EMU and trade: A PPML re‐assessment with intra‐national trade flows

• Author: Silviano Esteve‐Pérez,Salvador Gil‐Pareja,José Antonio Martínez‐Serrano,Rafael Llorca‐Vivero
• DOI: http://doi.org/10.1111/twec.12960
• Date: 01 October 2020
• Published date: 01 October 2020

2574

|

wileyonlinelibrary.com/journal/twec World Econ. 2020;43:2574–2599.

© 2020 John Wiley & Sons Ltd

Received: 31 October 2018

|

Revised: 17 March 2020

|

Accepted: 3 April 2020

DOI: 10.1111/twec.12960

ORIGINAL ARTICLE

EMU and trade: A PPML re-assessment with intra-

national trade flows

SilvianoEsteve-Pérez1

|

SalvadorGil-Pareja1

|

RafaelLlorca-Vivero1

|

José AntonioMartínez-Serrano2

1University of Valencia and INTECO, Valencia, Spain

2University of Valencia, Valencia, Spain

Funding information

Ministerio de Ciencia, Innovación y Universidades, Grant/Award Number: project RTI2018-100899-B-I00; Generalitat

Valenciana, Grant/Award Number: PROMETEO/102/2018

KEYWORDS

country-level heterogeneity, EMU, exports, gravity equation, imports

1

|

INTRODUCTION

Since Rose’s (2000) path-breaking study, a lot of studies have been carried out on the effect of cur-

rency unions on trade. Both Rose's striking finding that sharing a currency union more than triples

trade between countries and the creation of the euro have propelled an intense debate on this issue and,

in particular, about the effect of the Economic and Monetary Union (EMU) on trade.1 More than 50

papers have examined the effect of EMU on bilateral trade flows given that it is, by far, the most im-

portant monetary union.2 However, so far the results vary greatly across studies and even the most

recent articles provide mixed results. Whereas Glick and Rose (2016) and Larch, Wanner, and Yotov

(2018) find a positive EMU effect on trade, Mika and Zymek (2018) and Larch, Wanner, Yotov, and

Zylkin (2019) provide no evidence of a positive effect on trade. Therefore, the debate is still

ongoing.

1See Santos Silva and Tenreyro (2010) for a comprehensive review of the earlier empirical literature on the effect of currency

unions on trade. More recent papers on this area include Eicher and Henn (2011a), De Souza (2012), Campbell (2013), Roy

(2014), Glick and Rose (2016), Campbell and Chentsov (2017) and Larch el al. (2019). Rose (2017) reviews the impact of

EMU on trade.

2EMU refers to the process of harmonising the economic and monetary policies of EU Member States with the aim to create

the euro. EMU was accomplished in three stages. In this paper, EMU refers to the third stage of the process that came into

effect on 1 January 1999 for the eleven early joiners of the euro. The third stage involved the irrevocable fixing of exchange

rates and introduction of the single currency on the foreign exchange markets and for electronic payments.

|

2575

ESTEVE-PÉREZ ET al.

Furthermore, the literature of the EMU effect on trade up to date, with a few exceptions (see, e.g.,

Aristotelous, 2006; Baldwin & Di Nino, 2006; Faruqee, 2004; Gil-Pareja, Llorca-Vivero, & Martinez-

Serrano, 2003; Hashemi, 2016; Micco, Stein, & Ordonez, 2003), focuses on the overall EMU effect

without examining whether there are significant differences across the individual EMU members.

Besides, a caveat of all previous studies on the EMU effect on trade with the exception of Larch

etal.(2018) is that t hey rely on international trade flows only (omitting intra-national trade flows),

which prevents a proper assessment of both trade creation effects of the EMU among its members and

unilateral effects with third countries.

The contribution of this paper to the literature is threefold. First, this paper examines the EMU

impact on bilateral trade for the eleven early joiners (i.e., the countries that joined the EMU in 1999)

plus Greece relying, for the first time, on international and intra-national trade flows in line with the

microfoundations of the gravity equation of trade. Second, we properly assess the impact of EMU on

trade both with other EMU members and with third countries for each individual member country,

using also the latest developments in the empirical structural gravity literature. Finally, we analyse the

EMU impact across countries taking into account the direction of the trade flows (exports versus im-

ports).3 We further discuss these points in turn.

On the one hand, the inclusion of intra-national trade flows allows us to capture the possibility of

trade creation, that is, the replacement of domestic sales with trade with other EMU countries. As

noted by Larch etal.(2018), such possibility cannot be accounted for in studies that only employ in-

ternational trade flows.4 Additionally, the introduction of intra-national trade flows also enables us to

identify country-specific EMU effects on trade between EMU members and non-member countries.

All previous studies that try to estimate the EMU effect on trade with third countries cannot account

properly for multilateral resistance terms. This is because when relying only on international trade

flows, a multicollinearity problem precludes the estimation of the unilateral EMU membership effects

in the presence of the full set of exporter-time and importer-time fixed effects (the usual way to ac-

count for multilateral resistance terms in panel data sets).5

On the other hand, this paper accounts for heterogeneity of EMU effects across member countries.

The importance of heterogeneity across country members and within country pairs of a currency

union has been pointed out by the theoretical and empirical literature on integration processes (Chen

& Novy,2018). Heterogeneity across EMU members (even among the “early joiners”) is significant,

something that has undoubtedly raised important challenges during the 2008 crisis. The empirical ev-

idence suggests the existence of heterogeneity along several dimensions, such as the degree of devel-

opment of the economies (Santos Silva & Tenreyro, 2010), the size of the countries (Baldwin,2006),

the degree of product differentiation (Flam & Nordström, 2006, 2007) and the size and productivity

of the firms (Berthou & Fontagné, 2008; Hashemi,2016).

3Previous studies of the EMU effect on trade across countries estimate atheoretical log-linear versions of the gravity equation

ignoring multilateral resistance price terms (Anderson & van Wincoop,2003). In addition to the potential bias derived from

the omission of multilateral resistance terms, these studies do not account for other potential sources of bias derived from the

heteroscedasticity of the residuals, the existence of zeros in bilateral trade flows and the omission of intra-national trade

flows. This paper deals with all these issues simultaneously.

4Yotov (2012), Dai,Yotov, and Zylkin (2014), Bergstrand,Larch, and Yotov (2015), Anderson and Yotov (2016), Borchert

and Yotov (2017), Matoo,Mulabdic, and Ruta (2017), Heid,Larch, and Yotov (2017), Beverelli,Alexander, Larch, and

Yotov (2018) and Baier,Kerr, & Yotov (2019) also highlight the importance of taking into account the intra-national trade

flows in the estimation of the gravity equation of international trade.

5See Heid etal.(2017) and Beverelli etal.(2018) for a more formal discussion of the collinearity issues with country-specific

variables in structural panel gravity regressions.