Foreign assistance and emigration: Accounting for the role of non‐transferred aid

• Published date: 01 July 2020
• Date: 01 July 2020
• Author: Mauro Lanati,Rainer Thiele
• DOI: http://doi.org/10.1111/twec.12914

World Econ. 2020;43:1951–1976. wileyonlinelibrary.com/journal/twec

|

1951

© 2019 John Wiley & Sons Ltd

Received: 19 February 2019

|

Revised: 2 December 2019

|

Accepted: 19 December 2019

DOI: 10.1111/twec.12914

ORIGINAL ARTICLE

Foreign assistance and emigration: Accounting for

the role of non-transferred aid

MauroLanati1

|

RainerThiele2

1Migration Policy Centre (EUI), Robert Schuman Centre of Advanced Studies (EUI), Florence, Italy

2Kiel Institute for the World Economy, Kiel University, Kiel, Germany

Funding information

Stiftung Mercator, Grant/Award Number: PN 14-297

KEYWORDS

foreign aid, migration

1

|

INTRODUCTION

At least since the large movements of refugees and other migrants to the EU in 2015, many policy-

makers see the scaling-up of foreign aid as a key instrument to stem migrant inflows. The underlying

argument is that long-term development assistance can help address the root causes of migration

through the creation of earning opportunities, quality education and better public services, thereby

giving people an incentive to stay at home.

When it comes to assessing the impact of aid on emigration in order to verify whether policy-

makers’ claims are justified, a main challenge is to account for the heterogeneity of the relationship

between aid and migration that is to be expected because aid can serve many different purposes rang-

ing from support of civil society to the establishment of large-scale infrastructure. Recent empirical

research has provided some evidence in this regard by disaggregating foreign aid along various lines.

For example, Lanati and Thiele (2018b) hypothesise that, broadly speaking, foreign aid can either

raise incomes or improve public service provision within recipient countries. Using the distinction

suggested byClemens, Radelet, Bhavnani, and Bazzi (2012) between early-impact aid, which may

generate income growth in the short to medium term, and late-impact aid, which immediately affects

non-monetary dimensions of well-being but may lead to higher incomes only in the very long run,

Lanati and Thiele find that a rise in late-impact aid is associated with falling emigration rates. Gamso

and Yuldashev (2018a) compare the effects of rural and urban development aid on international mi-

gration. They find that countries that receive larger amounts of rural development aid have lower

emigration rates, which is mainly attributed to additional investments in agricultural sector capacity

building. By contrast, no significant link could be detected between aid to urban areas and migration.

Two further studies by Lanati and Thiele (2018a) and Gamso and Yuldashev (2018b) detect

differential impacts on migration across sectoral aid categories. Lanati and Thiele (2018a) investi-

gate the relationships between emigration rates and inflows of aid for social infrastructure, physical

1952

|

LANATI ANd THIELE

infrastructure and production sectors. All three aid categories have a statistically significant negative

effect on emigration rates, but only the impact of aid to the social sector is relevant in quantitative

terms. According to the estimates provided by Gamso and Yuldashev (2018b), emigration rates are

lower where governance aid is higher, whereas aid intended to promote economic or social develop-

ment does not affect emigration rates.

A common pattern that emerges from all these studies is that any major impacts of aid on migration

tend to run through improved public services that provide incentives for people to stay in their home

countries. There is no indication of empirically relevant income-enhancing effects of foreign aid that

might give rise to increased emigration by allowing would-be migrants to incur the costs of moving

to destination countries.

In this paper, we focus on another important dimension of heterogeneity that has so far been ne-

glected in the literature, namely whether or not the delivery of foreign aid is actually associated with

a transfer of resources to the recipient country. We depart from Qian’s (2015) observation that a sub-

stantial share of the foreign aid reported by OECD Development Assistance Committee (DAC) donors

is spent within their own borders. This so-called non-transferred aid is usually not considered a sepa-

rate analytical category, even though it can be expected to differ fundamentally from transferred aid

as concerns its impact on outcome variables such as emigration rates.1 In line with the discussion

above, aid that involves a transfer of resources to the recipient country may affect migration decisions

through raising individual incomes and/or improving the quality of public services. By contrast, in the

absence of a resource transfer would-be migrants do not experience such direct tangible benefits,

which implies that non-transferred aid is unlikely to be effective in tackling the root causes of migra-

tion. There may still be some indirect impacts of non-transferred aid on migration, for example if ex-

perts paid by donor governments provide useful advice that helps improve institutional quality in

recipient countries.

Our contribution to the literature is twofold. First, we update and extend Qian’s (2015) previous

account of the non-transferred aid delivered by OECD donors. In particular, we discuss in some

detail how spending on refugees within donor countries evolved after the recent refugee crisis, and

why such spending is regarded as part of international development cooperation. It turns out that

the surge in foreign aid since 2015 has largely been driven by steeply increasing in-donor refugee

costs. Second, we analyse whether and to what extent separating transferred aid from non-trans-

ferred aid qualifies previous estimates of the relationship between aid and migration. To the best of

our knowledge, we are the first to make this distinction in the empirical aid effectiveness literature.

Our regression results suggest that transferred aid has a markedly stronger impact on migration

than total aid including the non-transferred component. Future research will have to show whether

this carries over to other parts of the aid effectiveness debate such as the link between foreign aid

and economic growth. As expected, non-transferred aid itself does not appear to affect migrant

flows.

The remainder of the paper is structured as follows. In section 2, we provide an overview

of the composition and quantitative importance of non-transferred aid in OECD Development

Assistance Committee (DAC) donor countries, putting a focus on the 10 donors with the larg-

est aid disbursements. Section 3 first describes the econometric approach as well as the data

employed in the empirical analysis, and then presents and discusses the regression results.

Section4 concludes.

1 Note that the inaccuracies that may result from simply aggregating transferred and non-transferred aid are not restricted to

the estimates of the aid–migration relationship presented in this paper. A similar reasoning applies to large parts of the aid

effectiveness literature including the long-standing controversy on whether foreign aid raises economic growth.