Simulating World Trade in the Decades Ahead: Driving Forces and Policy Implications

• Date: 01 January 2017
• Published date: 01 January 2017
• Author: Alexander Keck,Lionel Fontagné,Jean Fouré
• DOI: http://doi.org/10.1111/twec.12479

Simulating World Trade in the Decades

Ahead: Driving Forces and Policy

Implications

Lionel Fontagn

e

1,2

, Jean Four

e

2

and Alexander Keck

3

1

PSE, University Paris 1, Paris, France,

2

CEPII, Paris, France and

3

Economic Research and Statistics

Division, WTO, Geneva, Switzerland

1. INTRODUCTION

IN the recent past –perhaps owing to perceptions of increased uncertainty in the wake of

the Great Recession and the related slowdown in the growth of world trade –a number of

long-term economic projections have been made, ranging from a time horizon of several

decades to almost one hundred years ahead, with some of these studies also touching upon the

prospects for international trade. These include qualitative, multidisciplinary analyses, such as

European Commission (2012), multisector partial equilibrium models, such as IMF (2012) as

well as computable general equilibrium (CGE) models, such as World Bank (2007). While the

studies make assumptions about productivity advances, which then shape sectoral develop-

ments, they do not employ an explicit model of economic growth. Other studies, such as Duval

and de la Maisonneuve (2010), Four

e et al. (2013), Asian Development Bank (2011) and

OECD (2012), provide economic growth projections over the long term. Only a few papers

have combined the two approaches, linking models of economic growth to sectoral simula-

tions. Petri and Zhai (2013) use the growth projections by the ADB (2011) as a baseline for

their CGE model. They focus on the economic prospects of the Association of Southeast Asian

Nations (ASEAN), China and India by the year 2030. Anderson and Strutt (2012) make predic-

tions for Asia at the same time horizon (2030), supplementing the growth predictions from

ADB (2011) with the ones from Four

e et al. (2013) for countries not contained in the former

study. Finally, Fontagn

e et al. (2013) combine CEPII’s macroeconomic model (MaGE) with

its multisectoral dynamic CGE model of the world economy (MIRAGE). This study considers

economic prospects up to the year 2100, making some simplifying assumptions about key

exogenous variables, and principally aims at providing a baseline for the evaluation of climate

change policies, although some sectoral and trade issues are discussed as well.

Our paper seeks to extend this relatively small, but rapidly expanding body of the literature

in a number of ways: first, using the methodology described in detail in Fontagn

e and Four

e

(2013), we combine an explicit economic growth model with a multisectoral model in order

to construct scenarios for the world as a whole (around 150 countries) up to a maximum time

horizon (year 2035), for which reasonable forecasts of a number of key variables (e.g. energy

prices, demographics) can be obtained from specialised institutions (e.g. IEA, 2011; UNPD,

The opinions expressed in this paper are not meant to represent the positions or opinions of the WTO

and its members and are without prejudice to members’ rights and obligations under the WTO. The

authors would like to thank Olivier Cadot, Lucian Cernat, Patrick Low, Marcelo Olarreaga, Fr

ed

eric

Robert-Nicoud and Robert Teh as well as conference participants at the University of Geneva, Univer-

sity of Mauritius, European Commission and WTO for helpful comments on this paper. All errors and

omissions remain our own.

©2016 John Wiley & Sons Ltd

36

The World Economy (2017)

doi: 10.1111/twec.12479

The World Economy

2011). To the best of our knowledge, our study is the only exercise at this extensive geo-

graphic and timescale, while remaining grounded in realistic assumptions about key economic

variables. Second, unlike other papers, we provide an entirely consistent modelling frame-

work, which makes the macroeconomic scenarios fully traceable throughout the sectoral simu-

lations. This is crucial to be able to identify the relative importance of key economic drivers

of future developments. Third, we undertake to examine to what extent the main trends that

have strongly altered the volume, composition, geography and nature of international trade in

recent times (WTO, 2013) are likely to continue or reverse over the next decades. This is

important in order to be able to identify challenges that may arise from possible changes to

existing trade developments. Finally, given the wide scope of our simulations, we are able to

determine the policy implications of different scenarios for different groups of countries,

which may allow policymakers to focus on the issues of principal concern to their constituen-

cies and evaluate their options.

The paper is organised as follows: Section 2 briefly describes the macroeconomic and trade

scenarios we implement and highlights key methodological advantages as well as limitations

of our approach. Section 3 examines the relative impact of the main fundamental economic

driving forces. In Section 4, we analyse whether, under these scenarios, major trends in trade

are likely to continue. In both of these sections, we discuss what the implications would be

for different groups of countries. Section 5 concludes providing an overview of major policy

challenges ahead and ideas for further work.

2. METHODOLOGY AND SCENARIOS

First, this section discusses the advantages and limitations of our dynamic CGE approach.

Second, the scenarios in terms of shocks to various economic and trade variables are intro-

duced.

a. Methodology and Limitations

This paper combines CGE modelling (MIRAGE) with the macroeconometric model

MaGE

1

to address the sectoral implications of the envisaged growth trajectory of national

economies.

2

The driving forces of the changes modelled with MIRAGE are the change in

countries’ GDP per capita, the (sectorally differentiated) gains in total factor productivity

(TFP) by these countries, their changing comparative advantage and evolving country sizes.

On the demand size, country size and GDP per capita determine the size of markets and

demand composition. With the reshaping of the world economy, the patterns of international

demand will be affected profoundly. Meanwhile, the capabilities of countries will change, as

developing economies accumulate human capital, invest in infrastructure and new production

capacities and progressively catch up in terms of efficiency. Finally, markets will change as

well as exporters, leading to a profound revamping of world trade patterns. MIRAGE is pro-

viding a detailed description of these changes for a given classification of countries and

regions and a given sectoral aggregation of traded goods and services. The transformations of

1

The versions used in this paper are, respectively, MIRAGE-e 1.0 (based on GTAP 7) and MaGE 2.1.

2

A methodological companion paper to the policy-oriented analysis presented here discusses the two

modelling frameworks used in great depth. See Fontagn

e and Four

e (2013). Both are also jointly referred

to in WTO (2013).

©2016 John Wiley & Sons Ltd

SIMULATING WORLD TRADE IN THE DECADES AHEAD 37