Vi Digital Library - Text Preview
Costa Rica: Trade Opening, FDI Attraction and Global Production Sharing
Working paper by WTO, 2011
Download original document (English)
Staff Working Paper ERSD-2011-08 Date: 23 May 2011
World Trade Organization
Economic Research and Statistics Division
Century Regionalism: Filling the gap between
century trade and 20
century trade rules
Graduate Institute, Geneva
Manuscript date: April 2011
Disclaimer: This is a working paper, and hence it represents research in progress. This
paper represents the opinions of the authors, and is the product of professional research. It is
not meant to represent the position or opinions of the WTO or its Members, nor the official
position of any staff members. Any errors are the fault of the author. Copies of working
papers can be requested from the divisional secretariat by writing to: Economic Research and
Statistics Division, World Trade Organization, Rue de Lausanne 154, CH 1211 Geneva 21,
Switzerland. Please request papers by number and title.
- 2 -
Century Regionalism: Filling the gap between
century trade and 20
century trade rules
Graduate Institute, Geneva
November 2010; this version April 2011
This paper weaves several sets of facts into an argument that: 1) today’s
trade is radically more complex, involving a “trade-investment-service
nexus”, 2) this 21st century trade demanded deeper disciplines which were
supplied by “21st century regionalism” while the WTO was otherwise
occupied, and 3) 21
century regionalism has quite different implications
for the world trading system than the traditional thinking suggests. The
paper also argues that the traditional thinking (building-stumbling-block
and Vinerian economics) is not up to the job of analysing 21
regionalism. An alternative framework is not provided, but elements a new
approach should encompass are discussed.
Keywords: Regionalism, multilateralism, WTO
JEL Classification: F13, F15
The paper was first presented at the WTO‟s “Workshop on PTAs and the WTO: A new era” held at the WTO
4 November 2010; http://www.wto.org/english/res_e/reser_e/wkshop_nov10_e/wkshop_nov10_e.htm. My
thanks to comments from the participants as well as seminar participants at Keio University and RIETI in
Tokyo, Georgetown University, the Indian Institute for Foreign Trade in Delhi, and the Taiwan WTO Center
(Chung-Hua Institution for Economic Research), and comments from Theresa Carpenter, Simon Evenett,
Patrick Low, Ted Moran, and Rod Ludema. Special thanks to Alan Maldic, Yose Damuri, Andy Lendle for help
with data analysis.
- 3 -
The last time multilateral trade rules were updated, Bill Clinton was in his first term of office,
data was shared by airmailing 1.4 megabyte HD floppy disks (few people had email), cell
phones looked like bricks and calling costs were measured in dollars per minute. Trade
mostly meant selling goods made in a factory in one nation to a customer in another. Simple
trade needed simple rules – a fact reflected in both multilateral and regional trade agreements.
Today‟s trade is radically more complex. The ICT revolution fostered an internationalisation
of supply chains, and this in turned created the “trade-investment-service nexus” at the heart
of so much of today‟s international commerce. Complex trade needs complex rules. As the
WTO was otherwise occupied, the incipient governance gap was filled by uncoordinated
developments elsewhere – primarily in deep regional trade agreements, bilateral investment
treaties, and autonomous reforms in emerging economies. The resulting package of deeper
disciplines – what could be called “21st century regionalism” – requires new thinking.
This paper marshals several sets of facts into an argument that:
Today regionalism is qualitatively different to that of the 1990s;
The traditional building-stumbling-block approach and Vinerian economics on which
it is premised are not up to the job of analysing this new regionalism; and
21st century regionalism has quite different ramifications for the world trading system
century regionalism did.
The facts are not new, having been documented by many observers. The aim of the paper is
to weave them together into an argument that points to a new way of thinking about
regionalism – specifically, its economic implications, its political economy determinants, and
its impact on the world trade system and the WTO.
In a nutshell, 21
century regionalism is not primarily about preferential market access as was
the case for 20
century regionalism; it is about disciplines that underpin the trade-
investment-service nexus. This means that 21
century regionalism is driven by a different set
of political economy forces; the basic bargain is “foreign factories for domestic reforms” –
not “exchange of market access”. As 21st century regionalism is largely about regulation
rather than tariffs, regulatory economics is needed rather than Vinerian tax economics.
century regionalism is a serious threat to the WTO‟s centrality in global trade
governance, but not for the reason suggested by the old building-stumbling-block thinking.
century regionalism is a threat to the WTO‟s role as a rule writer, not as a tariff cutter.
Plan of the paper
The next section discusses the complexity of 21
century commerce, how it arose and why.
The subsequent section, Section 3, discusses how this more complex trade created a demand
for more complex disciplines and how 21
century regionalism met these demands. Having
laid out the problem, Section 4 argues that a new analytic framework is needed to think about
century regionalism. Section 5 discusses features the new framework should display.
Section 6 considers the implication for the world trading system and presents some
- 4 -
2. EMERGENCE OF 21
Today‟s international commerce comprises complex, two-way flows of goods, services,
people, ideas, and investments in physical, human and knowledge capital – in addition to
trade in raw materials and final goods. There is nothing new about this from a qualitative
sense. It has been going on for decades. The 1957 Treaty of Rome and the 1965 US-Canada
Auto Pact, for example, were designed to encourage just such exchanges.
The novelty of 21
trade lies in the quantitative dimension. To illustrate what is really new, it
is useful to put 21
century trade into the broad context of globalisation.
2.1. Globalisation as two unbundlings
When sailing ships and horse carts were state-of-the-art transportation, only items with very
high value-to-weight ratios could be profitability shipped over anything but the shortest
distances. As a result, each village made most of what it consumed; production and
consumption were geographically bundled. Steam power changed this (O‟Rourke and
Williamson 1999, Chapter 3).
Railroads and steamships radically lowered transport costs, so production and consumption
could be unbundled geographically. Once unbundling was feasible, scale economies and
comparative advantage made it inevitable. This was globalisation‟s “first unbundling”
although it occurred in two waves punctuated by world wars and the Great Depression.
Global unbundling with local clustering
Globalisation‟s first unbundling generated a paradox. As production dispersed
internationally, it clustered locally (into factories and industrial districts). Better
transportation favoured scale economies that typically involved complex manufacturing
processes. This new complexity fostered local clustering. For example, consider a stylised
production process with several production bays. Coordination involves a continuous, two-
way flow among the bays of good, people, ideas, and investment in machines, training and
technology. The two-way flows never cease as continuous efforts to heighten productivity
keep the process in flux. Production clustered locally because proximity lowered the cost of
the two-way flows. A new distance-linked cost became important – what might be called
ICT revolution: 2nd unbundling’s equivalent of the steam revolution
Some of this coordination glue is related to communication. As telecommunications became
cheaper and surer from the mid-1980s, the coordination glue began to weaken. The price of
telephone calls plummeted, faxes became standard, cellular phone usage exploded, and the
telecommunication network became denser, more reliable and cheaper. Two other trends
interacted with cheaper communication costs – the spectacular fall in the price of computing
power (Moore‟s Law) and the equally spectacular rise in fibre optic transmission rates
(Gilder‟s Law). Long-distance information sharing was revolutionised as these developments
in telecoms were complemented by the rise of the internet – first email and then web-based
For estimates of trade flows back to 1870, see David, Meissner, and Novy (2011). For a detailed account of the
two waves of the first unbundling, see Baldwin and Martin (1999).
- 5 -
Figure 1: Growth of global internet hosts and phone lines, 1975 – 2011.
Source: WDI for phone and internet users; www.isc.org for internet hosts
Mobile and fixed-line telephone subscribers per employee
Mobile and fixed-line telephone subscribers (millions)
Internet users (millions)
Log number of internet hosts
The telecom and internet revolutions triggered a suite of information-management
innovations that made it easier, cheaper, faster, and safer to coordinate complex activities at
distance. Email, editable files (*.xls, *.doc, etc), and more specialised web-based
coordination software packages revolutionised peoples‟ ability to manage multifaceted
procedures across great distances. Working methods and product designs also shifted to make
production more modular and thus easier to coordinate at distance. Stages of production that
previously had to be performed in close proximity – within walking distance to facilitate
face-to-face coordinate of innumerable small glitches – could now be dispersed without an
enormous drop in efficiency or timeliness. Collectively, this is known as the information and
communication technology (ICT) revolution.
Figure 1, which displays several ICT, indicators, shows that there was an inflection point in
the growth of internet hosts in 1985 and in telephone subscribers in 1995. This suggests that
the coordination glue began to weaken sometime between 1985 and 1995.
As far as trade is concerned, the partial melting of the coordination glue meant that some
production stages that previously had to be within walking distance could now be dispersed
internationally. Once ICT made dispersion of production stages feasible, scale economies and
comparative advantage made it inevitable. This is globalisation‟s “second unbundling” – the
spatial unbundling of productions stages previously clustered in factories and offices.
radically changed the nature of international commerce giving rise to what might be called
the trade-investment-service nexus.
2.2. The trade-investment-services nexus: 21th century trade
unbundling did not end the need to coordinate production stages – it internationalised
it. International commerce became more complex. The result might be called 21
trade. The heart of 21
century trade is an intertwining of: 1) trade in goods, 2) international
investment in production facilities, training, technology and long-term business relationships,
and 3) the use of infrastructure services to coordinate the dispersed production, especially
services such as telecoms, internet, express parcel delivery, air cargo, trade-related finance,
customs clearance services, etc. This could be called the trade-investment-services nexus.
See Baldwin (2006a) for the original presentation of globalisation as two unbundlings and policy implications
for European social welfare states.
- 6 -
Figure 2: Schematic illustration of 20
1) Two-way flows of goods, ideas, technology,
capital, and technicians.
2) Investment and application of technical,
managerial and market know-how abroad.
Connecting factory & doing business abroad: The
The differences between 20
century trade are illustrated schematically in Figure 2.
The top panel illustrates 20
century trade; trade is dominated by goods made in factories in
one nation and sold to customers in another. There are complex two-way flows of goods,
people, and ideas (the double-headed arrows) but primarily within factories. The lower panel
century trade. Here factories and offices have been unbundled internationally
thus creating the trade-investment-service nexus where some of the complex two-way flows
that used to take place within factories and offices now take place across international
It is useful to think of the trade-investment-services nexus as being created by two distinct
Connecting factories, and
Doing business abroad.
Trade in parts and components, trade in infrastructure services, and foreign direct investment
are the most easily measured aspect of this multifaceted, multi-directional commerce, but
they are only the tip of the proverbial iceberg.
2.2.1. Indicators of the second unbundling
The “second unbundling” happened first among developed nations. Examples include US-
Canada or French-German trade in autos and auto parts in the 1970s (Hummels, Rapoport,
and Yi, 1998). The big change, however, came when the second unbundling accelerated
between developed and developing nations (Hanson and Feenstra 1997, and Ando and
Kimura 2005). The dominant factor here was the juxtaposition of the ICT revolution and
colossal wage discrepancies.
Many scholars have documented the nature of this „new‟ trade. For example, one very
obvious form of production unbundling is known as outward processing trade, or vertical
For firm-level evidence, see Ariu and Mion (2010).
- 7 -
specialisation trade as documented by Ishii and Yi (1997). Here intermediate inputs are
imported and used in goods that are subsequently exported. Amador and Cabral (2008) show
that this trade was more important in among European and North American nations up until
the early to mid-1980s . After that, it boomed North-South, but especially in Asia (Figure 3).
Figure 3: Outward processing trade, 1967 – 2005.
Source: Amador and Cabral (2008)
Source: Amador and Cabral (2008).
Internationalisation of the supply chain however is a much broader phenomenon than
outward processing trade.
In the case of East Asia, production unbundling has progressed to
a state where the term “Factory Asia” is widely used.
Development of Factory Asia can be tracked with the Asian Input-Output table maintained by
Japan‟s JETRO.6 In 1985, Factory Asia was simple. The manufacturing sectors of developing
East Asia – which meant all nations other than Japan – supplied most of their own
intermediates. And what they did not supply themselves, they imported from technologically
advanced nations, mostly Japan, the US and the EU. There was very little trade in
intermediates among them. By the 1990s, local sourcing of parts and components fell,
matched by an increase in imports from Japan, the US, Korea, Taipei China, Singapore, and
Moving ahead to 1995 and 2000, the trends generally continue, but by 2000, intermediate
trade among developing East Asian nations had become a significant phenomenon. Countries
like Thailand and China were no longer simply engaged in outward processing trade. They
were supplying considerable shares of intermediate inputs into the manufacturing sectors of
their fellow developing nations. Japan‟s and China‟s share of own-intermediates, however,
hardly budged during this time.
Evidence for these trends on a global scale can be found in a recent important paper, Johnson
and Noguera (2010). They document just how far supply chains have been internationalised.
Another important – and easily measured – facet of supply-chain internationalisation is
foreign direct investment. This also flourished at approximately the same time, namely the
mid-1980s and early 1990s. Figure 4 illustrates the case of Japanese auto and electrical
See, for example, Ando and Kimura (2005), Kimura, Takahashi, and Hayakawa (2007), Gaulier, Lemoine and
Unal-Kesenci (2007), and Athukorala (2005) in the East Asian case, and Dallas Fed (2002) or Feenstra and
Hanson (1996) on the North American case.
See Baldwin (2006b) for a more detailed analysis of Factory Asia using the Asian IO matrices.
- 8 -
machinery plants placed in East Asian nations. The evolution shows a clear acceleration from
1985 with another inflection point in the mid-1990s – mostly due to plants placed in China.
Figure 4: Number of Japanese auto and electrical machinery plants in East Asia, 1975 –
1975 1980 1985 1990 1995 2000 2004
Source: Fujita and Hamaguchi (2006).
3. FILLING THE GOVERNANCE GAP: 21
Simple commerce needs simple rules; complex commerce needs complex rules. When trade
meant factories in one nation selling goods to customers in another, international rules could
be simple – dealing primarily with border measures and a few „behind-the-border barriers‟
(BBBs) such as blatantly discriminatory national taxes and regulations. Multilaterally, the
1947 GATT embodied this simple set of disciplines. 20
century regional trade agreements
(RTAs) were even „shallower‟ – frequently addressing little more than tariffs and rules of
century trade became more complex, demands arose for more complex international
trade rules. This section argues that this demand was met by collection of uncoordinated
developments outside the WTO – a collection that might be called 21st century regionalism.
The nature of these new demands is discussed before considering how 21
regionalism filled the governance gap.
3.1. Demands for deeper economic integration rules
When it comes to governance, the critical difference between 20
century trade is the
trade-investment-services nexus. As discussed above, the nexus entails two elements, each of
which generated new demands for more complex international disciplines:
Doing business abroad.
When firms set up production facilities abroad – or form long-term ties with foreign suppliers
– they typically expose their capital as well as their technical, managerial and marketing
- 9 -
know-how to new international risks.
Threats to these tangible and intangible property rights
century trade barriers.
Connecting international production facilities.
Bringing high-quality, competitively-priced goods to customers in a timely manner requires
international coordination of production facilities via the continuous two-way flow of goods,
people, ideas and investments.
Threats to these flows became 21
century trade barriers.
Figure 5: The disciplines to underpin the “trade-investment-services nexus”
1) Cross border barriers: Threats to cross-border
movement of goods, ideas, capital, and people.
2) Behind the Border Barriers (BBBs): threats to
foreigners‟ property rights & local business climate.
21st century trade barriers
Figure 5 illustrates 21
century trade barriers schematically. Here are some examples.
The sharing of tacit and explicit technology and intellectual property is facilitated by
assurances that foreign knowledge-capital owners will be treated fairly and their
property rights will be respected.
Foreign investments in the training of workers and managers, physical plant, and the
development of long-term business relationships are facilitated by assurances on
property rights, rights of establishment, and anticompetitive practices.
Assurances on business related capital flows – ranging from new FDI to profit
repatriation – also helped foster the investment part of the trade-investment-services
Connecting factories often involves time-sensitive shipping, world class telecoms and
short-term movement of managers and technicians, so assurances on infrastructure
services are also important.
Tariffs and other border measures also matter – just as they mattered in the 20th
century but more so since the ratio of value added to value on individual shipment
falls as the production chain fragments, even though tariffs are applied to the value of
the goods as they cross borders.
As World Bank (2011) notes doing business abroad implicates “the laws, regulations and institutional
arrangements that shape daily economic activity.” The entails rules that establish and clarify property rights,
moderate the cost of resolving disputes, boost predictability of economic exchanges, and guard contractual
partners against abuse by public or private agents.
Tariffs on imported intermediates are one part of this. Coordinating international production also requires
assurances of world-class telecommunications, world-class goods transportation (especially express parcel
services and air cargo) and customs clearance, transportation and assured access for short-term visits by key
personal (managers and technicians), and capital and financial market openness to inward and outward
investment flows and profit repatriation.
- 10 -
This list suggests four types of 21
century trade barriers that were not barriers to 20
trade: competition policy (known as anti-trust in the US), movement of capital, intellectual
property rights beyond the TRIPs Agreement, and investment assurances.
Supply meets demand
century trade was welcomed by many nations. To encourage it, they wanted to remove
century trade barriers. This is the ultimate source of demand for more complex
international disciplines. Multinationals from advanced-technology nations were eager to
lower production costs by dispersing production and technology to the most cost-effective
locations. Developing-nation governments embraced 21
century trade as a fast lane to
industrialisation and growth. In short, the mutual interest in 21
century trade meant that the
rising demands for new discipline were met by willing suppliers – namely governments.
This complex trade first arose among rich nations (Figure 3). These nations‟ high levels of
property rights, legal transparency and availability of infrastructure services helped underpin
century trade even without WTO or RTA disciplines. The lack of deeper disciplines
became a first-order problem when 21
century trade started to involve emerging economies
with weaker domestic governance.
This is how 21
century regionalism first emerged. Emerging market economies– eager for
advanced technology factories – were willing to embrace disciplines on things that were not
traditionally considered to be trade barriers. As the WTO was otherwise occupied, the
demand for deeper discipline was filled by:
Bilateral Investment Treaties (BITs), and
Consider these in turn.
3.2. Deep RTAs
The distinction between deep and shallow RTAs was highlighted explicitly by Lawrence
(1996). He also noted its association with more complex trade and pointed out that it first
developed among developed nations in Europe and North America. From the mid-1990s,
deep RTAs spread rapidly to cover North-South trade. The US-Mexico component of
NAFTA and Europe‟s Euro-Med Association Agreements led the way. More recently, Japan
has joined the movement by signing deep Economic Partnership Agreements (EPAs) with the
large ASEAN economies. Note, however, that even South-South agreements have deepened
in the 21
- 11 -
Table 1: WTO-X and WTO+ provisions in regional trade agreements.
Pre-WTO 1995-2000 DDA era, post 2001
Customs 13 11 56
AD 12 8 53
CVM 4 5 52
Export Taxes 8 8 41
State Aid 10 9 34
TRIPs 6 4 41
GATS 7 2 39
STE 5 3 35
TBT 2 2 36
SPS 2 1 35
Public Procurement 5 0 32
TRIMs 6 2 31
Competition Policy 11 9 39
Movement of Capital 6 5 38
IPR 5 2 39
Investment 4 1 35
Source: World Trade Report, WTO (2011).The WTO+ and WTO-X classification is from Horn,
Mavroidis and Sapir (2010).
Recent research has started quantifying the depth of various RTAs based on the methodology
of Horn, Mavroidis and Sapir (2010). These authors read through all the US and EU
agreements and noted whether they contained WTO+ and WTO-X provisions, and whether
these provisions were legally enforceable. WTO+ provisions concern commitments that
already exist in WTO agreements but go beyond the WTO disciplines. WTO-X provisions
cover obligations that are outside the current WTO aegis. In all Horn and co-authors identify
14 WTO+ and 38 WTO-X policy areas. Yap, Medalla and Aldaba (2006) and Balboa (2008)
have done a similar exercise on Japanese EPAs.
The 2011 World Trade Report by the WTO Secretariat extends the Horn-Mavroidis-Sapir
method to 97 more RTAs. Of this, 33 involved the EU and 11 involved the US. The 43 other
RTAs were concluded by regional trading blocs and major trading powers such as ASEAN,
China and MERCOSUR. The sample of RTAs was chosen based primarily on their volume
of intra-RTA trade. The agreements investigated came into force during the 1958 to 2010
Using their new dataset, WTO (2011) concludes that RTAs have been getting increasingly
„deep‟ with most of the provision being legally enforceable. As they note: "The pattern
observed suggests that deepening commitments in these areas, i.e., going beyond
commitments in the WTO, continue to be a major driving force for recent RTAs.” The report
also points out that RTAs in the 21
century cover more WTO-X areas than earlier RTAs.
The main policy areas covered are competition policy, intellectual property rights, investment
and movement of capital. As these WTO-X provisions are largely regulatory in nature, their
growth is “testimony to the growing importance of behind-the-border measures in RTAs.”
One of the most striking results from the WTO‟s new work is the identification of a core of
four deeper disciplines that appear in over a third of RTAs, but which are not part of the
WTO‟s rulebook. These are competition policy (47% of all agreements), movement of capital
(39%), intellectual property rights not in the TRIPs Agreement (37%), and investment (31%).
- 12 -
Deep RTAs are also defined by what they are not. As it turns out, tariff preferences are no
longer very important at a global scale, although they are still important in certain sectors and
3.2.1. RTAs are no longer very preferential
While bilateral trade flows covered by RTAs account for about half of world imports, only
16.7% of world trade is eligible for preferences (Carpenter and Lendle 2010). The remaining
trade flows either have zero MFN tariffs (about 25% of world trade) so there can be no
preference, or they are excluded from preferential treatment by the terms of the RTA (about
9% of world trade). In fact, the largest traders have lowered their applied MFN tariffs to zero
on a wide swath of imports. The figure for the EU‟s external trade is 56%, while it is 43% for
the US, 48% for China, and 80% for Japan. The products where the large importers maintain
high tariff are routinely excluded from their RTAs, so again no preference arises. A second
important fact is that the margins of preferences are quite low on the 16.7% of world trade
that is preferential. In fact, less than 2% of world imports enjoy preferences over 10
As Table 2 shows, imports of the largest importers are not subject to huge preference
margins. Intra-EU trade is by far the most preferential with 9% carrying preference margins
over 10 percentage points. Preferences in this range granted by the US amount to 2% of its
imports; China and Japan grant such preference margins to essentially none of their imports.
The nations with the highest share of imports that are covered by large preferences (i.e. over
10% margins) are small nations that are heavily dependent on large neighbours. For instance,
16% of Mexican and 2% of Canadian imports receive over 10% margins, while the
corresponding number for Turkey is 2%. For these nations, and Chile, an important share of
imports received margins of between 5 and 10 percentage points.
Table 2: Margins of preference in 2008.
Share of imports according to Margin of Preference
Over 20% 20% to 10% 10% to 5%
World 1% 2% 7% 18% 69% $13.6
World (ex intra EU) 1% 1% 4% 11% 83% $9.8
Largest importers (over $500 billion)
EU (internal) 4% 5% 17% 38% 34% $3.8
EU (external) 0% 2% 3% 11% 82% $2.3
US 1% 1% 2% 22% 74% $2.1
China 0% 0% 2% 4% 93% $1.0
Japan 0% 0% 1% 5% 93% $0.7
Other top traders
Mexico 6% 10% 31% 1% 48% $0.30
Canada 0% 2% 26% 8% 65% $0.37
Chile 1% 3% 9% 40% 46% $0.18
Turkey 0% 2% 11% 27% 59% $0.19
Brazil 3% 4% 4% 1% 88% $0.17
Russia 1% 3% 2% 8% 85% $0.19
Indonesia 1% 1% 3% 20% 73% $0.07
These numbers are calculated without considering intra-EU trade. Taking world totals (including intra-EU
flows) Carpenter and Lendle (2010) calculate that 64% of world trade is covered by an RTA, 29.8% of world
trade is subject to preference margins, but only 3.9% of this enjoys margins over 10 percentage points.
- 13 -
Malaysia 1% 2% 1% 1% 92% $0.14
Thailand 1% 1% 1% 4% 93% $0.13
Australia 0% 0% 1% 12% 86% $0.19
Korea 0% 0% 1% 8% 90% $0.43
India 0% 0% 1% 4% 93% $0.22
Singapore 0% 0% 0% 0% 100% $0.24
Chinese Taipei 0% 0% 0% 0% 100% $0.23
Argentina 0% 0% 0% 5% 95% $0.15
Hong Kong 0% 0% 0% 0% 100% $0.37
Source: Author‟s calculations based on Carpenter and Lendle (2010) data.
The massive data demand for the calculations in Table 2 prevent Carpenter and Lendle
(2010) from covering all plurilateral RTAs beyond the EU and NAFTA. A rougher set of
facts, however, is available for a broader range of RTAs. Archarya, Crawford and Renard
(2010) calculated the share of various RTAs imports that are subject to zero MFN applied
rates and thus not eligible, de facto, for preferences.
The results, shown in Figure 6, confirm the basic fact that many so-called preferential trade
agreements are not in fact very preferential. The major North-North RTAs all have half or
more of their imports duty-free on an MNF basis. The developing nation RTAs, especially in
African and Latin America have much lower shares of MFN duty-free imports (the South
African Customs Union, SACU, and the Central American Common Market are exceptions),
but in most cases, the share of MFN duty-free imports has increased significantly since 1995
– with the increase for Mercosur being particularly noticeable.
Another excellent study illustrating the same basic facts is Fugazza and Nicita (2010). They
go one step further in considering interactions between preference margins and import
elasticities. The idea is to check whether the large preferences fall on goods where quantity
responds abundantly to small price differences. Despite this refinement, they come to the
same conclusion that tariff preferences are now rather small from a global perspective.
Figure 6: Share of imports with MFN zero tariffs, various RTAs, 1995 to 2008.
0% 20% 40% 60% 80% 100%
Source: Acharya, Crawford & Renard (2010)
Zero MFN tariff (% Total Imports)
Deep RTAs are not the only non-WTO route to establishing the disciplines required for 21
century trade. Equally important has been the explosion of Bilateral Investment Treaties.
- 14 -
3.3. Bilateral Investment Treaties
BITs have long attracted the attention of trade scholars, although much more from trade
lawyers than from trade economists (Sauvant and Sachs 2009).
Treaties (BITs) establish disciplines that govern interactions between private foreign
investors and host governments. As such, they are central to the investment part of the trade-
investment-services nexus which forms the core of 21
BITs are extremely common with about 2,500 in existence – almost five times the number of
RTAs. This number, however, overstates the coverage. Investment policy – unlike trade
policy – is not in the hands of EU but rather individual EU members, so, for example, 23 of
Turkey‟s 82 BITs are with EU members. Overall, more than 1,200 BITs have been negotiated
by individual EU members (TNI 2011).
As Figure 7 shows, the explosion of BITs coincided with the second unbundling (which took
off between 1985 and 1995). As international commerce became more complex – in
particular as the investment part of the trade-investment-services nexus developed, FDI
emitters signed BITs with FDI seekers. There may well have been a sort of domino or „race
to the bottom‟ element in the BIT proliferation, as FDI seekers felt they had to sign BITs to
guard their locational competitiveness against rivals who had already signed BITs with the
major source-nations of FDI.
The basic aim of BITs is to encourage FDI, often FDI related to the trade-investment-services
nexus. All the major FDI emitters – the Europeans, the US, Japan, etc. – have their own
model agreements. The US model is explicit and quite comprehensive, so it serves well as an
illustration of the basic features of a BIT.
Figure 7: Explosion of Bilateral Investment Treaties.
No. of Treaties Signed
Cumulative No. of Treaties Signed (right scale)
FDI in Billion US$
Source: UNCTAD’s World Investment Report, various issues.
BITs impact on FDI flows has been recently estimated by, for example, Egger and Merlo (2007).
See, for example, http://www.state.gov/documents/organization/117601.pdf.
- 15 -
The six basic goals of a typical US BIT are:
To assure non-discrimination. Investments abroad are to be treated as favourably as the
host party treats its own investors (national treatment) and third country investors (most
To discipline expropriation. Expropriations are to be limited and when they occur, they
must be compensated by prompt and adequate payment.
To assure transferability of investment-related funds. Capital and financial markets are to
be open so as to permit investment-related capital inflows and outflows without delay and at
market rate of exchanges.
To discipline the imposition of performance requirements. Host government policies that
tie the hands of investors – such as local content targets or export quotas, as a condition for
the establishment, acquisition, expansion, management, conduct, or operation of an
investment – are to be limited.
To assure investors ability to choose top managers. Host-government policies relating to
the choice or nationality of top management is to be limited, or prohibited.
To grant private investors the right to submit disputes with the host government to
international arbitration as opposed to local courts. This feature is one of the most
distinctive and it has lead to construction of a new jurisprudence completely independent of
the WTO‟s dispute settlement procedure. The main “court” is the International Centre for
Settlement of Investment Disputes (ICSID) based in Washington.
Not just North-South
Up to 1980, almost all BITs involved at least one old OECD nation, but BITs among OECD
nations were rare. For the most part, BITs were North-South agreements. Starting in the
1980s, with the trend accelerating in the 1990s, new OECD members (see the table note for a
list) and non-OECD nations began signing them. BITs are no longer only North-South
agreements. The second unbundling has seen a proliferation of South-South FDI which also
benefitted from BIT-disciplines and assurances.
This development is very much in line with the data from Japan‟s Asian Input-Output data
discussed above. In the 1980s, the supply chain was fairly simple – what was known as
“triangle trade” in Asia (METI 2005 Section3) – where Japan and the NIEs produce
intermediate goods, China and ASEAN import intermediate goods, assemble them into final
goods, and then export to the US and EU. As the second unbundling proceeded, the linkages
in Factory Asia became denser. The trade-investment-service nexus developed among
developing nations and thus arose a demand for the property assurances in BITs.
Table 3 shows the facts. The five panels in the table each show a decade starting in 1959. The
figures reflect that number of BITs signed in the relevant decade between partners in the
three groups – old OECD, new OECD and non-OECD. In the 1990s, almost 500 BITs were
signed among non-OECD nations. The BIT network is especially dense among the leading
East Asian emerging economies: China, Thailand, Indonesia, Malaysia and the Philippines.
There is another BIT network among the Andean nations.
As far as the old OECD nations are concerned, the flow of new BITs has tapered off
significantly, but South-South BITs continue to flourish. The world leader in BITs is
Germany with 147; the US is in 33
place with “only” 48.
- 16 -
Table 3: Number of BITs by decade and by partner pairings.
1959-69 Old OECD Members 1
New OECD Members 2
1970-79 Old OECD Members
New OECD Members 11
Non-OECD 67 1 9
1980-89 Old OECD Members 1
New OECD Members 30 2
Non-OECD 141 13 28
1990-99 Old OECD Members
New OECD Members 86 15
Non-OECD 606 202 485
2000-07 Old OECD Members
New OECD Members 12 3
Non-OECD 234 41 224
Source: Database of BITs compiled by ICSID http://icsid.worldbank.org/ICSID/FrontServlet.
Notes: Old OECD: Australia, Austria, Belgium-Luxembourg, Canada, Denmark, Finland, France, Germany,
Greece, Iceland, Ireland, Italy, Japan, Netherlands, New Zealand, Norway, Portugal, Spain, Sweden,
Switzerland, UK, US. New OECD: Czech Republic, Hong Kong, Hungary, Korea, Mexico, Poland, Singapore,
Slovak Republic, Turkey.
The 1990s explosion of BITs had two margins of expansion. The number of BITs signed per
nation increased, and the number of nations signing BITs increased. Before the second
unbundling started in the late 1980s, BITs were initially mainly between European FDI
emitters and developing nation FDI seekers. The UK and Japan signed their first treaties in
the 1970s and the US only in the 1980s. The list of developing country BIT signers expanded
rapidly in the 1990s and since many of them signed BITs with the major FDI emitters (the
big EU nations, the US and Japan) as well as with other developing nations in their region,
the number of new BITs in the 1990s top a thousand. By the end of the 1990s, almost all
WTO members were playing the BITs game.
Many developing nations actively seek the participation of foreign companies in their
economies and indeed set a central role for such participation in their industrialisation
strategies. As part of this, many have signed the deep RTAs and BITs as discussed above.
Even more, however, have embarked on unilateral policy reforms that improve their business
climates, especially for foreign firms in manufacturing or trade-oriented services. It also
improves the competitiveness of local manufacturing by lowering the cost of imported parts
Evidence for the broad reforms is difficult to marshal for many nations going back to the
early days of the second unbundling (1985-1995). One aspect, however, is easily measured –
tariffs. The spectacular tariff reductions in Latin America are shown in Figure 8. Clearly there
is substantial variation among nations, but a clear general towards unilateral tariff reduction
from the late 1980s. Note that now most Latin American tariff averages are around 10%. It
should be noted that Latin Americans where cutting tariffs in RTAs at approximately the
same rate as they were lowering their MFN applied rates.
- 17 -
Figure 8: Applied MFN tariff liberalisation in Latin America.
Average MFN applied tariffs AR
Source: IADB and ITC database.
Unilateralism has also proceeded in other developing nations. Some of this, especially in
Africa, was driven by IMF conditionality, but even nations not under such external pressure
lowered rates, as Figure 9 shows.
Figure 9: Average tariffs, South Asia, East Asia, Middle East & North Africa and Sub-
Source: World Databank, accessed April 2011
Developing Middle East & North Africa
Developing East Asia & Pacific
Note: Applied tariffs, all goods, simple average.
- 18 -
According to new evidence presented in WTO (2011), the global reduction in tariffs on parts
and components has exceeded the overall average. This is rough evidence of an association
between the 2
unbundling and autonomous tariff liberalisation.
4. NEW ANALYTIC FRAMEWORK NEEDED
Plainly regionalism has changed dramatically. How should we think about it? Can one
continue to use the traditional approach to regionalism that was established two decades ago
to understand shallow RTAs and 20
In the early 1990s, regionalism metastasised just as multilateral GATT talks slipped into a
four-year coma. These facts mandated an analytic approach that focused on regionalism
versus multilateralism. Krugman (1991a,b) Bhagwati (1991, 1993), and Summers (1991)
were the architects of the new approach, but it has come to be known by Jagdish Bhagwati‟s
bon mot “building blocks versus stumbling blocks”. Two assumptions form the pillars of the
It is useful to think of RTAs as being exclusively about preferential tariffs (non-tariff
measures are unimportant, or can be modelled as tariffs).
The identification of RTAs with tariff preferences (sometimes called the PTA-view) meant
that traditional thinking was based on Vinerian economics.
It is useful to frame the trade-off as regionalism-versus-multilateralism (unilateralism
can be ignored or folded into multilateralism).
In evaluating the tradeoffs, it is useful
to take regionalism as exogenous but the impact on multilateralism as endogenous.
On the political economy of unilateralism, see Garnaut 1991, Young 1996, Edwards and Lederman 1998,
Richardson 2001, and Sally 2008, Coates and Ludema (2001), Krishna and Mitra (2008), and very recently
Ludema, Mayda and Mishra (2010), Conconi and Perroni (2010), and Baldwin (2010).
All the pioneering works of Bhagwati, Krugman and Summers focused on tariffs, and virtually all the theory
articles stimulated by this early work assume they are only about preferential tariffs. Among the best-known are
Aghion, Antràs, Helpman (2007), Bond and Syropolous (1996), Freund (2000a,b), Grossman and Helpman
(1995), Krishna (1998, 2003). Levy (1997), McLaren (2002), Ornelas (2005a, b, c), Yi (1996), and Baldwin
Many adherents of the old view hold this premise so tightly that they insist on the using the term Preferential
Trade Agreements (PTAs) in lieu of the WTO‟s official moniker, RTAs.
When the building-stumbling-block paradigm was established, unilateralism was not widely appreciated; the
focus was on regionalism and its impact on the on-going multilateral negotiations (Uruguay Round). Subsequent
academic literature that endogenised the reaction to exogenous regionalism typically employed political
economy models where unilateralism never occurs, so unilateralism was excluded by modelling choices. For
example, Reizman (1985), Kennan and Reizman (1990), Krishna (1998), Freund (2000a, b), Limão (2006), and
Levy (1997). A later literature on “tariff complementarity” addresses this, but the results are routinely
characterised as informing the regionalism-versus-multilateralism choice that is central the building-stumbling-
block framing of the issues. See Bagwell and Staiger (1999a), Cadot et al. (1999), Freund (2000a), Yi (2000),
Ornelas (2005a, 2005c), Bond et al. (2004) and Saggi and Yildiz (2009).
For an early and very explicit statement see Bhagwati (1993 p. 3); Panagariya (2000 p.312) provides a recent
restatement. The seminal theory contributions – Limão (2006), Levy (1997), Freund (2000a), and Yi (2000) –
embrace this exogenous endogenous distinction. Some writers have softened the exogenous/endogenous
dividing line (e.g. Grossman and Helpman 1995, Krishna 1998). A later literature on “tariff complementarity”
addresses this, but the results are routinely characterised as informing the regionalism-versus-multilateralism
choice that is central the building-stumbling-block framing of the issues (instead of regionalism versus
unilateralism). See Bagwell and Staiger (1999a), Cadot et al. (1999), Freund (2000a), Yi (2000), Ornelas
(2005a, 2005c), Bond et al. (2004) and Saggi and Yildiz (2009).
- 19 -
Consider the evidence on whether these two pillars are valid for 21
4.1. Tariff preferences and Vinerian economics
Section 3.2.1 showed that it is factually incorrect to view RTAs as being mainly about
preferential tariffs. This directly suggests that the PTA-view of RTAs falls down when it
comes to 21
century regionalism. The reliance on Vinerian economics, however, could be
defended if the deep provisions in RTAs mimicked preferential tariffs. As we shall see, some
non-tariff measures in RTAs do act like preferential tariffs, but many do not. This assertion
can be demonstrated with empirical evidence and economic logic. To interpret the evidence
correctly, however, it is useful to review Vinerian logic.
Vinerian economics is nothing more than tax analysis when different rates are applied to
different suppliers. There are only three elemental effects. First is “Smith‟s Certitude”.17
Nations that receive tariff preferences see a higher export price and thus export more to the
preference giving nation. The preference-receiving nation (and especially its exporters) gains
from this “trade creation”. Second is “Haberler‟s Spillover”.18 Nations excluded from the
preferences see lower export prices (they lower prices to remain competitive even though
they still pay the tariff) and thus export less to the preference-granting nation. Excluded
nations (and especially their exporters) lose from this “trade diversion”. Third is “Viner‟s
Ambiguity”.19The preference-granting nation might or might not gain for the simple reason
that preferences create a new distortion (tax discrimination among foreign suppliers) while
removing another (tax discrimination between firms in the preference-granting and
A critical point in interpreting the evidence flows from the first two effects. If trade flows are
driven by preferential tariffs – or measures that are acting like preferential tariffs – we must
observe trade creation driving trade diversion.
This point is critical to interpreting the evidence, but not widely appreciated, so it is worth
considering some non-preferential trade reforms that could lead to trade creation without
trade diversion, or trade diversion that works in the „wrong‟ direction. For example, suppose
an RTA induces the integrating nations to apply identical safety standards for elevators, so
that it is easier for the partners to sell to each other. As product regulations are general
policies and not subject to rules of origin, it is logically possible that the RTA‟s regulatory
Smith (1776), as quoted in Pomfret (1997) puts it crisply: When a nation “exempt[s] the good of one country
from duties to which it subjects those of all other … the merchants and manufacturers of the country whose
commerce is so favoured must necessarily derive great advantage.”
The discussion in Haberler (1936) runs over several pages but is centred on page 384. Haberler‟s spillover
was certainly understood by scholars before Haberler (e.g. Bismarck used this aspect of customs union to
force/cajole many German-speaking states to join his unified Germany), but I assign it to Haberler since
Haberler‟s 1936 book shows that mainstream trade economists were confused about the theory of the second
best, illustrating why Viner‟s 1950 book was viewed as such a landmark.
Viner, who framed his arguments in words alone (he was blissfully ignorant of postwar mathematical and
diagrammatic analysis), couched his argument in the enduring but imprecise concepts of 'trade diversion' and
'trade creation'. Meade (1955) showed that the deep fundamentals turned on two effects: the trade volume effect
(i.e. changes in trade over a domestic-versus-border price wedge), and the trade price effect (i.e. the terms of
Specifically, the preference is why partner-nation exporters receive a higher price for their exports to the
preference-granting nations. This producer price rise is why they export more, and these additional exports – i.e.
trade creation – is what drives down domestic prices in the preference-granting nation. As third-nation exporters
must match the new lower price inside the preference-granting nation, they see their producer price fall and this
leads them to sell less (trade diversion). In short, trade creation drives trade diversion. Trade diversion without
trade creation must be driven by something other than preferential tariffs or things that can be analysed as if they
were preferential tariffs.
- 20 -
standardisation makes it easier for third nations to sell to the combined RTA market. This
could result in trade creation among the partners and “reverse trade diversion” for third
nations (or what is sometimes called external trade creation).
The empirical evidence on trade diversion
As argued above, the Litmus Test for the relevance of Vinerian economics is the
simultaneous observation of trade creation and trade diversion. If this is not observed, the
provisions in the RTA are not acting like tariff preferences and Vinerian economics is
A recent review, Freund and Ornelas (2010), note that most scholars estimate the increase of
imports among RTA partners and reduction of imports from third nations using the gravity
equation. This permits the researcher to control for other factors such as changes in income
and various idiosyncratic factors related to the year, the bilateral relationship, and the
individual nation. Summing up their review of the evidence, Freund and Ornelas write: “the
empirical literature is not entirely conclusive, it does suggest that trade diversion is not a
major concern, though in some agreements and sectors it may matter.” Nevertheless, most
studies find evidence for trade creation. Taken together this suggests that it may be more
correct to view RTAs as general trade liberalisations schemes than as discriminatory
One important study in this line, Magee (2008), uses data from the late 20
estimate the creation/diversion effects of 15 separate RTAs including NAFTA, the 1986
enlargement of the EU, and 1992 bilateral RTAs between the EU and several Central
European nations, Mercosur, ASEAN, and select RTAs among Latin American and among
African nations. Only 8 of the 15 RTAs are found to have been effective in the sense of
creating new trade among the partners (controlling for other factors).
The others did not
increase or decrease trade significantly. As trade creation is a prerequisite for preferential
tariffs having an impact, these 8 RTAs with positive trade creation are the only ones where
preferential tariffs might have been important. Of the 8 only 2 – the 1986 EU enlargement
and the EU Association agreements of the early 1990s are found to divert trade.
Moreover, Magee finds that NAFTA – the second largest RTA whose internal trade accounts
for about 8% of world trade – created trade and produced “reverse trade diversion” – that is
controlling for other factors NAFTA made it easier, not harder, for excluded nations to export
to the US, Canada and Mexico. This result strongly suggests that using Vinerian economics
to think about NAFTA‟s economic effects is misguided; NAFTA provisions do not seem to
be acting like preferential tariffs in the aggregate.
Estimates based on 21
century data – where we have direct evidence on the small size of
preference margins – show even weaker support for all-RTAs-are-PTAs view. A recent paper
estimates that most RTAs have produced the opposite of trade diversion (Archarya, Crawford
and Renard 2010).
The 8 are the 1986 EU enlargement, the Andean Community, Mercosur, the ASEAN FTA, the EU‟s and
EFTA‟s Association Agreements with Central and Eastern European countries, and NAFTA.
- 21 -
Figure 10: Recent estimates of trade creation and trade diversion.
-100% -50% 0% 50% 100% 150% 200% 250% 300% 350% 400%
Central European FTA
Estimated extra trade due to RTA
Source: Archarya, Crawford and Renard (2010)
Trade creation and diversion, selected RTAs
The estimated coefficients for trade creation (trade changes within the RTA), trade diversion
(changes in imports from third nations), and new exports to third nations are plotted in Figure
10. The most striking finding is that almost all the RTAs have lead to external trade creation,
i.e. reverse trade diversion – just the opposite of what would be expected if preferential tariffs
were the main change imposed by the RTAs. Indeed, in the only RTAs where there has been
large trade diversion, there has also been negative trade creation. Taken together, this
suggests that RTAs are acting more like general trade liberalisation schemes, or in the case of
COMESA and CARICOM as general trade restriction schemes.
These are critical pieces of evidence against viewing RTAs as being mostly or uniquely about
tariff preferences. If RTAs were only about preferential tariffs, Magee should have found
evidence of statistically significant trade diversion in all the RTAs where he also found trade
creation. The world still has preferential tariffs, and trade diversion does occur in some
sectors in some agreements, but the weight of empirical evidence now suggests that Vinerian
economics is not sufficient for understanding the economic impact of RTAs.
The second pillar of the old approach also has serious problems when it comes to 21
4.2. Regionalism vs multilateralism is inadequate
When the building-stumbling-block phrase was coined in 1991, the second pillar seemed a
reasonable assumption, i.e. that the issue was regionalism-versus-multilateralism with the
wave of regionalism taken as exogenous. After all, unilateralism was a rare bird in those
days. Virtually all tariff cutting had been done in multilateral trade negotiations (MTNs) or in
RTAs; reciprocity was critical to the politics of both. Moreover, the avalanche of RTAs (see
- 22 -
the light coloured bars in Figure 11) seemed to come out of the blue; taking the development
as exogenous followed naturally.
There are two major problems with the second pillar when it comes to 21
regionalism. The most obvious is the prevalence of unilateralism (demonstrated in Section
3.4). To put it differently, if the observed regionalism since 1994 had not been accompanied
by massive unilateral tariff cutting – i.e. the dark line in Figure 11 had been flat in the new
century instead of falling – the old approach might have been useful. No MTN occurred
while RTAs boomed so the old approach would have asserted that the RTAs were stumbling
blocks. But the world tariff average did fall in the new century, due mostly to unilateralism.
The second pillar could be restored if folding unilateralism into multilateralism was an
option. But this too falls down on the facts. What the world has seen is an explosion of
unilateral and regionalism, not unilateralism versus regionalism. Section 3 argued that
regionalism and unilateralism seemed to be jointly endogenous reactions to a common third
cause (the ICT triggered 2
unbundling). The old framing of the question misdirects attention
by highlighting a trade off that the world never faced.
The second major problem for the second pillar is posed by a different set of facts – the
historical parallelism of regionalism and multilateralism. For example, the US and Canadian
stances on starting FTA negotiations were shaped primarily by the outcome of a political
economy conflict inside each nation between national exporters (who would benefit) and
national import competitors (who would lose). Their stance on an MTN was shaped by the
same factors and actors. Given this, it is easy to understand the synchronicity of tariff-cutting
decisions. US and Canadian exporters were in the ascendency in 1986 and they got their
governments to embrace trade liberalisation regionally and multilaterally – the Canada-US
FTA talks and the Uruguay Round were both started in 1986.
Figure 11: Tariff liberalisation since 1947: RTAs, MTNs and unilateralism.
Sources: RTAs: WTO online databases & Hufbauer-Schott RTA database; tariffs: Clemson and Williamson (2004) up to 1988, then World DataBank (weighted tariffs all products)
New RTAs (number of
New WTO members
(number of nations)
GATT/WTO Round in
World average tariff (right
About half of world trade is now covered by an RTA of some form, but the NAFTA was the last big one; after
that, the RTA trade share was about 45% so the hundreds of RTAs expanded coverage by only 5% and many of
these excluded most high-tariff items.
- 23 -
In fact, synchronicity between regional and multilateral tariff cutting has been the hallmark of
the postwar period. In addition to the 1986, regional and multilateral tariff cutting marched in
tandem in 1964. That was the year the Kennedy Round was launched and it was also the year
the US and Canada started talks that led to the US‟s first postwar preferential trade
liberalisation arrangement – the US Canada Auto Pact – which grew into one of the world‟s
largest RTA outside of Europe (Keeley 1983 p.281). Two other GATT participants, Australia
and New Zealand, also signed an RTA in that year. 1972 again witnessed parallelism. The
EEC concluded talks for it first enlargement and bilateral FTAs with all non-acceding West
Europeans in 1972 – the same year that the US, EEC, and Japan jointly called for a new
GATT Round (Tokyo Round).
Box 1 for details.
Since 1994, no new progress has been made on multilateral tariff cutting, but a different form
of parallelism has continued. As Figure 11 shows, spreading regionalism has been associated
with rapid unilateral tariff cutting (also see Section 3.4).
These facts tell us that that it is
wrong to view the 1990s explosion of regionalism as exogenous.
This more fundamental analysis suggests that the old building-stumbling-blocks approach is
logically misstated – it asks about the correlation of two endogenous variables driven by
common factors (illustrated by the double-headed arrow in the diagram).
On the time of the Tokyo Round agreement, see Low (1993 p. 176).
The unilateralism can be seen as the continued fall of the world tariff average despite the lack of MTN cutting
and despite the fact that regionalism since 1995 has involved only small shares of world trade; see Section 3 on
- 24 -
Box 1: Parallelism in RTAs and GATT Rounds
4.2.1. The old approach’s logical incompleteness
Delving more deeply into the old approach‟s problems, it is clear that it is logically
incomplete. Surely government decisions on RTAs are also endogenous to political economy
forces, so it is logically incomplete to take regionalism as exogenous. A more complete
analysis is suggested by the endogenous trade policy literature (e.g. Grossman and Helpman
2002). This approach is premised on the assertion that special interest groups are the
fundamental determinants of trade policy. This approach has quite different implications for
how one thinks about regionalism and multilateralism.
The basic point can be illustrated with Figure 12.The three forms of tariff cutting are shown
at the right-side, namely regionalism, multilateralism, and unilateralism. The determinants of
all three forms of tariff reform are shown with the straight black arrows.
From the GATT‟s formation right up to the end of the 20th century, regional and multilateral liberalisations
have displayed a remarkable parallelism.
1947 saw the creation of the GATT and the Organisation for European Economic Cooperation
(OEEC) that coordinated preferential European liberalisation up to the end of the 1950s
(particularly of intra-European bilateral quantitative restrictions).
Annecy GATT Round was opened in April 1949 – the same year that European preferential
integration took off with the OEEC‟s European Payments Union.
The 1950-51 Torquay Round was launched just a few months after the famous Schuman
Declaration that is viewed by many as the birthday of today‟s European Union.
When European regional tariff cutting got serious from 1960, GATT tariff cutting restarted first
with the Dillion Round (1960-1961).
In 1964 – the year the Kennedy Round was launched – the US and Canada started talks that led to
the US‟s first postwar preferential trade liberalisation arrangement – the US Canada Auto Pact –
which grew into one of the world‟s largest PTAs (Keeley 1983 p.281). Two other GATT
participants, Australia and New Zealand, signed their own PTA.
1972 again witnessed parallelism between regionalism and multilateralism. The EEC concluded
talks for it first enlargement and bilateral FTAs with all non-acceding West Europeans. The same
year, the US, EEC, and Japan jointly called for a new Round, the Tokyo Round (Low 1993 p. 176).
1986 saw two big steps in preferential liberalisation and a new Round. The US and Canada
launched talks to broaden their 1965 sectoral arrangement into a deep FTA while Europe launched
its massive deepening preferential trade (Single Market programme). The Uruguay Round was also
launched in 1986 with an agenda that included many „deep‟ issues.
The parallelism broke down in the late 1990s. After some early post-Uruguay Round successes in 1997 –
e.g. the Information Technology Agreement that created global free trade in many ITC goods – multilateral
liberalisation stagnated while preferential trade liberalisation went from strength to strength. The many
years it took to launch the Doha Round and the decade of negotiations to date gave preferentialism a very
clear lead in the minds of many WTO members.
- 25 -
Figure 12: Endogenous-trade-policy approach to the tariff cutting.
Regional tariff cutting
Multilateral tariff cutting
Unilateral tariff cutting
Strength of pro-tariff cutting
& anti-tariff cutting special
Special interest politics shape the government‟s decision rule which in turn explains tariff
reform choices and the form they take (RTA, MTN or unilateral). The ultimate determinant
of a nation‟s stance on regionalism, multilateralism and unilateral is the strength of pro- and
Most writers in this tradition focus on the economic power of exporters and import-
competitors and how this power filters through to governmental decision rules via the
national political system.
In this sense, the economic power of export firms and import-
competing firms are the ultimate causes of trade policy choices. This applies equally to
regional, multilateral, and unilateral tariff cutting.
5. WHAT WOULD THE NEW APPROACH LOOK LIKE?
Up to the late 1980s regionalism was limited. Economists such as Jacob Viner, James Meade,
and Harry Johnson focused on narrow questions, chiefly: “Would a particular nation gain
from joining a free trade agreement?” New facts appeared in the 1990s and new thinking was
needed. The building-stumbling-block approach was developed to think about shallow RTAs
century has thrown up new facts and now a new framework for thinking about them
is needed. As the previous section argued, the 1990s approach is an inadequate framework for
thinking about 21
century regionalism and its implications for the world trade system. To
summarise, its focus on tariff preferences is not appropriate for 21
century regionalism. Its
focus on the regionalism-versus-multilateralism distinction is also amiss in a world where
unilateralism is a key driver of trade opening. Finally, the old approach was always logically
incomplete in that focused on correlations between endogenous variables (RTA and MTN
outcomes) rather than on the political economy roots of trade liberalisation.
Arguing that the old building-stumbling-block logic is not up to the job of understanding 21
century regionalism is much easier than finding a replacement. What the world needs is a
new framework that is as simple and compelling as the old one, but relevant to 21
regionalism. Such a framework has not yet emerged. It is useful, nevertheless, to list the
features it should include. The starting point is a discussion of the key features of deep RTA
5.1. Economics of deep RTA provisions
Much of 21
century regionalism concerns regulatory measures, not tariff measures. This
means that the fulcrum of the new thinking will be regulatory economics, not Vinerian
economics. Unfortunately, regulatory economics is harder and much less flexible than
An early reference is Cooper (1971 p.410), but also see Roesseler (1978), Blackhurst (1979), Baldwin (1980),
Moser (1990), or Hillman and Moser (1992). The best known reference is the trade-wars-trade-talks paper by
Grossman and Helpman (1995).
- 26 -
Vinerian economics since it often has to deal with industry-specific characteristics. For
example, in telecoms, network externalities are a dominate concern, but less so in the air
cargo. Asymmetric information is a prime concern in financial services but less so in capital
movement regulations. In some sectors, imperfect competition considerations are central
while in other sectors they are not. And the list goes on. This suggests that it may not be
possible even for the most brilliant economist to lay out the basic analytics as cleanly as Paul
Krugman did for the old thinking in his 1991 articles.
5.1.1. Hints from the lack of trade diversion
One critical fact that the new framework must explain is the presence of “reverse” trade
diversion, i.e. that RTAs seem to be making it easier for imports from partner and third
nations. As it turns out, many of the deep RTA provisions are not naturally thought of as
preferential despite their being reforms struck in the context of an RTA. Before turning to
specific deeper RTA provisions, consider the general source of the non-discrimination.
General considerations: Nationality of firms and pubic good features
As it turns out, it is hard, or even impossible, to write rules of origin for many deep
provisions. The trouble-making factors concern: 1) the difficulties of determining the
nationality of 21
century companies, 2) the public good nature of infrastructure services, and
3) the public-good nature of regulatory reform.
The first point is easy. In today‟s world, it is difficult to establish a company‟s nationality so
it is difficult to write deep RTA provisions that only apply to companies of any particular
nation. As we shall see in the examples below, this problem has frequently led RTAs to
define the affected firms by where they are incorporated. This allows firms from third nations
to free ride on bilateral opening by incorporating affiliates in one of the RTA nations.
The second point is more subtle. As argued in Section 3, deep RTAs typically foster the
trade-investment-services nexus by liberalising developing countries‟ infrastructure service
sectors, for example, telecoms. The RTA may be written in a way that gives telecom
companies from one partner an edge in providing telecom services in the market of the other.
In this sense it is discriminatory and may divert services trade from, say, a US telecom to,
say, a Japanese telecom. But as far as trade in goods is concerned, it matters little whether the
world-class telecoms are provided by a US or Japanese company. Trading firms from all
nations find it easier to sell to the developing country member of the RTA when
telecommunications work well. In this way, a deep RTA provision on telecom liberalisation
can act like a public good for exporters from all nations.
The third point is related. One important set of BBBs concerns health, safety and
environmental regulations. Given these are typically highly technical, governments
frequently ask the regulated industry to write the regulations (or at least gives them a big say
in their formulation). This naturally results in rules that favour incumbent domestic firms at
the expense of all others. To make the nation more attractive to foreign investment, the nation
may reform its regulations in the context of an RTA. For example, the RTA may commit the
nation to rely on international standards instead of national standards. The political economy
logic of this provision may be bilateral, but its effect can be multilateral. Disciplining
idiosyncratic regulations is likely to make it easier for exporters from all nations to sell to
- 27 -
Discrimination possibilities in six types of deep RTA provisions
To flesh out these general points, consider the relevance of Vinerian economics to the
analysis of six major areas of deeper RTA provisions: trade in services, government
procurement, competition policy, investment performance measures, technical barriers to
trade, and trade remedies.
Barriers to trade in services rarely occur at the border. Establishing discrimination – the sine
qua non of the applicability of Vinerian economics – is particularly difficult for two reasons.
First, it is hard to establish the origin of a service. Just think of how hard it would be to trace
the nationality of all the value added in a financial service or a telecom service.
government may turn from determining the origin of the service to determining the origin of
the service provider. This tactic, however, runs into the second intractable problem.
Corporate nationality is a hard thing to define precisely – especially for the large firms that
dominate global trade. In reaction, governments often turn to legalistic definitions such as the
nation where the company is incorporated. However once the rules-of-origin have reached
this point, they are easy to get around. For example, if the rule of origin in the Japan-
Malaysia EPA provides access only to banks registered in Japan or Malaysia, the rule
encompasses all the US and European banks that have affiliates incorporated in Tokyo.
Government procurement provisions are easy to subject to rules of origin as long as they
concern traded goods. For these provisions, Vinerian economics works fine. But much of
government procurement applies to services and here to rule-of-origin issues mentioned
above come to the fore.
Competition policy provisions are typically aimed at assuring partner companies that their
market access and/or investments will not be threatened by anti-competitive practices of
private or state-owned local firms. Competition policy, however, is difficult to subject to
rules of origin. Part of the problem is that it is difficult to establish the nationality of modern
corporations, as mentioned above. Another part stems from regulatory nature of most
competition policy. In most nations, the competition policy is based on rules delimiting
forbidden behaviours (e.g. price fixing) without regard to the nationality of the defendant or
plaintiff. Thus is Korea‟s competition policy is strengthened by the EU-Korea FTA, firms
from third nations benefit as well from the resulting increase in fairness of the Korean market
Investment performance provisions typically assure foreign investors that the host
government will not interfere with business decisions. It is difficult to make this
discriminatory for the reason mentioned above: difficulties in defining corporate nationality,
Moreover, governments often embrace such provisions as part of general move to a more
business-friendly investment climate and so end the practice for all foreign firms, not just
those coming from RTA partners. Indeed, developing nations may make these explicitly
nationality-neutral, so as to diffuse their reliance on FDI from their major source.
Technical barriers to trade (TBT) – such and health, safety, and environmental regulations
and standards – are also difficult to apply on a discriminatory basis, as the EU‟s experience
with its radical TBT liberalisation (the Single Market Programme) shows. TBTs are usually
justified on the basis of „good governance‟ criteria, like protecting consumer‟s health. Such
goals, however, admit little room for discriminating product by origin – either a product is
safe or it is not, regardless of where it is made.
See Baldwin, Evenett and Low (2009) for a more detailed analysis of specific deep RTA provisions.
This is a well-known problem and the same issue bedevils VAT authorities when it comes to de-taxing
- 28 -
Finally, RTA provisions concerning anti-dumping and anti-subsidy tariffs concern trade in
goods and so discrimination is possible using standard rules of origin. Here Vinerian
economics is appropriate.
To summarise, many but not all deeper provisions tend to act as general liberalisations rather
than discriminatory liberalisations because it is difficult or impossible to write rules of origin
for them that exclude third nations. The deep reasons are the difficulties in establishing the
nationality of modern corporations and of services as well as the public-good nature of the
many regulatory reforms in deep RTAs.
5.1.1. Political economy of 21st century RTAs
The new framework will also have to suggest a new political economy. The standard political
economy logic applied to trade liberalisation deals with tariffs and it focuses on an exchange
of market access.
This seems inadequate when explaining agreements whose main goal is to
establish disciplines that foster the trade-investment-services nexus. As discussed in Section
3.1, the nexus requires firms to connect factories and do business abroad. Deep RTAs provide
assurances on both fronts. The bargain in a 21
century RTA is “foreign factories in exchange
for domestic reforms”, not, “exchange of market access” as was the case for 20th century
RTAs. Of course, market access is still important, but the deep provisions are not really about
market access – they are about helping foreign companies connect production facilities
internationally, and do business locally.
Likewise, the winners and losers from 21
century RTAs invoke a much richer set of players,
so the new political economy framework will have to fit in a broader range of political
economy actors. For example, in many cases the losers from reform may be incumbent
corporations while the winners will be all other corporations, domestic and foreign.
An important implication of this recasting of the basic political deal is that not all nations can
drive such a bargain. Only nations that possess technology that they are willing to offshore
have the leverage to demand the massive domestic reforms that are so common in 21
To date, that basically boils down to US, the EU and Japan although Korea and Taipei, China
may also fall in this category. South-South FDI flows are rising, and South-South BITs along
with them, so this feature many change.
The new political economy framework must also account for all three forms of trade
liberalisation – multilateralism, regionalism, and unilateralism. In particular, the landmark
fact has to be the parallelism of RTAs, MTNs and unilateralism.
5.1.2. Parallelism as explained by the juggernaut effect
The parallelism of RTAs and MTNs up to the mid-1990s and the fact that the parallelism
continued after 1995 in the form of RTA and unilateralism provide important hints as to how
we should think about global trade liberalisation in general and 21
century regionalism in
particular. Indeed, the two forms of parallelism are clear indications that something is altering
the fundamental determinants of trade policy in a cyclic fashion. One line of thinking points
the finger to earlier trade liberalisation as the cause of the cyclic behaviour. This is illustrated
in Figure 13.
See, for example, Grossman and Helpman (1995), or Baldwin and Baldwin (1996).
- 29 -
Figure 13: Juggernaut and parallelism in RTAs, MTNs and unilateralism.
Regional trade liberalisation
Multilateral trade liberalisation
Unilateral trade liberalisation
Strength of pro-trade & anti-
trade special interest groups
Juggernaut & dominos
The basic idea is that tariff reforms themselves trigger feedback mechanisms that alter the
state of pro- and anti-liberalisation forces in each nation. Specifically, reciprocal trade
liberalisation strengthens pro-trade political economy forces and weakens those of anti-trade
As tariffs come down reciprocally – either multilaterally or regionally – import
competing sectors get smaller and typically less influential in trade policy formulation.
Similarly, as trade partners lower their tariffs, exporters grow in size and political strength
with the improved access to foreign markets. In short, this „juggernaut‟ feedback mechanism
asserts that reciprocal liberalisation tends to reshape the political economy landscape inside
each nation in a way that makes future liberalisation more likely.
The reciprocal liberalisation is typically phased in over 5 to 10 years and the necessary entry
(of exporters) and exit (of import-competitors) may take even longer. Under the juggernaut
approach, this is the source of the time dimension which leads to the episodic, synchronised
nature of trade liberalisation according to the juggernaut logic. Once the tariff-cutting ball
starts rolling, political economy momentum keeps it rolling until all tariffs in its path are
A second feedback mechanism is the so-called domino effect whereby the signing of one
RTA tends to induce the signing of more RTAs. The basic notion is that as FTAs reduce the
exports of third nations to the integrating nations, they also stimulate third-nation exporters to
engage in new political economy efforts to get their government to redress the new
discrimination. In many cases, third-nation governments respond by signing new FTAs with
one or both of the partners, who have recently integrated.
5.2. Political economy of unilateralism and the second unbundling
The new approach must give a central place to unilateralism as it is one of the three
mainstays of 21
century regionalism. Economists‟ thinking on unilateralism is not highly
developed, but a number of explanations have been offered, some of them focusing squarely
See Baldwin (1994), Staiger (1995), and Baldwin and Robert-Nicoud (2008).
Juggernaut is a mispronunciation of the Hindu deity of the Puri shrine, Jagannath, whose chariot – an
enormous and unwieldy construction – requires thousands to get rolling but once in motion, it is hard to stop.
See Baldwin (1994 p.73) for the first presentation of the idea, Baldwin and Robert-Nicoud (2009) for formal
modelling, and Fugazza and Robert-Nicoud (2010) for empirical evidence.
See Baldwin (1993) for the original formulation of the domino theory, Baldwin (1997) for an early
application, and Baldwin and Jaimovich (2009) for a formal model; empirical support is provided by Egger and
Larch (2008) and Baldwin and Jaimovich (2009). On the theory, also see Bond and Syropoulos (1996), Freund
(2000), Yi (1996), McLaren (2002), Levy (1997), and Krishna (1998).
- 30 -
on the trade-investment-service nexus.
These explanations focus on the suggestive fact that
unilateralism boomed at approximately the same time as supply chains started to stretch
across North-South borders.
One story is that emerging markets cut tariffs on parts and components to attract foreign
factories and got caught in a “race to the bottom” (Vézina 2010). Another turns on the logic
of effective rates of protection. Cutting tariffs on an imported input raises the level of
effective protection on downstream products (Cordon 1966). The political economy of this
explains why most nations have lower tariffs on imported intermediates than final goods –
this sort of tariff cascading is a way of increasing protection of the downstream activity (see
Cadot, de Melo and Olarreaga 2004). As the 2
unbundling proceeded, almost everything
became a traded intermediate input and so faced new pressures for liberalisation. While there
may be some merit in these ideas, it is clear that there is much work to be done on the
political economy and empirics of unilateralism.
5.2.1. Fiscal federalism and the WTO “border question”
When thinking about tariffs, it is clear that multilateralism is most efficient. But when it
comes to the deeper disciplines addressed by 21
century regionalism, the most efficient level
of governance is less clear. The theory of fiscal federalism should be part of the new
framework‟s thinking when it comes to the optimal allocation of policy setting.
There is a good analogy with what goes on in the EU where the question is: What policies
should be decided at the national level and which at the supranational level? The answer in
the EU is that the appropriate level depends upon the nature of the discipline. When it comes
to food standards, general guidelines are agreed at the EU level, but detailed regulations are
decided at the national level. Tariff and competition policy, however, are allocated to the EU
Similar thinking should be developed to make recommendations on which of the disciplines
that underpin the trade-investment-service nexus should be agreed at the multilateral level, at
the regional level, and at the national level.
6. IMPLICATIONS FOR THE WORLD TRADE SYSTEM
century regionalism is good news and bad news for the world trade system. The good-
news part is easy to explain. Trade liberalisation has progressed with historically
unprecedented speed in the 21st century, even when measured by a 20th-century-trade
yardstick like average tariffs (Figure 11). As a result, trade volumes have boomed, lifting
billions out of dire poverty. Twenty years ago, one could wonder whether regionalism would
be a building or stumbling block; now we know there were no stumbling blocks on the road
to zero tariffs. The road remained open and the world is driving down it as fast as ever. This
building-stumbling-block thinking, however, focuses attention on the wrong issues.
century regionalism is a threat to the world trade system, but the nature of the threat is
In the economics literature, most discussions of unilateralism consist of practical accounts of how and why
various nations undertook such measures (e.g. Garnaut 1991, Young 1996, Edwards and Lederman 1998,
Richardson 2001, and Sally 2008). The political economy theories that account for unilateralism include Coates
and Ludema (2001), Krishna and Mitra (2008), Baldwin (2005b), and very recently Ludema, Mayda and Mishra
(2010), Conconi and Perroni (2010), and Baldwin (2010). This section draws on the latter.
- 31 -
6.1. Regionalism’s real threat to global trade system
century regionalism has three parts: Deep RTAs, BITs, and unilateralism. Unilateralism
is not a systemic threat to the WTO – although it does make finishing Doha harder. Likewise,
the BITs have co-existed with the WTO for decades without any apparent harmful spillovers.
The real threat is that deep RTAs may undermine the WTO‟s central place in world trade
governance. But the threat to WTO centricity does not come on the tariff-cutting front; it
comes on the rules-writing front. As was shown above, the tariffs have come down a great
deal already, although they are still important in some sectors, especially agriculture.
More specifically, deep RTAs may undermine the WTO as the forum for agreeing new rules
– specifically the rules necessary to foster the trade-investment-services nexus that is the core
of today‟s international commerce. But why is the choice of forum a problem for global trade
governance? There are three main reasons to worry about the WTO being sidelined on the
rule writing front.
First, the basic WTO trade norms are almost universally accepted and respected – a very rare
thing (think of climate change, nuclear proliferation, or human rights). These norms are a
global public good of enormous, if unquantifiable benefit. The universality of the norms
stems in large part from the way they were promulgated – in multilateral negotiations where
the GATT/WTO consensus principle held sway. The new trade disciplines are being
promulgated in settings of massive power asymmetries – the deep RTAs signed by the US,
EU and Japan with small to medium sized developing nations. Lacking the legitimacy that
comes from multilateralism and consensus, it is not at all clear that the new norms will be
For example, some emerging markets – China, India and Brazil – are large enough to attract
foreign investment and technology without signing deep RTAs, and they have so far shunned
China in particular might decide to reject the rules – creating something like a “Cold
War of deeper trade disciplines”. This sort of distrust could spread beyond the new rules,
especially if China, India and Brazil feel that the US is practicing „competitive liberalisation‟
– trying to encircle them in a way that eventually confronts them with what might be seen as
an ultimatum. This outcome would be made more likely if the US reverts to its aggressive
unilateralism of the 1980s (the Plaza Accord and the Structural Impediments Initiative that
forced Japan to revalue and remove behind-the-border barriers) and 1970s (Nixon 10%
surcharge that forced Germany to revalue).
Second, a world where the WTO‟s importance starts to resemble that of UNCTAD – with all
the action going on in RTAs – is not a world that fosters multilateral cooperation on other
issues, such as trade-related policies that help with climate mitigation and adaption, or food
shortages linked to drought or floods. US, EU and Japanese interests will be served in the
short term, and the interests of small to medium emerging markets will likewise be served (if
not evenly), but where do Brazil, India and China fit in?
These nations are not in a position to set up their own systems of deeper disciplines for the
trade-investment-services nexus because they do not have advanced technology factories to
offshore in exchange for host-nation reforms (on the political economy of 21
regionalism, see Section 5.2). By the time their multinationals are ready to make major
outward pushes, the rules-of-the-road written by the US, EU and Japan will have been firmly
embedded into international commerce. More precisely, they will be embedded in the
domestic laws and regulations of all the host-nations that the Chinese, Indian and Brazilian
The EU-India agreement, for example, excludes many of the deeper disciplines in the EU‟s other RTAs.
- 32 -
companies will be looking at. Like it or not, Chinese, Indian and Brazilian companies will
have to play by the rules that are now being written by the US, EU, and Japan in agreements
that involve massive power asymmetries.
If Brazil, India and China play their assigned roles in this storyline, it may all work out
peacefully. But that is not the only outcome observed when such tactics were applied
historically. This is a world that starts to resemble the 19
Great Powers situation. That
episode of globalisation did not end well.
This is not the only scenario, of course. A whole system of trade and investment disciplines
has developed in the form of the BITs. Up to now, the BITs and its system of jurisprudence,
negotiations and politics does not seem to have undermined the WTO‟s authority on the
issues covered in the 1995 Marrakesh Agreement. But as international commerce becomes
ever more dominated by the trade-investment-services nexus, the WTO may be increasingly
sidelined when it comes to trade governance.
Third, the WTO‟s adjudication function is still working well, but any dispute settlement
system must walk on two legs. The judges can connect the dots for particular cases, but the
basic rules must be updated occasionally to match evolving realities. For example, the
Appellate Body finds itself ruling on issues like “zeroing” where the negotiated consensus is
disputed. If the basic rules applied by the Appellate Body are not updated, there is a serious
danger that the judges will overreach themselves, basing decisions on previous decisions that
were based on previous decisions. Similar challenges may arise when members ask the
Appellate Body to rule on 21
century climate subsidies and taxes based on rules negotiated
in the 1940s and last updated in 1994. The larger members may be tempted to take matters
into their own hands, applying sanctions based on unilateral law, not multilateral law.
Finally, if the WTO gets sufficiently sideline, it may prove very difficult to successfully
negotiate rule updates. Hereto, the GATT/WTO has always packaged rule-updating in
Rounds that were driven primarily by trade-liberalisation politics (juggernaut). If all the trade
liberalising action moves to the RTAs, WTO members will have to find a new way to
negotiate rule updates.
6.2. What is to be done?
One course of preventive action would be to work towards multilateralising the deeper
disciplines in RTAs. Many of the deep disciplines (e.g. opening up of the telecom sector, or
liberalisation of the air transport sector) are embodied in national laws. Many of these have a
public-good nature to them in the sense that they facilitate trade with all nations, not just the
members of the RTA which brought about the reform. Questions of consistency seem to be of
second order for such measures, but for other measures, e.g. intellectual property protection,
or investors‟ rights, it is not clear that the various deeper disciplines are compatible.
Distinguishing the various categories of disciplines is an important task for trade scholars and
governments. The WTO‟s centricity is not in peril if the various deep RTAs turn out to have
implemented reforms that are consistent with each other. Such disciplines, or at least the
basics, might be multilateralised with WTO agreements (like the GATS), or plurilateral
agreements like the Government Procurement Agreement. The disciplines that are creating
mutually inconsistent rules are more of a problem and need to be identified.
Part of this exercise will be to identify which deeper disciplines are more efficiently
organised at the global level and which are best set at the regional or national level. As
discussed above, economic theory on the allocation of tasks to various levels of government
(fiscal federalism) could be used to think about which of the deeper measures belong in the
- 33 -
WTO and which are more appropriately dealt with in RTAs and/or national legislation.
Again, this is an open question for trade scholars.
One hint in this direction is the near universality of certain provisions. Data from the WTO‟s
World Trade Report suggests that there are four core disciplines in deep RTAs that go
beyond WTO agreements. These are competition policy (covered by 47% of all agreements
in the WTO database), movement of capital (39%), intellectual property rights not in the
TRIPs Agreement (37%), and investment not covered by GATT 1994 (31%).
More modest versions of multilateralising the deeper disciplines can also be envisioned. The
WTO could develop some basic guidelines for deeper provisions in RTAs, akin to those on
tariffs and services in the GATT and GATS. For example, the GATS provides a few basic
guidelines for Services FTAs – e.g. FTAs should provide substantial sectoral coverage,
substantially eliminate discrimination in national treatment in the affected sectors, and raise
no barriers against third nations. Even these very basic guidelines are completely absent when
it comes to deeper provisions like competition policy, rights of establishment, FDI-linked
capital flows, etc. Since many of the deep RTAs are aimed at improving the investment
climate rather than providing discriminatory market access, many of the deep RTA
provisions already respect rules like those for Services FTAs. Perhaps then it might not be too
difficult to codify a set of guidelines in a WTO agreement, or plurilateral.
6.3. Concluding remarks
The rise of 21
century regionalism is not yet a disaster for the world trade system. It has kept
trade liberalisation and trade booming despite the WTO‟s slow progress. But the present
course of events seems certain to undermine the WTO‟s centricity – RTAs will take over as
the main loci of global trade governance. Over the past ten years, WTO members have “voted
with their feet” for the RTA option. Without a reform that brings existing RTA disciplines
under the WTO‟s aegis and makes it easier to develop new disciplines inside the WTO
system, the RTA trend will continue, further eroding WTO centricity and possibly taking it
beyond the tipping point where nations ignore WTO rules since everyone else does.
This scenario runs the risk that global trade governance drifts back towards a 19
Great Powers world. In the best of cases, the WTO would continue to thrive as the institution
that underpins 20
century trade flows. The Marrakesh agreements would form a „first pillar‟
of a multi-pillar trade governance system. All the new issues would be addressed outside the
WTO in a setting where power asymmetries are far less constrained. This is what has
happened with the BITs – they established a parallel system of disciplines without
substantially undermining the WTO‟s authority on Marrakesh disciplines. But this is not the
only scenario. It is also possible that the WTO‟s inability to update its rules gradually
undermines the authority of the Dispute Settlement Mechanism.
If the RTAs and their power asymmetries take over, there is a risk that the GATT/WTO
would go down in future history books as a 70-year experiment where world trade was rules-
based instead of power-based. It would, at least for a few more years, be a world where the
world‟s rich nations write the new rules-of-the-road in settings marked by vast power
asymmetries. This trend should worry all world leaders. In the first half of the 19
attempts by incumbent Great Powers to impose rules on emerging powers smoothed the path
to humanity‟s greatest follies – the two world wars.
See more detailed arguments on this point see Baldwin (2008) and Baldwin and Carpenter (2009).
- 34 -
Aghion, Philippe, Pol Antras, Elhanan Helpman (2007)."Negotiating free trade," Journal of
International Economics, Elsevier, vol. 73(1), pages 1-30, September.
Amador, João and Sónia Cabral (2008). “Vertical specialization across the world: A relative
measure”, Bank of Portugal Working paper, No. 10-2008.
Ando, Mitsuyo and Fukunari Kimura (2005). "The Formation of International Production and
Distribution Networks in East Asia," in T. Ito and A. Rose (eds) International Trade in East
Asia, NBER-East Asia Seminar on Economics, Volume 14, pages 177-216.
Ariu, Andrea and Giordano Mion (2010). "Trade in services: IT and task content," Research
series 201010-200, National Bank of Belgium.
Bagwell K, Staiger RW. (1999a). Regionalism and multilateral tariff cooperation. In
International Trade Policy and the Pacific Rim, ed. J Piggott, A Woodland, pp. 157-185. New
York: St. Martin‟s Press.
Balboa, Jenny (2008). “Negotiated Trade Liberalization in East Asia: Examining Japan‟s
Economic Partnership Agreement (EPA): Focusing on the Japan Philippines Economic
Partnership Agreement (JPEPA)”, Philippine APEC Study Center Network, Philippines
Institute for Development Studies working paper.
Baldwin, Richard (1993). „A domino theory of regionalism‟, NBER WP, 4465, published in
Baldwin, Haaparanta and Kiander (eds.), Expanding Membership in the European Union,
Cambridge University Press.
Baldwin, Richard (1994). Towards and integrated Europe, CEPR, London.
Baldwin, Richard (2002). “Asian Regionalism: Promises and Pitfalls,” in East Asian
Economic Regionalism: Feasibilities and Challenges, C. Ahn, R.Baldwin, and I. Cheong
(eds.), Springer, April 2005.
Baldwin, Richard (2004). “The spoke trap,” CNAEC Research Series 04-02, KIEP, Seoul.
Baldwin, Richard (2006a). “Globalisation: the great unbundling(s),” Chapter 1, in
Globalisation challenges for Europe, Secretariat of the Economic Council, Finnish Prime
Minister‟s Office, Helsinki, 2006, pp 5-47.
Baldwin, Richard (2006b). “Managing the Noodle Bowl: The fragility of East Asian
Regionalism,” CEPR DP 5561, eventually published in Singapore Economic Review, vol. 53,
issue 03, pages 449-478, April 2008.
Baldwin, Richard (2008). “The WTO tipping point”, VoxEU.org, 1 July 2008;
Baldwin, Richard (2010) "Understanding the GATT's wins and WTO's woes", CEPR Policy
Insight No.49, 5 June.
Baldwin, Richard and Robert Baldwin (1996). "Alternate approaches to the political economy
of endogenous trade liberalization”, with Robert E. Baldwin, European Economic Review,
40, pp 775-782.
Baldwin, Richard, Philippe Martin (1999). "Two waves of globalisation: Superficial
similarities and fundamental differences”, in H.Siebert (ed.), Globalisation and Labour,
J.C.B. Mohr for Kiel Institute, Tubingen, 1999. Also NBER WP 6904.
- 35 -
Baldwin, Robert (1970). Non-Tariff Distortions of International Trade. Washington, D.C.:
The Brookings Institution.
Baldwin, Robert (1980). “The Economics of the GATT” in Issues in International
Economics, edited by Peter Oppenheimer. Stocksfield, England and Boston: Oriel.
Bhagwati, Jagdish (1991). The World Trading System at Risk, Harvester Wheatsheaf, Hemel
Bhagwati, Jagdish (1993). “Regionalism and multilateralism: an overview,” in J. de Melo and
A. Panagariya (eds), New Dimensions in Regional Integration, Cambridge University Press,
Bond, Eric W., and Constantinos Syropoulos (1996a). “The size of trading blocs: Market
power and world welfare effects,” Journal of International Economics, Vol. 40, pp. 411–437.
Bond, Eric W., and Constantinos Syropoulos (1996b). “Trading Blocs and the Sustainability
of Inter-regional Cooperation,” forthcoming in Canzoneri, M., Ethier, W. and Grilli V. (eds.)
The New Transatlantic Economy, Cambridge University Press, Cambridge.
Bond, Eric W., and Constantinos Syropoulos and Alan Winters (1996). “Deepening of
Regional Integration and External Trade Relations,” CEPR Discussion Paper No. 1317.
Cadot, Olivier & Jaime de Melo & Marcelo Olarreaga, 2004. "Lobbying, Counterlobbying,
and the Structure of Tariff Protection in Poor and Rich Countries," World Bank Economic
Review, Oxford University Press, vol. 18(3), pages 345-366.
Carpenter, Theresa and Andreas Lendle (2010). “How Preferential is World Trade?” HEID
Working Paper 2010-32.
Coates, D., and Ludema, R., 2001, “A Theory of Trade Policy Leadership”, Journal of
Development Economics 65(1).
Conconi, Paola and Carlo Perroni (2010). “Conditional versus Unconditional Trade
Concessions for Developing Countries”, pdf dated September.
Cooper, Richard (1971). “Tariff issues and the Third World”, The World Today, Chatham
House publication, September.
Corden, W. M. (1966). "The Effective Protective Rate, the Uniform Tariff Equivalent, and
the Average Tariff," Economic Record 42, 200-216.
Dallas Fed (2002). “Maquiladora Industry: Past, Present and Future.” Federal Reserve Bank
of Dallas, El Paso Branch, Issue 2.
Egger, Peter, and V. Merlo. “The Impact of Bilateral Investment Treaties on FDI Dynamics.”
The World Economy, 30, 2007, 1536–1549.
Either, Willam (1996),'Multilateral roads to regionalism,' CIES DP, (University of Adelaide).
Ethier, William (2002). "Unilateralism in a Multilateral World," Economic Journal, Royal
Economic Society, vol. 112(479), pages 266-292.
Feenstra, Robert and Gordon Hanson (1997). "Foreign direct investment and relative wages:
Evidence from Mexico's maquiladoras," Journal of International Economics, Elsevier, vol.
42(3-4), pages 371-393, May.
Freund, Caroline & Ornelas, Emanuel, 2010. "Regional trade agreements," Policy Research
Working Paper Series 5314, The World Bank.
- 36 -
Freund, Caroline (2000a) “Different Paths to Free Trade: The Gains from Regionalism”.
Quarterly Journal of Economics 115:4, 1317-1341.
Freund, Caroline (2000b). “Multilateralism and the endogenous formation of preferential
trade agreements”, Journal of International Economics 52 (2000) 359–376.
Fugazza, Marco and Frédéric Robert-Nicoud (2010). "The Emulator Effect of the Uruguay
Round on US Regionalism," CEP Discussion Papers dp0973.
Garnaut, Ross (1991). Australia: A case study of unilateral trade liberalisation, in J. Bhagwati
and H. Patrick, Aggressive Unilateralism: America's 301 Trade Policy and the World Trading
System, University of Michigan Press.
Gaulier, Guillaume, Francoise Lemoine and Deniz Unal-Kesenci (2007). "China's emergence
and the reorganisation of trade flows in Asia," China Economic Review, vol. 18(3), pages
Grossman, G. and E. Helpman (1995). “The Politics of Free-Trade Agreements,” American
Economic Review 85 (1995b):667-690.
Grossman, Gene and Elhanan Helpman (2002). Interest Groups and Trade Policy , Princeton
and Oxford UK: Princeton University Press. Grubel, Herbert G.; Lloyd, Peter J. (1971). "The
Empirical Measurement of Intra-Industry Trade". Economic Record 47 (4): 494–517.
Haberler, Gottfried (1936). The theory of international trade with its applications to
commercial policy, W. Dodge & Company, New York.
Hillman, Arye and Peter Moser (1992). „Trade Liberalization as Politically Optimal
Exchange of Market Access‟ published in 1996 in The new transatlantic economy, Matthew
B. Canzoneri, Wilfred Ethier, Vittorio Grilli (eds), Cambridge University Press.
Horn, Henrik, Petros Mavroidis, and André Sapir (2010). "Beyond the WTO? An Anatomy
of EU and US Preferential Trade Agreements," The World Economy, Blackwell Publishing,
vol. 33(11), pages 1565-1588
Ishii, Jun and Kei-Mu Yi (1997). "The growth of world trade," Research Paper 9718, Federal
Reserve Bank of New York.
Johnson, Robert and Guillermo Noguera (2010). "Accounting for Intermediates: Production
Sharing and Trade in Value Added", manuscript.
Kimura, Fukunari, Yuya Takahashi, and Kazunobu Hayakawa (2007). “Fragmentation and
parts and components trade: Comparison between East Asia and Europe”, The North
American Journal of Economics and Finance, Volume 18, Issue 1, 1 Pages 23-40.
Krishna Pravin (2003). Are regional trading partners "natural"? J. Polit. Econ. 111:202-26
Krishna, Pravin (1998). “Regionalism and Multilateralism: A Political Economy Approach,”
Quarterly Journal of Economics, February 1998, Vol. 113, No. 1, Pages 227-250.
Krishna, Pravin and Mitra, Devashish (2008). "Reciprocated unilateralism in trade reforms
with majority voting," Journal of Development Economics, Elsevier, vol. 85(1-2), pages 81-
93. First circulated 2004 (NBER WP 10826).
Lawrence, Robert (1996): Regionalism, Multilateralism, and Deeper Integration (Brookings
Institution, Washington DC).
Levy, Philip, 1997, “A political-economic analysis of free-trade agreements,” American
Economic Review, Vol. 87, pp. 506–519.
- 37 -
Low, Patrick (1993). Trading free: The GATT and US trade policy, Twentieth Century Fund
Press, New York.
Ludema, Rodney, Anna Maria Mayda, and Prachi Mishra (2010). “Protection for Free? The
Political Economy of U.S. Tariff Suspensions”, June, EIIT Working Paper.
McLaren, John (2002). A theory of insidious regionalism. Q. J. Econ. 117:571-608
METI (2005). White Paper on International Economy and Trade 2005, Tokyo.
Moser, Peter (1990). The political economy of the GATT, University of St. Gallen PhD
thesis. St. Gallen.
O‟Rourke, Kevin and Jeffery Williamson (1999). “Transportation revolutions and commodity
market integration”, Chapter 3 in Globalisation and History, MIT Press, Cambridge.
Ornelas Emanuel (2005c).. Endogenous free trade agreements and the multilateral trading system.
Journal of International Economics 67:471-97
Ornelas, Emanuel (2005a). "Trade creating free trade areas and the undermining of
multilateralism," European Economic Review, vol. 49(7), pages 1717-1735, October.
Ornelas, Emanuel (2005b). "Rent Destruction and the Political Viability of Free Trade
Agreements," The Quarterly Journal of Economics, vol. 120(4), pages 1475-1506,
Panagariya, Arvind (2000). "Preferential Trade Liberalization: The Traditional Theory and
New Developments," Journal of Economic Literature, vol. 38(2), pages 287-331.
Richardson, Martin (2001) Unilateral Liberalisation in a Multilateral World, University of
Otago Economics Discussion Papers No. 0102, February 2001.
Roesseler, Freider (1978). “The rationale for reciprocity in trade agreements under floating
currencies”, Kyklos, Volume 31, Issue 2, pages 258–274, May.
Rostow, Walter (1960). The stages of Economic Growth: A non-communist manifesto,
Cambridge University Press.
Saggi K, Yildiz HM. 2009. Bilateralism, pure multilateralism, and the quest for global free
trade. Work. Pap., Southern Methodist University
Sally, Razeen (2008). “Globalisation and the Political Economy of Trade Liberalisation in the
BRIICS” pdf draft, http://www.ecipe.org/people/razeen-sally.
Sauvant, Karl and Lisa Sachs (2009). The Effect of Treaties on Foreign Direct Investment
Bilateral Investment Treaties, Double Taxation Treaties, and Investment Flows, Oxford
Summers, Larry (1991). "Regionalism and the World Trading System." Economic Review,
Federal Reserve Bank of Kansas City, issue Nov, pages 295–302.
UNCTAD (2010). World Investment Report, 2010. Geneva.
Vézina, Pierre-Louis (2010). "Race-To-The-Bottom Tariff Cutting," IHEID Working Papers
12-2010, The Graduate Institute of International Studies, July.
World Bank (2011). Doing Business 2011: Making a difference for entrepreneurs, World
Yi, S. (1996), 'Endogenous formation of customs unions under imperfect competition: Open
regionalism is good,' Journal of International Economics, 41, 151-175.
- 38 -
Young, Soogil (1996). “Political Economy of Trade Liberalization in East Asia", in ]effrey I.
Schott (ed.), The World Trading System: Challenges Ahead. Washington D.C.: Institute for