Aid, trade and migration: How do bilateral flows interact?

• Author: Audrey Menard,Aurore Gary
• Published date: 01 February 2018
• DOI: http://doi.org/10.1111/twec.12550
• Date: 01 February 2018

ORIGINAL ARTICLE

Aid, trade and migration: How do bilateral flows

interact?

Audrey Menard

1

|

Aurore Gary

2

1

Economics, Universite de Nantes, Nantes, France

2

Economics, CES, Universit

e Paris 1, Panth

eon-Sorbonne, Paris, France

1

|

INTRODUCTION

The possible damaging effect of immigration on unemployment is a source of anxiety for policy-

makers and citizens alike. Migration policies in most destination countries are generally strict,

entailing binding quotas on the number of migrants. However, even countries with official

immigration constraints still accept undesired (but legal) immigration. Family-reunification and

asylum-seekers policies can explain ongoing migration inflows to developed economies. The new

migration plan proposed by the European Union in June 2016 reassesses the European Commis-

sion’s position defended in 2005: promoting development in migrant-sending countries is now

intended to address a “root cause”of migration, and so allow donors to achieve their immigration

goals. For some politicians and academics, increasing aid flows and liberalising trade with

developing countries are “smarter solutions”to fight immigration as compared to restrictive migra-

tion policies (De Haas, 2007; Stalker, 2002). Aid, via international cooperation, is also expect ed to

bolster developing countries’efforts to regulate migration outflows.

Yet, as reported in Berth

elemy, Beuran, and Maurel (2009), aid flows towards low to medium-

income countries can in fact increase the demand for entry into the donor country, in turn render-

ing its migration policies more restrictive. Here, we extend their study in two ways: (i) we account

for bilateral trade relationships; and (ii) we examine the effects of unemployment in develo ped

countries on the amount of foreign assistance they deliver and on the number of migrants they

receive.

First, we observe that promoting free trade has been perceived as better than aid as a way to

favour development and, therefore, to decrease the economic drivers of emigration (Stalker, 2002;

Winters, McCulloch, & McKay, 2004). Second, we argue that the tightness of the labour market

(say, an increase in unemployment rates in the assistance-providing country) may not only exert

downward pressures on migration as suggested by Azam and Berlinschi (2009), but also affect the

amount of foreign aid it supplies. Governments in industrialised nations may be more likely to

heighten their migration restrictions when unemployment increases (due to a higher demand for

protection). Hanson and Spilimbergo (2001) showed that US controls at borders strengthen when

sectors employing immigrants face recession. Accordingly, migration quotas are affected by

changes in economic conditions in the destination country. But imposing binding quotas on the

DOI: 10.1111/twec.12550

World Econ. 2018;41:431–456. wileyonlinelibrary.com/journal/twec ©2017 John Wiley & Sons Ltd

|

431

number of migrants may not be sufficient to reduce pressure on the labour market (due to time or

legal restrictions). Hence, donors may want to (and indeed are asked to) expand their foreign assis-

tance to lessen the incentives for migrating.

Policy coherence has widely been recognised by OECD countries as a priority (Gary & Maurel,

2015; Manning & Hradsky, 2008). Surprisingly, however, interactions between the outcomes of

policies that relate developed to developing countries have not received more systematic attention.

This is precisely the goal of this paper. Our main contribution includes a greater understanding of

how aid, migration and trade affect each other. In particular, we shed some quantitative ligh t on

the possible effects of increasing aid levels and liberalising trade on the number of people migrat-

ing from the recipient to the donor economy.

Regarding quantitative evidence, studies are usually based on pairs of policy outcomes, rarely

focusing on a bidirectional relationship. Concerning the aid–migration relationship, we expect to

find evidence of bidirectional causation. First, donors can use foreign assistance as a policy instru-

ment for limiting inflows of migrants (Azam & Berlinschi, 2010), given that aid reduces income

differentials between origin and destination countries (Angelucci, 2004). Berth

elemy et al. (2009)

found that this effect holds in countries above a critical income threshold, equal to US$7,300 per

capita in PPP 2000 prices. Otherwise, aid flows increase migration pressures in OECD countries,

thereafter encouraging restrictive migration policies. Faini and Venturini (1993) confirmed that aid

encourages migration flows from relatively poor countries: aid increases revenues, allows citizens

to bear the financial burden of migrating, and helps disseminate information on the donor econ-

omy. Second, Anwar and Michaelowa (2006) and Lahiri and Raimondos-Møller (2000) have

explained how ethnic groups exert pressure on Northern governments to allocate more aid towards

their homeland.

We also expect that aid and trade would affect each other in both directions. Lundsgaarde,

Breunig, and Prakash (2007) evidenced a “displacement effect”from aid to trade. Imports from

developing countries reduce the amount of assistance, suggesting that extending trade with

developing countries is perceived as a better way to promote economic development (Winters

et al., 2004). On the other hand, aid can be tied to donors’exports (Canavire-Bacarreza, Nun-

nenkamp, Thiele, & Trive~

no, 2006; Claessens, Cassimon, & Van Campenhout, 2009; Dollar &

Levin, 2006): bilateral donors are prone to reward trade partners and new market opport unities.

Finally, aid efforts can support trade activities, in particular since the inception of the “Aid for

Trade”initiative, in which aid is associated with economic integration (see Cadot, Fernandes,

Gourdon, Mattoo, and Melo, 2014; Vijil, 2014 for details). Studies have found that up to one-

third of assistance delivered is spent directly on imports of goods from the donor economy

(Wagner, 2003).

Turning to the migration–trade relationship, international migration flows and international trade

can be negatively related and perceived as substitutes (Faini & Venturini, 1993). Accordingly, lib-

eralising trade should decrease incentives to migrate. However, if immigrants do prefer goods pro-

duced in their country of origin, migration and trade can easily be complements (Campaniello,

2014). In addition, migration can also reduce trade costs (see, for instance, Felbermayr and Toubal,

2012). Migrant networks can facilitate bilateral economic transactions thanks to lower transacti on

costs and higher cultural proximities between host and origin countries (Ehrhart, Le Goff, Emma-

nuel, & Singh, 2014; Rauch & Casella, 2003). In return, if bilateral trade increases between two

countries, migrants can be attracted to the exporting country in expectation of its growth potential

(Felbermayr, Grossmann, & Kohler, 2015).

Our main goal is to assess simultaneous interactions between aid, trade and migration flows

between developing (South) and developed (North) countries. This question is particularly relev ant

432

|

MENARD AND GARY