This Selected Issues paper on Sri Lanka reviews several issues that highlight both Sri Lanka’s accomplishments and their policy constraints amidst a protracted period of civil conflict and political instability. High intermediation costs have held back development of the financial sector and could also frustrate Sri Lanka’s quest for higher growth. The main constraints to achieving higher growth include the civil conflict, political instability, high fiscal deficits and inflation, and underdeveloped financial markets.

Abstract

This Selected Issues paper on Sri Lanka reviews several issues that highlight both Sri Lanka’s accomplishments and their policy constraints amidst a protracted period of civil conflict and political instability. High intermediation costs have held back development of the financial sector and could also frustrate Sri Lanka’s quest for higher growth. The main constraints to achieving higher growth include the civil conflict, political instability, high fiscal deficits and inflation, and underdeveloped financial markets.

IV. International Trade and Trade Potential in Sri Lanka1

A. Introduction

1. Sri Lanka’s external trade has performed well in recent years. While trade with India has expanded, it still accounts for only a small portion of Sri Lanka’s total exports. As the economic boom of its big neighbor is expected to continue in the foreseeable future, Sri Lanka should be able to benefit, contributing to economic growth and further trade expansion.

2. This chapter investigates Sri Lanka’s trade and trade potential, in particular with respect to trade with India. In the next section, we analyze Sri Lanka’s general trade openness and recent trade performance. In Section C, we focus on regional trade and quantify the geographic concentration of Sri Lanka’s exports. In Section D, we zoom in on bilateral trade with India and discuss the potential for expansion. Section E presents concluding remarks.

B. Some Characteristics of External Trade in Sri Lanka

3. Recent trade performance has been encouraging, but continues to rely heavily on a few commodities. After the global economic slowdown in 2000–2001, trade picked up at about the same pace as before (Figure IV.1). In the last two years, average export growth was about 10 percent, while imports grew roughly at 15 percent annually. The trade deficit, which historically has been hovering around 7–11 percent of GDP, was just over 11 percent in 2004. Traditionally, Sri Lanka has been relying heavily on a few commodity exports, notably tea and garments. The latter accounted for roughly 49 percent of the value of exports in 2004, although its share has come down from a peak of 54 percent in 2000. Recently, the growth of nongarment exports has picked up to 14 percent on average in 2003–2004, while garment exports grew just under 8 percent annually.

Figure IV.1.
Figure IV.1.

Sri Lanka Trade Performance

(1990=100, U.S. dollar value)

Citation: IMF Staff Country Reports 2005, 337; 10.5089/9781451972542.002.A004

4. Sri Lanka’s general openness to trade is low, but exports are higher than what would be expected. Table IV.1 compares actual trade openness to what is predicted by an openness model. A discussion of the features of this model and the regression results is presented in Annex IV.I. Trade openness in Sri Lanka appears to be less than would be expected for a country its size and stage of economic development, with overall imports and exports about 30 percent below the norm. Sri Lanka’s South Asian trading partners share the same characteristic, while a sample of other Asian countries (including China) show trade about 40 percent above the potential predicted by the model. Sri Lanka’s exports, on the other hand, are higher than would be expected (40 percent above potential); while exports of its regional trading partners are in line with the model’s prediction. Export performance of other Asian countries far exceeds their estimated export potential.

Table IV.1

Actual versus Potential Trade Openness, 1998–2003

article image
Sources: Direction of Trade Statistics; and Fund staff calculations.

South Asian Association for Regional Cooperation (SAARC) includes Bangladesh, Bhutan, India, Maldives, Nepal, Pakistan, and Sri Lanka.

C. Geographic Concentration of Exports

5. The united States and the European Union continue to be the primary export markets for Sri Lanka. The combined share of the two main industrialized trading blocks exceeds 60 percent of total exports, unchanged from a decade ago (Figure IV.2). Since 2000, the export share of the United States has been declining, but it remains the primary export market, with about 31 percent ($1.86 billion) of total exports, compared to 29 percent ($1.76 billion) for the European Union. Exports to India have become more important and now account for about 5 percent of total exports ($0.3 billion), up from less than 1 percent five years ago. Exports to South Asian countries other than India remain underdeveloped, as their share of total exports from Sri Lanka has stabilized at around 2 percent. The export share of other (East) Asian countries has declined to roughly 8 percent, down from about 13 percent a decade ago.

Figure IV.2.
Figure IV.2.

Trade Intensity Index, 1994–2003

Citation: IMF Staff Country Reports 2005, 337; 10.5089/9781451972542.002.A004

6. Analysis based on the trade intensity index reveals a more nuanced picture of Sri Lanka’s export concentration. The trade intensity index is a more sophisticated method to assess trade concentration as it takes into account the size of the import market.2 It measures the extent to which an importing country’s share of total Sri Lankan exports is large or small in relation to the importing country’s share of total world trade. According to this measure, Sri Lanka’s exports are highly concentrated with its South Asian trading partners, including India (Table IV.2). Export concentration with its two main trading partners, the United States and the European Union, turns out to be remarkably low when the share in total global trade of these two trading blocks is taken into account. Export concentration with India is high, and while the trade intensity index fell in the late 1990s, it increased during 2002-03 reflecting rapid growth of exports to India, before falling again in 2004. Export concentration with Bangladesh, Bhutan, and Nepal is much lower than with India (and the Maldives), underscoring the point that these potential markets for Sri Lankan exports are underdeveloped.

Table IV.2

Actual versus Potential Trade Openness, 1998–2003

article image
Source: Fund staff estimates.

EU-15 refers to the 15 members states of the European Union in the period prior to enlargement in 2004: Austria, Belgium, Denmark, Finland, France, Germany, Greece, Ireland, Italy, Luxembourg, the Netherlands, Portugal, Spain, Sweden and the United Kingdom.

7. The combination of exports at or above potential with overall trade below potential suggests that the import ratios of South Asian countries were relatively low during the observation period. This is part of the heritage of import substitution policies that have dominated the region in the past decades. East Asian countries, on the other hand, have long ago abandoned such policies and opened up their markets, as indicated by the fact that both overall trade and exports are above estimated potential. This illustrates the point that reducing barriers to trade would help Sri Lanka and the rest of the region to sustain higher growth.

D. Trade Potential with India

8. The pickup of exports to India is a recent phenomenon. Sri Lanka was slow to pick up on the large trade potential that its fast growing neighbor provides, but recently exports to India have been increasing rapidly (Figure IV.3). In the last two years, exports to India have almost doubled, contributing 12 percent to Sri Lanka’s overall export growth, even though exports to India are only 5 percent of the total. While exports to India from Bangladesh and Pakistan jumped in the early and late 1990s, respectively, exports from Sri Lanka did not take off until 2001, one year after the agreement to establish the Indo-Sri Lanka Free Trade Agreement (ISLFTA) came into effect. Exports from other (East) Asian countries to India have shown solid growth rates since the early 1990s, although not at the same pace as some of India’s South Asian trading partners. The rapid expansion of exports to India was not matched by equally high export growth to other South and East Asian countries—exports to these countries have been virtually flat in the past decade (Figure IV.4). The rapid growth in exports to India has resulted in a tripling of Sri Lanka’s share of Indian imports to 0.3 percent (Figure IV.5), despite a small drop in market share last year, when Indian imports rose by more than 35 percent.3 The recent pick up in exports to India has not resulted in higher export concentration with India (Table IV.2). In fact, Sri Lankan exports are now relatively less concentrated with India than a decade ago. This suggests that the expansion of Sri Lankan exports to India has not kept up with the overall increase in India’s imports.

Figure IV.3.
Figure IV.3.

India: Import Growth by Trading Partner

(1990=100, U.S. dollar value)

Citation: IMF Staff Country Reports 2005, 337; 10.5089/9781451972542.002.A004

Figure IV.4.
Figure IV.4.

Sri Lanka: Export Growth by Trading Partner

(1990=100, U.S. dollar value)

Citation: IMF Staff Country Reports 2005, 337; 10.5089/9781451972542.002.A004

Figure IV.5.
Figure IV.5.

Sri Lanka: Exports to India

(Market Share in percentage points and U.S. dollar value)

Citation: IMF Staff Country Reports 2005, 337; 10.5089/9781451972542.002.A004

9. Sri Lanka participates in regional economic cooperation and trade initiatives, but these have been largely ineffective in promoting regional trade. The main vehicle for regional economic integration is the South Asian Association for Regional Cooperation (SAARC), established in 1985. In 1995, member countries established the SAARC Preferential Trading Arrangement (SAPTA). This agreement, however, has been relatively ineffective in promoting regional trade, and so far regional integration has been slow due to political tensions and large differences in economic policies and regulations (see World Bank/IMF 2004). In 2004, SAARC leaders agreed to establish by 2006 the South Asia Free Trade Agreement (SAFTA), but private sector expectations are low after the experience with SAPTA.4 As regional economic integration in the context of SAARC stalled, Sri Lanka has focused more on bilateral trade agreements.

10. The free trade agreement with India has proven very successful in promoting bilateral trade, despite the limited grant of preferences. The agreement to establish the ISLFTA was signed in 1998 and came into effect in 2000.5 Since then, bilateral trade has expanded significantly, with Sri Lanka’s exports to India growing more than 300 percent since the agreement came into effect. This success comes despite the fact that trade preferences granted under the ISLFTA are limited (Baysan et al., 2004). Many of Sri Lanka’s main export products are excluded from preferential treatment via a negative list or quota limitations. When the agreement was signed, these products accounted for 46 percent of Sri Lanka’s exports to India. This suggests that the expansion of exports to India has been the result of diversification of Sri Lanka’s export base. In 2002, Baysan et al. (2004) estimate that 38 percent of the exports to India were products that were not exported to India at all prior to the ISLFTA.

11. India and Sri Lanka are in the process of expanding their bilateral economic cooperation and integration, which should help realize the trade potential in both goods and services. The success of the ISLFTA has sparked efforts to expand economic cooperation. Since 2003, the two countries have been negotiating to convert the ISLFTA into a Comprehensive Economic Partnership Agreement (CEPA). The objective is to go beyond establishing a free trade zone, and initiate efforts to increase cross border investment and economic and technological cooperation, which should support the growth of bilateral trade in goods, as well as health, tourism, and construction and transportation services.

12. The fast growing Indian market provides ample opportunities to further expand Sri Lanka’s exports. The Indian economy has grown faster than Sri Lanka’s two main trading partners and will continue to do so. During the period 1992–2004, GDP and import growth in India has outpaced the United States by 2 to 1 and the European Union by 3 to 1 (Figures IV.6 and IV.7). IMF staff medium-term forecasts suggest that this trend will continue, with growth in India picking up to 6.4 percent in the medium term, compared to 3.5 percent in the United States and 2.2 percent in the European Union. Import growth in India is projected at roughly 16 percent through 2009; more than twice the import growth rate in the United States and the European Union. In the long run, Deutche Bank projects that Indian GDP growth could average 5.5 percent annually through 2020, outpacing all OECD countries as well as China.6 The projected rapid expansion of the Indian economy indicates that there is ample scope for Sri Lanka to further expand exports and economic integration.

Figure IV.6.
Figure IV.6.

GDP Growth of Three Main Trading Partners, 1992–2009

Citation: IMF Staff Country Reports 2005, 337; 10.5089/9781451972542.002.A004

Figure IV.7.
Figure IV.7.

Growth of Imports of Three Main Trading Partners, 1992–2009

Citation: IMF Staff Country Reports 2005, 337; 10.5089/9781451972542.002.A004

13. A standard gravity model also reveals considerable potential for Sri Lanka’s exports to India to continue to grow rapidly. India’s economy and imports are growing so rapidly, Sri Lankan exports should be able to continue their rapid growth even if the intensity index for exports to India were to decline. To analyze Sri Lanka’s export potential with India we estimate a gravity model of bilateral trade to simulate potential growth (see Annex IV.I for details and the estimation results). We have estimated the model using data for all SAARC countries. The results of the simulation analysis are shown in Table IV.3. Despite the recent pick up in exports to India, Sri Lanka is not fully realizing its estimated export potential with India. According to the model, Sri Lanka’s potential exports to India are 0.45 percent of GDP ($354 million in 2004). This implies that last year, Sri Lanka’s exports to India were under performing by roughly 10 percent (about $30 million). Since 2000, the total export underperformance was about 33 percent ($280 million). Although, according to this analysis, Sri Lanka’s exports to India are still underperforming, there has been a significant improvement since the free trade agreement with India was signed.

Table IV.3

Sri Lanka: Actual and Potential Exports to India, 2000–2004

article image
Sources: WEO; and Fund staff calculations.

E. Concluding Remarks

14. India is projected to remain Sri Lanka’s fastest growing main trading partner. In recent years, Sri Lanka has successfully penetrated the Indian export market, but it has not yet fully realized its trade potential. The growth of exports, while rapid, has been less than the overall growth of India’s external trade, as indicated by the decline of the trade intensity index of Sri Lanka’s exports to India from 16.2 in 1994 to 15.l in 2004. Simulations with a regional gravity model for bilateral trade with India indicate that in 2004 Sri Lanka was underutilizing its trade potential with India, providing another indication that there is room for expansion of exports to India.

15. The ISLFTA that came into effect in 2000 has successfully promoted bilateral trade. In the last two years, exports to India have almost doubled and accounted for 12 percent of Sri Lanka’s overall export growth, even though exports to India are only 5 percent of Sri Lanka’s total exports. This success comes despite the fact that many of Sri Lanka’s main export products are excluded from the agreement via a negative list or quota limitations. This suggests that the growth of exports to India has been the result of diversification of Sri Lanka’s export base.

16. The success of ISLFTA bodes well for future export growth. Sri Lanka can build on the early success of the trade agreement with India as the two countries seek to expand their economic cooperation. The CEPA, which is being negotiated, could provide incentives to increase cross border investment and trade in services, just like the ISLFTA did for trade in goods. This increased opening of markets should help Sri Lanka to benefit from the rapid future growth of the Indian economy and realize its full trade potential.

17. Despite the scope for expanding regional trade discussed here, the focus on regionalism should not come at the expense of more liberal trade policies in general. Although regional trade integration and cooperation can support growth and economic development, higher and more sustainable growth rates can be achieved from broad trade liberalization (see Vamvakidis 1998). The growth experience of many East Asian countries also underscores the need for free trade beyond export promotion. In addition to deepening regional trade integration, Sri Lanka should therefore consistently pursue free trade policies. In this context the introduction of a low uniform tariff, abolishing the remaining 10 percent import duty surcharge, and reduction of some excessively high excise taxes, should benefit the external sector and promote growth.

Application of Trade Openness and Gravity Models to Sri Lanka’s Trade

Trade Openness Model

A country’s general degree of over- or under-trading can be assessed by estimating an openness model. The model aims to explain cross-country variations in trade openness by variations in the size of the economy, the stage of economic development, and country or region specific factors. Trade openness is defined as the sum of imports and exports as a share of GDP (the trade ratio), or, alternatively the exports to GDP ratio. In our estimation, we have used population as a proxy for the size of the country and GDP per capita as a proxy for economic development, while dummies account for region specific factors. For more details on this approach, see Rodrik (1998) and Freinkman et al (2004). Our specification is similar although we use an updated and expanded data set. The regression results are shown in Table IV.1.1 below. As expected, larger countries turn out to trade less, while more developed economies tend to be more open. East Asian countries trade more than would be expected based on their size and stage of economic development, while Latin American countries trade less than expected.

Table IV.1.1.

Regression Results for Trade Openness Model 1/

article image
Source: Fund staff estimates.

* stands for statistical significance at the 1 pecent level; ** at 5 pecent; and *** at 10 percent, respectively.

Gravity Model

Bilateral trade performance can be analyzed with a gravity model. The basis of the model is identical to the trade openness model, as cross-country variations in population and GDP per capita are used to explain variations in trade and export ratios. The gravity model adds as explanatory variables a proxy for the distance to the export market—in our case, we use the distance between the capital of the exporting country and India’s capital Delhi as a proxy. Although using the importing nation’s capital is a common approach, in this case by using the distance to Delhi as a proxy the gravity equation underestimates Sri Lanka’s trade potential with India given Sri Lanka’s close proximity to India’s southern trade centers. Using the distance to Chennai or Bangalore as proxies would predict an even higher export potential.

We follow the methodology used by Rodrick (1998) and Elborg-Woytek (2003), but estimate a standard model without additional explanatory variables other than the ones mentioned. Our sample includes pooled cross-country time series (1991–2004) for all SAARC countries. The regression results are shows in Table IV.1.2. It appears that, as expected from our estimation of the openness model, larger countries export less, while more developed economies tend to export more. The distance parameter has the expected negative sign; a larger distance between the two trading partners is associated with lower trade.

Table IV.1.2.

Regression Results for Trade Openness Model 1/

article image
Source: Fund staff estimates.

* stands for statistical significance at the 1 percent level.

STATISTICAL APPENDIX

Table 1

Sri Lanka: Gross Domestic Product and Expenditure Components, 2000–2004

article image
Source: Data provided by the Sri Lankan authorities.

Includes changes in stocks and investment by public corporations not financed through the government budget.

Including statistical discrepancy.

Table 2

Sri Lanka: Saving, Investment, and Current Account, 2000–2004

article image
Source: Data provided by the Sri Lankan authorities.

Total revenue minus current expenditure.

Includes investment by public corporations not financed through the government budget.3/

Includes net factor income and transfers from abroad.

Table 3

Sri Lanka: Gross Domestic Product by Industrial Origin at Current Prices, 2000–2004

article image
Source: Data provided by the Sri Lankan authorities.

Tea, rubber, and coconuts.

Table 4

Sri Lanka: Gross Domestic Product by Industrial Origin at Constant Prices, 2000–2004

article image
Sources: Data provided by the Sri Lankan authorities; and Fund staff estimates.

Tea, rubber, and coconuts.

Table 5

Sri Lanka: Trends in Principal Agricultural Crops, 2000–2004

article image
Sources: Sri Lanka Tea Board, Rubber Development Department; Coconut Development Authority; Department of Census and Statistics; Ministry of Agriculture; Paddy Marketing Board; National Fertiliser Secretariat, Plantation Companies; Central Bank of Sri Lanka.

Three major coconut kernel products only.

On a cultivation year basis.

20.9 kg. of paddy = l bushel of paddy.

Table 6

Sri Lanka: Consumption and Prices of Petroleum and Electricity, 2000–2004

article image
Sources: Data provided by the Sri Lanka authorities; Ceylon Petroleum Corporation; and Ceylon Electricity Board.

Including use for electricity generation.

End of period.

Price includes taxes.

Period average.

Unit cost of production including customs duty (all customs duties are charged to domestic sales), turnover taxes, and all other expenses.

Basic rate on household consumption of electricity between 50 and 500 kilowatt hours per month, excluding fuel surcharge levied on all users of electricity exceeding 150 Kwh per month, in Sri Lankan rupees per Kwh.

Basic rate on household consumption of electricity between 91 and 180 units in 2002.

More than 180 units.

Table 7

Sri Lanka: Price Indicators, 2000–2004 (Annual percentage change)

article image
Source: Data provided by the Sri Lanka authorities.

Based on market prices.

Low-income housing is under rent control.

Table 8

Sri Lanka: Selected Wage and Employment Developments, 2000–2004

article image
Source: Data provided by the Sri Lankan authorities.

Average of initial salary grades for non-executive and minor employees, skilled and non-skilled; excludes school teachers.

Includes employees of government ministries, provincial government, local government, school teachers, and defense personnel.

Includes universities, public corporations, boards, and state-owned banks.

Table 9

Sri Lanka: Labor Force, Employment, and Unemployment, 2000–2004

article image
Source: Department of Census and Statistics, Quarterly Labour Force Survey.

Average of first three quarters.

Less than General Certificate of Education (Ordinary Level).

Table 10

Sri Lanka: Employment by Economic Sectors, 2000–2004

article image
Source: Department of Census and Statistics, Quarterly Labor Force Survey.

Average of first three quarters.

Table 11

Sri Lanka: Summary of Central Government Operations, 2000–2005

article image
Sources: Data provided by the Ministry of Finance; and Fund staff estimates.

Includes Rs. 4,400 million net lending to CWE in 2003.