trade policies in the 1990s and the poorest countries

DURING THE 1960S AND 1970S, DEVELOP-ing economies exhibited severe trade-related distortions, including quantita-

tive restrictions on imports and exports, veryhigh tariffs, overvalued exchange rates, andadministrative controls on foreign exchangeallocation. Although growth remained rapidagainst a background of a favorable externalenvironment up to the first oil shock in 1973,policies of import control and substitution in-duced inefficiencies as well as rigidities in eco-nomic structure. Often, they resulted in peri-odic balance of payments crises. Subsequently,the failure of many countries to adjust ade-quately to the external shocks of the 1970sand early 1980s underlined the importance ofreforms designed to encourage responsivenessto market signals, improve the investment cli-mate, and to enhance export diversification.

One result was that most developing coun-tries implemented significant liberalization oftheir trade regimes during the late 1980s andthe 1990s. Their export growth acceleratedduring the 1990s, and kept pace with the 6percent per year expansion of world trade in volume terms. However, average per capitagrowth rates in developing countries as a groupremained well below those of the rich countriesin the 1990s. Though the giant low-incomecountries China and India embarked on marketreforms and grew rapidly, growth in a largenumber of small, poor countries was disap-pointing. This led many observers to questionthe success of liberalization programs.

This chapter reviews the export andgrowth performance of developing countriesin the 1990s, giving special attention to thepoorest economies. It reviews the decline intrade barriers during the 1990s, examines out-put and export trends, and analyzes domestictrade-policy constraints. Finally, it considersexternal barriers to developing countries’ ef-forts to accelerate their export growth.

The chapter reaches the following con-clusions:

• Trade regimes were significantly liberal-ized during the late 1980s and 1990s. De-veloping countries cut the average tariffrate by half, narrowed tariff dispersion inmany instances, and greatly reduced theincidence of nontariff barriers to trade.Most countries now rely on market forcesrather than administrative fiat to allocateforeign exchange, and black market pre-miums have declined significantly. Al-though the degree of trade protection isstill high in many developing countries,gross distortions in trade regimes havebeen greatly reduced.

• Despite the reforms and improved globaleconomic conditions, developing coun-tries’ average real per capita incomes in-creased by less than 1 percent per yearduring the 1990s, compared with morethan 2 percent in industrial countries.This outcome is partially affected by thepolitical shocks and various foreign and

1

Trade Policies in the 1990sand the Poorest Countries

2

Embargoed until Tuesday, December 5, 2 p.m. EST

Unpublished Proofs

civil conflicts. Eighteen developing coun-tries were severely affected by conflict,and as a group they suffered declines inper capita income of more than 1 percentper year. Incomes also fell in most transi-tion economies following the breakup ofthe Soviet Union. Excluding countries hitby these severe political shocks, develop-ing countries saw per capita incomes riseby 1.5 percent a year, about 1 per centfaster than in the 1980s.

• These countries also saw merchandise ex-port growth of 6.4 percent per year dur-ing the 1990s, about 2 percent faster thanduring the 1980s. Export outcomes werevery uneven, however. Regions that sawthe largest declines in trade barriers, in-cluding East Asia, South Asia, and LatinAmerica also saw the largest accelerationin exports. By contrast, growth in exportvolumes in Sub-Saharan Africa averagedonly 2 percent per year, in part becauseworld trade of the products Africa exportsgrew at half the rate of growth of worldtrade. Countries in Sub-Saharan Africa andin the Middle East and North Africa alsosaw market share decline in their tradi-tional exports.

• The poorest countries were those most af-fected by conflict and political shocks. Tenof the 32 low-income countries were af-fected by conflict. Average per capita in-come of the low-income, small countriesdeclined during the 1990s, but averaged 1 percent a year in real terms if countriesinvolved in conflict are excluded. This rep-resents a significant acceleration comparedwith the 1980s, but is still well below theaverage of middle-income countries. Ex-ports also accelerated in the small low-income countries not involved in conflict,but grew about 3.5 percent slower than inthe middle-income countries in the 1990s.

• Despite significant progress, many of thepoorest countries have not put in place thepolicies necessary to raise living standardsby improving (or even maintaining) ex-port shares in traditional markets and en-

couraging rapid diversification. Appreci-ated real exchange rates and high real ex-change rate volatility have often beenassociated with a muted export responseto trade liberalization and other reformmeasures. Per capita income growth wassignificantly faster in countries with rela-tively stable real exchange rates. Addition-ally, institutional weaknesses, such as the absence of effective duty exemption/drawback programs, coupled with theneed to use revenues from tariffs on inter-mediate and capital goods as a revenuesource, have acted as an effective tax onexports in many of the poorest developingcountries. Finally, weak export infrastruc-ture, inadequate ancillary export services,and high transport costs—often in part theresult of policy shortcomings—have leftmany of the poorest countries (particu-larly the landlocked ones) at a competitivedisadvantage on international markets.

• External barriers to exports from devel-oping countries, especially agricultural andlabor-intensive products, continue to im-pede the integration of the poorest eco-nomies into the world market. While theshare of developing countries in worldtrade of manufactures rose sharply in the1990s, their share of world trade in agri-cultural products and processed foods hasdeclined. In part this development is theresult of domestic policies that restrainagricultural exports. But high trade bar-riers imposed by industrial countries onagriculture and processed food imports,and agricultural subsidies in industrialcountries, are also important and have be-come even more important with the do-mestic policy reforms in developing coun-tries. These barriers particularly penalizerural areas where the majority of the poor in developing countries reside, andimpede growth in the poorest countries in areas of their comparative advantage.Various restrictions and subsidies in in-dustrial countries also hamper the poorestcountries’ efforts to diversify into down-

G L O B A L E C O N O M I C P R O S P E C T S

2 Embargoed until Tuesday, December 5, 2 p.m. EST

Unpublished Proofs

stream processing, higher value added,and faster-growing products. The poorestcountries are the least equipped to dealwith these external barriers because ahost of other domestic policy and institu-tional weaknesses inhibit their diversifica-tion into less restricted sectors.

Reductions in barriers to trade

Developing countries made substantialprogress in reforming their trade and ex-

change rate policies during the 1990s. Tariffswere cut, their dispersion declined in manycountries, fewer products were covered byquantitative restrictions (QRs), the number ofcountries allocating foreign exchange throughadministrative means (as measured by the In-ternational Monetary Fund) dwindled, and the black market premium narrowed. Althoughany one of these measures is an imperfect guideto the restrictiveness of the trade regime,1

taken together they show enormous progressin opening the developing economies to inter-national trade. Several studies have found thatincreased openness is associated with eco-nomic growth (box 2.1).

Tariffs. The average tariff rate in develop-ing countries has been cut by at least half inthe last 20 years, from 32 percent in the firsthalf of the 1980s to 15.6 percent in the secondhalf of the 1990s. Tariff reductions have beensignificant in most regions (figure 2.1), buthave been largest in South Asia (where, how-ever, they remain the highest of any region),Latin America, and East Asia. In many devel-oping countries, the degree of tariff dispersionhas also declined (table 2.1). Despite this prog-ress, tariffs remain high in many countries, av-eraging more than 15 percent in three of thesix geographical regions in 1996–98.

Latin American countries reduced tariffssubstantially during the 1990s. Colombia, forexample, slashed import tariffs by 65 percentin just one year in 1991, while Argentina andNicaragua reduced them from an average of110 percent to 15 percent in one bold move in1992 (Dornbusch and Edwards 1995). India

reduced its average tariff from 100 percent in 1986 to around 33 percent in 1998, andBangladesh’s average tariff fell from 82 per-cent to 24 percent during the same period.

Tariffs in Africa have also been reduced,though more moderately: the average un-weighted tariff remains at almost 20 percent.Some countries have embarked on rapid re-forms, however. Tariffs fell in Kenya from41 percent in 1980–85 to 13.5 percent in1996–99 and in Guinea from 76.4 percent in1978–80 to 10.8 percent in 1990–95. A fewcountries, such as Zimbabwe, increased theiraverage tariffs (from 10 percent in 1980–85 to22.7 percent in 1996–99), reflecting in part theconversion of an import surcharge into tariffs.3

The Middle East and North Africa region hasseen little reduction in average tariffs, althoughthe signing of trade agreements with the Euro-pean Union in recent years by several countriesin the region should eventually pave the wayfor significant reductions.

Nontariff barriers (NTBs). Developingcountries in all regions have substantially re-duced the coverage of NTBs (such as licensing,prohibitions, quotas, and administered pricing)during the 1990s (table 2.2).4 In many coun-tries where NTBs remain more or less preva-lent (for example, India and Korea), commit-ments have been made to liberalize them in thefuture (Michalopoulos 1999). NTB coveragehas been reduced significantly in a number ofLatin American countries, with the remainingNTBs mainly on agricultural products (Dorn-busch and Edwards 1995). Most countries inSouth Asia have reduced NTBs significantly,though India still has restrictions on a relativelylarge number of imports.

Exchange rate regimes. The 1990s alsosaw a general move toward market-based for-eign exchange regimes. In 1991, 66 countrieshad restrictions on payments for current ac-count transactions, but by 1995 only 28 did(table 2.3). Many countries also undertook sig-nificant exchange rate reforms. The averageblack market premium for developing coun-tries, one indicator of the restrictiveness of for-eign exchange allocations as well as of macro-

T R A D E P O L I C I E S I N T H E 1 9 9 0 s A N D T H E P O O R E S T C O U N T R I E S

3

Unpublished Proofs


economic imbalances, fell almost 70 percentbetween the 1980s and the 1990s.5

The average black market premium hasremained low or fallen further in all regions.East Asian countries have had low black mar-ket premiums in recent years because theywere early adopters of an outward-orienteddevelopment strategy. Latin America alsoachieved impressive gains in the late 1980s,and South Asia in the first half of the 1990s.In Sub-Saharan Africa and in the Middle East

and North Africa, the considerable reductionsin black market premiums in many countriesare masked by two outliers (table 2.4).

Reforms by income group. The trend to-ward lowering tariffs and eliminating foreignexchange restrictions is evident in both low-income and middle-income countries. Theaverage tariff rate for low-income countriesfell from almost 45 percent in the early 1980s to 20 percent in the late 1990s, only slightlyabove the average rate for middle-income



Unpublished Proofs

A large number of cross-country empirical studieshave documented a strong relationship between

trade and growth.2 Dollar (1992) finds that two sep-arate measures of trade restrictions (an index of realexchange rate distortion and an index of real ex-change rate variability) are negatively correlated withgrowth in developing countries. Sachs and Warner(1995) find that an openness index reflecting tariffand nontariff barriers, the black market premium,and the existence of commodity marketing boardshas a high and robust negative coefficient in growthequations. Ben David (1993) finds that income levelsin countries that joined the European Common Mar-ket have tended to converge toward the level of therichest partners. Similar studies have documented in-come convergence among states within the UnitedStates. Edwards (1997) finds a significant, positiverelationship between openness and productivitygrowth, which is robust to several different measuresof openness. In a recent study designed to identifymeasures of openness that cannot be said to be af-fected by economic growth, Frankel and Romer(1998) find that countries that are geographicallyclosest to centers of demand have grown faster.

The positive relationship between openness andgrowth has also been supported by numerous casestudies (Bhagwati 1978; Little, David, Scitovsky, and Scott 1970; and Papageorgiou, Choksi, andMichaely 1991), firm-level analyses (Bigsten andothers 1998; Clerides, Lach, and Tybout forthcoming;and Kraay 1997), as well as a combination of eco-nomic theory and anecdotal evidence dating back at

Box 2.1 Openness and growth—evidence,old and new

least to Adam Smith. But the statistical relationshipidentified by cross-country studies is still occasionallychallenged by researchers. For example, a recentanalysis (Rodriguez and Rodrik 1999) raises severalquestions about the findings, and argues that the ad-vocates of trade liberalization claim too much. Insome cases the indicators of openness that re-searchers use as measures of trade barriers are highlycorrelated with other sources of poor economic per-formance, including macroeconomic policy, or theyimperfectly reflect a country’s trade policy regime.For example, most of the explanatory power of theSachs and Warner index is derived from just two in-dicators—the black market premium and state mo-nopoly over major exports. These two indicators arecorrelated with a wide range of macroeconomic pol-icy and institutional factors other than trade open-ness, and thus yield an upward bias in the estimationof trade restriction effects. Rodriguez and Rodrik(1999) underline that tariff and nontariff barriers,two variables that directly measure trade openness,have little explanatory power when considered sepa-rately in the cross-country regression studies.

It is perhaps not surprising that the effects oftrade reform are difficult to isolate statistically, sincetrade liberalization is rarely implemented as a stand-alone policy measure, nor is such a course recom-mended. It has long been accepted that trade liberal-ization needs companion policies to be successful,often including other market reforms, macroeconomicstabilization, exchange rate adjustment, and adequatesafety nets. (See, for example, World Bank 1997b.)

countries (figure 2.2). By the late 1990s, mostof the low-income countries had eliminatedcurrent account restrictions and reduced theblack market premium to negligible levels.

Trends in trade and economicgrowth

The lowering of trade barriers and wide-spread adoption of market-based foreign

exchange regimes during the 1990s was ac-


5

Unpublished Proofs


Figure 2.1 Average unweighted tariff rates by regionPercent

Source: World Bank data, and World Trade Organization data.

0

South Asia Latin Americaand the

Caribbean

East Asiaand Pacific

Sub-SaharanAfrica

Middle Eastand

North Africa

Europe andCentral Asia

Industrialeconomies

10

20

30

40

50

60

70

1980–85 1986–90 1991–95 1996–98

Table 2.2 Frequency of total core nontariffmeasures for developing countries,1989–98

Region 1989–94 1995–98

East Asia and Pacific (7) 30.1 16.3Latin America and the Caribbean (13) 18.3 8.0Middle East and North Africa (4) 43.8 16.6South Asia (4) 57.0 58.3Sub-Saharan Africa (12) 26.0 10.4

Notes: Average number of commodities subject to nontariffmeasures as a percentage of total. Figures in parentheses arethe number of countries in each region for which data areavailable.Source: Michalopoulos 1999.

Table 2.1 Standard deviation of tariff rates

1990–94 1995–98

South Asia

Bangladesh 114.0 14.6India 39.4 12.7Sri Lanka 18.1 15.4

Sub-Saharan Africa

South Africa 11.3 7.2Malawi 15.5 11.6Zimbabwe 6.4 17.8

East Asia and Pacific

Philippines 28.2 10.2Thailand 25.0 8.9Indonesia 16.1 16.6China 29.9 13.0

Latin America

and the Caribbean

Argentina 5.0 6.9Brazil 17.3 7.3Colombia 8.3 6.2Mexico 4.4 13.5

Middle East and North Africa

Egypt, Arab Rep. of 425.8 28.9Tunisia 37.4 11.7Turkey 35.7 5.7

Notes: Country observations are for one year in the timeperiod noted above.Source: World Bank, World Development Indicators 1998;2000; WTO, Trade Policy Reviews.

Table 2.4 Average black market premium(percent)

Region 1980–89 1990–93 1994–97

Totala 82.0 78.2 20.3East Asia and Pacific 3.6 3.6 3.2Middle East and North Africa 165.6 351.6 46.5

Excluding outliersb 7.1 8.8 1.4Latin America and the Caribbean 48.7 13.1 4.4South Asia 40.8 45.1 10.1Sub-Saharan Africa 116.5 28.6 32.2

Excluding Nigeria 112.1 25.8 9.6

a. Sample of 41 developing countries.b. Algeria and the Islamic Republic of Iran.Source: World Bank data.

companied by an acceleration of trade in de-veloping countries. The improvement in the in-ternational environment in the 1990s, markedby lower interest rates and inflation but higherworld trade growth and non-oil commodityprices than in the 1980s, also contributed tothe acceleration of exports (table 2.5). More-over, developing countries faced a more stableglobal economy in the 1990s, as the volatilityof these key indicators declined. The growth ofdeveloping-country exports almost doubled inreal terms, to 6.2 percent a year in the 1990s,compared to 3.7 percent a year in the 1980s.6

Per capita income also accelerated, though moremodestly, from no growth in real terms in the1980s to 0.7 percent in the 1990s.

The observation that developing countriesas a group continued to exhibit relatively slowgrowth in per capita income in the 1990s—about one-third the rate of advance in the richcountries, despite large-scale liberalization and

export expansion, has sometimes been cited asevidence that increased integration with theglobal economy has not helped developingcountries, particularly the poorest. In fact, out-put declines in countries that were hit by se-vere political shocks, including the breakup of the Soviet Union and various external or internal conflicts, had a large effect on the developing-country aggregates, and the rise ofincomes in countries that did not suffer thesecalamities was considerable (table 2.6). Exclud-ing the transition economies and countriesinvolved in conflict, per capita GDP growthrates in developing countries averaged 1.5 per-cent per year during the 1990s, versus 0.4 per-cent in the 1980s, and export growth ratesaveraged 6.4 percent in the 1990s, comparedwith 4.3 percent in the 1980s. Furthermore,while developing countries became more openduring the 1990s, volatility declined in mostcountries (see box 2.2).



Unpublished Proofs

Table 2.3 Countries imposing restrictions on payments for current account transactions(percent)

Region 1980 1991 1995

East Asia and Pacific (9) 33 33 22South Asia (5) 100 100 40Middle East and North America (6) 67 67 33Sub-Saharan Africa (23) 85 83 39Latin America and the Caribbean (30) 44 60 17Europe and Central Asia (17) . . . 94 47Industrialized economies (12) 17 8 0Total (102) 55 65 27

Notes: Figures in parentheses are the number of countries in each regional grouping.Source: IMF, Exchange Arrangements and Exchange Restrictions, 1981, 1992, 1996.

With the exception of the countries intransition and conflict, there is little evidencethat developing regions that engaged in rapidtrade liberalization and companion reformssaw a deterioration in performance. In fact, de-veloping regions that saw the largest reductions

in tariffs during the 1990s also saw the mostrapid increases in GDP and exports.7 GDP andexport growth rates more than doubled inLatin America in the late 1980s and 1990s(table 2.7), a period that followed large ad-vances in trade liberalization as well as theadoption of important macroeconomic reformsand widespread privatization. South Asia andEast Asia also saw very sharp increases in ex-port growth rates and continued high GDPgrowth. By contrast, countries in Sub-SaharanAfrica and in the Middle East and North Af-rica made relatively less progress in reducingtariffs, and both GDP and export growth wasconsiderably slower than in the other regions.

Factors affecting developing country ex-ports. The acceleration of exports of develop-ing countries in the 1990s reflected highergrowth of world trade (itself a reflection inpart of the increased integration of developing


7

Unpublished Proofs


Figure 2.2 Merchandise export andGDP per capita growth in developingcountries in the 1990sPercent per year

7

6

5

4

3

2

1

0Developing countries

[84 countries]Developing countriesexcluding ECA andcountries in conflict

[59 countries]

Source: World Bank data.

Merchandise export growth, average 1990–98

GDP per capita growth, average 1990–98

Table 2.5 The international environment

1980–89 1990–98

U.S. Real LIBOR 4.8 2.4G-7 CPI (percent p.a.) 5.3 2.7G-7 Real GDP growth (percent p.a.) 2.8 2.1World Export Growth (percent p.a.) 4.4 6.4Oil price (index, 1987 = 100) 138.4 113.1Non-oil commodities price

(index, 1987=100) 115.6 123.0


Table 2.6 GDP and merchandise export growth rates(percent per year, in constant prices; group simple average)a

GDP per capita Merchandise exportsb

Number of1980–89 1990–98 1980–89 1990–98 countries

World 0.6 1.2 4.1 6.4 133Industrial countries 2.1 2.2 5.6 6.8 30Countries that lack consistent data –0.4 1.8 3.6 6.4 20Developing countries (excluding those

that lack consistent data) 0.3 0.7 3.7 6.2 83Europe and Central Asiac 2.6 –1.1 0.4 7.6 6Countries in conflict –1.1 –1.3 3.0 5.0 18Developing countries excluding ECA

and countries in conflict 0.4 1.5 4.3 6.4 59

a. All growth rates are estimated using least squares for the sample periods.b. Merchandise exports are deflated by constant 1987 U.S. dollar export prices.c. To give comparable data over 1980s and 1990s, the republics of the former Soviet Union are treated as one country.Source: World Bank data.

The increase in openness and acceleration ofgrowth during the 1990s were associated with

a small decline, and not an increase, in the volatilityof output and export growth in developing countries(see table). Output volatility increased significantly in countries affected by conflict and in Europe andCentral Asia (ECA), but fell for other developingcountries as a group and in three of the five geo-graphic regions. Although open economies are morevulnerable to external shocks, especially when theyare reliant on short-term capital flows (Easterly,Islam, and Stiglitz 2000), lower volatility in the1990s may have resulted from less severe global eco-nomic shocks than in the 1980s. The 1990s were, ofcourse, not tranquil for developing countries: theysaw the breakup of the Soviet Union, the Gulf Warfollowed by a global recession, the Tequila crisis, and

Box 2.2 Trends in volatilitythe global financial crisis of 1998–99. But the decadeof the 1980s was characterized by much larger oilshocks that contributed to two global recessions anddramatically affected economic volatility of oil ex-porters. Partly reflecting the oil shocks, the industrialcountries saw high inflation and international inter-est rates rose to high levels. The buildup of oil-related debts in the 1970s contributed to a massivedebt crisis in developing countries that lasted most ofthe decade.

Policies, as well as luck, may have contributedto the modest reduction in volatility in the 1990s,especially lower inflation in both industrial and de-veloping countries reflecting sounder macroeconomicpolicies. Increased integration with world marketsmay also have contributed to reduced volatility incountries where shocks originated domestically.

countries), as well as greater diversification oftheir export base and, for some developingregions, improved market share in traditionalmarkets (table 2.8).8 While South Asia, EastAsia, and Latin America saw both increases inmarket share in traditional markets and some

diversification, countries in the Middle Eastand North Africa and Sub-Saharan Africa sawsome diversification, but a large decline in mar-ket share in traditional markets.

Sub-Saharan Africa stands out because theregion experienced much slower growth of world



Unpublished Proofs

Volatility of growth rates

GDP per capita Merchandise exports Merchandise exports(1987 US$)a (1987 US$)a (import prices deflator)b

1980s 1990s 1980s 1990s 1980s 1990s

All countries 4.5 4.2 15.2 14.1 17.9 15.2Industrial 2.9 3.4 7.3 8.1 8.5 8.9Developing 5.0 4.4 17.6 16.0 20.8 17.1

Conflicts 6.3 7.3 21.3 28.9 25.6 30.0Europe and Central Asia 4.1 7.0 10.6 17.7 12.2 17.5Other countries 6.4 4.1 24.8 19.6 28.0 18.1Developing excluding ECA and countries

in conflict 4.8 3.6 17.3 13.2 20.3 14.4Sample 4.3 3.5 15.1 11.2 18.0 13.2

East Asia 3.8 4.6 10.4 10.8 12.8 9.1Middle East and North Africa 5.0 3.6 15.4 7.7 21.7 14.0Latin America and the Caribbean 4.3 2.7 12.6 13.0 15.6 14.6South Asia 2.1 2.6 11.8 8.2 14.0 10.0Sub-Saharan Africa 4.6 3.7 18.4 12.3 20.2 14.0

a. Standard deviation of growth rates expressed as a percentage in constant U. S. dollars.b. Standard deviation of growth rates of merchandise exports, deflated by import prices and expressed as a percentage.Source: World Bank data.

trade in its export basket in both decades thanother developing regions. Reflecting the contin-ued decline in primary commodity prices andlow-income elasticities of demand, world tradegrowth rates during the 1990s were less than 2 percent annually for exports of some of the poorest African countries, including Benin,Chad, Mali, and Mauritania, among others.

Trends in the poorest developing coun-tries. Excluding China and India, which saw

significant increases in per capita incomes inthe 1990s, per capita income declined in the1990s in the poorest developing countries as a group (figure 2.3). Since 1980, 10 of the 32low-income countries with consistent datahave been involved in foreign or civil wars (seebox 2.3). In 1999, one African in five lived ina country severely disrupted by conflict (WorldBank 2000a). Excluding countries affected byconflict, per capita GDP growth rates in small


9

Unpublished Proofs


Table 2.8 Decomposition of merchandise export growth for the sample countries(percent per year, current U.S. dollars)

1980/81–1989/90 1989/90–1997/98

World WorldTotal Trade Market Total Trade Market

Exports Growth Share Diversification Exports Growth Share Diversification

Total 3.9 4.1 –1.0 0.8 5.9 5.4 0.0 0.6East Asia

and Pacific 8.5 5.2 2.1 0.9 11.9 6.5 4.9 0.1Middle East and

North Africa 2.4 3.1 –1.3 0.6 3.0 6.2 –3.5 0.5Latin America

and theCaribbean 3.9 5.4 –2.2 0.8 8.3 6.7 0.6 0.9

South Asia 8.9 5.9 1.5 1.5 10.0 6.4 2.8 0.6Sub-Saharan Africa 1.0 2.9 –1.6 0.8 3.0 3.8 –1.2 0.6

Source: COMTRADE database, and authors’ calculations.

Table 2.7 GDP, services, and merchandise export growth rates(percent per year, in constant prices; group simple average)

ExportsPurchasing power of

GDPa Merchandisea Goods and servicesa GNFS exportsc

Number of1980–89 1990–98 1980–89 1990–98 1980–89 1990–98 1980–89 1990–98 countries

Totalb 3.1 3.8 4.3 6.4 4.5 6.3 2.2 6.2 59East Asia

and Pacific 5.4 6.3 5.7 14.5 5.9 14.3 4.1 13.7 7Middle East and

North Africa 3.6 3.4 5.3 6.0 5.5 5.0 0.2 3.3 9Latin America

and theCaribbean 1.8 3.6 3.9 9.1 4.8 8.1 1.6 8.8 14

South Asia 4.9 5.2 5.8 9.3 5.0 10.3 5.3 10.7 5Sub-Saharan

Africa 2.5 2.9 3.5 2.0 3.4 2.6 2.0 2.7 24

a. All growth rates are simple arithmetic averages of individual rates computed using least squares. Nonfactor services are de-flated by constant 1987 GDP deflator; other series are expressed in constant U.S. dollars.b. Refers to a sample of 59 developing countries.c. Denotes average annual growth rate of GNFS export deflated by merchandise import prices.Source: World Bank data.

low-income countries averaged 1 percent peryear during the 1990s, well below the 1.5 per-cent rate of growth achieved by middle-incomecountries (table 2.9).

Merchandise export volume growth ratesin the 22 small, low-income countries (exclud-

ing countries affected by conflict) averagedonly 4.1 percent per year during the 1990s, upfrom 2.9 percent in the 1980s but still wellbelow performance in the middle-income coun-tries. More than two-thirds of these small low-income countries (16 of 22) are in Sub-Saharan



Unpublished Proofs

Figure 2.3 Merchandise export growth and GDP per capita growth in poor developing countries in the 1990sPercent per year

–1

Low-income smallcountries [32 countries]

Low-income smallcountries excludingcountries in conflict

[22 countries]

Least developedcountries

[26 countries]

Least developedcountries excludingcountries in conflict

[16 countries]

0

1

2

3

4

5




There is a large body of literature documenting thenegative impact of wars and rebellions on a

country’s growth performance, and there is also evi-dence that economic factors also contribute to con-flict. Using a panel data set for 161 countries, Collierand Hoeffler (1999) find that among other factors,the risk of civil war is significantly and negatively re-lated to growth in per capita income. A reliance onexports of primary commodities is positively relatedto civil conflict. For a country with a primary com-modity export share of around 0.26, the risk of acivil war is around 23 percent (compared with 0.5percent for countries with no natural resource ex-ports), in part because the capture of resources either

Box 2.3 Economic factors contributing to conflictmotivated or made possible several rebellions. Dia-monds in Angola and Sierra Leone and drugs inColombia are some obvious examples.

Countries that are commodity-dependent are sub-ject to larger external shocks. Rodrik (1998) arguesthat while external shocks can lead to distributionalconflicts, societies that are cohesive and have stronginstitutions can adjust more easily to such shocks andavoid conflict. He compares Brazil, Korea, and Turkey,all of which suffered sharp trade declines during the1970s. Korea was the hardest hit, as it was more openthan either of the other two, but it recovered muchfaster. Adjustment was delayed in both Turkey andBrazil, and they witnessed considerable turmoil.

Africa. An important factor accounting for theslow growth of the poor countries’ exports wasslow growth of world trade in their traditionalexport baskets. Several of the poorest develop-ing countries lost market share in traditionalexports and were unable to diversify rapidlyenough into new products.

Summarizing, developing countries not inconflict and not in transition saw significantacceleration in incomes and exports in the1990s, following adoption of trade and com-panion reforms. The poorest small countrieswere among those most affected by conflictsand political shocks. The poorest small coun-tries that avoided these calamities performedbetter, but still not as well as the middle-income countries, on average. Thus, despite sig-nificant progress, it appears that policy regimesin many of the poorest countries are still in-adequate to improve or even maintain exportcompetitiveness in traditional markets or toencourage rapid diversification. The next twosections will discuss the key trade-related pol-icy impediments to rapid integration of thepoorest countries in the world economy, in-cluding trade barriers to their exports in indus-trial countries.

Weaknesses in domestic trade-related policies

Domestic policies that directly restrict ex-ports or deter investment in the export

sector remain important in most of the poorestdeveloping countries (Yeats and others 1997).Policies that tend to discourage investment in

the export sector include overvalued real ex-change rates, inconsistent or erratic macroeco-nomic policies, a high share of governmentspending in aggregate expenditure, excessivereliance on tariff revenues and on taxes onagriculture, and a variety of direct public inter-ventions in product and factor markets, espe-cially price controls and requirements that bankcredit be allocated to the public sector (WorldBank 1989, 1994, 2000a). Weak infrastructureand inadequate provision of ancillary servicesto exports also severely constrain export per-formance in many countries. As discussed inprevious sections, some of the most severe pol-icy distortions present in developing countriesin the 1980s are being alleviated: average in-flation and fiscal deficits have declined signifi-cantly (Easterly 2000); real exchange rateshave been adjusted in many cases; tariffs andquotas have been cut; and export taxes and re-strictions on agriculture have been reduced oreliminated in many countries.

This section examines three aspects oftrade-related domestic policies that remain im-portant impediments to integration of many ofthe poorest developing countries into the globaleconomy: overvalued and unstable real ex-change rates, high cost of imported inputs forexporters, and the higher costs of transporta-tion and other trade-supporting infrastructure.

Exchange rate management. A competi-tive and stable real exchange rate ensures thatproducing tradable goods is profitable, and tounderpin investor confidence in the export sec-tor. In many of the poorest countries, pro-tracted episodes of real exchange rate overval-


11

Unpublished Proofs


Table 2.9 Growth rates by income level for the sample countries(percent per year)

GDP Merchandise exports GDP per capitaNumber of

1980s 1990s 1980s 1990s 1980s 1990s countries

Total 3.1 3.8 4.3 6.4 0.4 1.5 59Low income (large countries) 7.9 8.4 6.7 14.1 6.1 6.9 2Low income (small countries) 2.7 3.6 2.9 4.1 –0.1 1.0 22Least developed 2.2 3.5 2.9 3.2 –0.6 0.8 16Middle income 3.0 3.6 5.1 7.4 0.5 1.5 35

Note: The sample (59 developing countries) excludes countries in conflict, transition economies, and those with limited data.Source: World Bank.

uation have impaired the competitive positionof local firms. At different times, they havecontributed to unsustainable trade deficits,and prompted increased tariffs, quantitativerestrictions, and foreign exchange controls infutile attempts to contain these deficits. Epi-sodes of overvaluation have also resulted inforeign exchange crises, economic instability,and the deterioration of growth over the longrun. At the same time, unsustainable macro-economic policies have also contributed to theovervaluation of real exchange rates (Bhagwati1978; Krueger 1978; Papageorgiou, Choksi, andMichaely 1991; and Rodrik 1996).

The effects of real exchange rate overval-uation have been studied extensively. Dollar(1992) shows that Africa and Latin Americahad relatively overvalued exchange rates andlarge real exchange volatility, while East Asiahad relatively undervalued exchange rateslower real exchange rate volatility, and morerapid growth. Using a wide variety of mea-surements and techniques, econometric stud-ies have fairly uniformly found that overval-uation and associated real exchange ratevolatility are negatively correlated with bothexport and GDP growth.9

The factors that lie behind protractedepisodes of real exchange rate overvaluation,despite awareness of their adverse implica-tions for growth, are varied. They include useof the exchange rate as a nominal anchor forthe stabilization program and difficulties inexiting from the peg; rigid exchange rateregimes that fail to accommodate secular dete-rioration in the terms of trade and adjust towidening productivity differentials; dispropor-tionate influence of urban elites who are con-sumers of imported consumer goods; concernsabout the effect on the urban poor dependenton imported food staples or energy; and aver-sion to incur the immediate fiscal costs reval-uating foreign currency liabilities.

Perhaps because of increased openness, inrecent years policymakers appear to havebecome more aware of the need to maintaincompetitive and stable real exchange rates. Theaverage exchange rate of developing coun-

tries depreciated during the 1990s (Easterly2000), and many countries adopted more flex-ible nominal exchange rate arrangements.10 Insome cases, for example the CFA countries(see box 2.4), these efforts resulted in a sharpturnaround in economic performance.

Several of the poorest countries failed toachieve stable and competitive real exchangerates during the 1990s. Overall, the averagevolatility of real exchange rates in the small,low-income countries increased. High exchangerate volatility was associated with weak eco-nomic performance. Among the countries inour sample (excluding countries affected byconflict), the poor countries that exhibited lowexchange rate volatility saw exports rise at over6 percent a year and per capita incomes rise at nearly 2 percent a year, much faster thancountries with high exchange rate volatility(figure 2.4).

As countries became more open to inter-national trade, the adverse effects of real ex-change rate volatility on stability and growthmay have become even more severe. In somecases, trade liberalization combined with sharpreal exchange appreciation severely impairedthe profitability of domestic firms.11 Ex-change rate overvaluation and high volatilityalso has a negative effect on investor con-fidence and can delay the supply response toliberalization.12 Figure 2.5 shows six coun-tries in Africa that moved to flexible exchangerates and liberalized their trade regimes duringthe early 1990s. These countries subsequentlyexperienced appreciation and high volatilityof the real exchange rates and low rates ofgrowth. Per capita income growth in these sixcountries averaged only 0.5 percent a year inthe 1990s, and their export effort (changes inmarket share and diversification) fell by 2.3percent a year.

The causes of high real exchange ratevolatility have been widely explored. They in-clude unstable or inconsistent macroeconomicpolicies, which result in inflationary pressuresand high fiscal and current account deficits.Many countries have tried to maintain stablenominal exchange rates, despite the formal



Unpublished Proofs

adoption of more flexible exchange rate re-gimes during the 1990s (Calvo and Reinhart2000).13 However, attempts to maintain nom-inal exchange rate stability in the face of highand variable inflation rates or various domes-tic or foreign shocks imply that the real ex-change rate will be volatile. For example, 17of the 26 countries that exhibited high real ex-change rate volatility were also rated by theWorld Bank as exhibiting relatively unstablemacroeconomic policies.14 Among the 22 low-income countries in our sample, only twomanaged to achieve both macroeconomic sta-bility and low real exchange rate volatility.

In addition, terms-of-trade shocks are animportant factor contributing to high real ex-change rate volatility in commodity-dependentcountries. Hausman and others (1999) cite thevolatility of capital flows as a driving force be-hind the volatility of exchange rates in manymiddle-income countries, particularly in con-junction with large debts denominated in for-eign currency.

Thus, the record of the 1990s underlinesthe difficulties that poor countries face in main-taining a stable and competitive real exchangerate, reflecting their thin foreign exchange markets, their vulnerability to weather-relatedshocks, and their dependence on primarycommodity exports that are subject to sharpchanges in price.15

Exchange rate policies are hampered bythe lack of clear signposts as to the exchangerate regime that minimizes real exchange ratevolatility. The choice of exchange rate regimewill depend on many factors, including thecurrency composition of public and privateexternal debt and domestic financial assets,the track record and credibility of the mone-tary authorities, and the nature of externaland internal shocks.16

Free trade status for exporters. High tar-iffs coupled with inefficiencies in customs andtax administration have increased the costs ofexporting from many developing countries.Access to inputs at world prices is critical to


13

Unpublished Proofs


the annual rate of increase in exports due to risingmarket shares and diversification, rose from –4.8percent per year in 1986–93 to 7.8 percent per yearin 1994–98.

Exchange rates and CFA countries

1986–93 1994–98

Average GDP growth 0.7 4.7Average REER index 99.4 63.0Export efforta –4.8 7.8

Note: The countries included are Burkina Faso, Cameroon,Central African Republic, Chad, Côte d’Ivoire, Gabon,Mali, Niger, Senegal, and Togo.a. The merchandise export growth rate achieved throughchanges in market share and export diversification.Source: IMF, International Financial Statistics.

The CFA countries provide one of the most strik-ing examples of the impact of exchange rate

overvaluation on growth and the gains from im-proved exchange rate policies. The currencies ofthese countries are tied to the French franc, and atvarious times the appreciation of the franc, deterio-ration in the terms of trade, and higher inflation inCFA countries contributed to a highly appreciatedcurrency that tended to depress export and GDPgrowth, the latter to less than 1 percent per year for1986 to 1993 (see box table). The CFA franc was de-valued in 1994, and GDP growth accelerated to 5percent per year from 1994 to 1998. Export growthalso accelerated, reflecting improved market share intraditional products and greater diversification, aswell as acceleration of world trade in the CFA coun-tries’ export products: export effort, which measures

Box 2.4 Exchange rate overvaluation in the CFAcountries

export competitiveness, so that high tariffs onintermediate and capital goods can reduce ex-ports, particularly of processed goods that re-quire substantial intermediate inputs. At thesame time, many developing countries dependon tariffs for an important share of govern-ment revenues. Thus governments often face adifficult tradeoff between the desire to encour-age export production through low tariffs onimported inputs and the need to generate suf-ficient revenues.

Successful exporters such as the Asiannewly industrializing countries, in the earlystages of their export drive, have enabled ex-port firms to obtain their inputs at world mar-ket prices through reimbursing tariff duties oninputs (duty drawback systems), providing ex-emptions on duties for exporters (duty exemp-tion systems), and setting up export process-ing zones where intermediate inputs can enterthe country free of duty.17 These systems havebeen effective in reducing or eliminating tariffson exporters’ inputs, while maintaining a tar-iff structure that meets the revenue needs ofthe government. These systems, however, re-quire considerable administrative resources to

ensure that reimbursement or exemption isgranted quickly and fairly.

By contrast, the poorest developing coun-tries have faced severe difficulties in ensuringduty-free access to inputs by exporters, for tworeasons. First, a substantial share of govern-ment revenues in many of these countries is gen-erated from taxes on international trade. To re-duce distortions affecting domestic productionand to simplify administration, tariff reform inmany poor countries involves increasing lowtariffs while reducing peak tariffs (Falvey andKim 2000; Harberger 1988). As a result, manyof the poor countries have significant tariffs onintermediate and capital goods (table 2.10). Al-though the average tariff rates for this sample of15 African countries are not high comparedwith those of many other developing countries(figure 2.1), the duties on capital and interme-diate goods do represent a considerable tax onexport production.18 These duties are essen-tially designed to collect revenues, rather thanto protect domestic production.19

Second, most of the poorer countries lackthe administrative resources required to estab-lish effective drawback/exemption schemes.



Unpublished Proofs

Figure 2.4 Real effective exchange rate volatility and growth in 1990sPercent per year

Countries with highvolatility in 1990s

[26 countries]

Countries with lowvolatility in 1990s

[33 countries]

Low-income smallcountries with highvolatility in 1990s

[13 countries]

Low-income smallcountries with lowvolatility in 1990s

[9 countries]

0

1

2

3

4

5

6

7

8

9

Source: World Bank data; IMF, International Financial Statistics.




15

Unpublished Proofs


Figure 2.5 Real effective exchange rate behavior in selected poor countries, 1993–99

60

70

80

90

100

Ghana Kenya

Malawi Tanzania

Zambia Nigeria

70

80

90

100

110

120

130

140

1993 1994 1995 1996 1997 1998

40

60

80

100

120

80

100

120

140

160

60

80

100

120

140

160

60

80

100

120

140

160

180

200

1993 1994 1995 1996 1997 1998

1993 1994 1995 1996 1997 1998

1993 1994 1995 1996 1997 1998

1993 1994 1995 1996 1997 1998

1993 1994 1995 1996 1997 1998

Note: 1990=100, upward movement indicates appreciation.Source: IMF, International Financial Statistics; World Bank data.

Attempts to establish such systems during the1990s were generally not successful. The re-quirement of examining individual importtransactions imposed a substantial administra-tive burden on exporters and the government.In many countries, the government failed toprovide (or significantly delayed) promised re-imbursements of duties paid on imported in-puts.20 Even exporters’ value added taxes werenot reimbursed consistently. Ongoing work on15 reforming Sub-Saharan African countriesshows that none had a working duty drawbacksystem. Mauritius operated a successful exportprocessing zone program, while Cape Verdeand South Africa had low duties on intermedi-ate goods. The 12 other countries imposedsignificant duties on inputs to export produc-tion without effective means of compensatingexporters.

There is no easy solution for poor coun-tries attempting to ensure duty-free access toinputs by exporters.21 Further efforts are re-quired to strengthen the administration of cus-toms and to increase reliance on other sourcesof revenues such as income taxes. However,such efforts take time. In the short term, de-pending on the structure of export activity (forexample, degree of processing) and the level of institutional development, governmentsneed to decide whether drawback/exemptionsystems or free trade zones are feasible, or al-ternatively whether reductions in intermediateand capital goods tariffs are necessary to en-courage export production. Since all but five ofthe countries in table 2.10 have total collectionrates of less than 10 percent it is possible to

move to a tariff structure that has zero or verylow rates for capital and intermediate goodsalong with much lower rates for consumergoods (coupled with reduced exemptions).

Infrastructure for trade expansion. Weakinfrastructure constrains economic growth inmany of the poorest countries and has a severeimpact on trade performance. A firm’s abilityto compete in the world economy depends inpart on the cost and availability of supportingservices, such as transport, communications,and finance. While some of the problems thesecountries face in providing adequate infra-structure and other services are to some extentthe result of location, technological deficien-cies, and a lack of capital, many of these prob-lems result from restrictive regulatory policiesthat limit competition.22

Many of the least developed countries haveweak and expensive service suppliers. Amjadi,Reincke, and Yeats (1996) conclude that hightransport costs in low-income African countriesare a more important trade barrier than tar-iffs.23 These high costs add to the cost of ex-porting, and (other things being equal) lower itsprofitability in many of these countries. Coun-tries in Africa normally absorb all, or most, ofthe transport charges for penetrating externalmarkets. Africa’s net freight and insurance pay-ments in 1990/91 were about $3.9 billion, ap-proximately 15 percent of the total value of the region’s exports. For developing countriesas a whole similar payments averaged 5.8 per-cent, about one-third of Africa’s ratio. Thus, inorder to be competitive, an average producer inAfrica has to be 10 percent more efficient than



Unpublished Proofs

Table 2.10 Tariffs in selected African countries

Average nominal Trade weighted Collection Share oftariff average Rate tariff revenues

All goods 19.8 16.2 10.3 100.0Capital goods 13.7 13.2 7.8 12.9Intermediate goods 16.9 14.2 8.1 42.4Consumer goods 23.6 15.3 10.9 22.4

Note: Textiles, energy, and passenger cars are not included in capital, intermediate, and consumer goods. Data are for 1996 forBenin, Burkina Faso, Cameroon, Cape Verde, Côte d’Ivoire, Ghana, Malawi, Mauritius, Senegal, South African Customs Union,Tanzania, Uganda, Zambia, and Zimbabwe; data for Mali are for 1997.Source: World Bank data.

firms in other developing countries. For the 10landlocked countries in Africa, average netfreight and insurance payments in 1990 were42 percent of their total exports, eight times theaverage for other developing countries.

Limão and Venables (1999) also find thathigher transport costs and weak infrastructureexplain a significant portion of Africa’s poortrade performance. This is especially true forlandlocked countries. Exporters from land-locked economies face final city destinationsthat are on average four times further from thesea than that of exporters from coastal eco-nomies. The median transport cost for land-locked countries is 58 percent higher than themedian for coastal countries. Delays and coor-dination problems at the border, higher insur-ance costs due to uncertainty and delays, anddirect charges made by transit countries alsoadd to the transport costs of landlocked coun-tries. Improving transport infrastructure inlandlocked countries and their transit coun-tries can dramatically reduce costs and increasetrade flows. Improving a country’s worldwiderank from the 75th percentile to the 50th per-centile in the distribution of infrastructurequality would double the volume of trade.

Poor availability of communications andenergy can also constrain exports. Unreliableservice can be even more damaging to com-petitiveness than high costs. Production stop-pages, missed delivery dates, and lack of re-liable communications make it difficult tocompete and become part of global produc-tion networks. Despite liberalization of the fi-nancial systems in many countries, access tocredit is limited, interest rate spreads are high,and in many countries weakness in regulatoryinstitutions have kept financial sector reformsfrom having their anticipated benefits.

Reforms are being undertaken in manydeveloping countries to liberalize the policyregimes for these sectors. Foreign participa-tion is being allowed to improve the level oftechnology and efficiency, and also to generatethe financing for the required investments.The potential gains from these reforms and re-sulting investments are very large. For exam-

ple, allowing exporters to charter their ownvessels in Côte d’Ivoire halved the costs ofshipping bananas to the United States, and re-duced cocoa freight costs one-quarter (WorldBank 2000a). Improving the operating effi-ciency of the Nacala rail line through Mozam-bique, with little new investment, would in-crease the GDP of Malawi by 3 percent. Theseefforts may have a significant role in improv-ing export performance over the next decade(AfDB 1999). Donors also are engaged in theefforts to help the low-income countries toovercome the major institutional impedimentsthat inhibit their integration into the worldeconomy (see box 2.5).

Protection in industrial countries

Import restrictions and subsidies in industrialcountries limit the growth of developing

countries’ exports by supporting less efficientproduction in industrial countries. Export sub-sidies to industrial country producers furtherlimit the expansion of developing-country ex-ports to third markets. Moreover, these poli-cies make it difficult for developing countriesto diversify into products for which world de-mand is high or increasing, and in line withtheir evolving competitive advantage.

Industrial-country import restrictions. Al-though the average tariffs in the Quad (Can-ada, European Union, Japan and the UnitedStates) countries range from only 4.3 percentin Japan to 8.3 percent in Canada, their tariffsand trade barriers remain much higher onmany products exported by developing coun-tries (Finger and Laird 1987). The UruguayRound contributed to a sharp decline in theuse of nontariff barriers (NTBs) in the OECD.In the Quad, only 1.2 percent of tariff lines are subject to NTBs. However, most of theNTBs are found in agriculture (tariff quotas,for example)25 and textiles and clothing (Multi-fiber Arrangement), where developing coun-tries have a comparative advantage (Fingerand Schuknecht 1999).

Products with high tariffs in Quad coun-tries include (1) major agricultural staple food


17

Unpublished Proofs


products, such as meat, sugar, milk, dairy prod-ucts, and chocolate, where tariff rates fre-quently exceed 100 percent; (2) tobacco andsome alcoholic beverages; (3) fruits and vegeta-bles—including 180 percent for above-quotabananas in the European Union and 550 per-cent and 132 percent for shelled groundnuts in Japan and the United States, respectively;(4) food industry products, including fruit juices,canned meat, peanut butter, and sugar confec-tionery, with rates exceeding 30 percent inseveral markets; and (5) textiles, clothing, andfootwear, where tariff rates are in the 15 to 30

percent range for a large number of products.These are sectors in which developing countrieshave a comparative advantage.

For example, in the United States only 311of 5,000 tariff lines are above 15 percent.26

Yet 15 percent of exports from least developedcountries to the United States face these tariffs(Hoekman, Ng, and Olarreaga 2000). Thus,there might be considerable potential for theleast developed countries to increase their ex-ports if U.S. tariffs were reduced.

Some of the highest tariff rates in indus-trial countries are applied to products that are



Unpublished Proofs

The least developed countries (LDCs) are particu-larly disadvantaged by a lack of capacity to pro-

vide the minimal public and private trade facilitationservices and infrastructure required to support inter-national trade.24 The Uruguay Round agreement in-cluded promises by industrial countries to providetechnical assistance to help developing countriesstrengthen their trade services, but these promiseswere not binding, and the reality has been disappoint-ing (Michalopoulos 1999; Wang and Winters 2000).

The Integrated Framework (IF) was establishedin 1996 to increase the effectiveness and efficiency oftrade-related technical assistance to the countries, inpart by strengthening coordination among participat-ing agencies and ensuring that technical assistance isdemand driven. Participating agencies are the WorldTrade Organization, the International MonetaryFund, the International Trade Center, the United Na-tions Development Programme, the United NationsConference on Trade and Development, and theWorld Bank.

By the summer of 2000, 40 LDCs had com-pleted the first step of the IF process (a needs assess-ment). The six international agencies responded, in-dicating areas in which they could assist. Thisprocess revealed little overlap in the activities of theagencies and substantial needs that required addi-tional financing. Progress in organizing IF roundta-bles to mobilize donor resources proved difficult; by

Box 2.5 The integrated framework for least developed countries

August 2000 only five roundtables had taken place.In only one case (Uganda) did the roundtable lead tothe commitment of new funds from donors.

An independent review of the IF, completed inJune 2000, highlighted that recipients expected addi-tional funding, while donors were hoping to increasethe effectiveness of technical assistance through im-proved coordination among agencies. The reportnoted a lack of clear priorities, weak administration,and a lack of donor resources. The review recom-mended that trade needs be better integrated in na-tional development strategies, that steps be taken tostrengthen the secretariat and coordination func-tions, and that a trust fund be established for IFactivities. Although the memberships and governingbodies of the core IF agencies broadly support thefirst four suggestions, to date there has been littleconcrete support for the trust fund proposal. Effortsare ongoing to mobilize the required resources.

The six agencies are committed to continue todeliver trade-related technical assistance to least de-veloped countries within their existing resource con-straints, pursuant to their mandates and competence.The World Bank’s activities in trade policy, trade fa-cilitation, and export development are primarily un-dertaken at the country level and based on the coun-try’s development strategy. Consequently, the Bank’sefforts in mainstreaming trade will be driven by thepriorities identified by each country’s government.

typically exported by developing countries.For example, almost $26 billion of exportsfrom developing countries in 1999 to theworld were products that would have facedtariffs above 50 percent in the Quad countries.Only about $5 billion of that sum was actuallyexported to the Quad countries. On the otherhand, Quad countries imported about $50 bil-lion of the same goods, most of it from otherindustrial countries.27 This suggests some po-tential for developing countries to expand ex-ports of these products to the Quad if the tar-iffs were lowered. Although $26 billion is onlyabout 2 percent of total developing-countryexports, individual developing countries (in-cluding among the least developed countriesin the above example) may be more affected.

The tariff rates in table 2.11 do not reflectthe Generalized System of Preferences (GSP)and other preferential schemes operated byQuad countries. These schemes tend to reducethe tariff rates applicable to imports of theseproducts from some developing countries.North-South preferential trade agreementssuch as NAFTA or the Euro-MediterraneanAgreements also provide duty-free entry forsome developing countries.28 However, in mostcases where tariff peaks are present, the sensi-tivity of domestic industry to imports excludevarious products from preferential schemeslimit the amount that can be imported underthe preferential rates, or restricts the number ofcountries that are eligible. For example, theUnited States has no GSP or least-developedcountry preference for products facing tariffsof more than 50 percent.29 The EU limits pref-

erential margins and imposes country or sectorquotas, or both (Michalopoulos 1999).

The potential for growing exports in therestricted categories is illustrated by those de-veloping countries that have managed to accel-erate their agricultural exports through freetrade arrangements with industrial countries(see chapter 1 for a discussion of the benefits ofNorth–South regional trade agreements). Forexample, since Mexico joined NAFTA in 1994,the value of its agricultural exports has in-creased by 15 percent per year—twice the rateof the previous five years. After joining the EUin 1986, Portugal’s agricultural exports grewby 9.2 percent per year in nominal U.S. dollars,and Spain’s exports grew by 10.9 percent. Inthe prior five years, both Portugal and Spainhad seen declining export earnings from agri-culture. These experiences suggest that furthertrade liberalization in agriculture is likely tohave a major effect on developing countries’abilities to increase agricultural exports.

Developing-country trade barriers. Devel-oping countries’ exports are subject to muchhigher trade barriers in other developing coun-tries than in industrial countries. The averagetariff developing countries face in their exportsof manufactures to other developing countriesis 12.8 percent, more than three times the av-erage tariff on their manufactured exports toindustrial countries (table 2.12) (Hertel andMartin 1999). Tariffs in developing countries


19

Unpublished Proofs


Table 2.12 Average tariff rates byimporting and exporting region(percent)

Importing region

High-income DevelopingExporting region countries countries

ManufacturesHigh-income 0.8 10.9Developing 3.4 12.8World 1.5 11.5

AgricultureHigh-income 15.9 21.5Developing 15.1 18.3World 15.6 20.1

Source: Hertel and Martin 2000.

Table 2.11 Developing-country exports toQuad countries facing tariffs of more than50 percent

Average most favored nation tariff (%) 113Range of tariffs (%) 50–343Exports to Quad (billion US$) 5.0Share of developing countries in

total imports of Quad (%) 10Exports to the world (billion US$) 26.6

Source: OECD, 2000b; UN COMTRADE; and staff calculations.

Table 2.13 Producer support estimates for OECD countries

Percent US$ billion

1986–88 1997–99 1986–88 1997–99

Lower-income OECD countries 37.5 28.7 28.8 39.4Australia, New Zealand 9.5 4.5 1.8 1.4Canada, United States 29.5 18.5 47.5 47.8European Union 44.0 44.0 95.2 116.6Other non-EU OECD 71.0 66.7 7.8 7.8Japan 67.0 61.0 53.6 53.1OECD 41.8 34.9 234.8 266.2

Source: OECD 2000a.

are somewhat higher in agricultural markets aswell. These high tariffs are becoming all themore important to developing countries, as theshare of South–South trade in their exports hasrisen from about 26 percent in 1980 to 40 per-cent in 1999.

Industrial-country agricultural subsidies.Agricultural subsidies in industrial countrieslimit the growth of developing countries’ agricul-tural exports by supporting inefficient producers.Furthermore, surplus agricultural products havebeen exported at a loss, further reducing the op-portunities for many developing-country pro-ducers in third markets. Estimates of agriculturalproducer support declined as a share of grossfarm receipts between the mid-1980s and themid-1990s. Over the past few years, however,when agricultural prices have declined, subsidieshave actually increased.30 By 1999 the averageproducer support estimate reached 40 percent,almost equal to the average of 1986–88 (ta-ble 2.13). Virtually all OECD countries, exceptNew Zealand and Australia, have increased theirsupport levels. The biggest increases in the rate ofprotection took place in lower income OECDcountries but the largest subsidy in absoluteterms was given by the EU. During 1997–99, theaverage annual value of subsidies was about 60percent of total world trade in agriculture, andalmost twice the value of agricultural exportsfrom developing countries.31

The impact of trade restrictions on di-versification and trade growth. The lowering of industrial country trade barriers has com-bined with improved efficiency in developing

countries to spur rapid export growth. Themore successful middle-income countries have achieved rapid progress by increasing exportsof labor-intensive manufactures, in part replac-ing production in industrial countries. This de-celeration or decline in industrial countries’share in labor-intensive manufactures has al-lowed developing countries to increase theirexports at a much higher rate than the growthof world income.32 By allowing the least-costproducers to capture a greater share of de-mand, the efficiency of global production hasincreased. Despite remaining quotas and otherrestrictions on many labor-intensive products(such as textiles), the shares of developingcountries’ manufacturing exports in worldtrade (figure 2.6) and in the consumption of theQuad countries have increased. But, even in1995 the shares were not high—ranging from6.8 percent of consumption in Canada and theUnited States to only 3.4 percent in Japan.

Tariff escalation in industrial countrymarkets has restricted the market access of de-veloping country producers of finished goods,thus hampering industrialization.33 The Uru-guay Round has made some progress in re-ducing the degree of overall tariff escalation.Yet, in a number of sectors (such as food pro-cessing) that are of particular interest to de-veloping countries, high levels of tariff escala-tion are still present (box 2.6). Many productsare also protected by some form of quota. Assome of these quotas are allocated on the basisof historical trade shares, new and more effi-cient countries cannot enter these markets.



Unpublished Proofs

Developing country exporters of agricul-tural products have not achieved even the lim-ited penetration of industrial country marketsthat occurred in manufactures. The share of de-veloping countries’ agricultural exports in worldtrade have actually decreased (figure 2.7). Therelatively low level of developing country agri-cultural exports can be attributed in part togreater protection and subsidies in agriculturethan in manufacturing. Higher protection al-lows only the most efficient agricultural pro-ducers in developing countries to enter industri-alized country markets and relatively moreinefficient producers in industrial countries tomaintain their market share. The success ofmany developing countries in products that facelower protection and subsidies, such as cut flow-ers from Africa, and more stable trade shares infruits and vegetables also suggest that if protec-tion in agriculture is lowered, many of the poor-est countries could expand their exports.34

Implications for the poorest countries.While external constraints are not the primaryreason for slow export growth and decliningterms of trade of the poorest countries;35 nev-ertheless, industrial country trade restrictionsand subsidies are having an adverse impact on

growth and poverty reduction. Many of thepoorest countries are still primarily agricul-tural exporters, 40 to 60 percent of their pop-ulation (and the majority of the poor—WorldBank 2000c) lives in rural areas, and expan-sion of agricultural exports is one of their fewavenues to accelerating growth (at least in themedium term), given their level of technologyand human capital base. Many of the poorestcountries made progress during the 1990s inremoving domestic policy constraints on agri-cultural exports; however, policy reforms andinvestments in rural areas, which are neces-sary for poverty alleviation, are unlikely toyield significant improvements unless the de-mand for many of these products can be ex-panded through exports to world markets.

Numerical estimates of gains from re-duced agricultural protection vary consider-ably, but reducing protection could have aparticularly important impact on the poorestcountries. Ianchovichina, Mattoo, and Olar-reaga (2000) show that if the all the Quadcountries gave free trade access to the low-income African countries, their net exportswould increase about 6 percent. The negativeimpact of this expansion on other developing


21

Unpublished Proofs


Figure 2.6 Imports of manufactures from developing countries as a percentage ofapparent consumptionPercent

0

European Union United States and Canada Japan

1

2

3

4

5

6

7

8

1980 1985 1990 1995

Source: UNCTAD 1996.

Food manufacturing products are subject to high tariff levels and a high degree of escalation.

In the European Union and Japan, fully processedmanufacturing food products face tariffs almosttwice as high as tariffs on products in the first stage of processing. In Canada, tariffs on fullyprocessed food products are 13 times higher than for products in the first stage of processing. Highnominal tariffs on goods with high degrees of pro-cessing imply even higher rates of effective protec-tion (because the tariff rate is a relatively large per-centage of value added).

Partly because of these trade restrictions, thepenetration of developing-country food exports hasbeen limited, is much lower than average manufac-turing, and has been declining in some markets.

Tariff escalation in food manufacturing(percent)

Canada EU Japan

First stage 3 15 35Semiprocessed 8 18 36Fully processed 42 24 65

Source: WTO (several years), Trade Policy Review—QuadCountries.

and on industrial countries would be negligi-ble, because the absolute size of the increase isvery small in comparison to the world trade.36

Almost all the gains come from the expansionof agricultural trade. Hertel, Hoekman, andMartin (2000) report that a 40 percent reduc-tion in industrial countries’ agricultural tariffsand export subsidies by 2005 (a less radicalassumption than in the paper cited above)would increase income in most developing re-gions by less than 1 percent.

However, one point needs to be underlinedwhen discussing the trade barriers facing theexports of the poorest counties. Most of thesecountries have weak export-supporting infra-structure and skills that make it difficult toswitch from domestic market to exporting, andvice versa, in response to changes in relativeprices, or to diversify out of traditional com-



Unpublished Proofs

Figure 2.7 Share of developing countriesin world tradePercent

Source: FAOSTAT, World Bank data.

0

10

20

30

40

1970 1975 1980 1985 1990 1995

Agriculture

Manufactures

Box 2.6 Food processing

Imports of processed food from developingcountries as a percentage of apparentconsumptionPercent

0

European Union United Statesand Canada

Japan

0.5

1

1.5

2

2.5

3

3.5

4

1980 1985 1990 1995

Source: UNCTAD 1996.

modities facing artificially constrained mar-kets. These weaknesses are reflected in littleexport-market diversification, imperfect capitalmarkets that do not provide adequate andtimely credit, low capitalization of many enter-prises, small numbers of firms, and a very diffi-cult environment for transactions. The fact thatmany of these countries have a history of policyreversals and mismanagement implies that thecredibility that must underpin investment islacking. Thus they are in a much weaker posi-tion than some of the more successful countriesto overcome industrial countries’ barriers totheir traditional exports through diversification.

Notes1. For example, import-weighted calculations of

average tariffs may understate the importance of veryhigh rates that block all imports, while the number ofQRs may not accurately reflect their impact on trade.Moreover, data availability for these series is unevenboth over time and across countries. Nevertheless, theyare preferable indicators of trade policy than manyused in the literature (see Edwards 1997). Some ofthem (such as share of trade to GDP, or growth of ex-ports) are endogenous variables that depend on loca-tion, country size, and other policies; some rank traderegimes as perceived by users (the Heritage index),while others measure the degree of trade distortionsonly at one point in time (the Sachs-Warner index).Most of these measures have even more limited coun-try and time period coverage.

2. See Barro and Sala-i Martin 1995 for a review ofthe econometric literature. Sensitivity of results to bothvariables and specifications are highlighted by Levine andRenelt 1992 and Sala-i Martin 1997a, 1997b).

3. One concern is that in many developing coun-tries tariff bindings have been applied to only a smallpercentage of tariff lines, and wherever applied, theseceilings in general have been higher than the actual ap-plied tariffs. In a sample of 42 countries, the boundceilings were 49 percent, while the average applied ratewas only 18 percent. The large gap between tariff bind-ings and actual levels may make the government sus-ceptible to protectionist pressures and creates uncer-tainty about the future direction of reforms.

4. These averages need to be used with caution, asdata are available for only a few countries in each re-gion and may not exist for both the time periods underconsideration.

5. The black market premium is used here to mea-sure the administrative allocation of foreign exchangefor trade purposes. However, a large black market pre-

mium may also reflect the degree of overvaluation aswell as capital account restrictions.

6. This chapter uses simple averages, which giveequal weight to individual countries regardless of size.These figures differ from the weighted average growthrates reported in chapter 1 and the annex, which givegreater weight to bigger countries. China and India inAsia, and South Africa and Nigeria in Africa, dominatethe regional averages. Weighted averages are appropri-ate for measuring the total regional increase in GDP orexports, while simple averages are appropriate in ana-lyzing the impact of policy on growth (since the rela-tionship between policy and performance is just as rel-evant for a small country as for a large one).

7. Tariff reductions are clearly not the only deter-minant of export growth rates, and some regions withrelatively high tariff rates (such as South Asia) hadrapid growth in exports. Nevertheless, the reductionsin tariff rates during the 1990s were, on average, asso-ciated with rapid export growth.

8. To measure the role of world trade conditions asopposed to other factors that determine export perfor-mance, export growth (in nominal dollars) is decom-posed into three components using COMTRADE data:(a) growth caused by the expansion of world trade forproducts defined as traditional exports in the initial pe-riod, (b) gains in the share of individual countries inworld trade in these commodities, and (c) growth attrib-utable to diversification. COMTRADE data are notidentical to the trade data used in the rest of the report.They are based on the reported imports from OECD andother reporting countries and cover about 90 percent ofworld trade. Because most countries in Sub-SaharanAfrica and the Commonwealth of Independent States(CIS) do not report their imports fully, they underesti-mate intra-African and intra-CIS trade. The decomposi-tion into growth in traditional markets, increased mar-ket share, and diversification depend on the definition ofthe traditional or initial export basket. In table 2.8 theexport baskets of 1979–80 and 1989–90 were taken andall exports above US$5 million were classified as major.Growth of new exports and exports that were less than$5 million during the initial period are classified as di-versified. (See Yeats 1998, and Ng and Yeats 2000 for asimilar exercise.)

9. See Dollar 1992; Elbadawi 1998; Ghura andGrenness 1993; Razin and Collins 1997; and Sekkatand Varoudakis 2000. Edwards and Savastano (1999)have a critical review of the measures of exchange ratemisalignment. See also Hinkle and Montiel 1999 andWilliamson 1994 for a detailed analysis of the mea-surement issues and economic implications of ex-change rate misalignment.

10. According to IMF, 97 percent of its members in1970 were classified as having a pegged exchange rate;by 1999, this share had come down to very low levels


23

Unpublished Proofs


(Calvo and Reinhart 2000). As of 1999, of the 185countries reporting their exchange rate arrangements,84 had pegged regimes, 75 had floating rates, and 26managed a system of limited flexibility. Of the countrieswith floating rates, 27 maintained a managed float, and48 had an independent float (Shatz and Tarr 2000).

11. The fears of deindustrialization by manyAfrican countries are usually attributed to trade re-form, while the exchange rate behavior probablyplayed an even more important role.

12. Some studies have found a negative relation-ship between indicators of real exchange rate volatilityand private investment (Aizenman and Marion 1995,1996). Similar results were obtained by Darby and oth-ers 1998 for a group of five OECD countries and byServen 1998 for developing countries. Other studieshave found no relationship between real exchange ratevolatility and aggregate investment (Bleaney 1996;Goldberg 1993; Ramey and Ramey 1995).

13. They cite various evidence, including thelower volatility of nominal exchange rates in develop-ing countries than in several industrial countries,higher volatility of foreign exchange reserves thanwould be expected under floating rates, that countriesuse nominal interest rates to smooth exchange ratefluctuations, and that exchange rate rigidity occurs inthe face of commodity price shocks.

14. The World Bank prepares ratings of develop-ing countries’ macroeconomic policy stances using aconsistent framework, for the purpose of cross-countrycomparisons. Although exchange rate volatility macro-economic stability ratings are highly correlated, the re-lationship is not perfect. There are twenty-one coun-tries which either had high exchange rate volatility andwere rated as having stable macroeconomic policies orvice versa. Only 16 countries managed to achieve bothmacroeconomic stability and low real exchange volatil-ity. These countries also achieved much higher percapita GDP growth rates than other countries.

15. Thin markets lead to large changes in ex-change rates due to small changes in demand and sup-ply of foreign exchange.

16. A recent IMF study on exchange rate regimescomes to the same conclusions (Mussa and others 2000).

17. In many countries, initial policies did not in-clude broad trade liberalization, but what can be called“compensatory” policies for exports. These range fromelaborate input coefficients for duty exemption systemin Korea and duty drawback systems in Taiwan to ex-port processing zones, and so on, which gradually gaveway to elimination of NTBs, lower tariffs, and openingup of the domestic economy. Mauritius relied on an ef-fective free trade zone, while the success of Bangladeshgarments is due to a well-working bonded warehousescheme, which is a form of duty exemption system.

18. There has been significant tariff reductionsince 1996 in most of these countries. Also, there arelarge differences among the countries included (thetrade-weighted tariff ranges from 25 percent in Zim-babwe to about 7 percent in the South African Cus-toms Union).

19. Many of these products do not have local sub-stitutes, and it is unlikely that they will be produced inthese countries over the medium term. In that sense,they are pure revenue tariffs.

20. See Nash and Fouratan 1997 for the experi-ences in Africa, and Rajapatirana 1997 for Latin Amer-ica and the Caribbean.

21. Just lowering tariffs on intermediate and cap-ital goods without corresponding reductions in tariffson consumer goods will increase the effective rate ofprotection for domestic producers, and thus may im-pair the efficiency of resource allocation.

22. Most infrastructure services in developingcountries are supplied by governments or public de-partments with administrative barriers to entry. Thislack of competition leads to higher costs, poor mainte-nance, and a lack of investment for technological up-grading. Opening up many of these services to compe-tition under a transparent regulatory environment willlower costs, attract foreign and private capital, and im-prove delivery of services. For examples of recent pri-vatization and demonopolization efforts and their pos-itive impact on performance, see AfDB 1999.

23. In a sense these high costs act as an exporttax, increasing the costs of production over those ofsuppliers that do not pay these duties.

24. See various needs assessment papers issuedthrough the Integrated Framework for Trade-RelatedTechnical Assistance to Least Developed Countriesprocess. They include references to the state of standardsand conformity assessment systems and infrastructureneeds in these countries (www.ldcs.org/index.htm).

25. Under tariff quotas, a fixed quantity of prod-ucts from specified countries can be imported at alower tariff and anything above that level is subject tothe normal tariff.

26. The average most favored nation (MFN) tar-iff for these 311 lines is 21 percent; for the least devel-oped countries it is 18 percent.

27. It is not clear why so few of these products areimported into Quad countries. But the significant dif-ference between the United States’ (86 percent) and theEuropean Union’s (9 percent) share of developing coun-tries’ exports suggest that intra-EU trade makes up thedifference.

28. Note that GSP and North-South free tradearrangements imply some trade diversion that will ben-efit some developing countries, but hurt others. Forth-



Unpublished Proofs

coming work by Hoekman, Ng, and Olareaga (2000)argues that these preferences are not very extensive.

29. The NAFTA preference is 50 percent for thesetariffs.

30. The producer support estimate (percent PSE)is calculated as the annual monetary value of produc-tion-related support to agricultural producers as ashare of gross farms receipts (OECD 2000a).

31. The recent proposal by the EU to grant duty-free access to all exports from least developed countriesand the move to income support rather than direct sup-port to production should improve market access for de-veloping countries.

32. UNCTAD (TDR 1999) has shown that inlow- and medium-skill manufactures, small changes inimport penetration ratios significantly increase the ex-port growth rates for these products from developingcountries.

33. Tariff escalation means that the tariff raterises with the level of processing.

34. For example, developing countries’ share infruits and vegetables trade, which is less protected, hasnot declined over this period.

35. Previous sections described the impact of con-flicts and trade-related policies on developing coun-tries’ performance. In addition, expansion of agricul-tural exports (such as tropical commodities) that aresubject to few trade restrictions in industrial countriestends to reduce prices and hence the terms of trade.

36. The effect on agricultural exports is muchgreater. The model reduces the exports of other prod-ucts that faced lower tariffs because of factors movingto the expanding agricultural sector.

ReferencesAfDB (African Development Bank). 1999. African De-

velopment Report 1999: Infrastructure Develop-ment in Africa. New York: Oxford University Press.

Aizenman, Joshua, and Nancy Marion. 1995. “Volatil-ity, Investment and Disappointment Aversion.”Working Paper 5386. National Bureau of Eco-nomic Research, Cambridge, Mass.

———. 1996. “Volatility and the Investment Re-sponse.” Working Paper 5841. National Bureauof Economic Research, Cambridge, Mass.

Amjadi, Azita, Ulrich Reincke, and Alexander Yeats.1996. “Did External Barriers Cause the Margin-alization of Sub-Saharan Africa in World Trade?”World Bank Discussion Paper 348. World Bank,Washington, D.C. March.

Barro, Robert J., and Xavier Sala-i Martin. 1995. Eco-nomic Growth. Cambridge, Mass.: MIT Press,McGraw Hill.

Ben David. 1993. “Equalizing Exchange: Trade Liber-alization and Income Convergence.” QuarterlyJournal of Economics. 108(3).

Bhagwati, Jagdish. 1978. Foreign Trade Regimes andEconomic Development: Anatomy and Conse-quences of Exchange Control Regimes. Cam-bridge, Mass.: Ballinger.

Bigsten, Arne, and others. 1998. “Exporting and Effi-ciency in African Manufacturing.” (Paper pre-sented for the Conference on Policies for Compet-itiveness in Manufacturing in Sub-Saharan Africa,Johannesburg, South Africa. 6–7 November.)

Bleaney, Michael. 1996. “Macroeconomic Stability, In-vestment and Growth in Developing Countries.”Journal of Development Economics 48: 461–77.

Calvo, Guillermo A., and Carmen Reinhart. 2000.“Fear of Floating.” University of Maryland, Col-lege Park, Md., and National Bureau of EconomicResearch, Cambridge, Mass. Draft. May.

Clerides, Sofronis, Saul Lach, and James Tybout.Forthcoming. “Is Learning by Exporting Impor-tant? Micro-Dynamic Evidence from Colombia,Mexico, and Morocco.” Quarterly Journal ofEconomics.

Collier, Paul, and Anke Hoeffler. 1999. “Greed andGrievance in Civil War.” Working Paper 2355.World Bank, Washington, D.C.

Darby, Julia, and other. 1998. “The Impact of Ex-change Rate Uncertainty on the Level of invest-ment.” Discussion Paper 1896. Centre for Eco-nomic Policy Research, London.

Dollar, David. 1992. “Outward-Oriented DevelopingEconomies Really Do Grow More Rapidly: Evi-dence from 95 LDCs, 1976–85.” Economic Devel-opment and Cultural Change 40: 523–44. April.

Dornbusch, Rudiger, and Sebastian Edwards, eds. 1995.Reform, Recovery and Growth. Latin America andthe Middle East. Cambridge, Mass.: National Bu-reau of Economic Research.

Easterly, William. 2000. “The Lost Decades . . . andthe Coming Boom? Policies, Shocks, and Devel-oping Countries” Stagnation 1980–1998.” WorldBank, Washington, D.C. May.

Easterly, William, Roumeen Islam, and Joseph E.Stiglitz. 2000. “Explaining Growth Volatility.”World Bank, Washington, D.C. January. Processed.

Edwards, Sebastian. 1997. “Openness, Productivityand Growth: What Do We Really Know?” Work-ing Paper 5978. National Bureau of EconomicResearch, Cambridge, Mass. March.

Edwards, Sebastian, and Miguel A. Savastano. 1999.“Exchange Rates in Emerging Economies: WhatDo We Know? What Do We Need to Know.”Working Paper 7228. National Bureau of Eco-nomic Research, Cambridge, Mass. July.


25

Unpublished Proofs


Elbadawi, Ibrahim. 1998. “Real Exchange Rate Policyand Non-Traditional Exports in Developing Coun-tries.” Research for Action 46. UNU World Insti-tute for Development Economics Research (UNU/WIDER), The United Nations University, Helsinki.

Falvey, Rod, and Cha Dong Kim. 2000.”Timing andSequencing Issues for Trade Liberalization inAfrica.” In Trade and Fiscal Adjustment in Africa,edited by David Bevan, Paul Collier, NormanGemmell, and David Greenaway. Great Britain:Macmillan Press Ltd.

Finger, J. Michael, and Sam Laird. 1987. “Protection inDeveloped and Developing Countries: An Over-view.” Journal of World Trade Law 21 (6).

Finger, J. Michael, and Ludger Schuknecht. 1999.“Market Access Advances and Retreats: theUruguay Round and Beyond.” Policy ResearchWorking Paper. World Bank, Washington, D.C.

Frankel, Jeffrey A., and David Romer. 1999. “DoesTrade Cause Growth?” The American EconomicReview 89 (3): 379–99. June.

Ghura, Dhaneshwar, and Thomas J. Grenness. 1993.“The Real Exchange Rate and MacroeconomicPerformance in Sub-Saharan Africa.” Journal ofDevelopment Economics 42: 155-1–74.

Goldberg, Linda S. 1993. “Exchange Rates and Invest-ment in United States Industry.” Review of Eco-nomics and Statistics 75: 575-5–88.

Harberger, Arnold C. 1988. Reflections on UniformTaxation. Processed.

Hausman and others. 1999. “Financial Turmoil andthe Choice of Exchange Rate Regime.” WorkingPaper, Inter-American Development Bank, Wash-ington, D.C.

Hertel, Thomas, Bernard Hoekman, and Will Martin.2000. “Towards a new round of WTO negotia-tions: issues and implementation for developingcountries.” (Paper presented at the ABCDE meet-ing. World Bank, Washington D.C. April.)

Hertel, Thomas, and Will Martin. 1999. “Would De-veloping Countries Gain From Inclusion of Man-ufactures Trade in the WTO 2000 Negotiations?”(World Bank Paper presented at the Conferenceon the Millennium Round, WTO, Geneva. Sep-tember 20–21.)

———. 2000. “Liberalising Agriculture and Manufac-turers.” The World Economy 23 (4): 455–69. April.

Hinkle, Lawrence E., and Peter J. Montiel. 1999. Ex-change rate misalignment: Concepts and Mea-surement for Developing Countries. Washington,D.C.: World Bank.

Hoekman, Bernard, Francis Ng, and Marcelo Olar-reaga. 2000. “Problems Created by Protection inthe Markets for Developing Country Exports.”World Bank, Washington, D.C. July.

Ianchovichina, Elena, Aaditya Mattoo, and MarceloOlarreaga. 2000. “Duty-Free Access for LDCs’

Exports: How Much Is It Worth and Who Pays?World Bank, Washington, D.C. June.

IMF (International Monetary Fund). 1981. ExchangeArrangements and Exchange Restrictions. Wash-ington, D.C.

———. 1992. Exchange Arrangements and ExchangeRestrictions. Washington, D.C.

———. 1996. Exchange Arrangements and ExchangeRestrictions. Washington, D.C.

———. 1998. Trade Liberalization in IMF-SupportedPrograms. February 1998. IMF, Washington,D.C.

———. International Financial Statistics. Washington,D.C.

Kraay, Aart C. 1997. “Exports and Economic Perfor-mance: Evidence from a Panel of Chinese Enter-prises.” Development Research Group, WorldBank, Washington, D.C. Processed.

Krueger, Anne O. 1978. Liberalization Attempts andConsequences. A Special Conference Series onForeign Trade Regimes and Economic Develop-ment. Volume X. Cambridge, Mass.: National Bu-reau of Economic Research; New York. BallingerPub. Co.

Levine, Ross, and David Renelt. 1992. “A SensitivityAnalysis of Cross-Country Growth Regressions.”American Economic Review 82: 942–63.

Limão, Nuno, and Anthony Venables. 1999. “Infra-structure, geographical disadvantage and trans-port costs.” Policy Research Working Paper 2257.Development Research Group—Trade, World Bank,Washington, D.C. December.

Little, Ian, Malcolm David, Tibor Scitovsky, and Mau-rice Fitzgerald Scott. 1970. Industry and Trade inSome Developing Countries. Oxford, U.K.: Ox-ford University Press.

Michalopoulos, Constantine. 1999. “Trade Policy andMarket Access Issues for Developing Countries:Implications for the New Millennium Round.”Policy Research Working Paper 2214. World Bank,Washington, D.C.

Mussa, Michael, and other. 2000. “Exchange rateregimes in an increasingly integrated World Econ-omy.” Occasional Paper 193. International Mon-etary Fund, Washington, D.C. Also available on(http://www.imf.org/EXTERNAL/PUBS/CAT/longres.cfm?sk&sk=3518.0).

Nash, John, and Faezeh Foroutan, eds. 1997. TradePolicy and Exchange Rate Reform in Sub-Saha-ran Africa. National Centre for DevelopmentStudies, The Australian National University.

Ng, Francis. 1997. “A Profile of Tariffs, Para-tariffs,Non-Tariff Measures, and Economic Growth inDeveloping Countries.” World Bank, Washing-ton, D.C.

Ng, Francis, and Alexander Yeats. Forthcoming. “Onthe Recent Trade Performance of Sub-Saharan



Unpublished Proofs

African Countries: Cause for Hope or More ofthe Same?” Africa Region Technical Paper, WorldBank, Washington, D.C.

Organisation for Economic Co-Operation and Devel-opment (OECD). 1996. “Indicators of Tariff andNon-Tariff Trade Barriers.”

———. 2000a. “Agricultural Policies in OECD Coun-tries: Monitoring and evaluation 2000.” Agricul-ture and Food, Paris.

———. 2000b. “Tariffs and Trade: OECD Query andReporting System.” CD Rom. Paris.

Papageorgiou, Demetrios, Armeane M. Choksi, andMichael Michaely, eds. 1991. Liberalizing ForeignTrade in Developing Countries—The Lessons ofExperience. Washington, D.C.: World Bank; Ox-ford: Blackwell.

Rajapatirana, Sarath. 1997. Trade Policies in LatinAmerica and the Caribbean: Priorities, Progressand Prospects. San Francisco, Calif.: Interna-tional Center for Economic Growth.

Ramey, George J., and Valerie A. Ramey. 1995.“Cross-Country Evidence on the Link betweenVolatility and Growth.” American Economic Re-view 1138–51.

Razin, Offair, and Susan M. Collins. 1997. “Real Ex-change Rate Misalignments and Growth.” Work-ing Paper 6174, National Bureau of EconomicResearch, Cambridge, Mass. September.

Rodriguez, F., and Dani Rodrik. 1999. “Trade Policyand Economic Growth: A Skeptic’s Guide to theCross-national Evidence.” May, Discussion Paper.Centre for Economic Policy Research, London.

Rodrik, Dani. 1996. “Understanding Economic PolicyReform.” Journal of Economic Literature 24 (1).March.

———. 1998. “Where did all the Growth Go? ExternalShocks, Social Conflict and Growth Collapses.”Working Paper 6350, National Bureau of Eco-nomic Research, Cambridge, Mass.

Sachs, Jeffrey D., and A.Warner. 1995. “Economic Re-forms and the Process of Global Integration.”Brookings Papers on Economic Activity 1–118.

Sala-i-Martin, Xavier. 1997a. “I Just Ran Four MillionRegressions.” Working Paper 6252. National Bu-reau of Economic Research, Cambridge, Mass.

———. 1997b. “I Just Ran Two Million Regressions.”American Economic Review, Papers and Proceed-ings 87 (2): 178–83.

Sekkat, Khalid, and Aristomene Varoudakis. 2000.“Exchange Rate Management and ManufacturedExports in Sub-Saharan Africa.” Journal of De-velopment Economics 61 (1). February.

Serven, Luis. 1998. “Macroeconomic Uncertainty andPrivate Investment in LDCs: An Empirical Investi-gation.” Washington, D.C. World Bank. Processed.September.

Shatz, Howard J., and David G. Tarr. 2000. “ExchangeRate Overvaluation and Trade Protection: Lessonsfrom Experience.” Policy Research Working Paper2289, World Bank, Washington, D.C. February.

———. 1996. “Handbook of International Trade andDevelopment Statistics.” Geneva.

UNCTAD, 1999. Trade and Development Report,Geneva.

Wang, Zhen Kun, and L. Alan Winters. 2000. “Putting‘Humpty’ Together Again: Including DevelopingCountries in a Consensus for the WTO.” PolicyPaper 4. Centre for Economic Policy Research,London.

Williamson, John (ed). 1994. Estimating EquilibriumExchange Rates. Washington, D.C: Institute forInternational Economics.

World Bank. 1989. Sub-Saharan Africa: From Crisis toSustainable Growth. A Long-Term PerspectiveStudy. Washington, D.C.: World Bank.

———. 1994. Adjustment in Africa: Reforms, Results,and Road Ahead. World Bank Policy ResearchReport. New York: Oxford University Press.

———. 1995. Global Economic Prospects and the De-veloping Countries. Washington, D.C.: World Bank.

———. 1996. “Trade Policy Reform in DevelopingCountries since 1985.” Discussion Paper 267.Washington, D.C.

———. 1997a. Global Economic Prospects and theDeveloping Countries. Washington, D.C.: WorldBank.

———. 1997b. World Development Indicators. Wash-ington, D.C.

———. 1998. World Development Indicators. Wash-ington, D.C.

———. 2000a. Can Africa Claim the 21st Century?Washington, D.C.: World Bank.

———. 2000b. Trade Blocs: A World Bank Policy Re-search Report. New York: Oxford UniversityPress.

———. 2000c. World Development Report: Povertyand Development. Washington, D.C.

———. 2000d. World Development Indicators. Wash-ington, D.C.: World Bank.

WTO (World Trade Organization). Trade Policy Re-views—Quad Countries. Geneva.

———. 1998. Trade Policy Reviews for 31 DevelopingCountries 1995–98. Geneva.

Yeats, Alexander. 1998. “Have Policy Reforms Influ-enced the Trade Performance of Sub-SaharanAfrican Countries.” Trade Research Team. WorldBank. Washington, D.C.

Yeats, Alexander J., with Azita Amjadi, Ulrich Reincke,and Francis Ng. 1997. Did Domestic PoliciesMarginalize Africa in International Trade? Direc-tions in Development, World Bank, Washington,D.C.


27

Unpublished Proofs


Unpublished Proofs



Annex 2.1 Sample countries in various charts and tables

Within core sample

ECA Countries CountriesAll countries Least- Low-income with high with low

developing Countries (excluding developed small REER REERcountries in conflict Yugoslavia) Core sample countries countries volatility in volatility in

No. Region [105] [19] [6] [59] [16] [22] 1990s [26] 1990s [33]

1 Africa Angola Angola2 Africa Burundi Burundi3 Africa Benin Benin Benin Benin Benin4 Africa Burkina Faso Burkina Faso Burkina Faso Burkina Faso Burkina Faso5 Africa Botswana Botswana Botswana6 Africa Central Central Central Central Central

African African African African AfricanRepublic Republic Republic Republic Republic

7 Africa Côte d’Ivoire Côte d’Ivoire Côte d’Ivoire8 Africa Cameroon Cameroon Cameroon9 Africa Congo Congo

10 Africa Comoros11 Africa Cape Verde12 Africa Djibouti13 Africa Ethiopia Ethiopia14 Africa Gabon Gabon Gabon15 Africa Ghana Ghana Ghana Ghana16 Africa Guinea17 Africa Gambia Gambia Gambia Gambia Gambia18 Africa Guinea-Bissau19 Africa Equatorial

Guinea20 Africa Kenya Kenya Kenya Kenya21 Africa Liberia Liberia22 Africa Lesotho23 Africa Madagascar Madagascar Madagascar Madagascar Madagascar24 Africa Mali Mali Mali Mali Mali25 Africa Mauritania Mauritania Mauritania Mauritania Mauritania26 Africa Mauritius Mauritius Mauritius27 Africa Malawi Malawi Malawi Malawi Malawi28 Africa Namibia29 Africa Niger Niger Niger Niger Niger30 Africa Nigeria Nigeria Nigeria Nigeria31 Africa Rwanda Rwanda32 Africa Sudan33 Africa Senegal Senegal Senegal34 Africa Sierra Leone Sierra Leone35 Africa São Tomé &

Principe36 Africa Swaziland37 Africa Seychelles

Islands38 Africa Chad Chad Chad Chad Chad39 Africa Togo Togo Togo Togo Togo40 Africa Tanzania Tanzania Tanzania Tanzania Tanzania41 Africa Uganda Uganda42 Africa South Africa South Africa South Africa43 Africa Democratic Democratic

Republic of Republic ofthe Congo the Congo

44 Africa Zambia Zambia Zambia Zambia Zambia45 Africa Zimbabwe Zimbabwe Zimbabwe46 East Asia China China China47 East Asia Fiji48 East Asia Indonesia Indonesia Indonesia Indonesia

Unpublished Proofs


29Embargoed until Tuesday, December 5, 2 p.m. EST

Annex 2.1 Sample countries in various charts and tables (continued)

Within core sample



No. Region [105] [19] [6] [59] [16] [22] 1990s [26] 1990s [33]

49 East Asia Republic of Republic of Republic ofKorea Korea Korea

50 East Asia Malaysia Malaysia Malaysia51 East Asia Philippines Philippines Philippines52 East Asia Papua

New Guinea53 East Asia Thailand Thailand Thailand54 East Asia Myanmar Myanmar Myanmar Myanmar Myanmar55 ECA Former Former

Soviet Union Soviet Union56 ECA Bulgaria Bulgaria57 ECA Former Former

Czechoslovakia Czechoslovakia58 ECA Hungary Hungary59 ECA Poland Poland60 ECA Romania Romania61 ECA Yugoslavia, Yugoslavia

federal federalRepublic Republicof (Serbia/ of (Serbia/Montenegro) Montenegro)

62 LAC Argentina Argentina Argentina63 LAC Bolivia Bolivia Bolivia64 LAC Brazil Brazil Brazil65 LAC Barbados66 LAC Chile Chile Chile67 LAC Colombia Colombia Colombia68 LAC Costa Rica Costa Rica Costa Rica69 LAC Dominican Dominican Dominican

Republic Republic Republic70 LAC Ecuador Ecuador Ecuador71 LAC Guatemala Guatemala72 LAC Guyana73 LAC Honduras Honduras Honduras74 LAC Haiti Haiti75 LAC Jamaica Jamaica Jamaica76 LAC Mexico Mexico Mexico77 LAC Nicaragua Nicaragua78 LAC Panama79 LAC Peru Peru80 LAC Paraguay Paraguay Paraguay81 LAC El Salvador El Salvador82 LAC Suriname83 LAC Trinidad and

Tobago84 LAC Uruguay Uruguay Uruguay85 LAC Venezuela, Venezuela, Venezuela,

Rep. Bol. de Rep. Bol. de Rep. Bol. de86 MNA Turkey Turkey Turkey87 MNA United Arab

Emirates88 MNA Kuwait Kuwait89 MNA Morocco Morocco Morocco90 MNA Bahrain91 MNA Algeria Algeria Algeria

Unpublished Proofs



92 MNA Egypt, Egypt, Egypt,Arab Arab ArabRep. of Rep. of Rep. of

93 MNA Iran, Islamic Iran, Islamic Iran, IslamicRep. of Rep. of Rep. of

94 MNA Iraq Iraq95 MNA Jordan Jordan Jordan96 MNA Oman Oman Oman97 MNA Saudi Arabia Saudi Arabia Saudi

Arabia98 MNA Syrian Arab Syrian Arab

Republic Republic99 MNA Tunisia Tunisia Tunisia100 MNA Yemen, Yemen,

Rep. of Rep. of101 South Asia Bangladesh Bangladesh Bangladesh Bangladesh Bangladesh102 South Asia India India India103 South Asia Sri Lanka Sri Lanka Sri Lanka Sri Lanka104 South Asia Nepal Nepal Nepal Nepal Nepal105 South Asia Pakistan Pakistan Pakistan Pakistan

Annex 2.1 Sample countries in various charts and tables (continued)

Within core sample



No. Region [105] [19] [6] [59] [16] [22] 1990s [26] 1990s [33]

trade policies in the 1990s and the poorest countries

Documents