logo logo

New World Bank study on regional and bilateral trade agreements

Global Economic Prospects 2005
Trade, Regionalism, and Development

World Bank
November 2004


agreements (RTAs) is fundamentally altering
the world trade landscape. The
number of agreements in force now surpasses
200, and it has risen sixfold in just two
decades. Today more than one-third of global
trade takes place between countries that have
some form of reciprocal RTA. [1] The European
Union (EU) and United States are playing a
prominent role in this proliferation (figure 1).

This report addresses two questions:

• What are the characteristics of agreements
that strongly promote-or
hinder-development for member countries?

• Does the proliferation of agreements
pose risks to the multilateral trading
system, and how can those risks be

Identifying What Works: Open Regionalism

RTAs are often one component of a larger
political effort to deepen economic relations
with neighboring countries. [2] As such,
they can create opportunities to expand trade
through joint action to overcome institutional
as well as policy barriers to trade. At a basic
level, it is often easier to motivate reciprocal reductions
in border barriers when the participants
are fewer and the policymakers feel more
in control of outcomes. Moreover, RTAs have
the flexibility to pursue trade-expanding
policies not addressed well in multilateral trading
rules. Trade agreements therefore usually
go beyond slashing tariffs to include measures
to reduce trade impediments associated with
standards, customs and border crossings, and
services regulations-as well as broader rules
that improve the overall investment climate.
Finally, these agreements often form cornerstones
of larger economic and political efforts
to increase regional cooperation. RTAs can
help motivate and reinforce broader reforms in
domestic policy; they can be designed to contribute
to a political environment that is more
conducive to stability, investment, and growth.

Not all agreements create new trade and
investment. Those RTAs with high external border
protection are particularly susceptible to the
adverse effects of trade diversion (figure 2). In
fact, a statistical analysis based on findings from
several econometric studies suggests that many
agreements cost the economy more in lost trade
revenues than they earn, because they discriminate
against efficient, low-cost suppliers in nonmember
countries. Of course, this finding does
not take into account the potential dynamic
gains, the positive effects associated with services
liberalization, or any of the benefits from
adopting new regulations. But it does underscore
the point that regional agreements carry
risks that merit close scrutiny by would-be

As agreements proliferate, a single country
often becomes a member of several different
agreements. The average African country
belongs to four different agreements, and the
average Latin America country belongs to
seven agreements. This creates a “spaghetti
bowl” of overlapping arrangements (figure 3).
Each agreement has different rules of origin, different tariff schedules, and different periods
of implementation, and together they complicate
customs administration. Customs agents
report that it takes longer to process goods
covered by preferential arrangements, and
longer processing times drive up the cost of
trade. In general, the longer the delays in customs,
the smaller the role of trade in GDP.

So what characteristics lead to expanded
trade and development? A prerequisite for
the success of any trade policy is that it be
integrated into a sound domestic policy framework.
It is virtually impossible for entrepreneurs
to take advantage of new opportunities-
whether they originate in market access
through an RTA, through a multilateral agreement,
or other sources-if the domestic investment
climate is not supportive. Macroeconomic
stability, basic property rights, and
adequate infrastructure regulation are all key.
Indeed, trade agreements can reinforce positive
elements in the domestic reform program by
anchoring policy to the agreement itself. But
an RTA cannot substitute for sound domestic

With prerequisites in place, the RTAs most
likely to increase national incomes over time
are those designed with:

• Low external MFN tariffs,
• Few sectoral and product exemptions,
• Nonrestrictive rules-of-origin tests that
build toward a framework common to
many agreements,
• Measures to facilitate trade,
• Large ex-post markets,
• Measures to promote new cross-border
competition, particularly in services, and
• Rules governing investment and intellectual
property that are appropriate to the
development context.

Low external tariffs and wide coverage
minimize the risks of trade diversion, while
nonrestrictive rules of origin allow for increased
trade. The practice of excluding many
agricultural products is common, and it can
limit development payoffs. Trade facilitation
measures, though worthwhile in and of themselves,
receive more policymaker attention when they are embedded in an RTA, and they
often have positive trade-creating effects for
all trade partners.

Well designed agreements are of limited
value if they are not implemented, and many
RTAs have more life on paper than in reality.
Weak implementation often afflicts South-
South agreements. Monitoring mechanisms
are often inadequate and do not receive the
sustained high-level political attention necessary
to drive institutional improvements in,
for example, adherence to tariff reduction
schedules, customs, and border crossings.
Against these benchmarks of success, it is
difficult to give universally high marks to any
single category of agreement. In general,
North-South agreements score better on implementation
than South-South agreements.

Because North-South agreements can integrate
economies with distinct technological
capabilities and other different factor proportions,
and because they usually result in larger
post-agreement markets, the potential gains
are usually greater. However, tighter rules of
origin, more restrictive exclusions for particular
sectors (such as agriculture), and a preoccupation
with rules not calibrated to development
priorities can undercut these benefits
(figure 4). North-South agreements, particularly
those with the United States, have been
more effective in locking in new services liberalization;
they have pressed intellectual property
rights beyond World Trade Organization
(WTO) rules; and expanded the sphere of investment
protections; but they contain few
provisions to liberalize the temporary movement
of labor.

Some South-South agreements are better
at focusing on merchandise trade, minimizing
exclusions, adopting less restrictive
rules of origin, and lowering the border costs.
For example, the Caribbean Community
(CARICOM) and the Common Market of
Eastern and Southern Africa (COMESA) have
had some success in reducing border costs. But
in general, South-South agreements have not
adhered to implementation schedules, and
they suffer from their small market size and
economic similarity. And like the North-South
agreements, South-South agreements rarely
provide for the temporary movement of labor.

Consequences for the Multilateral System

The development consequences of RTAs are
not limited to their effects on members-
they also have cumulative effects on the multilateral
system. In one sense, RTAs are a step toward
greater openness in the whole system, by
promoting more trade and generating new domestic
constituencies with an interest in openness.
Moreover, some regional trade policies
are effectively nondiscriminatory, such as measures
to improve customs, speed transactions at
ports or border crossings, or in some cases open
services markets. These measures can complement
unilateral and multilateral policies.

However, this view overlooks the effects
that RTAs can have on excluded countries.
Preferences for some countries mean discrimination
against others. Indeed, the General
Agreement on Tariffs and Trade (GATT),
borne out of the sad experience of discrimination
in the prewar years, was founded on the
principle of nondiscrimination. Today, the
adverse consequences for the excluded countries
are much less severe than at GATT’s
inception, because tariffs and other barriers
have come down sharply, mitigating the exclusionary
effects of regional arrangements.

The exception-and it is not trivial-is agriculture.
Another mitigating factor is that
many countries excluded by trade agreements
between the United States and the EU enjoy
some degree of preferential access through
voluntary preference schemes, such as the
Generalized System of Preferences (GSP),
America’s Growth and Opportunity Act
(AGOA), and the EU’s Everything But Arms
(EBA) program. To be sure, these programs
lack the certainty of market access that MFN
agreements and RTAs provide, because preferences
are voluntary and subject to political
whim, but they do mitigate the effects of exclusions
for selected, very low-income countries.
Finally, some developing countries-the
spokes in the hub-and-spoke analogy-are
signing bilateral agreements with each other
and with other hubs.

Inevitably some countries get left out of
trade agreements, either because they are not
favored politically, because they cannot afford
the costs of many separate negotiations, or because
their neighborhood is less open. Countries
as diverse as Bolivia, India, Mongolia,
Pakistan, and Sri Lanka do not enjoy the same
level of access to the United States or the EU
as Chile, Jordan, or Mexico, and they see their
trade diminished when bilateral agreements
are signed.

RTAs can also undercut the incentives of
governments to press for multilateral liberalization,
which would improve global trade
rules. This study finds little evidence that
major players in the current WTO negotiations
have changed their negotiating positions
or retreated from the multilateral process,
even as they avail themselves of regional trade
deals. However, as the discussions become
politically difficult, the risk is ever present that
even they will abandon multilateralism in
favor of “satisficing regionalism.” One
consequence of the spread of regional agreements
is that many poorer developing countries
have diverted scarce negotiating resources
to regional negotiations at the expense
of more active participation in the Doha discussions.

The average developing country belongs
to five separate RTAs and is negotiating
more all the time. In the future, will countries
that now enjoy preferences fight multilateral
liberalization, or even oppose further regional
liberalization, to keep their privileged market
access? A few small developing countries are
indeed likely to lose advantages in preferential
markets, and they may scuttle a deal if their legitimate
concerns are not addressed.

The Importance of Doha to Open Regionalism

The policy solution to these twin concerns-
the need to design regional agreements
that create trade and regional agreements
that have minimal exclusionary
effects-comes together in the form of low
MFN tariffs and other border barriers. An
agreement that lowers border protection
around the world promotes open regionalism
by mitigating trade diversion. At the same
time, it would diminish the exclusionary effects
of discriminatory preferences built into
regional agreements. The first order of business
for the international community is to accelerate
progress on the Doha Agenda and to
fill in the blanks of the August 2004 framework
agreement with reductions in protection,
especially for products produced by the
world’s poor.

For Developing Countries, a Three-Part Strategy

Developing countries wishing to harness
trade to their development strategy
should see regional integration as one element
in a three-pronged strategy that includes unilateral
liberalization, multilateral liberalization,
and regional liberalization.

Historically, unilateral liberalization, which
is usually linked to a broader program of domestic
reform, has accounted for most of the
reductions in border protection. Most comprehensive
trade reforms among large countries
(Argentina, Brazil, and China in the early
1990s, and more recently, India) were primarily
unilateral reforms that were undertaken to
increase the productivity of the domestic economy.
The same process took place in many
small countries as well. In fact, of the 21 percentage
point cuts in average weighted tariffs
of all developing countries between 1983 and
2003, unilateral reforms account for roughly
two-thirds of the reduction. Tariff reductions
associated with the multilateral commitments
in the Uruguay Round accounted for about
25 percent, and the proliferation of regional
agreements amounted to about 10 percent of
this reduction (see figure 5).

Autonomous liberalization promotes global
competitiveness by lowering costs of inputs, increasing
competition from imports to drive productivity
growth, and integrating the national
economy into the global economy. Autonomous
trade reform is, ironically, more important than
ever in the presence of RTAs; low border barriers
minimize the risks of trade and investment
diversion. Low external barriers promote trade
in world markets, and this is highly correlated
with increases in intraregional trade, irrespective
of the presence of an RTA.

Multilateral liberalization leverages domestic
reforms into increased market access
around the world. Developing countries collectively
stand to gain much more in the WTO
arena than in any smaller regional market.
Moreover, this multilateral forum is the only
place that developing countries, working together,
can press for more open markets in
agriculture and can seek disciplines on tradedistorting
agricultural subsidies and on contingent

Some have argued that RTAs can be an alternative
to multilateral liberalization. They
are not. Gains for all developing countries
from these agreements, even under the most
generous of assumptions, are usually only a
fraction of those from full multilateral liberalization.
Of course, if one of the partner countries
is a high-income, large-market economy,
and if most other countries are excluded from
preferential access, the countries signing the
first trade agreement may benefit individually
and substantially-but those benefits wither
as new countries sign additional agreements.

In fact, the scenarios in this study show that
all developing countries would collectively
lose if they were all to sign preferential agreements
with the Quad (Canada, the EU, Japan,
and the United States) (figure 6). Therefore,
developing countries have a powerful collective
interest in an effective Doha Agenda-
even if they all are scrambling to gain preferential
market access to the Quad.

Forging policies on open regionalism is the
third component of trade policy strategy. Desirable
as multilateral liberalization is, the
Doha Round is likely to realize only part of its
development potential. For some types of policy,
collective regional actions may be the first,
best course, and may result in effective nondiscriminatory
benefits. [3] For example, RTAs can
reduce regional political tensions, take advantage
of scale economies in infrastructure provision,
and lead to joint programs to improve
border crossings or to motivate liberalization
in services. But countries should sign on with their eyes wide open. The lessons of this study
(and others before it [4]) are that, much as with
unilateral or multilateral policies, design and
implementation determine the ultimate effects.
It is important to use trade policy to leverage
domestic reforms that promote growth. For
South-South agreements, it is essential that the
focus be on some combination of full trade
liberalization behind low external border
protection, greater services deregulation and
competition, and proactive trade facilitation
measures that together positively affect both
intra- and extra-regional trade.

High-Income Countries and Development

High-income countries, in order to realize
their broad development objectives, must
intensify their efforts to realize the development
promise of the Doha Agenda. This has
the potential to open up trade, particularly in
agriculture, in a way that would benefit lowincome
groups around the world. Because
the high-income countries are the large
players in the system, they have a special interest
in-and responsibility for-using effective
multilateral reforms to discipline the discretionary
aspects of the regional agreements.

Allowing developing countries to concentrate
scarce negotiating resources on the multilateral
agenda may require that high-income
countries decelerate their efforts at expanding
RTAs. Irrespective of the pace of new agreements,
high-income countries could consider
the following rules of thumb when designing
agreements to promote development. First, reducing
the extensive exclusions for agriculture
would transfer the income gains to rural areas
in participating developing countries. Second,
adopting more common and nonrestrictive
rules of origin across agreements would
reduce the administrative barriers that often
undermine agreements and that increase the
burden on customs administration. Third,
working with prospective partners to ensure
that new regulations regarding investment and
intellectual property are appropriate to the
level of development would reduce risks of
undue enforcement costs. Finally, providing
trade-related technical assistance, not only in
the implementation phase but also in the negotiating
phase, would promote greater liberalization
of services and lower MFN tariffs.

Acting Collectively to Mute the Effects of Discrimination

To minimize the discriminatory effects of
RTAs at the multilateral level, all countries
must assume greater responsibility for maintaining
the multilateral system. The international
community, working through the WTO,
should revisit Article V of its charter. If the
stated disciplines cannot be enforced in the
near term for collective political reasons, then
increasing transparency and information
should become a priority. At present, the WTO
collects little if any information updating specific
provisions, their implementation, and the
trade consequences. It even fails to take advantage
of extant public monitoring efforts in specific regions, which could inform their data
collection effort. Collecting and publishing
specific information on RTAs would allow
members that find themselves excluded to
challenge these agreements in the court of public
opinion. Even the more modest goal of
transparency will require building a new consensus
and providing the staff of the WTO
with more resources than they have currently

Nonetheless, WTO members should consider
enhancing the existing rules to ensure
that regional agreements have positive development
and systemic outcomes. This could include
(based on a modest tightening of current
practice) setting quantitative indicators that
define “substantially all trade.” It could include
efforts to simplify and harmonize the
rules of origin that are applied to both developed
and developing countries. These items
are on the Doha Agenda and may be ready for

Organization of This Study

As is customary, chapter 1 of this study presents
the World Bank’s view of the global
economy. The short-term section analyzes the
main forces shaping the global outlook and the
implications for developing countries; the longterm
analysis focuses on structural changes
in the global economy that will affect poverty
rates and the prospects for attaining the Millennium
Development Goals. A novel feature of
this year’s report is the introduction of a companion
online feature (see www.worldbank.
), where the reader can find additional
information on regional trends and commodity
prices, and tools to design scenarios to
his or her own specifications.

Chapter 2 introduces the issues associated
with regional trade agreements and provides
an overview of regional trading trends.
Subsequent chapters focus on the content and
consequences of regional agreements for trade
creation (chapter 3), trade facilitation
(chapter 4), and services, investment, intellectual
property rights, and labor mobility
(chapter 5). Chapter 6 returns to the issue of
making regional agreements more compatible
with a nondiscriminatory multilateral system.

The full report is available here


[1Negotiated as bilateral or multicountry treaties,
regional trade agreements grant members assured preferential
market access, usually at zero tariffs for eligible
products. Following WTO convention, the term
“regional trade agreement” includes both reciprocal
bilateral free trade or customs areas and multicountry
(plurilateral) agreements. These are distinct from nonreciprocal
voluntary agreements, such as the generalized
system of preferences (GSP). Also, for statistical
purposes, unless otherwise noted, intra-EU trade is excluded
from quantitative trade analysis. The EU is defined
as including the 15 countries that belonged to the
union before its enlargement in 2004.

[2See Devlin and Estevadeordal (2004) and Schiff
and Winters (2003), among others.

[3See Robert Lawrence (1997), who develops the
idea of subsidiarity as applied to regional agreements.

[4See Schiff and Winters (2003).

 source: WB