The changing geography of world trade: Projections to 2030

1. Introduction

Rapid economic growth in Asian and other emerging economies is shifting the global economic and industrial centre of

Journal of Asian Economics 23 (2012) 303323

Article history:

Received 5 April 2011

Received in revised form 17 October 2011

Accepted 26 February 2012

Available online 3 March 2012

JEL classication:

D58

F13

F15

F17

Q17

Keywords:

Global economy-wide model projections

Asian economic growth

SouthSouth trade

Intra- and extra-regional trade intensity and

propensity indexes

Rapid economic growth in Asia (and some other emerging economies) has been shifting

the global economic and industrial centres of gravity away from the north Atlantic, raising

the importance of Asia in world trade, and boosting SouthSouth trade. This paper

examines how trade patterns are likely to change in the course of continuing economic

growth and structural changes in Asia and the rest of the world over the next two decades.

It does so by projecting a core baseline for the world economy from 2004 to 2030 and

comparing it with alternative scenarios, including slower economic growth rates in the

North, slower productivity growth in primary sectors, and prospective trade policy

reforms in Developing Asia, without and with policy reforms also in the North and in

SouthSouth trade. Projected impacts on international trade patterns, sectoral shares of

GDP, openness to trade, and potential welfare gains from reforms are highlighted, in

addition to effects on bilateral trade patterns as summarized by intra- and extra-regional

trade intensity and propensity indexes. The paper concludes with implications for regional

and multilateral trade policy.

2012 Elsevier Inc. All rights reserved.

Contents lists available at SciVerse ScienceDirect

Journal of Asian Economicsgravity away from the north Atlantic, and accelerated globalization is causing trade to grow much faster than output,especially in Asia. Together these forces are raising the importance of the emerging countries in world output and trade, aswell as boosting SouthSouth trade. Asias share of global merchandise trade has doubled since 1973, to just over 30 percent,with its exports growing at three times the rate for the rest of the world over the past decade; and China is now the worldslargest exporter, followed by Germany and the United States (WTO, 2010).

This paper explores how trade patterns might change over the next two decades in the course of economic growth andstructural changes in rapidly emerging economies and the rest of the world. In particular, it examines possible changes in theimportance of Developing Asias intra-regional trade and its trade with other developing country regions under variousscenarios. It does so using indexes of intra- and extra-regional trade intensity and propensity (as dened by Anderson &The changing geography of world trade: Projections to 2030

Kym Anderson a,b,*, Anna Strutt c,a

aUniversity of Adelaide, AustraliabCEPR, United KingdomcUniversity of Waikato, New Zealand

A R T I C L E I N F O A B S T R A C TNorheim, 1993) in additional to bilateral trade shares.

* Corresponding author at: School of Economics, University of Adelaide, Adelaide, SA 5005, Australia. Tel.: +61 8 8303 4712; fax: +61 8 8223 1460.

E-mail addresses: [email protected] (K. Anderson), [email protected] (A. Strutt).

1049-0078/$ see front matter 2012 Elsevier Inc. All rights reserved.doi:10.1016/j.asieco.2012.02.001

There are several new speculative studies on the future of growth convergence, from both academics (e.g., Rodrik, 2011;Spence, 2011) and major consulting rms (e.g., Citi, 2011; PwC, 2011). The present study is more formal in the sense that itmakes use of neoclassical trade and development theory together with historical experience insofar as they are captured inthe equations and behavioural parameters of a standard applied general equilibrium model of the global economy. The paperbegins by outlining the economy-wide modelling methodology and database we use in projecting world markets from 2004to 2030. It then presents results that emerge from our core projection, followed by variants on that according to alteredassumptions about growth rates and trade policies. Some caveats are then presented, before the nal section drawsimplications (a) for policies that can affect endowment, productivity and income growth rates and (b) for regional and global

K. Anderson, A. Strutt / Journal of Asian Economics 23 (2012) 303323304trade policies.

2. Modeling methodology and database

In this study we employ the GTAP model (Hertel, 1997) of the global economy and Version 7.1 of the GTAP database whichis calibrated to 2004 levels of production, consumption, trade and protection (Narayanan & Walmsley, 2008). The GTAPmodel is perhaps the most widely used model for economy-wide global market analysis, in part due to its robust and explicitassumptions. It is based on neoclassical general equilibrium theory that ensures changes in trade patterns will be a functionof changes in one or more of the following: (i) factor endowments, (ii) technologies of production, (iii) demand patterns, (iv)infrastructure affecting determinants of trade costs such as transport and communications, and (v) price-distorting policies(see, e.g., Leamer, 1987; Limao & Vanables, 2001). Economic growth is generated by factor accumulation and technologicalimprovements in production and distribution, and the resulting growth in incomes affects demands for various productsaccording to the nonhomothetic preferences of consumers. Previous empirical validation work using the GTAP model (e.g.,Gelhar, Hertel, & Martin, 1994), and a survey of econometric research by Martin (2007) suggest changes in factorendowments are more important than changes in demand. However, when very populous, resource-poor economies rapidlyindustrialize, as with China and India, their import demand for raw materials might be expected to play a signicant role inchanging comparative advantages.

In its simplest form, the GTAP model assumes perfect competition and constant returns to scale in production. Thefunctional forms are nested constant elasticities of substitution (CES) production functions. Land and other naturalresources, labor (skilled and unskilled), and produced physical capital substitute for one another in a value added aggregate,and composite intermediate inputs substitute for value-added at the next CES level in xed proportions. Land is specic toagriculture in the GTAP database, and is mobile amongst alternative agricultural uses over this projection period, accordingto a relatively high Constant Elasticity of Transformation (CET) which, through a revenue function, transforms land from oneuse to another. In the modied version of the GTAP model we use, natural resources including coal, oil and gas, are specic tothe sector in which they are mined. Aggregate regional employment of each productive factor is xed in the standard macro-economic closure, though we use exogenous projections to model changes in factor availability over time. Labor andproduced capital are assumed to be mobile across all uses within a country, but immobile internationally, in the long-runmodel closure adopted.

On the demand side there is a regional representative household whose expenditure is governed by a CobbDouglasaggregate utility function which allocates net national expenditures across private, government, and saving activities. Thegreatest advantage of this household representation is the unambiguous indicator of economic welfare dictated by theregional utility function.1 Government demand across composite goods is determined by a CobbDouglas assumption (xedbudget shares). Private household demand is represented by a Constant Difference of Elasticities (CDE) functional form,which has the virtue of capturing the non-homothetic nature of private household demands, calibrated to replicate a vectorof own-price and income elasticities of demand (Hertel, McDougall, Narayanan, & Aguiar, 2008). In projecting to 2030 wemodify these elasticities for developing country crops and animal products for rapidly growing economies so they moreclosely match the income elasticities for these products in currently higher-income countries (following Yu, Hertel, Preckel,& Eales, 2004).2

Bilateral international trade ows are handled through Armington (1969) specication by which products aredifferentiated by country of origin. These Armington elasticities are the same across regions but are sector-specic, and theimportimport elasticities have been estimated at the disaggregated GTAP commodity level (Hertel, Hummels, Ivanic, &Keeney, 2007). For present purposes, where we are dealing with long-term changes, we follow the typical modelling practiseof doubling the short-to-medium term Armington elasticities. The regional balance of trade is determined by therelationship between regional investment and savings and investment can be allocated either in response to rates of return,

1 Altering taxes in the GTAP model does not imply a reduction in government revenue and expenditure, as government expenditures are not tied to tax

revenues. A tax reduction, for example, leads to a reduction in excess burden, so regional real income increases and real expenditure including government

expenditure may also rise.2 This is but one of several differences between the present projection exercise and that reported in Anderson and Strutt (2011a). Other renements

include updating the projections of GDP, population, unskilled labor, skilled labor and produced capital. We also now assume that land as well as other

natural capital endowments change slightly over time, and we alter the macro closure to allow investment to respond to changes in rates of return.

Furthermore, the initial database is augmented with estimates of distortions to agricultural prices in developing countries in 2004, based on Valenzuela and

Anderson (2008).

with capital markets in short run equilibrium, or in xed shares across regions so that it moves in line with global savings. Forpresent purposes, we allow investment to respond to changes in rates of return.

The GTAP Version 7.1 database divides the world into 112 countries/country groups, and divides each economy into 57sectors: 20 for agriculture, food, beverages and tobacco, 6 for other primary goods, 16 for manufactures and 15 forservices. For most modelling tasks, including this one, it is necessary for the sake of both computational speed anddigestion of model outputs to restrict the number of regions and sectors. In the present study we initially aggregate to 33countries/country groups and to 26 sector/product groups, as shown in column 2 of Appendix Tables A1 and A2. We thenfurther aggregate to 14 regions and just 4 sectors for most tables presented in this paper, as dened in column 1 of thoseAppendix Tables.

3. Core projection of the database to 2030

We project the GTAP databases 2004 baseline for the world economy to provide a new core baseline for 2030 by

K. Anderson, A. Strutt / Journal of Asian Economics 23 (2012) 303323 305assuming the 2004 trade-related policies of each country do not change. However, over the 26-year period we assume thatnational real GDP, population, unskilled and skilled labor, capital, agricultural land, and economically extractable mineralresources (oil, gas, coal and other minerals) grow at exogenously set rates, summarized in Appendix Table A3. The exogenousgrowth rates for GDP, investment and population are based on the latest (August 2011) ADB projections over the next twodecades, supplemented by World Bank data for real GDP and investment growth for the period to 2010, along with CEPII datafor population growth to 2010 and for regional projections of GDP, investment and population not readily available in theADB dataset (Foure, Benassy-Quere, & Fontagne, 2010).3 For projections of skilled and unskilled labor growth rates, we drawon Chappuis and Walmsley (2011). We estimate historical trends in agricultural land from FAOSTAT (summarized inDeininger & Byerlee, 2011) and in mineral and energy raw material reserves from BP (2010) and US Geological Survey (2010)and assume that past annual rates of change in farmland and fossil fuel reserves since 1990 continue for each country overthe next two decades.4 For other minerals, in the absence of country-specic data, the unweighed average of the annual rateof growth of global reserves for iron ore, copper, lead, nickel and zinc between 1995 and 2009 for all countries is used (fromUS Geological Survey, 2010). These rates of change in natural resources are summarized in the last ve columns of AppendixTable A3.

Given those exogenous growth rates,5 the model is able to derive implied rates of total factor productivity and GDP percapita growth. For any one country the rate of total factor productivity growth is assumed to be the same in each of its non-primary sectors, and to be somewhat higher in its primary sectors. Higher productivity growth rates for primary activitieswere characteristic of the latter half of the 20th century (Martin & Mitra, 2001), and are necessary in this projection if realinternational prices of primary products (relative to the aggregate change for all products) are to rise only modestly.6 Theimplied TFP growth rates for all sectors are shown in the rst column of Appendix Table A4,7 and the international priceconsequences for the core simulation are depicted in the rst column of Appendix Table A5.

It should be noted that the extent to which productivity growth rates are higher in each primary sector than in othersectors is the same for high-income and developing countries, and is the same for all crop and livestock industries withineach countrys farm sector. Since overall TFP growth is higher for developing than high-income countries in Appendix TableA4, this means we are assuming agricultural TFP growth is higher for developing than high-income countries on average.That is consistent with recent (if not earlier) experience: Ludena, Hertel, Preckel, Foster, and Nin (2007, Table 2) estimate thatagricultural TFP annual growth during 19812000 averaged 1.3 percent globally and only 0.9 percent for high-incomecountries (but during 19611980 those rates were 0.6 and 1.4 percent, respectively).

3.1. Consequences for size and sectoral and regional compositions of GDP and trade

The differences across regions in rates of growth of factor endowments and total factor productivities, and the fact thatsectors differ in their relative factor intensities and their share of GDP and that demands are nonhomothetic, ensure that the

3 World Bank and CEPII data are compiled from Chappuis and Walmsley (2011).4 Past reserves data are from BP (2010). For coal, however, production data are used since reserves data are not available. The growth rates for Vietnams

oil and gas, along with Thailands coal, provided implausibly high projections for the future, so they were modied.5 There is much uncertainty in macroeconomic projections over this kind of timeframe. See, for example Garnaut (2011) for some discussion on the

uncertain nature of GDP, population and energy projections. The wide variance between projections by think tanks (e.g., from Citi, 2011; PwC, 2011;

Subramanian, 2011) is therefore not surprising. Even the current size of the Chinese economy is under serious question (Feenstra, Ma, Neary, & Rao, 2011).6 We chose that calibration because it is consistent with the World Bank projections over the next four decades (see van der Mensbrugghe & Roson, 2010).

An alternative in which agricultural prices fall, as projected in GTAP-based projection studies in the late 20th century (e.g., Anderson, Dimaranan, Hertel, &

Martin, 2007), is considered unlikely over the next two decades given the slowdown in agricultural R&D investment since 1990 and its consequent delayed

slowing of farm productivity growth (Alston, Babcock, & Pardey, 2010). It is even less likely for farm products if fossil fuel prices and biofuel mandates in the

US, EU and elsewhere are maintained over the next decade. Timilsina et al. (2010) project that by 2020 international prices will be higher in the presence

versus the absence of those biofuel mandates for sugar (10 percent), corn (4 percent), oilseeds (3 percent), and wheat and coarse grains (2.2 percent), while

petroleum product prices will be 1.4 percent lower.7 In the core baseline, these TFP estimates are endogenously determined. However, in the simulations modelling lower worldwide primary sector

productivity, the TFP estimates are exogenous while GDP is endogenous.

structures of production, consumption and trade across sectors within countries, and also between countries, is going to bedifferent in 2030 than in 2004.

In particular, the faster-growing developing economies (especially those of Asia) will account for considerably larger sharesof the projected global economy over the next two decades. Their aggregate share of world GDP is projected to rise from 20percent in 2004 to 41 percent in 2030, and for just Developing Asia from 11 to 28 percent. Western Europes share, meanwhile, isprojected to fall from one-third to less than one-quarter. Population shares change much less, with the developing countriesshare rising from 80 to 83 percent but Developing Asias component falling a little, from 55 to 53 percent between 2004 and2030. Thus per capita incomes converge considerably, with the ratio of the high-income to developing country average morethan halving, from 16 to 7 between 2004 and 2030. In particular, the per capita income of Developing Asia is projected to risefrom 20 to 53 percent of the global average over the projection period (bottom rows of Appendix Table A6).

8

K. Anderson, A. Strutt / Journal of Asian Economics 23 (2012) 303323306When global value added is broken down by sector, the changes are more striking. This is especially so for China: by 2030 it isprojected to return to its supremacy as the worlds top producing country not only of primary products but also of manufactures(Table 1) a ranking China has not held since perhaps the mid-19th century when rst the UK and then (from 1895) the US wasthe top-ranked country for industrial production see Allen (2011, Fig. 2) and also Bairoch (1982) and Crafts and Venables (2003).

The Asian developing country share of global exports of all products almost doubles, rising from 21 to 39 percent between2004 and 2030 (Table 2). Chinas share alone grows from 6.7 to 18.4 percent. Note, however, that the growth of Chinas exportshare is entirely at the expense of high-income countries, as the export shares for almost all other developing-countryregions in Table 2 also grow. The groups import share also rises, although not quite so dramatically: the increase forDeveloping Asia is from 18 to 34 percent.9

The developing country share of primary products in world exports rises somewhat and continuing its pattern of thepast three decades (Martin, 2007) its share of manufactures in world exports rises dramatically over the projection period(doubling in Asias case, as does its services share Table 3). The developing country share of primary products in worldimports rises substantially though (Table 4), almost all of which is due to Developing Asias expected continuing rapidindustrialization.10 Developing Asia and other developing countries increase their share in total world imports by more thanhalf, and even by one-third in manufactures. In fact we understate that phenomenon because of the high degree ofaggregation of manufacturing industries in the version of the GTAP model we use here.

Given the political sensitivity of farm products, regional shares of global trade in just agricultural and food products areshown in Table 5(a). The developing country share of exports of those goods is projected to remain relatively unchanged.However, that country groups share of global imports of farm products rises dramatically. The source of that rise is mainlyChina. It is possible that China will seek to prevent such a growth in food import dependence in practice, by erectingprotectionist barriers at least for food staples.11

As for the sectoral shares of national trade and output, the consequences of continuing Asian industrialization are againevident: primary products are less important in developing country exports and considerably more important in theirimports, and conversely for non-primary products, with the changes being largest in Developing Asia. The opposite is true forhigh-income countries (Tables 6 and 7), which may seem surprising but recall that what one part of the world imports theremaining part of the world must export to maintain global equilibrium. The agricultural shares of output (GDP at constantprices) decline in the more-rapidly industrializing parts of the world, especially South and Southeast Asia and China, whilerising slightly in some other regions (Appendix Table A8).12

3.2. Consequences for SouthSouth and other bilateral trade

Turning now to bilateral trade patterns, the extent of SouthSouth trade as a share of global trade is projected to morethan double by 2030, rising from 13 to 30 percent. The share of NorthNorth trade in global trade, by contrast, is projected tofall, from 51 to 27 percent. Even so, the share of high-income countries exports to Developing Asia in global trade is slightlyhigher by 2030, and the share of Developing Asias exports to high-income countries in global trade is far higher, at 19percent in 2030 compared with 12 percent in 2004 (Table 8(a and b)). The latter is not surprising, given that the share ofDeveloping Asias exports in world trade almost doubles over this projection period, thanks to not only its high GDP growthrate but also its high trade-to-GDP ratio (rst two columns of Table 9).

Trade indexes may be used to take into account changes in regional shares of global trade and in openness. Two inparticular are used by Anderson and Norheim (1993): an intensity index, and a propensity index. The export trade intensity

8 Using producer expenditure on value added in each sector.9 Capital ows explain the difference between each regions global export and import shares.

10 Recall, though, that we are assuming no change in agricultural (or other) trade policies over the projection period. Perhaps a more likely scenario,

especially for rapidly growing Asia, would be a steady rise in agricultural protection to slow the decline in food self sufciency as has happened over the

past 50 years in the most-advanced Asian economies (see, e.g., Anderson, 2009).11 In a longer companion paper focusing just on agriculture (Anderson and Strutt, 2011b), we provide two scenarios in which developing country food self-

sufciency declines less. One involves faster grain productivity growth in Asias emerging economies; the other includes agricultural protection growth in

developing countries. Their latter certainly raises food self-sufciency in the South, but at the expense of reducing food consumption, overall economic

welfare and hence food security there.12 For discussion of impacts on Latin American countries, please see Anderson and Strutt (2012).

K. Anderson, A. Strutt / Journal of Asian Economics 23 (2012) 303323 307Table 1

Regional shares of global value added by sector, 2004 base, 2030 core and 2030 with full liberalization.

Agric. and food

(percent)

Other primary

(percent)

Manufactures

(percent)

Services

(percent)

Total

(percent)

(a) 2004 base

W. Europe 26.9 13.0 33.8 33.0 32.1

E. Europe 4.6 8.0 2.0 2.2 2.5

US and Canada 17.0 14.0 26.5 35.2 32.0

ANZ 2.0 2.6 1.3 1.9 1.8index is dened in value terms as the share of country is exports going to country j [xij/xi] divided by the share of country jsimports (mj) in world imports (mw) net of country is imports (mi). That is,

Ii j xi j=xi

mj=mw mi(1)

If the importer j is a country group and country i is part of country group j, it is necessary to subtract country is importsfrom mj (the numerator of the second expression in square brackets in Eq. (1)), since country i does not export to itself. If the

Japan 6.9 1.4 11.5 12.6 11.7

China 9.4 8.9 8.9 2.8 4.4

ASEAN 4.3 6.1 3.0 1.3 1.9

Pacic Islands 0.1 0.1 0.0 0.0 0.0

Rest E. Asia 1.8 1.0 4.1 2.7 2.9

India 7.5 2.6 1.2 1.1 1.6

Rest S. Asia 1.8 0.6 0.3 0.4 0.5

Central Asia 0.4 1.2 0.1 0.1 0.2

Latin America 9.1 7.3 5.0 3.7 4.4

M.E. and Africa 8.2 33.2 2.3 2.8 4.0

High-income 57.4 39.0 75.1 84.8 80.1Developing 42.6 61.0 24.9 15.2 19.9

of which Asia: 25.3 20.5 17.6 8.6 11.5

Total 100.0 100.0 100.0 100.0 100.0(b) 2030 core

W. Europe 15.7 7.5 22.9 25.3 22.9

E. Europe 3.4 7.8 2.1 2.6 3.0

US and Canada 12.4 7.5 19.9 32.1 26.8

ANZ 1.7 2.0 1.0 2.2 1.9

Japan 3.4 0.3 6.3 9.5 7.9

China 24.8 22.6 25.3 8.5 13.6

ASEAN 5.2 6.9 4.8 2.2 3.2

Pacic Islands 0.0 0.1 0.0 0.1 0.1

Rest E. Asia 1.4 0.7 4.5 3.4 3.2

India 11.7 4.7 3.2 3.7 4.3

Rest S. Asia 2.7 1.4 0.6 1.0 1.1

Central Asia 0.5 1.2 0.1 0.2 0.3

Latin America 7.8 9.1 5.8 4.9 5.6

M.E. and Africa 9.3 28.1 3.5 4.3 6.2

High-income 36.5 25.1 52.1 71.8 62.5Developing 63.5 74.9 47.9 28.2 37.5of which Asia: 46.3 37.7 38.5 19.1 25.7

Total 100.0 100.0 100.0 100.0 100.0(c) 2030 core with full liberalization of trade

W. Europe 15.0 7.7 22.4 25.2 22.7

E. Europe 3.6 7.9 1.9 2.6 2.9

US and Canada 12.9 7.5 19.3 31.4 26.3

ANZ 2.2 2.0 0.9 2.2 2.0

Japan 3.0 0.3 6.5 9.6 7.9

China 23.7 22.3 26.2 8.7 13.7

ASEAN 5.8 7.1 4.9 2.4 3.4

Pacic Islands 0.1 0.1 0.0 0.1 0.1

Rest E. Asia 1.4 0.7 4.9 3.6 3.4

India 10.1 4.5 3.5 3.8 4.2

Rest S. Asia 2.7 1.4 0.7 1.1 1.1

Central Asia 0.5 1.2 0.1 0.2 0.3

Latin America 10.0 9.1 5.4 4.9 5.7

M.E. and Africa 9.2 28.3 3.4 4.4 6.2


Total 100.0 100.0 100.0 100.0 100.0

Source: Derived from the authors GTAP Model results.

Table 2

Shares of world exports and imports of all goods and services, 2004 and 2030.

Exports (percent) Imports (percent)

2004 2030 2004 2030

W. Europe 42.3 25.2 42.5 28.1

E. Europe 3.6 4.4 3.2 3.5

US and Canada 13.7 9.5 18.8 16.6

ANZ 1.3 1.2 1.4 1.6

Japan 6.1 2.4 5.1 4.5

China 6.7 18.4 5.7 15.5

ASEAN 6.0 9.4 5.2 7.4

Pacic Islands 0.1 0.1 0.1 0.1

Rest E. Asia 6.3 6.0 5.3 5.6

India 1.0 3.1 1.2 3.6

South Asia 0.4 1.2 0.5 1.2

Central Asia 0.4 0.5 0.4 0.5

Latin America 5.4 7.8 4.7 4.8

M.E. and Africa 6.8 10.9 5.8 6.9

High-income 67.0 42.7 71.1 54.3Developing 33.0 57.3 28.9 45.7of which Asia: 20.8 38.7 18.4 34.0

Total 100.0 100.0 100.0 100.0


Table 3

Regional sectoral shares of global exports of all products, 2004 and 2030.

Agric. and food

(percent)

Other primary

(percent)

Manufactures

(percent)

Services

(percent)

Total

(percent)

(a) 2004

W. Europe 2.9 1.0 29.4 9.1 42.3

E. Europe 0.2 0.9 1.9 0.5 3.6

US and Canada 0.9 0.4 9.4 3.0 13.7

ANZ 0.3 0.2 0.5 0.3 1.3

Japan 0.0 0.0 5.5 0.6 6.1

China 0.2 0.1 6.0 0.5 6.7

ASEAN 0.4 0.4 4.4 0.7 6.0

Pacic Islands 0.0 0.0 0.0 0.0 0.1

Rest E. Asia 0.1 0.0 4.9 1.3 6.3

India 0.1 0.0 0.7 0.2 1.0

Rest S. Asia 0.0 0.0 0.3 0.1 0.4

Central Asia 0.0 0.2 0.1 0.0 0.4

Latin America 0.9 0.8 3.0 0.6 5.4

M.E. and Africa 0.4 3.3 2.2 0.9 6.8


Total 6.6 7.4 68.2 17.8 100.0(b) 2030 core

W. Europe 2.3 1.7 16.2 5.0 25.2

E. Europe 0.2 2.1 1.6 0.5 4.4

US and Canada 1.3 1.1 5.6 1.6 9.5

ANZ 0.3 0.5 0.2 0.1 1.2

Japan 0.0 0.0 2.1 0.2 2.4

China 0.0 0.0 16.4 1.9 18.4

ASEAN 0.6 1.0 6.8 1.0 9.4

Pacic Islands 0.0 0.0 0.0 0.0 0.1

Rest E. Asia 0.1 0.1 4.9 1.0 6.0

India 0.2 0.2 1.9 0.8 3.1

Rest S. Asia 0.1 0.0 0.8 0.3 1.2

Central Asia 0.1 0.4 0.0 0.0 0.5

Latin America 1.2 1.8 4.1 0.7 7.8

M.E. and Africa 0.5 5.0 3.1 2.2 10.9


Total 6.9 14.0 63.6 15.5 100.0


K. Anderson, A. Strutt / Journal of Asian Economics 23 (2012) 303323308


Regional sectoral shares of global imports of all products, 2004 and 2030.

Agric. and food

(percent)

Other primary

(percent)

Manufactures

(percent)

Services

(percent)

Total

(percent)

(a) 2004

W. Europe 3.1 2.5 28.2 8.6 42.5

E. Europe 0.3 0.4 2.1 0.5 3.2

US and Canada 0.9 1.6 13.7 2.7 18.8

ANZ 0.1 0.1 1.0 0.3 1.4

Japan 0.5 0.8 2.8 1.0 5.1

China 0.2 0.5 4.4 0.6 5.7

ASEAN 0.3 0.4 3.7 0.8 5.2

Pacic Islands 0.0 0.0 0.1 0.0 0.1

Rest E. Asia 0.3 0.6 3.6 0.8 5.3

India 0.1 0.3 0.6 0.2 1.2

Rest S. Asia 0.1 0.0 0.3 0.1 0.5

Central Asia 0.0 0.0 0.3 0.1 0.4

Latin America 0.4 0.2 3.5 0.7 4.7

M.E. and Africa 0.7 0.2 3.9 1.0 5.8


Total 6.9 7.6 68.3 17.2 100.0(b) 2030 core

W. Europe 1.8 1.8 18.4 6.1 28.1

E. Europe 0.3 0.5 2.1 0.6 3.5

US and Canada 0.6 1.6 12.0 2.4 16.6exporter i is a country group, an approximation can be calculated by excluding only 1/nth of is imports from mw in thedenominator of the second expression in square brackets in Eq. (1), where n is the number of countries in the exporter group;and in the case where i = j, also multiply mj (the numerator of the second expression in square brackets in Eq. (1)), by (n 1)/n. The weighted average of Iij across all j is unity; and the more Iij is above unity, the more intense is the bilateral traderelationship between i and j.

A change in trade openness within a region may result in so much net trade creation that, even though the index ofintensity of region is trade with other regions falls, there is a rise in the regions propensity to trade outside its own regionbecause of an increase in the value of its trade with other regions as a proportion of is GDP. To capture the combined effect ofthese two changes, in openness and in extra-regional trade intensity, Anderson and Norheim (1993) dene the index of thepropensity to export extra-regionally (Pij,) as

Pi j xi j=Gi

mj=mw mi ti Ii j (2)

where Gi is is GDP and ti is the ratio of is total exports to is GDP. Unlike Iij, Pij does not have a weighted average across all j ofunity.

Table 10 summarizes the trade intensity indexes. It suggests that, in the absence of trade policy changes, the intensity ofintra-Developing Asia trade will decline between 2004 and 2030, from 1.84 to 1.22. The intensity of that regions exports todeveloping countries of other regions will remain at around 0.75, and the intensity of other developing countries exports toDeveloping Asia will rise a little, from 0.98 to 1.11. The estimated indexes of intensity of the two developing regions exportsto high-income countries switch over that period, rising from 0.82 to 0.92 for Developing Asia and falling from 0.92 to 0.82for other developing countries. Meanwhile, the intensity of exports from high-income countries to the two developingregions changes by less than 3 percent and remains below unity. In short, the intensity of trade between the two developing

ANZ 0.1 0.0 1.2 0.3 1.6

Japan 0.3 0.5 2.8 0.9 4.5

China 1.8 5.2 7.7 0.8 15.5

ASEAN 0.5 0.7 5.3 0.8 7.4

Pacic Islands 0.0 0.0 0.1 0.0 0.1

Rest E. Asia 0.2 0.8 3.7 0.8 5.6

India 0.1 1.9 1.3 0.3 3.6

Rest S. Asia 0.2 0.3 0.6 0.1 1.2

Central Asia 0.0 0.0 0.3 0.1 0.5

Latin America 0.3 0.4 3.4 0.7 4.8

M.E. and Africa 0.9 0.4 4.8 0.9 6.9


Total 7.2 14.2 63.8 14.8 100.0


Ta

K. Anderson, A. Strutt / Journal of Asian Economics 23 (2012) 303323310Exports Imports

2004 base 2030 core

base

2030 slower GDP

and K growth

2030 slower

prim TFP

2004

base

2030 core

base

2030 slower GDP

and K growth

2030 slower

prim TFP

Regional shares of world (including intra-EU) trade (percent)

W. Europe 43.6 32.6 31.8 30.2 45.6 24.8 23.9 23.9

E. Europe 3.2 3.6 3.5 4.0 4.3 3.7 3.4 4.2

US and Canada 14.3 18.5 17.8 19.2 12.4 8.5 8.1 8.4

ANZ 4.7 4.9 4.6 4.1 1.0 1.0 0.9 0.9

Japan 0.4 0.6 0.6 0.6 7.2 3.8 3.7 3.8

China 3.6 0.5 0.5 0.2 3.6 25.5 26.0 25.1

ASEAN 6.7 8.8 9.3 10.3 4.4 7.3 7.7 8.2

Pacic Islands 0.2 0.2 0.2 0.1 0.2 0.2 0.2 0.2

Rest E. Asia 0.9 1.4 1.4 1.6 3.9 3.4 3.5 3.3

India 1.4 2.3 2.4 2.7 0.8 1.6 1.6 1.6

Rest South Asia 0.6 0.8 0.8 0.9 1.1 3.2 3.2 2.3

Central Asia 0.5 0.8 0.8 0.7 0.4 0.5 0.5 0.5Regcc

incotr

4

ea

Soble 5

ional shares of world trade in agricultural and food products and agricultural self sufciency, 2004 base and 2030 core and slower growth scenarios.ountry groups is projected to grow by 2030 while that between the developing and high-income aggregate country groupshanges little.However, because of the high trade-to-GDP ratios of Asian countries (see the rst two columns of Table 9), the estimateddex of propensity of Developing Asias trade with other regions increases, from 0.38 to 0.42 for exports to high-incomeountries, while remaining at 0.35 for exports to other developing countries. There is a large increase in the propensity ofther developing countries to export to rapidly-growing Developing Asia though, from 0.32 to 0.48, in contrast to the otherade propensity indexes which change very little (Table 11).

. Alternative projections to 2030

The above core projection is but one of myriad possibilities, so in this section we explore several others and compare theirconomic consequences with those just summarized for 2030. Specically, the following two alternative growth scenariosre considered, plus a set of scenarios involving varying degrees of trade policy reform:

A one-third reduction in rates of growth of GDP and capital in (and hence slower growth in the volume of trade of) high-income countries (the North);

Latin America 13.6 17.2 17.9 17.3 5.5 4.5 4.6 4.4

M.E. and Africa 6.3 7.9 8.3 7.9 9.6 12.3 12.7 13.1

High-income 66.2 60.2 58.4 58.2 70.5 41.8 40.0 41.3Developing 33.8 39.8 41.6 41.8 29.5 58.2 60.0 58.7of which Asia: 13.9 14.7 15.4 16.6 14.3 41.5 42.6 41.2

Total 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0

Baseline scenarios Trade reform scenarios

2004 2030 core 2030 slower

GDP and

K growth

2030 slower

prim. TFP

ASEAN + 6,

no agric

ASEAN + 6,

with agric

ASEAN + 6

MFN

Full Lib. South

South Lib.

(b) Agricultural self-sufciency (excluding other food products)a

W. Europe 0.94 1.05 1.06 1.01 1.06 1.04 1.07 1.01 1.06

E. Europe 0.94 0.95 0.97 0.94 0.95 0.95 0.94 0.91 0.95

US and Canada 1.04 1.20 1.21 1.19 1.20 1.19 1.23 1.24 1.20

ANZ 1.45 1.64 1.66 1.56 1.63 1.82 1.68 1.75 1.65

Japan 0.81 0.83 0.83 0.82 0.82 0.75 0.65 0.65 0.83

China 0.97 0.83 0.83 0.84 0.82 0.82 0.80 0.79 0.82

ASEAN 0.97 0.85 0.85 0.83 0.84 0.87 0.86 0.86 0.84

Pacic Islands 0.92 0.90 0.90 0.88 0.91 0.91 0.92 2.05 0.87

Rest E. Asia 0.77 0.79 0.79 0.76 0.78 0.84 0.81 0.77 0.77

India 1.01 1.00 1.00 1.00 1.00 1.02 0.99 1.00 1.01

Rest S. Asia 0.96 0.85 0.86 0.90 0.86 0.86 0.87 0.85 0.85

Central Asia 1.04 1.09 1.08 1.05 1.09 1.08 1.08 1.10 1.09

Latin America 1.10 1.23 1.23 1.29 1.23 1.23 1.26 1.40 1.23

M.E. and Africa 0.93 0.92 0.92 0.93 0.92 0.92 0.92 0.91 0.91

High-income 0.97 1.09 1.10 1.06 1.09 1.09 1.11 1.09 1.09Developing 0.98 0.91 0.92 0.93 0.91 0.92 0.91 0.93 0.91of which Asia: 0.96 0.87 0.87 0.87 0.87 0.87 0.85 0.85 0.86

Total 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00

urce: Derived from the authors GTAP Model results.

a Agricultural self-sufciency ratio excludes other (processed) food products.

Table 6

Sectoral shares of national exports, 2004 and 2030.

Agric. and food

(percent)

Other primary

(percent)

Manufactures

(percent)

Services

(percent)

Total

(percent)

K. Anderson, A. Strutt / Journal of Asian Economics 23 (2012) 303323 311

4

coTre

o

So

1

si

Th

ofW. Europe 9.0 6.8 64.3 20.0 100.0

E. Europe 5.7 47.5 35.9 11.0 100.0

US and Canada 13.5 11.6 58.2 16.7 100.0

ANZ 29.4 45.1 16.0 9.4 100.0(bHigh-income 6.5 3.9 69.5 20.1 100.0

Developing 6.7 14.6 65.6 13.1 100.0

of which Asia: 4.4 3.6 78.7 13.3 100.0

Total 6.6 7.4 68.2 17.8 100.0

) 2030 coreRest S. Asia 10.2 1.7 70.7 17.3 100.0

Central Asia 8.4 53.1 26.7 11.8 100.0

Latin America 16.6 15.1 56.5 11.8 100.0

M.E. and Africa 6.1 48.0 32.5 13.5 100.0(a) 2004

W. Europe 6.8 2.3 69.4 21.5 100.0

E. Europe 5.8 26.6 52.7 14.8 100.0

US and Canada 6.9 3.1 68.4 21.7 100.0

ANZ 23.3 18.1 35.4 23.3 100.0

Japan 0.5 0.1 90.1 9.3 100.0

China 3.5 1.2 88.6 6.7 100.0

ASEAN 7.4 6.2 74.3 12.2 100.0

Pacic Islands 17.1 25.2 31.9 25.7 100.0

Rest E. Asia 1.0 0.2 78.4 20.4 100.0

India 9.4 4.8 67.6 18.3 100.0Slower total factor productivity (TFP) growth in primary sectors, so that real international prices for agricultural, mineral andenergy products by 2030 are much more above 2004 levels than in the core projection and thus more consistent with theprojections of some international agencies that specialize in those markets instead of with the World Banks projections;Full liberalization of goods trade barriers following the formation of a free trade area among ASEAN + 6, the six being China,Japan and Korea plus India, Australia and New Zealand (see Kawai & Wignaraja, 2010), under three versions of thisinitiative to show how much a broadening of the product and country involvement brings the potential benets ever-closer to those from global trade reform; andPartial liberalization of SouthSouth goods trade barriers aimed at bringing them down to the same as the average tariffsfaced in SouthNorth trade in 2004 for those products for which the latter are lower (most non-farm goods see AppendixTable A7(a)).

.1. Reduced rates of GDP and capital growth

If it is assumed the rates of growth to 2030 of GDP and capital in high-income countries are one-third lower than in there projection, it slows the structural transformation of output towards non-primary sectors in Developing Asian countries.his slowdown in high-income countries reduces their share of world GDP (by three percentage points) and raises thelative GDP per capita of developing countries by one-seventh.13

The impact of slower high-income country growth on bilateral trade patterns can be seen from Table 8: it raises the sharef intra-Developing Asia trade in global trade by one percentage point, and of all SouthSouth trade by more than

Japan 1.7 1.7 88.6 7.9 100.0

China 0.2 0.1 89.3 10.5 100.0

ASEAN 6.5 10.4 72.1 11.0 100.0

Pacic Islands 13.0 36.4 34.7 15.9 100.0

Rest E. Asia 1.6 0.8 80.5 17.1 100.0

India 5.3 6.8 61.1 26.9 100.0

Rest S. Asia 4.5 0.7 66.3 28.5 100.0

Central Asia 9.7 77.4 8.4 4.5 100.0

Latin America 15.4 23.2 52.3 9.1 100.0

M.E. and Africa 5.0 46.3 28.5 20.1 100.0


Total 6.9 14.0 63.6 15.5 100.0


3 If one were to assume that this one-third reduction in growth in high-income countries also caused growth to slow by half that extent (that is, by one-

xth) in developing countries, there would be almost no change in either the regional shares of world GDP or relative GDPs per capita projected for 2030.

is is because developing countries are assumed in the core scenario to continue to grow somewhat faster than high-income countries, so even if their rate

growth were to slow proportionately less, it would translate to a slowdown in GDP by a similar number of percentage points.

K. Anderson, A. Strutt / Journal of Asian Economics 23 (2012) 303323312Table 7

Sectoral shares of national imports, 2004 and 2030.

Agric. and food

(percent)

Other primary

(percent)

Manufactures

(percent)

Services

(percent)

Total

(percent)

(a) 2004

W. Europe 7.4 5.9 66.4 20.3 100.0

E. Europe 9.0 11.0 64.9 15.1 100.0

US and Canada 4.5 8.5 72.8 14.1 100.0

ANZ 4.9 3.9 73.2 18.0 100.0

Japan 9.6 16.3 55.0 19.1 100.0

China 4.4 8.5 77.1 10.0 100.0

ASEAN 5.8 6.8 72.7 14.7 100.0

Pacic Islands 11.8 0.8 69.0 18.4 100.0

Rest E. Asia 5.1 11.5 68.3 15.1 100.0

India 4.5 28.5 52.0 15.0 100.0

Rest S. Asia 13.5 6.9 64.2 15.5 100.0

Central Asia 7.7 5.5 63.4 23.5 100.0

Latin America 7.9 4.6 73.4 14.0 100.0

M.E. and Africa 11.4 3.5 67.9 17.2 100.0


Total 6.9 7.6 68.3 17.2 100.0(b) 2030 core

W. Europe 6.4 6.6 65.4 21.6 100.0

E. Europe 7.6 14.1 61.2 17.0 100.0

US and Canada 3.7 9.6 72.0 14.7 100.02 percentage points (which is also the amount by which the share of NorthNorth trade in global trade is lowered). Similarly,regional shares of the worlds agricultural trade shift away from high-income countries a little more (Table 5). If that slowergrowth of high-income countries did not slow down developing country GDP growth, then all but $66 billion of the net globalcost of the slowdown ($5.7 trillion per year in 2030) would be borne by high-income countries (Table 12).

4.2. Slower TFP growth in primary sectors

The core projection sets higher TFP growth rates for some primary product sectors than for other sectors such that averagereal international prices for agricultural and other primary products by 2030 are no more than 9 and 25 percent, respectively,above 2004 levels (column 1 of Appendix Table A5). That is quite different from what was experienced in the 20th century, whenreal primary product prices traced a long-run downward trend (apart from the 1973 and 1979 OPEC cartel-induced jumps in theprice of fossil fuels). In the past decade, however, those prices have been rising, and price projections of several internationalagencies suggest they will be well above 2004 levels in the next decade or so (FAO/OECD, 2010; IEA, 2010; Nelson et al., 2010).Hence this second alternative scenario, in which we assume the addition to the TFP growth rate of 2.5 percentage points per yearfor forestry and shing is reduced to 1 percentage point. For mining, agriculture and lightly processed food the productivitydifferential in the core projection is smaller, but it too is reduced, by 1 percentage point. These amendments lead to realinternational prices for farm products in 2030 to be 25 instead of just 9 percent above those in 2004, and those for other primaryproducts to be 101 instead of 25 percent above 2004 levels (see Appendix Table A5 for details by product).

The higher prices more than compensate for lower farming and mining productivity such that the share of primaryproducts in GDP is somewhat higher in this scenario than in the core projection. This does not lead to developing countriesbeing more food self-sufcient though, or to much change in their share of global trade in farm products or in bilateral tradepatterns. It does, however, raise considerably the share of GDP that is traded by each region (Table 9), due simply to thehigher prices of primary products.

ANZ 4.2 2.7 72.2 20.8 100.0

Japan 6.1 10.9 63.3 19.7 100.0

China 11.8 33.4 49.8 4.9 100.0

ASEAN 7.1 9.6 72.1 11.1 100.0

Pacic Islands 10.5 2.6 67.4 19.6 100.0

Rest E. Asia 4.4 15.0 66.4 14.2 100.0

India 3.1 52.9 36.8 7.2 100.0

Rest S. Asia 18.6 21.8 51.6 8.0 100.0

Central Asia 6.4 4.9 63.8 24.9 100.0

Latin America 6.7 8.4 70.5 14.4 100.0

M.E. and Africa 12.8 5.5 68.8 13.0 100.0


Total 7.2 14.2 63.8 14.8 100.0



Shares of bilateral trade of high-income and developing countries in global trade, 2004 base and 2030 core and slower growth and trade reform scenarios.

Exporter Importer

High-income Developing Asia Other developing Total

(a) 2004 base

High-income 51.2 9.1 6.7 67.0Developing Asia 12.2 6.9 1.7 20.8Other developing 8.0 2.2 2.0 12.2Total 71.4 18.2 10.4 100.0

(b) 2030 core sim

High-income 26.9 10.8 4.9 42.7Developing Asia 19.4 15.9 3.4 38.7Other Developing 8.3 7.1 3.2 18.6Total 54.7 33.7 11.6 100.05. Trade liberalization scenarios

Each of the following trade liberalization scenarios is compared from the 2030 core baseline. These scenarios are intendedsimply to be indicative of potential gains that may be possible, given the anticipated size and structure of global markets in 2030.In reality, some of these agreements are already partially implemented and others, if implemented, will be staggered over time.

5.1. Full liberalization of ASEAN + 6 trade barriers versus global free trade in goods

If the ASEAN free trade area were extended to the additional countries included in ASEAN + 6 (China, Japan, South Korea,India, Australia and New Zealand) and their goods trade were to be liberalized fully, that could go a long way towardsgenerating the benets that could come from global goods trade liberalization. This is because the global shares of thatexpanded bloc of countries in 2004 would rise from the ASEAN alone share of 2 to 34 percent for GDP for the ASEAN + 6 blocin 2030 and similarly from 6 to 37 percent for exports (Table 2 and Appendix Table A6). But as with all such regional trading

(c) 2030 slower GDP and capital growth for HICs

High-income 24.7 10.8 4.9 40.4Developing Asia 19.4 17.0 3.8 40.2Other Developing 8.2 7.6 3.5 19.4Total 52.3 35.5 12.3 100.0

(d) 2030 slower primary TFP growth


(e) 2030 core sim plus ASEAN + 6 preferential libn without agric


(f) 2030 core sim plus ASEAN + 6 preferential libn, all goods


(g) 2030 core sim plus ASEAN + 6 MFN libn, all goods


(h) 2030 core sim plus global MFN libn, all goods


(i) 2030 core sim plus partial SouthSouth trade libn


(j) 2030 core sim with constrained trade surplus for China and trade decit for the US



Ta

K. Anderson, A. Strutt / Journal of Asian Economics 23 (2012) 303323314Baseline scenarios Trade reform scenarios

2004 2030 core 2030 slower

HIC growth

Slower primary

TFP growth

ASEAN + 6

No Ag.

ASEAN + 6 ASEAN + 6

MFN

Full Lib. SS Lib.Expav

So

Ta

In

So

Ta

In

Soble 9

orts plus imports of goods and services as a proportion of GDP, 2004 base and alternative 2030 simulations.greements, the potential benets depend on the extent to which all trade is freed up. Hence we present results from threeersions of this initiative plus global goods trade reform. Those three variants are:

All merchandise trade except for agricultural goods is freed within the expanded bloc (that is, on a preferential basis, with nochange to barriers to trade with other countries),

W. Europe 0.66 0.70 0.70 0.79 0.70 0.70 0.71 0.75 0.70

E. Europe 0.69 0.78 0.78 0.90 0.77 0.78 0.79 0.89 0.77

US and Canada 0.27 0.31 0.31 0.35 0.31 0.31 0.31 0.33 0.30

ANZ 0.39 0.44 0.45 0.49 0.47 0.49 0.48 0.49 0.43

Japan 0.26 0.29 0.29 0.33 0.32 0.33 0.32 0.33 0.28

China 0.77 0.82 0.81 1.29 0.91 0.92 0.99 1.02 0.91

ASEAN 1.48 1.59 1.58 1.85 1.75 1.77 1.79 1.80 1.72

Pacic Islands 0.96 0.96 0.95 1.05 0.95 0.95 0.97 1.29 1.10

Rest E. Asia 1.05 1.10 1.09 1.28 1.15 1.17 1.21 1.26 1.20

India 0.37 0.46 0.45 0.62 0.55 0.58 0.66 0.67 0.61

Rest S. Asia 0.51 0.66 0.65 0.83 0.65 0.65 0.66 0.86 0.81

Central Asia 0.99 1.02 1.01 1.13 1.02 1.02 1.03 1.12 1.06

Latin America 0.49 0.58 0.57 0.71 0.59 0.59 0.59 0.67 0.64

M.E. and Africa 0.80 0.87 0.86 0.94 0.86 0.86 0.87 1.03 0.96

High-income 0.44 0.48 0.49 0.55 0.49 0.49 0.50 0.52 0.48

Developing 0.77 0.83 0.82 1.07 0.89 0.90 0.94 1.01 0.93

of which Asia: 0.90 0.89 0.88 1.22 0.98 1.00 1.05 1.09 1.00

Total 0.51 0.61 0.62 0.73 0.64 0.64 0.66 0.71 0.65


ble 10

tra- and extra-regional trade intensity indexesa for Developing Asian countries, other developing countries and high-income countries, 2004 and 2030.

High-income countries Developing Asian countries Other developing countries

(a) Benchmark database, 2004

High-income 1.08 0.73 0.94

Developing Asia 0.82 1.84 0.76

Other developing 0.92 0.98 1.55

(b) Core baseline, 2030

High-income 1.17 0.74 0.97



(c) Full global MFN tariff liberalization, 2030

High-income 1.17 0.80 0.94


Other Developing 0.89 1.06 1.25


a For denitions of the intensity indexes, see text in Section 3.2.

ble 11

tra- and extra-regional trade propensity indexesa for Developing Asian countries, other developing countries and high-income countries, 2004 and 2030.

High-income countries Developing Asian countries Other developing countries

(a) Benchmark database, 2004

High-income 0.23 0.15 0.20



(b) Core baseline, 2030

High-income 0.24 0.15 0.20



(c) Full global MFN tariff liberalization, 2030

High-income 0.26 0.18 0.21




a For denitions of the propensity indexes, see text in Section 3.2.

Table 12

Effects on welfare and GDP of liberalizing trade in Asia and globally, 2030.

Slower HIC gr. ASEAN + no ag ASEAN + with ag ASEAN + MFN Global MFN SouthSouth libn

a


pledbslgthcothgb

So

1

ef

frOther 28.0 5.3 4.3 30.1 100.6 97.0Total 5697.6 16.3 59.8 166.4 384.0 107.4(a) Change in welfare (in 2004US$ billion per year)

W. Europe 1787.0 16.0 14.8 26.4 76.3 38.8E. Europe 453.4 0.3 0.3 8.8 24.2 8.4US and Canada 2758.4 11.2 14.9 16.4 12.2 33.1ANZ 238.3 1.4 12.1 1.0 3.9 1.6Japan 395.2 21.8 39.4 31.6 33.0 13.9China 24.8 2.0 2.4 22.0 19.9 5.7ASEAN 10.7 31.1 40.8 42.5 50.1 32.7Pacic Islands 0.1 0.1 0.3 0.1 1.2 0.7Rest E. Asia 1.3 9.1 15.2 28.9 41.8 37.1India 1.6 13.2 8.1 2.1 13.2 5.3R. South Asia 0.2 3.1 3.4 0.6 4.6 3.8Central Asia 2.5 0.0 0.2 1.3 3.2 4.0Latin America 4.1 2.2 3.4 1.5 28.9 26.4M.E. and Africa 23.8 3.0 0.9 28.6 71.7 70.6High-income 5632.2 4.2 22.1 84.1 149.6 78.9Developing 65.5 20.6 37.8 82.3 234.5 186.3of which: Asia 37.5 25.8 42.0 52.2 133.9 89.3All merchandise trade including agricultural goods is freed within the expanded bloc,All merchandise trade including agricultural goods is freed within the expanded bloc and also with the rest of the world (thatis, on an MFN basis).

The economic welfare effects of those reforms are summarized in Table 12. If the ASEAN + 6 initiative was purelyreferential and the reform excluded farm products, the global gains would be only $16 billion a year by 2030. That would bess than the gain to Developing Asia ($26 billion per year), with high-income countries losing $4 billion and othereveloping countries $5 billion. Were agriculture not to be excluded from the deal, the global gains would nearly quadrupleut most of them would be enjoyed in the Western Pacic, and non-Asian developing countries as a group still would beightly worse off. Were those reforms by ASEAN + 6 to be on an MFN basis (that is, removed for trade not only within theroup but also with non-members), the global gains would more than double again, to $166 billion per year by 2030, whilee gain to Asias developing countries would be about one-quarter higher at $52 billion. In that case non-Asian developinguntries also would be better off, by $30 billion. For all three sets of countries those welfare benets are less than half whatey would be if all countries of the world were to remove their barriers to goods trade. Such an extreme reform wouldenerate welfare gains of $384 billion per year globally by 2030, made up of $150 billion for high-income countries, $134illion for Developing Asia, and about $100 billion for other developing countries.14

ASEAN + no ag ASEAN + with ag ASEAN + MFN Global MFN SouthSouth libn

(b) Change in real GDP (%)

W. Europe 0.04 0.02 0.03 0.60 0.11E. Europe 0.01 0.00 0.12 0.98 0.01US and Canada 0.01 0.01 0.02 0.14 0.03ANZ 0.10 0.12 0.17 0.16 0.04Japan 0.06 0.47 0.58 0.58 0.03China 0.13 0.13 0.45 0.53 0.21

ASEAN 0.77 0.94 2.04 2.21 0.93

Pacic Islands 0.07 0.24 0.09 2.68 1.30Rest E. Asia 0.20 0.61 0.86 1.18 0.80

India 0.13 0.21 1.28 1.37 0.96R. South Asia 0.15 0.15 0.01 1.38 1.00Central Asia 0.03 0.01 0.03 0.59 0.43Latin America 0.06 0.07 0.05 0.44 0.44M.E. and Africa 0.09 0.09 0.01 1.13 0.59High-income 0.01 0.05 0.10 0.41 0.06Developing 0.08 0.17 0.51 0.94 0.55of which Asia 0.17 0.32 0.85 1.05 0.58

Other 0.08 0.08 0.03 0.75 0.51World 0.02 0.09 0.24 0.59 0.15


a As measured by an equivalent variation in income.

4 That estimated global welfare gain from freeing all of the worlds merchandise trade is small both in absolute terms and compared with the welfare

fect of a GDP slowdown in high-income countries (shown in column 1 of Table 11). However, as the caveats in the next section make clear, those gains

om global trade reform are very much lower-bound estimates.

The impacts of such reforms on the worlds bilateral trade pattern are similar for all four reform options (Table 8). Notsurprisingly, preferential trade reforms raise intra-Developing Asian trades share of global trade, by 3.5 percentage points.That drops back a point if ASEAN + 6 remove their barriers to trade with all partners rather than with just bloc membercountries. However, SouthSouth trade as a share of global trade is 32 percent in both cases (compared with 29.6 in the coreprojection).

K. Anderson, A. Strutt / Journal of Asian Economics 23 (2012) 303323316Each of these reform scenarios adds to globalization, as captured by the share of world GDP traded. In the core simulationthat is 61 percent globally, but it is 64 percent with ASEAN + 6 preferential trade and 71 percent with global free trade: andfor Developing Asia those numbers are 89, 100 and 109 percent, respectively (last two rows of Table 9).

The consequences of freeing all global trade for bilateral trade intensity and propensity indexes are shown in part (c) ofTables 10 and 11. Compared with the 2030 baseline, such reform would further strengthen the intensity and propensityindexes of Developing Asias exports to other developing countries, raising the intensity index from 0.75 to 0.91 and thepropensity index from 0.35 to 0.50. The propensity of other developing countries to export to Developing Asia also wouldincrease, from 0.48 to 0.51. Such reform would alter the shares of global GDP by sector, but compared with the changes due toeconomic growth from 2004 to 2030, the changes in shares by 2030 due to full global reform would be relatively small(different by no more than one percentage point in most cases compare parts (b) and (c) of Table 1).

5.2. Partial liberalization of SouthSouth trade barriers

In this scenario, we explore the impacts of intra-South import tariffs being reduced to the average level of tariffs imposedon exports from the South to the North. Tariffs imposed by the South in the initial database tend to be relatively high,particularly those applied to other South countries (see Appendix Table A7(b)). To model the impact of reduced intra-Southtariffs, the SouthNorth average tariffs for each commodity become the ceiling tariff for SouthSouth trade, with anybilateral tariffs already lower than these remaining unaffected (that exception being predominately farm products seeAppendix Table A7(a)). This tariff reduction is nearly as benecial to developing countries as a group as is a move to globalfree trade: while it hurts high-income countries slightly, a lowering of barriers to SouthSouth trade even to just the levelsprevailing in South-North trade could bring four-fths of the gain to developing countries as would ow from a freeing of allcountries goods trade (Table 12). Furthermore, it gives an even larger boost to the share of SouthSouth trade in global trade,raising it to 36 percent compared with the core projection of 30 percent (Table 8(b) and (i)).

6. Caveats

As with the results from all other economy-wide projections modelling, it is necessary to keep in mind numerousqualications. One is that for the core projection we have assumed all trade-related policies and trade costs remain unchangedthrough the projection period. As noted above, the former is somewhat unrealistic when that simulation suggests populousdeveloping countries such as China and India would become far more dependent on food imports by 2030.15 We have alsoassumed trade expenses in the form of transport and communications costs do not change, even though they have been fallingduring the current wave of globalization. If they were to continue to decline over time, those countries that have been tradinglittle and in only a few products because they are small, remote or have poor transport infrastructure will be able to trade more and thus would gain doubly if governmental barriers were to be lowered at the same time as trade costs fall (Venables, 2004).

A second assumption is that we do not constrain trade imbalances over time. However, even in the initial database, theseare huge for the United States and China, and some argue that these imbalances are unlikely to be sustained over time (seeFeldstein, 2011; Garnaut, 2011, among others). Given that the large and rapidly growing Chinese economy is an importantdriver of some of the changes we model, we tested the sensitivity of key results to determine how they might change in analternative scenario where Chinas trade surplus is constrained. In particular, we considered a fairly extreme alternativescenario in which the trade surplus for China and the trade decit for the United States are essentially eliminated over thenext two decades.16 Since we simply wished to test the robustness of our ndings to this possibility, we did not modify otherassumptions from the core baseline, including the GDP growth and capital accumulation rates, or trade balances in otherregions. The importance of China in global exports will naturally reduce if it is not able to continue its huge trade surpluses,while the importance of the United States in global exports will increase if it no longer runs large trade decits. Bilateral tradeows, particularly for the United States and China, will thus be impacted fairly signicantly by this modied assumption.There will also be repercussions for trade ows with other regions, including somewhat lower exports from most DevelopingAsian regions to China, due to Chinas reduced need for intermediate imported components if its net export ows grow less.Table 8(j) reports the impacts of this simulation in terms of inter-regional trade ows. Our results suggest total intra-developing Asia trade would be approximately 0.6 percent less, as a result of the importance of China in this region. However,

15 For an exploration of the impact on the above results of assuming a counterfactual that involves agricultural protection growth in developing countries,

see Anderson and Nelgen (2011).16 Changes in the trade balance are accomodated by allowing saving rates to reduce in China and increase in the US, given the relationship SI = XM,

consistent with the projections in Garnaut (2011). We note that the trade balance needs to be xed relative to regional income to preserve homogeneity in

the GTAP model, therefore it was necessary to iterate to drive the actual trade balances close to zero.

the percentage of total intra-developing Asian trade excluding China is only 0.1 percent lower than the core baseline, and wend that the relative contribution to global trade ows of trade between other developing regions changes by less than 0.2

K. Anderson, A. Strutt / Journal of Asian Economics 23 (2012) 303323 317percent. If global growth rates are not impacted by the trade rebalancing, these ndings suggest that this modication of thetrade balance assumption does not change the overall pattern of our main ndings. In particular, the Souths share in globaltrade still more than doubles, from 12.8 in 2004 to 28.6 percent in 2030 (compared with 29.6 percent in our core baselinethat does not constrain trade imbalances).

Third, we have aggregated the model into just 26 sectors/product groups. This leads to gross underestimation not only ofthe gains from trade reform but also of the extent to which rms can take advantage of intra-industry trade (not to mentionon-going fragmentation of global production into ever-more pieces whose location is footloose (Hanson, Mataloni, &Slaughter, 2005), which is not captured in our results).

Fourth, we have assumed constant returns to scale and perfect competition rather than allowing rms to enjoy increasingreturns and some degree of monopoly power for their differentiated products. This too leads to underestimates of the welfaregains from trade reform (Krugman, 2009). The fact that opening an economy exposes monopolistic rms to greatercompetition generates gains from trade reform that could be quite substantial in terms of reducing rm mark-ups, accordingto numerous country case studies (see, e.g., Krishna and Mitra, 1998 on India, Pavcnik, 2002 on Chile).

Fifth, where consumers (including rms importing intermediate inputs) value a greater variety of goods, or a greaterrange of qualities, intra-industry trade can grow as a result of both economic growth and trade policy reform, but that too isnot taken into account in the above analysis.

Sixth, in the trade reform scenarios we have not allowed domestic policies also to be reformed (apart from agriculturalsubsidies), even though it is typical for trade reforms including in the context of signing regional trade agreements to bepart of a broader program of microeconomic policy reform. Recent studies show, for example, that when labor markets arefreed up at the same time as trade, they can have very different welfare and bilateral trade effects than if those factor marketsremain inexible (Helpman & Itskhoki, 2010; Helpman, Marin, & Verdier, 2008). That is true also when nancial marketreforms are considered, not least because the inclusion of nancial markets allows an additional set of inuences on realexchange rates (see, e.g., McKibbin & Stegman, 2005). Hoxha, Kalemli-Ozcan, and Vollrath (2009) examine gains fromnancial integration and nd that a move from autarky to full integration of nancial markets globally could boost realconsumption by 7.5 percent permanently, even assuming no accompanying productivity gains. National case studies ofreform to services trade more generally also nd gains several times those from goods trade reform (e.g., Dee, Hanslow, &Pham, 2003; Konan & Maskus, 2006; Rutherford & Tarr, 2008). However, methodologies for estimating the extent of andeffects of globally removing barriers to services and factor ows between countries are far less developed thanmethodologies applied to trade in goods (Francois & Hoekman, 2010).

Seventh, the savings in bureaucratic costs of administering trade barriers, in traders costs of circumventing barriers, andin lobbyists costs of rent-seeking to secure or maintain trade-distorting policies are all non-trivial. Since they are notcaptured in the above modelling, this is another reason to consider the above welfare results from trade policy reform aslower-bound estimates.

Eighth, the standard GTAP model used here is comparative static and it does not measure the additional dynamic gainsfrom trade reform. Dynamic gains arise in numerous ways. One of the more important is through encouragement of themore-efcient rms to take over from the less efcient in each country (Bernard, Jensen, Redding, & Schott, 2007; Melitz,2003; Melitz & Ottaviano, 2008; Treer, 2004). Another way is through multinational rms sharing technologies andknowledge across countries within the rm (Markusen, 2002). Offshoring is yet another mechanism through whichheterogeneous rms are affected by trade liberalization, including via re-locating from small to larger nations (Baldwin &Okuba, 2011). The greater competition that accompanies trade reform also can stimulate more innovation (Aghion & Grifth,2005), leading to higher rates of capital accumulation and productivity growth (Lumenga-Neso, Olarreaga, & Schiff, 2005).

In short, the aggregate welfare gains from freeing up trade are likely to be far bigger than the estimates reported heresuggest, but their distribution, and the estimated bilateral patterns of global trade and relative GDPs of nations by 2030, alsomay be somewhat different if an empirical model with all of the above features had been available and used.17 We also notethat in the current modelling, we are not able to explicitly explore implications for poverty alleviation or environmentaloutcomes and their consequent impact on economies.

7. Conclusions

Should relatively rapid economic growth in Asia and to a lesser extent in other developing countries continue tocharacterize world economic development to the extent suggested above, the Souths share of global GDP and trade willcontinue to rise steeply over the next two decades. More particularly, the share of SouthSouth trade in global trade isprojected to more than double in the core projection, from 13 to 30 percentor to 32 percent if GDP and capital growth in theNorth were to be one-third slower than in the core projection (or if ASEAN + 6 opened up or all goods trade were to be freedglobally), and to 28 percent if there were to be a 1 percentage point slowdown in annual rate of productivity growth in theworlds primary industries.

17 For more on the challenges of enhancing standard global economy-wide models in these ways, see Francois and Martin (2010).

Our results suggest that a slowdown in economic growth in the North need not damage Developing Asia and otherdeveloping economies so long as the latter can keep growing. One important way to immunize developing economy growthfrom a slowdown in the North is through greater openness to SouthSouth trade. We nd that lowering barriers to SouthSouth trade, even to just the levels prevailing in SouthNorth trade, could bring 80 percent of the welfare gains to developingcountries as would ow from a freeing of all countries goods trade.

The results also suggest developing countries need not wait for a multilateral trade agreement to benet from freer trade: anagreement by members of the prospective ASEAN + 6 bloc to free their trade on an MFN basis could generate for Asia almost 40percent of the benet that is estimated to ow if the whole world so liberalized. Since Doha is likely to generate only a tinyfraction of the global gains from full trade reform (Anderson & Martin, 2006; Laborde, Martin, & van der Mensbrugghe, 2011),freeing up Asian trade under a broad regional agreement has the potential to bring even higher gains than Doha.

Acknowledgements

The authors are grateful for helpful interactions with Benno Ferrantini, Tom Hertel, Jayant Menon, Ernesto Valenzuela,Terrie Walmsley and Joseph Zveglich, and for funding support from the Asian Development Bank, the Australian ResearchCouncil, the Rural Industries Research and Development Corporation and Waikato Management School.

Appendix A

See Tables A1A8.

Table A1

Aggregations of regions in the GTAP Model.

Aggregations of regions Modelled regions Description Original GTAP regions

W. Europe WesternEurope EU27 and EFTA AUT BEL CYP CZE DNK EST FIN FRA DEU

GRC HUN IRL ITA LVA LTU LUX MLT NLD

POL PRT SVK SVN ESP SWE GBR CHE

NOR XEF BGR ROU

E. Europe Russia Russia RUS

RestEEurope Other Europe ALB BLR HRV UKR XEE XER TUR

US and Canada USA USA USA

Canada Canada CAN

ANZ Australia Australia AUS

NewZealand New Zealand NZL

Japan Japan Japan JPN

China China China CHN

ASEAN Singapore Singapore SGP

Indonesia Indonesia IDN

Malaysia Malaysia MYS

Philippines Philippines PHL

Thailand Thailand THA

Vietnam Vietnam VNM

RestSEAsia Cambodia, Laos, Brunei, Myanmar, KHM LAO XSE

K. Anderson, A. Strutt / Journal of Asian Economics 23 (2012) 303323318Timor Leste

Pacic Islands PacicIslands Pacic Countries XOC

Rest E. Asia HongKong Hong Kong HKG

SouthKorea South Korea KOR

Taiwan Taiwan TWN

RestNEAsia North Korea, Macau, Mongolia XEA

India India India IND

R. South Asia Pakistan Pakistan PAK

Bangladesh Bangladesh BGD

RestSAsia Afghanistan Bhutan Maldives, Nepal,

Sri Lanka

LKA XSA

Central Asia CentralAsia Arm Azeb Geo Kaz Kyr Taj Tkm Uzbek KAZ KGZ XSU ARM AZE GEO

Latin America Mexico Mexico MEX

Argentina Argentina ARG

Brazil Brazil BRA

RestLatAmer Other Latin America XNA BOL CHL COL ECU PRY PER URY VEN

XSM CRI GTM NIC PAN XCA XCB

M.E. and Africa ME_NAfrica Middle East and North Africa IRN XWS EGY MAR TUN XNF

SthAfrica South Africa ZAF

RestSSAfrica Sub-Saharan Africa NGA SEN XWF XCF XAC ETH MDG MWI

MUS MOZ TZA UGA ZMB ZWE XEC BWA XSC

Source: Authors compilation from www.gtap.org.

High-income countries (the North) are dened as the rst ve country groups in the table (i.e. the regions of WEurope, EEurope, USC, ANZ and Japan). The

rest are dened as developing countries (the South), of which China, ASEAN, PacicIslands, Rest E Asia, South Asia and Central Asia make up Developing

Asia in our analysis.

Table A2

Aggregations of sectors in the GTAP Model.

Aggregations of commodities Modelled commodities Description Original GTAP sectors

Agric. and food Rice Paddy and processed rice pdr pcr

Wheat Wheat wht

Fruit_Veg Vegetables, fruit, nuts v_f

Oilseeds Oil seeds osd

Sugar Raw and processed sugar c_b sgr

Cotton Plant-based bres pfb

Grains Other cereal grains gro

OtherCrops Other crops ocr

Beef_Sheep Beef and sheep ctl wol cmt

Pork_Chicken Pork and chicken oap omt

Dairy Dairy products rmk mil

OtherFood Other processed food vol ofd b_t

Other primary Fish_Forest Forestry and shing frs fsh

Coal Coal coa

Oil Oil oil

Gas Gas gas

OthMinerals Other minerals omn

Manufactures Text_App_Lea Textiles, apparel and leather tex wap lea

MotorVehicle Motor vehicles and parts mvh

Electronics Electronic equipment ele

OtherLtMan Other light manufacturing lum ppp fmp otn omf

HeavyManuf Heavy manufacturing p_c crp nmm i_s nfm ome

Services Utiliti_Cons Utilities and construction wtr cns

Elect_Gas Electricity and gas distribution ely gdt

Trade_transp Trade and transport trd otp wtp atp

OthServices Other Services cmn o isr obs ros osg dwe

Source: Authors compilation from www.gtap.org.

Table A3

Average annual GDP and endowment growth rates, 20042030 (percent).

GDP Population Unskilled

labor

Skilled

labor

Produced

capital

Agric. land Oil Gas Coal Other

minerals

W. Europe 1.48 0.14 1.09 1.50 1.60 0.28 2.81 0.77 2.51 2.07E. Europe 3.51 0.02 0.57 1.49 4.03 0.23 2.64 0.12 1.86 2.07US and Canada 2.09 0.82 0.17 1.59 1.75 0.20 1.00 0.14 0.19 2.07Australia and NZ 2.78 1.07 0.31 1.89 1.59 0.56 1.49 6.10 3.55 2.07Japan 0.92 0.21 1.45 0.98 0.40 1.14 0.00 0.00 9.34 2.07China 8.05 0.29 0.03 2.88 7.62 0.36 0.40 4.85 5.62 2.07ASEAN 5.25 0.97 0.45 3.67 5.95 0.17 1.31 1.48 11.71 2.07

Pacic Islands 3.66 1.72 2.30 1.88 3.86 0.19 1.54 1.21 0.15 2.07

Rest E. Asia 3.47 0.31 0.45 2.20 3.11 0.87 0.00 0.00 1.59 2.07India 7.88 1.18 1.37 4.03 7.27 0.04 0.24 0.00 4.93 2.07Rest S. Asia 7.23 1.36 1.99 4.93 8.14 0.10 0.27 2.18 2.26 2.07Central Asia 4.09 0.46 0.67 1.07 4.38 0.29 2.81 0.77 2.51 2.07Latin America 3.81 0.92 0.78 3.32 4.85 0.22 3.29 0.34 5.15 2.07ME and Africa 4.55 1.92 1.04 4.16 5.46 0.05 1.27 3.64 1.89 2.07

High-income 1.73 0.27 0.55 1.49 1.56 0.33 2.07 0.40 0.26 2.07Developing 5.28 1.03 0.52 3.28 5.81 0.09 1.48 2.24 5.57 2.07of which Asia: 6.24 0.78 0.26 2.93 6.26 0.16 0.72 0.93 5.93 2.07Total 2.45 0.88 0.37 1.68 2.65 0.17 1.67 1.23 2.50 2.07

Source: Authors assumptions (see text for details).

Table A4

Implied annual growth in total factor productivity for non-primary sectors,a 20042030.

2030 core (percent, using 2004

GDP values as weights)

2030 slower primary TFP (percent, using 2004 GDP

values as weights)

A B C A B C D

W. Europe 0.7 1.7 3.2 0.7 0.7 1.7 0.3E. Europe 1.2 2.2 3.8 1.2 1.2 2.2 0.3

US and Canada 1.0 2.0 3.5 1.0 1.0 2.0 0.0

Australia and NZ 1.4 2.4 4.0 1.4 1.4 2.4 0.5

Japan 1.0 2.0 3.5 1.0 1.0 2.0 0.0

China 2.9 3.9 5.5 2.9 2.9 3.9 2.3


Table A4 (Continued )

2030 core (percent, using 2004

GDP values as weights)

2030 slower primary TFP (percent, using 2004 GDP

values as weights)

A B C A B C D

ASEAN 1.3 2.4 3.9 1.3 1.3 2.4 0.4

Pacic Islands 0.6 1.6 3.1 0.6 0.6 1.6 0.5Rest E. Asia 1.7 2.7 4.2 1.7 1.7 2.7 0.9

India 3.1 4.2 5.7 3.1 3.1 4.2 2.6

Rest S. Asia 2.2 3.2 4.7 2.2 2.2 3.2 1.4

Central Asia 1.8 2.8 4.4 1.8 1.8 2.8 1.0

Latin America 0.7 1.7 3.2 0.7 0.7 1.7 0.4ME and Africa 0.8 1.8 3.4 0.8 0.8 1.8 0.2High income 0.9 1.9 3.4 0.9 0.9 1.9 0.1Total developing 1.7 2.7 4.3 1.7 1.7 2.7 0.9

Developing Asia 2.4 3.4 4.9 2.4 2.4 3.4 1.7

Total World 1.1 2.1 3.6 1.1 1.1 2.1 0.1


a The above TFP growth rates are those implied for the non-primary sectors by the GDP and factor growth rates in Appendix Table A3, with the following

assumptions about primary sector TFP growth. Primary sector TFP rates were exogenously set higher than those for the non-primary sectors to the following

extent in the core projection for all countries, with the aim of ensuring only modest growth in international relative prices for those products (shown in

Appendix Table A5): 1% for agriculture, lightly processed food and other minerals, 0% for fossil fuels, and 2.5% for the forestry and shing sector. In the

slower primary TFP growth scenario, the increment for forestry, shing is reduced to 1% p.a., the increment for agriculture and lightly processed food is

removed, and productivity growth in fossil fuels is assumed to be 1% lower than in non-primary sectors. For the trade reform scenarios, the core projections

TFP growth assumptions are maintained.

Column heading letters refer to:

A: non-primary sectors (and extractive sectors in the core baseline)

B: agriculture, lightly processed food and other minerals

C: forestry and shing

D: extractive sectors (coal, oil and gas)

Table A5

Cumulative changes in international prices, 20042030.

Price relative to global average output price change across all sectors, percent

Core

2030 sim

Slower

growth HICs

Slower

TFP growth

ASEAN +

no ag

ASEAN +

with ag

ASEAN +

MFN

Global

MFN

South

South libn

Rice 9.7 10.0 22.1 0.5 0.8 3.2 2.1 0.5Wheat 14.6 15.6 48.3 0.4 1.1 5.7 5.0 0.8Coarse Grains 22.0 23.3 61.2 0.1 0.6 1.5 3.4 0.1Fruit and Veg. 40.8 43.1 85.7 0.2 0.9 4.5 3.5 0.5Oilseeds 21.4 22.9 63.8 0.3 2.1 2.4 2.1 0.6Sugar 2.0 2.1 5.3 0.3 1.2 3.5 2.8 0.0Cotton 30.5 32.2 67.5 1.3 2.0 2.3 5.3 1.0Other Crops 12.8 13.3 48.8 0.2 1.2 2.0 1.6 0.6Beef and sheep 1.7 2.2 13.3 0.3 0.4 0.5 0.2 0.2Pork and chicken 12.7 13.7 24.6 0.0 0.7 2.5 3.0 0.1Dairy 2.1 2.0 7.9 0.3 0.8 0.3 1.1 0.3Other food 4.3 4.9 12.3 0.1 0.7 1.8 1.9 0.3Forest and sh 22.2 24.5 198.2 0.5 0.4 0.8 1.2 1.0Coal 12.3 14.5 9.6 0.2 0.6 0.4 0.5 0.7Oil 35.5 37.0 102.7 0.2 0.5 2.6 0.6 3.5

Gas 10.8 1.0 54.6 0.8 0.3 0.2 1.2 2.4Other minerals 23.8 27.6 91.7 0.6 0.9 1.2 0.6 2.1Text., App. and Lea. 3.8 4.2 8.1 0.6 0.8 2.0 1.9 0.0Motor vehicle 0.0 0.2 3.8 0.2 0.1 0.4 1.0 0.4Electronics 5.2 6.0 13.6 0.1 0.1 0.4 0.3 0.4Other Light Man. 0.9 1.2 1.2 0.2 0.1 0.4 0.4 0.1Heavy Manuf. 1.6 1.5 6.9 0.0 0.1 0.2 0.3 0.4Utiliti and Const. 1.0 0.9 1.1 0.1 0.2 0.2 0.1 0.1Elect and Gas 5.7 6.2 7.1 0.1 0.0 0.2 0.2 0.1Trade and Transp 1.5 1.5 6.8 0.1 0.1 0.4 0.3 0.0Other Services 2.3 2.3 8.2 0.0 0.1 0.4 0.5 0.2Aggregate prices

Agric. and food 8.9 9.7 24.5 0.0 0.8 1.7 1.3 0.3Other primary 24.7 25.3 100.8 0.1 0.3 0.9 0.1 2.0Manufactures 0.2 0.4 0.9 0.1 0.0 0.4 0.5 0.2Services 1.8 1.8 7.0 0.0 0.1 0.4 0.4 0.1


K. Anderson, A. Strutt / Journal of Asian Economics 23 (2012) 303323320

Table A6

Regional shares of world real GDP and population, and GDP per capita relative to world average, 2004 and the core projection for 2030 (percent).a

2004 GDP share 2030 GDP share 2004 population share 2030 population share 2004 GDP/capita 2030 GDP/capita

W. Europe 33.0 22.7 7.8 6.4 422.7 357.9

E. Europe 2.5 2.9 4.7 3.7 54.3 79.1

USC 30.9 24.9 5.1 5.0 604.2 502.0

ANZ 1.8 1.7 0.4 0.4 479.6 444.0

Japan 11.4 6.8 2.0 1.5 569.3 457.8

China 4.1 14.4 20.4 17.3 20.0 83.3

ASEAN 1.9 3.5 8.6 8.7 22.4 40.5

PacicIslan 0.1 0.1 0.1 0.2 38.2 37.4

RestEAsia 2.9 3.3 1.6 1.4 178.1 239.1

India 1.6 5.3 17.0 18.1 9.2 29.3

RestSouthAsia 0.5 1.3 5.8 6.5 7.8 20.3

CentralAsia 0.2 0.3 1.1 0.8 17.8 34.0

LatinAmerica 5.3 6.7 8.6 8.6 61.2 77.8

MEast_Africa 4.0 6.2 16.8 21.7 23.9 28.4

High-income 79.6 59.0 20.0 16.9 398.7 349.9

Developing 20.4 41.0 80.0 83.1 25.5 49.3

of which Asia: 11.1 28.1 54.6 52.8 20.4 53.2

Total 100.0 100.0 100.0 100.0 100.0 100.0


a 2004 prices.

Table A7

Average import tariffs by sector and region, 2004.

North (percent) Rest of South (percent) World (percent)

(a) Tariffs faced by the South when exporting to the North, to other South regions, and to the World

Rice 88.2 20.2 32.0

Wheat 14.6 4.6 6.4

Coarse grains 20.8 21.8 21.5

Fruit and Veg 11.6 11.2 11.5

Oilseeds 2.7 17.2 10.4

Sugar 86.1 16.4 44.8

Cotton 0.3 1.5 1.1

Other crops 5.8 10.7 7.0

Beef and sheep 58.2 6.8 31.5

Pork and chicken 16.8 9.2 14.1

Dairy 30.2 7.9 12.4

Other food 6.8 16.6 10.9

Forest and sh 2.2 5.8 3.9

Coal 0.0 3.1 1.3

Oil 0.1 3.9 1.5

Gas 0.1 0.6 0.2

Other minerals 0.1 2.1 0.9

Text., App. and Lea. 8.7 13.1 10.1

Motor vehicle 2.3 13.0 5.5

Electronics 0.7 1.8 1.1

Other Light Manuf. 1.4 7.4 3.3

Heavy Manuf. 1.3 6.5 3.9

Average across all goods and services 2.5 6.0 3.9

Exporting region Importing region

North South World

(b) Average tariffs imposed by aggregate region

North 1.04 5.71 2.15

South 2.46 6.05 3.86

World 1.45 5.86 2.73

Source: Derived from the authors GTAP Model baseline.


K. Anderson, A. Strutt / Journal of Asian Economics 23 (2012) 303323322Table A8

Sectoral shares of value added at constant (2004) prices, 2004 and 2030.

Agric. and food (percent) Other primary (percent) Manufactures (percent) Services (percent) Total (percent)

(a) 2004

W. Europe 4.8 1.3 18.1 75.8 100.0

E. Europe 10.5 10.3 14.1 65.1 100.0

US and Canada 3.0 1.4 14.3 81.3 100.0

ANZ 6.3 4.6 12.1 77.0 100.0

Japan 3.3 0.4 16.9 79.4 100.0

China 12.0 6.4 34.4 47.3 100.0

ASEAN 12.6 10.0 26.4 51.0 100.0

Pacic Islands 5.9 8.8 12.8 72.6 100.0

Rest E. Asia 3.5 1.1 25.0 70.3 100.0

India 27.5 5.2 13.5 53.8 100.0

Rest S. Asia 22.3 3.8 10.6 63.3 100.0

Central Asia 14.1 21.2 9.0 55.6 100.0

Latin America 11.8 5.3 19.9 62.9 100.0

M.E. and Africa 11.6 26.3 9.9 52.3 100.0


Total 5.7 3.2 17.3 73.9 100.0(b) 2030 core

W. Europe 5.0 2.1 16.0 76.9 100.0

E. Europe 8.2 12.7 12.8 66.3 100.0

US and Canada 3.3 1.8 12.6 82.3 100.0

ANZ 6.6 7.0 8.4 78.0 100.0References

Aghion, P., & Grifth, R. (2005). Competition and growth: Reconciling theory and evidence. Cambridge MA: MIT Press.Allen, B. (2011). Global economic history: A very short introduction. London: Oxford University Press.Alston, J. M., Babcock, B. A., & Pardey, P. G. (Eds.). (2010). The shifting patterns of agricultural production and productivity worldwide. Ames, IA: Midwest Agribusiness

Trade Research and Information Center, Iowa State University.Anderson, K. (2009). Distorted agricultural incentives and economic development: Asias experience. The World Economy, 32(March (3)), 351384.Anderson, K., Dimaranan, B., Hertel, T., & Martin, W. (1997). Economic growth and policy reforms in the APEC region: Trade and welfare implications by 2005. Asia-

Pacic Economic Review, 3(April (1)), 118.Anderson, K., & Martin, W. (Eds.). (2006). Agricultural trade reform and the Doha development agenda. London/Washington DC: Palgrave Macmillan/World Bank.Anderson, K., & Nelgen, S. (2011). Whats the appropriate agricultural protection counterfactual for trade analysis? In W. Martin & A. Mattoo (Eds.), The WTOs Doha

agenda: An assessment. London: Centre for Economic Policy Research for the World Bank. Ch. 13.Anderson, K., & Norheim, H. (1993). Is world trade becoming more regionalized? Review of International Economics, 1(June (2)), 91109.Anderson, K., & Strutt, A. (2011a). Asias Changing Role in World Trade: Prospects for SouthSouth Trade Growth to 2030, Background Paper for Chs. 2 and 4 of the

Asian Development Banks Asian Development Outlook 2011. Since published as ADB Economics Working Paper 264, Manila, July.Anderson, K., & Strutt, A. (2011b). Agriculture and Food Security in Asia by 2030, Background Paper for an Asian Development Bank Institute project on ASEAN,

China and India: A Balanced, Sustainable, Resilient Growth Pole, mimeo, Tokyo, October.Anderson, K. & Strutt, A. (2012). South Americas Contribution to World Food Markets: GTAP Projections to 2030, paper for a IATRC-IAAE Pre-Congress Symposium

on Globalization, Macroeconomic Imbalances and South Americas Potential to be the Worlds Food Basket, Iguazu, Brazil, 18 August.Armington, P. (1969). A theory of demand for products distinguished by place of production. IMF Staff Papers 16, 159178.Bairoch, P. (1982). International industrialization levels from 1750 to 1980. Journal of European Economic History, 11, 269331.Baldwin, R., & Okuba, T. (2011). International trade, offshoring and heterogeneous rms. Cambridge, MA: NBER Working Paper 16660, January.Bernard, A. B., Jensen, J. B., Redding, S. J., & Schott, P. K. (2007). Firms in international trade. Journal of Economic Perspectives, 21(Summer (3)), 105130.BP. (2010). BP statistical review of world energy. London: British Petroleum.Chappuis, T., & Walmsley, T. L. (2011). Projections for World CGE Model Baselines. West Lafayette IN: GTAP Research Memorandum No. 22, Center for Global Trade

Analysis, Purdue University. September.Citi. (2011). Global growth generators: Moving beyond Emerging Markets and BRIC, Global Economics View. New York: Citigroup Global Markets. 21 February.Crafts, N., & Venables, A. J. (2003). In M. Bordo, A. Taylor, & J. Williamson (Eds.), Globalization in historical perspective (pp. 323364). Chicago: University of Chicago

Press for the NBER.

Japan 3.4 0.6 13.5 82.5 100.0

China 5.6 4.3 38.4 51.7 100.0

ASEAN 8.2 12.0 27.5 52.4 100.0

Pacic Islands 4.7 8.9 13.7 72.6 100.0

Rest E. Asia 2.8 1.5 23.6 72.1 100.0

India 13.5 3.8 15.0 67.7 100.0

Rest S. Asia 10.9 2.3 13.2 73.6 100.0

Central Asia 11.7 22.9 5.2 60.2 100.0

Latin America 9.5 7.5 20.3 62.8 100.0

M.E. and Africa 9.7 19.1 11.6 59.6 100.0


Total 5.8 4.5 18.9 70.8 100.0


Dee, P., Hanslow, K., & Pham, D. T. (2003). Measuring the cost of barriers to trade in services. In T. Ito & A. O. Krueger (Eds.), Services Trade in the Asia-Pacic Region.Chicago: University of Chicago Press for the NBER.

Deininger, K., & Byerlee, D. (2011). Rising global interest in farmland: Can it yield sustainable and equitable benets? Washington DC: World Bank.FAO/OECD. (2010). FAO/OECD agricultural outlook 20102019. Paris/Rome: OECD/FAO.Feldstein, M. S. (2011). The role of currency realignments in eliminating the US and China current account imbalances. Cambridge MA: NBER Working Paper 16674.

January.Feenstra, R., Ma, H., Neary, J. P., & Rao, D. S. P. (2011). Who Shrunk China? Puzzles in the measurement of real GDP London: CEPR Discussion Paper 8592. October.Foure, J., Benassy-Quere, A., & Fontagne, L. (2010). The world economy in 2050: A t

Documents

The changing geography of world trade: Projections to 2030