Formation of interdependent regional trade agreements and production networks

• Published date: 01 October 2017
• DOI: http://doi.org/10.1111/twec.12489
• Author: Innwon Park,Soonchan Park
• Date: 01 October 2017

ORIGINAL ARTICLE

Formation of interdependent regional trade

agreements and production networks

Innwon Park

1

|

Soonchan Park

2

1

Division of International Studies, Korea University, Seoul, Korea

2

Department of Economics and International Trade, Kongju National University, Kongju, Korea

1

|

INTRODUCTION

The number of regional trade agreements (RTAs) has been increasing over the last two decades.

As of November in 2015, the WTO has been notified of 413 RTAs. Among them, 18 (4.4%) are

enlarged RTAs and 395 (95.6%) are new RTAs.

1

The domino theory of regionalism introduced by

Baldwin (1993) is a common explanation of why RTA formation has been growing. Countries

excluded from an RTA seek membership in existing RTAs for a net positive trade creation effect;

alternatively, they may be compelled to negotiate new RTAs among non-members, in order to

counteract the intimidation of exclusion.

Unlike the extensive body of theoretical literature devoted to countries’choices in forming

RTAs, empirical research on the formation of RTA—featuring a focus on the possible enlargemen t

of existing RTAs and the competitive formation of new RTAs among non-members—has only

recently gained traction. Baier and Bergstrand (2004), Egger, Egger, and Greenaway (2008) and

Magee (2003) empirically analyse endogenous determinants of RTA formation and predict the like-

lihood of country-pairs forming RTAs. They mainly consider the participating countries’econom ic

characteristics and highlight positive trade creation between members of a freer trade bloc as a

main incentive to RTA formation. On the other hand, Baldwin and Jaimovich (2012) emphasise

the worrisome trade diversion effect as another incentive to members of an RTA at the expense of

third countries.

The third-country effect on the probability of a country-pair having an RTA explains the forma-

tion of interdependent RTAs.

2

Chen and Joshi (2010) empirically test the third-country effect on

the formation of RTAs by measuring the net welfare effect and find that not only the participating

countries’economic characteristics but also third-country effect play an important role in those

countries’decisions to establish new RTAs. Egger and Larch (2008) test the third-country effect

that relates to enlarged RTAs and to new RTAs by applying spatial econometric methods. They

find that the enlargement of existing RTAs is more likely to occur, compared to the possible for-

mation of new RTAs; they also find that interdependent links will be weakened by higher trade

costs, measured in terms of bilateral distance, and strengthened by bilateral trade volume. Baier

1

See the WTO website: http://rtais.wto.org/UI/publicsummarytable.aspx.

2

As in Baier, Bergstrand, and Mariutto (2014), RTA interdependence is defined as the effect of other RTAs on the probabil-

ity of a country-pair creating an RTA.

DOI: 10.1111/twec.12489

2032

|

©2017 John Wiley & Sons Ltd wileyonlinelibrary.com/journal/twec World Econ. 2017;40:2032–2055.

et al. (2014) simultaneously estimate the two sources of RTA interdependence with a formal theo-

retical model and successfully decompose RTA interdependence into cases of RTA enlargement

(“own-FTA”) and the formation of new RTAs (“cross-FTA”). They find that, as sources of RTA

interdependence, own-FTA effects are much stronger than cross-FTA effects.

3

The existing empirical literature that tests the formation of interdependent RTAs persists in tra-

ditional gravity factors. To answer this question, Baier et al. (2014) test the third-country effects of

market size, measured in terms of GDP, on the deepening interdependence of RTA; Egger and

Larch (2008), meanwhile, estimate the positive probability of RTA formation as driven by cheaper

trade costs and as measured in terms of bilateral distance. The current study proposes production

networks (PNs) as another important driving force of the formation of interdependent RTAs.

Production networks may increase the trade creation effect and reduce the trade diversion effect

of RTAs, and aid in the proliferation of RTA formation. Fast-growing network trade will enhance

economic interdependence among interconnected countries, and the deepening interdependence will

generate more RTA-related benefits over time (Arndt, 2001; Ramondo & Rodr

ıguez-Clare, 2013).

However, to date, no empirical research attempt has been made to test the direction of linkage

from deepening PN to RTA formation. Orefice and Rocha (2014) is related to an empirica l study

of the two-way positive relationship between PNs, measured in terms of the import values of parts

and components, and deeper integration, measured in terms of a set of indices; however, they do

not directly investigate the bilateral relationship between PNs and probability of forming RTAs.

We attempt to fill this research gap by applying a qualitative choice econometric regression

model (probit) to panel data covering bilateral country-pairs among 147 countries between 2000

and 2010. More specifically, this study analyses whether deepening PNs, measured in terms of

intermediate-good trade volume, increase the likelihood that interdependent RTAs will form with

considering the third-country effect. In addition, we compare the member-specific network effects

of RTAs by classifying each record of the entire sample in terms of its level of economic develop-

ment, developed versus developing countries. The possible endogeneity problem is carefully exam-

ined by applying an instrumental variable (IV) regression method.

This paper is organised as follows. Section 2 introduces the bilateral link between the formation

of interdependent RTAs and deepening PNs by constructing a theoretical model. Section 3

describes the model specifications and data used for an empirical experiment. Section 4 sum-

marises the empirical findings, and Section 5 concludes this research.

2

|

PRODUCTION NETWORKS AND FORMATION OF RTA:

THEORETICAL CONSIDERATIONS

2.1

|

Production networks, trade and RTA: observations

The PN can be measured by the following three approaches: (i) investigating arm’s lengt h transac-

tions between independent firms and intra-firm trade of multinational enterprises (Anderson & Fre-

driksson, 2000; Ando & Kimura, 2003; Braunerhjelm, 1998; Hanson, Mataloni, & Slaughter, 2001;

Slaughter, 2000); (ii) quantifying vertically linked production processes across multiple countries

using world input–output tables (Hummels, Ishii, & Yi, 2001; Johnson & Noguera, 2012a, 2012b;

Koopman, Wang, & Wei, 2014); and (iii) calculating trade share of intermediate goods using

3

They define the “own-FTA effect”as the impact on the net welfare gains of a free trade agreement (FTA) between the two

countries involved because of already established FTAs—and the “cross-FTA effect,”as an impact on the net welfare gains

of an FTA between the pair—or due to other FTAs existing in the rest of the world (Row).

PARK AND PARK

|

2033