Paper for "Brazil as an Export Economy"
St. Anthony's College, Oxford
December 7, 1998
© Peterson Institute for International Economics
In this paper we ignore recent adverse events, notably the defeat of "fast-track" authority in the United States and the threatened financial crisis in Brazil. Our focus instead is on the longer term prospects for trade and investment between Brazil and the United States, assuming that the FTAA is ratified in 2005. We base our forecasts on the NAFTA experience of the past decade. We examine three aspects: the possible expansion of merchandise trade flows; the possible compositional change in Brazilian exports; and the possible growth of US foreign direct investment in Brazil. Our approach is to examine both developments in the NAFTA over a period of about a decade, and the comparative density in 1997 of trade and investment relations between the United States and Mexico, and the United States and Brazil. We conclude with a brief review of trade barriers that are most contentious.
Substantial economic integration already existed in North America prior to the entry into force of NAFTA in January 1994. Much of this economic integration reflected economic reforms and increased openness, especially in Mexico in the early 1990s, but also in Canada in the late 1980s. In a sense, the NAFTA in 1994, and the Canada-US Free Trade Agreement in 1989, were the icing on a range of pro-market macro economic and micro economic policies embraced by the participating countries. If and when the FTAA enters into force, it too will cap a longer and deeper commitment by Brazil and other Latin American countries to pro-market policies.
Mexican Exports to its NAFTA Partners
Growth of Mexican Exports
In 1996, almost all of Mexico's exports had the North American market as their destination. This is remarkable considering that a decade earlier only about 56 percent of Mexican exports were sent to Canada and the United States (Table 1).
Mexican merchandise exports to the United States increased by a factor of 3.9 between the 1982-86 average and the 1992-96 average, by comparison with a growth factor of 1.4 for Mexican exports to the rest of the world (excluding exports to Canada). A closer look at the annual growth rates of Mexican exports to the United States shows that, with the exception of 1995 and 1996, exports to the United States grew substantially faster than Mexican merchandise exports to the rest of the world.
A similar story can be told for Mexican exports to Canada: they increased rapidly from a very small base. Between 1982-86 and 1992-96, exports to Canada increased by a factor of 5.3. Much of the growth in Mexican exports to Canada occurred in the early 1990s, partly because of a "getting to know you" phenomena associated with the NAFTA process.
Composition of Mexican Exports
A decade ago Mexican exports to the United States were dominated by natural resources (Table 2). In the five-year period from 1982 to 1986, mineral fuels on average accounted for 52 percent of Mexican export to the United States. Adding another 12 percent for agricultural exports, exports based on natural resources made up 64 percent of total Mexican exports to the United States. During the same time period, only 30 percent of exports were manufactured goods.
A decade later the composition of Mexican exports to the United States looks quite different. In the period from 1992 to 1996, the dominant component of Mexican exports to the United States is manufactured goods—making up almost 77 percent of the total. Meanwhile the share of exports based on natural resources has diminished to 17 percent of the total.
A similar pattern emerges for Mexican exports to Canada (Table 3). A decade ago, in the period from 1982-1986, natural resources accounted for 74 percent of Mexican exports to Canada. Manufactures made up only about 21 percent of exports. A decade later compositional shares were practically reversed.
Canadian Exports to its NAFTA Partners
Growth of Canadian Exports
Between 1982-86 and 1992-96, Canadian exports to the United States grew significantly faster than Canadian exports to the rest of the world: the growth factors were 2.2 and 1.6 respectively (Table 4). However a closer look at developments shows that most of the differential growth occurred in anticipation and in the immediate aftermath of the Canada-US FTA. Indeed, Canadian exports increased by 48 percent between 1986 and 1992. By the time NAFTA was implemented the growth burst had passed.
Until the 1995 peso crisis, Canadian exports to Mexico grew significantly. The annual growth of exports was higher than the growth of exports to the United States or to the rest of the world. Over the decade as a whole, 1982-86 to 1992-96, the growth factor reached 3.5.
US Exports to its NAFTA Partners
Growth of US Exports
In 1997, Mexico surpassed Japan in becoming the United States' second largest trading partner after Canada.1 Since 1992, with the exception of the crisis year 1995, US exports to Mexico have grown substantially faster (an annual average growth rate of about 13 percent) than US merchandise exports to the rest of the world. The growth factor for the decade as a whole, 1992-96 vs. 1982-86, was much higher for US exports to Mexico than US exports to the rest of the world, 5.3 vs. 2.3 (Table 7). To a large extent, the rapid export growth reflected Mexican liberalization launched by President Miguel de la Madrid and continued by President Carlos Salinas, the sponsor of NAFTA.
For the United States and Canada, NAFTA represented an extension and acceleration of the Canadian-US FTA which entered into force in 1989. US exports to Canada grew about the same as US exports to the rest of the world (excluding Canada and Mexico), over the decade 1982-86 to 1992-96, shown by growth factors of 2.3 in both cases.
Composition of US Exports
The changing composition of US exports to Mexico over the past decade echoes the compositional shifts of exports in the other direction, but in a more subdued fashion (Table 8). The share of manufactures as a percentage of US exports to Mexico increased from 61 percent between 1982 and 1986 to 75 percent a decade later. Meanwhile, the share of exports based on natural resources decreased from about 13 percent in the 1982-1986 period to about 8 percent for 1992-1996.
Trade liberalization has not dramatically affected the composition of US exports to Canada. The composition of United States exports to Canada (Table 9) as well as Canadian exports to the United States (Table 6) have changed only slightly between 1982 and 1996. Manufactures dominated total trade between these two countries for both periods under consideration.
Foreign Direct Investment in Mexico
Prior to the conception of NAFTA, US firms had invested heavily in Mexico, encouraged by the liberalization of Mexico's trade and investment regimes beginning in the 1980s. By the time NAFTA was negotiated and entered into force, a substantial amount of cross-border regional production integration had already taken place. Thus, regional integration in North America was as much investment-led as policy-led. In fact business support was a critical driver in both the Canada-US FTA and NAFTA. This support was predicated as much on investment possibilities as on trade expectations. Moreover, the sharp growth in trade integration between Mexico and the United States was propelled by foreign direct investment.
The stock of FDI by the US firms in Mexico increased by a factor of 5.1 between 1982 and 1997 (Table 12). However, over a similar 15-year period, 1980-95, world FDI in Mexico increased by a factor of 7.9. By contrast with trade growth, direct investment under the NAFTA was not particularly "loaded" towards North America. However between 1982 and 1994, US affiliates operating in Mexico increased their share of Mexican exports to the United States threefold, to reach over one fifth of total Mexican exports (Table 13). Nearly all of these exports were shipped on an intra-firm basis.
As early as the 1970s, US auto firms operating in Mexico were pushed by various Mexican decrees to increase their intra-firm exports. However, it was not until trade liberalization began to take hold in Mexico in the late 1980s that intra-firm trade expanded rapidly in the auto sector and elsewhere. This expansion in turn propelled further trade and investment liberalization in the 1990s.
The Prospects for US - Brazil Trade and Investment
Growth of Brazil - US Trade
Brazilian exports to the United States increased by a factor of 1.3 between 1982-86 and 1992-96 (Table 10). This modest increase in part reflected trade liberalization in Brazil—the restructuring of the economy away from import-substitution towards export-led growth. Over the same period, however, Brazilian exports to Latin America increased at a much faster rate than exports to the United States: for Latin American destinations, the growth factor was 3.8. Faster growth mirrored economic integration between Brazil and its Mercosul partners, Argentina, Uruguay and Paraguay. Thus, by 1992-96 period, 24 percent of Brazilian exports were absorbed by Latin American countries, while 20 percent of exports were destined to the United States.
The experience of Mexican-US trade under NAFTA suggests that realization of an FTAA could lead to a growth burst of Brazilian exports to the United States (and vice versa). Two-way trade might double over levels that would otherwise be reached. The growth spurt could last for about five years; possibly beginning a few years before the FTAA was signed.
Composition of Brazilian exports to the United States
In terms of trade composition, Brazil is halfway between Canada and Mexico. Brazil started the 1980s with a higher share of manufactured exports to the United States than Mexico, and the subsequent compositional swing was more muted. The share of manufacturing exports rose from 48 percent of total Brazilian exports to the United States in the 1982-86 period to about 69 percent a decade later (Table 11). Meanwhile, the share of exports based on natural resources decreased from 40 percent to about 16 percent.
The experience of Mexico suggests that, with FTAA, the swing towards manufactures in US-Brazil trade would become more pronounced. The most difficult sectors in the FTAA negotiation will probably be agricultural products, especially orange juice, sugar, and tropical produce. But over the long haul, the big merchandise trade growth will be in manufactured items.
Foreign Direct Investment in Brazil
Foreign direct investment in Brazil has expanded sharply, but not as fast as in Mexico. The stock of US FDI in Brazil expanded, by a factor of 3.8 between 1982 and 1997, while the total FDI stock from all countries expanded by a factor of 5.6. The historical growth comparison with Mexico suggests that an FTAA accord could perhaps lead to a 30 percent expansion of FDI in Brazil, both from the United States and other home countries.
Recent trade policy and investment strategy changes in Brazil will promote investment. Brazil and Argentina alike see the Mercosul as a means of increasing investment, both from domestic and international sources. Since taking office on January 1995, President Cardoso has emphasized privatization and deregulation and has ended legal discrimination against foreign-owned companies. One of President Cardoso' s goals is to entice foreign companies through improved legal, political and administrative structures.
US affiliates operating in Brazil nearly doubled their share of Brazilian exports to the United States between 1982 and 1994, to about one fifth of the total (Table 15). Almost all of the exports by United States affiliates are on an intra-firm basis (in 1994, the figure was 94 percent). This indicates that many of the United States transnational corporations in Brazil have integrated their operations between plants in Brazil and the United States. So far, however, FDI in Brazil has neither driven US-Brazilian trade nor prompted a strong push for the FTAA.
Comparative Trade and Investment Densities—Brazil and Mexico
In 1997, two-way trade between the United States and Mexico was six-times larger than two-way trade between the United States and Brazil (Table 17). The respective sizes of the Brazilian and Mexican economies do not account for this huge difference in two-way trade with the United States. In fact, GDP and GDP per capita taken by themselves would point to larger US-Brazil than US-Mexico trade—for the simple reason that Brazil has the larger economy and the higher per capita GDP. However, apart from preferential trade liberalization under the NAFTA, other factors help explain the fact that US-Mexican two-way trade flows are much larger than US-Brazilian flows.
In a recent IIE study, Jeffrey Frankel used a gravity model to identify several important factors that contribute to larger bilateral trade flows: in addition to GNP and GNP per capita, other important determinants include: geographical proximity, a common border, a common language, and a common trade bloc.2 Based on Frankel's gravity model, the difference in distance between the United States and Mexico, on the one hand, and the United States and Brazil on the other, would account for two-way US-Mexico trade 3.37 times larger than two-way US-Brazil trade. The common US-Mexico border would account for a further increase in US-Mexico trade of 1.65 times US-Brazil trade. However, the larger Brazilian GDP would mean that Brazilian two-way trade would be 1.94 times larger, and the higher per-capita Brazilian GDP would imply an additional increase of 1.03 times US-Mexican two-way trade.
All-in-all, adjusting for these assorted differences in size and geography, Frankel's gravity model parameters predict that two-way US-Brazil trade flows would be $56 billion if Mexico were suddenly attributed with all the characteristics of Brazil but kept its membership in the NAFTA.3 The fact that US-Brazilian trade flows were in fact $25 billion in 1997 suggests that the FTAA-effect is potentially quite large—perhaps a doubling of trade.
By contrast with the large difference in potential US-Brazil trade densities with an FTAA, there is rather little difference in potential investment densities based on current comparisons between Brazil and Mexico. Per unit of GDP and per capita, US FDI in Brazil is already similar to US FDI in Mexico. Of course the Brazilian economy is twice as large as Mexico's, and Brazil is the hub of the large Mercosul region. With an FTAA, some further expansion of investment could be expected, but the investment boost would not likely be as large as the trade boost. However, one might expect US and other foreign affiliates based in Brazil to increase their shipments to the United States and Europe.
Brazilian Concerns about US Trade Barriers
While the United States has relatively low tariffs by world standards, Brazil has long complained that US trade barriers rank among the highest in sectors where Brazil has a comparative advantage, notably steel, textiles, footwear, citrus, tobacco and sugar. Brazilian officials contend that US trade barriers cost the country as much as $5 billion in lost exports every year—a very large figure if correct.4
Among Brazil's chief concerns are US quotas and non-tariff barriers affecting agricultural goods. In 1997, the US quota system for sugar imports allocates 14 percent of the total imports (300,000 tons) to Brazil, making it the second largest exporter of sugar to the United States after the Dominican Republic. However, this is still a fraction of Brazil's annual production of over 14 million tons and Brazilian sugar exporters estimate that, if the quota were lifted, Brazilian sugar exports could increase tenfold from 300,000 tons to about 3 million metric tons annually.5 US tariffs on frozen concentrated orange juice (FCOJ) have long been an issue of friction between the United States and Brazil. Since 1995, US most-favored-nation (MFN) tariffs on FCOJ have been cut as a result of negotiated Uruguay Round commitments in which the US agreed to reduce the tariff by 15 percentage points. That still leaves the MFN tariff at about 15 percent ad valorem. Consequently, Brazil has lost market share to Mexico in imports of FCOJ due to NAFTA's preferential tariff treatment. Following the passage of NAFTA in 1994, Brazil's share of the US orange juice market decreased to 84 percent and Mexico's increased to nearly 12 percent. A FTAA would eliminate the tariff advantage given to Mexico under NAFTA and allow Brazil to recapture some of the FCOJ market. Various US sanitary and phytosanitary measures pose additional barriers to Brazilian agricultural exports.
Because many of Brazil's chief complaints about US trade barriers pertain to agriculture, Brazilian officials pushed for an agriculture negotiating group in the FTAA talks. Despite opposition from the United States, Brazil was successful in securing a specific working group on agriculture at the San Jose Ministerial meeting in March 1998.
Other sources of friction are the US antidumping and countervailing duties on imports from Brazil, in particular iron and steel. Although the United States has introduced no new antidumping and countervailing duty orders against Brazil in recent years, many duties imposed in the 1980s and early 1990s remain in place (Tables 19 and 20) and new duties could be imposed as a result of the recent round of steel antidumping cases. Brazilian authorities claim that antidumping measures on steel alone cost the industry as much as $80 million in duties annually, and probably a much larger figure in lost exports.6 The Brazilian government objects to the antidumping orders, arguing that, following privatization of the steel industry in 1993, Brazilian steel prices are low not because of subsidization but because of low production costs.
US concerns about Brazilian trade barriers focus on relatively high tariff rates, but also include non-tariff barriers in the areas of government procurement, investment, services, and information technology. The US negotiating strategy has called for early tariff cuts in the FTAA. Given its concern over US non-tariff barriers, however, Brazil objects to early tariff cuts in isolation from an over-arching agreement on issues such as antidumping, dispute settlement, and subsidies.7 For the moment, no over-arching agreement is in sight and trade negotiators are preparing the groundwork for a major push after the year 2000.
As Brazil and the United States get into the thick of FTAA talks, the distinction between trade glue and trade growth should be remembered. The trade glue of the FTAA will be the resolution of difficult sectors and topics, illustrated by FCOJ, sugar, auto tariffs and antidumping duties. There is an imperfect overlap between these sectors and topics on the one hand and the likely realm of FTAA-inspired trade growth, namely manufactured goods and business services.
Tables 1-20 [pdf]
2. Frankel, Jeffrey A. 1997. Regional Trading Blocs in the World Economic System. Institute for International Economics. Washington, D.C.
3. This figure is calculated as actual US-Mexico 1997 two-way trade ($157 billion) divided by 3.37 (for the distance difference as between Kansas-Mexico City, 2,000 kilometers, vs. Kansas-Sao Paulo, 10,500 kilometers); divided by 1.65 (for the common US-Mexico border) times 1.94 (for the larger Brazilian GDP, see Table 16) and times 1.03 (for the higher Brazilian GDP per capita, again see Table 16). Chain-multiplication gives a factor of 2.78 for dividing actual 1997 US-Mexico two-way trade to calculate hypothetical trade if Mexico took on all the attributes of Brazil (but kept its membership in NAFTA). These calculations are based on the coefficients in the last column of Frankel's table 4.2, pp.62-63, op.cit. The result is $56 billion of hypothetical two-way US-Brazil trade if Mexico were geographically and economically transplanted to Brazil, but remained in NAFTA.