当前位置：文档库 › Economic Implications of Asian Integration

Economic Implications of Asian Integration

Global Economy Journal Volume8,Issue32008Article1

Economic Implications of Asian Integration Joseph F.Francois?Ganeshan Wignaraja?

?Johannes Kepler University Linz and CEPR,joseph.francois@jku.at

?Asian Development Bank,gwignaraja@https://www.wendangku.net/doc/c014575644.html,

Economic Implications of Asian Integration?

Joseph F.Francois and Ganeshan Wignaraja

Abstract

The Asian countries are once again focused on options for large,comprehensive regional integration schemes.In this paper we explore the implications of such broad-based regional trade initiatives in Asia,highlighting the bridging of the East and South Asian economies.We place emphasis on the alternative prospects for insider and outsider countries.We work with a global general equilibrium model of the world economy,benchmarked to a projected2017sets of trade and production patterns.We also work with gravity-model based estimates of trade costs linked to infrastructure,and of barriers to trade in services.Taking these estimates,along with tariffs,into our CGE model,we examine regionally narrow and broad agreements,all centered on extending the reach of ASEAN to include free trade agreements with combinations of the northeast Asian economies(PRC,Japan,Korea)and also the South Asian economies.We focus on a stylized FTA that includes goods,services,and some aspects of trade cost reduction through trade facilitation and related infrastructure improvements.What matters most for East Asia is that China,Japan,and Korea be brought into any scheme for deeper regional integration.This matter alone drives most of the income and trade effects in the East Asia region across all of our scenarios.The inclusion of the South Asian economies in a broader regional agreement sees gains for the East Asian and South Asian economies.Most of the East Asian gains follow directly from Indian participation. The other South Asian players thus stand to bene?t if India looks East and they are a part of the program,and to lose if they are not.Interestingly,we?nd that with the widest of agreements,the insiders bene?t substantively in terms of trade and income while the aggregate impact on outside countries is negligible.Broadly speaking,a pan-Asian regional agreement would appear to cover enough countries,with a great enough diversity in production and incomes,to actually allow for regional gains without substantive third-country losses.However,realizing such potential requires overcoming a proven regional tendency to circumscribe trade concessions with rules of origin, NTBs,and exclusion lists.The more likely outcome,a spider web of bilateral agreements,carries with it the prospect of signi?cant outsider costs(i.e.losses)both within and outside the region. KEYWORDS:regionalism,Asia FTAs,ASEAN,preferential trade,gravity model of services trade,trade costs and infrastructure

?This paper represents the opinions of the authors,and is not meant to represent the position or opinions of any organizations with which they are af?liated.Thanks are due to Rosechin Ol?ndo for ef?cient research assistance.

1. I NTRODUCTION

In the wake of extended delay in the Doha Round of WTO trade talks, a myriad of possible bilateral and regional combinations is now on the table. Indeed, even if WTO talks conclude successfully, it is unlikely they would yield any substantive impact on the general pattern of Asia protection patterns. (See Francois, van Meijl and van Tongeren 2005.) This has added more fuel to the fire driving Asian negotiations.

In this article, we offer a comprehensive examination of regional and sub-regional FTA pairings, inclusive not only of tariffs, but also of trade facilitation and services liberalization. This contrasts with the current literature, which is focused on goods. Both our facilitation and services experiments are built from gravity-based econometric estimates of trade costs. Because the collective impact of a set of FTAs can contrast sharply with what assessments of individual FTAs imply (Francois, McQueen and Wignaraja, 2005), the comprehensive approach offered here also provides insight into the differential impact of individual Asian FTAs, a collective wave of such FTAs, and a comprehensive regional approach to liberalization. We place emphasis on the alternative prospects for insider and outsider countries. Interestingly, we find that with a true Pan-Asian focus, the insiders benefit substantively in terms of trade and income while the aggregate impact on outside countries is negligible. Broadly speaking, an Asian regional FTA would appear to cover enough countries, with a great enough diversity in production and incomes, to actually allow for regional gains without a major price measured in substantive third-country losses. Realizing such potential though requires overcoming a proven regional tendency to circumscribe trade concessions with rules of origin, non-tariff barriers (NTBs), and exclusion lists. The more likely outcome, a spider web of bilateral agreements, carries with it the prospect of significant outsider costs both within and outside the region.

The article is organized as follows. As background, in Section 2 we

provide a summary of the existing literature on FTAs (actual and prospective) in the region. In Section 3 we offer an overview of the model and database. In Section 4 we spell out policy scenarios, linking them to the underlying patterns of production and trade. We conclude in Section 5.

2. B ACKGROUND

There is a growing body of literature on the impact of FTAs in Asia using global computable general equilibrium (CGE) models. This interest can be attributed to the proliferation of bilateral and plurilateral FTAs in Asia in recent years. Table 1 provides a broad overview of this literature. From the table, it can be seen that the focus of this research has been devoted to FTAs covering East Asian economies

Francois and Wignaraja: Economic Implications of Asian Integration

Published by The Berkeley Electronic Press, 2008

while the literature on FTAs involving South Asian economies or pan-Asian FTAs between East Asia and South Asian economies is more limited. This body of research raises question such as: will an East Asia FTA, or South Asian FTA or even a Pan-Asian FTA create gains for members or not? Will non-FTA members lose? And what sectors will gain or lose within members and non-members? There is currently intense debate in Asian policy circles on these questions and possible adjustment strategies needed to deal with countries and sectors that may lose though FTA formation.

By relying on a simulation approach that combines data and prospective

scenarios in a structured manner to analyze the economic effects of policy changes on due to the formation of an East Asia FTA, CGE models have emerged as an important tool for shedding light on these issues. The CGE models used have varied somewhat in their underlying economic structure, behavior of agents and focus, but while the theoretical structure varies, commonly these models build on the Global Trade Analysis Project (GTAP) database. The primary focus of such policy scenarios has been on the removal of price distortions against imports that arise from existing trade barriers (in particular merchandise tariffs). Most studies have used the standard GTAP model 1 with constant returns to scale in production, perfect competition, and the Armington assumption (or some variant of GTAP) while a few have adopted CGE models with firm-level imperfect competition.

Four major findings from the formation of an East Asian FTA emerge

from this literature (for a selection see Ballard and Cheong, 1997; Urata and Kyota, 2003; Gilbert et al. 2004; Lee et al., 2004; and Zhang and others, 2006): (a) all the East Asian countries involved would collect welfare gains; (b) the countries that are excluded are much more likely to suffer welfare losses; (c) production of sectors with a comparative advantage increases; and (d) an East Asian FTA is a step toward multilateral liberalization.

Studies, however, differ in their estimates of welfare gains to members

and losses to non members from an East Asia FTA depending on the type of CGE model used, data source and baseline year. An early study by Ballard and Cheong (1997), using a CGE model with firm-level imperfect competition, indicated that both an APEC FTA and an East Asian FTA would generate gains for all members even without the participation of the USA and Japan. They also estimate that developing nations in Asia are expected to gain more when the USA joins the FTA than when Japan joins. Urata and Kyota (2003) estimate that an East Asia FTA will generate welfare gains for members from the highest of 12.5 % of GDP

See Hertel (1997). For more details about the current standard GTAP model see https://www.wendangku.net/doc/c014575644.html,. For gravity-based analysis of Asian preferences, see Cabalu and Alfonso (2007) and Manchin and Pelkmans-Baloing (2007).

Global Economy Journal, Vol. 8 [2008], Iss. 3, Art. 1

https://www.wendangku.net/doc/c014575644.html,/gej/vol8/iss3/1

T ABLE 1. Selected Studies on the Impact of Asian FTAs

Study

Model

and

Baseline

Parameters and Assumptions Impact of FTA Scenarios

Ballard and Cheong (1997) GTAP

Model

Base year

of 1992

Uses 1994 GTAP database; data

disaggregated into 9 regions and

5 sectors.

Assumes removal of all tariffs

and non-tariff barriers between

all members of FTA. Uses two

models with different

assumptions: 1) perfect

competition (costs explained by

Armington assumption), and 2)

firm-level imperfect

competition.

Notes:

ASEAN countries include

Indonesia, Malaysia,

Philippines, and Thailand only.

Newly Industrialized Economies

(NIEs) include Hong

Kong,PRC; Korea; Singapore;

and Taipei,PRC.

East Asia FTA includes ASEAN

countries, PRC, NIEs, and

Japan.

Welfare effects of East Asian FTA

and APEC FTA, and global

liberalization scenarios from

perfectly-competitive model (EV as

% of GDP under each scenario):

?ASEAN

(0.54, 0.92, 1.41)

?PRC

(0.45, 1.40, 1.72)

?Japan

(0.02, 0.97, 1.08)

?Newly Industrialized Economies

(1.12, 3.72, 3.75)

?Rest of the world

(-0.02, -0.06, 0.33)

?USA

(-0.03, 0.13, 0.15)

Welfare effects of East Asian FTA,

APEC FTA, and global

liberalization scenarios from the

imperfectly-competitive model (EV

as % of GDP under each scenario):

?ASEAN

(3.49, 3.06, 2.36)

?PRC

(6.68, 3.07, 1.98)

?Japan

(-2.40, 2.71, 2.19)

?Newly Industrialized Economies

(7.58, 13.35, 12.78)

?Rest of the world

(-0.15, -0.29, 1.35)

?USA

(-0.13, 0.42, -0.04)

Francois and Wignaraja: Economic Implications of Asian Integration Published by The Berkeley Electronic Press, 2008

Study

Model

and

Baseline

Parameters and Assumptions Impact of FTA Scenarios

Urata and Kiyota (2003) GTAP

Model

Base year

of 1997

Uses GTAP database (version

5); data disaggregated into 20

countries and 21 sectors.

Assumes removal of tariff and

non-tariff barriers among East

Asian economies.

Notes:

East Asia FTA covers all East

Asian countries and economies.

Estimated effects of an East Asian

FTA (EV as % of GDP):

?Thailand (12.54)

?Viet Nam (6.61)

?Singapore (3.69)

?PRC (0.64)

?Japan (0.19)

?USA (-0.09)

?EU (-0.02)

Gilbert, Scollay and Bora (2004)GTAP

Model

Base year

of 1997

Uses the pre-release version of

GTAP database (version 5); data

disaggregated into 26 regions

and 20 commodities.

Assumes removal of all import

tariffs on a preferential basis

between members, with each

member maintaining its own

initial extra-RTA tariffs.

Notes:

APEC FTA assumes MFN

liberalization.

Welfare effects of PRC-Japan-Korea

FTA, ASEAN+3 FTA, and APEC

FTA (EV as % of GDP under each

scenario):

?Thailand (-0.2, 1.6, 1.0)

?Viet Nam (-0.6, 3.1, 4.8)

?Singapore (-0.2, 2.5, 1.9)

?PRC (0.0, 0.0, 0.2)

?Japan (0.1, 0.1, 0.4)

?Korea (0.7, 0.7, 0.7)

?USA (0.0, 0.0, 0.0)

?EU (0.0, 0.0, 0.1)

Lee, Roland-Holst and van der Mensbrugghe (2004)LINKAGE

Model

Base year

of 1997

Uses the GTAP database

(version 5.2); data disaggregated

into 9 regions and 18 sectors.

Assumes gradual removal of

bilateral tariffs and export

subsidies of the relevant sectors

among the member countries

over the period 2005-2010.

Notes:

ASEAN countries include

Indonesia, Malaysia,

Philippines, Singapore,

Thailand, and Viet Nam only.

ASEAN+3 does not include

Taipei,China.

Welfare effects of ASEAN+3 FTA

in 2015 (EV in $ Bn):

?ASEAN countries (41.8)

?PRC and Hong Kong,PRC

(102.3)

?Taipei,China (-5.4)

?Japan (66.3)

?Korea (30.1)

?USA (-0.9)

?EU-15 (6.8)

?Rest of the world (-9.8)

4 Global Economy Journal, Vol. 8 [2008], Iss. 3, Art. 1

https://www.wendangku.net/doc/c014575644.html,/gej/vol8/iss3/1

Study

Model

and

Baseline

Parameters and Assumptions Impact of FTA Scenarios

Zhang and others (2006) GTAP

Model

Base year

of 2001

Uses the GTAP database

(version 6); data disaggregated

into 87 regions and 57 sectors.

Assumes elimination of all tariff

and non-tariff barriers on trade

in goods and some trade

facilitation programs for the

member countries.

Notes:

FTA scenarios assume tariff and

NTB elimination for goods, and

trade facilitation.

East Asian FTA would increase

overall GDP of East Asian countries

by 1.2% and economic welfare by

$104.6 billion.

Welfare effects of East Asian FTA

(EV in $ Bn):

?ASEAN countries (37.6)

?PRC, Japan, and Korea (66.9)

?ASEAN+3 (104.6)

Bandara and Yu (2003) GTAP

Model

Base year

of 1997

Uses 1997 GTAP database; data

disaggregated into 12 regions

and 17 industries.

Performs two opposite policy

simulations: 1) unilateral trade

liberalization scenario assumes

removal of all import tariff and

export duties of all South Asian

countries; and 2) preferential

trade liberalization scenario

assumes removal of all tariffs

and export duties between South

Asian countries but not between

other regions.

Notes:

SAFTA scenario assumes 100%

tariff cut as opposed to actual

tariff concessions given by

SAFTA members during the

final round of tariff reductions

in 1998.

GTAP database disaggregates

South Asia into four regions:

India, Sri Lanka, Bangladesh,

and rest of South Asia.

Welfare effects under unilateral

trade liberalization scenario (EV in

$ Mn):

?ASEAN (94.6)

?Japan (438.7)

?India (2331.9)

?Sri Lanka (83.9)

?Bangladesh (173.9)

?Rest of South Asia (-511.1)

?NAFTA (2509.0)

?EU (1125.1)

Welfare effects under SAFTA,

South Asia-ASEAN, and

multilateral trade liberalization

scenario (EV in $ Mn under each

scenario):

?ASEAN

(-70.1, 3039.5, 7324.3)

?Japan

(-156.6, -33.3, 33638.1)

?India

(756.2, -1313.4, 3521.3)

?Sri Lanka (4.1, -29.8, 274.4)

?Bangladesh

(-41.2, -151.9, 288.9)

?Rest of South Asia

(52.3, -791.1, 96.7)

?NAFTA

(-113.9, -42.2, -6091.5)

?EU (-169.9, -396.4, 9097.4)

Francois and Wignaraja: Economic Implications of Asian Integration Published by The Berkeley Electronic Press, 2008

Study

Model

and

Baseline

Parameters and Assumptions Impact of FTA Scenarios

Mohanty and Roy (2004) GTAP

Model

Base year

of 1997

Uses GTAP database (version

5); data disaggregated into 14

regions and 26 sectors.

Simulates three scenarios: 1)

removal of tariff and non-tariff

barriers between Japan,

ASEAN, PRC, India, and Korea

(JACIK) countries; 2) free

movement of investments within

JACIK countries; and 3) free

movement of investments and

skilled labor within JACIK

countries.

Notes:

GTAP database disaggregates

South Asia into two regions:

India and rest of South Asia.

ASEAN includes Indonesia,

Malaysia, Philippines,

Singapore, and Thailand only.

Welfare effects under Scenario 1, 2,

and 3 (EV in $ Mn):

?Thailand

(4409.8, 4594.7, 5799.7)

?Singapore

(2292.5, 1786.7, 1741.4)

?Indonesia

(3760.3, 3993.9, 6968.1)

?PRC

(6326.5, 7100.0, 16327.7)

?Japan

(107625.7, 111807.0, 150695.2)

?Korea

(13042.9, 13317.4, 14075.7)

?India

(6971.3, 7378.6, 9937.0)

?JACIK (147417.6, 153155.7,

210440.9)

?Rest of South Asia (not shown)

Plummer and Wignaraja (2006) GEMAT

Model

Base year

of 2001

Uses GTAP database (version

6); data disaggregated into 19

countries and 14 sectors.

Assumes removal of tariff

barriers between FTA members.

Notes:

ASEAN includes Indonesia,

Malaysia, Philippines,

Singapore, Thailand, and Viet

Nam only.

Welfare effects of multiple bilateral

FTAs in Asia, Asia-wide FTA, and

APEC FTA scenarios (EV in $ Mn

under each scenario):

?ASEAN

(8869, 10907, 8341)

?Northeast Asia

(-1219, 35713, 56734)

?Rest of Asia

(-101, 1355, -1560)

?USA

( -1371, 3263, 12035)

?EU

(-1021, -1413, -3047)

6 Global Economy Journal, Vol. 8 [2008], Iss. 3, Art. 1

https://www.wendangku.net/doc/c014575644.html,/gej/vol8/iss3/1

Study

Model

and

Baseline

Parameters and Assumptions Impact of FTA Scenarios

Siriwardana (2003) GTAP

Model

Base year

of 1997

Uses GTAP database (version

5); data disaggregated into 11

regions and 20 sectors.

Performs two liberalization

scenarios: 1) assumes removal

of all bilateral tariffs between

South Asian countries but not

between other countries; 2)

creation of customs union by

eliminating all tariffs between

South Asian regions and

adopting common external tariff

against all other countries in the

world.

Notes:

GTAP database disaggregates

South Asia into four regions:

India, Sri Lanka, Bangladesh,

and rest of South Asia.

Welfare effects under South Asia

FTA and South Asia Customs Union

scenarios (EV in $ Mn):

?PRC

(-680.55, -743.40)

?Japan

(-4008.51, -4111.84)

?India

(3046.62, 4995.84)

?Sri Lanka

(261.96, 1466.11)

?Bangladesh

(90.47, 1043.15)

?Rest of South Asia

(579.83, 4062.39)

?NAFTA

(-6434.40, -24333.60)

?EU

(-6434.40, -18950.50)

Bchir and Fouquin (2006) MIRAGE

Model

Uses GTAP database

(version 6).

Performs two policy

experiments: 1) assumes that

ASEAN removes its tariffs

bilaterally with PRC, India,

Japan, and Korea (with and

without exclusion of sensitive

products); and 2) assumes

removal of tariffs between

ASEAN, PRC, India, Japan, and

Korea (with or without

exclusion of sensitive products).

Welfare effects under bilateral

agreements between ASEAN and

PRC, India, Japan, and Korea in

2015 (EV as % change):

?ASEAN (2.18)

?PRC (-0.12)

?Japan (0.18)

?Korea (-0.40)

?India (-0.32)

?South Asia (-0.05)

?USA (0.00)

?EU-25 (-0.01)

Welfare effects under a single global

agreement between ASEAN and

PRC, India, Japan, and Korea in

2015, (EV as %):

?ASEAN (1.43)

?PRC (-0.27)

?Japan (0.41)

?Korea (1.64)

?India (-0.37)

?South Asia (-0.12)

?USA (0.00)

?EU-25 (-0.02)

Francois and Wignaraja: Economic Implications of Asian Integration Published by The Berkeley Electronic Press, 2008

for Thailand and 6.6% for Vietnam to the lowest of 0.19% for Japan and 0.64% for the PRC. They find modest welfare loses for non-members of -0.02% for the EU, -0.09% for the USA and -0.29% for Australia/New Zealand. Also using a GTAP-based model, Gilbert et al. (2004) find that an East Asia FTA will produce higher welfare gains for members than a narrower PRC-Japan-Korea FTA indicating that broadening FTAs brings benefits. They report lower welfare gains from an East Asia FTA for Vietnam (3.1%) and Thailand (1.6%) than Urata and Kyota (2003). Most recently, Zhang and others (2006) report GTAP simulations confirming the common results that all members gain from an East Asian FTA. They estimate that such and FTA would increase the overall GDP of East Asian countries by 1.2% and economic welfare by $104.6 billion. From their LINKAGE CGE model, Lee et al. (2004) show significantly higher welfare gains from an East Asia FTA for PRC+Hong Kong (4%) and Japan (1.6%), notable gains for Korea (3.7%) and ASEAN as a group (4%) and welfare losses for the rest of the world of under -0.2%.

By comparison, the available studies suggest mixed views about the

impact of an FTA involving only South Asian economies and one between selected East Asian and South Asia countries. Using GTAP, Siriwardena (2003) compares the effects of an FTA and a customs union for South Asian countries. He finds that the South Asian FTA scenario (with full trade liberalization internally) brings gains to all members and loses to non-members but that the customs union entails bigger gains for members as well as bigger loses to non-members. Not surprisingly perhaps, the region’s largest and most competitive economy, India gains the most ($3.1 billion in the FTA scenario). However, Bandara and Wu (also using GTAP) find lower gains for India ($756 million) from a South Asia FTA scenario, negligible gains for Sri Lanka and the rest of South Asia, and losses for Bangladesh. Likewise, Bandara and Yu (2003) provide a pessimistic assessment of an ASEAN-South Asia FTA. ASEAN as a whole is likely to see modest gains ($3 billion) and all the South Asian economies including India incur welfare loses. Non-members (e.g. EU and USA) also lose.

With an opposite result, Mohanty, Pohit and Roy (2004) argue that an East

Asia-India FTA (i.e., ASEAN+3 and India FTA which they call JACIK) will bring gains to members of between $147.4 billion (liberalization of trade barriers only scenario) to $210.4 billion (liberalization of barriers to trade, investment and labor). In their scenarios all members benefit, with Japan witnessing the largest gains ($108 billion), PRC and India (under $7 billion each) and Philippines the least ($1 billion). Interestingly, the authors do not provide details of how the normally technically difficult barriers to investment and labor are incorporated into their model. Nor do they provide estimates for the effects of JACIK on non-members. The work of Bchir and Fouquin (2006) on an East Asia-India FTA, relying on the MIRAGE CGE (also built around GTAP), suggest that non-8

Global Economy Journal, Vol. 8 [2008], Iss. 3, Art. 1

https://www.wendangku.net/doc/c014575644.html,/gej/vol8/iss3/1

members see small loses ranging from -0.02% for the EU -0.12 for the rest of South Asia and -0.16 for Russia. Interestingly, they also find that Asia’s giant economies lose from an East Asia-India FTA while ASEAN, Japan and Korea gain. The losers include both the PRC (-0.27%) and India (-0.37).

Finally, drawing on a GEMAT CGE model (a variant of the LINKAGE

model), Plummer and Wignaraja (2006) investigate the relative economic effects of various possible FTA scenarios – a fragmented scenario of bilateral FTAs and ASEAN to depict the current East Asian policy reality, an Asia-wide FTA (including Northeast Asia, ASEAN and South Asia) and an APEC FTA. Compared to the others, the fragmented FTA scenario leads to lower welfare for all. An Asia-wide FTA generates gains of $48 billion for the region and all members gain but Northeast Asian economies gain disproportionately. Meanwhile, the APEC FTA generates larger gains of 64 billion for Asia. As expected, Northeast Asia and US members gain but non-members like South Asia and the EU lose.

3. T HE M ODEL AND D ATA

We turn to a brief overview of the global CGE model used here. As is standard in the literature, the model is characterized by a global input-output structure (based on regional and national input-output tables) that explicitly links industries in a value added chain from primary goods, over continuously higher stages of intermediate processing, to the final assembling of goods and services for consumption. Inter-sectoral linkages are direct, like the input of steel in the production of transport equipment, and indirect, via intermediate use in other sectors. The model captures these linkages by modeling firms' use of factors and intermediate inputs. In terms of structure, the model is a version of the basic one employed by Francois, van Meijl, and van Tongeren (2005) to assess the Doha Round. The data, however, reflect a more current (and projected) economic landscape. The most important aspects of the model can be summarized as follows: (i) it covers all world trade and production; (ii) it includes intermediate linkages between sectors; (iii) and it allows for trade to affect capital stocks through investment effects. The last point means we model medium to long-run investment effects. (See Francois, McDonald and Nordstrom 1999.)

3.1 M ODEL D ATA AND THE B ENCHMARK

Our data come from a number of sources. Data on production and trade are based on national social accounting data linked through trade flows (see Reinert and Roland-Holst 1997). These social accounting data are drawn directly from the GTAP dataset, version 6.3. (Dimaranan and McDougall, 2002). The GTAP

Francois and Wignaraja: Economic Implications of Asian Integration

Published by The Berkeley Electronic Press, 2008

version 6 dataset is benchmarked to 2001, and includes detailed national input-output, trade, and final demand structures. The basic social accounting and trade data are supplemented with trade policy data, including additional data on tariffs and non-tariff barriers. We have projected the database to 2007, and through to 2017, using macroeconomic projections from the World Bank (circulated through the GTAP consortium) combined with macroeconomic outlook data from the IMF.2

The 2007 projection includes the phase-out of the Agreement on Textiles

and Clothing (ATC) quotas in 2005, as well as remaining WTO commitments under the Doha Round and the enlargement of the EU from 15 to 27 Members. The data on tariffs are taken from the WTO's integrated database, with supplemental information from the World Bank's recent assessment of detailed pre- and post-Uruguay Round tariff schedules and from the UNCTAD/World Bank WITS dataset. All of this tariff information has been mapped to activity (GTAP) sectors. Services trade barriers are based on the gravity model estimates described in the annex. These estimates are also discussed in the next section. We also work with the schedule of PRC WTO accession commitments. While the basic GTAP dataset is benchmarked to 2001, and reflects applied tariffs actually in place in 2001, we of course want to work with a representation of a post-Uruguay Round world. We also want to include the accession of PRC, the enlargement of the EU, as part of the baseline. Our 2017 projection is based on the 2007 policy baseline. The social accounting data have been aggregated to 35 sectors and 36 regions. The sectors and regions for the 35x36 aggregation of the data are given in Table 2.

3.2 T HEORETICAL STRUCTURE

We turn next to the basic theoretical features of the model. In all regions there is a single representative, composite household in each region, with expenditures allocated over personal consumption and savings (future consumption) and over government expenditures. The composite household owns endowments of the factors of production and receives income by selling them to firms. It also receives income from tariff revenue and rents accruing from import/export quota licenses (when applicable). Part of the income is distributed as subsidy payments to some sectors, primarily in agriculture.

Projection involves imposing changes in labor force and capital stocks, as well as World Bank/IMF projections for national income growth. A set of Hicks-neutral productivity parameters are then solved for, consistent with these macroeconomic projections. Relevant policy changes (like tariff changes linked to China’s accession to the WTO, and the ATC phaseout) are also included in the database projections.

Global Economy Journal, Vol. 8 [2008], Iss. 3, Art. 1

https://www.wendangku.net/doc/c014575644.html,/gej/vol8/iss3/1

T ABLE 2.

Model sectoring scheme

Regions

Sectors

Australia New Zealand Other Oceania PRC

Hong Kong, China Japan Korea

Taipei,China Other East Asia Indonesia Malaysia Philippines Singapore Thailand Vietnam

Other Southeast Asia Bangladesh Cambodia India Pakistan Sri Lanka

Other South Asia Central Asia Canada

United States Mexico

Latin America EU27 EFTA Turkey Russia

Other Europe

North Africa & Middle East South Africa

Sub-Saharan Africa

Grains

Horticulture Oil Seeds Sugar Cotton Beef Dairy

Vegetable Oils

Other Primary Agriculture Other Processed Foods Beverages & Tobacco Forestry Fisheries Mining Textiles Clothing Leather

Paper, Pulp, Printing Petrochemicals

Chemicals, Rubber, Plastics Iron and Steel

Non-Ferrous Metals Motor Vehicles

Electrical Machinery Other Machinery Other Manufactures Utilities Construction Trade Services Transport Services Communications Financial Services Insurance

Other Business Services Other Services

On the production side, in all sectors, firms employ domestic production

factors (capital, labor and land) and intermediate inputs from domestic and foreign sources to produce outputs in the most cost-efficient way that technology allows. Perfect competition is assumed in production sectors, where products from different regions are assumed to be imperfect substitutes in accordance with the so-called "Armington" assumption.

Francois and Wignaraja: Economic Implications of Asian Integration

Published by The Berkeley Electronic Press, 2008

Prices on goods and factors adjust until all markets are simultaneously in

(general) equilibrium. This means that we solve for equilibriums in which all markets clear. While we model changes in gross trade flows, we do not model changes in net international capital flows. Rather our capital market closure involves fixed net capital inflows and outflows. This does not preclude changes in gross capital flows.3 To summarize, factor markets are competitive, and labor and capital are mobile between sectors but not between regions. All primary factors, labor, land and capital are fully employed within each region.

We also include a dynamic link, whereby changes in investment,

following from policy changes, lead to changes in installed capital stocks and hence ultimately to production and trade volumes. This is based on the Solow model-based approach as outlined in Francois, McDonald, and Nordstrom (1999). Conceptually, as we are working with a projected baseline, these dynamic effects can be though of as including induced investment effects along an alternative path to the 2017 benchmark, wherein we have implemented the policy changes in time for investment effects to be realized in the 2017 equilibrium.

4. P OLICY L ANDSCAPE , S CENARIOS , AND D ISCUSSION

We next turn to our analysis of regional integration initiatives between East Asia and South Asia. This includes a broad overview of trade structure and policy in the region, our scenarios, and the impact of those scenarios on our baseline.

4.1 T RADE AND TRADE POLICY IN THE REGION

The regional share of international trade in Asia varies widely across regions. This was true in 2001, and also holds in our projected 2007 and 2017 benchmarks. Figure 1 provides a picture of this pattern for the countries of the region, while more detailed information is provided in Tables 3 and 4. The countries of East Asia (including Northeast and Southeast Asia) are in general much more integrated, in a regional sense, than are the countries of South Asia. Indeed, the difference is striking. For most countries in the region, more than 40% of their exports are destined for Asian markets (see Figure 1). From Table 3, it is clear that most of this trade is destined for East Asia. Indeed, for many countries in the region, this share is projected to rise. Oceania (primarily Australia and New Zealand), while not technically part of the region geographically, is closely tied economically, with the vast majority of its exports going to the region. With projected economic growth through 2017, this dependence only increases.

See the Hertel el al (1997) discussion on macroeconomic closure, especially in this class of model. The present approach facilitates welfare analysis.

Global Economy Journal, Vol. 8 [2008], Iss. 3, Art. 1

https://www.wendangku.net/doc/c014575644.html,/gej/vol8/iss3/1

T ABLE 3

Note: “East Asia” in the table includes both Northeast Asia (PRC, Japan, Taiwan, Korea) and Southeast Asia.

In contrast to East Asia, South Asia economies are much less closely tied

to their own region. Trade shares with Asia are generally well below 30 percent. From Tables 3 and 5, it is also clear that most of this trade is not actually with South Asia. Ironically, while there have been regional initiatives in South Asia, and not so much between South and East Asia, the bulk of South Asian regional exports go to East Asia rather than to South Asia. These results point to both a relatively low degree of integration within the region, and also to the potential for gains from liberalization initiatives that span the two sub-regions.

Table 5 provides a breakdown of MFN protection as of 2004/5 (from the

WITS database 4) on a trade-weighted basis. This provides a sense of the scope for gains from liberalization in merchandise trade. In general, import protection is higher in South Asia than in East Asia. This explains part of why East Asian trade relationships are deeper, and also why South Asian trade is also biased to

Available at https://www.wendangku.net/doc/c014575644.html,.

Direction of Exports

East Asia

South Asia East Asia South Asia East Asia South Asia Australia

0.4600.0290.4990.0370.6140.059New Zealand 0.3360.0130.3720.0190.4400.029Other Oceania 0.3360.0080.3600.0080.4250.008China

0.3650.0140.3480.0130.3360.012Hong Kong, China 0.3890.0140.4220.0130.4470.012Japan 0.3790.0100.3880.0090.4020.009Korea

0.4090.0190.4180.0170.4130.015Chinese Taipei 0.4310.0160.4460.0160.4630.016Other East Asia 0.2670.0150.2930.0150.3540.012Indonesia 0.4820.0380.4600.0360.4910.040Malaysia 0.4700.0340.4860.0350.5100.045Philippines 0.4720.0040.4780.0040.4610.004Singapore 0.4700.0390.4910.0350.5060.029Thailand 0.4340.0210.4420.0190.4590.016Vietnam 0.4170.0050.4380.0050.4650.004Cambodia

0.1450.0060.1650.0060.1550.006Other Southeast Asia 0.5270.0750.5480.0600.5960.041Bangladesh 0.0680.0140.0830.0160.0730.015India 0.2060.0390.2160.0370.2490.032Pakistan 0.1650.0420.1260.0320.1100.028Sri Lanka

0.1000.0310.1040.0330.0790.030Other South Asia

0.127

0.2050.1200.2220.1220.309

20072017

2001

Francois and Wignaraja: Economic Implications of Asian Integration

Published by The Berkeley Electronic Press, 2008

T ABLE 4

Note: “East Asia” in the table includes both Northeast Asia (PRC, Japan, Taiwan, Korea) and Southeast Asia.

East Asia. Of course, the relative size of the economies in the two regions also helps to explain this regional bias.

Table 6 provides a similar picture, only for services. This is based on our

estimates of services trade barriers for cross-border trade, as discussed in the annex. Unlike goods, in services there is not that much regional difference in the pattern of protection. In general, our estimates are that protection is much higher for goods than for services, and that this holds for countries in East Asia as well as South Asia.

The broad picture that emerges from this overview of the trade and trade

protection data is that East Asia is more integrated than South Asia, that South Asia itself has deeper trade ties with East Asia than with itself, and that import protection for merchandise explains part of this pattern. We now turn to an assessment of a set of stylized regional integration schemes.

Source of Imports

East Asia

South Asia East Asia South

Asia East Asia South

Asia Australia

0.3910.0110.4150.0130.4680.018New Zealand 0.2660.0110.2910.0110.3660.016Other Oceania 0.3030.0100.3260.0120.3980.016China

0.5370.0100.5320.0110.5160.015Hong Kong, China 0.6340.0130.6390.0160.6510.024Japan 0.3800.0100.4070.0110.4530.012Korea

0.4000.0110.4040.0110.4090.012Chinese Taipei 0.5110.0070.5280.0070.5670.010Other East Asia 0.6180.0260.6190.0270.6280.034Indonesia 0.4910.0210.4930.0220.5040.025Malaysia 0.5740.0150.5870.0170.6140.021Philippines 0.5340.0100.5310.0100.5610.017Singapore 0.5420.0130.5850.0150.6560.023Thailand 0.5270.0170.5450.0180.5750.026Vietnam 0.5180.0160.5210.0170.5360.021Cambodia

0.8440.0110.8470.0120.8320.015Other Southeast Asia 0.7550.0200.7680.0220.7750.026Bangladesh 0.5020.1370.5010.1400.4980.160India 0.2780.0130.2810.0140.2860.016Pakistan 0.2800.0330.2780.0310.2580.031Sri Lanka

0.4370.1150.4400.1180.4570.139Other South Asia

0.353

0.1870.347

0.1930.359

0.217

2001

2007

2017

Global Economy Journal, Vol. 8 [2008], Iss. 3, Art. 1

https://www.wendangku.net/doc/c014575644.html,/gej/vol8/iss3/1

F IGURE 1

Francois and Wignaraja: Economic Implications of Asian Integration

Published by The Berkeley Electronic Press, 2008

T ABLE 5

T ABLE 6

China 4.9 5.30.1India 13.915.40.3Sri Lanka 7.37.90.1Nepal 14.614.70.2Pakistan 12.213.10.2Singapore 0.00.00.2Thailand 5.4 6.50.2

source: WITS integrated database.

Services Trade Barriers

PRODUCER SERVICES

OTHER NON-TRADE SERVICES

Hong Kong, China 00Japan 27.020.6Korea

15.720.6Chinese Taipei 14.310.7Other East Asia 11.910.8Indonesia 12.512.2Malaysia 8.39.3Singapore 0.00.0Philippines 8.312.8Thailand 6.6 5.6Vietnam 20.836.1Cambodia

20.836.1Other Southeast Asia 46.346.4Bangladesh 25.229.0India 26.032.6Pakistan 30.034.5Sri Lanka

22.020.9Other South Asia

7.49.4

source: author estimates

Global Economy Journal, Vol. 8 [2008], Iss. 3, Art. 1

https://www.wendangku.net/doc/c014575644.html,/gej/vol8/iss3/1

4.2 S CENARIOS

In most of what follows we examine three cores scenarios as follows:

? Scenario 1 – An ASEAN+3 FTA - involves a regional free trade

agreement between the members of the ASEAN block and PRC, Japan, and Korea (i.e. an East Asia FTA).

? Scenario 2 – ASEAN+3 and India-FTA – extends Scenario 1’s free trade

agreement to include India (i.e an East Asia and India FTA). Hence, South Asia’s largest economy is included but the rest of South Asia is excluded.

? Scenario 3 – ASEAN+3 and South Asia FTA extends Scenario 2’s free

trade agreement further to include the rest of South Asia and implements full free trade across South Asia itself (i.e. an East Asia-South Asia FTA ).

We will also (in less detail) look at sub-regional scenarios involving South Asia. All three of the core scenarios involve free trade in merchandise goods (i.e. tariffs as represented in Table 5), free trade in services (based on estimates in Table 6), and trade cost reductions equal to 2.5 percent of the cost of trade. Trade cost reductions can follow from trade facilitation measures that streamline the administrative cost of clearing goods across borders. Recent estimates place these costs at anywhere from 6 percent to 30 percent of the costs of goods traded (Francois, Hoekman and Manchin 2006; Manchin and Pelkmans 2007: Manchin 2006). They can also follow from improvements to trade-related infrastructure. Indeed, recent estimates suggest that for North-South trade, variations in trade-related infrastructure explain more of the sample variations in goods trade than does trade policy itself (Francois and Manchin 2007). Table 7 reports estimates of the trade cost savings that would follow from a 1% and 5% improvement in the general quality of trade-related infrastructure, based on Francois, Manchin, and Pelkmans (2007). From the estimates in the table, a 5% improvement would yield a 2.5% trade cost savings, on average. Broadly speaking, the 2.5% trade cost reduction in our scenarios is meant to capture a regional trade initiative that includes both administrative improvements (so that goods move more quickly and with less paperwork) and some investment in physical infrastructure in the poorer countries in the region. Japan, in particular, has emphasized the infrastructure potential of regional schemes. From Manchin and Pelkmans (2006), this seems a conservative estimates of the benefits from a simple streamlining of administrative barriers in the region, let alone other trade cost reduction measures.

Francois and Wignaraja: Economic Implications of Asian Integration

Published by The Berkeley Electronic Press, 2008

T ABLE 7

4.3 B ROAD W ELFARE AND T RADE E FFECTS

Tables 8-17 summarize the results of our three core experiments. All results are reported relative to the 2017 baseline simulation. Broadly speaking, the scenario with the widest regional FTA coverage implies global income gains of $260.9 billion in 2001 dollars, or approximately 0.5 percent of global income. This follows from a $263.9 billion gain for insiders, and a loss of $3 billion for outsiders. Interestingly, for the narrower FTAs there are losses for South Asian economies in the range of -0.3 to -0.5 percent of GDP, while for regional outsiders in all cases the extra-regional losses are generally quite small. This suggests a pattern that we will see in the sections on sector effects and on regional direction of trade, of apparent dominance of the results by trade creation and gains from trade, rather than trade diversion and losses, under our broad FTA scenario.

North-South Sample 0.150.280.43 2.15

country estimates Australia 0.250.060.32 1.58Bangladesh 0.180.550.73 3.65Cambodia 0.180.550.72 3.62China 0.190.480.67 3.36Hong Kong, China 0.250.060.31 1.55India 0.180.530.71 3.56Indonesia 0.200.430.62 3.12Japan 0.27-0.020.25 1.23Korea, Rep.0.240.140.38 1.90Lao PDR 0.180.530.71 3.57Malaysia 0.220.260.48 2.40New Zealand 0.250.090.34 1.70Pakistan 0.180.500.69 3.45Philippines 0.200.410.61 3.04Singapore 0.250.050.31 1.54Thailand 0.210.310.53 2.63Vietnam 0.180.570.74 3.71source: Francois, Manchin, and Pelkmans-Baloing (2007)

Global Economy Journal, Vol. 8 [2008], Iss. 3, Art. 1

https://www.wendangku.net/doc/c014575644.html,/gej/vol8/iss3/1