localized unemployment caused by import competition coming on top of a general output contraction caused by the fall in the large component of domestic demand.

Assuming a restoration of domestic stability and demand growth, political sustainability of the trade liberalization experiment will, on the other hand, depend on avoiding serious external disequilibria. The somewhat uneventful behavior of imports after the liberalization measures suggests that a major surge in manufactured import demand is not to be expected as a consequence of the announced tariff cuts. This somewhat puzzling development places Brazil in sharp contrast with respect to liberalization episodes in Mexico and in Turkey, for instance, where rapid and very significant increases in openess followed the implementation of import liberalization. The recession is certainly a factor in reducing import propensities in 1990 and 1991 and certainly, in a recovery, income effects will add a strong stimulus to the demand for manufactured imports. Yet, the level of imports to be reached after the recovery and the elimination of some trade restrictions still in force - including, for instance, "non-border" NTBs, such as local content requirements, and the poor working of ports - is still uncertain. In any event, the likelihood of balance of payments strains has been played down given that, on the one hand, export growth seemed positively affected by the static efficiency gains caused by the fall in import protection and, on the other, the buoyant state of the capital account. In such conditions, the Central Bank managed the exchange rate as to avoid significant real depreciation, thus ridding the country from the vicious circle of inflation and fiscal disequilibrium typical of the 1980s. Yet, to sustain this situation of slowly increasing export orientation and heavy capital inflows it is vital that Brazil restores a sound government financial position. Failing this, there is the threat that, as frequently happened in the past, trade restraints would come to be used to fight external disequilibria created by macroeconomic inconsistencies. However, if the right macroeconomic policies are adopted this threat is likely to disappear over time as the country's productive structure adapts to the new pattern of specialization induced by relatively low tariffs, thus generating coalitions against the reimposition of the old structure of protection.

Therefore, the longer the period of external stability following liberalization, the greater the possibility of consolidating the current trade policy reform. The crucial horizon is the next three years, the period in which the announced tariff cuts will be enforced. Within this time frame, it is essential to restore sustainable output growth with macroeconomic stability so as to reduce both structural adjustment costs and the risk of balance of payment problems. These imperatives require, of course, a successful stabilization effort, but also underline the links between the future of Brazil's trade policy and the restoration of adequate volumes of foreign financial assistance

4.2) Trade policies: the regional dimension

Sustaining the announced trade liberalization experiment is not, however, a sufficient condition for progress in the minilateral dimensions of trade policy. As far the MERCOSUL is concerned, two further conditions will have to be met. The first is the revision of the too fast track for free trade in goods set in the Assunción Treaty. The decision to replace the previous product by product negotiations by an across the board linear tariff elimination by 1994 was basically established on the basis of political judgement regarding two points. First - a notion which had been shown to be true in the Brazilian trade reform programme - that linear, non-discriminating, cuts reduced the incentives to the formation of organised lobbies against integration. Second, that such a radical move would impart credibility to the liberalizing intentions of the member governments. This decision lacked, however, any serious technical backing and, given the existing productivity differentials in some sectors, it is difficult to believe that it will be met in full, as programmed. The likely outcome will be a revision of the 1994 deadline to fit the structural adjustment problems of sensitive products to a more realistic time frame. On the other hand, in the case where efficiency differences result from technological capability or other forms of firm specific advantage and not from locational advantages, this adjustment process may be speeded up by bilateral direct investment flows which take advantage of the interesting statute of "binational enterprises", created in the Buenos Aires Treaty and ratified in Assunción, through which Mercosur members grant national treatment to any firm whose majority control is held by MERCOSUL residents.

The second, and perhaps more difficult, is the attainment of stability in the cruzeiro-austral bilateral real exchange rate. Given the arbitrary management of quantitative trade restraints, the wide variations in exchange rates shown in Chart 1 have not in the past significantly influenced bilateral trade flows, which were basically determined by fluctuations in activity levels. However, after the enforcement of free bilateral trade as envisaged in the MERCOSUL, changes in relative prices are bound to play a much more important role in determining trade flows. Thus, wide changes in competitiveness generated by exchange rate fluctuations will place an additional - and, if they happen to have the amplitude of the past, unbearable - strain on the political sustainability of the integration process.

[chart1 here]

To gauge the impact of these fluctuations on relative competitiveness note that, ignoring transportation costs, the changes in the relative price of an import from Argentina and its domestic substitute in the Brazilian market can be decomposed into the change in the real cruzeiro-austral exchange real rate and the fall in the Brazilian tariff during integration. Thus, changes in competitiveness of Brazilian products competing with Argentine imports of the order of 20%, created by the elimination of the average 25% Brazilian tariff, can be enormously magnified by a repetition of historically registered real exchange rate appreciations - sometimes of over 100%, as shown in Chart 1. Attaining bilateral exchange rate stability will, of course, require macroeconomic stabilization both in Brazil and Argentina. This may sound far fetched in the near future, but there can be little doubt that this is a necessary component of the success of the MERCOSUL project.

Although the crowded MERCOSUL agenda formally conditions the timing and extent of Brazil's response to the Bush Initiative, the lack of any significant response to the FTA offer can to a large extent bey explained by the uncertainty surrounding the specifically Brazilian interest in these negotiations from a national standpoint. To evaluate the prospective gains to Brazil in of joining an FTA with the United States one has to take into account the net effect of two distinct sets of impacts. There are, on the one hand, two certain negative effects. The first is the trade diversion effect caused by the erosion of Brazil's preferential margins in MERCOSUL markets with the lowering of trade barriers of other MERCOSUL members against competing American imports. The second is that, as the United States is not the most efficient supplier of all Brazilian imports not coming from MERCOSUL (as witnessed by the large share of non-US suppliers in Brazilian non-MERCOSUL imports) the gains from bilateral liberalization are lower than those of multilateral liberalization. This efficiency loss may not be negligible as even after the process of unilateral trade liberalization is completed by 1994, the Brazilian multilateral tariff applied against third party imports will remain relatively high. A positive effect, on the other hand, will result from the trade preferences in the US market granted under the hemispheric FTA, which can be especially high if they affect products currently behind high tariff or non-tariff barriers. Thus, the only economic rationale for a Brazilian interest in the Bush Initiative, that is, for a decision to move faster in bilateral trade liberalization than in its (multilateral) domestic trade liberalization reform, is to gain discriminatory access to the large US market. The crucial determinant of Brazil's interest in the Bush Initiative is, therefore, the extent of discriminatory access to be gained in the US market.

The evaluation of such prospective gains can only be made today on very a priori grounds for at least two reasons. The first is that they depend on the still uncertain outcome of the Uruguay Round, especially in the key areas of tariffs and safeguards. Without knowing the height of the post-Uruguay Round US tariff and, especially, whether VERs would come under GATT discipline and, if so, whether "selectivity", i. e. discrimination, in the application of temporary quantitative restraints will be allowed, it is simply impossible to assess the extent of discrimination granted by free access to the American market. The second is the credibility of the US offer of free market access. Indeed, even after the Round is concluded, there are grounds to believe that access in "sensitive" areas such as those today affected by VERs or where the tariff is relatively high will continue to be restricted. The acid test here will be the first FTA negotiations involving Mexico and, perhaps, Chile, whose outcomes are still well into the future.

All in all, under the present circumstances it seems safe to bet that Brazil will not press for any early comprehensive bilateral liberalization deal between MERCOSUL and the United States and that she holds the power to restrain isolated moves by their sub-regional partners. Thus, in the foreseeable future, given the weight of Brazil and Argentina in total South American trade, greater hemispheric integration - that is, the rise in North-South American trade as a proportion of total hemispheric trade - will continue to depend more on maintaining higher rates of hemispheric growth and the momentum of national or sub-regional trade liberalization policies in South America than on the rise of hemispheric trade discrimination against third partners, as envisaged in the Bush Initiative.

4.3) Trade policies: the multilateral dimension

Brazil's participation in the GATT will continue to be marked by her traditional concern with system-wide issues and multilateral disciplines which, as argued above, stem from structural factors such as the country's diversified export interests and geographically scattered trade pattern. However, Brazil's negotiating style in the final stages of the current Round is bound to be influenced by the recent liberalizing changes in the direction of her trade and foreign direct investment policies. These changes shall allow Brazilian negotiators to reduce their former necessarily defensive style - centered on the preservation of S&D status, especially relating to the application of Article XVIIIb, and the concern with the introduction of GATT disciplines on new themes, especially services - while keeping an agressive stand on the discussion of the backlog issues. This change from a confrontational attitude on issues squarely placed along the North-South divide, as S&D and trade in services, from which resulted a not inconsiderable isolation of Brazil as a "hard liner" in the early stages of the Round, will place Brazilian negotiators in a more comfortable and, perhaps, more influential position to influence the more complex coalitions formed around the backlog issues.

A move towards a "softer" stand on the formerly sensitive issue of S&D has already been made with the pledge not to invoke Article XVIIIb, a move which is a natural result of the replacement of administrative controls by tariffs as the instrument of protection in Brazil. Another areas in which former conflict between Brazil and, especially, the United States, can be reduced by foreign investment policy reforms under discussion in Brasilia are TRIMS and services, as the most contentious aspects of the discussion of a framework for GATT disciplines on these new themes was the question of national treatment of foreign-owned concerns. The only remaining area of conflict along North-South lines is likely to be TRIPS where Brazil may continue to side with relatively industrialised developing countries without strong national innovation systems, which see the issue of international regulation of industrial property and technology transfer under a different perspective than that of the innovating countries.

Brazil's stand on the old (backlog) issues is likely to show little change and, as indicated above, the remaining stages of the Round will witness a strong Brazilian voice in several crucial coalitions sustaining GATT's liberal principles. On safeguards and market access, siding with the other NICs pressing for greater GATT discipline and surveillance in their application. On tropical products, siding with other developing producers against tariff escalation in the OECD. On agriculture, siding with other Cairns Group members and the United States against the CAP and protectionism.


5. New directions in FDI and technology transfer policies

The shaping of Brazilian integration into the world economy in the years to come is certainly an outcome of the future course of the trade liberalization experiment, but is is bound to be affected by policies adopted in the areas of foreign direct investment and on rules and decisions on issues - such as intelectual property and technology transfer rules - associated to high tech industries. Strategic decisions on these areas are implicit in the on going redefinition on Brazil's industrial policies and in the broader debate on the redefinition of a new development model. Most likely, to judge from past experience, these choices are to be affected by pragmatic assessments of recent trends in North-South investment flows and associated international industrial restructuring trends, and the opportunities being open in the global industrial economy for semi industrialized countries. Notwithstanding, no illusions are entertained as to the effects of Brazilian industrial policy decisions on fundamental features of the country's integration into the world economy. Past experience as well as especialized literature teach that the effectiveness of specific measures of incentive is basically determined by their alignment with major trends in international investment on a worldwide scale, under relevant historical circumstances. Regarding foreign direct investment in particular, it is well known that the regulatory regime is only one of many considerations relevant to investment decisions taken by foreign companies, whose strategies are governed by what is commonly called the "investment climate", or the "market fundamentals" in the country in question. In this connection, one is led to believe that macroeconomic considerations, or the investment climate affecting domestic and foreign firms, may be much more relevant to effective decisions than wedges introduced by industrial policy. Although it is recognized that regulatory restrictions may inhibit direct foreign investment, deregulation does not necessarily bring about positive effects. This is obvious in the case of Brazil because continuing instability in the country's macroeconomic situation makes it hard to believe that there will be any appreciable recovery in direct investment in response to deregulation of foreign capital in Brazil.

Many delicate issues enter the domain of policy redefinitions affecting foreign direct investment and in most these areas one cannot say Brazilian dispositions are by any means restrictive or deviant with respect to international practices. This is the case of rules for debt-to-equity swaps, export processing zones, sectorial restrictions to foreign capital (risk contracts for oil prospecting, mining, transportation, finance and informatics) and discrimination in government procurement and credit from official institutions. Other areas in which somewhat restrictive rules prevailed have been subject to changes determined by the new industrial policy orientation. This is the case of rules of access to Stock Exchanges, which were considerably liberalized, tax provisions affecting dividends and capital repatriation, which were relaxed, and difficulties in dealing with foreign investments intensive in intangible assets. The latter is actually the area in which uncertainties are the greatest, mostly because procedures are highly discretionary and rules are by no means explicit.

Two government agencies are crucial to regulate transactions involving intangible assets: the INPI (Natinal Institute of Industrial Property), if some form of technology transfer contract is involved, and FIRCE, because it necessaily involves a remittance abroad and may involve the mechanism of the profit remittance law. In the latter case, FIRCE may decide, for instance, that the intangible assets should be deducted from the concept of capital relevant for the taxation of dividends' remittance. The justification is most usually the idea that it is difficult to have a proper valuation of an intangible asset. Yet, this is really the same type of problem that existed, and continues to exist, with direct investment in merchandise, in compliance with the mechanism introduced in the mid-1950s by the famous Central Bank (then SUMOC) Instruction 113. According to this norm, it was, and still is, possible to bring equipment into the country and - with no foreign exchange transaction involved - turn it into registered capital for purposes of remittance of dividends. There is indeed a problem with the valuation of equipment, principally when second hand. This evaluation is currently carried out by the Board of Industry and Trade - CIC, formerly the Foreign Trade Bureau - CACEX, and always involves some form of negotiation. There is no reason to assume that a similar procedure cannot be adopted for intangible assests, with the intervention, for instance, of INPI. In other words, it makes sense to update Instruction 113 for the 1990s by making it applicable to intangible assets.

An interesting example of the difficulties created by the regulatory stance towards intangibles occurs in the case of services. Existing restrictions on the import of services - regulatory or natural, as many services cannot be marketed internationally - would form natural incentives for direct investment in Brazil in such sectors, but hindrances to capitalization of intangible assets tend to make registered capital very small as a basis for remittances, so discriminates very clearly against such activities. These difficulties faced by direct foreign investment in the services sector in Brazil should be viewed with concern because international investment in services is taking up an increasing proportion of total direct international investment. Indeed, it has been repeatedly noted in recent asssessments of international investment trends that conventional forms of equity investment have become less important in an internationalized world where the technological frontier is moving very rapidly and that there should be an increased role for "new forms" of relationship between MNCs and host countries - joint ventures, licensing, technical assistance, marketing agreements and so on. The integrated MNC should remain a dominant force in technology and information-intense industries, but probably will not be the typical foreign investment of the future". Increasingly one should expect MMCs to reduce the cost and the commercial risks associated with direct investment and to play a role as intermediaries on the input side technological innovation) as well as on the output side (control over markets through marketing a agreements)".

In addition to the difficulties in handling transactions with intangible assets created by FIRCE, we may find the ones associated to the regime that governs technology transfer and foreign exchange and tax treatment given to royalty payments as determined in Profit Remittance Law. Since the original formulation of this Law, payment of royalties between head offices and subsidiaries has been banned (Article 14). The Law was concerned with tranfer of profits disguised as royalties, a legitimate concern if we consider that in 1985 some 81% of total receipts registered in the US balance of payment as royalties arose from percentage-based intra-firm transactions, probably at transfer prices. Yet the outright ban is somewhat radical solution for the problem; it might be more sensible to lift it and set up an "a posteriori" supervisory mechanism to monitor transfer prices.

Another set of issues relates to supervising technology transfer contracts. In view of imperfections in technology markets and assymetries in bargaining power between Brazilian and foreign companies, there is a clearcut need for Government intervention to lubricate channels of technology transfer and in function of antitrust considerations, insofar as that, at the international level, it is common for technology transfer contracts to be scrutinized on these bases, as they tend to conceal unfair business practices. Yet, INPI activities with registration of technology contracts have been very restrictive, due to its overlap with generic foreign exchange restrictions generated by the Central Bank. Interestingly, the INPI's operational criteria seem to have accurately transplanted the notion of locally-manufactured similar products to the field of intangible assets, and consequently the logic of import substitution, in assessing contracts covering kown-how, technical cooperation and specialized services. Basic criteria for approval of such contracts involve judging the local availability of similar services or technology. Thus, the INPI has been quite successful in saving foreign exchange by squeezing technology contract expenses down to negligible amounts - US$ 135 million in 1988 - which is a sad benchmark for a country that spends so little on technological development.

Acid criticisms have been leveled at INPI activities also in the area of trademarks and patents. It has been noted in this area that inadequate protection of intelectual property together with INPI practices has discouraged technology transfer and investment in research and development by Brazilian companies. This topic is complex and warrants a brief digression. In sectors where technological innovation takes place at a fast pace, and at the same time, inventors are not able to make full use of the fruits thereof, standards of competition are greatly affected by both the speed of technical progress and the existing opportunities for imitation through clones. In the OECD countries, there exists regulation of competition aiming at curbing unfair competition, in licensing and intellectual property regulations, as well as in the definition of standards and interfaces.

It is interesting to notice that these issues appear in even more complex form in the debate on international regulation of intellectual property, such as occurred in the Uruguay Round. In countries such as Brazil, where the technological infrastructure is precarious, there is much incentive to immitation where appropriability of innovations is small and development costs and risks are high. This gives rise to the natural trend towards the production of clones shown by high-technology industries in Brazil. However, besides the benefits conferred by the possibility of immitation, there are nevertheless costs involved in "dependent insertion" in the form of the risk of finding sources of immitation suddenly being cut off. This results in concern over "technological protectionism" by the developed countries, i.e., the possibility of increased appropriability of innovations through stricter systems covering intellectual property. The balance between costs and benefits of this "dependent" stance varies according to the sector in question. This can be illustrated by the contrasting responses given by Brazil to pressures to regulate intellectual property in the pharmaceutical industry and for software. In this latter case, where Brazil acceded to provide IPR protective regulation, Brazilian companies were already in a position to innovate (as barriers to entry in terms of overheads for research and development activities are small), and thus were eventually interested in protection of their own rights. In addition, they pushed to obtain protection mechanisms against foreign competition, such as: payment of royalties by foreign companies at the tourism foreign exchange rate, market reservation in distribution, approval of contracts in technology transfer agreements based on an examination of similarity with locally-manufactured goods, and approval of the import of single copies by Brazil's National Monetary Council.

In the case of pharmaceutical patents, however, even under the threat of strong trade sanctions by the US government, the Brazilian authorities did not alter their position of not recognising patents on pharmaceutical products and processes. In this case, Brazilian companies' market share was very small, and their technological efforts not suffciently large to make them interesed in protecting their results. It is clear that the possibility of copying hinders these efforts, as no company can make expenditures on development of technology in competition with powerful multinationals in a sector where there is a possibility of concentrating on clones. It is thus quite natural that recognition of patents is not in the interests of Brazilian companies in the sector, a phenomenon that is also noted in many other countries Thus, only recently, in a comprehensive deal aimed at reducing bilateral trade tensions with the United States at the onset of the Collor government, did Brazil agreed to recognise these patents.


6. Conclusions

The unilateral (domestic), minilateral (regional and hemispheric) and multilateral components of Brazil's trade strategy are not independent and must fit into a coherent whole. In the absence of radical dislocations in the world economy and of unexpected developments in the multilateral trade system, Brazil's trade strategy for the 1990's is bound to be conditioned by the extent and pace of her current trade liberalization experiment. It will be the success in sustaining the departure from the old barrage of foreign exchange controls and implementing the tariff cuts scheduled for the next few years which will (i) allow sustaining a positive - and, given Brazil's relative size, crucial - contribution to regional integration, as well as (ii) confer greater legitimacy to her participation in the GATT. By the same token, failure and reversal of domestic trade policy reform will delay regional integration and contribute to an increasing isolation of Brazil's influence on multilateral trade questions and, indeed, in other multilateral fora. The successful implementation of the domestic trade liberalization programme is, however, far from granted unless real exchange rate stability and sustained growth is achieved which, in turn, requires restoring macroeconomic stability.

It should also be noted that an active and positive participation in the regional, hemispheric and multilateral arenas of international trade policy presents no contradiction of objectives, but that, by virtue of Brazil's diversified pattern of trade and the overwhelmingly importance of non-Mercosur markets for Brazilian exports, concern with systemic issues and emphasis on multilateral principles of non-discrimination will continue to dominate Brazil's international stand on international trade policy issues.

Finally, as far a FDI is concerned, Brazil's long-standing legislation on profit remittances is not likely to change significantly. However, issues relating to non-equity investment and the related aspects of technology transfer should pass through important revision and modernisation.

7. References

M. de P. Abreu & W. Fritsch (1989) "Obstacles to Brazilian Export Growth and the Present Multilateral Trade Negotiations" in J. Whalley (ed.), Developing Countries in the World Trading System: country studies from a Ford Foundation Project (vol. 2). Macmillan, London, 1989.

R. Erzan & A. Yeats (1991), "Prospects for United States-Latin American Free Trade Areas. Empirical Evidence Concerning the View from the South", mimeo, The World Bank, Washington.

G. H. B. Franco (1992) "O regime cambial e a conta capital: notas sobre o caso brasileiro" Centro de Estudos Monetarios Lationamericanos, Mexico.

W. Fritsch (1991) "Brazil's Trade Strategies for the 1990s", paper presented at the International Forum on Latin American Perspectives, IDB/OECD Development Centre, Paris, 21st-22nd November.

_____ & E. M. Modiano (1988) "A restrição externa ao crescimento econômico brasileiro: uma perspectiva de longo prazo" Pesquisa e Planejamento Econômico 18 (2).

_____ & G. H. B. Franco (1992) "Política comercial no Brasil: passado e presente" Pensamiento Iberoamericano 21, enero-junio.

(1991a) "The reform of Brazilian industrial and trade policies: an interim assessment. A report prepared for ECLA", mimeo, Rio de Janeiro.

_____ & _____ (1991b) Foreign Direct Investment and Industrial Restructuring in Brazil: Its impact on industrial restructuring Paris: OCED Development Center.

_____ & _____ (1991c) "Efficient Industrialization in a Technologically Dependent Economy: The Current Brazilian Debate" in Competition and Economic Development Paris: OECD.

R. Iglesias (1991) "A Política Cambial da Argentina e do Brasil no Período 1970-89" in P. Motta Veiga (ed.) Cone Sul: a economia política da integração Rio de Janeiro: FUNCEX.

G. A. Maciel (1978) "Brazil's Proposal for the Reform of the Gatt System" The World Economy, 1 and 2, January.

R. B. Neves (1985) Exportações e Crescimento Industrial no Brasil Rio de Janeiro, IPEA-INPES.

_____ & H. C. Moreira (1987) "Os incentivos às exportações de manufaturados", Pesquisa e Planejamento Econômico 17(2).

OECD (1988) The Newly Developing Countries: challenge and opportunity for OECD industry Paris: OECD.