VIII Export Specialization and Economic Growth

Eliot Kalter, Steven Phillips, Manmohan Singh, Mauricio Villafuerte, Rodolfo Luzio, and Marco Espinosa-Vega
Published Date:
June 2004
  • ShareShare
Show Summary Details
Mauricio Villafuerte

Claims in the recent economic literature that natural resource-based exports have a negative impact on long-term economic growth have raised some concerns about the structure and evolution of Chile’s exports.1 Exports in Chile have been very dynamic in recent decades but have decelerated somewhat in recent years, while the structure of exports continues to be dominated by natural resources. This section asks whether the structure of Chilean exports has worked in favor of or against a sustained economic growth and looks into how the country could continue to improve its current “export model.” It argues that

  • Chile has not been subject to the “curse” of natural resources endowment, because it has avoided the main factors that the economic literature identifies as reasons for a negative impact of natural resource-based exports on economic growth.

  • Chile has developed several export industries based on its natural resource endowments, with important spillovers to economic activity, a combination of static comparative advantages with knowledge and innovation.

  • To increase overall productivity and the diversification of Chilean exports, government policies should concentrate on improving education levels.

  • The government should continue to deepen its trade liberalization efforts, with the current “open regionalism” strategy being an adequate mechanism to increase the market access of Chilean exports.

  • The recent weak export growth can be traced to the difficult external environment.

This section first examines the reasons why natural resources in Chile have not been detrimental to growth, the factors behind the recent evolution of exports, and the link between exports and growth in Chile. It concludes by discussing potential trade specialization patterns through the creation of new comparative advantages.

Exports of Natural Resources and Growth in Chile

Can natural resource-based exports help sustain economic growth in Chile? This question is motivated by (1) an economic literature that reports a negative impact of natural resource abundance on economic growth rates (due in part to limited spillovers from natural resource-based export products); and (2) the recent deceleration of export growth in Chile. This section presents arguments that contrast these facts with the Chilean experience.

The “Curse” of Natural Resources

The idea that natural resource endowment could work against economic development is not a new one, but has received increased attention in recent years in light of cross-country empirical research. In particular, Sachs and Warner (1995) found a negative relationship between natural resource exports and growth rates of economic activity, though their findings are not uncontroversial.2

The explanations offered by the theoretical literature that postulates a negative impact of natural resource exports on growth can be broadly grouped as follows:

  • Political economy dynamics by which interest groups fight to capture the rents from natural resources. These “voracity effects” (which are particularly acute in the case of government-controlled natural resources) reveal themselves in and “rent-seeking” distortionary activities, economic inefficient policies in taxation, general,3 and lead to a bad allocation of resources and, hence, lower economic growth.

  • Explanations focused on the productive structure of the economy, stressing the inability of natural resource exports to generate key linkages among activities generating spillovers on aggregate output. The development literature of the 1940s and 1950s made the case against resource-based growth on (1) the premise of a secular decline in world prices of primary exports relative to manufactures; (2) poor potential for productivity growth of natural resource sectors; and (3) small “forward and backward linkages” from primary exports to the rest of the economy.4

  • Dutch Disease. A boom in natural resource exports leads to an appreciation of the real exchange rate that in turn produces a reallocation of factors of production away from other tradables. In the long run, this process would increase the dependence on natural resource exports and, hence, limit the sources of economic growth.5

The Chilean Case

Chile has not been subject to the factors identified by the literature as a “curse” of natural resources, namely (1) the exports of natural resources have not crowded out other promising sectors; (2) the relative importance of nonprocessed natural resource exports has consistently fallen; (3) productivity in natural resource-based sectors has increased in the last 25 years; and (4) new export products have been developed, partly on the basis of technological spillovers from natural resource exports (as will be shown below).

Indeed, the recent copper mining boom (in the early and mid-1990s) did not have a negative impact on other export sectors. According to several studies, it produced a very limited crowding out of investments from other sectors and, in fact, had an overall positive impact on the Chilean economy.6 A key factor behind this outcome was the opening of the copper industry to foreign investment, which has expanded Chile’s mining production and has avoided the crowding out of (limited) domestic capital from other sectors. At the same time, the existence of foreign ownership has helped to limit the impact of increased copper exports on the real exchange rate (Dutch Disease) and to share the risks of international price fluctuations. Another positive factor for Chile has been the application of a prudent fiscal policy that has led to increased government savings in times of high copper prices. The latter attests to the strong institutional framework in place in Chile to prevent the capture of economic policy by interest groups.7

There has been a steady (though slow) diversification of Chilean exports away from nonprocessed natural resources.8 Overall, there is a considerable preponderance of natural resource-based exports, but there has been a steady increase in the share of products with a more technological content due to the diversification of the country’s export basket. In fact, an analysis of merchandise exports reveals a steady but slow diversification over time (Table 8.1).9 The Herfindahl index on export concentration is still higher in Chile than in representative Latin American countries (Table 8.2), but the gap has been falling. In addition, Chile’s export markets have diversified over time, with an increased share from the United States and Asia (Table 8.3). Furthermore, Chile has displayed some ability to reallocate its exports from less dynamic to more dynamic markets.10

Table 8.1.Evoluation of Export Concentration Indicators
Number of items > 0.5 percent of total exports23242828
Share in exports of
Top 5 items67.561.756.554.5
Top 10 items77.674.567.265.9
Top 20 items86.685.578.376.0
Herfindahl index0.
Sources: UN COMTRADE database; and author’s calculations.
Sources: UN COMTRADE database; and author’s calculations.
Table 8.2.Selected Countries: Herfindhal Index of Export Concentration
Sources: UN COMTRADE database; and author’s calculations.
Sources: UN COMTRADE database; and author’s calculations.
Table 8.3.Chile: Evoluation of Export Market Shares(In percent of total)
Average Annual
198019902001Growth Rate
Latin America24.412.522.56.1
United States12.617.019.68.8
European Union38.738.125.74.5
Sources: Central Bank of Chile; and author’s calculations.
Sources: Central Bank of Chile; and author’s calculations.

Recent Evolution of Chilean Exports

One of the concerns about the potential contribution of natural resource-based exports to future growth in Chile refers to the recent slowdown in export growth, and whether it signals structural factors that will also limit export growth in the future.

In fact, the recent evolution of noncopper exports can be clearly linked to the trends of its standard determinants, and there has not been any anomalous break in export behavior. GDP growth of export partners and export prices started to decelerate in the mid-1990s, and the Asian crisis clearly affected Chile’s export growth.11 By contrast, the real exchange rate appreciation in the early 1990s was linked to a deceleration in noncopper exports, but the depreciation in the late 1990s would have helped the recovery in noncopper export growth rates. Noncopper exports and investment flows also appeared to be closely interrelated. A simple econometric estimation of a noncopper export function (Table 8.4) suggests that the weaker export growth in recent years can be explained by the deterioration in external demand and export prices (more than offsetting gains from a more depreciated real exchange rate).12

Table 8.4.Regression on Noncopper Export Growth
Dependent variable: D (X_NC)
Method: Least squares
Sample (adjusted): 1978–2002
VariableCoefficientStandard Errort-StatisticProb.
D (Y_XB (−1))4.2271.0124.1760.001
D (REER (−1))−0.3750.138−2.7140.014
R-squared0.808Mean dependent var0.086
Adjusted R-squared0.743S.D. dependent var0.095
S.E. of regression0.048Akaike info criterion−2.990
Sum squared resid0.042Schwarz criterion−2.649
Log likelihood44.380F-statistic12.591
Durbin-Watson stat2.339Prob (F-statistic)0.000
Residual tests: Breusch-Godfrey = 1.31 (0.52); ARCH = 0.78 (0.38); Jarque-Bera = 2.48 (0.29).
Source: author’s calculations.
Source: author’s calculations.

That behavior, however, does not imply that Chile should passively accept changes in external market conditions. International markets are very dynamic, with increasingly tougher demand requirements. This issue is discussed below.

Exports and Growth

One of the concerns raised by the literature on natural resource-based exports is founded on a perception of limited links with overall economic activity. This section examines this issue both from an aggregate perspective and by looking at sectoral-level evidence in Chile. The latter offers the strongest evidence of the positive link between natural resource-based exports and economic growth in Chile.

Economic theory argues that openness should increase the level and growth rates of income,13 however the empirical evidence on the causality link between exports and growth is mixed. Export expansion to foreign markets, by improving resource allocation and production efficiency, can raise the steady-state level of income. On the other hand, exports can become a transmission channel for externalities due to the increased exposure to foreign markets. The endogenous growth models emphasize concepts such as diffusion of technology and learning by doing as mechanisms that would allow countries to achieve higher steady-state growth rates. Thus, exports allow access to imported capital goods with the latest technological improvements and are themselves forced to innovate to keep and expand access to increasingly demanding foreign markets and create so-called dynamic comparative advantages.14 The empirical evidence on a cross-country basis tends to confirm this link,15 but it is less conclusive on a country-by-country basis due in part to endogeneity between exports and GDP.

In Chile, exports of goods and services have contributed annually about 2 percentage points to GDP growth since 1974, with the contribution climbing to about 2 ½ percentage points since the 1990s (Table 8.5). The correlation between exports and GDP growth rates was very high only in the early 1990s, when investment growth rates were historically high. By contrast, in part of the 1980s there was a boost to exports coupled with internal adjustment as a result of the debt crisis, and in recent years the deterioration in terms of trade and in access to international financial markets limited overall GDP growth. The contribution of exports to GDP growth in Chile partly explains the positive growth differential between Chile and other Latin American countries in recent decades (Figure 8.1), which in turn is the result of Chile’s significantly larger share of exports in GDP. However, while Chile’s integration to the world economy is high by regional standards, the country still lags behind fast-growing economies in South-East Asia (Table 8.6).

Table 8.5.Real GDP and Export Volume Growth, Average Per Period(In percent)
Real GDPExport GoodsExport of Goods

and Services

to Growth
Source: Central Bank of Chile.
Source: Central Bank of Chile.

Figure 8.1.Differences Between Contribution of Exports to Growth in Chile and Largest Latin American Economies

(Three-year moving averages)

Source: IMF, World Economic Outlook.

Table 8.6.Selected Economies: Exports and Total Trade Ratios to GDP, Average 1998–2000(In percent)
X/GDPX + M/GDPPer Capita GDP

(In thousands of

U.S. dollars)
New Zealand244714,080
Taiwan Province of China448313,057
Czech Republic501045,301
Source: IMF, World Economic Outlook.
Source: IMF, World Economic Outlook.

As regards the export-led growth hypothesis, a basic cointegration analysis for the period 1975–2002 could not confirm its validity for Chile. Looking at the evolution of exports and GDP over time it can be noticed that in certain periods a strong expansion in exports coexisted with weak nonexportable production; more recently, the expansion in export volumes has not produced a significant rise in employment because technological improvements increased average labor productivity and has been accompanied by sharp falls in export prices (which affected national income growth).16 An econometric evaluation detected a cointegrating relationship for a system comprising GDP, gross capital formation, and volume of exports, but exports of goods were not found to be weakly exogenous in that estimation (Table 8.7).17 These results suggest a (long-run) feedback from output (and investment) to exports.18

Table 8.7.Chile: A Cointegrating Analysis of GDP and Exports
Null Hypothesis Summary Test Statistics
r = 0r = 1r = 2
Max-Eigenvalue statistic27.38*10.407.66
5 percent critical value25.5418.9612.25
Trace statistic45.45*18.067.66
5 percent critical value42.4425.3212.25
Weak exogeneity test statistics
LR statistic1.935.02*5.03*
p value[.16][.03][.02]
Multivariate statistics for testing stationarity
LR statistic16.24**14.13**13.69**
Statistics for testing significance of a given variable
LR statistic13.42**11.44**13.13**
Note: The VAR includes two lags on each variable, an intercept, and dummies for 1982 and 1985. The estimation period is 1974–2002. *(**) denotes rejection of the hypothesis at the 5 percent (1 percent) level.Source: author’s calculations.
Note: The VAR includes two lags on each variable, an intercept, and dummies for 1982 and 1985. The estimation period is 1974–2002. *(**) denotes rejection of the hypothesis at the 5 percent (1 percent) level.Source: author’s calculations.

However, sectoral-level data offer the strongest evidence of the link between exports and economic growth in Chile based on spillovers from certain natural resource-based exports to other levels of their production chain and to the development of new export products. A case in point within the agriculture sector is the export of fruits,19 where Chile was able to transfer, adapt, and extend technologies developed in other countries (with initial assistance from the Corporacion de Fomento de la Produccion (CORFO)). Afterward, the private sector carried out further innovations at all levels of its production chain (particularly backward linkages) and also disseminated those innovations to other sectors with export potential.

After the trade liberalization process, the manufacturing sector undertook an extensive restructuring centered on (mainly) simpler-technology activities linked to Chile’s natural resource endowments.20 Evidence at the firm level suggests that there had been some technological dynamism, especially in product “engineering” and in adaptation to increasingly rigorous international demand requirements thanks to specialized human capital and the support of specialized institutions (such as the private nonprofit organization Fundacion Chile). Specifically, Fundacion Chile’s initiatives for the development and transfer of new commercial technologies have contributed to the dynamism of agro-industry, fishing, and forestry sectors, including the salmon industry.21 The latter expanded rapidly in the mid-1990s together with its long chains of upstream and downstream activities (e.g., to feed the fish and distribute medicines, specialized equipment, and rafts in which salmon are grown). A similar process has marked the development of the wine industry, another high-technology sector based on foreign investment and technologies that has gained international recognition and has increased its exports from less than US$100 million in the early 1990s to about US$650 million in 2003.

The Development of New Comparative Advantages

How can the value and diversification in Chilean exports be increased? This section argues that the combination of static comparative advantages with knowledge and innovation should lead to increased value added and diversification, together with a stable macroeconomic framework that provides the right incentives to develop growth-enhancing export sectors. Government supporting policies should be concentrated on improving education levels and on deepening the trade liberalization efforts, particularly by ensuring the access of Chilean products to world markets.

Chile’s specialization in natural resource-based exports is in line with its comparative advantage. Chile has plentiful natural resources distributed in extensive latitudinal and altitudinal ranges, a small domestic market, and high transport costs to major international markets. Those structural characteristics have initially led to a specialization in goods with relatively simple production technologies and with relatively high value-added per unit weight.22 At the same time, and as reviewed in the previous section, there has been a gradual increase in the technological content of exports, and this process has led to the inception of more (and new) differentiated products for international markets. Nonetheless, Chile’s export basket varies markedly depending on the geographic destination:23 there is more natural-resource content in the exports to industrial countries. The latter might be explained by similar natural resource endowment with nonindustrial countries but also by a positive correlation between tariffs and value-added content in Chile’s export markets.

There would seem to be benefits from developing new comparative advantages to further diversify the export base and to better respond to changing world demand patterns. There is broad agreement in the literature on the negative correlation between export concentration and growth, partly on account of larger volatility in terms of trade and in the real exchange rate. Even though comparative advantages are key determinants in export patterns, they are not necessarily static and can evolve over time based on the relative supplies of production factors and relative productivity levels. De Ferranti and others (2002) stress that it is not so much what a country produces, but how it is produced that matters, and that production factors such as knowledge and education should help raise productivity growth, even if a country specializes in “traditional” sectors. In the presence of abundant natural resources, high levels of human capital might not only avoid the crowding out of productive factors to other sectors, but even lead to increased rates of growth, since a well-educated labor facilitates the movement of workers across economic activities and allows the inception of new industrial activities linked to natural resources (Bravo-Ortega and De Gregorio, 2002).

International experience shows the key role of natural resources in the economic success of some natural resource-abundant countries. Many industrial countries based their initial development (and to some extent continue to do so) on their abundant natural resources. The United States’ industrial success can be considered as one of the most natural resource-rich nation that made a gradual transition to resource-rich manufacturing industries.24 Exploitation of minerals in the United States, as in Australia more recently, was the main driving force of growth and industrialization for more than a century.25 The Nordic countries (which share some characteristics with Chile in terms of geography, market size, and export orientation) also have become highly competitive exporters of manufactures: Sweden and Finland have become leading exporters of telecommunications equipment; Norway has specialized in engineering and shipping services. Two closely interrelated factors stand out behind the success of their development strategies: openness of the economies and high investment in human capital. The latter helped those countries to absorb technological progress to transform resource-based activities into industries with higher productivity levels.

By contrast, inward-oriented policies and high tariffs have not helped economic growth. Irwin (2002) shows that inward-oriented policies and high tariffs were not a critical factor behind the late nineteenth century growth experience of high-income countries. Noland (2001) suggests that industrial policies made at most a minor contribution to the recent growth experience in East Asia, and that most of its success came from good macroeconomic policies, export orientation, and investment in human capital and in efficient social infrastructure. In Latin America, industrial policies and/or import substitution seemed to work at the beginning but started to disintegrate by the 1960s, partly because of the pressures of interest groups to keep incentives indefinitely; furthermore, several Latin American countries used receipts from natural resource exports to finance industrial policies and protectionism, causing a systematic overvaluation of local currencies and a misallocation of resources that might explain the negative association of natural resource abundance and growth rates.

Accumulation of Human Capital and Technological Innovation

Chile’s market-oriented policies,26 its stable macroeconomic framework, and its open foreign direct investment (FDI) regime should continue to provide adequate incentives to develop growth-enhancing export sectors, but the country could make the most out of those exports by developing abilities to innovate and absorb innovation. The general framework of policies provides incentives for an efficient allocation of resources. However, the technological content of Chile’s export basket is still not very high. A channel to absorb technology and to innovate is FDI; in this area Chile has been actively signing bilateral treaties with its main investor countries and regions, while the free trade agreements with large external partners could help Chile become a “hub” for multinational corporations to access larger neighboring Latin American countries.27 However, there is evidence that foreign companies establishing in outward-oriented countries tend to use production technologies that fit the host country’s comparative advantage.28

The accumulation of human capital is the key channel of technological innovation and assimilation, and an essential requirement for a succesful exportled economic growth.29

  • There has been significant progress in educational attainment of the labor force, but even though Chile is in the upper range of educational attainment of Latin America’s labor force (7.9 years of schooling in 2000, Figure 8.2), it still lags compared with natural resource-abundant member countries of the Organization for Economic Cooperation and Development (OECD) (11.1 years) and East Asian tigers (9.7 years).30

  • Schooling attainment levels do not reveal the adequacy of education: a study on adult literacy skills31 showed that 50 percent of Chile’s labor force has a low level of basic reading comprehension and only 20 percent achieved the level considered “adequate.”32

  • An analysis of the quality of current education based on the results of the 1999 international study of student achievement in mathematics and science (TIMSS) suggests relatively low quality (even adjusting for income levels, Figure 8.3): Chile ranked thirty-fifth among the 38 countries that participated in the study.33 Part of the poor outcome in international tests can be traced to Chile’s higher income inequality (Figure 8.4).34

Figure 8.2.Selected Economies: Educational Attainment and Income

Sources: Barro and Lee (2002); and IMF, World Economic Outlook.

Figure 8.3.Selected Economies: TIMSS Mathematics Results and Income, 1999

Sources: Trends in International Mathematics and Science Study (1999); and IMF, World Economic Outlook.

Figure 8.4.Selected Economies: TIMSS Mathematics Results and Inequality in Income, 1999

Sources: Trends in International Mathematics and Science Study (1999); and World Bank, World Development Indicators.

The government could help improve the quality of education by updating curricula, improving teachers’ skills, tracking performance, and improving accreditation systems.35 Total (public and private) spending is high compared with countries with similar income levels. The fact that Chile spends relatively more but achieves lower outcomes than, for example, Mexico, suggests that the emphasis should be placed on increasing efficiency of spending and quality of education. The Chilean government has been undertaking several efforts in this area, including an increased focus on preschool education, increase in school hours, the review of curricula, and the introduction of standardized admission tests to tertiary education. The World Bank is also assisting in the design of programs to better match labor force skills with the needs of a knowledge-based economy. Increased accountability by public (municipal) and subsidized (through the voucher system for education) private schools should help tackle educational inequalities across income strata; an initiative that might also be considered is to make variable the value of vouchers to broaden students’ choices and increase competition within the educational system.

Some other microeconomic reforms might also help encourage private sector innovation.36 The government could transfer most of its research execution (often disconnected from the needs of the private sector) to the private sector (through bidding processes), strengthen intellectual property rights, and improve patent approval procedures.

Appendix I: Evolution of Exports in Chile

Exports in Chile have been very dynamic in recent decades, but have decelerated somewhat in recent years. Since 1980, the value of exports has tripled in real terms and the volume of exports has quintupled, substantially above world trends. The increase in the volume of exports has averaged more than 9 percent in the past 30 years, and has been broadly based (Table 8.8). However, the expansion of exports has decelerated recently. Growth in export value (in real terms) has been far more volatile owing to sharp fluctuations in export prices.

Table 8.8.Export Volume and Average Growth Rates(In percent)
Agriculture, hunting, forestry, and fishing12.08.44.2
Total exports9.010.27.1
Source: Central Bank of Chile.
Source: Central Bank of Chile.

The structure of exports has been dominated by natural resources. As shown in Table 8.9 (which includes refined copper as a nonprocessed natural resource, ignoring embedded technological improvements), exports of natural resources and processed natural resources have accounted for more than 85 percent of total exports.

Table 8.9.Export Structure(In percent of total)
Natural resources74.870.257.551.1
Processed natural resources18.225.433.534.6
Other industrial products6.
Sources: Central Bank of Chile; and author’s calculations.
Sources: Central Bank of Chile; and author’s calculations.

The performance of exports has partly resulted from an export-oriented policy framework that included an aggressive trade liberalization process, a competitive real exchange rate, and several (temporary) export support schemes.

  • Starting in the mid-1970s, exports expanded following an aggressive trade liberalization process (through reduced tariffs and elimination of quantitative restrictions) and the utilization of excess capacity in the export sector.

  • After the debt crisis of the early 1980s, exports experienced an increased dynamism on account of a substantial real depreciation and the introduction of several export support schemes, including subsidies (simplified drawback system for nontraditional exports, delayed payment of tariffs on imported capital goods, and tax incentives for the forestry sector), financing facilities (through CORFO), institutional support for export activities (ProChile and Fundacion Chile), and incentives for foreign direct investment (Chapter 19 on debt-equity swaps).

  • In the 1990s, exports benefited from the consolidation of a sound macroeconomic framework and a trade policy strategy geared toward the negotiation of regional trade agreements to improve market access for Chilean exports (in the context of the phasing out of export subsidies as agreed with the World Trade Organization (WTO)).

Appendix II: Chile’s Trade Policy Strategy and Recent Agreements

Starting in the mid-1970s, Chile pursued an export-oriented development strategy, eliminating the antiexport bias of trade policy by means of aggressive unilateral trade liberalization.37 Tariffs were lowered sharply and set at a uniform rate for all imports except a few agricultural products that were subject to price bands. The uniform tariff was reduced from 15 percent to 11 percent in June 1991, and even further to 6 percent over the period 1999–2003. The unilateral trade liberalization—together with macroeconomic stability, a “competitive” real exchange rate, and several specific export promotion programs38—contributed to the significant expansion and diversification of exports.

Chile has actively participated in multilateral trade negotiations within the framework of the WTO and in 1995 incorporated the WTO Agreements into its domestic legislation. The government considers that subscription to the WTO, despite limited progress in the liberalization of important sectors (e.g., agriculture, textiles, and clothing), benefits Chile’s exports by strengthening the dispute settlement mechanism, adopting a common set of rules that reduce trade distortions, improving the quality of information regarding market access, and promoting investment flows through clear rules for trade in services and in intellectual property matters.39 Domestically, the adoption of WTO-consistent legislation should reduce the possibility of adopting discriminatory protectionist policies and reinforce trade-related institutions. The WTO-tariff bound agreed to by Chile is 25 percent for most products, 31.5 percent for some agricultural products (wheat, wheat flour, oils, and dairy products), and 98 percent for sugar.

Since the early 1990s, Chile’s trade policy has focused on the negotiation of a number of regional (both bilateral and plurilateral) integration agreements (RIAs). This policy has been characterized by the authorities as “open regionalism,” complementary to the multilateral trade strategy and aimed at ensuring niches for Chilean exports.40 Under the auspices of the Latin American Integration Association (LAIA), Chile signed economic complementarity agreements with Mexico (1991), Republica Bolivariana de Venezuela (1993), Colombia (1993), Ecuador (1994), Mercosur (1996), and Peru (1998), and a partial-scope agreement with Bolivia (1993).41 Chile and signed concluded a free trade Canada agreement with free for negotiations trade agreements in 1996,42 with the European Union (EU), the United States, and Korea in 2002, and with the European Free Trade Association (EFTA) in early 2003.

The Rationale for Regionalism

In the economic literature, regionalism entails second-best type policies and no general theorem exists to support it. According to the standard trade theory, there is no additional benefit that could not be achieved through unilateral liberalization.43 Assuming perfect competition, small economies, constant returns of scale and no transport costs, efficiency in production and consumption is achieved when tariffs (and nontariff barriers) are zero, even if trading partners do not reciprocate. In this context, increased exports from RIAs do not increase welfare as the resulting producer surplus would be zero. Growth enhancement through the adoption of technological knowledge (following the literature based on endogenous growth theory) could be achieved through unilateral trade liberalization and not only by agreements with developed countries (major producers of technological knowledge). In addition, uniform tariffs (which do not exist in practice under the presence of RIAs) have the advantages of equal effective protection rates across sectors, simplicity that reduces business costs, lower customs administration costs (no need for rules of origin), and lower discretion (with lower incentives for corruption).

Regarding specific costs of RIAs, obviously the most important is related to trade diversion, as preferential treatments might lead to welfare-reducing distortions in production and consumption. The cost of trade diversion depends on the level of overall protection and the advantages that RIAs’ partners are given. For instance, RIAs with small, developing countries are likely to reduce welfare (as shown for the agreement with Mercosur by Harrison, Rutherford, and Tarr (2002)) as they tend to be based on political considerations. On the other hand, developing countries have limited bargaining power in their negotiations of RIAs with the United States and the EU. Hence, some conditions to get an RIA might be tough in areas such as intellectual property rights (e.g., patent protection on pharmaceuticals before 1991 given the abundance of generic products) and environmental and labor standards. The latter might reduce the flexibility of the economy, which precisely might be needed to facilitate the adjustment to the RIA. Another cost of RIAs involves the need to establish rules of origin to avoid the problem of trade deflection; those rules can in turn lead to additional trade diversion and welfare loss as they are typically hard to enforce, particularly in the presence of overlapping RIAs.44

Chile’s policymakers support the regionalist approach on several grounds:

  • Improved market access for Chilean exports and economies of scale that lead to increased efficiency.45 Through several RIAs Chile will have lowered its preferential tariff to all its major trading partners, with an effect similar to that of unilateral trade liberalization, but with better market access for its export products. In addition, negotiations would allow access for products with higher value added, which tend to be subject to tariff escalations, a factor that has complicated the diversification of exports away from natural resources.46

  • Protection against potential establishment of protectionist measures, thereby ensuring the stability of exports. For example, the Generalized System of Preferences (GSP) gives preferential access to the U.S. market but benefits have to be renewed from time to time and can be eliminated unilaterally. RIAs normally establish dispute settlement mechanisms to increase confidence in the security of market access.

  • RIAs in Chile result in relatively low trade diversion costs, since Chile has a low uniform tariff and RIAs do not include nontariff restrictions to the rest of the world.

  • Regionalist strategy is a response to limited progress in multilateral negotiations sponsored by the WTO.

  • Strategy is a response to the formation of trade blocs. Being excluded is costly as in the case of the North American Free Trade Agreement, since Mexico and Canada are the most important competitors of Chile as providers of natural resources and processed natural resources to the U.S. market.47 By contrast, when signing several RIAs a country becomes a “hub,” and its partners become “spokes,” if they have not signed free trade agreements with each other. In that case, investors would prefer to invest in the hub, reaching the spokes from there.

  • The negotiation of RIAs tends to involve multidimensional aspects complementary to trade relations. For example, in the agreement with Canada (1996) both countries decided to eschew antidumping actions against each other. In addition, RIAs normally include specific provisions to promote and protect investment flows between trading partners. The latter would benefit Chile as a recipient of foreign direct investment flows from developed countries and as an investor in other South American countries.

  • Negotiation of RIAs could be adapted to the perceived sectoral and regional costs and benefits. In fact, it could be argued that the sequencing of Chile’s RIAs (by concentrating on neighboring countries first) allowed Chile to strengthen some of its industries before signing agreements with developed countries (Selaive, 1998). In addition, longer phase-out periods were agreed on for sensitive agricultural products.48

  • RIAs with developed countries should raise credibility of the overall policy framework and governance, helping to differentiate Chile from other countries.

Recent Regional Trade Agreements

As mentioned earlier, Chile concluded negotiations for free trade agreements with the EU, the United States, and Korea in 2002, and with the EFTA in early 2003. These countries account for about half of Chile’s exports.

Chile’s agreement with the EU is the first one signed by a Latin American country and also the most advanced trade deal the EU has ever negotiated with a nonmember country. This means that, in addition to better access for Chile’s export products, there would be an incentive for European investors to establish businesses in Chile as an entry to the other Latin American countries. Under this agreement, about 85 percent of Chilean exports to the EU would enjoy zero tariff from the beginning, and this share would increase to 96 percent after 4 years and to 99.7 percent after 10 years.49 An important feature is the phaseout of tariff escalation clauses that historically have hindered diversification of Chilean exports to the EU.50 Before the agreement, the fishing and wine industries were facing 13 percent and 10 percent tariffs, respectively, while agro-industrial exports faced tariffs of 40 percent. The agreement also covers investment and services, establishes a dispute resolution mechanism, and reduces sanitary and phytosanitary inspections on Chilean exports by accepting the reports from the relevant Chilean regulatory agencies. The trade agreement with the EU took effect in February 2003. Tariff reductions agreed to for goods have already been applied, while those for services and investment would take effect only after the agreement’s approval by the parliaments of each EU member country. A general equilibrium study reported by the central bank estimated the potential direct impact of the agreement on Chilean exports and GDP at 3.2 percent and 0.5 percent, respectively.51

The agreement with the United States includes trade, a dispute resolution mechanism, services, investment, and sections on environmental and labor standards. The agreement was signed in June 2003 and was ratified by the United States and Chilean congresses, and went into effect in January 2004. The agreement on trade consolidates the market access conditions for Chilean exports (including GSP conditions) and aims at eliminating all tariffs in a maximum period of 12 years, including on agricultural and textile products. About 87 percent of Chilean exports to the United States would enjoy zero tariff from the beginning, and about 95 percent will be completely liberalized in 10–12 years. All escalation tariffs on Chilean exports will be gradually phased out, while tariffs on textile products are to be eliminated immediately. On capital flows, the agreement states that foreign investors will only be able to protest against restrictive measures by the Chilean government (on payments and transfers as well as on inflows) only a year after their implementation; this feature would allow Chile to put in place transitory restrictive measures to regulate the movement of capital flows. Consultation mechanisms will be established on environmental and labor issues, with the commitment of both countries to comply with current relevant regulations in that area that are consistent with internationally approved standards. Based on estimates by the Chilean Ministry of Foreign Relations, the potential direct impact of the trade agreement (assuming complete liberalization) on Chilean exports and GDP would be 16 percent and 2 percent, respectively.

The free trade agreement with Korea in October 2002 is the first one agreed upon between an Asian and a Western economy, and between transpacific members of the Asia-Pacific Economic Community (APEC). It is still being drafted into law; there is no specific timetable for the legislation to be discussed in both countries’ legislatures, but it would likely be enacted into law in 2004. Tariffs on Chilean exports would be eliminated in six steps over the next 13 years, with tariffs on 41 percent of exports being eliminated immediately and on 97 percent of exports in the next 7 years.52 This should benefit Chilean exporters vis-à-vis their main competitors in the Korean market, that is, Canada and New Zealand. In addition, the current export basket to Korea is not very diversified and the negotiated tariff reductions would open opportunities for new exports in the agriculture, forestry, fishing, wood, and chemical sectors. This agreement would also open opportunities for Korean investments in Chile aimed at an expanded Latin American market.

The agreement with EFTA also includes a dispute resolution mechanism, removal of antidumping measures, and sections on services and investment. The trade agreement involves the immediate abolition of tariffs for more than 90 percent of Chilean exports.


For example, Sachs and Warner (1995), Auty (1990), and Gelb and others (1988).

Lederman and Maloney (2002) argue that the findings of Sachs and Warner are probably due to unaccounted for country-specific effects and do not hold when applied to other periods of time or after dealing with endogeneity issues. By contrast, they emphasize the negative correlation between export concentration and growth.

Tornell (1999).

A key policy implication of this view was the “Prebisch hypothesis,” which called for reduced dependency on natural resource exports through a state-led inward-looking industrialization policy based on high tariff and quota barriers (import-substitution industrialization).

Of course, a limited endowment of production factors (particularly human capital) would exacerbate the impact of the exploitation of natural resources on other sectors, but this process could self-perpetuate if the abundance of natural resources slows human capital accumulation by making schooling more expensive. See Asea and Lahiri (1999).

Ilades-Georgetown University and Gerens, Ltd. (1996). This finding is shared by Lagos (1997) and Spilimbergo (1999).

See Section II on Chilean institutions and in particular the description of the country’s approaches to maintaining fiscal discipline.

See Appendix I on the evolution of exports in Chile in recent decades.

The latest numbers reflect partly the recent (sharper) fall in copper prices.

For example, Cabezas (2003) showed that one-third of the reduction in exports to Argentina in 2002 was reallocated to other export destinations.

Average of percent changes of data for Chile’s individual trading partners weighted by their share in total exports of goods, as calculated in the IMF’s World Economic Outlook database.

A least-squares estimation was made on first differences since the variables involved were found to be nonstationary and no cointegration relation was detected for the period under analysis (1975–2002). A battery of tests was performed to confirm the stability of parameters and to ensure that the residuals were well behaved. This finding was confirmed in an analysis by Baeza (2003).

See Berg and Krueger (2003) on openness and growth.

See Giles and Williams (2000) for a survey on the empirical literature on the export-led growth hypothesis.

Gallego and Loayza (2002) found a positive impact of openness on output growth in Chile based on a cross-country panel data estimation.

See Central Bank of Chile (2002).

Johansen’s procedure for a two-lag (based on Schwarz and Hannan-Quinn criteria) VAR system. Similar results were obtained for a system including noncopper exports volume.

These results contrast with Agosin (1999) time-series analysis for the period 1960–95.

Jarvis (1992).

Pietrobelli (1998) and Alvarez and Fuentes (2003).

Fischer (2001). The salmon industry has helped the establishment of newer fishing industries such as for turbot, abalone, and white sturgeon.

Sachs and Larrain (2001).

Zechner (2002) and Ffrench-Davis (2002).

Irwin (2002).

Wright and Czelusta (2002).

Chile ranked twentieth among 80 countries in the Global Competitiveness Report 2002–2003, thirteenth in the specific macroeconomic environment index, and nineteenth in the public macroeconomic environment index, and nineteenth in the public institutions index (first in Latin America). More recently, the government established a special tax treatment (no double taxation) for firms using Chile as a regional investment platform.

See Appendix II on Chile’s trade policy strategy and the characteristics of recent free trade agreements.

De Ferranti and others (2003).

The critical importance of educational levels in achieving higher growth rates was highlighted in a cross-country panel data regression by Gallego and Loayza (2002): increases in average years of schooling and in the quality of education in Chile to the top 10 percent of the world would lead to a rise in per capita growth rates of 2 percentage points, to 6 percent.

Brunner and Elacqua (2003) provide an assessment of human capital in Chile.

IALS, published by the OECD, as mentioned in Arellano (2001).

To “cope with the demands of everyday life and work in a complex society.”

In the microeconomic competitiveness ranking of the GCR, Chile was placed about sixtieth in quality of mathematics and science education and of public schools among 80 countries surveyed.

The differences in performance take place between schools, which in turn are segmented by socioeconomic characteristics. Cuadernos de Economía (2002).

World Bank (2004) and De Ferranti and others (2003).

See World Bank (2004).

Until the early 1970s, Chile had a heavily regulated economy based on an import-substitution development strategy. In fact, the average tariff in 1973 was 105 percent and ranged from 0 percent to 750 percent, in addition to some quantitative restrictions.

For a detailed review of those policies, see Agosin (1999) and Macario (1998).

World Trade Organization (1997).

World Trade Organization (1997).

The agreements with Mexico, Venezuela, Colombia, and Ecuador aimed at establishing free trade for about 95 percent of all tariff lines by 2000. The agreement with Mercosur would establish free trade for at least 75 percent of tariff lines by early 2006 and free trade for all tariff lines by 2014. The agreements also included provisions for the liberalization of investments, establishment of transparent rules for applying safeguard clauses, and commitments to harmonize export incentives.

The agreement with Canada, which entered into force in July 1997, provided for the elimination of most tariffs by January 2003, with the exception of tariffs on some “sensitive” agricultural products, which would be phased out until 2014.

Velasco and Tokman (1993).

Krueger (1997).

The unilateral trade liberalization strategy implicitly assumes that the tariff and nontariff barriers facing the exports of a country are very low or nonexistent (Wonnacott and Wonnacott, 1981).

Harrison, Rutherford, and Tarr (2002) show that the strategy of combining free trade agreements with Canada, Mexico, the United States, the EU, Mercosur, and the rest of South America produces welfare gains for Chile that are many multiples of the value of unilateral free trade if it were to attain tariff-free access to all these markets.

Campero and Escobar (1992), Zechner (2002), and Schiff (2002).

Hachette and Morales (1996) provide evidence that the agricultural sector would be the most affected (both on the positive and negative sides) by free trade agreements, but they also note the capacity that the agricultural sector had shown in the past to restructure itself.

Chilean exports to the EU could enjoy this treatment only up to certain quotas of total export to the EU area, but these quotas are so large (larger than the total trade with the EU), and growing at about 10 percent a year, that it is unlikely they could ever become binding. The remaining share (0.3 percent) would be subject to a revision clause.

For instance, export concentration levels to the EU are far higher than to the rest of the world, with exports of processed natural resources (particularly refined copper) accounting for about 73 percent of total exports to the EU.

General equilibrium model from “Sustainable Impact Assessment” SIA-Chile, referred to in the Central Bank’s Monetary Policy Report (January 2003). However, these estimates are static and conservative since they do not consider other channels, such as better resource allocation, reduced transaction costs, lower un-certainty on policies, and a better business climate, all of which would tend to motivate investment in physical and human capital and productivity-enhancing technological exchanges.

Both parties agreed on a list of sensitive products that will be exempted altogether from the free trade agreement.

    Other Resources Citing This Publication