5. Liberalization of Trade: Why So Much Controversy?

5.
Liberalization of Trade: Why So Much Controversy?
H. Tang and Ann Harrison
Economists have long recognized the gains from international trade; the study of these gains
is where modern economics began. Over the centuries, gains from international trade have
been a powerful integrating force that brought together remote parts of the world and
different civilizations, helped disseminate knowledge and ideas, and shaped the course of
regions or nations. With rapid reductions in transport and communication costs, this trend
accelerated in the 19th century. At the beginning of the 20th century it reached unprecedented
high levels, from which it declined, however, following the two World Wars, the 1929 Crisis,
and a world-wide increase in protectionism.
A reversal in protectionism started after World War II among industrialized countries, and in
the 1970s among developing countries. In the 1990s, trade reforms expanded or were
consolidated in South Asia, East Asia, Latin America, Eastern Europe and, to a lesser extent,
in Africa and the Middle East. Despite positive results and high expectations, controversies
about the economic and social effects of trade liberalization have persisted. Why have some
trade liberalizations ended up in reversals, and why have others brought about prosperity,
opportunities, and economic diversification? Why has global integration been so vocally
opposed by some? Is there still a role for protection of infant industries in growth strategies?
Does trade liberalization cause growth, or is it growth that increases the openness of an
economy? Does trade liberalization increase vulnerability and poverty? Or does it help
reduce poverty instead? Has import substitution helped accelerate growth? Or has it reduced
it? These issues will no doubt take time to be settled. However, from the experience and
academic research of the 1990s this chapter finds sufficient evidence to draw five lessons:
First, openness to trade has been a central element of successful growth strategies—all
countries that sustained growth over time have also reduced trade barriers. This empirical
regularity is confirmed both by the long term historical record and by the experience of the
1990s. Similar to country experiences with macroeconomic stability (Chapter 4), the
experience of the 1980s put in sharp contrast the performance of countries that responded to
shocks by increasing outward orientation (East Asian countries) and those that did not (Latin
America in the 1970s, Africa, and most countries of the Middle East and North Africa).
Influenced by this diversity in results, during the 1980s and 1990s, most developing countries
significantly reduced tariffs on imports and dismantled other forms of impediments to trade:
quantity restrictions, trade monopolies, and marketing boards. As in the case of
macroeconomic reforms, however, the results varied and, in general, fell short of
expectations—whereas a reduction in barriers to trade helped efficiency and growth in many
cases (East and South Asian countries, Botswana, Chile, Mauritius, Tunisia ), it failed to do
so in many others—simply because reducing barriers to trade are only one of the measures
needed for countries to expand their participation in international trade.
One explanation for the diversity in outcomes, and this is the second lesson to draw from the
1990s, is that trade is an opportunity, not a guarantee. Trade reform in some countries
brought about few gains in terms of export expansion or increased economic growth, while it
147
created social and economic adjustment costs. These country experiences do not imply that
less trade reform would have been desirable, but that trade reform must be done sensibly, as
part of an effective growth strategy. Liberalization of trade in Argentina in the 1980s and
1990s, and in Chile in the early 1980s, for example, was accompanied by an appreciation of
the real exchange rate that reduced the competitiveness of domestic industries, and incentives
to exports—with adverse consequences for the balance of payments and real economy. In
many countries of Eastern Europe in the 1990s, trade was liberalized while property rights
were not well defined, and the institutional base for a market economy was not well
developed. These, and other institutional issues preventing the free movement of resources,
often meant that trade reforms did not expand economic opportunities—restricting them
instead (Freund 2004).
In short, while trade integration can strengthen an effective growth strategy, it cannot ensure
its effectiveness. Other elements are needed, such as sound macroeconomic management,
building trade-related infrastructure, and trade-related institutions, economy-wide
investments in human capital and infrastructure, or building strong institutions. Because
these reforms are often difficult to implement, there has been excessive emphasis on trade
policy alone, rather than as a component of an overall growth strategy. In addition to freeing
markets and ensuring the institutional foundation of a market economy, there may also be a
need for governments to address market failures that impede a supply response. Identifying
which industries warrant special treatment is a highly risk- laden challenge, however, and the
experience of the last few decades is riddled with examples of attempts at correcting market
failures which became more costly than the failures themselves. At the same time, there has
been learning in how to structure interventions in a manner that can reduce the risks of
capture and failure.
In part because successful trade reforms are introduced in conjunction with complementary
policies and initiatives, it is difficult empirically to identify the growth effect of trade policy
alone, compared with the growth impact of other policy initiatives, and to disentangle
whether trade causes growth or growth causes trade. As an economy accumulates physical
and human capital, shifts its comparative advantage towards more capital- intensive activities,
and becomes internationally competitive on a wider range of goods and services, it will
inevitably trade more. But is higher trade the result or the cause of its growth? Most likely
both processes are at work.
Third, there are many possible paths to open an economy. The challenge for policymakers is
to identify which is the path best suited to their political economy, institutional constraints,
and initial conditions. As these vary from country to country, it is not surprising that
countries that have succeeded in promoting growth through trade liberalization have
followed a variety of policy approaches. The heterogeneity in country experiences regarding
the timing and pace of reforms is striking. Different countries have opened up different
sectors at different speeds (for example Bangladesh and India); others have achieved partial
liberalization through the establishment of export processing zones (for example China and
Mauritius); and yet others have combined unilateral trade reforms (Estonia) with
participation in regional trade agreements (for example Mexico and countries in Central and
Eastern Europe that have now joined the European Union).
148
Fourth, the distributive effects of trade liberalization are diverse—and not always pro-poor.
Trade reforms were expected to be pro-poor because in most societies the relatively wealthy
and urban classes have been more successful at using protection for their own benefit. The
expectation was that trade reform would increase the incomes of the unskilled. Yet evidence
from the 1990s on the relationship between trade reforms and poverty is to date mostly
indirect. Even in instances where trade policy has reduced poverty, there are still distributive
issues, and one important policy lesson is that countries need to help those affected move out
of contracting (import-competing) sectors into expanding (exporting) sectors. This is an issue
relevant to both developing and industrialized countries.
Fifth, the preservation and expansion of the world trade system hinges on its ability to strike
a better balance between the interests of industrialized and developing natio ns. It has long
been recognized that the world trade system, while more supportive of development than at
the beginning of the 1990s, still remains biased against the poor. The 1991 WDR had already
documented the differential treatment of industrialized countries’ imports. Notwithstanding a
decade of significant expansion of international trade, global markets are most hostile to the
products the world's poor produce—agriculture, textiles, and labor-intensive manufactures—
and problems of escalating tariffs, tariff peaks, and quota arrangements systematically deny
the poor market access and skew incentives against adding value in poor countries. These
problems are embedded in the remaining structure of protection in both industrial countries
and developing countries (the latter both due to their own anti-export biases and to higher
barriers to trade in developing country markets), and they can be addressed through
collective actions. Those actions should best be done through the Doha Round and the WTO.
Although there is a role for non-reciprocal preferences and for reciprocal regional
approaches, this comes at a cost to excluded countries, is arbitrary and political, and thus not
first best in terms of generating the right incentives for investment.
Evidence for the first of the lessons above will be presented in Country Note 2 and is not
repeated here. After a brief historical overview, this chapter discusses the evidence for the
four lessons above.
1.
From import substitution to export orientation to trade liberalization
Protection of domestic industries has a long history. In the 12th century, for example, to
maintain the competitive edge of their textile industries, Flanders and England restricted the
movement of experienced weavers. In the 13th century, England enacted laws restricting
types and origin of fabrics certain individuals could wear. Some of these laws aimed at
identifying individuals’ social origins through their clothes, but the rationale for others was
the protection of domestic industries. In the 17th century, a new law prohibited wearing
imported silk and calicoes, which provided impetus to England’s calico-printing, silk, and
cotton- linen industries, and hence restricted then- flourishing imports from France, India,
China, and Persia. In 16th and 17th century France, the state promoted selected industries,
through import protection, direct ownership, or subsidies, as did Japan later during the Meiji
period. While protection of domestic industries took various forms—such as subsidized
capital, monopoly or monopsony rights—protection from imports was the most widely used
149
and became particularly important after the start of the industrial revolution. During the
1800s and first half of the 1900s, tariffs on imports in industrial countries were as high as 3050 percent (Table 5.1).
Table 5.1: Tariff rates in industrial countries, 1820-1987
Kind of goods
and country
or region
Manufactures
Austria
Belgium
Denmark
France
Germany
Italy
Netherlands
Spain
Sweden
Switzerland
United
Kingdom
United States
Average
All goods
Australia
Canada
Japan
United States
Average
1820
1875
1913
1925
1930
1950
1987
18
9
14
20
13
18
4
41
20
9
16
15
10
21
20
22
6
41
16
14
5
24
14
18
11
3
18
26
25
11
9
7
9
10
50
15-20
9-10
15-20
12-15
4-6
8-10
3-5
15-20
3-5
4-6
0
23
5
3
7
40
22
40-50
11-14
25
17
37
19
48
32
14
16
7
7
45
14
4
41
16
17
20
40
18
14
13
38
14
13
19
45
17
9
4
13
6
8
6
6
23
21
23
11
7
7
30
10
7
30
21
46
63
21
19
7
7
7
7
Source: World Development Report 1991.
The infant industry argument was one among several used in the 1950s and 1960s to justify
growth strategies that were based on industrialization-cum- import substitution. Three others
were at least as influential. The first was the view that the income elasticity for goods
produced by developing countries was low, implying that as incomes in industrialized
countries rose, expansion of developing countries’ exports of agricultural products and raw
materials faced declining relative prices. Second, many policymakers in the 1950s and 1960s
were skeptical of free markets. Influenced by the depression of the 1930s, the positive role of
governments in the recovery of the US and reconstruction of Europe, and the apparent
successful development of the former USSR, these individuals believed that government
forces were needed to help allocate resources for long-term economic development. Import
protection was one of the tools towards that goal. Third, as discussed in Chapter 2, in the
1950s and 1960s capital accumulation, which could be encouraged through import
protection, was seen as central for growth.
Many developing countries pursued import substitution industrialization strategies in the
three decades that followed World War II. Yet three developments raised doubts on the longrun effectiveness of growth strategies based on import protection. The first was that, starting
in the 1960s, Korea and Taiwan, China, adopted export-oriented growth strategies that not
only yielded superior results in terms of economic performance, but also helped these two
economies much better withstand the severe interest rate and oil price shocks of the 1970s.
150
The contrasting experiences of Korea and Brazil provide a clear example. Korea responded
to the 1970s rising real interest rates and oil prices by further opening its economy and
encouraging exports through a more competitive real exchange rate and improved access to
imported inputs for exporters. Brazil responded by further closing its economy with a view to
embarking on the “second phase of import substitution” focused on substitution of capital
goods, expansion of public investment, and further rises in tariffs. The result, inter alia, was
appreciation of the exchange rate. All these were policies that proved detrimental to exports,
proved unsustainable in the long run, and had to be reversed in the following two decades.
By contrast, Korea’s policies established the basis for three decades of rapid and sustained
growth, and the country has joined the ranks of industrialized nations. In retrospect, it is clear
that import substitution strategies overlooked the fact that the size of domestic markets set
limits to the expansion of domestic industries, that manufacturing exports could be part of a
viable growth strategy, and that substitution of imports could only be a phase, possibly short,
of an industrialization process.
Second, support from special interests led to levels of protection in developing countries
during the 1970s and 1980s that were unprecedented in industrial countries, even after
accounting for the natural protection afforded by higher transport costs in the 19th century
and the beginning of the 1900s. In India, for example, maximum tariffs were as high as 350
percent in 1991 and similarly high tariff levels were common in other developing countries.
In many cases, tariffs were not the main constraint on imports. Until the early 1990s, for
example in India, Turkey, and Brazil, laws prevented imports of goods that could be
produced domestically, regardless of the cost of domestic production, implying a de facto
infinite level of protection. This protection was reinforced by numerous non-tariff barriers to
imports, and administrative allocation of foreign exchange. High tariffs, administrative
restrictions, rationing of foreign exchange and of import licenses created high returns to rent
seeking, a complex web of vested interests, and an environment that stimulated corruption
and weakened national institutions. Vested interests benefiting from such levels, and type, of
protection, proved a formidable force against reduction of protection, and its maintenance
well beyond the time it could be beneficial. The vulnerability of the state to capture by vested
interests, and to the misuse of its discretion, discredited import substitution strategies even
among those economists who believed in the strategic importance of protection in the initial
phases of industrialization.
Third, implementation of growth strategies based on import substitution proved to be much
more difficult in practice than in theory, and the practical and political aspects of
implementation often negated most of the expected gains (Balassa 1971; Little, Scitovsky,
and Scott 1970). Essentially, the level of distortions became inconsistent with the goals of
these strategies: central planning allocation of external resources, and high nominal tariffs
often provided negative protection to emerging activities, protection to activities with
negative value added, and contributed to misallocation and underutilization of capital in
capital scarce economies. Overvaluation of the exchange rate resulting from import
restrictions discouraged exports and penalized agriculture—further reducing the size of the
market for import-competing industries.
151
As a result, during the 1980s and 1990s virtually all developing countries followed the
examples set by Singapore, Hong Kong, Korea, and Taiwan: encouraging exports and
reducing levels of protection. Industrialization based on import protection was gradually
discredited and, starting in the mid-1980s, most developing countries sought to reduce levels
of import protection and liberalize trade. Chile and Sri Lanka were among the first
liberalizers, starting already in the 1970s. Argentina and Uruguay followed shortly thereafter.
Trade liberalization resumed and expanded in the 1990s, leading to increased integration of
developing economies in world trade.
At the beginning of the 1990s, many developing countries had already embarked on or were
starting ambitious economic reforms. While some of these reforms were unilateral, others
were accomplished in the context of multilateral trade agreements such as the Uruguay
Round. Important components of those reforms included large tariff reductions and
elimination of quotas, as well as the relaxation of restrictions on foreign investment. As
discussed in Chapter 3, this was at the origin of a strong expansion in international trade.
Export revenues for developing countries doubled between 1990 and 2000, rising from 12.5
to almost 25 percent of GDP. Using export shares in GDP as a measure of globalization,
developing countries are now more integrated with the world economy than high income
countries (Figure 5.1). Merchandise export growth for developing countries quadrupled in
the 1990s, rising to an annual rate of 8.5 percent from growth rates of less than 2 percent in
1980s.
Figure 5.1: Export shares of GDP
(percent)
25.0
Developing countries
22.5
20.0
17.5
15.0
High income countries
12.5
10.0
1980
82
84
86
88
1990
92
94
96
98
2000
2
Source: GEP 2003, World Bank data.
The decade of the 1990s saw a tremendous increase in the global integration in goods,
services and investment flows (Table 5.2). The largest increase was in investment flows, with
foreign direct investment (FDI) as a share of GDP rising by more than four times between
152
1990 and 2000. Developed countries experienced a much larger increase in FDI flows, which
rose more than five times as a share of their GDP, while developing countries experienced a
more modest but still sizeable increase of such flows.
Table 5.2: Global integration, 1980-2000
Exports and Imports of Goods and
Services as a share of GDP (in
current US$)
1980-85
1990
2000
avg
Developing countries
Developed countries
41.0
40.7
39.2
39.1
55.3
46.4
FDI flows as a share of GDP
(in percent)
1980
1990
2000
0.3
0.6
0.8
1.0
2.6
5.1
Note: A nominal measure is used here because of the difficulty in obtaining price deflators for services trade.
Regardless of whether a real or a nominal measure is used, there was still a large increase in trade integration on
the “goods” side in the 1990s. The analysis in the rest of the chapter regarding goods trade uses “real”
measures, with nominal values deflated by the relevant price indices.
Source: Trade and foreign direct investment flow figures from IMF Balance of Payments Statistics;
GDP from World Bank World Development Indicators.
In terms of absolute magnitudes (as measured by the share of GDP), the big story of the
1990s is trade integration, especially for developing countries. By 2000, exports and imports
of goods and services had reached more than half (55 percent) of developing countries’ GDP
in the aggregate, surpassing the developed country aggregate share of 46 percent. The
increase in trade integration was also much greater for developing countries, as both groups
had started the 1990s with nearly identical trade shares of 39 percent. And, despite the
increasing importance of services trade (it rose from 8.1 to 9.6 percent of GDP in developed
countries, and 7.6 to 9.4 percent of GDP in developing countries between 1990 and 2000),
goods trade integration dominated the globalization scene in the 1990s (Table 5.3).
Table 5.3: Exports and imports of goods and services as shares of GDP
(in current US$)
Export and Import Shares
of GDP
Goods
Services
Goods and services
Developed countries
1980
34.2
7.7
41.8
1990
31.0
8.1
39.2
Developing countries
2000
36.8
9.6
46.4
1980
37.1
7.9
45.1
1990
31.6
7.6
39.2
2000
45.9
9.4
55.3
Source: Trade figures from IMF Balance of Payments Statistics; GDP from World Bank World
Development Indicators.
Equally remarkable was the shift in the composition of developing country exports. As a
group, developing countries moved beyond their traditional specialization in agricultural and
resource exports into manufactures trade. Exports of manufactures grew at nearly twice the
rate of agriculture, and now constitut e nearly 80 percent of exports from all developing
countries. Countries that were low- income in 1980 managed to raise their exports of
manufactures from roughly 20 percent of their total exports to more than 80 percent (Figure
153
5.2). As a result, many grew quickly and entered the ranks of today’s middle- income
countries. The middle- income group of 1980 also increased their manufactured share, but
somewhat less rapidly, to reach nearly 70 percent. 1
Figure 5.2: Developing countries: important exporters of manufactures
In middle income countries, manufactures make up 70 percent of
exports…
Middle-income countries
’ share of world exports,
1981
80
70
-2001 (percent)
Manufacturing
exports (%)
60
50
Resources
exports (%)
40
30
20
Agricultural
exports (%)
10
0
1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001
1
These changes were not just due to declines in the prices of agricultural and resource commodities re lative to
manufactures—the strong shift in the composition of exports shows up even when price changes are removed.
Further, it was not just due to a few, large high-growth exporters such as China and India. Excluding China and
India, the share of manufactures in developing country exports grew from one-tenth in 1980 to almost twothirds in 2001. It increased sharply, but not equally, in all regions. The laggards included Sub-Saharan Africa
and the Middle East and North Africa, which have yet to reach 30 percent. Many countries, particularly the
poorest, remain dependent on exports of agricultural and resource commodities.
154
… and in low income, manufactures make up 80 percent of exports
Low-income countries’ share of world exports, 1981-20001 (percent)
90
Manufacturing
exports (%)
80
70
60
Resources
exports (%)
50
40
30
Agricultural
exports (%)
20
10
0
1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001
Source: World Bank, Global Economic Prospects (2004.
The rising tide of exports did not lift all boats. Forty-three countries had no increase on
average in their merchandise exports between 1980 and 2000. 2 They were plagued by civil
conflict, and engaged in politically motivated trade embargoes—factors that were often
complicated by inept governance. Possibly as a result of these factors, or for independent
reasons, many of these countries also relied excessively on one or two primary commodities.
The poor performance of these countries can be traced to a number of explanations, of which
trade policy is only one factor—consistent with the themes emerging from this chapter.
Services trade was particularly important for some developing countries in the 1990s in
accelerating either their trade integration, or income growth, or both. India and Mauritius are
notable examples of countries that experienced much faster growth in services than goods
trade, on both the import and export fronts. Between 1990 and 2000, services exports (in
current dollars) from Mauritius grew at more than 8 percent per annum, compared with 2
percent per annum for goods exports; the same figures for India were 15 and 9 percent,
respectively. In India, software exports constituted 40 percent of services exports in 2000-013
(around 10 percent of exports of all goods and services), which was by far the largest service
export component. In Mauritius, communications and computer services experienced the
fastest growth during the 1990s, tripling in dollar terms over the decade, and were the only
category of service exports to increase its share of services exports (from 17 to 23 percent).
Industrial countries imposed important restrictions on labor movements. In addition to goods
and services, the global integration of labor has been emerging as an important issue in the
globalization agenda during the 1990s. In 2001, remittances from permanent as well as
2
World Bank (2003b), p. 63.
Software exports were $6.2 billion in 2000/2001 according to the Trade Policy Review of India, 2002, by the
World Trade Organization.
3
155
temporary migrants provided some US$ 71 billion to developing countries, nearly 40 percent
more than all official development assistance and significantly more than net debt flows to
developing countries’ income. 4
However, such remittances were concentrated in a few developing countries; their importance
for developing countries as a group has actually been declining over the decade, falling from
slightly above 4 percent of all foreign exchange receipts to slightly below. 5 Yet the decline in
remittances is difficult to identify since so much is unofficial. To many parts of the world,
unofficial remittances far outweigh official ones—the 4 percent figure should hence be seen
as a crude approximation. Regardless of the exact amount, existing payments systems make
remittances difficult and costly, especially in and to Africa and Central America.
The small share of remittances in foreign exchange receipts for developing countries,
however, is more indicative of industrial country restrictions on labor movements than
anything else. Allowing temporary labor movements could play an important role as a
poverty reduction instrument: if the temporary movement of labor up to 3 percent of the total
labor force in rich countries were permitted, developing countries would stand to gain as
much as $160 billion in additional income. 6 However, virtually all GATS commitments have
focused on the first three “modes” of international service delivery rather than mode 4, which
is labor. Mode 4, which involves the temporary movement of labor to provide services,
accounts for only 1.4 percent of services trade (Figure 5.3). To date, little has been done to
loosen conditions governing the temporary movement of natural persons supplying services.
The lack of liberalization in labor services has been particularly costly to developing
countries, whose comparative advantage lies in the export of medium and low-skilled, laborintensive services.
4
World Bank (2003b), p.xxiii.
From 1990 to 2000, income from migrant workers overseas (including workers’ remittances and employees’
compensation) as a share of foreign exchange receipts (measured as exports of goods, services and workers’
income) fell from 4.3 to 3.8 percent for all developing countries. Conceptually it makes sense to compare
income from migrant workers with receipts from exports of goods and services since labor could be viewed as
one form of a country’s service exports. Almost all of the drop could be attributed to the decline in migrant
workers’ income in Egypt5 , which in 1990 had enjoyed the largest amount of this income in nominal terms in
the developing world. The decline in migrant workers’ income in Egypt during the 1990s was Gulf War-related.
Excluding Egypt, the ratio fell from 3.7 to 3.6 percent over the decade. The countries where incomes from
migrant workers have become quite important—ranging between 20 to 46 percent of total foreign exchange
receipts in 2000—and where such income had increased significantly over the 1990s (increases ranging from 10
to 46 percent) include Albania, Ecuador, Jamaica, Jordan, Nicaragua, Sudan and Uganda. At the same time,
however, there are also countries that experienced large declines in such incomes ranging from 10 to 30 percent;
these include Benin, Cape Verde, Egypt, Lesotho, and Pakistan.
6
Walmsley and Winters (2003).
5
156
Figure 5.3: Temporary labor mobility, underused mode of trade in services
(Value of world trade in services by mode (percent)
Mode 4 (movement of natural persons)
1%
Mode 1 (cross
-border supply)
28%
Mode 3 (commercial
presence)
57%
14%
Mode 2 (consumption
abroad)
Source: IMF Balance of Payments Yearbook
Source: World Bank, Global Economic Prospects 2004.
2.
Trade: one element in successful growth strategies
Developing countries shifted to more liberal trade policies in the 1990s. By the early years of
the decade, the superiority of outward orientation over import substitution as a development
strategy was generally accepted by researchers and policymakers alike. 7 The fall of
communism in central and Eastern Europe, together with the collapse of the former Soviet
Union, gave further credence to the superiority of outward orientation over inward
orientation as an economic strategy. Many countries had already started liberalizing their
trade regimes before the 1990s, while others followed suit or escalated their efforts during
the decade, including hitherto very highly protected and inward- looking economies like
India. Trade liberalization was also adopted widely by countries in sub-Saharan Africa (SSA)
as a key instrument to reverse the hitherto dismal growth performance in many of them.
As in the case of macroeconomic reforms, however, the results varied and, in general, fell
short of expectations—whereas openness has he lped efficiency and growth in many cases
(East and South Asian countries, Botswana, Chile, Mauritius, Tunisia ), it failed to do so in
many others. One important determinant of the success of different trade reforms was the
presence of complementary policie s.
7
See Krueger (1997) and Baldwin (2003) for expositions on the evolution of economic thinking over this issue
during the second half of the twentieth century.
157
The need for complementary policies as part of a successful growth strategy
The rapid growth in trade integration and the fundamental changes in trade patterns that
occurred during the 1990s were achieved through unilateral and multilateral trade reforms, in
conjunction with a number of complementary reforms and investments. World Bank
estimates indicate that tariff cuts in industrial countries accounted for about one-third of the
improvement in market access and tariff cuts in the developing countries themselves
accounted for two-thirds of the improvement in market access for the developing countries. 8
While trade reforms both at home and abroad were one critical factor in explaining the
increase in global integration, this section highlights other developments and policies that
played important complementary roles, in particular, macroeconomic policies and traderelated infrastructure and institutions.
Although this chapter focuses on only two sets of complementary policies, many other
factors contributed to the rise in trade integration in the 1990s. Average educational levels
and capital stock per worker rose sharply throughout the developing world. Improvements in
transport and communications, in conjunction with developing-country reforms, allowed the
production chain to be broken up into components, with developing countries playing a key
role in global production sharing. Increases in foreign direct investment also played a role
(Box 5.1). Foreign investment grew dramatic ally during the 1990s, bringing capital to
developing countries, augmenting the total supply of capital per worker, and bringing
connections with other elements in the network of global production sharing.
The need for macroeconomic stability
The presence of macroeconomic stability was an important component of successful
outcomes from trade reforms (Thomas and Nash 1992; Nash and Takacs 1998).
Macroeconomic stability entails low levels of inflation and a stable and competitive
exchange rate. Exchange rate volatility creates a risky business environment in which there
are uncertainties about future profits and payments. These risks are especially exacerbated in
the many developing countries that have not developed financial instruments for hedging
against foreign exchange risk. Some of the most severe cases of exchange rate
mismanagement and attendant highly volatile real exchange rates in the 1990s were found in
Ghana, Kazakhstan, Malawi, Nigeria, Zambia, Russia, and a few other CIS countries in
Central Asia.
Proper exchange rate management requires, inter alia, appropriate sequencing of trade
reforms and capital account liberalization. In particular, sequencing of trade reforms prior to
capital account liberalization is important, since the large inflows of capital that generally
accompany the latter could lead to a large appreciation of the real exchange rate, with the
negative consequence of large import surges that destabilizes domestic industries and the
balance of payments.
8
World Bank (2003b), p. 76.
158
Box 5.1: The impact of foreign direct investment on growth
Countries welcome FDI for many reasons. Capital-scarce countries benefit from the infusion of a less
volatile source of capital. Greater investment financed by incoming FDI should also translate into
higher growth. Foreign investors are expected to provide employment opportunities, better wages and
working conditions, and more training. In many countries, foreign firms and joint ventures are given
special treatment in the expectation that these firms will transfer new technology and knowledge to
domestic workers and firms. What does the evidence suggest?
There are two approaches to studying the impact of FDI on host countries: cross-country studies,
which examine the relationship between FDI and growth, and micro studies which focus primarily on
the impact of foreign firms on host country wages and test for the possibility of technology transfer.
Cross-country exercises capture the direct effect of FDI (higher productivity levels/growth rates
exhibited by foreign affiliates) and the contribution of FDI to capital accumulation, as well as
possible technology spillovers. The cross country evidence is mixed, in part because incoming FDI as
a share of GDP is typically quite small. A recent cross-country study using data for 72 countries for
1960-95 (Levine and Carkovic 2003) finds no evidence that FDI exerts a positive impact on economic
growth independent of other growth determinants (openness, black market premium, financial
development, initial income, years of schooling). However, Bosworth and Collins (1999) find that
FDI, by raising total factor productivity, raises a country’s rate of output growth. Borenzstein et al.
(1998) find that FDI both added to capital accumulation as well as raised the efficiency of investment,
but only when the host country has a minimum threshold of human capital, an indicator of absorptive
capacity. The Borenzstein study is consistent with more recent evidence that suggests that FDI can
promote growth if the country has complementary institutions such as developed financial markets
(Alfaro et al. 2003) or is open to trade (Balasubramanyam et al. (1996)).
There are also a number of studies using micro-data that analyze the role of FDI in promoting
technology transfer and raising host country wages. The evidence on firm-level analyses of intraindustry spillovers is mixed. In explaining differing micro-level evidence on the impact of FDI on
growth, a recent World Bank Report showed that absorptive capacity can explain differences in
ability to absorb technology from foreign firms (GDF 2000). Accordingly, Malaysia and Taiwan
(China) have high absorptive capacity, and FDI was found to have a positive relationship to
productivity, while Morocco, Uruguay, and Venezuela have low absorptive capacity, and FDI was not
found to have a positive impact on productivity. However, the issue is not just the absorptive
capacity. These findings could reflect an immediate (and short-run) shock from foreign entry
resulting in local firms losing market shares versus the delayed (but possibly long-term) effect of
knowledge spillovers. The loss in market share is one explanation for the lack of spillovers within the
same sectors identified by Aitken and Harrison (1999), Haddad and Harrison (1993), Djankov and
Hoekman (2000), Konings (2001), and Damijan et al. (2003).
However, what all these firm level studies agree on is that foreign affiliates are more productive than
indigenous firms. While part of these results could reflect the fact that foreign firms acquire more
efficient domestic enterprises, anecdotal evidence also suggests that local firms acquired by foreign
investors undergo restructuring and improve their performance as a result of the takeover. This direct
effect should not be ignored as its magnitude may be significant. Other evidence also suggests that
foreign enterprises pay higher wages (Aitken et al. (1997), Harrison and Scorse (2003)).
In sum, while quite a lot of evidence points to FDI’s positive contribution to growth, there is no
consensus on the issue, and in particular no consensus on causality. Regardless of whether FDI
independently contributes to growth, it is clear that policies and institutions that are important for
159
growth would also be the ones that would attract FDI as well as enhance the impact of FDI on
growth. Therefore, countries should focus on such policies and institutions rather than narrowly on
how to attract FDI.
India provides a good example of proper sequencing of its trade reforms as well as
maintenance of an appropriate macroeconomic framework, both of which contributed to its
impressive export and growth performance in the 1990s. 9 A significant real depreciation of
the real exchange rate had preceded the first liberalization measures in the early 1990s. This
served to increase export incentives and cushioned the impact of lower import barriers on the
domestic industry. Also, trade liberalization had preceded opening of the capital account.
And, since 1992, India’s real effective exchange rate has remained at more or less the same
level, which has facilitated trade reforms.
Zambia, on the other hand, provides a good illustration of how macroeconomic instability
can undermine potentially positive effects of structural reforms. Despite undertaking
substantial trade and other structural reforms in the early 1990s which had resulted in one of
the most liberal trade regimes in Africa, Zambia’s export performance has been lackluster.
One of the reasons for this outcome is an unstable macroeconomic environment, with high
inflation and high real interest rates, as well as a highly volatile real exchange rate resulting
from external shocks (large declines in 1995 and 1997 in the prices of copper, Zambia’s main
export) and poor management of the se shocks. 10
Malawi provides another example where macroeconomic instability undermined export and
growth performance. The country had not experienced a sustained period of macroeconomic
stability during the 1990s. Inflation was generally high and on top of that, volatile, with rates
ranging from 9 percent to more than 80 percent, and averaging 31 percent for the entire
period. This had resulted in an overvalued and highly volatile real exchange rate, both of
which had seriously undermined domestic production, investment, and exports. The
manufacturing sector experienced a major contraction, falling by 9 percent during 1995 and
1996. These developments hindered the export diversification efforts of Malawi, which has
continued to be highly concentrated in tobacco. 11
The need for trade-related infrastructure and institutions
Successful trade integration requires supporting trade-related infrastructure and institutions.
This is the so-called “behind-the-border” agenda, which in principle could include virtually
all aspects of the development agenda, as they are undoubtedly all important for engendering
a supply side response. 12 The importance of institutions is illustrated by the comparison
below of Jamaica and Mauritius (Box 5.2). Two additional elements that have emerged as
9
World Bank (1994).
World Bank (2003g).
11
World Bank (2003d).
12
From Tsikata (2003), which summarized the findings of Diagnostic Trade Integration Studies undertaken
during 2001-03 for twelve least developed countries (Burundi, Cambodia, Ethiopia, Guinea, Lesotho,
Madagascar, Malawi, Mali, Mauritania, Nepal, Senegal, and Yemen).
10
160
important constraints are transport infrastructure and institutional capacity for meeting
products standards.
Box 5.2: Jamaica and Mauritius – a tale of two islands
Starting at nearly the same per capita GDP in 1984, the income levels of Jamaica and Mauritius began
to diverge significantly such that by 2000, per capita GDP in Mauritius had more than doubled from
its 1984 level in real terms (rising from $1,951 to $4,160 in 1995 US dollar terms), while that in
Jamaica had more or less stagnated (rising from $1,925 to only $2,150 in 1995 US dollar terms). Over
the period 1984-2000, per capita GDP grew at around 0.7 percent per annum in real terms in Jamaica,
compared with 4.8 percent in Mauritius. This is a dramatic difference in growth performances,
especially given the many similarities between the two countries.
To start with, the two countries enjoyed similarities in natural endowments and historical legacies.
Both are island economies, have tropical climates, are subject to shocks of nature (hurricanes in
Jamaica and cyclones in Mauritius), and are former British colonies with English as the official
language. Their economic structures are similar: around 6 percent of GDP is from agriculture; around
one-third from industry; and the remaining 60 percent or so from services. Sugar cane is widely
grown in both countries. Both countries enjoy trade preferences, in particular the Lome Convention
that granted them access to the EU sugar market and the US sugar program that granted them access
to the US sugar market. They both established export-processing zones (EPZs) where the apparel
sector is well-represented, with the primary impetus being provided in both cases by East Asian
investors which were bound by the quota limitations under the Multi-Fiber Agreement.
The disparate growth performances between the two countries cannot be attributable to differences in
trade performance. Between 1985 and 2000, Mauritian exports grew 3.9 percent per annum in real
terms, while Jamaican exports grew 3.6 percent per annum in real terms. In terms of trade shares of
GDP, in fact, Jamaica did much better than Mauritius, particularly in terms of goods trade (see table
below). This reflects the much better growth performance in Mauritius, as well as the fact that much
of this growth cannot be attributed to trade.
Real goods and services trade integration, Jamaica and Mauritius, 1980-2000
(percent)
Jamaica
Mauritius
Goods
1980
65
98
1990
57
109
2000
79
71
Services
1980
1990
39
36
26
38
2000
55
37
The analysis of several factors that are generally accepted as either fundamental or important to
growth rejected the following as possible reasons behind the growth divergence. First, initial per
capita incomes, which were practically the same in the mid-1980s. Second, education, as Jamaica
surpasses Mauritius in a host of education enrollment indicators for the two countries in 1990 and
2000. Third, FDI, as all through the 1990s Jamaica enjoyed hig her FDI as a share of GDP than
Mauritius. Fourth, geography, as Jamaica is much closer to the US and the EU than Mauritius is to
either, with the EU and the US being the two main markets that have granted preferential access to
the two countries. In fact, with the exception of the first factor, the performance of the two countries
in all the other factors would indicate that Jamaica should surpass Mauritius in growth.
The two key factors that are left that could explain the difference in growth performances are
institutional quality and macroeconomic stability. An in-depth case study of the Mauritian growth
161
performance (Subramanian and Roy 2001) has already pointed out that the Mauritius growth miracle
cannot be explained by its trade performance, that Mauritius is a “super-grower but not a super
trader.” According to that study, domestic export subsidization policies (through duty-free access to
imported inputs, tax incentives and much greater labor market flexibility in the export processing
zones) and preferential trade access (for sugar and textiles) together only manage to neutralize the
anti-export bias arising from high import protection to the economy, with the latter being the bigger
contributor. The outcome of such neutralization of the anti-export bias is a trading performance that is
average, but not exceptional (like the East Asian tigers’ performance). The study then attributes the
exceptional growth performance of Mauritius to other factors including superior institutions
(democracy and strong participatory institutions), and ethnic diversity which provided it with
important linkages with the rest of the world (68 percent of the population is Indian) and which forced
the need for economic balance to preserve the cash cow (that is, sugar), and the need for participatory
political institutions that were important for maintaining stability, law and order, rule of law, and
mediating conflict.
It appears that institutions could explain the difference in growth performances between Jamaica and
Mauritius. According to the six indicators on institutional quality compiled by Kaufmann et al.
(2002), Mauritius outperforms Jamaica in all but one: it does better in government effectiveness,
political stability, rule of law, control of corruption, and voice. Regulatory quality is the only
institutional measure in which Jamaica is superior to Mauritius. Recent World Bank analysis (World
Bank, 2003c) indicates that the rule of law is particularly a problem in Jamaica, with crime and
violence costing at least 4 percent of its GDP (excluding dynamic costs). The other major problem is
the lack of a social/political compact in Jamaica (unlike that in Mauritius), although there appears to
be progress on this front recently with the labor unions agreeing with the Government on limit their
wage increases in response to the grave economic situation.
Aside from institutions, another likely factor explaining the difference between the growth
performances of Mauritius and Jamaica is their different macroeconomic performances. Mauritius
outperformed Jamaica in macroeconomic stability for the two decades from 1980 to 2000, both in
terms of the level of inflation, as well as the stability and competitiveness of the real exchange rate
(see figures below).
Inflation, in percent
Real effective exchange rate, 1990=100
80
180
70
160
60
140
50
40
120
30
100
20
80
10
60
Jamaica
Mauritius
Jamaica
20
00
19
98
19
96
19
94
19
92
19
90
19
88
19
86
19
84
19
82
1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000
19
80
0
Mauritius
The same recent World Bank analysis indicated that the poor management of adverse macroeconomic
developments in the 1990s in Jamaica seemed to have more than offset the potentially positive effects
of the substantial trade (and capital account) liberalization of the early 1990s. In particular, the crisis
in the financial sector in the mid-1990s exacerbated an already deteriorating fiscal performance,
which led to a huge increase in debt to GDP after 1996/97. This has negative affected private sector
confidence, government investment, interest rates, and growth.
162
For exporters in some developing countries, country- level evidence has revealed transport
infrastructure as the single most important component of cost. The main issues related to
transport are lack of competition and inadequate investments. Transport costs are further
escalated by fees and checkpoints (formal and informal). Poor transport in particular affects
agricultural producers (mainly smallholder farmers and herders) who have difficulty
accessing markets, both domestic and external. 13 Even in the absence of regulatory
distortions, such localized markets may feature little competition and may fail to realize the
economies of scale or scope. The result is typically a vicious cycle of low productivity and
low profitability. Given the importance of agriculture in these poor countries (with shares of
GDP ranging from 15 to 52 percent), such constraints severely limit their growth potential.
And given that in most countries the majority of the poor reside in the rural areas, these
constraints also have seriously negative effects on poverty.
Malawi is a country where exceptionally high internal and external transport costs have
weakened the competitiveness and profitability of firms and farmers. Transport costs are by
far the single largest factor in the high costs paid by Malawian farmers. Although Malawi is
an efficient producer of sugar, domestic transport costs account for 15 percent or more of
local consumer prices, and for sugar exports, regional and international transport costs add
nearly 50 percent to the ex- mill production costs. 14 Part of the reason for high external
transport costs is geographic: Malawi is landlocked, and its links to ports are long and
uncertain. On the internal side, the lack of competition in the road transport sector (where
restrictions are placed on foreign operators) and high transport taxes (for example, tires carry
30 percent duty and a 20 percent surtax) are some of the main factors contributing to high
transport costs. In addition, Malawi is also one of the most costly countries in the world for
air transport. A variety of landing fees, temporary service permits, fees to charge batteries
and pump tires—as well as costs of aviation fuel—are generally higher in Malawi than
elsewhere.
There has been a proliferation of, and rising stringency in, products standards in international
trade in recent years. While this is occurring in official channels, it is increasingly driven by
consumer and commercial interests and magnified by advances in technology and, more
recently, security concerns. Consumers and civil society organizations are demanding stricter
food standards, the origin of which had to do with some highly publicized food scares (such
as “mad cow disease” in the UK). Private sector players in the developed countries—major
food retailers, food manufacturers and restaurant chains—have been adopting codes of
practice, standards, and other forms of supply-chain governance as part of their commercial
strategies of differentiation. Increasingly, middle- income and some low- income countries are
also raising their product standards, in part through the investments undertaken by
multinational supermarket or restaurant chains and competitive responses by local firms.
Finally, new food safety standards in industrialized countries are serving to shape the
expectations of developing country consumers, especially those with higher incomes and in
urban areas.
13
The discussion in this paragraph is taken from Jaffee and Sutherland, who reviewed the same 12 studies
mentioned in the previous footnote and summarized the agricultural sector analyses in these studies.
14
World Bank (2003e), p. 49.
163
A recent World Bank study15 suggests that this trend is not going away and, furthermore, that
the prospects are dim for applying “special and differential treatment” that would permit
poorer countries to meet lower requirements. In this light, the study suggests that developing
countries need to focus on developing and improving their food safety and agricultural health
management systems in order to reposition themselves competitively and to enhance their
export performances.
Building such capacity is not beyond the reach of developing countries, and indeed some of
them—including among the poorest ones—have managed to meet exacting international
standards to access high- value markets in industrialized countries. Examples include
Peruvian exports of asparagus to the US and the EU; and fish products in low- income
African countries that meet the EU standards of hygiene with respect to processing, transport,
and storage. Countries that do well in meeting strict standards are generally those where the
private sector is well-organized and the public sector is well- focused and supports the efforts
of exporters. Programs tailored to meet the institutional capacity of developing countries
include outgrower programs for smallholder farmers, systems of training and oversight for
small and medium-sized enterprises through associations and groups, or twinning and
regional networking for small countries.
Cross-country evidence
The evidence from the 1990s suggests that a variety of factors—a stable investment climate,
greater market access, complementary macroeconomic policies and unilateral or multilateral
trade reforms—reinforced each other in developing countries that successfully integrated into
the global economy. The fact that successful trade reforms were frequently pursued within a
broader reform agenda provides one explanation for why there is no significant relationship
between changes in import shares and tariff reductions shown in Figure 5.4. In principle, we
would expect to see falling tariffs leading to rising import shares. However, without taking
into account changes in physical and human capital investments, changes in market access
and individual country macroeconomic policies, it is not surprising that tariff reductions
alone show no significant association with changes in import shares. 16
15
Jaffee and Henson (2004).
Only tariffs were used to analyze trade policy in this section because of the very limited availability of data
on non-tariff barriers (NTBs) (both for countries as well as over time). However, this should not bias the
analysis since changes in NTBs in countries have largely been in line with changes in tariffs.
Changes in tariffs are defined as the percent change of (1+tariff), with tariffs being the simple average of
statutory tariffs. In Figure 5.4, the correlation coefficient is -0.18 and not statistically significant. Note that these
are simple correlations to illustrate the associations between the two variables only, without attribution to
causality.
16
164
Figure 5.4: Changes in import shares of GDP and changes in tariffs , 1990-2000
Percent change in real M/GDP
5
4
3
2
1
0
-0.5
-0.4
-0.3
-0.2
-0.1
-1
0
0.1
0.2
0.3
Percent change in relative price impact of tariffs
Note: Changes are for the ten-year period 1990-2000, not per annum.
Table 5.4 also illustrates the large range of integration responses to the reduction in tariffs in
the 1990s. There were large and positive trade integration responses in the countries that had
begun the 1990s with very high tariffs and reduced them the most. The table shows a much
wider range of responses for countries that began the decade with more moderate levels of
tariffs and lowered them further. One explanation consistent with this evidence is that at
more moderate levels of protection, other changes in the economy play an increasingly
important role in explaining changing trade shares. These include macroeconomic policies,
changes in “behind the border” barriers such as transport costs and health and safety
regulations, and changes in the quality of institutions.
165
Table 5.4: Tariff reductions and changes in goods trade integration
Reductions in
<1 times
tariffs, late 1980s
- late 1990s
40-70%
20-30%
Pakistan, Burkina
Faso
Change in integration, 1990-2000
1-1.5 times
1.5-2 times
India
Benin, Ecuador, Kenya,
Peru, Thailand
10-20%
Egypt, Iran,
Republic of Congo,
Mauritania,
Indonesia, Turkey,
Mauritius, Zambia Uganda, Venezuela
0-10%
Tanzania,
Paraguay,
Senegal
0-2% increase
Mozambique
2-10% increase
Tunisia
>10% increase
Syrian Arab
Republic
Bolivia, Chile, Cote
d'Ivoire, Jamaica,
Malaysia, Nigeria, South
Africa
Bangladesh
China
>2 times
Sudan
Argentina, Colombia,
Philippines
Costa Rica, El Salvador,
Guatemala, Nicaragua,
Sri Lanka
Ghana, Nepal
Madagascar, Trinidad
and Tobago
Mexico
Jordan, Morocco, Oman,
Saudi Arabia
Note: Trade integration, defined as the share of export plus import shares of GDP, is measured in real terms and
excludes services trade.
Interestingly, Figure 5.5 shows that tariff reductions in the 1990s had a much greater impact
on developing countries’ export shares than on their import shares. 17 This is consistent with
so-called “Lerner symmetry”, whereby taxing imports or exports has the same effect on
international trade. A tariff on imports is equivalent to a tax on exports, 18 and hence reducing
tariffs promotes exports. Together, Figure 5.4 and Figure 5.5 suggest one important avenue
through which trade reforms lead to growth: they reduce the anti-export bias in an economy,
thereby promoting growth by encouraging exports.
17
In Figure 5.5, the correlation coefficient is –0.25 and statistically significant between the relative price impact
of tariff changes and changes in real export shares.
18
Lerner’s symmetry in the two-good case can be illustrated as follows: Px/Pm(1+t) = [Px/(1+t)]/Pm, where
Px=price of exports; Pm=price of imports; t=tariff.
166
Figure 5.5: Changes in export shares of GDP and changes in tariffs, 1990-2000
Percent change in real X/GDP
2.5
2
1.5
1
0.5
0
-0.5
-0.4
-0.3
-0.2
-0.1
-0.5
0
0.1
0.2
0.3
-1
Percent change in relative price impact of tariffs
Note: Changes are for the ten-year period 1990-2000, not per annum.
Some lessons
Tariff reductions constitute only one element of a reform program for generating the
expected gains from integration. Positive growth responses to trade reforms are most likely
when trade reforms are part of a comprehensive strategy, with macroeconomic stability as
one of the key other elements. Other important complements include trade-related
infrastruc ture and institutions, investments in physical and human capital, greater access to
developed and developing country markets, and a sound rule of law.
Reaffirming lessons from earlier decades, trade reforms needed to be undertaken in the
presence of macroeconomic stability. These risks are especially exacerbated in countries
where financial instruments for hedging against foreign exchange risk are not developed,
which is the case in many developing countries. Proper exchange rate management requires,
inter alia, appropriate sequencing of trade reforms and capital account liberalization. In
particular, sequencing of trade reforms prior to capital account liberalization is important,
since the large inflows of capital that generally accompany the latter could lead to large
appreciation of the real exchange rate, with the negative consequence of large import surges
and destabilization of domestic industries and the balance of payments.
The evidence discussed in the preceding section highlighted one important mechanism
through which trade reforms affect growth: by reducing tariffs, trade reforms encourage
exports. What in general are major determinants of export growth? Empirical analysis
suggests that a number of factors mattered for export performance in the 1990s, including
macroeconomic stability, relative price changes arising from changes in tariffs and
167
government effectiveness. 19 Macroeconomic stability refers to the stability of the real
effective exchange rate as measured by the standard deviation, and average inflation. Tariff
reductions were found to be important for export growth in the 1990s, consistent with the
larger impact of tariff changes on export rather than import shares of GDP discussed earlier.
Government effectiveness reflects the combined perceptions of the quality of public service
provision, the quality of the bureaucracy, the competence of civil servants, the independence
of the civil service from political pressures and the credibility of the government’s
commitment to policies. 20 Estimation conducted for this chapter suggests that export growth
was higher in the 1990s in countries with greater macroeconomic stability, countries that
reduced tariffs more, and countries that had more effective government, as defined above.
In addition to the cross-country evidence described above, a number of detailed case studies
have generated some stylized facts on the determinants of export activity. A series of papers
using detailed plant- level data have shown that the manufacturing firms that move into
exporting are frequently the most productive in an economy. Consequently, policies that
encourage investments in human and physical capital, and that support technological change,
are likely to promote export growth. Evidence for Morocco suggests that many exporters are
new enterprises, which suggests that policies that encourage new plant entry and at the same
time facilitate exit for inefficient enterprises are likely to play an important role. Evidence
from Mexican and Indonesian censuses suggests that exporters are likely to use skilled labor,
which suggests that policies supporting the development of human capital are important.
Plant- level studies and anecdotal evidence also point to the importance of foreign investors in
helping developing country exporters to break into new markets. The latest studies also
control for the possibility of reverse causality, taking into account the fact that foreign firms
may create or take over most efficient firms (Aitken, Hanson, and Harrison 1997). Even if
the importance of foreign investment is difficult to identify in cross-country studies (Box 5.1
and forthcoming Technical Note Number 2), plant- level studies provide ample evidence that
foreign ownership has been associated with export activity. Studies on Indonesia, Mexico,
and Morocco show that joint ventures and foreign owned plants are significantly more likely
to export than other types of enterprises. Although the mechanism is not completely clear,
foreign firms are likely to provide knowledge of foreign markets and customer preferences,
as well as access to new technology and financing opportunities.
Infant industry protection
While import substitution policies have been largely discredited, the need to address market
failures preventing the development of internationally competitive industries in developing
countries has continued to be of concern to policymakers and academics. Suggestions have
been made to grant temporary modest levels of import protection where there is a
demo nstrated need (Williamson 2004). Other suggestions have focused on the right form of
protection, and the fact that inducements in the form of import protection could be
19
Technical Note Number 2, forthcoming as part of this volume, will provide a full discussion of the
regressions and regression results.
20
The government effectiveness indicator is compiled by Kaufmann et al. (2003) based on 17 separate sources
of subjective data on perceptions of governance constructed by 15 different organizations.
168
counterproductive. It has been pointed out that incentives should be in the form of subsidies
to the initial entrants rather than an import duty, as the duty (through the rent it provides)
would encourage other firms to acquire the knowledge discovered by the initial firms faster,
thus discouraging the initial firm’s investment in knowledge acquisition in the first place
(Baldwin 2003).
As discussed in Chapter 1, other suggestions have been to approach development as a
process of “self discovery” (Hausmann and Rodrik 2002). Specifically, according to this
approach, the key challenge that a country faces in its process of transforming into a modern
economy is learning what it is good at producing. The initial entrepreneur who discovers
what the country should specialize in can capture only a small part of the social value that
this knowledge generates, as other entrepreneurs will quickly emulate such discoveries. In
light of these externalities, this type of entrepreneurship will typically be undersupplied and
economic transformation delayed. Hence, there is a role for government involvement both to
provide incentives to induce such investments, as well as to exert discipline in pruning
investments that have turned out to be high cost ex post.
A key challenge in pursuing such strategy is to structure the right combination of incentives
(inducements) and discipline (competitive pressures, resistance against special interests).
Some of the world’s most successful economies during the last four decades (South Korea
and Taiwan, China since the early 1960s, China since the late 1970s) prospered by pursuing
policies that combined inducements for investment and risk taking while, at the same time,
gradually expanding competitive pressures that ensured efficient allocation by investors. The
contrast between the successful experiences of Asia with the generally failed experiences of
Latin America provides a useful illustration. During the industrial drives of South Korea and
Taiwan in the 1960s and the 1970s, export subsidies were provided contingent on export
performance. This strategy allowed policymakers to distinguish firms and sectors that were
highly productive from those that were not. The kind of subsidies provided included
supplying inputs, providing working capital, imposing import restrictions, and—in the case
of the Taiwanese textile industry in the 1950s—buying up the resulting production. Local
production grew spectacularly as a result. But the government also pruned non-productive
firms subsequently. By contrast, when Latin American countries followed import substitution
strategies in the 1960s and the 1970s, governments provided incentives without sufficient
discipline, with the result that too many low productivity firms operated alongside the high
performers. When discipline came along in the 1990s through trade openness and domestic
competition, however, there was too little promotion (Hausmann and Rodrik 2002). The
absence of a right balance between promotion and discipline on the part of Latin American
countries resulted in industrial performance that fell short of the East Asian countries during
all these decades.
Contrary to common perception that Chile pursued policies exclusively supporting market
discipline, the reality is that inducements have been part of that country’s success in the last
decades. Chile appears to be the exception among Latin American countries by striking the
right balance of inducements and discipline in promoting a domestic industry (Box 5.3).
169
Box 5.3: Behind Chile’s success – a less than orthodox approach
Chile, with its economic success, has often been touted as a miracle of free-market economics. In
reality, public -private collaboration strategies have played a very important role in fostering structural
change and stimulating non-tradit ional activities in the country. Indeed, the Chilean Government has
had a hand in the early stages of development of all major non-traditional exports.
Chile’s two largest export items after copper, fruits and salmon, have both benefited from privatepublic sector partnership. The foundations of the fruit industry were laid in the early 1960s through
the efforts of the Corporacion de Fomento , University of Chile and the National Institute of
Agricultural Research (INIA). The INIA, established in 1964 and staffed with highly-salaried skilled
researchers, had initiated the fruit research program from the start. The public sector carried out much
of the development of scientific personnel and knowledge to achieve technological transfer;
identification and planting of new varieties suitable for foreign markets; improvements in orchard and
post-harvest management; and the development of the infrastructure necessary to export fruit to
foreign markets. Private investment and exports took off after the reforms of the mid-1970s once
uncertainties regarding land reform, macroeconomic stability, and labor militancy were resolved.
They were further boosted by the sharp real depreciation of the currency in the mid-1980s.
The salmon industry, which generates $600 million in exports and provides jobs for more than
100,000 people in this country of 15 million, also benefited significantly form public interventions. It
was created single -handedly by Fundacion Chile , a non-profit institution created by the Chilean
Government in 1976. Fundacion Chile brought the technology of salmon farming to Chile, adapted it
and made it commercially viable, formed private sector businesses to use it, and eventually sold its
participation to Japanese investors at great profit.
Sources: Rodrik and Hausmann (2003); Ocampo (2004); and Washington Post, January 21, 2004.
Identifying the conditions that make it possible to assist new activities is not easy. Rodrik and
Hausman (2003) emphasize the importance of creating an institutional architecture that
resists the pull of special interests. While remaining agnostic on how this can be done, they
emphasize the importance of political leadership from the top, either through the guidance of
the country’s leaders and/or through a high coordination council. Whatever institutions are
employed to support new activities, they must be transparent and accountable. Otherwise,
selective support is likely to evolve into a new mechanism for supporting private interests in
the name of public gain.
Promotion of ne w activities should conform to a set of design principles, which include the
following: (1) incentives should be provided only for new, “sunrise” activities, not sunset
ones; (2) there should be clear benchmarks for success or failure; (3) there must be a
predetermined end to support (a so-called sunset clause); (4) public support should target
activities such as worker training or infrastructure investment, rather than specific sectors
such as electronics; (5) subsidized activities should provide clear potential for externalities;
(6) agencies involved in these activities should be autonomous enough to avoid capture by
private interests, yet maintain links with the private sector to maximize economywide gains.
This is not a prescription for creating new state enterprises, promoting existing activities, or
giving governments authority to expand their bureaucratic reach. Clearly, the institutional
and administrative requirements for success are formidable.
170
Evidence on trade reform and economic growth
As discussed in Chapter 2, on average, developing countries expanded exports more and
enjoyed higher per capita income growth than developed countries in the 1990s. 21 Yet by the
end of the 1990s, skepticism began to emerge on the necessity of trade openness for growth,
in large part because of the mixed outcomes of trade liberalization on growth and poverty
reduction. Many countries—as many as 43 were identified by the World Bank (GEP 2003)—
were left behind in the process of globalization. Many of them were low- income countries. In
SSA, countries such as Malawi and Zambia experienced poor trade and growth outcomes
despite quite extensive liberalization. These different outcomes led many policymakers,
academics, and human rights activists to repeatedly ask the question: what is the relationship
between trade reform and economic growth?
Academic researchers have been debating the merits of openness and its association with
growth for at least several decades (Box 5.4). The difficulties associated with finding a robust
statistical relationship between trade integration, trade policy, and growth can be illustrated
with Figure 5.6 and Figure 5.7. The first figure shows that countries with positive export
growth grew faster than other countries in the 1980s and 1990s. Although there is no
question that rising export shares are correlated with higher GDP growth, disentangling the
direction of causation is difficult. For example, fast growing countries that are becoming
more productive as a result of investments in human and physical capital—investments that
possibly result from institutional reforms not associated with trade liberalization—are also
likely to have rising export shares.
Figure 5.6: Integration with global markets and growth
Average annual per capita growth, 1980-99 -
3.5
3.0
Decreasing export
share in GDP
Increasing export
share in GDP
2.5
2.0
1.5
1.0
0.5
0.0
21
For developing countries as a group, the growth of real GDP per capita rose from 0.8 percent per annum
during 1980-90 to 1.9 percent in the decade following, surpassing income growth in developed countries which
actually dropped from 2.5 percent per annum in the 1980s to 1.7 percent in the 1990s.
171
Source: World Bank, Global Economic Prospects 2003.
Figure 5.7 shows a negative relationship between tariff reductions and growth in GDP per
capita in the 1990s.
Figure 5.7: Reductions in tariffs and changes in per capita GDP growth, 1 990s
Changes in per capita GDP growth
0.15
0.1
0.05
0
-0.15
-0.1
-0.05 -0.05 0
0.05
0.1
0.15
-0.1
-0.15
-0.2
-0.25
-0.3
Changes in tariffs
Countries that reduced their tariffs most also grew more quickly. However, the relationship is
not very strong nor statistically significant. This is not surprising given that other
determinants of growth—those mentioned in the preceding section—have not been
accounted for. Yet, even taking into account the impact of other policies, there is still a weak
relationship between openness and growth. 22 Other new research that focuses on the
relationship between trade reform and economic growth in the 1990s also finds that trade
reforms are associated with higher growth, but the strength of the results varies across
different studie s (Box 5.4) For example, Dollar and Kraay (2002, 2004); Lee, Ricci, and
Rigobon (2004); and Alcala and Ciccone (2004) all show a positive relationship between
trade and growth, whereas Rigobon and Rodrik (2004) get mixed results. One frequently
cited recent study is Wacziarg and Welch (2003), who find a positive relationship between a
22
Properly identifying the causal impact of changes in trade policies on growth needs to take into account other
factors associated with GDP growth, and the fact that changes in trade policy are often driven by other
changes —such as GDP growth. This means that trade policy should be “instrumented” or represented with
measures which affect trade policy but are not correlated with GDP growth. Since most reforms are driven by
initial protection levels, one way to get around the problem is to “instrument” the changes in tariffs in the 1990s
with initial tariffs during the 1986 through 1990 period. The latter was found to explain 36 percent of the
changes in tariffs during the decade—countries with high tariffs in the late 1980s and early 1990s reduced
tariffs by a higher percentage, while countries with already low tariffs reduced them less. The results, reported
in Background Note Number 1, also control for some other policies that affected growth in the 1990s, including
exchange rate policies, government consumption, and inflation. Although the results suggest that reducing
tariffs encourages growth, the results are not very robust.
172
composite measure of economic reforms and economic growth, but that relationship is not
significant for the 1990s. 23
Box 5.4: The trade and growth debate
The debate among economists and policymakers over the relationship between trade and growth has
risen to prominence during the last few years due on the one hand to the mixed growth outcomes of
developing countries that have undergone extensive trade liberalization and, on the other hand,
differences over data, econometric techniques and model specifications among professional
economists. On the analytic al side, the resurgence of interest in the 1990s among economists on the
impact of trade on growth can be attributed to the significant improvements in endogenous growth
theory as well as to the availability of more comprehensive data and new econometric techniques.
According to the new growth theory, attributed to Paul Romer (1986), Robert Lucas (1988) and Gene
Grossman and Elhanan Helpman (1991), whether import protection raises or lowers the growth rate
depends on the pattern of imports and exports. As a matter of theory, therefore, the relationship
between trade and growth is ambiguous, which is widely accepted by economists on both sides of the
debate. The issue is hence an empirical one, and that has become the focus of the debate in the last
few years.
The launching of the debate can be attributed to Rodriguez and Rodrik (RR) and Harrison and
Hanson (HH) who in two 1999 papers reviewed a number of empirical studies in the 1990s. While
HH showed that the Sachs and Warner (1995) study reflected the gains from macrostability rather
than trade reform, RR reviewed a number of studies, including Dollar (1992), Sachs and Warner
(1995) and Edwards (1998). RR expressed skepticism “that there is a strong negative relationship in
the data between trade barriers and economic growth, at least for levels of trade restrictions observed
in practice,” viewing “the search for such a relationship futile.” A unique feature of the HH and RR
analysis was their use of the various authors’ actual data sets in undertaking various tests of
robustness of their results. HH and RR criticized the empirical studies reviewed on data grounds, on
model-specification grounds, and on grounds of econometric techniques. Data problems included,
among others, the use of poor measures of trade barriers (including the World Bank’s classification of
trade regimes which they criticized as subjective as in Edwards’ paper), and the use of measures that
are highly correlated with other sources of bad economic performance such that what is captured is
policy failure that is not necessarily due only to trade policy (as in Dollar’s and Sachs and Warner’s
papers). Separately, Rodrik also criticized one of the more recent papers on the topic, Dollar and
Kraay (2001) on data and model-specification grounds. The data problem arises from the combination
of policy measures (tariff averages) with outcome measures (imports as a share of GDP). The model
specification problem arises from regressing income on trade shares when both are endogenous
(outcome variables).
Notwithstanding these criticisms, it would be safe to say that most authors agree on the following.
First, that trade protection is not good for economic growth
. Even RR themselves stated in their
paper that they do not think and have seen no credible evidence supporting the notion that trade
protection is good for economic growth, at least for the post-1945 period. Second, that trade
openness by itself is not sufficient for growth
, which is what RR was arguing in their paper: that
researchers and policymakers have been overstating the systematic evidence in favor of trade
openness, when what is really necessary is to further identify the connection between trade and
economic growth.
23
Nor do they isolate the role of trade policy per se, but look at the composite measure including exchange rate
reforms. Their work is done in a panel context, since they measure the impact of changes in trade policy on
economic growth.
173
More recent efforts to identify the gains from trade continue to get mixed results(see Baldwin (2003)
for a survey). Frankel and Romer (2000), for example, show a positive impact of trade on income,
using a clever technique that takes into account the endogeneity of trade policy by allowing trade
flows to depend on geographic proximity. However, their specification does not control for
unobserved differences across countries and the results are not robust to the inclusion of region
effects Other papers that find positive effects of trade on growth, using a variety of new econometric
techniques and new instrumental variable approaches, include Frankel and Rose (2003); Dollar and
Kraay (2002, 2004); Lee, Ricci and Rigobon (2004); and Alcala and Ciccone (2004). Other authors
get mixed results, including Rigobon and Rodrik (2004) and Wacziarg and Welch (2003). Wacziarg
and Welch use panel techniques that identify the impact of growth from following the same country
over time. The advantage of using panel techniques is that these approaches control for differences
across countries which are unobservable. Although Wacziarg and Welch (2003) find a generally
positive relationship between opening up to trade and economic growth, they find that this
relationship is not significant for the 1990s. In addition, Wacziarg and Welch use a composite
measure of openness, which reflects the impact of many different policies, an approach which has
been heavily critiqued by HH and RR.
Source: Baldwin (2003).
The fact that countries that have successfully integrated into the world economy have
followed different approaches and adopted a range of complementary policies makes it
difficult to pin down the exact statistical relationship between trade policy, trade integration,
and growth. The academic debates on whether openness to trade causes higher growth are
riddled with problems of measurement, reverse causation (faster growing countries tend to
open their markets more quickly), and omitted variable bias (countries that successfully
lower tariffs also adopt other complementary policies). Despite the difficulties in interpreting
country experiences during the1990s, however, most economists agree that trade
liberalization is important for growth over the long run.
Yet trade liberalization by itself is not enough for growth. 24 Recent studies show that trade
policy is most likely to be associated with positive outcomes when it is combined with a
favorable economic environment. 25 A recent World Bank study shows that lack of
regulations can undermine growth effects of trade, but that in countries with effective
regulation, the effects of trade reforms are positive for growth. 26 These academic studies
reinforce the case study and cross-country evidence presented earlier that emphasizes the
importance of complementary policies to maximize the gains from trade.
24
See, for instance, Rodrik (1997b).
Wacziarg and Welch (2003) say that “…preexisting institutional environment of countries, the extent of
political turmoil, the scope and depth of economic reforms, and the characteristics of concurrent
26
See Bolaky and Freund (2004). The authors measure excessive regulation using a World Bank survey on
labor regulations and business entry regulations. They find that the benefits of expanding trade (as measured by
trade shares) are offset by excessive regulations in the most regulated economies in the 1990s.
25
174
3.
Trade liberalization, poverty, and income distribution
Despite expected gains for the economy as a whole in the longer term, trade and investment
reforms generate both winners and losers in the short run. 27 The critical question is the
following: do the short-run costs of trade reform fall disproportionately on the poor? Is
globalization associated with changes in within- or across-country inequality? Economists in
the 1990s expected trade and foreign investment reforms to help developing countries reduce
poverty. Trade liberalization was expected to increase demand for goods produced by the
poor or low-skilled workers in developing countries. Why? Developing countries were
expected to have an abundance of unskilled labor, leading to a comparative advantage in
producing and exporting goods that used unskilled labor. Trade reforms were also expected
to increase the prices of the agricultural products produced by the poor and to reduce prices
for goods consumed by the poor. Yet if increasing globalization is in fact associated with
increasing demand for skilled labor in developing countries, then trade reforms could lead to
higher poverty and greater inequality. The effects could be worse in the short run if trade
reforms lead to unemployment and greater job instability.
Opening up to trade could affect poverty through several different avenues. One avenue is
direct: if opening up to trade is associated with higher growth, then economic growth may be
associated with a decline in poverty. This argument rests on two assumptions: first, that
opening up to trade leads to higher growth and second, that growth raises incomes of the
poor as much as it raises the incomes of the rich.
What actually occurred? There is widespread evidence that aggregate growth does reduce
poverty. 28 In other words, GDP growth is generally neutral with respect to changes in income
distribution, which means that all individuals gain. However, there is no clear evidence on
whether GDP growth is “pro-poor”, which would mean that the poor gain proportionally
more than the rich. To achieve GDP growth which is biased towards the poor, policies to
promote low- income households would probably have to be explicitly followed.
Trade, aggregate growth, and poverty reduction
To the extent that trade liberalization is associated with growth in the long run, trade reforms
should be associated with reductions in poverty. As discussed earlier in this chapter, while
the direction of causality is statistically not clear, the cross-country evidence suggests a
positive relationship between trade reform and long run growth. China and India are
prominent examples of countries that have experienced tremendous increases in trade
integration and growth, as well as large reductions in poverty. From 1980 to 2000, real per
capita GDP grew at an annual average of 8.3 percent in China and 3.6 percent in India, while
27
See Winters et al. (2004) and Goldberg and Pavcnik (forthcoming) for comprehensive surveys.
See, for example, the survey papers by Berg and Krueger (2003) and Winters et al. (2004), as well as papers
of Dollar and Kraay (2001). The general conclusion of these papers is that growth, including growth that is due
to trade openness, has no impact on income distribution, and therefore that income growth has led directly to
poverty reduction on average(that is, growth increases the incomes of the poor by just as much as it increases
the incomes of everybody else on average).
28
175
trade integration (of goods and services in real terms) doubled from 23 to 46 percent of GDP
in the former, and grew by 60 percent from 19 to 30 percent in the latter. Over this period,
both countries experienced massive reductions in the incidence of poverty—from 28 to 9
percent between 1978 and 1998 in China, and from 51 to 27 percent between 1977-78 and
1999-2000 in India. 29 These large reductions in the incidence of poverty in the two countries
have served to reduce or mitigate overall inequality in the world over the last 10 to 20
years—since a large share of the world’s poor live in these two countries—even though at
the same time inequality has risen within both countries. 30
Trade reforms can also affect poverty indirectly. Winters et al. (2004) identify a number of
important channels : (1) trade reforms could affect employment opportunities and wages of
the poor; (2) reforms also affect the prices that poor consumers pay for the goods that they
buy; (3) trade reforms could reduce government revenues and in turn social expenditures that
particularly affect the poor; and (4) liberalization could increase income instability as well as
workers’ chances of becoming poor. Even if aggregate poverty falls or remains constant,
many households may move into or out of poverty as a result of liberalization
An emerging literature using household level data suggests that the effects of trade
liberalization via changes in factor and goods prices can lead to poverty reduction. For
instance, a recent study of trade liberalization in Argentina using household survey data 31
found that Mercosur has benefited the average Argentine household across the spectrum of
income distribution. Furthermore, the same study also finds that Mercosur has had a pro-poor
bias: on average, poor households have gained more than middle- income households from
Mercosur, while the impact on rich families is positive but not statistically significant. The
reason behind these results is that Argentine trade policy has protected the rich over the poor
prior to the reforms, and granted some protection to the poor after the reforms.
While some studies have found short-run costs to employment from trade reforms, others
have found that trade reforms increased employment over the long run, as expanding sectors
created new employment opportunities. A study by two World Bank economists found that
trade explains much of the decline of Singapore’s unemployment rate, from more than 9
percent in the 1960s to close to 2 percent in the late 1990s. A study of 18 countries in Latin
America and the Caribbean for the period 1970-96 found that trade liberalization had a
negative, though small, direct effect on employment. 32 The negative effect was found to be
compounded in countries where the real exchange rate appreciated as a result of capital
inflows that had followed the economic reforms. Similar results were obtained in a study on
Brazil for 1990-97.33 This found that although trade liberalization had a negative—though
relatively small—short-term effect on employment, the more labor-intensive output mix that
resulted over the long run increased employment. Appreciation of the real exchange rate
during the period also contributed to this negative employment effect by encouraging imports
and undermining exports.
29
Asian Development Bank (2000), cited by Bhagwati and Srinivasan (2002).
Bhalla (2001); Ravallion (2003); and Sala-i-Martin (?).
31
Porto (2003).
32
Marquez and Pages (1997).
33
Moreira and Najberg (2000).
30
176
Much larger negative effects on output and employment have been found in some African
countries. One study for Kenya, Tanzania, and Zimbabwe 34 found that the majority of firms
responded to import competition pressure by contracting rather than upgrading aggressively.
Among the suggested reasons for such behavior are the firms’ lack of preparation for
competition, absence of policies to promote technological improvement (especially among
small and medium enterprises), and poor technological and human infrastructure.
These mixed results were echoed in a recent report that reviewed the impact of trade reform
in the 1990s on countries in Latin America and the Caribbean (De Ferranti et al. 2001). The
authors found losses in employment in previously protected industries and gains in
employment in others. Argentina lost much of its automobile industry while seeing an
expansion in more sophisticated chemicals and capital- and labor- intensive manufactures.
Brazil lost much of its cereals industry to Argentina under Mercosur, and its manufacturing
industry suffered more generally. Costa Rica lost much of its labor- intensive manufacturing
processes to Mexico after NAFTA, but also saw a substantial increase in manufactures of
computer chips. In each case substantial numbers of workers lost their jobs, and some
experienced either very long periods of unemployment, or large wage losses, or both. The
report makes the point that such dislocations are transitional and do not imply a permanent
increase in the unemployment rate, as shown in the experiences of Chile and Mexico. Chile
experienced double-digit rates of unemployment for several years after liberalization, but
from 1986-97 it had among the lowest such rates in the region. Mexico’s present rate of
unemployment is roughly at its traditional level, despite that country’s dramatic economic
integration with the US. Although most of these studies find that the unemployment effects
of trade liberalization tend to be temporary, even short-term costs can be high in human
terms. Such costs must be addressed through a variety of policy approaches, including
stronger social safety nets, in order to ensure that trade reforms succeed.
Finally, no direct evidence relates trade liberalization to social spending, another avenue
through which liberalization could affect income distribution. The available evidence,
relating mostly to the 1980s 35 , suggests that many trade reforms had no revenue costs. Some
of the main reasons were that temporary tariff surcharges were introduced when quantitative
restrictions were removed, and that changes in the import/export base arising from the trade
reforms had enhanced revenues. For example, one study that examined the impact of Kenyan
liberalization between 1989-99 (entailing halving the simple average import duty rate over
the period and abolishing import licensing requirements and foreign exchange controls)
found increases both in duty as a share of imports, and in import duty revenues as a share of
GDP (Glenday 2000, cited in Winters et al. 2002). The study attributed this increase in
revenues to the expansion of the revenue base, tighter exemption management, increased
duty rates on certain products (oil and agricultural commodities), a shift in imports to the
higher duty classes and possibly also improvements in customs administration and the
introduction of a pre-shipment program.
34
35
Lall (1999).
See Winters et al. (2002) for studies cited.
177
Even in cases where revenues were cut, available evidence suggests that reductions in public
expenditures important to the poor need not occur. There are alternative sources of revenues
(though caution needs to be exercised to ensure that replacement taxes do not hurt the poor)
and, with political will, social spending and in partic ular that oriented towards the poor, may
be shielded.
Regarding the effect of trade liberalization on vulnerability and income volatility, a recent
World Bank study36 reviewing the experience in three East Asian countries (Indonesia,
Korea, and Thailand) finds that their opening up to trade in the late 1980s and early 1990s
did not have strong negative effects on poverty and vulnerability. The study did indicate that
it remains an open question whether openness made the 1997-98 financial crisis much more
serious than the shocks that had hit the three countries in the 1980s, and suggested that this is
an area that required more analysis. In sum, it is difficult to make any general statements
about how trade reforms affect poverty.
Trade liberalization and inequality
The evidence on the relationship between trade liberalization and income distribution also
produces a complex picture. Simple predictions based on general equilibrium trade models
suggest that since developing countries have a comparative advantage in producing goods
using unskilled labor, globalization should have led to more equity, as the wages of unskilled
workers rose. Economists in the 1980s expected that trade reforms would lead developing
countries to increase exports of goods that use unskilled labor (such as textiles), raising the
returns to unskilled labor and reducing inequality. However, the emerging evidence suggests
that inequality within both rich and poor countries has been increasing. Can this increase in
inequality be attributed to globalization?
The short answer is that there is still a lack of a comprehensive understanding of the forces
behind the global increase in inequality. One popular hypothesis is that technological
change—which may or may not be associated with opening up to trade—has led employers
to demand more skilled labor. This phenomenon, referred to as skill-biased technical change,
has occurred in both developed and developing countries. Some economists argue that the
demand for more skilled workers is unrelated to trade liberalization, since this same trend has
been documented in services that are not traded on world markets. Others, however, argue
that skill-biased technical change is itself an outcome of globalization.
One reason why trade reforms may be associated with increasing inequality is that many
countries have traditionally protected the sectors that use unskilled labor. In Colombia,
Mexico, Morocco, and Poland, for example, protection prior to trade reforms was higher in
sectors that used more unskilled labor (such as textiles and apparel). Another explanation for
why trade reform may be associated with increasing inequality is that exporters—who benefit
from trade reforms—need to hire skilled workers in order to succeed in world markets. A
number of studies have shown that exporters are more likely to use a high proportion of
skilled workers. This means that as countries turn to exporting, the demand for skilled
workers will rise, increasing their wages relative to unskilled workers. Foreign firms in
36
World Bank (2003a).
178
developing countries also tend to hire more skilled workers relative to domestic firms. In
Mexico, increasing inequality is most evident in the border region—the region most affected
by increasing trade with the United States.
Nevertheless, the evidence on trade liberalization and wage inequality remains inconclusive.
In Argentina, Brazil, Costa Rica, the Dominican Republic, and Mexico, the industries that are
most exposed to international competition pay the highest wages. It is difficult to separate the
impact of globalization from technical change, since the adoption of new technologies could
be stimulated by external competition via trade. As an example, in Mexico, the tripling of
manufactured exports during the 1990s has been associated with increased rates of adoption
of modern production technologies, an acceleration of productivity growth, a relative
increase in the demand for skilled labor, and an increase in inequality.
There is no evidence that liberalization permanently worsens income distribut ion. There is
evidence that it has been associated with—at times significant and prolonged—adjustment
costs in the form of employment losses. In several instances, however, such adjustment costs
were exacerbated by other factors such as an appreciation of the real exchange rate that had
accompanied or followed trade liberalization (as in some Latin American countries). In the
case of Mexico, trade integration through NAFTA, while reducing poverty, has also
increased income inequality between regions due to their different initial conditions—regions
with lower per capita GDP and higher telephone density grew faster, while regions with high
public employment grew more slowly (Perry 2003).
To summarize, it is likely that trade reform in the 1990s was accompanied by falling poverty
and rising inequality. There is no systematic evidence that social expenditures for the poor
suffered as a result of revenue losses from trade reforms. Despite employment gains over the
longer run, some individuals were hurt by trade reforms in the short run. Governments need
to help the disadvantaged by strengthening social safety nets and by providing education and
training for the unskilled. Yet the administrative and institutional capacity required to design
and successful implement safety nets that are well-targeted and that avoid leakages is
daunting. The latter eludes even industrialized countries (as attested by the continued
protectionism in these countries), suggesting that more innovative approaches to trade
reforms and trade reform assistance packages may be needed.
4.
Different paths to trade reform
There are different ways to pursue trade liberalization. Some countries, such as Hong Kong
and Singapore, achieved spectacular success through unilateral trade reforms. In the 1990s, a
number of successful reformers—including Bangladesh, China, and India—pursued an
approach that focused first on obtaining an export supply response through limited trade
liberalization. Another approach that has produced successful trade integration is to combine
unilateral or multilateral trade reforms with participation in regional trade agreements, as in
the case of Mexico joining the North America Free Trade Agreement (NAFTA), and the
Central and Eastern European countries participating in the Europe Agreements with the
European Union (EU) member countries.
179
Conversely, there are countries such as Malawi and Zambia that pursued across-the-board
rapid liberalization but, due to a combination of poor macroeconomic management and poor
management of trade policies, experienced resounding failure. Other countries adopted yet
another approach. Korea and Taiwan, for instance, established a complex set of
administrative measures that produced highly export-oriented economies; China established
special economic zones, while India pursued its own unique approach to liberalization.
In this section, we document the great diversity of country experience in promoting growth
through trade liberalization. One element is common to almost all of the success stories:
though the approaches were diverse, they all either explicitly or implicitly promoted export
growth. Exporters were provided with incentives to ensure that selling on international
markets was as attractive as domestic sales. This required establishing a regime that offset
the anti-export bias, and in turn required an effectively functioning bureaucracy to implement
the offsetting regulation—like “indirect duty drawbacks” in Korea. Since the institutional
capacity that is required to implement offsetting regulation is frequently absent, classic trade
liberalization—through low, uniform tariffs and the elimination of quantitative restrictions—
a is generally prescribed. At the end of this section, we discuss the unique country
circumstances that made it possible to compensate exporters in countries that did adopt
conventional trade reforms.
China and India
China provides an example of a model of partial trade liberalization, which it pursued
through a dual-track approach. Special economic zones (SEZs)—one of the drivers in
China’s export and growth success—were set up in the 1980s to provide the firms established
within them access to duty-free imports of inputs. Firms outside the SEZs faced a much more
protected trade regime with much higher tariffs on imports—the average tariffs they faced
were 56 percent in 1982, going down only to 44 percent in 1991 and 16 percent by 2000
(Lardy 2002).
China began with partial trade reform through the establishment of SEZs. 37 China began with
the establishment in 1980 of four SEZs in two coastal provinces (Guangdong and Fujian),
which were selected for their geographic location. Their success led to the addition in 1984
of 14 coastal cities (including Shanghai) as “coastal open cities,” which were given authority
similar to that of the SEZs. By 1992, most cities along the Yangtze River and the borders of
China were also granted special privileges as coastal cities, with Shanghai being granted even
more autonomy. These developments, in turn, spurred the establishment of numerous
“development zones” in many inland cities that extended tax benefits and autonomy to
foreign and domestic investment. In many cases, such zones were established without the
approval of the central government. 38 Within two years, 1991 to 1993, foreign direct
37
Information in this paragraph is from Qian (2000).
The autonomy given to local governments in China is a very important factor in this development. This
autonomy is provided in the form of the “fiscal contracting system” introduced between 1980 and 1993, under
which provincial governments are provided incentives to build up local economies and their own revenue bases.
Specifically, the incentives arise from allowing the provinces to keep the lion’s share of the increases in
revenues at the margin—data from the reform period of 1982-91 show that the correlation coefficient between
the provincial budgetary revenue and expenditure is 0.75, compared to 0.17 in the pre -reform period of 1970-79
38
180
investment into China rose nearly eight times, from US$ 4.4 billion to US$ 28 billion (from
1.2 to 6.4 percent of GDP). In 1993 China became the second largest destination for FDI,
next to the US.
The SEZs enjoy lower tax rates, and are granted greater authority in approving foreign
investment projects than other regions. Within the SEZs, foreign firms are allowed to invest,
and directly import and export. Duties on imported equipment and materials are low for
exporting firms. The removal of these administrative barriers had nearly as great an effect in
spurring trade as the tariff reductions, which did not really begin until the 1990s. Exports
grew at an annual average of 15 percent in the decade of the 1980s, and 19 percent in the
decade of the 1990s. 39
India provides another model of partial liberalization. Unlike China, where liberalization was
initiated only in the SEZs, India liberalized trade across the economy, although it liberalized
one sector at a time. India launched a coherent trade reform program in 1991, after piecemeal efforts at liberalizing trade during the 1980s. This program has continued to date, with
some faltering during 1997-2001.40 The reforms entailed concurrent reductions of what had
been among the highest non-tariff barriers (NTBs) and tariffs in the world. A large reduction
in NTBs and the streamlining of a very complex import licensing regime came early in the
reform program, while tariffs were reduced in a phased manner, with reductions still
continuing to date. Currently, the maximum customs tariff for non-agricultural goods is 30
percent and is scheduled to be reduced to 20 percent in 2004. 41 Capital and intermediate
goods imports were liberalized first, and consumer goods (which were effectively banned)
not until several years later. It was not until 2001 that all consumer goods imports were
liberalized. 42
This sequencing of trade liberalization, which entailed earlier liberalization of capital and
intermediate goods and much steeper reduction in tariffs for some of them, was intended to
deter the deferment of investments that might occur if domestic producers expected further
reduction in capital goods tariffs. 43 It had the following results: the liberalization of capital
and intermediate goods contributed to a rapid supply (export) response: exports grew 20
percent in dollar terms within three years of the start of the reform program. The strong
export supply response provided impetus for a continued response, not least because the
export receipts that were generated alleviated the pressures on the balance of payments.
(Qian 2002). Another study (Jin, Qian, and Weingast 2001) found that such incentives were indeed
significant—both for the growth of employment of non-state enterprises as well as in the reform of state
enterprises.
39
Qian (2002); Jin, Qian, and Weingast (2001).
40
A result of the increasing import competition from East and Southeast Asian countries that devalued their
currencies in the aftermath of the Asian financial crisis.
41
These tariffs underestimate true import competition since there are also specific tariffs.
42
Although imports of several agricultural goods, making up 40 percent of Indian agricultural GDP, continue to
be controlled by state trading enterprises.
43
World Bank (1994).
181
A role for regional agreements?
Some countries have managed to achieve strong integration and growth outcomes by
adopting unilateral or multilateral trade reforms combined with participation in regional trade
agreements. In the 1990s, the central and eastern European countries signed Europe
Agreements with the European Union, and Mexico joined the North America Free Trade
Agreement (NAFTA). Signing on to RTAs provides these countries with access to the
markets of their fellow members, and can also help improve the quality of domestic
institutions. In the case of the central and eastern European countries, World Bank (2000a)
shows that the institutional harmonization aspect of the Europe Agreements has been very
important for the successful trade integration and growth outcomes of these countries. In
particular, the “deeper integration” aspects of the agreements entailing harmonization of
investment policies, regulatory rules and institutions with members of the EU has encouraged
export-oriented foreign direct investment into these countries. In Mexico, a recent study44
shows that NAFTA has had positive effects on trade, FDI, technological transfer, and
growth, and is also associated with improvements in manufacturing total factor productivity.
However, the successful outcomes in the central and eastern European countries and Mexico
are thus far the exceptions rather than the rule : evidence suggests that as many as half of
regional trade agreements are substantially trade-diverting. There are also significant
economic losses to the economies excluded from North-South trade agreements (which is the
vast majority of the South), via trade and investment diversion.
Evidence suggests that for developing countries, signing on to regional trade agreements with
developed countries, particularly large developed countries, is most useful. It has certainly
been beneficial for the central and eastern European countries and Mexico. 45 But country
case studies also suggest that signing on to RTAs (even with developed countries) will not
generate positive export and growth responses unless the countries themselves also pursue
other necessary economic, political, and social reforms. The experience of the EU accession
countries in the 1990s shows that simply signing on to the Europe Agreements provided no
guarantee that there would be benefits. Benefits only accrued to those countries that were
also undertaking the necessary economic, political, and institutional reforms to transform
their economies into market-based ones. 46 This was amply demonstrated in the case of
Bulgaria and Romania, both of which signed Europe Agreements in 1993, in advance of
several other accession countries, but have lagged behind in the transition process and fared
much worse in economic performance compared to some of the latecomers (in particular
Estonia and Slovenia, which signed such agreements in 1995 and 1996, respectively).
Further, a study on NAFTA 47 found that although the agreement has contributed to
institutional harmonization between Mexico and the US in the areas covered by the
44
Lessons of NAFTA, World Bank (2003).
World Bank (2000c).
46
Much of the benefits came in the form of export-oriented FDI from the EU member countries. World Bank
(2000a).
47
Perry et al. (2003).
45
182
agreement—in particular intellectual property rights, investor protection, and environmental
standards—it did not help narrow the institutional gaps outside of these areas that are
important for income convergence between the two countries, especially in the areas of rule
of law and corruption.
Most importantly, regional trade agreements can divert attention away from the multilateral
WTO process, and result in higher costs than benefits for the developing countries. 48 Recent
experience with the Free Trade Area of the Americas, the Central American Free Trade
Agreement, and the US-Australia Free Trade Agreements suggests that regionalism is not
likely to do much to address the key market access priorities for developing countries: tradedistorting agricultural support in the North, contingent protection, and liberalization of
temporary movement of service suppliers. Furthermore, the high costs of negotiating such
agreements also divert resources away from such larger multilateral issues.
Political economy of trade reforms
The success of trade reforms, whether via the complex administrative type of the East Asian
tigers, or the special economic zone or export processing zone (EPZ) type adopted by China
and Mauritius, or the approach adopted by Bangladesh and India, is not automatic, and there
are certainly countries where such strategies have failed. The failure of EPZs, for instance,
has been the norm rather than the exception, particularly in Sub-Saharan Africa but also
elsewhere. 49
A factor that was clearly important for the trade reforms adopted by China and India was the
credibility of reforms, both in terms of their not being reversed and also in their being timebound. Another factor was strong institutions. In the case of Mauritius, democracy and strong
participatory institutions have been suggested as important reasons for success. 50
Political economy considerations need to be taken into account when designing a trade
reform program in order to ensure the sustainability of reforms. This is of great importance in
stimulating private investment. If the expectation is that the reforms will be reversed, the
private sector will not invest. There is also a vicious circle here, as a positive supply response
would in turn garner social and political support to maintain, as well as to further,
liberalization. The key elements on the political economy front are (1) to ensure that the costs
of adjustment arising from reforms are eased; and (2) that reforms are credible.
Easing the costs of adjustment
Easing the costs of adjustment is clearly important to generate social and political support for
reforms. Ensuring that safety nets are adequate to compensate losers is one way of easing
48
Discussion of this point is taken from World Bank (2004); see also Stiglitz (2004).
The pursuit of a strategy that relies on compensating exporters for import tariffs—depending on how it is
implemented—could contravene WTO rules where export subsidies are involved. One suggestion by
Williamson (2004) is for the WTO to allow temporary protection in developing countries for industries they
wish to nurture so long as there is a pre-specified timetable for the removal of such protection, with the
discipline taking the form of WTO sanctions when countries do not abide by the timetable
50
Subramanian and Roy (2001).
49
183
adjustment costs. However, as discussed earlier, a more efficacious way is to design a reform
program that minimizes adjustment costs. China and Mauritius provide good examples in this
regard. China minimized political opposition to the reforms ex ante—not many vested
interests opposed the SEZs because these were set up outside the scope of central planning
and did not disrupt planned production and allocation. At the same time, the approach also
maximized political support for maintaining the reforms ex post as the number of winners
grew over time.
In Mauritius, too, new profit opportunities were created at the margin while leaving old
opportunities undisturbed. The upshot was that there were no identifiable losers. Mauritius
partially liberalized trade by establishing export processing zones (EPZs) and segment ing the
labor market (Subramanian and Roy 2001). Labor market rules were much less stringent in
the EPZs than elsewhere in the economy. Until the mid- to late-1980s, employers had greater
flexibility in discharging workers in the EPZ sector. This approach worked because
allowances had to be paid before dismissing workers, and advance notification of
retrenchment to a statutory body was not required, and the conditions of overtime work were
more flexible. In the 1980s, EPZ wages were about 36-40 percent lower than wages in the
rest of the economy, with the differentials narrowing to 7-20 percent in the 1990s. Aside
from acting as a subsidy to exports, the segmentation of the labor market also prevented the
expansion of the EPZs from driving up wages in the rest of the economy and disadvantaging
the import-substituting industries.
Ensuring credibility
Reform credibility is an area that could benefit from much more understanding and analysis,
not just for trade reforms but for reforms in general. The following discussion, based on
country examples, highlights some ways in which credibility of trade reforms can be
achieved.
In the first instance—and at the very least—reforms should be publicly communicated so that
economic agents are aware of them and respond accordingly. In a well-known example of a
“failed” case of reforms, the lifting of export restrictions on cashew nuts in Mozambique,
there was very little communication to those directly affected by the reforms—the cashew
nut farmers—about the reforms (McMillan, Rodrik, and Welch 2002). Few farmers were
aware that substantial reforms were going on, and this diminished the benefits they received
from the reforms since much of the price increase that resulted from the reforms went to the
traders. This also meant that the supply response was stymied. If the farmers had been made
aware of the reforms, they could have strengthened their bargaining power vis-à-vis the
traders, making it difficult for the latter to pay low prices. Finally, public communication of
the reforms would also diminish the possibility of reform reversals, thereby boosting their
credibility.
Credibility could also be promoted by undertaking measures that are less easy to reverse than
price changes. In the Mozambique nut case, the supply response was poor not just because
farmers did not benefit from much of the price increase as mentioned, but also because
entrepreneurs in the cashew nut processing industry did not make investments to improve
184
efficiency (which themselves could also enhance the irreversibility of the reforms) despite
the higher farm gate prices, probably because they expected the reforms to be reversed. Had
the price reforms been accompanied by non-price reforms, the credibility of the overall
reform program would have been strengthened. Non-price reforms that would have helped
promote the supply response in the Mozambique case could have come as various forms of
government interventions. Government investment in transport, for instance, would help
make the marketing sector more competitive and cashew farming more profitable. The
government could also intervene to provide traders and farmers with access to credit—in the
case of the former to promote competition in cashew marketing and in the latter to encourage
adoption of improved technologies for cashew growing. Finally, such non-price interventions
by themselves would also signal to the public the government’s commitment to the reforms
and hence strengthen the ir credibility.
Yet other ways of promoting credibility include the establishment of institutions such as
India’s Tariff Commission, which is charged with the design and implementation of the trade
reform program and has a tenure that outlasts governments. Such long tenure helps enhance
the credibility of reforms, as it diminishes private sector expectations that the reform program
will be reversed by successive governments. Finally, credibility could also be achieved
through signing on to regional trade arrangements that lock in reforms.
5.
Issues of differential market access
The 1990s saw the continuation of trade liberalization around the world that had begun a
decade before. Average unweighted tariffs were cut by half in both developed and
developing countries, from around 8 to 4 percent in the former, and from around 26 to 13
percent in the latter (Table 5.5). There was also a substantial reduction in non-tariff barriers
and significant progress in moving toward market-based foreign exchange regimes. The tariff
reductions in the 1990s left tariffs higher on agriculture than manufacturing goods in both
developed and developing countries, and tariff escalation in both types of goods and both
groups of countries (Table 5.6).
The relatively low average tariffs of developed countries, however, mask the sometimes high
protection in these countries in the form of tariff peaks, tariff escalation, specific duties, and
production subsidies, with such protection being much more pronounced in agriculture than
in manufacturing. 51 The large numbers of the poor who live in rural areas and work in
agricultural production in developing countries mean that such protectionism exacerbates
poverty in the world. The issue of protectionism in agriculture, in particular in cotton, has
risen in prominence in multilateral trade talks, to the extent that it was one of the main
reasons for the failure of the most recent round of World Trade Organization (WTO) talks in
Cancun in September 2003. Since then, Brazil has gone to the WTO with charges that the US
subsidies on cotton52 are inconsistent with WTO obligations, and the WTO ruling on April
2004 affirmed Brazil’s charges.
51
52
World Bank (2003b).
US subsidies on cotton amounted to US$ 3.7 billion in 2002 (three times US aid to Africa).
185
Table 5.5: Average unweighted tariffs , by country group, 1980-2000
(percent)
1980-85
1988-90
1993-95
1998-00
High
Developing South
Income
Asia
Africa
8.0
7.7
6.8
4.0
30.0
24.8
21.6
16.3
33.0
25.5
19.0
13.2
65.0
64.4
36.1
22.9
Latin
America &
Caribbean
31.0
23.3
13.2
10.7
E.Asia
Middle East &
North Africa
E.Europe &
Central Asia
29.0
16.6
14.4
9.6
28.0
17.1
18.2
15.7
15.0
12.8
8.9
9.8
Note: These are statutory tariffs (most-favored-nation tariffs), which reflect countries’ trade policy stance.
Actual tariffs are lower because of preferences and various exemptions.
Source: WTO CD ROM 2000 and WTO Trade Policy Review, various issues, 1995-2001.
Table 5.6: Average unweighted tariffs , by product group
(percent)
Developin
g
countries
Developed
countries
Agricultural product
Average 1st
Semistage
processe
d
17.4
16.9
21.5
Fully
processe
d
25.4
5.5
11.0
4.6
8.3
Industrial product
Average 1st
stage
12.7
9.9
Semiprocesse
d
10.9
3.3
2.9
3.4
Fully
processed
4.8
15.2
Note: The average agricultural tariffs for developed countries do not include the ad valorem equivalent of
specific tariffs; if the latter we re included, the figures would be much higher.
Source: WTO CD ROM 2000 and WTO Trade Policy Review, various issues, 1995-2001.
Today, trade in manufactures as well as in agriculture is still impeded. Although tariffs on
manufacturing in rich countries are on average lower than in developing countries, the types
of goods exported by poor countries face higher tariffs in the rich countries (Table 5.7). For
example, exporters of manufactures from industrial countries face, on average, a tariff of 1
percent on their sales to other industrial countries, while exporters from developing countries
pay anywhere from 2 percent if they are from Latin America (where the North America Free
Trade Agreement weighs heavily) to 8 percent if they are from South Asia. In agriculture, the
industrial countries impose an average 15 percent tariff on imports from other industrial
countries, whereas the rates on imports from developing countries range from 20 percent
(Latin America) to 35 percent (Europe and Central Asia).
Overall, rich countries collect from developing countries about twice the tariff revenues per
dollar of imports that they collect from other rich countries. Protection also takes forms other
than tariffs—among them quotas, specific duties, and contingent protection measures such as
antidumping duties. As with tariffs, these measures tend to be used more frequently against
labor- intensive products from developing countries. The quota arrangements in the WTO
Agreement on Textiles and Clothing still shackle the exports of many poor countries, and
while these are scheduled to be removed [at the start of 2005], rich countries to date have
freed up only 15 percent of trade, obliging them to implement major changes at the end of the
186
phase- in period. Average antidumping duties are seven to ten times higher than tariffs in
industrial countries, and around five times higher in developing countries. Today’s protection
remains heavily concentrated in the most politically sensitive areas—textiles, clothing, other
labor- intensive manufactures, and agriculture—in both rich and poor countries.
Table 5.7: Rich countries levy higher tariffs on poor countries’ exports
(1997 protection rates facing exporters in each region, in percent)
Importing Region
Exporting Region
East.
Asia
Europe
and
Central
Asia
Latin
Americ
a
Middle
East
South
Asia
SubSahara
n Africa
Industri
al
33.3
31.0
24.2
42.1
23.0
16.6
26.7
43.7
30.3
36.4
36.0
43.4
34.6
20.3
20.1
15.5
23.8
14.8
14.9
13.7
14.4
65.4
45.3
55.3
50.3
76.4
41.1
39.1
16.4
38.4
34.2
29.7
31.8
27.7
30.9
24.0
19.0
12.7
24.7
18.9
11.0
33.6
15.3
30.5
35.1
20.4
23.4
25.8
23.6
East.
Asia
Europe
and
Central
Asia
Latin
America
Middle
East
South
Asia
SubSaharan
Africa
Industria
l
7.4
8.2
6.4
4.3
5.4
7.1
4.4
9.6
13.8
6.4
6.7
11.5
11.0
6.1
8.5
15.1
11.4
15.4
8.8
13.6
11.7
10.4
12.2
8.6
8.9
11.4
10.2
6.1
25.2
28.1
25.8
19.4
33.6
19.0
27.6
12.2
14.5
12.8
11.9
11.7
17.4
20.6
1.0
5.1
5.9
2.1
6.0
8.1
4.2
Agriculture
Industrial
East Asia
Europe and Central Asia
Latin America and
Caribbean
Middle East
South Asia
Sub-Saharan Africa
Nonagriculture
Industrial
East Asia
Europe and Central Asia
Latin America and
Caribbean
Middle East
South Asia
Sub-Saharan Africa
Source: Weighted averages calculated using GTAP Version 5 Database (www.gtap.org). Most-favored-nation
rate except for major free-trade blocs such as the European Union and the North America Free Trade Area.
The differential protection by industrialized countries against developing country exports has
been recognized for some time (WDR 1991). Even in after large, unilateral trade reforms by
developing countries in the 1990s, differential treatment by industrialized countries still
constrains the expansion of trade by developing countries, particularly the poorest. China has
been able to successfully negotiate rapid export growth in a trading system that restricts free
exchange of many agricultural and labor- intensive goods, but many other countries have not
achieved the same success. And even high-export growth countries such as China as well as
less spectacularly successful developing country exporters would benefit from further
liberalization of developed country markets. To continue the momentum towards greater
global integration, high- income countries must further open their markets to developing
187
country exports. Industrial countries levy higher tariffs on imports from developing countries
than from other rich countries—a fact that must be addressed in the upcoming Doha round of
negotiations. In addition, developing countries have been hampered in critical areas,
including access for their agricultural and textile exports, and restrictions on temporary and
permanent movements of their labor force.
References
Acemoglu, Daron, Simon Johnson, and James A. Robinson (2001),“The Colonial Origins of
Comparative Development: An Empirical Investigation,” American Economic Review (91) 5.
Ahmed, Belal (2001), “The Impact of Globalization on the Caribbean Sugar and Banana Industries,”
The Society for Caribbean Studies Annual Conference Papers, Vol. 2.
Aitken, Brian J., and Ann E. Harrison (1999), “Do Domestic Firms Benefit from Direct Foreign
Investment? Evidence from Venezuela”, American Economic Review 89.
Alfaro, Laura, Areendam Chanda, Sebnem Kalemli-Ozcan, and Selin Sayek (2000), “FDI and
Economic Growth: The Role Of Local Financial Markets,” Working Paper 01-083, Harvard
Business School.
Alcala and Ciccone (2004), Quarterly Journal of Economics.
Allen, Donald S., and Michelle T. Gyles (1995), “Trade, Growth and Capital: A Case Study of
Jamaica.” Working Paper 1995-012B. Federal Reserve Bank of St. Louis.
Anderson, J., and D. Marcouiller (1999), “Trade, Insecurity, and Home Bias.” NBER Working Paper
7000. Cambridge, MA: National Bureau of Economic Research.
Baier, Scott L., and Jeffrey H. Bergstrand (2001), “The Growth of World Trade: Tariffs, Transport
Cost, and Income Similarity,” Journal of International Economics 53.
Balasubramanyam, V.N., M. Salisu, and D. Dapsoford (1996), “Foreign Direct Investment and
Growth in EP and IS Countries”, Economic Journal 106.
Baldwin, E. Robert (2003), “Openness and Growth: What's the Empirical Relationship?” NBER
Working Paper No. 9578. Cambridge, MA: National Bureau of Economic Research. Forthcoming
(2004) in Robert E. Baldwin and L. Alan Winters, eds., Challenges to Globalization: Analyzing
the Economics. Chicago, IL: University of Chicago Press.
Berezlin, Peter, Ali Salehizadeh, and Elcior Santana (2002), “The Challenge of Diversification in the
Caribbean,” IMF Working Paper No. 02/196. Washington DC: International Monetary Fund.
Berg, Andrew and Anne Krueger (2003) “ Trade, Growth and Poverty: A Selective Survey,” IMF
Working Paper No. 03/30. Washington DC: International Monetary Fund.
Bhagwati, Jagdish N. (1978), Foreign Trade Regimes and Economic Development: Anatomy and
Consequences of Exchange Control Regimes. Cambridge, MA: Ballinger Press.
Bhagwati, Jagdish N., and T. N. Srinivasan (2002), “Trade and Poverty in the Poor Countries,”
American Economic Review, 92 (2): 180-183.
Bhagwati, Jagdish N. (2002), “Globalization and Appropriate Governance.” 2000 WIDER Annual
Lecture. World Institute for Development Economics Research, the United Nations University,
Helsinki.
Bhattasali, Deepak, Li Shantong, and Will Martin (2002), “Impacts and Policy Implications of WTO
Accession for China”. Mimeo. World Bank.
188
Blomstrom, M. (1986), “Foreign Direct Investment and Productive Efficiency: the Case of Mexico”,
Journal of Industrial Economics 15.
Blomstrom, Magnus, Robert Lipsey, and Mario Zejan (1994), “What Explains the Growth of
Developing Countries?” In William J. Baumol, Richard R. Nelson, and Edward N. Wolff, editors,
Convergence of Productivity: Cros-national Studies and Historical Evidence. Oxford and New
York: Oxford University Press.
Bloom, David E., Ajay Mahal , Damien King, Aldrie Henry-Lee, and Philip Castillo (2001),
“Globalization, Liberalization, and Sustainable Human Development: Progress and Challenges in
Jamaica.” UNCTAD/UNDP Global Programme Occasional Paper.
Bineswaree, Bolaky, and Caroline Freund (2004), “Trade Regulations, and Growth” Working
Paper. March. [need institution]
Borensztein, Eduardo, Jose De Gregorio, and Jong-Wha Lee (1998), “How Does Foreign Direct
Investment Affect Economic Growth?” Journal of International Economics 45, June.
Bosworth, Barry, and Susan M. Collins (1999), “Capital Flows, Investment, and Growth”. Tokyo
Club Papers, Vol. 12.
Carkovic, Maria , and Ross Levine (2003), “Does Foreign Direct Investment Accelerate Economic
Growth?” Mimeo (October). University of Minnesota.
Chuang, Yih-Chyi, and Chi-Mei Lin (1999), “Foreign Direct Investment, R&D, and Spillover
Efficiency: Evidence from Taiwan’s Manufacturing Firms”, Journal of Development Studies 35
(4).
Clark, Ximena, David Dollar and Alejandro Micco (2002, “Maritime Transport Costs and Port
Efficiency” Working Paper No. 2781, the World Bank, February.
Clerides, Sofronis K., Saul Lach, and James R. Tybout (1998), “Is Learning by Exporting Important?
Micro-dynamic Evidence from Colombia, Mexico, and Morocco,” Quarterly Journal of
Economics 113 (3): 903-47.
Collins, Susan, and Robert Z. Lawrence, eds. (2000), “Brookings Trade Forum 1999.” Washington,
DC: Brookings Institution Press.
De Ferranti, David, Guillermo E. Perry, Daniel Lederman, and William Maloney (2001), “From
Natural Resources to the Knowledge Economy: Trade and Job Quality”. Washington DC: World
Bank.
Dollar, David, and Aart Kraay (2001), “ Trade, Growth, and Poverty.” Mimeo (June). Washington
DC: World Bank.
Dollar, David and Aart Kraay (2003), “Institutions, Trade, and Growth: Revisiting the Evidence,”
Working Paper No. 3004. Washington DC: World Bank.
Easterly, William (2003), “National Policies and Economic Growth: A Reappraisal.” CGD Working
Paper No. 27, May. Washington DC: Center for Global Development.
Feenstra, Robert C., Dorsati Madani, Tzu-Han Yang, and Chi-Yuan Liang (1999), “Testing
Endogenous Growth in South Korea and Taiwan,” Journal of Development Economics 60: 317341.
Feenstra, Robert, and Hiau Looi Kee (2003), “On the Measurement of Product Variety in Trade.”
Mimeo. University of California and World Bank.
189
Bolaky, Bineswaree, and Caroline Freund (2004), “Trade, Regulations, and Growth.” Mimeo.
University of Maryland and World Bank.
Funke, Michael, and Ralf Ruhwedel (2001), “Export Variety and Export Performance: Empirical
Evidence from East Asia,” Journal of Asian Economics 12: 493-505.
Ghirmay, Teame, Richard Grabowski, and Subhash C. Sharma (2001), ”Exports, Investment,
Efficiency, and Economic Growth in LDCs: an Empirical Investigation,” Journal of Applied
Economics 33: 689-700.
Glenday, G. (2000), “Trade Liberalization and Customs Revenues: Does Trade Liberalization Lead to
Lower Customs Revenues? The Case of Kenya.” African Economic Policy Discussion Paper No.
44. John F. Kennedy School of Government, Harvard University.
Penny Goldberg and Nina Pavcnik (2004), “Trade, Inequality and Poverty: What Do We Know?
Evidence from Recent Trade Liberalization Episodes in Developing Countries”, NBER Working
Paper No. 10593. Forthcoming in Susan Collins and Carol Graham, eds., Brookings Trade
Forum. Washington, DC: Brookings Institution Press.
Haddad, Mona, and Ann Harrison (1993), “Are there Positive Spillovers from Direct Foreign
Investment?” Journal of Development Economics 42.
Hausmann, Ricardo, and Dani Rodrik (2002),“Economic Development as Self-Discovery.” National
Bureau of Economic Research Working Paper No. 8952. Cambridge, MA: National Bureau of
Economic Research.
Hoekman, Bernard, Constantine Michalopoulos, Maurice Schiff, and David Tarr (2001), “Trade
Policy Reform and Poverty Alleviation.” Developed as part of the World Bank’s Poverty
Reduction Strategy Sourcebook (December). Washington, DC: World Bank.
Heron, Tony (2002), “Export Processing Zones and Policy Competition for Foreign Direct
Investment: The Offshore Caribbean Development Model,” July.
Huang, Yasheng, and Tarun Khanna (2003), “Can India Overtake China?” Foreign Policy, July.
International Monetary Fund (IMF) (2001), “Jamaica: Staff Report for the 2001 Article IV
Consultation”. Washington, DC: IMF.
________ (2002), “Jamaica Staff Report for the 2002 Article IV Consultation”. Washington, DC:
IMF.
________ (2003), “Redrafting the Reform Agenda”, Finance and Development, September.
Washington, DC: IMF.
Jaffee, Steven, and Spencer Henson (2004), “Agro-Food Exports from Developing Countries: the
Challenges Posed by Standards”. Prepared as chapter for book edited by Ataman Aksoy,
forthcoming, World Bank.
Jaffee, Steven, and Jeanette Sutherland (2003), “ Integrated Framework Trade Diagnostic Studies:
Findings and Lessons from Agricultural Sector Analysis.” Mimeo. Washington, DC: World
Bank.
Jin, Hehui, Yingyi Qian, and Barry R. Weingast (2001), “Regional Decentralization and Fiscal
Incentives: Federalism, Chinese Style .” Mimeo. Stanford University.
Kanbur, Ravi (2001), “Economic Policy, Distribution, and Poverty: The Nature of Disagreements,”
World Development.
Kaufman, Daniel, Aart Kraay, and Massimo Mastruzzi (2003), “Governance Matters III: Governance
Indicators for 1996-2002”. Working Paper No. 3106. Washington, DC: World Bank.
190
Kaufman, Daniel, Aart Kraay, and Pablo Zoido-Lobaton (2002), “Governance Matters II: Updated
Indicators for 2000/01.” Working Paper No: 2772. Washington, DC: World Bank.
Kee and Hoon (forthcoming, Journal of Development Economics)
Kokko, Ari (1994), “Technology, Market Characteristics, and Spillovers”, Journal of Development
Economics 43.
Kokko, Ari, Ruben Tansini, and Mario C. Zejan (1996), “Local Technological Capability and
Productivity Spillovers from FDI in Uruguayan Manufacturing Sector”, Journal of Development
Studies 32.
Krueger, Anne O. (1978), Foreign Trade Regimes and Economic Development: Liberalization
Attempts and Consequences. Lexington, MA: Ballinger Press.
________ (1983), Trade and Employment in Developing Countries. Chicago, IL: University of
Chicago Press.
________ (1997), “Trade Policy and Economic Development: How We Learn,” American Economic
Review 87 (1).
Krugman, Paul (1998), “Firesale FDI”. Working Paper, Massachusetts Institute of Technology.
Lall, S. (1999), The Technological Response to Import Liberalization in Sub-Saharan Africa. London:
Macmillan Press.
Lardy, Nicholas R. (2002), Integrating China into the Global Economy, Washington, DC: Brookings
Institution Press.
Lau, Lawrence J., Yingyi Qian, and Gerard Roland (2000), “Reform without Losers: An
Interpretation of China’s Dual Track Approach to Transition,” Journal of Political Economy 108
(1).
Lee, Ricci and Rigobon (forthcoming, Journal of Development Economics 2004)
Levinsohn, J. (1999), “Employment Responses to International Liberalization in Chile ,” Journal of
International Economics 47:321-344.
Lim, Ewe-Ghee (2001), “Determinants of, and the Relation between, Foreign Direct investment and
Growth: a Summary of Recent Literature”. IMF Working Paper /01/175. Washington DC:
International Monetary Fund.
Marques, Gustavo, and Carmen Pagés (1998), “Trade and Employment: Evidence from Latin
America and the Caribbean,” IDB Working Paper No. 366. Washington DC: Inter-American
Development Bank.
McMillan, Margaret, Dani Rodrik , and Karen Horn Welch (2002), “When Economic Reform Goes
Wrong: Cashews in Mozambique”. Mimeo (July).
Michaely, Michael, Demetris Papageorgiou, and Armeane Choksi (1991), Liberalizing Foreign
Trade: Lessons of Experience in The Developing World . Cambridge, MA: Blackwell.
Mishra, Banamber, and Matiur Rahman (1998), “Exports, Foreign Aid, Exchange Rate and Economic
Growth: Empirical Evidence From 28 Low-Income LDCs” Scandinavian Journal of Development
Alternatives 17: 21-31.
Moran, T. (1998), “Foreign Direct Investment and Development: The New Policy Agenda for
Developing Countries and Economies in Transition.” Washington, DC: Institute for International
Economics.
191
Moreira, Mauricio Mesquita, and Sheila Najberg (2000), “Trade Liberalization in Brazil: Creating or
Exporting Jobs?” Journal of Development Studies (36) 3: 78.
Nash, John, and Wendy Takacs, eds. (1998), Trade Policy Reform: Lessons and Implications.
Washington, DC: World Bank.
Panagariya, Arvind (2003), “Miracles and Debacles: Do Free-trade Skeptics have a Case?”
Economics Working Paper Archive, WUSTL, International Trade Series No. 0308013.
Perry, Guillermo, Daniel Lederman, William Maloney, and Luis Serven (2003), “Lessons from
NAFTA for Latin America and the Caribbean.” Latin America and the Caribbean Region, World
Bank, April.
Porto, Guido (2003), “Using Survey Data to Assess the Distributional Effects of Trade Policy.”
Mimeo. Washington, DC: World Bank.
Prasad, Eswar S., Kenneth Rogoff, Shang-Jin Wei, and M. Ayhan Kose (2003), “Effects of Financial
Globalization on Developing Countries: Some Empirical Evidence.” Occasional Paper No. 220.
Washington, DC: International Monetary Fund.
Pritchett, Lant (1996), “Measuring Outward Orientation in LDCs: Can it be Done?” Journal of
Development Economics 49.
Qian, Yingyi (2000), “The Process of China’s Market Transition 1978-98: The Evolutionary,
Historical and Comparative Perspective.” Stanford University, April.
Qian, Yingyi (2002), “How Reform Worked in China.” Department of Economics, University of
California, Berkeley, August.
Radalet, Steve, Cristina Garcia, and Mumtaz Hussain (2000), “Trade and Exports.” Background paper
examining the state of the Andean Region for the Andean Competitiveness Project, April.
Razin, Assaf, Efraim Sadka, and Chi-wa Yuen (1999), “Excessive FDI under Asymmetric
Information.” NBER Working Paper No. 7400. Cambridge, MA: National Bureau of Economic
Research.
Ravallion, Martin (2001), “Growth, Inequality and Poverty: Looking Beyond Averages,” World
Development 29 (11): 1803-1815.
________ (2003), “The Debate on Globalization, Poverty, and Inequality: Why Measurement
Matters”. Development Research Group Working Paper 3038. Washington DC: World Bank.
Redding, S. , and Venables, A. (2003), “Geography and Export Performance: External Market Access
and Internal Supply Capacity”. CEPR Discussion Paper No. 3807. London: Centre for Economic
Policy Research.
Rigobon and Rodrik (2004)
Rodríguez, Francisco, and Dani Rodrik (2000), “Trade Policy and Economic Growth: A Skeptic ’'s
Guide to the Cross-national Evidence,” Electronic Working Papers No. 9901, University of
Maryland, Department of Economics.
Rodrik, Dani (1992), “The Limits of Trade Policy Reform in Developing Countries,” Journal of
Economic Perspectives (6) 1.
________ (1997a), “TFPG Controversies, Institutio ns, and Economic Performance in East Asia,”
NBER Working Paper No. 5914. Cambridge, MA: Natio nal Bureau of Economic Research.
________ (1997b), “Trade Policy and Economic Performance in Sub-Saharan Africa. Mimeo.
Harvard University.
192
________ (1999), “The New Global Economy and Developing Countries: Making Openness Work.”
Policy Essay No.24. Washington, DC: Overseas Development Council.
________ (1999a), “Where Did All the Growth Go?” Journal of Economic Growth 4, December.
________ (2002), “Trade Policy Reform as Institutional Reform”. In B. M. Hoekman, P. English, and
A. Mattoo. Development, Trade, and the WTO: A Handbook. Washington, DC: World Bank.
________ (2003), “Growth Strategies”, CEPR Discussion Paper No: 4100. London: Centre for
Economic Policy Research.
Rodrik, Dani, editor (2003), In Search of Prosperity: Analytic Narratives on Economic Growth.
Princeton, NJ: Princeton University Press.
Rodrik, Dani, and Ricardo Hausmann (2003), “Discovering El Salvador’s Production Potential.”
Mimeo (September).
Rodrik, Dani, Arvind Subramanian, and Francesco Trebbi (2002), “Institutions Rule: the Primacy of
Institutions over Geography and Integration in Economic Development.” NBER Working Paper
No. 9305. Cambridge, MA: National Bureau of Economic Research.
Shatz, Howard, and Anthony Venables (2000), “The Geography of International Investment.” In
Gordon L. Clark, Meric S. Gertier, and Maryann P. Feldman, editors, The Oxford Handbook of
Economic Geography. Oxford: Oxford University Press.
Stiglitz, Joseph, and The Initiative for Policy Dialogue on behalf of The Commonwealth Secretariat
(2004), “An Agenda for the Development Round of Trade Negotiations in the Aftermath of
Cancun.”
Subramanian, Arvind, and Devesh Roy (2001), “Who Can Explain the Mauritian Miracle:Meade,
Romer, Sachs, or Rodrick?” African Department Working Paper No. 01/116. Washington, DC:
International Monetary Fund.
Summers, Lawrence H. (1999), “Distinguished Lecture on Economics in Government: Reflections on
Managing Global Integration,” Journal of Economic Perspectives 13 (2).
Thomas, Vinod, and John Nash (1991a), Best Practices in Trade Policy Reform. Oxford; New York;
Toronto and Melbourne: Oxford University Press for the World Bank.
________ (1991b), “Reform of Trade Policy: Recent Evidence from Theory
and Practice”, World Bank Research Observer 6 (2): 219-40.
________ (1992), “Trade Policy Reform: Recent Evidence from Theory and
Practice”. In R. Adhikari, C. Kirkpatrick, and J. Weiss, eds., Industrial and Trade Policy Reform
in Developing Countries. Contemporary Issues in Development Studies. Manchester and New
York: Manchester University Press and St. Martin’s Press.
Tsikata, Yvonne (2003), “Integrating into the Global Economy: Challenges, Experiences, and Policy
Lessons from the Integrated Framework”. Mimeo. Washington, DC: World Bank.
Wacziarg, Romain , and Karen Horn Welch (2003), “Trade Liberalization and Growth: New
Evidence.”Stanford University Working Paper. [reference to be updated]
Williamson, John (1990), “What Washington Means by Policy Reform.” Chapter 2 in John
Williamson, Latin American Adjustment: How Much Has Happened? [add publisher, place.]
Updated November 2000.
_________ (2000), “What Should the Bank Think about the Washington Consensus?” Paper prepared
as background to the World Bank’s World Development Report 2000, July.
193
________ (2004), “The Washington Consensus as Policy Prescription for Development.” Paper
prepared for lecture at the World Bank, January.
Winters, Alan L., Neil McCulloch, and Andrew McKay (2004), “Trade Liberalization and Poverty:
the Evidence So Far”, Journal of Economic Literature XLII (March): 72-115.
World Bank (1993), The East Asian Miracle: Economic Growth and Public Policy. New York, NY:
Oxford University Press.
________ (1994), India: Issues in Trade Reform. Washington, DC: World Bank.
________ (1997a), “China Engaged: Integration with the Global Economy.” China 2020 Series,
September. Washington, DC: World Bank.
________ (1997b), Global Development Finance, Washington, DC: World Bank.
________ (2000a, Progress Toward the Unification of Europe, Washington, D.C.
________ 2000b, Securing Our Future in a Global Economy, Latin American and Caribbean Region,
Washington, D.C.
________ (2000c, Trade Blocs, Policy Research Report, Washington, D.C.
________ (2001, Global Development Finance, Washington, D.C.
________ (2002, Globalization, Growth and Poverty: Building an Inclusive World Economy, Policy
Research Report, Washington, D.C.
________ (2003a, East Asia Integrates, Washington, D.C.
________ (2003b, Global Economic Prospects 2004, Washington, D.C.
________ (2003c, Jamaica Country Economic Memorandum, Washington, D.C.
________ (2003d, Malawi Country Economic Memorandum: Policies for Accelerating Growth ,
Washington, D.C., January.
________ (2003e, Malawi Integrated Framework Diagnostic Trade Integration Study, Washington,
D.C.
________ (2003f, Trade Policies in South Asia: An Overview, Volumes I and II (Discussion Draft),
Washington, D.C., May.
________ (2003g, Zambia: The Challenge of Competitiveness and Diversification, Washington,
D.C.
________ (2003h, Zambia Country Economic Memorandum: Policies for Growth, Diversification
and Poverty Reduction, Washington, D.C., September.
________ (2004, “Leveraging trade for development: the World Bank Research Agenda”, progress
report to the Board of Directors, the World Bank.
World Trade Organization (WTO) (1998), “Jamaica Trade Policy Review.” Geneva: WTO.
194