I. Introduction and Summary

This paper assesses the potential for the export sector to play a larger role in promoting economic growth in Central America, Panama, and the Dominican Republic (CAPDR) through increased intra-regional and global trade integration.² CAPDR has grown in recent decades, but while there has been considerable variation across countries, on average growth has been lower than in comparator countries, and it is expected to continue to be modest in most countries in upcoming years. In this context, Swiston and Barrot (2011) found that CAPDR has significant scope to raise growth through supply-side reforms—especially reforms to the domestic financial system and key product markets. While the authors found that CAPDR countries rank relatively well in terms of de jure external openness (tariff rates and explicit restrictions on transactions), this paper expands their analysis by examining empirically CAPDR’s outward orientation—the extent to which the export sector is a driver of long-term growth—to ascertain areas for improvement.

CAPDR countries have actively pursued trade agreements and other regional integration projects in the last decade, but there is still much on the agenda. There has been substantial progress on bilateral and multilateral trading arrangements (MTA’s), including entry into force of the Dominican Republic-Central America-United States Free Trade Agreement (CAFTA-DR) in 2006–2009 and the signature in June 2012 of an Association Agreement (AA) with the European Union (EU). Implementation of regional projects to improve transportation and electricity linkages has also accelerated. Import tariffs have been substantially reduced, helping to spur trade both within and outside of the region, but implementation of policies that would support trade more broadly has lagged, such as investing in infrastructure, creating a customs union, and harmonizing regulations. While the full impact of recent initiatives will only materialize over a period of decades, this report takes stock of their status, describes the benefits and challenges they could bring, and identifies priorities for further progress.

The indexes of outward orientation developed in this paper suggest that CAPDR’s export sector is less favorably positioned to support long-term economic growth than similar countries in other regions. Overall exports-to-GDP ratios have hit a plateau since their peak in the late 1990’s, while trade in comparable countries in other regions is both higher than CAPDR and has continued to increase. A long time series of highly-detailed data on trade by product and partner is used to construct indexes of outward orientation for over 160 countries, which facilitates analysis of the composition of exports across countries and over time. The indexes cover the geographic scope of trading relationships, involvement in global production chains, and technological sophistication of exports, three concepts that the literature has emphasized are important for economic growth. The indexes show that CAPDR countries have increased their participation in global production chains and the sophistication of their exports, but continue to lag other trade blocs like the EU and members of the Association of Southeast Asian Nations (ASEAN). In addition, CAPDR could increase exports by penetrating the substantial untapped markets for the products that CAPDR already exports.

The paper finds a strong link between the indexes of outward orientation and both exports and economic growth, underscoring the potential benefits to CAPDR of efforts to deepen integration. An empirical model finds that trade logistics and outward orientation are among the key determinants of exports to GDP across countries. The analysis finds that CAPDR’s exports would be 10 to 20 percent of GDP higher if its logistics and outward orientation were at the levels of large Latin American countries, ASEAN, or the EU. Using a dynamic panel model, the paper finds that outward orientation is also a driver of economic growth. Growth in CAPDR economies could be higher by an average of 0.8 to 1.6 percent per year if its export sector were as dynamic as these comparator regions.

This paper has six sections: Section II describes CAPDR’s historical growth performance and the medium-term growth outlook. Section III discusses implementation of regional integration initiatives and highlights areas for further progress. Section IV presents data on the level and composition of the region’s exports, constructs indicators of outward orientation, and compares CAPDR’s indicators to those of other regions. Section V estimates models of exports to GDP and economic growth and illustrates CAPDR’s potential for improvement if policies to facilitate trade were implemented. Section VI concludes.

II. Historical and Prospective Economic Performance

From a long-term perspective, living standards in CAPDR have risen, but the income gap with advanced economies has not changed significantly. Real GDP per capita at purchasing power parity (PPP) has more than doubled since 1960, but the gap with the United States or the EU has not been reduced (Figure 1). Among other emerging markets, ASEAN members were less wealthy than CAPDR in the 1960’s, but have grown rapidly and now have a per capita income twice as high as CAPDR, while even large Latin American (LA5) economies have grown slightly faster.³ Income per capita in CAPDR as a whole declined to 16 percent of the U.S. level in 2005–2009, from 19 percent in the early 1960’s, despite growth in the Dominican Republic and Panama of about 3 percent per year over that period.

View Full Size

Economic Growth in CAPDR

Citation: IMF Working Papers 2012, 234; 10.5089/9781475510843.001.A001

• Download Figure
• Download figure as PowerPoint slide

Growth in the coming years is expected to moderate compared to the pre-crisis period. Growth was buoyant during the global boom years (4 percent per year from 2004–2007), though it remained lower than in other emerging markets (5 percent in the LA5 and 6 percent in ASEAN). Real GDP growth forecasts of IMF staff for 2012–17 suggest that this acceleration was temporary for the majority of economies in the region, and, while growth will continue to vary across countries, in most cases it is expected to decelerate to more typical pre-boom growth rates in the years ahead (Figure 1).⁴ These forecasts imply a continued slow rate of convergence in most of the region.

This reflects, in part, a reversion to less buoyant external demand and unfavorable terms of trade. Growth of external demand is projected to remain below that experienced during the boom years and below that of most South American countries (Figure 1), as CAPDR trades far more with the United States than with fast-growing Asian economies. Similarly, while the prices of some key agricultural exports have increased since 2007, they have been more than offset by higher prices for petroleum and other raw materials imports, such that the overall terms of trade deteriorated substantially in the 2000’s. The terms of trade are not expected to improve substantially from current levels, weighing on growth in CAPDR while benefitting the resource-intensive economies of South America.

One potential factor in the long-term growth performance of CAPDR countries is that they have been relatively less open than other small countries. Exports-to-GDP ratios tend to be higher in smaller countries as, given the absence of a sizable domestic market, producers take advantage of economies of scale through international trade. Studies have found a link between small country size and high openness (Alesina and Wacziarg, 1998), and have also suggested that openness substitutes for country size in promoting economic growth (Alesina, Spolaore, and Wacziarg, 2000 and 2004). The outward orientation of CAPDR countries over the last half century has been below that of the typical small country, as a regression of exports to GDP on country size shows that CAPDR’s exports have been below the norm for countries of its size, while those of ASEAN and the EU have been above the norm (Figure 2).⁵

View Full Size

Export-to-GDP Ratios, Average 1960–2009

Citation: IMF Working Papers 2012, 234; 10.5089/9781475510843.001.A001

• Download Figure
• Download figure as PowerPoint slide

Cross-country experience suggests that raising exports could play an important role in boosting long-term growth. There has been a strong correlation over the last half-century between countries with high economic growth and their openness as proxied by the deviation of non-oil exports from the predicted value in the above model (Figure 3). This suggests that CAPDR’s economic performance could improve with a more dynamic outward orientation.

View Full Size

Openness and Growth

Citation: IMF Working Papers 2012, 234; 10.5089/9781475510843.001.A001

• Download Figure
• Download figure as PowerPoint slide

In this context, spurring export growth would require deepening intra-regional integration and establishing new trade relationships outside the region. Deeper intra-regional integration would allow the combined market to take advantage of economies of scale and could spur investment, exports, and growth. Taken together, the CAPDR economies would be about the size of Colombia or Peru, making them a more attractive destination for potential investment (Figure 4). However, even as a combined market CAPDR would be a fraction of the size of the U.S. and Mexican economies, underscoring the importance of integration with major economies nearby. CAPDR’s trade links with the United States are strong, but trade with Mexico is very low given its proximity and economic size, suggesting that there could be unexploited opportunities for deeper integration. In addition, CAPDR has not expanded its trade links with the rest of the world over the last decade, pointing to the need to broaden trading relationships outside the region (Figure 5).

View Full Size

Economic Size and Outward Orientation ^1/

Citation: IMF Working Papers 2012, 234; 10.5089/9781475510843.001.A001

^1/ The relative size of the bubble corresponds to the size of the economy in terms of PPP GDP.

• Download Figure
• Download figure as PowerPoint slide

View Full Size

CAPDR: Exports by Destination

Citation: IMF Working Papers 2012, 234; 10.5089/9781475510843.001.A001

• Download Figure
• Download figure as PowerPoint slide

III. Trade Integration Agreements and Regional Initiatives

Following substantial work towards regional trade integration over the last half century, in recent years CAPDR has emphasized multilateral initiatives that underpin integration with the rest of the world. This section reviews their implications for trade, and the benefits and challenges that they bring.

The main MTA’s and initiatives promoting regional integration include:⁶

The Central American Common Market (CACM) began in 1960 and includes Costa Rica, Guatemala, Honduras, El Salvador, and Nicaragua, with Panama expected to join in 2012. It aims to create a free trade area for goods through tariff elimination for the member countries, establish a common external tariff, create a customs union, and eliminate non-tariff barriers to trade through harmonizing trade regulations.

CAFTA-DR, the free trade agreement that includes the five CACM countries, and adds the Dominican Republic and the United States, came into force upon ratification in signatory countries between 2006 and 2009. Panama and the United States have also signed a free trade agreement that will enter into force in October 2012. CAFTA-DR aims to establish a free trade area for goods, and extends liberalization to services and other areas important for trade, such as investor protection. It aims to reduce tariff and non-tariff barriers, including through boosting the transparency and efficiency of customs.

The EU-Central America Association Agreement (AA) was signed in June 2012. It includes CAPDR and the 27 EU countries. It still needs to be ratified by the legislatures of CAPDR countries and the European Parliament before it is implemented. The AA aims at creating a free trade area in goods and services through a reduction of tariffs and non-tariff barriers, and seeks to reinforce cooperation in customs to promote regional integration in CAPDR. The agreement also extends into other important trade-related areas, such as intellectual property rights and competition policy.

The free trade agreement between Mexico and the five CACM countries was signed in late 2011. It has to be ratified by the national legislatures before implementation. It will replace the bilateral trade agreements Mexico has with Costa Rica, El Salvador, Guatemala, and Honduras.

The Mesoamerican Project, launched in 2008, aims to improve regional infrastructure and trade logistics. It includes CAPDR, Belize, Colombia, and Mexico (Box 1).

The CACM has made advances in promoting intra-regional trade through tariff reductions but progress has been slower in harmonizing trade regulations and building a customs union. Only a few products (including coffee and sugar) are not traded freely in the region, and a majority of products have a common external tariff. Overall, tariffs have declined significantly in recent years (Figure 6) and there have been advances in harmonizing technical rules on trade and customs regulations. Nonetheless, non-tariff barriers remain an obstacle to trade in the region, particularly in the area of sanitary and phytosanitary (SPS) regulations. Moreover, there is still no integrated or common customs system and although countries continue to work towards this gradually, often national legislation has not been harmonized to regional rules. Guatemala and El Salvador signed an agreement in 2008 to create a customs union, but progress has been slow.

View Full Size

CAPDR: Average Effective Tariff Rates

Citation: IMF Working Papers 2012, 234; 10.5089/9781475510843.001.A001

• Download Figure
• Download figure as PowerPoint slide

CAFTA-DR should facilitate the region’s access to the U.S. market. It expands and makes permanent the bilateral trade preferences formerly extended unilaterally by the United States under the Caribbean Basin Initiative.⁷ Duties were eliminated on most goods immediately and the remaining products will have tariffs phased out over five to twenty years, depending on the product and country. Central American countries have a significantly larger number of products under gradual phase-out of tariffs than the United States, particularly for sensitive agricultural products, to allow more time for domestic producers to adjust to U.S. competition, while those goods protected under the CACM are not subject to liberalization. CAFTA-DR introduced new rules of origin (or cumulation of origin) that apply the same favorable tariff treatment to inputs and value-added in production from any member country, thus encouraging co-production arrangements across countries (González, 2005; Jaramillo and Lederman, 2006).⁸ For some manufactured goods, duty-free inputs will be allowed from certain third countries. These rules foster vertical integration of production chains. In addition, CAFTA-DR secures for its members market access for cross-border services, including telecommunications, finance, and tourism, and requires the regulatory authorities in Central American countries to improve transparency. The agreement also aims to reduce technical barriers to trade (TBT), including by providing technical assistance for SPS issues; liberalizes government procurement; enhances intellectual property and investment rights; and addresses enforcement of competition, labor, and environmental standards.

The AA with the EU will enhance links between the two regions, while reinforcing regional integration within Central America. The AA will make permanent the zero tariffs Central American exporters already have under the Generalized System of Preferences (GSP+), and extends these to bananas, raw cane sugar, and shrimp. The agreement will largely eliminate tariffs for manufactured goods, fisheries, and agriculture over ten to fifteen years, while certain sensitive agricultural products will remain protected. The agreement has rules of origin that allows members to use inputs from other members and share production in the elaboration of goods. Like CAFTA-DR, the AA intends to boost integration by developing harmonized trade regulations moving towards international standards in customs, and offering technical assistance to do so. In addition, it attempts to promote regional integration by aiming to introduce a single customs declaration and duty for the region and a regional competition authority.

Despite this progress, challenges to trade integration persist, including problems with external tariff harmonization, complex and unharmonized regulations, inadequate infrastructure and logistics, and lack of coordination among customs administrations.

A significant amount of tariffs are not harmonized. The Secretariat of Economic Integration for Central America (SIECA) states that 95.7 percent of tariff lines are harmonized but this figure is lower because it does not include vehicle lines, the temporary disharmonization of external tariffs under CAFTA-DR (Figure 7), and differences in rules of origin due to different bilateral free trade agreements in the region (de Gavidia, 2011).⁹

View Full Size

Programmed Harmonization of Tariffs under CAFTA-DR

Citation: IMF Working Papers 2012, 234; 10.5089/9781475510843.001.A001

Source: SIECA.

• Download Figure
• Download figure as PowerPoint slide

Trade rules can be complicated and vary across CAPDR countries. Rules of origin under CAFTA-DR can be highly complex and hard to apply. For example, determinations for rules of origin can be lengthy, U.S. importers at times have opted to pay tariffs rather than wait for favorable determinations, and sometimes CAFTA-DR exporters do not understand customs procedures (Hornbeck, 2012). TBT persist as not all technical rules have been harmonized and those that have been have not always been applied homogeneously across countries due to capacity problems. Similar SPS problems persist with deficiencies in the administrations of some countries due to a lack of regulations (or lack of harmonization) and technical capacity constraints.

Delays at customs in the region abound because of problems with security, governance, the electronic exchange of information, coordination, low capacity, and inadequate financing. Trade is also hindered by rising rates of crime and violence.

Inadequate infrastructure, particularly in the areas of roads, railroads, ports, and electricity results in CAPDR scoring lower on average on trade logistics compared to other regions and in higher transport costs (Figure 8). Weak logistics is due to low levels of investment spending and maintenance and a lack of coordination between governments in CAPDR. Regional legislation to facilitate transit is also stalled. Although the Mesoamerican Project has made good progress in integrating the distribution of electricity in the region, individual member countries need to boost their supply capability (Guasch, Rojas-Suarez, and Gonzales, 2011).

View Full Size

CAPDR: Trade Logistics, Infrastructure, and Customs

Trade logistics in CAPDR have improved but lag behind other regions. The regulatory and security environment are also behind comparators.

Citation: IMF Working Papers 2012, 234; 10.5089/9781475510843.001.A001

1/ rebased to 1-100 scale.Sources: World Bank; World Economic Forum; and authors’ calculations.

• Download Figure
• Download figure as PowerPoint slide

In light of these challenges, policies should focus on improving infrastructure and logistics, harmonizing trade rules and establishing a customs union, and strengthening regional coordination.

Trade rules, tariffs, and customs union: The remaining trade restrictions on products could be removed; customs requirements, inspection procedures, SPS and other technical standards harmonized; and a common external tariff fully implemented. Efforts should be made to harmonize existing free trade agreements regarding rules of origin and cumulation of origin which would also facilitate vertical integration. More technical assistance and training needs to be provided to officials and firms in order to effectively implement harmonized trade rules, which can be very technical, and to strengthen the tax collection capacity and efficiency of customs administrations. Eventually, to have a full customs union, import tax and tariff rules would need to be standardized and the collection of import duties centralized.

Infrastructure: Construction and maintenance of roads, railways, ports, and electricity should be increased, as public investment in the region only averaged about 4 percent of GDP in 2011. Governments facing financing constraints could utilize public-private partnerships to finance some projects, provided that they are undertaken within sound frameworks that minimize fiscal risks. The Mesoamerican Project should continue to coordinate projects, improve regional transport linkages, and attract financing.

Regional coordination: A regional approach is called for to address these issues, as well as to improve security and governance, particularly in border control and customs administration. Institutions for regional integration should be enhanced and a strategy on integration should be drafted in order to establish common objectives, coordinate efforts, and monitor implementation.

IV. The Profile of Capdr’s Exports

This section examines the level and orientation of CAPDR’s exports in light of the regional integration initiatives and areas for further progress described above. It finds that while both total and intra-regional exports have risen in recent decades, total exports are lower than in similarly-sized countries in other regions and intra-regional trade is not as high as within some other regional trading blocs. Fast-growing, highly-integrated regions have broadened the geographic scope of their trading relationships, increased participation in global production chains, and raised the technological sophistication of the basket of goods for export, providing lessons for CAPDR countries.

A. Overall Exports of Goods and services

Exports have gradually become more important for CAPDR economies, now accounting for about 40 percent of GDP on average. Exports have generally been on an increasing trend, rising sharply in the 1990’s but then leveling off in the last decade (Figure 9). Exports of goods still predominate, at over 30 percent of GDP, while exports of services have risen steadily to 10 percent of GDP. Panama is the most open country in the region, with exports at 70 percent of GDP (a significant portion is comprised of re-exports from the Colón Free Zone and transportation services related to the canal), while exports are less than 30 percent of GDP in the Dominican Republic, El Salvador, and Guatemala (Figure 10).

View Full Size

CAPDR: Exports to GDP

Citation: IMF Working Papers 2012, 234; 10.5089/9781475510843.001.A001

• Download Figure
• Download figure as PowerPoint slide

View Full Size

CAPDR: Exports of Goods and Services

Citation: IMF Working Papers 2012, 234; 10.5089/9781475510843.001.A001

• Download Figure
• Download figure as PowerPoint slide

However, CAPDR’s exports are lower and have risen less quickly in recent years than in countries of similar size in other regions of the world. To focus the comparison on examples that are relevant for CAPDR, the experience of small countries (those with a population of between 1 million and 30 million) in regions that include several emerging markets with per capita incomes close to those in CAPDR is examined—the Association of Southeast Asian Nations (ASEAN), the EU, and the Middle East and North Africa (MENA). See Table A.1 in Appendix I for a full list of the countries included. Exports in these regions are higher than those of CAPDR, ranging from 50 percent of GDP for MENA to 100 percent of GDP for ASEAN. Over the last decade, exports have risen faster than GDP in these countries, while CAPDR’s exports-to-GDP ratio has been nearly flat (Figure 11).

View Full Size

Exports of Small Countries

Citation: IMF Working Papers 2012, 234; 10.5089/9781475510843.001.A001

• Download Figure
• Download figure as PowerPoint slide

Intra-regional integration has risen, but the experience of other regions suggests there is room for further increases. Trade within CAPDR amounted to 6.9 percent of GDP on average in the 2000’s, up from 4.8 percent in the 1990’s. This is well behind small ASEAN and EU countries (Figure 12). It needs to be recognized, however, that the EU comprises a much larger region, which also includes several large countries. For that reason, the comparison was expanded to include the largest economies in close geographical proximity to the other blocs. Thus, for CAPDR the extended region includes the United States and Mexico; for ASEAN, Japan and China; and for MENA, the EU. CAPDR’s exports improve in this comparison because goods exports to the United States are 14% percent of GDP on average (though trade with Mexico is less than 1 percent of GDP despite its proximity and economic size). Overall, CAPDR’s intra-regional and extended regional trade lags small ASEAN and EU countries, suggesting that there is potential to increase trade integration with its neighbors.

View Full Size

Intra-Regional Exports of Small Countries

Citation: IMF Working Papers 2012, 234; 10.5089/9781475510843.001.A001

Note: Extended regions are USA for CAPDR, JPN and CHN for ASEAN, and EU for MENA.

• Download Figure
• Download figure as PowerPoint slide

B. Outward Orientation Indexes

To summarize a country’s degree of integration and qualitative aspects of its goods exports, three indexes capturing aspects of outward orientation were calculated.¹⁰ Various analysts have found these aspects to be beneficial to development. An overall index of outward orientation was then calculated by taking the product of these indexes. The three indexes are:

Market Penetration Index: Brenton and Newfarmer (2007) found that countries with dynamic exports were often set apart by their penetration of a high proportion of global markets for their products. They found that high ‘birth rates’ of individual export relationships are less important determinants of market penetration than low ‘death rates’. This reflects the ability of exporters to sustain trade relationships over time, which helps drive higher export growth at the country level. The market penetration index (MPI) represents the share (on a scale of 0 to 100) of world import markets in which a country participates, weighted by its export basket. The MPI first calculates for each product the imports of each country as a proportion of the global imports of that product.¹¹ For each exporter, it sums for each product the import shares of the countries it exports to. The overall MPI is the weighted average across products based on each product’s importance in that country’s export basket.

Vertical Integration Index: This measures intra-industry trade in a manner similar to Grubel and Lloyd (1971), capturing the extent to which the product categories of a country’s imports overlap with those of its exports (on a scale of 0 to 100). A higher value of the index seeks to capture greater involvement in global production chains in which a small transformation is performed on imported inputs before they are exported. This trade can have benefits in terms of technology transfer, product discovery, taking advantage of niches in which a country has a comparative advantage, and entry into broad networks of trade relationships (Brenton, Newfarmer, and Walkenhorst, 2009). Grubel and Lloyd (1971) summed intra-industry trade and expressed it as a share of the sum of exports and imports, adjusting the result for any overall trade imbalance. Given this paper’s focus on exports, intra-industry trade is calculated as a share of exports, with no trade imbalance adjustment.

Product Sophistication Index: This paper follows Hausmann, Hwang, and Rodrik (2007) to construct export sophistication indexes by assigning a productivity score to each product based on the average income level (GDP per capita at PPP) of the exporters of that product. This is justified by the tendency of high-income countries to specialize in more sophisticated high-technology products, which supports their higher average wages and reflects their higher average productivity levels. Aggregating these scores across the export basket of a given country yields a measure of the sophistication of that country’s exports.¹² The authors found the sophistication indexes to be positively related to higher rates of economic growth across countries.

These indexes were calculated using a long time series of highly-detailed data broken down by product and trading partner, with broad geographic coverage. Many studies examining detailed trade data have used the World Trade Flows dataset produced by Feenstra and others (2005). This study uses the World Trade Flows dataset and an update and expansion produced by Asmundson (2012). Both these sources draw on the United Nations’ COMTRADE database. It extends back to 1962, is based on a highly-detailed product classification spanning over 700 products (SITC2 at the 4-digit level of disaggregation), and covers bilateral trade flows by product between nearly all countries of the world.¹³

These indexes suggest that CAPDR has significant room to improve the geographical and technological composition of its exports. These indexes tend to trend upward over time as trade relationships have multiplied and the degree of processing trade and product sophistication have generally increased (Figure 13). The overall index of outward orientation of EU countries was relatively high in the 1960’s and has risen further since the 1990’s, coinciding with a period of deepening integration. The graphs also include the Asian Tigers (Singapore from ASEAN along with Hong Kong, Korea, and Taiwan) and show the rapid rise in the outward orientation indexes of these fast-growing countries. CAPDR, the ASEAN countries, and the LA5 had similar overall indexes in the 1960’s, but the latter two have climbed since the 1990’s, leaving CAPDR behind.

View Full Size

CAPDR: Assessing Outward Orientation

Citation: IMF Working Papers 2012, 234; 10.5089/9781475510843.001.A001

• Download Figure
• Download figure as PowerPoint slide

On the geographic side, the market penetration index (MPI) points to substantial opportunities for CAPDR to raise exports by increasing shipments of existing products to new trading partners. The MPI for CAPDR countries is less than 20 on average. In other words, CAPDR ships its products to countries that account for less than 20 percent of global imports of the products that CAPDR exports (as a point of reference, the United States accounted for 15 percent of global imports from 2005–2009 and a similar share in earlier periods). This is significantly less than the LA5, European Union, or ASEAN. Country size does not explain this low penetration, as the MPI’s of CAPDR countries are lower than that of countries of similar size in Europe and Asia.

CAPDR economies could also increase integration into global production chains. While an imperfect proxy, the Vertical Integration Index (VII) does seem to reflect the extent of countries’ involvement in global processing chains, as export-oriented countries such as the Asian Tigers have among the highest scores. The VII’s of CAPDR countries have risen on average since the 1980’s and are in line with those of the LA5 and ASEAN, but are much lower than those of the EU. The challenge is to leverage CAPDR’s entry into these trading relationships into opportunities to raise the value added of exports by exploiting the transfer of advanced technology and demonstrating comparative advantage in successively more sophisticated industries.

The technological content embodied in CAPDR’s exports has increased, but remains lower than that of other regions. While the approach used to calculate Product Sophistication Indexes (PSI’s) has its limitations (the United States is a large exporter of grain, for example),¹⁴ there is a high correspondence between a country’s PSI and other measures of sophistication such as those contained in the World Economic Forum’s Global Competitiveness Index (World Economic Forum, 2011).¹⁵ The sophistication of CAPDR’s products has improved gradually over the span of decades, but remains behind the LA5, ASEAN, and emerging market countries in the EU.

CAPDR’s low product sophistication ranking is reflected in the preponderance of agricultural products and basic manufactured consumer goods. The share of agricultural goods has declined since the early 1990s, but they remain among the most important products (Figure 13). Most of CAPDR’s exports of manufactures are basic consumer goods, as opposed to high-technology products or capital goods. Costa Rica and the Dominican Republic are two relative success stories in this regard, as they have increased exports of capital goods and other high-technology goods.

There is also room to increase the sophistication of services exports. Sophistication indexes for services exports (excluding exports of government services) were calculated following the same methodology as was employed for goods.¹⁶ The time span of the data is shorter and the classification far less detailed, covering only ten categories, which reduces the dispersion of the sophistication indexes across countries compared to those for goods.¹⁷ Nevertheless, the relative rankings both across regions and among CAPDR countries are in line with those for goods. CAPDR’s services exports, which are concentrated in travel (which includes tourism) and transportation, are less sophisticated on average than those of ASEAN, the EU, or the LA5 (Figure 14).¹⁸ Within CAPDR, Costa Rica has made strides in improving the sophistication of its services exports by finding niches in information technology and other business services (such as call centers), and Panama has a relatively more sophisticated mix because of its financial services industry.

View Full Size

Sophistication of Services Exports

Citation: IMF Working Papers 2012, 234; 10.5089/9781475510843.001.A001

• Download Figure
• Download figure as PowerPoint slide

V. Outward Orientation and Economic Performance

This section illustrates CAPDR’s potential for higher exports and growth by drawing lessons from regions with stronger logistics and a higher degree of outward orientation. Two models are estimated—one relating exports-to-GDP ratios to logistics and outward orientation, and a second relating economic growth to outward orientation. The results underscore the close link between logistics, outward orientation, and growth of exports and economic activity. They show that CAPDR could realize substantial gains by improving trade logistics and enacting policies to widen the scope of its trade relationships, facilitate participation of exporters in global production chains, and encourage innovation and the adoption of more stringent quality standards to improve the sophistication of exports.

A. Exports Model

A substantial proportion of the variation in ratios of exports of goods and services to GDP across countries is explained by country size and geography. More populous countries tend to trade less as a share of their economy because of scale effects (Alesina and Wacziarg, 1998; Alesina, Spolaore, and Wacziarg, 2000). Geographic factors such as country size and landlocked or island status could affect trade ratios through transportation costs. The impact on total exports of natural resource endowments, such as petroleum or precious metals, is unclear. Exports of the natural resource product directly raise total exports, but could lower exports of non-natural resource products by appreciating a country’s real exchange rate and reducing the competitiveness of those sectors, such that the overall impact is uncertain. Given the importance of exports of petroleum products for many countries and the ready availability of data, oil exports are included as an explanatory variable in a model of non-oil exports, such that the expected sign is negative.

A model of non-oil exports to GDP with the standard controls finds that country size and income are key determinants. The model estimated here is a modified version of the cross-sectional model used by Alesina and Wacziarg (1998) for overall openness, run on data averaged over five-year periods to minimize cyclical factors.¹⁹ The level of income is lagged to avoid endogeneity, and the export deflator for each country is included to capture the impact of fluctuations in export prices. Regional dummies are also included. The equation under analysis is:

$\begin{array}{cc}{X}_i=\alpha +\beta Z_i+\gamma D_i+{\epsilon }_i& \left(1\right)\end{array}$

Where i’s index countries, X is the ratio of non-oil exports to GDP, Z is a matrix of explanatory variables, D is a matrix of country dummies, and β and γ are the associated vectors of coefficients. Table 1 shows the results for 2005–2009, run on a cross-section of about 100 countries determined by data availability.²⁰ Column (1), which includes only non-policy variables, shows that, as expected, richer countries tend to trade more and larger countries tend to trade less. The East Asian regional dummy is strongly significant, reflecting the region’s strong emphasis on export-led growth, and implies that East Asian countries export about 25 percent of GDP more than countries in other regions with similar characteristics. The European Union dummy is only marginally significant, and other regional dummies, including that for CAPDR, are not significant. Other explanatory variables are not significant and are not shown to save space.

Table 1.

The Determinants of Exports to GDP

Notes: The regressions are run on data averaged over the five-year period. Robust T-statistics are in brackets. Variables significant at the 1 percent level are denoted by three asterisks, those at the 5 percent level by two asterisks and those at the 10 percent level by one asterisk.

Table 1.

The Determinants of Exports to GDP

Dependent variable: log of ratio of non-oil exports of goods and services to GDP, 2005-2009
		(1)	(2)	(3)	(4)
Current account openness			0.888**	0.384	0.0398
	(index)		[2.175]	[1.007]	[0.0962]
Index of outward orientation				0.330***	0.247***
	(log)			[4.069]	[2.872]
Logistics performance index					2.813***
	(log)				[3.713]
Lagged income		0.470***	0.419***	0.194**	-0.125
	(log of per capita PPP GDP)	[6.261]	[5.479]	[2.124]	[-0.914]
Size		-0.119***	-0.119***	-0.157***	-0.192***
	(log of area times population)	[-5.982]	[-6.004]	[-6.578]	[-7.839]
Island		-0.272*	-0.363**	-0.291**	-0.241*
	(dummy)	[-1.850]	[-2.611]	[-2.078]	[-1.779]
CAPDR		0.198	0.0595	0.216	0.245
		[0.675]	[0.205]	[0.759]	[0.977]
East Asia		0.968***	1.060***	0.701***	0.409*
		[3.611]	[3.972]	[3.097]	[1.826]
European Union		0.271*	0.169	-0.143	-0.224
		[1.775]	[1.096]	[-0.816]	[-1.313]
Latin America and Caribbean		-0.106	-0.0592	0.0653	0.140
		[-0.439]	[-0.266]	[0.348]	[0.699]
Sub-Saharan Africa		0.335	0.392	0.529**	0.241
		[1.059]	[1.217]	[2.154]	[0.954]
Intercept		0.0245	0.147	1.016	0.510
		[0.00681]	[0.0400]	[0.298]	[0.165]
Other controls: Oil exports, export deflator, landlocked dummy.
Number of countries		101	101	100	97
Adjusted R-squared		0.546	0.571	0.645	0.743

Notes: The regressions are run on data averaged over the five-year period. Robust T-statistics are in brackets. Variables significant at the 1 percent level are denoted by three asterisks, those at the 5 percent level by two asterisks and those at the 10 percent level by one asterisk.

View Table

Table 1.

The Determinants of Exports to GDP

Dependent variable: log of ratio of non-oil exports of goods and services to GDP, 2005-2009
		(1)	(2)	(3)	(4)
Current account openness			0.888**	0.384	0.0398
	(index)		[2.175]	[1.007]	[0.0962]
Index of outward orientation				0.330***	0.247***
	(log)			[4.069]	[2.872]
Logistics performance index					2.813***
	(log)				[3.713]
Lagged income		0.470***	0.419***	0.194**	-0.125
	(log of per capita PPP GDP)	[6.261]	[5.479]	[2.124]	[-0.914]
Size		-0.119***	-0.119***	-0.157***	-0.192***
	(log of area times population)	[-5.982]	[-6.004]	[-6.578]	[-7.839]
Island		-0.272*	-0.363**	-0.291**	-0.241*
	(dummy)	[-1.850]	[-2.611]	[-2.078]	[-1.779]
CAPDR		0.198	0.0595	0.216	0.245
		[0.675]	[0.205]	[0.759]	[0.977]
East Asia		0.968***	1.060***	0.701***	0.409*
		[3.611]	[3.972]	[3.097]	[1.826]
European Union		0.271*	0.169	-0.143	-0.224
		[1.775]	[1.096]	[-0.816]	[-1.313]
Latin America and Caribbean		-0.106	-0.0592	0.0653	0.140
		[-0.439]	[-0.266]	[0.348]	[0.699]
Sub-Saharan Africa		0.335	0.392	0.529**	0.241
		[1.059]	[1.217]	[2.154]	[0.954]
Intercept		0.0245	0.147	1.016	0.510
		[0.00681]	[0.0400]	[0.298]	[0.165]
Other controls: Oil exports, export deflator, landlocked dummy.
Number of countries		101	101	100	97
Adjusted R-squared		0.546	0.571	0.645	0.743

Notes: The regressions are run on data averaged over the five-year period. Robust T-statistics are in brackets. Variables significant at the 1 percent level are denoted by three asterisks, those at the 5 percent level by two asterisks and those at the 10 percent level by one asterisk.

The model finds a strong link between outward orientation and exports. Column (2) adds current account openness and finds a positive impact on exports, as expected, while other results are similar.²¹ Column (3) adds the overall index of outward orientation presented in the previous section and shows that it is a significant determinant of exports to GDP. An increase of one standard deviation (15 points on a 100 point scale) is estimated to raise exports by 11 percent of GDP. The statistically significant impact of current account openness is not robust to the inclusion of outward orientation.

Better logistics are also associated with higher exports. Column (4) of Table 1 adds the overall LPI and finds that it is a highly significant determinant of exports. An increase of one standard deviation in the LPI is estimated to raise exports by 12 percent of GDP. The coefficient on outward orientation remains strongly significant and implies that a one standard deviation increase would raise exports by 7 percent of GDP. When controlling for the profile of exports and the quality of logistics, the level of income is no longer a significant driver of exports to GDP. The East Asia dummy is only marginally significant and implies that East Asian countries only export 8 percent of GDP more than similar countries in other regions. These results suggest that much of the observed tendency for richer and East Asian countries to trade more is explained by their better logistics and their more aggressive outward orientation.

These results indicate that CAPDR has the potential to substantially boost exports. Applying the coefficients from column (4) of Table 1, the export ratios of CAPDR countries would be higher by 10 percent of GDP, on average, if their logistics and outward orientation levels were comparable to those of the LA5, and 20 percent of GDP higher, on average, if they were at EU levels.²² In such a scenario, exports to GDP in CAPDR would exceed those of small MENA countries and be on par with those of small EU countries (Figure 15).

View Full Size

Exports of Small Countries

Citation: IMF Working Papers 2012, 234; 10.5089/9781475510843.001.A001

• Download Figure
• Download figure as PowerPoint slide

CAPDR’s potential for higher exports, especially within the region, is corroborated by other studies. A study employing a different methodology (Marcelo, Stokenberga, and Schwartz, 2010) reaches similar conclusions. Using a gravity model, they estimate that intra-regional trade could double and trade with the United States and EU could increase by a third if adjacency and time-distance factors were improved to EU levels. Indeed, while intra-regional trade would remain below that of small ASEAN and EU countries in such a scenario due to CAPDR’s smaller economic size, exports including the extended region would be similar (Figure 16).

View Full Size

Intra-Regional Exports of Small Countries

Citation: IMF Working Papers 2012, 234; 10.5089/9781475510843.001.A001

Note: Extended regions are USA for CAPDR, JPN and CHN for ASEAN, and EU for MENA.

• Download Figure
• Download figure as PowerPoint slide

B. Growth Model

To evaluate how outward orientation affects economic growth, the indexes were added to a growth model including the standard control variables. Swiston and Barrot (2011) estimated a growth model on data since 1960 using System GMM, a dynamic panel method, and applied the results to CAPDR. They found that investment in physical and human capital, macroeconomic stability, demographics, and initial income were significant determinants of long-run growth rates. In addition, structural reforms, as summarized by a battery of 13 indicators, were important. This paper uses the same model but adds the index of outward orientation and each subcomponent in turn, in order to examine which characteristics of goods exports are important for growth.²³

In addition to the usual factors associated with economic growth, the index of outward orientation and its subcomponents are all associated with a significantly higher rate of economic growth (Table 2). The results for the standard variables are in line with those in Swiston and Barrot (2011). The estimated impact of structural reforms remains statistically and economically significant, as do the other factors mentioned above. In addition, the index of outward orientation is found to be a significant determinant of growth. An increase of one standard deviation in the index is estimated to raise the long-run growth rate by 1.3 percentage points. Market penetration and vertical integration are found to be statistically significant and export sophistication marginally significant.²⁴ The results suggest that these indexes are good proxies for the aspects of a country’s outward orientation that are beneficial for economic growth.

Table 2.

The Impact of Outward Orientation on Growth

Notes: The regressions are run on data from 1965-2009 in non-overlapping five-year periods. Export structure variables are entered as the log of an index in which the United States equals 100 as a reference value. Time dummies for each period are included but not shown. T-statistics based on Windmeijer (2005) robust standard errors are in brackets. Variables significant at the 1 percent level are denoted by three asterisks, those at the 5 percent level by two asterisks and those at the 10 percent level by one asterisk.

Table 2.

The Impact of Outward Orientation on Growth

Dependent variable: growth in PPP GDP per capita
		(1)	(2)	(3)	(4)
Index of outward orientation		0.691**
	(log)	[2.474]
Market penetration			2.149**
	(log)		[2.097]
Vertical integration				0.951***
	(log)			[2.716]
Product sophistication					2.403*
	(log)				[1.792]
Lagged income		-2.165***	-2.365***	-1.711***	-1.629***
	(log of PPP GDP per capita)	[-4.808]	[-3.771]	[-4.658]	[-3.757]
Capital formation		2.452***	1.861*	2.735***	3.091***
	(log of investment to GDP)	[2.662]	[1.773]	[3.364]	[4.529]
Human capital		0.860**	1.021**	0.847**	0.260
	(log of advanced schooling)	[2.188]	[2.083]	[2.008]	[0.512]
Inflation		-2.500*	-1.966	-2.679**	-2.277**
	(log of 1 plus inflation rate)	[-1.974]	[-1.424]	[-2.194]	[-1.992]
Structural reforms		4.318**	6.604***	3.724*	4.925***
	(index)	[2.104]	[3.233]	[1.724]	[2.848]
Labor force		0.604***	0.655***	0.648***	0.666***
	(as share of population, log)	[3.794]	[4.001]	[4.268]	[5.362]
Intercept		6.424	1.479	-0.0761	-8.886*
		[1.168]	[0.314]	[-0.0191]	[-1.905]
Observations		698	698	698	698
Number of countries		79	79	79	79
Number of instruments		23	23	23	23
Hansen test p-value		0.787	0.226	0.743	0.440
AR(2) test p-value		0.425	0.145	0.512	0.523

Notes: The regressions are run on data from 1965-2009 in non-overlapping five-year periods. Export structure variables are entered as the log of an index in which the United States equals 100 as a reference value. Time dummies for each period are included but not shown. T-statistics based on Windmeijer (2005) robust standard errors are in brackets. Variables significant at the 1 percent level are denoted by three asterisks, those at the 5 percent level by two asterisks and those at the 10 percent level by one asterisk.

View Table

Table 2.

The Impact of Outward Orientation on Growth

Dependent variable: growth in PPP GDP per capita
		(1)	(2)	(3)	(4)
Index of outward orientation		0.691**
	(log)	[2.474]
Market penetration			2.149**
	(log)		[2.097]
Vertical integration				0.951***
	(log)			[2.716]
Product sophistication					2.403*
	(log)				[1.792]
Lagged income		-2.165***	-2.365***	-1.711***	-1.629***
	(log of PPP GDP per capita)	[-4.808]	[-3.771]	[-4.658]	[-3.757]
Capital formation		2.452***	1.861*	2.735***	3.091***
	(log of investment to GDP)	[2.662]	[1.773]	[3.364]	[4.529]
Human capital		0.860**	1.021**	0.847**	0.260
	(log of advanced schooling)	[2.188]	[2.083]	[2.008]	[0.512]
Inflation		-2.500*	-1.966	-2.679**	-2.277**
	(log of 1 plus inflation rate)	[-1.974]	[-1.424]	[-2.194]	[-1.992]
Structural reforms		4.318**	6.604***	3.724*	4.925***
	(index)	[2.104]	[3.233]	[1.724]	[2.848]
Labor force		0.604***	0.655***	0.648***	0.666***
	(as share of population, log)	[3.794]	[4.001]	[4.268]	[5.362]
Intercept		6.424	1.479	-0.0761	-8.886*
		[1.168]	[0.314]	[-0.0191]	[-1.905]
Observations		698	698	698	698
Number of countries		79	79	79	79
Number of instruments		23	23	23	23
Hansen test p-value		0.787	0.226	0.743	0.440
AR(2) test p-value		0.425	0.145	0.512	0.523

Notes: The regressions are run on data from 1965-2009 in non-overlapping five-year periods. Export structure variables are entered as the log of an index in which the United States equals 100 as a reference value. Time dummies for each period are included but not shown. T-statistics based on Windmeijer (2005) robust standard errors are in brackets. Variables significant at the 1 percent level are denoted by three asterisks, those at the 5 percent level by two asterisks and those at the 10 percent level by one asterisk.

These results imply that CAPDR should prioritize putting in place the conditions to facilitate broadening the scope of its trading relationships, exploiting niches in global production chains, and enhancing the sophistication of exports. Applying the above results to CAPDR countries suggests that growth would be higher by 0.3 to 1.8 percentage points per year if its export sector were as dynamic as the LA5 and by 1.1 to 2.5 percentage points if it were as dynamic as the EU (Figure 17). Some of the potential gains may lie outside the scope of policy, in natural resource endowments, for example, while others are related to deep-seated structural factors such as the business environment or workforce training. However, the discussion in Section III implies that policies could play a key role, even if the impact were only discernible over the long run.

View Full Size

CAPDR: Effect on Growth from Improving Export Structure

Citation: IMF Working Papers 2012, 234; 10.5089/9781475510843.001.A001

• Download Figure
• Download figure as PowerPoint slide

VI. Conclusions

CAPDR is making substantial progress in pursuing deeper intra-regional and global integration. MTA’s with the United States and EU, among others, will likely support exports and economic growth by reducing or eliminating tariffs on most products with key trading partners, introducing cumulation of origin rules that facilitate integration of production across countries, increasing the efficiency of customs administration, reducing non-tariff barriers, and lowering barriers to foreign investment. This should create opportunities for investment, spurring productive linkages across countries, technology transfer, and export sophistication. The Mesoamerican Project’s emphasis on improving infrastructure could help the region to capitalize on potential economies of scale and encourage private investment. The IMF has facilitated trade integration in the region, including through providing technical assistance on customs issues and urging countries to raise levels of public infrastructure.

Notwithstanding this progress, cross country experience suggests that CAPDR’s export sector could play a more important role in supporting long-term growth. CAPDR’s exports have increased, but remain lower than those of similarly-sized economies in other regions. According to the indexes of outward orientation constructed in this paper, the breadth of geographic trading relationships, depth of vertical integration through international trade, and degree of technological sophistication of exports in CAPDR are less conducive to higher exports and growth than in fast-growing, export-oriented economies.

CAPDR’s exports and growth could be substantially higher with improved logistics and a more aggressive outward orientation. In particular, the model of exports-to-GDP ratios used in this paper suggests that exports could be 10 to 20 percentage points of GDP higher with improved logistics, a broader geographic scope of trading partners, deeper integration into global production chains, and increased technological sophistication of exports. In addition, our growth model finds that if these conditions were in place, the average CAPDR country could raise its long-term growth rate by 0.8 to 1.6 percentage points per year.

While the full impact of recent initiatives has yet to materialize, these findings underscore the importance of tackling remaining obstacles to integration. Effective implementation of the MTAs is critical for reaping their potential benefits. In this context, policymakers should prioritize building a customs union, harmonizing trade rules, and increasing the capacity of firms and officials to implement these rules. This will require overcoming divergent national legislation and institutional capacity, strengthening coordination, and working within budget constraints. In addition, addressing deep-seated structural issues such as regional infrastructure links, the security situation, and governance will require high-level political coordination and increased funding.