Thepoliticaleconomyofrefugeemigrationandforeignaid PDF

The Political Economy of Refugee
Migration and Foreign Aid

This page intentionally left blank
The Political Economy of
Refugee Migration and
Foreign Aid
Mathias Czaika
© Mathias Czaika 2009
All rights reserved. No reproduction, copy or transmission of this
publication may be made without written permission.
No portion of this publication may be reproduced, copied or transmitted
save with written permission or in accordance with the provisions of the
Copyright, Designs and Patents Act 1988, or under the terms of any licence
permitting limited copying issued by the Copyright Licensing Agency,
Saffron House, 6-10 Kirby Street, London EC1N 8TS.
Any person who does any unauthorized act in relation to this publication
may be liable to criminal prosecution and civil claims for damages.
The author has asserted his right to be identified as the author of this
work in accordance with the Copyright, Designs and Patents Act 1988.
First published 2009 by
PALGRAVE MACMILLAN
Palgrave Macmillan in the UK is an imprint of Macmillan Publishers Limited,
registered in England, company number 785998, of Houndmills, Basingstoke,
Hampshire RG21 6XS.
Palgrave Macmillan in the US is a division of St Martin’s Press LLC,
175 Fifth Avenue, New York, NY 10010.
Palgrave Macmillan is the global academic imprint of the above companies
and has companies and representatives throughout the world.
Palgrave® and Macmillan® are registered trademarks in the United States,
the United Kingdom, Europe and other countries.
ISBN-13: 978–0–230–57688–9 hardback
ISBN-10: 0–230–57688–5 hardback
This book is printed on paper suitable for recycling and made from fully
managed and sustained forest sources. Logging, pulping and manufacturing
processes are expected to conform to the environmental regulations of the
country of origin.
A catalogue record for this book is available from the British Library.
A catalog record for this book is available from the Library of Congress.
10 9 8 7 6 5 4 3 2 1
18 17 16 15 14 13 12 11 10 09
Printed and bound in Great Britain by
CPI Antony Rowe, Chippenham and Eastbourne
This work is dedicated to those who currently live as refugees
and asylum seekers in desperate situations around the globe,
whether living in a camp, a retention centre or somewhere
else, but certainly in insecurity and fear.
May your voices be heard and your prayers answered.
This page intentionally left blank
Contents
List of Tables ix
List of Figures xi
Abbreviations xii
Acknowledgements xiv
1 Introduction 1
1.1 Research objective 1
1.2 Research outline 5
2 Forced Migration Patterns and Politics 10

2.1 Introduction 10
2.2 Migration characteristics 11
2.3 International actors for refugee protection 16
2.4 Refugee Politics 19
3 International Refugee Burdens 25

3.1 Introduction 25
3.2 Assessing the refugee burden to host countries 26
3.3 Methodology of a refugee burden index 30
3.4 Concluding remarks 40
4 Determinants of Forced Displacement:

The Case of Aceh, Indonesia 43
4.1 Introduction 44
4.2 Conflict history and displacement in Aceh, Indonesia 46
4.3 Theoretical background of displacements 50
4.4 Empirical analysis of displacement in Aceh 52
5 The Political Economy of Refugee Migration 67

5.1 Introduction 67
5.2 The Model 70
5.3 Counter-asylum policies 77
vii
viii Contents
6 Asylum Cooperation among Asymmetric Countries 88

6.1 Introduction 88
6.2 Literature review 90
6.3 The analytical framework 94
6.4 Asylum cooperation in the European Union 104
7 Refugee Movements and Aid Politics 114

7.1 Introduction 114
7.2 Aid and refugee migration policies 115
7.3 Aid allocation patterns and statistics 118
7.4 UN diplomacy and aid allocation 123
8 Aid Allocation and Asylum Migration 131
8.2 Aid allocation literature review 132
8.3 The model 135
8.4 Comparative statics 137
8.5 Empirical evidence 138
9 Refugee Movements and Aid Responsiveness 146

9.2 Refugee movements: some policy options 148
9.3 Empirical analysis 151
10 Conclusion 164
10.1 Summary 164
10.2 Policy implications 166
Appendix 170
Notes 181
Bibliography 192
Index 203
Tables
2.1 International migration stocks (2005) and

flows (2000–05), in 1000 12
3.1 Correlation matrix of the RCI components 33
3.2 Top 20 burdened countries in 2003 37
3.3 Top 5 intra-regionally most burdened countries
in 2003 38
3.4 Inter-regional Refugee Burden Index in 2003 40
3.5 Inter-regional Refugee Burden Matrix
(gross/net stocks) in 2003 41
4.1 Conflict and population change in
the Aceh sample, 1999–2002 54
4.2 Probit estimation: conflict incidence 58
4.3 Determinants of change in village population 60
4.4 Quantile regressions of population change 61
4.5 Determinants of population change by
conflict occurrence 62
5.1 Simulation: aid transfers and refugee migration 85
6.1 Asylum cooperation: equal population quota, EU-27 108
6.2 Asylum cooperation: equal economic burdens, EU-27 110
7.1 Sector composition of total ODA grants, 3-year average 122
7.2 Bilateral and multilateral aid allocation 127
7.3 Bilateral donor-specific coefficients 128
8.1 Bivariate correlates of German ODA and
explanatory variables 141
8.2 German aid allocation 142
9.1 Aid allocation determinants, 1992–2003, 18 donors,
148 recipients 157
9.2 Categorization of aid donors 160
A.1 Specifications of the Refugee Capacity Index 170
A.2 Descriptive statistics, Aceh sample 171
A.3 Derivations of the refugee migration model 173
A.4 Descriptive statistics, EU-27 174
A.5 Descriptive statistics, UN speeches sample 175
A.6 Data description and sources, UN speeches sample 175
ix
x Tables
A.7 Descriptive statistics, German aid allocation sample 176

A.8 Data description and sources, German aid
allocation sample 177
A.9 Descriptive statistics, aid panel dataset 178
A.10 Correlation matrix, aid panel dataset 179
A.11 Data definitions and sources, aid panel dataset 180
Figures
1.1 Global distribution of refugees, in millions 3

4.1 Aceh districts 49
4.2 Perceived value of displacement 51
5.1 Cross-border migration duration of
persecuted individuals 73
5.2 The refugee migration decision: who, where,
and how long 75
5.3 Liberalization of self-reliance in the
first asylum country 79
5.4 Declining asylum recognition rate in the
Western asylum country 80
5.5 Increasing income level in the country of origin 84
5.6 Self-financing proactive asylum policy 87
6.1 Asylum policy among countries with
asymmetric spillovers 98
6.2 Asylum policy outcomes among countries with
asymmetric costs 98
7.1 Donor motivations of ODA provision,
EU public poll, 2007 118
7.2 Net disbursement of ODA, OECD/DAC countries 119
7.3 Top bilateral donor states in 2006 120
7.4 Regional distribution of ODA, 1996 and 2006 121
8.1 Asylum applications in EU countries (in 1000) 133
xi
Abbreviations
BIS Bank for International Settlements

BMZ Bundesministerium für Wirtschaftliche Zusammenarbeit
und Entwicklung (German Ministry for Economic
Development and Cooperation)
BPS Badan Pusat Statistik (Statistics Indonesia)
CEAS Common European Asylum System
CIA Central Intelligence Agency
CRS Creditor Reporting System
DAC Development Assistance Committee
DPKO Department for Peace-Keeping Operations
EC European Commission
ECHO European Community Humanitarian aid Office
ELRFI Ethno-Linguistic-Religious Fractionalization Index
ERC Emergency Relief Coordinator
ERF European Refugee Fund
EU European Union
FDI Foreign Direct Investment
GAM Gerakan Aceh Merdeka (Acehnese Freedom Movement)
GDPI Gross Domestic Product Index
HDI Human Development Index
IASC Interagency Standing Committee
ICARA International Conference on Assistance to Refugees in
Africa
ICRC International Committee of the Red Cross
IDP Internally Displaced Person
ILO International Labour Organization
IMF International Monetary Fund
IOM International Organization for Migration
NGO Non-Governmental Organization
OAU Organization for African Unity
OCHA Office for the Coordination of Humanitarian Affairs
ODA Official Development Assistance
OECD Organization of Economic Cooperation and Development
OLS Ordinary Least Squares
xii
Abbreviations xiii
PDI Population Density Index

PFI Political Freedom Index
PODES Potensi Desa (Village Potential Census)
PQLI Physical Quality of Life Index
PSI Political Stability Index
RBI Refugee Burden Index
RBS Refugee Burden-Sharing
RCI Refugee Capacity Index
TNI Tentara Nasional Indonesia (Indonesian military
forces)
TPOC Total Persons of Concern
UN United Nations
UNDP United Nations Development Program
UNHCHR United Nations High Commissioner for Human
Rights
UNHCR United Nations High Commissioner for Refugees
UNICEF United Nations International Children’s Emergency
Fund
UNRWA United Nations Relief and Works Agency for Palestine
Refugees
WB World Bank
WFP World Food Program
Acknowledgements
To whom shall I give more praise and thank for HIS support
than to my heavenly Father! He is and was my solid founda-
tion in all my struggles and times of thinking, writing, and
abandoning. He is and was my friend and encourager, my
shining light in times of joy and doubt. I praise you, Jesus!
I would like to express my sincere gratitude to my supervisor, Prof.

Günther Schulze. His detailed and constructive comments as well as
all his encouragement and understanding have provided a good basis
for the present book.
Furthermore, I am grateful to my (former) colleagues and friends
from the University of Freiburg, Krisztina Kis-Katos, Eva Deuchert,
Antonio Farfán Vallespín, Juliane Fliedner, Grischa Perino, Martin
Keller, Michael Trost, Inga Müller, Amy Mayer and Marcus Marktanner
for all of the discussions, collaboration, assistance, proofreading and
fun. I also appreciated all of the helpful comments of the partici-
pants in the Wednesday Doktorandenseminar at the University of
Freiburg. I am thankful also to Amy McDonough for proofreading
this book. However, I would like to state that any errors in this thesis,
be they imprecise quotations or basic misrepresentations, are unin-
tentional and my responsibility alone.
My warm thanks are due to Giseltraud Herbert for all her kindness
and helpful assistance throughout the past five years.
Special thanks to all of my friends from Students for Christ (SfC)
for their interest, encouragement and prayers.
I owe my loving thanks to my wonderful parents Hartmut and
Lydia, for enabling my university education, and particularly, for
their love and support throughout the years.
Finally, I want to express my special thanks to my beloved wife
Catheline. She helped me to concentrate on completing this thesis
and supported me in many ways, especially by her prayers. Without
her help, encouragement and understanding, this book could not
have been completed.
xiv
1
Introduction
If you don’t help the third world,

[...] then you will have these poor people in your society.
(Paul Rasmussen, Prime Minister of Denmark,
Copenhagen Social Summit 1995)
This book addresses the causes and incidences of refugee movements

and the policy responses of Western asylum and aid-giving coun-
tries. In particular, it is focused on the positive and normative aspects
of asylum migration and its management by Western industrialized
countries. Thereby, asylum policy and aid policy are the main instru-
ments of interest in this work. The book focuses explicitly on the
following research questions: What are the driving factors of forced
migration movements? How can refugee burdens be assessed and
compared across different types of host countries? Who are the actors
in international refugee protection and management, and how do
they act? Is cooperation among asylum countries necessary, and pos-
sible? And finally, how does the phenomenon of international refu-
gee movements, and specifically how do asylum seekers, influence
the aid allocation politics of Western industrialized countries?
The remainder of this introductory chapter describes the back-
ground of these research objectives and also provides an outline of
the subsequent analytical parts.
1.1 Research objective
The relevance of this work is self-evident. In 2005, the United Nations

estimated about 190 million people living in a country in which
1
2 Refugee Migration and Foreign Aid
they were not born (United Nations, 2006b). During the past decades,
large migration movements originated from countries and regions
ranging from those characterized by economic and social despair or
political insecurity to countries that are more stable, secure and
economically developed. Continuously increasing disparities in the
level of human and social security, on the one side, and improving
facilities for communication and transportation, expanding social
networks, as well as the emergence of a commercial trafficking indus-
try, on the other side, have led to this steadily growing number of
foreign-born populations in developed and developing host coun-
tries. This ‘age of migration’ (Castles & Miller, 2003) has caused a
growing concern, mainly in Western industrialized states, which are
major destinations for all types of migrants, and has brought the
topic of international voluntary and involuntary (forced) migration
to the top of national and international agendas.
Generally, it is difficult to distinguish between forced migration
and voluntary economic migration, since it is usually arbitrary to
state that people move voluntarily or by force. Basically, voluntary
migration movements, which are purely economically motivated, do
not largely originate from the least developed regions or countries
and poorest populations, but rather from those that are better
equipped. This phenomenon, known as the ‘migration hump’, is
mainly due to constraints in economic resources, which prevents the
poorest from raising the funds necessary to emigrate. During the
past decades, these economic constraints have become even more
important for emigration decisions since international (or, transcon-
tinental) migration often requires substantial resources for smug-
gling or trafficking agents.
In contrast to pure economically motivated migration, this book
is focused on refugee migration, which is by definition forced and
involuntary. With respect to this group of migrants, fear, insecurity
and individual or mass persecution in times of violent conflicts in
many of the most fragile and economically marginalized countries
of the developing world have triggered increasing numbers of dis-
placed people. For this subpopulation, the determinants that cause
their (involuntary) emigration are comprised of a complex mixture
of political, social and economic factors, of which the individual
psychological perception of ‘threat’ is only one, but most often the
decisive and final factor. Concerning the various push factors at the
Introduction 3
origin site, one can distinguish those rather related to political inse-
curity such as human rights violation, persecution of societal sub-
groups, or violent conflict, and those related to social and economic
conditions such as indigence, unemployment, lack of access to edu-
cation, health or social services, or environmental and natural deg-
radation. Analyzing and disentangling this whole range of driving
forces of large displacements in conflict-ridden situations is one task
of this book. A case study on the conflict-affected province of Aceh,
Indonesia, attempts to identify and to categorize this mixture of
conflict-related and traditional, socioeconomic determinants.
Many persecuted people of this world are not able to migrate at all,
or only a small distance, because of lack of resources. Those who are
able to emigrate mostly move either only internally or to a neighbor-
ing country. Consequently, a number of relatively stable countries in
Asia, the Middle East, South America, Central and East Europe or
Southern Africa have increasingly become preferred destinations for
asylum seekers and refugees. However, although the majority of
forced migrants, that is refugees, asylum seekers, or internally dis-
placed persons (IDPs), are in fact compelled to stay close in their
origin sites, Western industrialized regions such as the EU and
Northern America are also first choice destinations for many forced
migrants from the southern hemisphere (Figure 1.1 and Table 2.1).
20
18
16
14
12 Rest of World
10 Asia
8 Africa
6 Western Europe
4 Northern America
2
0
1980
1982
1984
1986
1988
1990
1992
1994
1996
1998
2000
2002
2004
Figure 1.1 Global distribution of refugees, in millions

Note: Figure is based on data provided by UNHCR (2005).
Particularly during the 1990s, the rise in international refugee

migration has become a major challenge to many host countries,
and as a consequence, Western countries, but also destination coun-
tries in the region of origin, introduced various restrictions intended
to impede and dissuade people from accessing their territory for
the purpose of seeking asylum. Industrial states often acknowledge
the benefits of immigration when it meets the economic require-
ments of the domestic labor market. However, policymakers and the
general public increasingly perceive refugee (asylum) migration as a
means to bypass legitimate immigration controls.
As claiming refugee status allows the claimant to stay in the asy-
lum country, at least for the years while the application is processed,
large-scale asylum procedures generate significant costs for adminis-
tration and provision of social care, public services and legal advice.
This burden of hosting refugees has led to a growing reluctance
in most host societies to accept further inflows, particularly in
Western asylum countries. However, is the often stated perception of
Western asylum countries correct that they carry the bulk of the
global refugee burden? This question is addressed by generating a
more objective measure to make refugee burdens comparable across
countries without having explicit data on the actual political, social
and economic costs of hosting a large refugee population.
The political economy of refugee management in Western states is
mainly driven by general concerns about the apparently negative
impact of asylum migration inflows on unemployment, social wel-
fare systems, security and national identity. Despite various benefits
from immigration that are put forward either by Western firms and
employers with respect to economic gains or by liberal protagonists
for sociocultural and humanistic reasons, there remains nevertheless
considerable resistance in society and in the political discourse to any
liberalization of migration and asylum policies. Since the late 1980s,
when the problem of unwanted immigration and asylum flows began
to emerge, primarily the EU, the US, Canada, and Australia, have
been at the forefront of implementing more restrictive policies to
scale down immigration pressures. A wide spectrum of measures has
been introduced designed to limit the numbers of asylum seekers,
and the costs of hosting them and of processing their claims.1
Based on the experience of the past decade in Western countries, a
more restrictive asylum policy in one country leads to a diversion of
Introduction 5
asylum flows to other countries, and consequently, to a race to

the bottom of asylum standards, and thus, to an increase of irregu-
lar (undocumented) migration flows into Western countries. As a
result, unidimensional restrictive asylum and immigration policy
approaches seem rather inadequate to curtail the trend of rising
South–North asylum migration movements. Since migration flows
are mostly self-perpetuating by network effects, they are also
unavoidable. Deficiencies of restrictive asylum policies require alter-
natives to these reactive asylum policy approaches. For instance, it is
often argued that promotion of economic, political and social devel-
opment in poor refugee-sending countries could curb emigration.
In particular, development aid and other financial transfers are dis-
cussed as alternative instruments to reduce refugee migration rather
than restrictive asylum systems and stricter border controls. In this
book, we distinguish basically two impact variants of such transfers:
first, financial transfers could compensate for non-private spillovers
that are generated by refugee-hosting countries (that is, internaliza-
tion of cross-border externalities), and thus, produce a more efficient
allocation of refugees. However, under what conditions do countries
actually have incentives to act cooperatively and to join an interna-
tional burden-sharing regime? And second, foreign aid has the
potential to work as an instrument that curbs the socioeconomic
and also political causes for emigration. However, do countries actu-
ally allocate aid for that purpose of migration prevention? If yes, to
what extent are aid policies of Western donor countries influenced
by the refugee movements that increasingly take place in the devel-
oping world? Furthermore, do Western donor countries distinguish
between different types of forced migrants? These questions are dis-
cussed in the subsequent chapters.
1.2 Research outline
The remainder of the book proceeds as follows.2 Chapters 2 and 7

provide the respective background information for the two major
topics of this book: refugee migration movements and the allocation
of foreign aid. While Chapter 2 focuses on the characteristics of
migration and its underlying causes as well as on the actors and
their asylum policy approaches, Chapter 7 primarily explores the
motivations and patterns of foreign aid allocation policies.
In Chapter 3, the incidence and distribution of the current global

refugee population is investigated. The focus is put on the develop-
ment of an appropriate measure (that is, an index) that makes
refugee-hosting burdens more comparable across countries. We elab-
orate particularly on the question of how large populations of refu-
gees affect host countries that are highly heterogeneous with respect
to their political, economic, and social ‘performance’. These three
categories are often the most relevant areas when host countries are
confronted with refugee inflows. Since burdens in these areas are not
directly measurable, for example, in terms of monetary costs, this
analysis corrects for cost heterogeneity by indexing the political,
economic and social performance of the host country when provid-
ing asylum to a refugee population.3 This methodology generates
implicit performance-based refugee burdens that are comparable
across host countries. As a result, a Refugee Burden Index displays a
more appropriate picture of the actual extent of both global and
intra-regional distribution of refugee burdens, and thus, the unequal
distribution across asylum countries.
Chapter 4 elaborates on the determinants of forced displacement
by studying the effects of the protracted conflict in the Indonesian
province of Aceh on population movements across villages. As a
starting point, this chapter explores the correlates of this long-
lasting armed conflict. We then identify the determinants of dis-
placement based on various conflict-related and socioeconomic push
and pull factors. The study, based on village-level data, disentangles
the driving forces behind changes in village populations during the
most severe years of the Acehnese civil war (between 1999 and 2003).
It is shown that the driving forces of displacement are not only the
immediate conflict factors like violence and fear of death, but that a
significant part of the population also left their homes because of
other political, economic, cultural and religious factors. Thus, tra-
ditional migration causes are still relevant factors in explaining large
displacements in times of conflict and war. This case study partly
unravels the complexity of forced migration determinants in a con-
flict environment and contributes empirical evidence to the scarce
quantitative literature on conflict-induced migration movements.
Chapter 5 models the individual migration decision of persecuted
people. It analyzes the driving forces of the magnitude, composition
and duration of refugee movements that are originally caused by
Introduction 7
conflict and persecution. We presume that the migration decision is

based on intertemporal optimization by each persecuted individual.
It is shown that ceteris paribus asylum seeking towards Western coun-
tries is often a phenomenon of the rather less persecuted people.
Furthermore, since all host countries are interested in a sustainable
reduction of their potential or actual asylum burdens, Western coun-
tries and conflict-neighboring host countries in the region of origin
are likely to end up in a ‘race to the bottom’ of restrictive asylum
policies. As an alternative, this analysis shows that under some cir-
cumstances proactive refugee-related aid transfers are an effective
instrument to relieve Western countries of their asylum pressure.
Whether financial transfers among asylum countries are an appro-
priate instrument for instigating a cooperative international refugee
regime is the task of Chapter 6. Two alternative international refugee
regimes with and without financial cooperation schemes among
heterogeneous asylum countries are analyzed. It becomes evident
that refugee protection levels are generally suboptimal because of
the absence of an institutionalized and efficient refugee cooperation
regime. Asylum provision as a ‘public good’ causes non-cooperative
deflection and free-riding of (potential) refugee-hosting states.
Furthermore, the analysis indicates that, first, asylum cooperation is
rather attainable among less heterogeneous countries, and second,
financial transfers are, under some circumstances, able to enforce
and stabilize a cooperative international asylum regime. These impli-
cations of the theory are applied on the context of the enlarged
European Union. At least for the (currently) 27 member states of the
EU, this exercise indicates that asylum cooperation is attainable
under the present voting system, and even more so under the pro-
posed voting principle outlined in the EU reform treaty.
Based on the policy implications of the previous chapters,
Chapters 8 and 9 analyze whether the normative claim of a proac-
tive refugee-related aid policy is reflected in actual bilateral aid-
giving patterns of Western aid donors. The objective is to
disentangle bilateral aid patterns and to identify the most relevant
aid allocation determinants of Western donors. Particular emphasis
is put on the influence of refugee (asylum) migration on aid trans-
fers towards both refugee-sending and conflict-neighboring first
asylum countries. In these empirical studies, only aid flows for the
period after the Cold War are addressed. This coincides temporally
with the period of increasing asylum migration movements towards

Western donor countries.
In Chapter 8, the focus is put on the means that seems most effi-
cient in triggering donor countries to respond sustainably to urgen-
cies: asylum migration to Western donor countries. The objective of
the analysis is to identify the influence of asylum seekers and other
foreign-born immigrants on the bilateral allocation of development
aid funds. In contrast to most other empirical studies that focus pri-
marily on domestic economic and political variables relating to the
recipient countries, this analysis is rather interested in the role of
domestic politics in the donor country, without ignoring relevant
aspects of the recipient countries. In the literature of political econ-
omy, the influence of interest groups has received wide attention.
In this study, the relevant interest groups are the permanently set-
tled foreign-born immigrants as well as the asylum applicants that
newly enter the Western donor country. Both groups seem to be
interested in increasing aid transfers in favor of their home country.
This hypothesis is studied in the case of Germany, since it is both
one of the largest donors of bilateral aid and a main haven for asylum
seekers and other immigrants with a significant stock of foreign-
born population and a continuous inflow of asylum seekers. The
analysis shows that bilateral asylum burdens have a significant effect
on the amount of bilateral aid transfers, and particularly, when the
respective recipient countries have – besides the influence of asylum
seekers – a significant stock of permanent expatriates living in the
donor country.
Chapter 9 continues this discussion and analyzes the impact that
different characteristics of refugee movements have on the alloca-
tion decisions of bilateral donors of (long-term) development and
(short-term) emergency aid. The evidence indicates that short-term
aid is indeed used in severe situations of violent conflicts and natural
disasters, and therefore is predominantly allocated to the home
countries of refugees. Furthermore, donor states respond to all types
of forced migration (that is internally displaced persons, cross-border
refugees or international migrants seeking for asylum in donor states)
by increasing their humanitarian aid efforts. With respect to the
allocation of long-term development assistance, however, donor
states target their aid funds primarily to the countries of origin of
refugees, particularly in the case where these refugees have become
Introduction 9
asylum applicants in the donor country itself. Further disaggregation

indicates that this effect is stronger for the donor countries that are
main havens for asylum seekers.
Finally, Chapter 10 summarizes the main insights of these differ-
ent studies and discusses some further policy implications. Although
the complexity of the conflict–refugee–aid linkages permits only
modest statements, it becomes clear that there is an unambiguous
need for alternatives given the increasing trend of restrictive asylum
and immigration policies. More proactive international root-cause
strategies for the countries in trouble and burden-sharing initiatives
for the overflowing refugee-hosting countries are unavoidable to sus-
tainably address refugee migration movements.
2
Forced Migration
Patterns and Politics
Abstract
This chapter provides a general overview on migration characteris-

tics, such as different types and causes of migration. Furthermore,
the main actors in migration and refugee politics are introduced
and the most relevant alternatives in international and national ref-
ugee policy are briefly described.
2.1 Introduction
Migration is a global phenomenon. There is no country in the world

that does not register either transnational or internal migration
movements. The ever increasing integration of world markets works
as a major driving force of economically motivated migration move-
ments. Yet, individuals do not only move for economic reasons.
Refugees (or, forced migrants) constitute a substantial part of the
worldwide migration stocks and flows. By definition, they do not
(primarily) leave their home village or country in order to find better
economic living conditions for themselves and their families, but
because they seek protection from suppression, persecution and vio-
lation of their human rights. In general, states try to control migra-
tion by various measures that restrict immigration.4 However, the
intensity of efforts to limit immigration varies significantly across
states; some countries have more liberal immigration laws than oth-
ers. In principle, all national states try to manage migration flows; no
country has completely open borders. Nevertheless, international
10
Forced Migration Patterns and Politics 11
migration continues to increase, and policymakers in many states

suggest that more international cooperation is needed to arrest the
trend of continuously increasing numbers of all types of (interna-
tional) migrants. This is particularly critical in view of the proceed-
ing integration of world markets, increasing wealth and income
inequalities, as well as political insecurity in many states of the
developing world. This chapter begins with an overview of the
characteristics, the pattern, and the causes of the migration phenom-
enon. Next, the implications for national and international migra-
tion policies are discussed, though primarily focusing on the
subgroup of forced migrants, that is, refugees, asylum seekers and
internally displaced persons (IDPs).
2.2 Migration characteristics
The global migration pattern is difficult to describe and can only

roughly be estimated due to a lack of quality data on global migra-
tion. Even in many Western industrialized countries, the quality of
migration-related data is insufficient and fragmentary (Crisp, 1999).
For instance, there is a lack of individual data available on the age,
origin and educational background of migrants, as well as informa-
tion on the (intergenerational) characteristics of (re)integration pro-
cesses. Data collected on global migration stocks and flows are still
sketchy; there are deficits in the national statistics, in addition to dif-
ferent definitions and collection methods across countries. Therefore,
clear-cut statements on the statistics of the global migration pattern
must be interpreted with caution (Crisp, 1999). This data deficiency
is also problematic with regard to the formulation of appropriate pol-
icy strategies, since without reliable data, the planning and evalua-
tion of political programs and initiatives can hardly be accurate.
Nevertheless, keeping this fact in mind with regard to international
migration numbers, it can be stated that the migration pattern has
changed strongly since the end of the Cold War.
By the end of 2005, almost 200 million people lived temporarily
or permanently in countries where they were not born (United
Nations, 2006b). This corresponds to around three percent of the
world population (Table 2.1). Between 2000 and 2005, industrialized
countries have had an average annual migration net inflow of
approximately 2.6 million people stemming from the developing
Table 2.1 International migration stocks (2005) and flows (2000–05), in

1000
Population Migrant stock Net annual inflow

(per capita) (per capita)
Developed countries 1,211,265 115,397 2,622

(9.5%) (2.2%)
Developing countries 5,253,484 75,237 –2,622
(1.4%) (–0.5%)
Source: United Nations (2006b): Population Division.
world, which is around 0.2 percent of the developed countries’

combined population (United Nations, 2006a). In Western industri-
alized countries, these migration numbers are often perceived as
much higher, although they rose only moderately compared to pre-
vious decades.
Besides modestly increasing migration movements, migration
forms have become more varied and more complex. It is difficult for
governments and international organizations to distinguish between
the different migration forms, for example, to distinguish between
economic migrants and refugees, or between temporary, permanent
and repeated immigration, as well as legal and illegal (irregular)
immigration.
2.2.1 Migration types

Disaggregating global migration movements indicates roughly
three major bases of the total stock. First, subsequent immigration
(family reunification) of relatives of former immigrants (or refugees)
living now permanently in the host country constitutes the major
part of the whole migration stock (United Nations, 2005a). In most
Western countries, subsequent immigration of family members is
limited to spouses and children of migrants with permanent/
unlimited residence rights.5 Some Western countries generously per-
mit the influx of relatives that do not belong to the actual core
family. It is expected that subsequent immigration of relatives can
facilitate integration in the host country. Other countries, particu-
larly in East and Southeast Asia, are stricter and refuse immigration
of family members in principle.
The second most important migration form is labor migration,

that is, entry for the purpose of a limited or durable practice of an
economic activity (United Nations, 2005a). This group consists of,
for example, unskilled seasonal workers in the agriculture, produc-
tion or service industries (tourism). In many countries certain indus-
tries would be no longer competitive without such economic
migrants. However, labor migrants may also be highly qualified
employees such as technicians, scientists or managers. In most coun-
tries, but not in the classical immigration countries like the US,
Canada, Australia or New Zealand, labor migrants receive only a lim-
ited residence and work permit that is often restricted to a certain
employer–employee relationship. In recent years, some European
states have also begun to provide durable job market-related immi-
gration rights in order to fill structural gaps in their labor force.6
The third category of migrants, which is of particular interest to
this book, is the group of asylum seekers and refugees. Together, they
constitute around seven percent of global migration stock (UNFPA,
2006). The number of refugees worldwide decreased from 16.3 million
people in 1993 to approximately 12.7 million in 2005, of which
8.4 million were assisted by United Nations High Commissioner for
Refugees (UNHCR) and an additional 4.3 million were under the
responsibility of the United Nations Relief and Works Agency for
Palestine Refugees (UNRWA) (UNHCR, 2003, 2006c).7 However, the
total number of persons who left their country or region of origin
because of fear of persecution, political suppression or violent con-
flicts was even larger. In 2006, UNHCR was responsible for more
than 32 million persons – in addition to the already specified refu-
gees in the narrow sense, there were 596,000 asylum seekers, and
734,000 refugees returned to their home countries. Additionally,
internally displaced persons (IDPs) constituted another 12.8 million
persons by the end of 2006 (UNHCR, 2006c).8
Since the mid-1990s, the number of refugees has decreased and the
number of internally displaced persons increased. This is primarily
because of changing flight causes: on the one hand, there were more
internal civil conflicts with forced displacements as part of warfare,
and on the other hand, asylum policies of most host countries have
become more restrictive, which made it more difficult for refugees to
access safe and developed countries for the purpose of asylum.
2.2.2 Migration causes

The above outlined rather modest increase in total migration num-
bers established two stylized facts of migration research. First, despite
often very desperate living conditions, most individuals or groups
remain in their country of origin and do not emigrate or flee their
poor situation (Fischer et al., 1998; Straubhaar, 2000). And second,
international migrants and refugees do not usually belong to the
poorest group of people in their country of origin. Instead, they are
rather the more wealthy ones, which have the financial resources
and/or family or other social networks for immigrating into the des-
tination country (Bhagwati et al., 1984).
Migration decisions are complex because they often have far-
reaching consequences. Basically, the individuals decide upon leav-
ing their familiar living environment, that is their families and
neighborhood, for an unfamiliar situation abroad. However, the
migration-determining factors are generally more comprehensive.
Traditional migration theory differentiates between factors that take
place on various socioeconomic and political levels. These are usu-
ally separated into push factors, which are effective in the country or
region of origin, and pull factors, which proceed from the potential
destination. Since these factors are most often mutually interdepen-
dent, they can hardly be separated. In general, the core causes of
migration movements are economic, political, sociocultural, demo-
graphic or ecological, or a mixture of these.
Economic factors are certainly the most important reason for emi-
gration. Demographic, ecological, ethnical or political migration
motives frequently become effective only in interaction with eco-
nomic motives. Even refugees can have economic reasons for escap-
ing their origins, for example, if their economic bases have been
destroyed by violent conflicts (see Chapter 4). Mixed motivations
make it difficult to differentiate between economic and non-
economic migration. Refugees are often presumed to give political
reasons for their subjacent economic motives (Neumayer, 2005b).
Thus, this so-called ‘asylum-migration nexus’ reflects the fact that
forced (that is involuntary) migrants are often hard to distinguish
from those with pure economic motives.
Besides economic reasons, political migration factors are often as
important as economic factors. International refugee law defines the
obligation for protection of transnational migrants according to the
distinction of political and non-political migration causes (Hathaway,

1990). The strongest political pressure is exerted on potential host
countries when migrants originate from countries in violent conflict
and war. However, other aspects of persecution and human rights
violations are added to the various definitions of a refugee.9 In most
cases conflict and violence lead not only to transnational migrations,
but also to mass displacements within the country. These internally
displaced persons (IDPs) are often in a worse situation than cross-
border refugees, because international refugee laws are not valid for
their protection. Loescher & Milner (2005) state that the political
inaction of international actors in terms of their lack of interest in
intervening in situations of armed conflict and human rights viola-
tions and in assisting host countries are the major causes for the
increasing number and durations of protracted refugee situations
around the developing world.
Social, cultural and religious migration factors are closely related
to political push factors. Emigration can be forced if, for example,
social, cultural, or economic practices of minorities are rejected by
significant other groups within a society and the government can-
not alleviate these tensions. On the pull side, similar cultural and
ethnical backgrounds, such as common language between the origin
and destination country, facilitate migration movements substan-
tially. Networks among ethnical communities in the host country
offer often crucial support to new immigrants, for example, by pro-
viding accommodation or work after the entry (for example, Massey
et al., 1998). Obviously, these effects hold for both legal and irregular
entry and stay.
Demographic development is often mentioned as another
migration-inducing factor. However, demographic pressure becomes
effective only if certain economic, social, or ecological conditions
are also present. Particularly for developing countries, population
growth can induce both internal and transnational migration
movements depending on its rate and regional distribution. For most
industrialized countries, however, pull effects are more relevant
because of decreasing fertility rates and a rising life expectancy.
Aging societies with contracting populations in the long-run suffer
from a diminishing labor force. Therefore, many industrialized
countries discuss possibilities to abate these structural gaps by tem-
poral or permanent (high-skilled) immigrants (Borjas, 1995).
Environmental factors have increasingly become a cause for emi-

gration. Obviously, this applies particularly to natural disasters such
as floods, earthquakes, droughts or hurricanes. Mass movements
induced by such factors are usually temporally limited and the (envi-
ronmental) refugees normally remain in the region or in the coun-
try. However, other human caused environmental damages are often
even more serious in their consequences. These are, for example, the
long-term loss of field and pasture land because of excessive land use
and the misuse of water resources, as well as the amassment of seri-
ous floods caused by ecological degradation of the soils. Other
environment-related migration factors can be environmental degra-
dations by military activities, wide environmental pollution, expan-
sion of industrial or agricultural manufacturing areas, or resources
shortages (for example, drinking water).10
2.3 International actors for refugee protection
2.3.1 UNHCR and its historical background

National states and international organizations arrange their res-
ponsibilities and competencies for migration movements according
to whether these take place voluntarily or by force – that is whether
they are migrants, or refugees according to the UN Refugee
Convention of 1951/1967. The Convention does not obligate the sig-
natory states to grant individual asylum to a refugee but only not to
expel or to reject entry into a country, in which his/her life or liberty
would be endangered (‘non-refoulement principle’).
In 1949, the Office of the United Nations High Commissioner for
Refugees (UNHCR) was established. The mandate of the organization
is to provide adherence to the principles of refugee protection. It pro-
vides long-term assistance in terms of refugee protection and post-
conflict repatriation and reintegration. It is the most important
human rights and advocacy organization for refugees and, more
recently, also for internally displaced persons (Loescher, 2002).
Within the UN system, UNHCR plays the most important role in ref-
ugee protection matters. The mandate of UNHCR, the UN Convention
on the Status of Refugees, was adopted in July 1951 (UNHCR, 2000). At
that time, the Convention was only applied to individuals who
became refugees before 1950. The signing states were given the
possibility to limit their obligations to only European refugees, while

internally displaced persons were generally excluded (UNHCR,
2000). The spatial and temporal restrictions of the 1951 Convention
were only waived with the ratification of the supplementary protocol
of 1967, since during the 1960s and 1970s UNHCR expanded its
operations to extra-European refugee situations in many regions
of the world. As a consequence, UNHCR became the undisputed
international authority in matters of refugee protection, and began
advising governments in asylum-related legal questions (UNHCR,
2000).
In the 1980s, the situation changed again as refugees became
involved in the deputy wars in Asia, Africa and Latin America. Some
conflict parties recruited refugees as fighters and used refugee
camps as military replenishment or retreat areas. At the same time,
the maintenance of refugee camps took up substantial resources and
organizational capacities, as many refugee situations became pro-
tracted and showed no prospect of imminent solution. Furthermore,
increasing numbers of refugees sought asylum in Western states.
As a consequence of this and the fact that the Western states
increasingly considered the work of UNHCR to be unsuitable for the
solution of refugee crises, the Western governments endeavored
to solve refugee and asylum problems themselves on bi- and multi-
lateral levels, further diminishing the role of UNHCR (UNHCR,
2000).
Since the end of the Cold War, Western governments’ perception
of refugees has changed. Refugees are no longer regarded as a strate-
gic instrument in the struggle between the two geo-political blocks,
but have been increasingly perceived as a burden. The large refugee
movements of the 1990s fuelled the fear of political destabilization
and became a major issue of international and regional security
aspects.11 Starting from 1991, UNHCR intervened in internal affairs
of states affected by refugee disasters, in a way that would have been
inconceivable during the Cold War.12 UNHCR expanded its assis-
tance to ever more conflict-affected populations, in particular to
local populations and internally displaced persons. During the
1990s, the number of persons for whom UNHCR were responsible
had almost doubled (UNHCR, 2000). This made UNHCR the domi-
nant actor in forced migration matters, not only in the developing
world but also in collaboration with industrialized states.
2.3.2 Other international organizations

Beside UNHCR, there are several other UN special organizations
and other international organizations working in the field of refugee
issues.13 For instance, since 1997, the Office for the Coordination of
Humanitarian Affairs (OCHA) has come under the direction of the
UN Under-Secretary for Humanitarian Affairs, who is also the
Emergency Relief Coordinator (ERC) and the chairman of all impor-
tant international humanitarian organizations of the comprehensive
Interagency Standing Committee (IASC). In 1993, the office of the
United Nations High Commissioner for Human Rights (UNHCHR)
was established. The task of the High Commissioner is to supervise
and promote the adherence of human rights, which also includes
issues of forced displacement. The United Nations Development
Program (UNDP) steers and coordinates development projects
throughout the developing world and is also active in a number of
refugee disasters, contributing to the reconstruction of local infra-
structure, the promotion of economic development and employment
opportunities, fighting poverty, and thus, also to the long-term solu-
tion of refugee crises. Since 1992, the mandate of UNICEF also
includes the assistance of displaced children and women, who belong
to the most vulnerable group in many refugee situations. Apart from
establishing initiatives for the improvement of legal protection for
children and women, UNICEF is also an operational actor in refugee
situations. The task of the World Food Program (WFP) is to supply
food and to guarantee basic nourishment, and furthermore, to
improve the capability of refugees for self-reliance, for example, by
financing food-for-work programs. The Department for Peace-
Keeping Operations (DPKO) is responsible within the UN secretariat
for the planning and execution of different types of peace missions.
In recent years, given the growing complexity of refugee situations
and the larger range of auxiliary operations, cooperation and coordi-
nation of humanitarian assistance among relief organizations and
military actors has become much more important (Angenendt,
2002).
Beside these UN organizations, there are many other international
organizations active in the refugee assistance. For instance, the
International Organization for Migration (IOM), which carries out
assistance and repatriation of refugees and asylum seekers on behalf
of governments, or the International Committee of the Red Cross
(ICRC), which ensures humanitarian protection of the civilian pop-

ulation in armed conflicts. Other important actors are also regional
organizations like the European Union, which has become one of
the most important donors of international humanitarian assistance
in recent years through its office for humanitarian assistance (ECHO),
or the Organization for African Unity (OAU), which has contributed
to the political resolution of some refugee crises in sub-Saharan
Africa. Finally, there are many international and national non-
governmental organizations (NGOs), which also carry out important
humanitarian assistance and relief for refugees.
2.3.3 National states

The role of national states as actors in international refugee and asy-
lum issues has altered, although they retain their role as important
and often essential actors. States still are the key actors in asylum
migration and in tackling the root causes of migration for durable
solutions to refugee situations. National states are essential actors in
efforts to protect the human rights of involuntarily displaced popu-
lations (Mertus, 1998). However, the sovereignty of the nation state
when it comes to entry of their own territory and the control of their
borders is a matter of consensus in the international community of
national states. Several multilateral agreements and conventions on
human rights basically urge nation states to grant asylum to forced
migrants for humanitarian reasons, but discretionary policies of
national governments in asylum and immigration matters are gener-
ally accepted (Sassen, 1998). For instance, during past periods, the US
government considered refugee and asylum policy too important to
let the United Nations control it, and they sought to limit the scope
and independence of UN agencies. In this respect peace missions
and refugee interventions in specific regions are much more gener-
ously funded and supported by the US and other Western states
when their respective geopolitical interests are significantly affected
(Loescher, 2002). Thus, the focus is subsequently primarily on the
role and interests of national states and their bilateral activities.
2.4 Refugee politics
As the international economic integration of markets proceeds,

incentives and opportunities for different types of transnational
migration movements will continue to increase, and individual

cross-border mobility is expected to rise for all regions of the world.14
However, in contrast to the relatively unconstrained mobility of
goods and markets, international mobility of people is still tightly
constrained by restrictive immigration and asylum policies that
restrain a large share of potential migration that otherwise would
take place (Chiswick & Hatton, 2002). In the face of increasing
numbers of irregular immigrants, asylum seekers and refugees,
Western countries have become increasingly interested in preventing
large exoduses from developing countries. Policymakers are becom-
ing more and more interested in the political and economic root
causes of these unwanted population movements. Münz & Weiner
(1997) point out three possibilities for Western states to react:
One is to seek better instruments to control, to establish and

enforce stricter border regimes and restrictive migration and refu-
gee laws, to try to repatriate those who enter unlawfully or do not
qualify for asylum. A second is to accept and absorb a certain
number of refugees and economic migrants and to address the
social, economic and political consequences posed by the flows. A
third response is to develop prevention and intervention strate-
gies towards countries that actually or potentially generate mass
emigration. (Münz & Weiner, 1997, p. ix)
Subsequent chapters are primarily concerned with all of these pol-

icy options. International burden-sharing regimes and the potential
for international cooperation in refugee protection are analyzed as
well as the potential or actual interest of Western destination coun-
tries for migration prevention policies, focusing here on financial
and technical aid as major economic instruments. In the next sec-
tion, we introduce these two pillars of a comprehensive migration
and asylum policy; later chapters delve more into these areas of refu-
gee politics.
2.4.1 Burden-sharing and international cooperation

Policymakers of many developed and developing countries acknowl-
edge that the global migration challenges are not brought under con-
trol without international cooperation and intervention. Numerous
states around the world perceive themselves as overburdened and
have substantial problems designing adequate immigration and

integration policies. Western governments are trapped by competing
interests when formulating their immigration and asylum policies.
Enterprises urge for a more liberal migration policy in order to over-
come job market bottlenecks, while the general public is concerned
about additional job market competition, internal security and
national identity. Therefore, Western industrialized countries try to
establish and to unilaterally improve national capacities for control-
ling and managing migration movements. However, they also strive
for common, international regulations for improving regional coop-
eration in migration matters (for example, common immigration
laws as well as a harmonization of asylum policies within the
European Union).
In fact, why is refugee burden-sharing and international coopera-
tion desirable? As Chapter 3 will show, refugee burdens are very
unequally distributed both internationally between the Southern
and Northern hemisphere as well as among developed and develop-
ing states. These unequally distributed refugee burdens are primarily
caused by standard free-riding incentives and the pure self-interest
of national states, as well as the relative attractiveness of the pull fac-
tors in the host countries. The public good character of refugee pro-
tection refers primarily to the non-excludability and non-rivalry in
‘consuming’ security and political stability benefits spilling over
from refugee-hosting states (Suhrke, 1998). However, some authors
argue instead that states in fact generate excludable benefits, which
are specific to the host country, and thus, asylum provision is more
accurately characterized by a joint product (Sandler, 1992; Betts,
2003). In this case, state-specific benefits might be generated through
international reputation gains or a host country’s ideological and
humanitarian objectives (Thielemann, 2003a). A joint product model
predicts that provision of refugee protection, in terms of the number
of accepted refugees on its own territory, is proportional to the
excludable benefits accruing to the asylum-providing country.
Beyond country-specific self-interests, national states might follow
norms of solidarity and humanity. States that integrate such princi-
ples in their asylum policy have then an increased willingness to
accept asylum seekers as refugees (Thielemann, 2003b). A third rea-
son for (cross-country) unequal refugee burdens may refer to struc-
tural differences in migration-related pull factors, which determine
the relative attractiveness of potential host states for asylum seekers

and other migrants. Beside standard economic factors such as income
and employment opportunities, other non-economic determinants
like a state’s reputation of being liberal in asylum and immigration
policies, having a generous welfare system, or well-established social,
ethnic, and cultural networks might often be even more important
pull factors (see the discussion in Chapter 8).
Generally, because of sovereignty, national states are not obliged to
agree to any international cooperation schemes, and thus, some
often refrain from any refugee protection contributions. Many states,
however, participate to some extent in a latent burden-sharing in
order to satisfy norms of solidarity, geopolitical self-interests, or the
rationale of an ‘insurance’ against future mass inflows by stabilizing
a fragile region (Schuck, 1997).
Several (explicit) refugee burden-sharing regimes are discussed,
ranging from rather ‘soft’ options like the harmonization of asylum
policies and legislations (partly realized by the Amsterdam Treaty
(1999) in the EU) to rather ‘hard’ options such as financial cost-
sharing with overburdened host countries (Acharya & Dewitt, 1997;
Thielemann, 2005; Betts, 2006), or even asylum quotas and a physi-
cal ‘refugee reallocation’ (as discussed in Chapter 6). Some of these
policy options increase efficiency by equalizing marginal costs across
countries for asylum provision and stop the general ‘race to the bot-
tom’ in asylum standards observed since the early 1990s. However,
burden-sharing without tackling the root causes of unwanted emi-
gration is generally considered as insufficient and not sustainable for
any asylum regime.
2.4.2 Migration prevention and aid assistance

Policies that control and restrict migration are costly, and other
policy approaches may therefore prove to be more effective in reduc-
ing unwanted immigration. Supplementary to the outlined, rather
reactive, measures of asylum provision, root cause strategies and
migration preventive policies may be more effective in the long term.
These include conflict resolution by diplomatic or military interven-
tions, intensive humanitarian assistance before, during and after a
conflict and refugee crisis, and enhanced development promotion
(for example, by trade, FDI, or aid). Obviously, identification of the
root causes and formulation of appropriate policies is a complex task
and requires country-specific analysis. One policy instrument for

tackling the root causes of mass emigration, development assistance,
is at the heart of the subsequent analysis.
Western countries have an ongoing interest in the potential of
development cooperation as effective migration prevention policy.15
Improving the economic, social, and institutional conditions in
the countries of origin by an extended provision of financial and
technical aid is central in such proactive migration prevention
approaches.16 Aid transfers may reduce migration propensities by
improving economic and governance conditions of emigration coun-
tries. However, some authors alert to the contrary effect. For instance,
de Haas (2006) argues that such interventions may rather accelerate
economic-motivated emigration and secondary refugee move-
ments. This argument is related to the ‘migration hump’ theory,
which states that migration propensities of the very poor are gener-
ally lower than those of the better off. However, empirical evidence
of this nonlinear effect of economic development on emigration is
rather ambiguous (Rotte et al., 1997; Neumayer, 2005b). Similarly,
interventions for the promotion of democracy and good governance
institutions are effective only in the long-run, but for the short-
run, they might stimulate further upheaval and refugee outflow
(Schmeidl, 2001; Nyberg-Sørensen et al., 2003; Neumayer, 2005b).
Furthermore, the effectiveness of aid interventions in reducing
the risk of violent conflict is highly disputed. While some authors
argue that under certain conditions, particularly in some later peri-
ods of post-conflict situations, aid is effective in reducing conflict
risks (Collier et al., 2003; Collier, 2007), other authors argue that
some types of aid might even increase the probability of conflict
occurrence (Esman & Herring, 2001). This debate about the effec-
tiveness of aid with respect to the mitigation of conflict risk and
emigration pressure is still unresolved, however, progress in gover-
nance structures and economic development are definitely impor-
tant elements of a comprehensive international refugee policy.
Simple evidence shows that almost all refugees originate from coun-
tries that are ranked in the lowest two quintiles of global poverty and
which experience eminent levels of internal violence (Münz &
Weiner, 1997). Since poverty and violence are often interrelated, any
aid policy approach needs to tackle both issues simultaneously to be
effective. However, proper and clear evidence on the effectiveness of
aid for emigration-prevention purposes is still lacking. Nevertheless,

several donor countries have re-designed their development policies
and reallocated their aid budgets for migration management objec-
tives. Chapters 8 and 9 further discuss this issue and provide evi-
dence about the influence of different aspects of refugee and asylum
migration on the decision of donor governments to re-formulate
development policies and to reallocate their respective aid transfers.
3
International Refugee Burdens
Abstract17
Many developing, as well as developed, countries provide temporary

asylum to a large population of refugees and most of these host
countries claim to be overburdened. The effective burden a country
has to shoulder is difficult to determine since the absolute number of
refugees must be evaluated in the light of country-specific condi-
tions in the host country. This chapter provides a methodology that
makes refugee burdens more comparable on a cross-country basis.
Taking into account different aspects of a host country’s capacity
we provide a Refugee Burden Index that is based on proxy indi-
cators representing the economic, socio-demographic and politico-
institutional environment. This methodology is applied on a sample
of 174 countries, revealing the extent of inequalities in the pattern of
refugee burden-sharing, both globally and regionally.
3.1 Introduction
There is an ongoing debate in academia18 as well as in politics19 focus-

ing on national contributions to the international regime of human
protection and financial assistance to refugees and asylum seekers.
Citizens and policymakers of many refugee-hosting countries espe-
cially in the industrialized world perceive their country to be over-
burdened and disadvantaged in comparison to other countries.
However, most refugee-producing countries – which are in most
cases conflict-affected countries-, are situated in the developing
25
world and it is usually their neighboring countries (first asylum

countries) that must bear the lion’s share of the global refugee bur-
den in absolute terms. In 2003, developing regions hosted some
6.5 million refugees (67 percent), whereas the developed world pro-
tected some 3.2 million refugees (33 percent) (UNHCR, 2004d).
Although the largest share of the hosted asylum-seeking and refugee
population is borne by the low and least developed countries, some
of the high-income OECD countries also provide refuge to a quite
impressive number of people.
In public discussions, the question of whether a country is rather
overburdened or not is most generally answered by subjective
feelings instead of more analytical reasoning.20 For establishing the
ongoing Refugee Burden-Sharing (RBS) debate on a more solid and
impartial foundation, this study provides a methodological frame-
work to make refugee and asylum seeker burdens internationally
more comparable. The RBS debate lacks in a common and widely
accepted measure reflecting the status quo of the relative refugee bur-
den. For achieving wide acceptance, a refugee burden measure should
account for fairness, or equity, as well as performance aspects.
In the next section we discuss the extent to which refugees and
asylum seekers constitute a burden for host countries by assessing
their impact on a host country’s society and economy. In Section 3.3
we develop the methodology for a standardized and internationally
applicable Refugee Burden Index (RBI), which takes into account not
just the absolute number of hosted refugees and asylum seekers but
also includes country-specific capacities in categories such as eco-
nomic performance, population absorbance capacity, and societal
and political stability issues. The fourth section contains the applica-
tion of that measurement tool based on a sample of 174 developing
and developed countries of which 145 countries were refugee-
hosting countries in 2003. A country-specific and a region-specific
RBI enable us to make inferences concerning the ‘degree of overbur-
den’ a country or a region in the sample experienced in 2003.
3.2 Assessing the refugee burden to host countries
For assessing the ‘burden’ of asylum seekers and refugees to host

societies, which literally implies that there is a negative impact (net
costs) of hosting refugees for a receiving country, we must break
International Refugee Burdens 27
down this burden into different core factors. That is, we should
clearly distinguish the differences between asylum seekers and refu-
gees, as well as the differences between developed and developing
countries as host societies. Although most countries of the world
have signed the 1951 Convention relating to the Status of Refugees and
its extending 1967 Protocol, the receiving countries’ asylum systems
are not at all standardized. The systems differ in their process of
adjudicating the claims for refugee status, for example, systems may
differ in the status decision duration, the opportunities for second or
third appeals and the naturalization procedure. Therefore, we must
be cautious with both cross-country comparisons and the asylum
and refugee data itself (Crisp, 1999; Hovy, 2001). Secondly, it would
be naïve not to take into account the very heterogeneous composi-
tion of host societies. Western industrialized countries with well
established immigration institutions, powerful economies and stable
societies are not directly comparable to developing countries with
severe economic, political and bureaucratic deficiencies. That is,
however, the task of this study: to make refugee burdens assessable
across heterogeneous host countries.
The impact of asylum seekers and refugees tends generally to be
larger in those countries in which asylum seekers as well as refugees
account for a higher proportion of total population. Among industri-
alized countries there is considerable variation in the numbers of
arriving asylum seekers (Neumayer, 2004). However, even if asylum
seekers account for a non-negligible share of the population, they are
rather unlikely to have significant economic, demographic, environ-
mental and other impact on the host society at least in developed
countries. This means that asylum seekers may
affect greatly the public opinion about immigration if they are

perceived as a growing population, even if not yet a significant
one in numerical terms. (Martin et al., 2005, p. 101)
Thus, whether in fact or just perceived, asylum seekers and refu-

gees constitute at least a gross burden to the receiving and hosting
countries in economic, fiscal, political, institutional, societal and
environmental terms.
Direct fiscal costs for maintaining an asylum adjudication system,
detention, care and maintenance, return of rejected applicants and
so on are actually the most readily measurable impacts. These costs

vary immensely across countries. One of the few studies elaborating
the cost of asylum processing and social security estimated the fiscal
costs per asylum seeker in 1994 to be US$ 16,596 in Denmark,
US$ 10,299 in Sweden and US$ 4,622 in Austria (Jandl, 1995), whereas
Ghana, for instance, estimates its annual costs per refugee to be
about US$ 348 (US$ 29 per month) (Betts, 2006). Asylum costs vary
across host countries to the degree that they keep asylum seekers in
reception or detention centers, maintain them in refugee camps or
disperse them throughout the host country, granting them work per-
missions or providing generous social support and so on In fact,
there is no international standard for hosting asylum seekers and
refugees and international refugee law does not explicitly prescribe
how to host asylum seekers and refugees adequately.
Generally, discussions on asylum adjudication costs fall short by
focusing predominantly on gross expenditures, not taking into
account the possible positive economic impact of asylum seekers on,
for example, tax revenues. Depending on the host country’s policy
on granting work authorization based on the demographic and edu-
cational characteristics of refugee populations, asylum seekers might
be net contributors to the host country’s fiscal budget (Martin et al.,
2005). However, even if eligible, Borjas (1994) points out for the US
that at least in the short run, asylum seekers and refugees have worse
economic outcomes (wages) compared to other labor migrants. In
the long run, however, the outcomes of these immigrant groups may
adjust to be similar (Borjas, 1982).
Some studies elaborating on labor market participation rates in
developed countries find on average a higher unemployment rate
among refugees compared to the local population.21 The impact of
immigrants on the host country’s labor market depends highly on
whether immigrants are substitutes or complements to the native
labor force. Having different characteristics in terms of skills and
education, immigrants that are complements to native labor endow-
ment should rather be beneficial and generating a positive impact on
employment and earnings if they are allowed to join the workforce.
Furthermore, in the long run for industrialized host countries, immi-
grants may fill demographic gaps caused by aging populations.
Positive impacts are often not extensively realized by the native
society of a host country and therefore asylum has been one of the
most controversial issues in discussions about immigration. Public
perception of asylum migrants is mainly determined by the immigra-

tion motivations of the asylum seekers (economic vs. forced), cultural
and linguistic differences to the resident populations or unfamiliar
social and cultural practices of the newcomers tending to create ten-
sion either among immigrant groups themselves or between asylum
migrants and the native population. Sometimes, these tensions are
capitalized on and reinforced by political leaders who expect to ben-
efit from a strict stance towards unwanted arrivals. Though in some
particular cases, a large influx of forced migrants may even be wel-
comed by local residents if the situation the arrivals flee from is well
known and the media have already ‘generated some altruism’.
While, an influx of a large foreign population generally destabi-
lizes the social structure of a society, traditional immigration coun-
tries have typically developed a capability for integrating foreigners
without weakening the social cohesion. Nevertheless, large-scale
refugee inflows always stress local capacities even in cases of well-
established formal and informal institutions. The higher the perfor-
mance of these institutions, the more stable the local community is
and the less likely that asylum seeker constitute a threat for the host-
ing country.
For many host countries, especially in the developing world,
another factor which is crucial in assessing the impact of refugees is
the length of their stay. The average duration of major refugee situa-
tions increased from 9 years in 1993 to 17 years in 2003 (Loescher &
Milner, 2005).22 Long-lasting or protracted refugee situations are
not only a waste of life time and resources of the refugees, but have
also serious consequences for the host countries by nurturing insta-
bility, tensions and even conflicts. Large, neglected and frustrated
populations living in obscurity and on subsistence-level pittances
are easy to recruit for armed groups and thereby jeopardizing the
societal and political situation of the host country and region
(Loescher & Milner, 2005).
The above-mentioned factors vary dramatically between the
refugee-hosting countries of the world and it seems presumptuous
to compare refugee capacities of highly developed Western countries
with the weak institutionalized developing countries of, for exam-
ple, sub-Saharan Africa. However, we venture to tackle this challenge
in our next section and develop a simple quantitative measure for
making large-sample cross-country refugee burden comparisons
assessable.
3.3 Methodology of a refugee burden index
In a large-scale refugee or asylum situation, host countries are placed

under tremendous pressure in providing facilities for a high number
of people.23 Reception countries are faced with the responsibility
of ensuring the protection of human rights to refugees as required by
UN Refugee Convention 1951/67, as well as providing for their basic
personal economic and social needs such as food and other essential
supplies, medical care, accommodation, education and security.
A country’s capacity to satisfy these needs differs enormously
between the industrialized and the developing world. Probably the
best proxy for measuring the capacity to meet these rather economic
requirements is the income per capita of a country. We assume that
the higher a country’s GDP per capita, the higher its capability to
provide basic economic and social resources to a refugee population
and the lower the burden per refugee to the host country.
However, the effective burden of hosting refugees and asylum
seekers is more than solely affording the economic requirements of
refugees. In fact, if a large refugee population is received in an area or
country marked by poverty, ethno-societal tensions or even civil
strife, the political and social environment of the hosting country
may deteriorate, giving rise to latent or open conflict (Loescher &
Milner, 2005). The potential for destabilization caused by the pres-
ence of refugees is often higher in countries with weak governance
institutions and democratic structures. A RBI therefore must be a
surrogate measure, incorporating all of these important capacity
dimensions necessary for providing a safe, appropriate, and non-
destabilizing refuge to forced migrants. Before such an RBI itself is
composed, several sub-indices have to be created for each of those
dimensions; that is, representing economic conditions, population
absorbance capabilities, and the quality of governance and demo-
cratic institutions.
3.3.1 Refugee capacity index

3.3.1.1 Ability-to-care
The economic condition of a refugee-receiving country is highly
important for indicating the country’s resource capacity for bearing
a large refugee population. We use income per capita as an indicator
of a country’s ability to care for refugees, such as providing food,
health and education services, maintenance, security and bureau-

cratic services. Thus, income per capita serves as a proxy for all
economic capacity aspects in supporting refugees and asylum seek-
ers. Therefore, we use the income index GDPI as an ability-to-care
indicator, which simply reflects the logarithm of a country’s income
per capita. For compressing the distribution range, the income per
capita of a country is adjusted by the logarithm of the actual value
and corrected by minimum (US$ 100) and maximum (US$ 40,000)
goalpost values, that is, the GDPI value is zero if a country’s per cap-
ita income is US$ 100 per annum.24 This implies that the country’s
marginal economic capacity for hosting one additional refugee is
zero. An income near the maximum value of US$ 40,000 results in a
GDPI value of one, implying that a country’s marginal economic
capacity for hosting one more refugee is ‘infinite’.
3.3.1.2 Socio-political acceptance of immigrants

We also distinguish between two further factors influencing the
level of acceptance of immigrants. First, we use a modified popula-
tion density measure, the ratio of population per arable land, as an
indicator of the capacity of a country to absorb new arrivals. It is
often argued that the ability of a host country to absorb refugees is a
function of both the national population size, as well as the available
arable land which is often an important means of sustaining the
nutrition of refugee families in developing countries (Jacobsen, 1997;
Kibreab, 1997). Refugees are likely to integrate less easily in a host
society where space at their disposal is limited. Population density
seems, therefore, an adequate measure for representing both space
limitation and population strain. In countries with a high popula-
tion pressure caused by, for example, a high population growth or
large former inflows of migrants, an additional large influx of refu-
gees may over-stress a country’s absorbance capacity and lead to
social tensions. As a maximum goalpost value we use one million
persons per square kilometer of arable land, which is roughly the
population density of some microstates. We use logarithms for
compressing the range of the scale at the high end, which gives
densely populated countries a relatively lower weight. Thus, a
Population Density Index (PDI) of close to zero represents a highly
crowded country with very limited additional capacity and space
resources for further refugees.
Secondly, host societies often perceive refugees as a threat to their

country’s stability and their major societal values (Weiner, 1993).
Thus, the degree of acceptance for new migrants might be deter-
mined by the existing ethnic composition of a country, that is we
assume that a more homogeneous (that is less fractionalized) popu-
lation is more able to cope with further immigrants as new members
of society (see, for example, Annett, 2000). To represent this societal
dimension we use an ethnicity dataset by Alesina et al. (2003) that
provides country-level data on the shares of major ethnic, linguistic
and religious groups. Given these shares, we calculate the ethno–
linguistic–religious fractionalization index ELRFI as an equally
weighted composite Herfindahl index. The values of the ELRFI score
within a range of zero to one. A value near one represents a highly
homogeneous ethnic, linguistic and religious structure, implying a
high sociocultural stability, enabling the society to cope with new
immigrants.25
3.3.1.3 Politico-institutional performance

In many host countries the societal tensions potentially caused by
the presence of refugees cannot be adequately articulated because
of the absence of well-developed democratic channels and good
governance institutions. State performance in terms of providing
capable democratic processing and efficient governance structures is
extraordinarily important in managing societal stability, particu-
larly in times of a large influx of asylum seekers and refugees. For
indicating the quality of democratic institutions, we use the dataset
on political rights and civil liberties of FreedomHouse (2005).26 These
data are originally scaled from 1 (best) to 7 (worst), and we rescale
them so that the sub-index for political freedom, PFI, ranges from
zero to one.27 A PFI near one represents well-established democratic
institutions, which imply robust capacities for transforming societal
and cultural conditions and developments into representative politi-
cal outcomes.
Furthermore, the presence of high-quality governance institutions
is an essential factor in the determination of whether a country has
the capacity to ensure political stability. A large-scale influx of asy-
lum seekers and refugees demands well-functioning and efficient
governance institutions to prevent a destabilization of a society’s
structure. Taking this into account, we introduce the governance
and political stability index PSI for representing a country’s gover-

nance capacity. We use the political stability measure of a dataset
provided by Kaufmann et al. (2003) for reflecting this issue.28 The
data are originally scaled from –2.5 to +2.5 and we rearrange them to
a scale from zero to one.
For calculating an aggregate index reflecting the core dimensions
of economic, societal, political and institutional aspects relevant for
hosting refugees, we generate the refugee capacity index (RCI) by
aggregating all of the above-mentioned sub-indices. For reasons of
simplicity and the lack of more convincing alternatives, we assess all
three dimensions x (economic, societal and politico-institutional)
the same weight ax and generate the RCI for country i as:
RCI i ␣GDPIGDPI i ␣PDI PDI i ␣ELRI ELRI i ␣PFI PFI i ␣PSI PSI i

(3.1)
economic societal polit–institutional
Table 3.1 displays the Pearson correlation coefficients for each

component of the RCI based on our dataset of 174 countries. Except
for the PDI all sub-indices of the RCI are more or less highly posi-
tively correlated to each other, which reveals collinearity between
the different dimensions. For our purposes this is not a problem but
rather an asset because this diminishes the importance of the weights
ax we assign to each component and consequently, assigning the
same weight to each one seems less ad hoc.29
Table 3.1 Correlation matrix of the RCI components
PDI ELRI PFI PSI RCI
GDPI –0.11 0.42* 0.55* 0.69* 0.89*

(–1.42) (6.15) (8.70) (12.48) (25.18)
PDI –0.24* –0.12 –0.09 0.01
(–3.22) (–1.56) (–1.15) (0.19)
ELRI 0.20* 0.38* 0.54*
(2.68) (5.33) (8.42)
PFI 0.58* 0.78*
(9.44) (16.15)
PSI 0.82*
(19.07)
Notes: t-values in parentheses; critical t-value: 1.974 (5 percent level).

Source: (*) significant at the 5 percent level.
Through aggregation, the RCI is likewise scored between zero and

one. The higher the score of the RCI the higher is a country’s overall
capacity for receiving and hosting a relatively large refugee popula-
tion. The RCI composites the core factors which are essential for
appropriate refugee management and, furthermore, it allows cross-
country comparisons concerning the general capacity of host coun-
tries for receiving additional refugees.
3.3.2 Refugee gaps

We are now able to calculate hypothetical country-specific refugee
populations based on an egalitarian burden-sharing concept. This
equity concept requires that the numbers of refugees for a specific
country should be proportional to the country’s population, that
is, the applied burden-sharing rule implies equal cross-country
per capita ratios of refugees. These ratios are adjusted by some
country-specific economic, political and social capacity factors.
Taking this into account, we are able to determine country-specific
refugee gaps. We define a refugee gap as the quantitative deviation of
the ‘de facto’ from the ‘should be’ refugee population of a country.
Before introducing the methodology of the refugee gap, we should
give a definition of our understanding of the expressions ‘adjusted
de facto’ and ‘adjusted should be’, respectively. We define an equita-
ble number of hosted refugees, that is a ‘should be’ refugee burden,
as a share s out of the total global refugee population, which is equiv-
alent to a country’s national population share out of the total global
population. However, these ‘fair’ shares are adjusted (indexed adj) by
the above-mentioned country-specific factors. These factors repre-
sent the asylum provision capacities derived in the previous section
and expressed in a country’s RCI. The fair and adjusted burden of a
country i in terms of the ‘should be’ refugee population ref adj,fair is
eventually given by:
ref iadj,fair siadj,fair ¦ ref idf (3.2)

i
sifair RCI i ¦ i sifair RCI i

1
with siadj,fair
and sifair popi ¦ i popi .
1
Basically, such an exemplary adjusted ‘should be’ figure could be

used as a benchmark for an international refugee burden-sharing
regime based on fairness and host country performance. To deter-

mine the virtual refugee burdens of all countries in the sample, we
must compare the current de facto (indexed by df) situation with
this hypothetical benchmark. Therefore, we calculate in an analo-
gous manner the adjusted de facto refugee population ref adj,df for a
country i. For this we adjust a country’s de facto share of the total
global refugee population by its value of the RCI:
ref iadj,df siadj,df ¦ ref idf (3.3)

i
¦ i sidf RCI1
1
with siadj,df sidf RCI
1

i i
and sidf ref idf ¦ i ref idf .

1
Before calculating the actual RBI, which expresses a country’s

effective refugee burden level, we must calculate the virtual gap
between the ‘fair’ figures of eq. (3.2) and the ‘de facto’ figures calcu-
lated in eq. (3.3). The so-called ‘adjusted de facto’ refugee burden gap
for country i refers to:
gapiadj,df = ref iadj,df − ref iadj,fair . (3.4)
We calculate in Section 3.3.4 the Gap 03 value, which states to what

extent the number of persons seeking refuge deviates from the calcu-
lated ‘should be’ number. Then, we can consider a country with a pos-
itive Gap 03 as ‘overburdened’ (at least in 2003). Otherwise, a country
is below its equitable number of refugees to host and it would still have
capacity to host more refugees without becoming overburdened.
3.3.3 Refugee Burden Index

Based on the above methodology, the RBI for country i can be speci-
fied as:
gapiadj,df
RBI i (3.5)
ref iadj,fair
The RBI is scaled from –1 up to positive infinity. An RBI score of –1

stands for a totally unburdened country in terms of having not
hosted any refugee or asylum seeker in 2003. A score of zero stands
for a fairly burdened country, that is, the country hosted exactly the
number of refugees and asylum seekers it should equitably host.

And, for instance, a score of 2.5 reflects an overburden level of
250 percent, that is, this country hosts 2.5 times more refugees and
asylum seekers than the calculated fair number demands.
For our application of this concept we use aggregate figures docu-
menting the number of refugees and asylum seekers by the end
of 2003 in 174 developed and developing countries provided by
UNHCR (UNHCR, 2004d). Of these, 145 hosted either refugees or asy-
lum seekers or both in 2003. We can aggregate these two figures
because there is no overlap between the two categories. Including only
these two categories in our calculations is perhaps rather narrow,
however they are the best documented. UNHCR’s broadest figure of
a country’s refugee population is the ‘Total Persons of Concern’
(TPOC), taking additionally into account the number of Internally
Displaced Persons (IDPs), returned refugees, returned internally dis-
placed persons and others. However, the figures used to calculate the
TPOC are vague and rather rough estimates because those people
are often not assisted and therefore not registered by UNHCR.
Therefore, our calculations are based on the rather narrow aggregate
measure of refugees and asylum seekers.30
3.3.4 Results
We now apply the previously outlined concept of a RBI to our set of
countries, consisting of 145 hosting countries of refugees and asylum
seekers and a further 29 non-hosting countries (in 2003).31
Table 3.2 depicts the results for the top 20 most burdened refugee
receiving countries.32 Among the top 20 there are sixteen countries
from the African continent, which reveals the main region of con-
cern. In contrast, only two European countries (Armenia included)
were among the top 20. In 2003, the absolute runaway of the
ranking is Armenia, which was overburdened with approximately
42 times its respective equitable number, corresponding to an abso-
lute excess burden of around 210,000 persons. Also severely overbur-
dened were Djibouti (Gap 03 of about 29,000 persons) and Serbia &
Montenegro (Gap 03 of about 278,000 persons), which were both
shouldering more than 20 times their equitable number.
The refugee population of these countries is rather homogeneous,
that is these countries act as first asylum countries for citizens of
neighboring countries involved in any latent or open conflict, for
Table 3.2 Top 20 burdened countries in 2003
Country RBI 03 Gap 03
Armenia 42.15 210,443

Djibouti 29.82 28,990
Congo (Brazzaville) 26.80 114,419
Serbia & Montenegro 23.21 278,408
Guinea 23.07 216,535
Chad 20.88 188,897
Zambia 17.92 240,417
Tanzania 17.30 740,204
Liberia 16.96 51,266
CAR 15.96 63,622
Sierra Leone 13.93 78,197
Uganda 11.33 294,811
Burundi 9.37 67,216
Congo (Kinshasa) 9.09 380,956
Iran 8.73 860,747
Pakistan 7.90 1,297,087
Kenya 6.40 247,716
Gabon 6.22 14,049
Cote d’Ivorie 5.88 96,888
Guinea-Bissau 5.40 8,581
example, Armenia for Azerbaijani, Djibouti for Somalis and

Ethiopians, Serbia & Montenegro for Croatians and people from
Bosnia-Herzegovina. However, most Western European countries act
rather as third countries of asylum and operate as safe havens for a
whole variety of nationalities. Therefore, overburdened countries
like, for example, Sweden (RBI 03 of 2.67, rank 29), the Netherlands
(RBI 03 of 2.30, rank 31), Denmark (RBI 03 of 2.16, rank 32) and
Switzerland (RBI 03 of 1.89, rank 34) host a very heterogeneous refu-
gee and asylum-seeking population, reflecting the particular role
such countries play in the international refugee burden-sharing sys-
tem. There were four other Western European countries that can also
be considered as overburdened, they were Norway (RBI 03 of 1.07),
Germany (RBI 03 of 0.83), Austria (RBI 03 of 0.46) and the UK (RBI 03
of 0.41).
Furthermore, we can apply the concept of the RBI not just globally
on all countries of the sample, but we are also able to calculate
intra-regional refugee burdens.33 As Table 3.3 illustrates, each world
Table 3.3 Top 5 intra-regionally most burdened

countries in 2003
Region RBI 03 Gap 03
Europe
Armenia 43.96 297,648
Serbia & Montenegro 24.23 394,084
Sweden 2.82 87,782
Bosnia-Herzegovina 2.77 23,367
Netherlands 2.44 120,700
Asia-Pacific
Nepal 33.14 156,444
Thailand 4.18 80,183
Papua New Guinea 3.73 5,600
Australia 3.31 27,484
New Zealand 1.80 2,811
Americas
Canada 2.80 123,171
Belize 2.37 703
Costa Rica 1.75 8,953
United States 1.07 415,853
Ecuador –0.09 –1,110
Caswaname
Iran 1.35 506,135
Pakistan 1.15 717,623
Iraq 0.25 29,173
Saudi Arabia 0.11 17,921
Algeria –0.28 –53,551
Southern Africa
Zambia 8.46 224,562
Namibia 1.05 8,148
Angola –0.46 –14,707
Malawi –0.46 –12,223
South Africa –0.48 –79,180
West Africa
Guinea 9.69 163,464
Liberia 6.98 37,902
Sierra Leone 5.63 56,811
Cote d’Ivorie 2.05 60,866
Guinea-Bissau 1.84 5,261
Central Africa
Congo (Brazzaville) 0.96 43,174
Chad 0.54 51,723
Continued
Table 3.3 Continued
Region RBI 03 Gap 03
Tanzania 0.29 131,010

CAR 0.20 8,227
Burundi –0.27 –20,304
East Africa
Djibouti 4.62 18,779
Uganda 1.25 135,837
Kenya 0.35 56,516
Sudan –0.02 –3,104
Ethiopia –0.51 –149,526
region has regional safe havens for refugees and asylum seekers.34
These countries were much more burdened in terms of the number
of hosted refugees than other countries of the same region.
In Africa, each of the four regions has its own main havens hosting
refugees ‘from the neighborhood’ and which were overburdened up
to around 1,000 percent of their equitable share! In Southern Africa,
Zambia (RBI 03 score of 8.46) hosted around 158,000 refugees from
Angola; in West Africa, Guinea (RBI 03 score of 9.69) hosted around
160,000 refugees from Liberia; in Central Africa and the Great Lakes
region, Congo (Brazzaville) (RBI 03 score of 0.96) hosted about 81,000
refugees from the Democratic Republic of Congo, and the already
mentioned case of Djibouti (RBI 03 of about 4.62).
Discrepancies arise between the global and regional RBI 03 scores,
reflecting the inequalities that exist between different regions
(Table 3.4). For countries that belong to globally overburdened
regions, such as Africa and Caswaname,35 the intra-regional RBI 03
score is lower than their respective global RBI 03 score. For instance,
Guinea has an RBI 03 score on a global basis of 23.07, but calculated
on a regional basis only of 9.69. On the other hand, countries, which
were part of an under-burdened region like Asia-Pacific or the
Americas, experience a higher RBI 03 on a regional basis, compared
to the score on a global basis. This is the case, for example, for
Australia, which has an RBI 03 score of –0.32 on a global basis, but a
score of 3.31 on a regional basis. Countries of under-burdened regions
have higher RBI scores calculated on a regional basis compared to
Table 3.4 Inter-regional Refugee Burden Index in 2003
Region Refugees (de facto) Refugees (fair) Gap 03 RBI 03
Europe 1,629,681 1,698,171 –68,490 –0.04

Asia-Pacific 798,469 5,079,545 –4,281,077 –0.84
Americas 649,948 1,757,084 –1,107,136 –0.63
South Africa 418,003 209,061 208,942 1.00
West Africa 690,073 306,534 383,539 1.25
Central Africa 1,729,803 122,011 1,607,792 13.18
East Africa 1,017,107 185,494 831,613 4.40
Africa 3,854,985 823,100 3,031,885 3.68
Caswaname 3,194,777 769,959 2,424,818 3.15
those calculated on a global basis. This is due to the different total

population accounts, global versus regional, applied in the two
exercises.
In general, these results reveal a noticeable inter-regional discrep-
ancy of refugee burdens. As Table 3.4 shows, particularly Asia-Pacific
and the Americas are under-burdened in an inter-regional view;
Europe is slightly under-burdened, whereas Africa (and especially
Central Africa) and Caswaname are highly overburdened.
Table 3.5 itemizes the regional refugee burdens exposed in
Table 3.4. For instance, Europe was a net source region for refugees
and asylum seekers towards Asia-Pacific and especially the Americas,
whereas it was a net recipient from the other regions. Although the
Americas was the only region that was a net recipient of refugees and
asylum seekers, vis-à-vis all other regions it was still under-burdened
(RBI 03 score of –0.63). Another under-burdened region, Asia-Pacific
(RBI 03 score of –0.84), was a net refugee-producing region particu-
larly towards the Americas.
Generally, these figures display two important insights: first, the
global burden of refugee hosting is primarily shared among the
countries within a particular region, and second, unequal distribu-
tion of refugee population is a phenomenon that is observed within
as well as between the various regions of this world.
3.4 Concluding remarks
The often-voiced statements by policymakers and citizens concern-

ing the national burdens in terms of the number of received asylum
Table 3.5 Inter-regional Refugee Burden Matrix (gross/net stocks) in 2003
Caswaname East Central West South Americas Asia/ Europe in

Africa Africa Africa Africa Pacific from
27,816 2 0 5 90 215,198 20,068 1,485,863 Europe

83 0 41 138 30 100,430 828,544 143,113 Asia/Pacific
13 0 0 2 1 82,130 658 12,621 Americas
21 11 202,119 60 217,749 3,106 54 18,301 South Africa
1,146 3 244 443,841 475 33,437 1,043 36,853 West Africa
179 32,253 943,273 51,479 90,100 15,047 224 39,977 Central Africa
96,778 905,279 152,451 329 9,054 74,774 11,602 114,207 East Africa
3,236,293 46 35 26,009 161 98,893 38,337 472,214 Caswaname
x Europe
x –123,045 Asia/Pacific
x –99,772 –202,577 Americas
x 3,105 24 18,211 South Africa
x 415 33,435 905 36,848 West Africa
x 51,235 –112,019 15,047 183 39,977 Central Africa
x 120,198 326 9,043 74,774 11,602 114,205 East Africa
x –96,732 –144 24,863 140 98,880 38,254 444,398 Caswaname
seekers and hosted refugees provides the motivation for this analysis.
We focus the effective burden of refugee-hosting countries and inves-
tigate the overall picture of the global and regional distribution of
the global refugee population. For this, we provide a methodology
for indicating a cross-country pattern of effective refugee burdens.
‘Effective’ in this context means that country-specific factors that
are significant in determining a country’s refugee capacity are taken
into account. The resulting RBI reflects several major dimensions
that seem reasonable for measuring a country’s capacity for hosting
refugees. This RBI methodology is applicable on all levels of interest:
we calculate the RBI for a sample of 174 countries that are categorized
into eight global regions. This affords cross-country and cross-
regional comparisons about the actual burdens expressed in a degree
of overburden.
The main critique of this concept is twofold. First, the results are
driven mainly by the applied egalitarian equity concept. The appli-
cation of an alternative concept would imply, at least slightly, differ-
ent results. The second shortcoming is the ad hoc selection of the
different economic, political-institutional and social indicators that
were used for representing a country’s refugee-hosting capacity, and
the determination of their relative importance. Thus, simplicity may
be both an advantage and disadvantage of this concept.
Finally, this concept of a RBI may enrich the Current RBS debate.
In this RBS debate, of which UNHCR’s Convention Plus initiative
(established in 2003) is only one part, discussions about resettle-
ment, voluntary repatriation and local integration efforts, supported
by international aid allocation for refugees, require a clear under-
standing of the status quo concerning the global refugee situation
and its impact on heterogeneous host countries. Countries differ in
their capability to receive asylum seekers as well as in their willing-
ness and ability to share refugee burdens physically, but also finan-
cially. If such refugee-related financial transfers from donor countries
to other refugee-hosting countries would also be taken into account,
the ‘net refugee-related burden’ would be lower for recipients and
higher for donors. Such a combined measure of physical and finan-
cial contributions to the provision of the global or regional public
good ‘asylum’ would probably change the RBI ranking significantly.
4
Determinants of Forced
Displacement: The Case
of Aceh, Indonesia
Abstract36
The purpose of this chapter is to identify the determinants of dis-

placement behavior based on various push and pull factors at the
village level. The study concentrates on changes in village popula-
tion during three years of civil conflict (1999–2002) in Aceh,
Indonesia. The empirical analysis is based on a unique dataset from
two census rounds of the Indonesian Village Potential Census
(PODES). It uses data on around 5,200 Acehnese villages and relates
village-level population change to conflict variables, geographic pat-
terns and traditional socioeconomic determinants of migration. By
applying quantile regressions, the push (outflow) factors and the pull
(inflow) determinants of migration can also be distinguished. We
identify the following factors as the main determinants of the Aceh
migration pattern in this period: First, conflict clashes induced large
rearrangements of the population between villages in highly affected
districts, as well as strong village emigration from the geographically
remote regions in Central Aceh towards the less conflict-affected
coastal industrial areas. Besides conflict factors, an (ongoing) rural–
urban migration process, driven by socioeconomic factors, has taken
place during the conflict period. Second, there is also evidence
that security considerations, such as the presence of police in a vil-
lage or neighborhood, were emigration-reducing (or, immigration-
inducing). Third, although the presence of ethnic-Javanese has not
43
been a primary cause of conflict incidence, their intimidation by

the rebel movement has led to a significant outflow, primarily from
conflict-affected villages in Central Aceh. These results reveal that,
beside a conflict-induced fear of violence, population movements
in Aceh have also been an outcome of traditional migration
determinants.
4.1 Introduction
Worldwide, the number of international and intra-state conflicts has

fallen dramatically since the end of the Cold War. This has led to a
continuous reduction in the world’s refugee population over the past
years. However, internal displacement has not been reduced to the
same extent that cross-border refugee movements have. Globally, at
the end of 2005, about 23.7 million internally displaced persons
(IDPs) were seeking refuge within their own conflict-affected
countries (IDMC, 2006a). Obviously, large displacements are induced
by armed conflicts and grave human rights violations, but beyond
this, the role of other political, economic or social determinants of
internal displacement is still unresolved.
Although some hypotheses on the determinants of forced migra-
tion decisions have been put forth in the refugee studies literature,
the relevance of traditional socioeconomic migration determinants
has not been systematically analyzed at a disaggregated level. At
the cross-country level, most studies emphasize that violence is the
major push factor of forced migration flows, indicating that institu-
tional or economic factors have a relatively small impact (see, for
example, Schmeidl, 1997; Moore & Shellman, 2004). However,
using Colombian household data, Engel & Ibánez (2007) find that,
even in a conflict environment, economic incentives play an
important role in household displacement decisions, although the
impact of economic incentives is less strong where violence levels
are high.
This study identifies the determinants of displacement during a
period of conflict in Aceh, Indonesia. The Aceh conflict was politi-
cally motivated and arose between the Indonesian military forces
TNI (Tentara Nasional Indonesia) and the Acehnese Freedom Movement
GAM (Gerakan Aceh Merdeka). This conflict was particularly intense
between 1999 and 2004. During this period, large-scale displacements
Determinants of Forced Displacement 45
were a recurrent feature, and the number of displaced persons has

been estimated at more than 500,000 (IDMC, 2006b).
Our main task is to investigate the major determinants of these
internal displacements during three years of severe clashes (1999 to
2002) in the Aceh province. For this purpose, we use data on 5,211
Acehnese villages from two rounds of the Indonesian Village Potential
Census PODES 2000 and 2003 (BPS, 2000, 2003). Unlike empirical
analysis based on household data, this village-level dataset enables
us to focus on the village-specific determinants of population
changes, which makes this analysis unique.
Although we cannot directly interpret our results as explaining
the individual migration decision of household members, we are
able to identify the most relevant push and pull factors at the village
level. This level of aggregation also allows us to cover almost the
whole area affected by the Acehnese conflict (91.3 percent of
Acehnese villages). We distinguish between the effects of conflict-
related and more traditional socioeconomic determinants of net
population change. Furthermore, and in addition to standard OLS
estimation, we apply a quantile estimation technique, which enables
us to separate the push (outflow) factors from the pull (inflow) fac-
tors driving the change in Acehnese village population stocks during
this period of civil conflict.
Obviously, violence and displacement are strongly linked, and this
relationship is reflected in our results: the presence of conflict, which
is captured by both reported conflict variables and by district con-
trols, reduces net population increase. Beyond this, we are also able
to identify several socioeconomic factors as driving forces of popula-
tion displacement, indicating an ongoing rural–urban migration
movement within the province of Aceh. We find that population
displacement in Aceh is not only caused by the fear of violence, but
also by traditional socioeconomic migration variables. This corrobo-
rates the results of Engel & Ibánez (2007) from their study of
Columbian households.
The remainder of this chapter is structured as follows. The next
section briefly reviews the Acehnese conflict history and its effects
on population displacement in the past decade. Then, we shortly
review the traditional migration literature and discuss its main impli-
cations for our empirical analysis. The empirical section contains an
introduction to the dataset, an explanation of the empirical strategy,
and a discussion of the results of the regression analysis. Some final

remarks conclude this chapter.
4.2 Conflict history and displacement

in Aceh, Indonesia
Indonesia’s recent history is characterized by persistent conflicts and

population displacements in various provinces. Frequent flare-ups of
violence in the different provinces originate from a mixture of eth-
nic, religious, and social causes. The conflict in Aceh was strongly
politically motivated and has been carried out between rebel fighters
and the Indonesian state, thus, it is more closely related to the ‘clas-
sical’ type of civil war. Of all of the clashes since Indonesian inde-
pendence in 1945, the armed conflict in the Aceh region has been
one of the most persistent conflicts, lasting for more than three
decades.
Since the mid-1970s, the Aceh Freedom Movement (GAM) grew
steadily and gained considerable power. This provoked frequent
clashes between the GAM, military forces and paramilitary groups,
culminating in the period between 1999 and 2003. During the
period of political liberalization that followed the resignation of
President Suharto in early 1998, Aceh experienced a short period
of political détente. However, following the riots in the city of
Lhokseumawe, which was one of the hardest-hit areas in Aceh, the
announced process of demilitarization was significantly slowed
(Schulze, 2004; IDMC, 2006b).
In early 1999, frustrated by the lack of substantive changes,
Acehnese student activists initiated a campaign for a referendum
on Aceh’s political status, which rapidly gained support throughout
the province. In mid-1999, military troops and security forces
killed tens of pro-independence Acehnese demonstrators, and plans
for renewed counter-insurgency operations were announced. In
this context of deteriorating conditions, the election of President
Abdurrahman Wahid increased the mobilization of support for
independence in Aceh. In November 1999, a pro-referendum rally
drew an estimated two million supporters (about one half of the
whole Acehnese population) and brought the province to a stand-
still (Sidel, 1999). In 1999, large numbers of Acehnese began fleeing
their homes in response to military and police actions, or out of fear
of being involved in clashes between the security forces and the

GAM. The months preceding Indonesia’s general election in October
1999 saw a dramatic increase in the number of IDPs (IDMC,
2006b).
In early 2000, the GAM announced the rebels’ willingness to
negotiate a ceasefire if military operations, including roadblocks,
door-to-door searches, and other actions to locate GAM members,
were discontinued. In May 2000, the GAM and the Indonesian
authorities agreed on a humanitarian pause. However, on June 1st, a
day before this pause was to take effect, more than 6,000 people fled
their homes in North Aceh because of renewed fighting (IDMC,
2006b). The following months saw continued sweeping operations
as well as ongoing clashes, causing displacement and unrest.
In April 2001, Indonesia’s efforts to end the separatist rebellion in
Aceh entered a new phase with the launch of a military offensive
against the GAM rebels. The majority of victims of this offensive
were civilians, and severe atrocities were committed by both sides.
The district of Central Aceh was most severely hit during this period,
during which hundreds of people were killed by the GAM, the mili-
tary or local militias (ICG, 2002). In this phase, around 32,000 peo-
ple fled from Central Aceh and sought refuge in adjacent regions
(UNDP, 2006). In early 2002, representatives of the GAM and the
Indonesian government agreed to turn the armed conflict into a
political dispute with the involvement of other Acehnese groups. In
December 2002, an agreement on cessation of hostilities was con-
cluded in Geneva, which resulted in a significant drop in the level of
violence.
However, in early 2003, the ceasefire failed and severe clashes
again followed in Aceh. The Indonesian government reacted by
implementing martial law on May 9th, 2003. Since then, enforced
military operations have led to widespread human rights violations.
Thousands of civilians have fled their homes or have been forcibly
relocated by the military (AI, 2004). Martial law introduced a new
round of violence in Aceh, during which the internal displacement
of population into designated villages or camps emerged as a strategy
of war (Hedman, 2005). In this phase, forced displacement has also
been openly used by the military for separating GAM members from
their civilian base. Counter-insurgency operations have relied exten-
sively on recruiting civilians to join militias, civilian defense groups,
and military auxiliary units. These groups are also reported to have
carried out severe human rights violations (AI, 2004).
Between the introduction of martial law in May 2003, and the eve
of the tsunami earthquake in December 2004, an estimated 2,300
people were killed in struggles between the Indonesian government,
the militias, and the GAM (HRW, 2005). In the same period, around
150,000 persons became internally displaced. After the tsunami
earthquake on 26 December 2004, which killed over 100,000 people
and displaced over 500,000 people, a ceasefire was installed (HRW,
2005). In August 2005, the Indonesian government and the GAM
signed a Memorandum of Understanding bringing this 30-year old
conflict to a preliminary end.
According to conservative estimates, the aggravated fighting
and violence since 1999 resulted in more than 500,000 internally
displaced persons (IDPs) and refugees (IDMC, 2006a). The Acehnese
conflict has led to two distinct patterns of displacement. First,
within the province, local people have been temporarily displaced
when their villages were under attack, and have been sheltered
in mosques or community halls, particularly alongside the two
main roads that run along the North and East coasts, and along
the South and West coasts. These displaced persons have usually
remained inside their home province. They typically returned to
their villages within a few weeks and started reconstructing their
houses and livelihoods. The four main areas of displacement
within Aceh were North Aceh, East Aceh, Central Aceh and West
Aceh (Figure 4.1). In the period from 1999 to 2000, the average
length of stay of the IDPs in the various sites was quite short. Later
in 2001, many people who fled their villages had to stay away from
their homes for periods ranging from several months up to almost
two years due to the destruction of their houses and the loss of
other assets (Ramly, 2005). Second, although ethnic-Javanese
were not officially considered by the GAM as targets, there were
thousands of ethnic-Javanese who left Aceh, fleeing mostly to
the neighboring province of North Sumatra or to Java (UNOCHA,
2003; Schulze, 2004). The largest number of such displacements
occurred in 2001. By September 2002, there were about 178,000
IDPs outside of Aceh, most of whom found refuge in North Sumatra.
Only a small number of Javanese sought refuge within Aceh
(Ramly, 2005).
6°0'0''N 96°0'0''E 98°0'0''E 6°0'0''N

SABANG
Banda Aceh Map of Aceh, Indonesia

Sigli
ACEH BESAR Lhokseumawe
BIREUEN ACEH UTARA

PIDIE [NORTH ACEH]
ACEH JAYA ACEH TENGAH ACEH TIMUR

[CENTRAL ACEH] [EAST ACEH]
ACEH BARAT
[WEST ACEH] Takengon Langsa
NAGAN RAYA ACEH

TAMIANG
Meulaboh GAYO LUES
4°0'0''N ACEH 4°0'0''N
BARAT DAYA
[SOUTHWEST ACEH]
Medan
INDIAN ACEH
OCEAN TENGGARA
[SOUTHEAST ACEH]
Tapaktuan ACEH
SUMATERA
SELATAN
[SOUTH ACEH] UTARA
[NORTH SUMATRA]
ACEH
SINGKIL
IM
S
E
U
LU
E
2°0'0''N 2°0'0''N
0 25 50 100 Kilometers
96°0'0''E 98°0'0''E
Provincial Capital
Other Towns
District Boundaries
Note: Simeulue, Sabang,
Lhokseumawe, and Langsa
all have district status.
Note: Map boundaries and locations

are approximate. Geographic features
and their names do not imply official
endorsement or recognition by the UN.
I N D O N E S I A
© 2003 by EAST-West Center
www.eastwestcenter.org
Figure 4.1 Aceh districts

Note: Reprinted by permission of East-West Center.
The subsequent empirical analysis concentrates on net popula-

tion change at the village level for the period between fall 1999 and
fall 2002. Thus, we are able to quantify the effects of the upsurge in
violence that preceded the preliminary ceasefire in December 2002,
but we do not examine the wave of forced displacement that fol-
lowed the introduction of martial law in May 2003.
4.3 Theoretical background of displacements
Displacement in the context of civil conflict is a consequence of

the presence or the threat of a violent attack, and not a voluntary
migration decision in a narrow sense. While many individuals or
whole households flee out of a conflict area to save their lives, we
also observe many people who do not leave their homes to seek ref-
uge. There are at least two explanations for this phenomenon. First,
violence is not randomly targeted, that is some individuals or groups
within the population are more prone to be violently targeted by
armed groups, which make these people more disposed to flee than
others. Second, when deciding upon staying or leaving, individuals
or households do not only take into account security factors, but
other traditional (socioeconomic) determinants as well. These two
hypotheses are not mutually exclusive. Whether targeted individu-
als or households prefer to stay at home also depends on the degree
of risk aversion. As socioeconomic determinants are expected to play
a significant role in explaining forced migration, we refer briefly to
the main implications of the traditional migration literature with
respect to these determinants.
In traditional rural–urban labor migration models the rate of
migration is higher, the larger the urban–rural wage gap, and the
higher the perceived probability of finding a job in the modern sec-
tor (Harris & Todaro, 1970). In the new economic migration litera-
ture, Stark (1991) and other authors refine these early migration
models adding new variables, such as income uncertainty, relative
deprivation (Faini, 1996), and human capital investment in children.
Households spread their risks in structurally different markets by
pooling and sharing their incomes afterwards. This is regarded as an
insurance against uncertain income flows from specific markets to
smooth families’ intertemporal income and consumption (Ghatak
et al., 1996). Thus, uncertainty plays an important role in any
migration decision: in pure economic migration as well as in

conflict-induced displacement. When considering forced displace-
ment, insecurity creates additional costs that modify the expected
outcome, and diminish the relevance of other socioeconomic migra-
tion determinants.
When deciding on displacement, individuals or households com-
pare alternative sites and choose the site promising the largest net
benefits. Thus, early rational-choice models on migration decisions
compared alternative locations by calculating the present value dif-
ference of individual income reduced by migration costs. Migration
is then a result of higher expected net benefits at the reception site
(Sjaastadt, 1962). In the context of a conflict, net returns to displace-
ment are determined by the difference between origin and recep-
tion site benefits and the influence of insecurity and fear of
persecution. Origin site incomes as well as migration costs are
directly biased by threats, direct violence, and disruption (Figure 4.2).
The influence of the economic push and pull factors on the per-
ceived value of displacement is mitigated by the impact of the con-
flict environment, but it is not necessarily eliminated. If economic
factors do not play any role in a violent conflict environment,
Origin site net benefit Reception site net benefit
Expected income
(assets, wages, remittances, etc.)
Conflict environment
(Perception of insecurity, fear) Expected income
(wages, self-employment, asylum and
job probability, remittances,
networks, etc.)
Expected displacement costs

(Opportunity and transport costs.
survival risk, etc.)
Figure 4.2 Perceived value of displacement

a complete population outflow might occur. However, we generally

do not observe this.
Social networks also play an important role in explaining the size
of a population outflow and the duration of stay (Carrington et al.,
1996). According to this approach, migration costs decrease with the
number of migrants already settled in the destination country.
Established networks of previous (economic) migrants can strongly
influence the displacement decision by providing housing, support
in finding employment and other contacts.
Generally, young, economically active people have stronger incen-
tives to migrate, since their discounted net benefits are larger because
of their longer planning horizons (Todaro & Smith, 2006). In con-
flict environments, young adults, male and female, are the most
probable targets for threat, violence, and forced recruitment, which
increases the likelihood for displacement of this group of the
population.
Thus, displacement is driven by both the (non-random) targeting
of violence and by economic considerations. In many cases, the fear
of violence and non-survival triggers non-transitory displacement
by reinforcing the fundamental push factors that drive the rural–
urban migration. The relative importance of violence and economic
factors is a priori unknown and must be addressed by empirical
analysis. The next section investigates the driving forces of poten-
tially conflict-induced displacement at the village level, by consider-
ing net population changes in the Aceh province during a period of
conflict.
4.4 Empirical analysis of displacement in Aceh
4.4.1 Data source and descriptive statistics

The data used in this study are based on the Village Potential Census
PODES (Potensi Desa) of the BPS (Badan Pusat Statistik) Statistics
Indonesia (BPS, 2000, 2003). This census collects information on a
regular basis at the lowest administrative level from all Indonesian
villages and urban neighborhoods.37 The information is based on the
responses of the village heads and includes a wide range of socioeco-
nomic indicators on population, economic activities, infrastructure,
and also on village security. We use data from two subsequent rounds
of PODES (2000 and 2003) that were collected in the fall of 1999
and 2002. We consider only those Acehnese villages (5,211, or

93.1 percent of all) where a match between the two census rounds
has been possible.
The term ‘conflict’ generally has no clear-cut definition. It could
describe severe atrocities with significant casualties and damages as
well as minor clashes. In our dataset, conflict incidence at the village
level is captured by two different questions. In PODES 2003, village
officials have been asked for the first time whether the village has
experienced any conflict during the previous year. Additionally, they
were asked to state the number of casualties (conflict-related deaths
or injuries) that occurred during the last year as a result of conflicts.
The questionnaire did not give further guidance to exactly what
events would define a conflict, and hence misreporting of conflict,
depending on the interpretation of each village head, cannot be
excluded (Barron et al., 2004). Nevertheless, village heads are very
well informed about the presence and extent of civil conflict in their
own village. The information that we can draw from a village census
is much more general than data from questionnaires targeting selec-
tive and small-scale household samples. Based on the data for con-
flict occurrence, the variable Conflict is set to one if the village head
has reported the occurrence of a civil conflict, and zero otherwise. A
potential measurement error might also arise because we observe
conflict incidence in the village only during the last year before the
survey, while our population change variable refers to the period of
the last three years. Nevertheless, our reported conflict incidence is
arguably still capturing a general proneness to conflict, as it is signif-
icantly higher in those regions that are known to have been most
severely affected by the Acehnese conflict. In order to reduce poten-
tial measurement errors, we also code an alternative measure of con-
flict labeled as Conflict cluster, which is set to one if at least 20 percent
of the villages within a subdistrict (kecamatan) reported conflict in
the previous year. In this way, we also capture the effects of conflict
clusters within the neighborhood while excluding villages with iso-
lated (and potentially minor) conflicts. Additionally, we include dis-
trict controls in our regressions to also capture part of the differences
in conflict intensity between districts. According to these defini-
tions, between 1999 and 2002, around a quarter of the sample vil-
lages have reported the presence of a conflict, while more than one
third of the villages belonged to a conflict cluster (Table 4.1).38
Table 4.1 Conflict and population change in the Aceh sample, 1999–2002
Total No Conflict Conflict

conflict cluster
Total pop. in 1999 3,684,021 2,684,672 999,349 1,381,138

Abs. pop. change –651 25,208 –25,859 –38,068
Rel. pop. change (%) –0.0002 0.94 –2.59 –2.70
Sample villages (N) 5,211 3,972 1,239 1,807
(76.2%) (23.8%) (34.7%)
Notes: Own calculations based on PODES 2000 and 2003. The sample covers around
93.1 percent of all Acehnese villages.
We define our main dependent variable as net Population change

between the two survey rounds in 1999 and 2002, which measures
the absolute change in village population in hundreds of inhabit-
ants. Table 4.1 shows that villages which were involved in a conflict
or were located in a conflict cluster during the year preceding the
second survey round (around one third of the villages), lost on
average around 2.7 percent of their population between 1999 and
2002.39
Based on our sample, we estimate for the period of 1999–2002 a
net population outflow from the Aceh province of about 176,600
persons. This figure is based on an average annual population growth
rate of 1.46 percent (documented for Aceh by BPS (n.d.) for 1990–
2000), which implies an increase in population of about 4.4 percent
over these three years. This would amount to an increase of 163,700
persons in our sample. Instead, total population in the Aceh sample
decreased by about 650 persons (see Table 4.1). These numbers indi-
cate a net outward migration of about 164,400 persons in our sample
villages (93.1 percent of all villages), and are in accordance with other
estimates (Ramly, 2005).
4.4.2 Empirical strategy

In our empirical analysis we focus on the determinants of net popu-
lation change in order to quantify the effects of conflict as well as
the effects of other socioeconomic variables. We run the regressions
both for all villages and separately for villages with and without con-
flict during the last year. Additionally, by applying quantile regres-
sions we distinguish between the push and pull factors of migration.
As a first descriptive step, we relate the reported conflict occurrence

to the same set of explanatory factors as in our subsequent regres-
sions of population change.
4.4.2.1 Estimation models

In our conflict regressions we model the unobserved levels of con-
flict intensity Ci* in village i as a latent variable, dependent on the
vector of explanatory variables Xi, the unknown vector of parameters
b, and the normally distributed error term di. Conflict occurrence Ci
is our observed binary variable which is set to one if the village has
been involved in a conflict during the previous period,
Ci 1 if ( Xi' i ! 0), and zero otherwise, (4.1)
which is estimated by a probit model. Based on the resulting coeffi-

cients we calculate the marginal effects of each explanatory variable
on the probability that a conflict in a village has occurred, evaluated
at the sample mean.
In the second set of regressions, the dependent variable is the
population change DPopi that occurred between 1999 and 2002 in
village i, which is explained by a set of conflict variables and other
migration-related variables Xi:
'Popi oCi Xi' i . (4.2)
We estimate the vector of the coefficients g0 and g by two alterna-

tive procedures. First, we report estimates from the standard OLS
technique, both for all villages and for the subsamples of conflict
and non-conflict villages. Second, we apply quantile regressions that
minimize the sum of the absolute residuals where values above
(below) a given quantile receive weights that are proportional
(inversely proportional) to the quantile that is to be estimated. We
report estimates from quantile regressions around the first (Q = 0.25),
second (Q = 0.50), and third (Q = 0.75) quartile. The great advantage
of this technique is that it enables us to distinguish between the
push and pull factors of migration. Considering the lowest and high-
est quartiles is especially useful as these two quartiles roughly coin-
cide with villages with considerable population outflow and inflow,
respectively. For instance, when running the regression around the
lowest (Q = 0.25) quartile, our parameter estimates reflect the push

factors of migration to a larger extent since observations on villages
with a larger than predicted net population decrease (that is, villages
with a population outflow considerably larger than expected) receive
a threefold weight.40 In the probit and OLS specifications we cluster
standard errors on the subdistrict level; by doing this we allow for the
correlation of error terms for villages within the same subdistrict.
4.4.2.2 Main control variables

Basically, changes in village population are either caused by differ-
ences in the fertility–mortality ratio or by (forced) migration move-
ments.41 In all sets of regressions, we apply – in addition to the
conflict variables – the same set of socioeconomic, political and geo-
graphic controls that could be related to one of these two channels
or to both.42 By using explanatory variables mainly from PODES
2000, we are able to reduce problems of reverse causality to a consid-
erable extent.
Besides investigating the role of conflict for forced migration flows,
this analysis will test whether a strong rural–urban migration pat-
tern is still present in times of major conflict. As our main control for
economic opportunities we include the indicator variable Urban,
which is based on the official classification of villages and neighbor-
hoods by the Indonesian state, and proxies the availability of public
services and the structure of economic activity in the village. The
urban dummy is strongly correlated with other measures of economic
structure (for example, the share of families in the village living pri-
marily from agriculture, or the share of village land that is in non-
agricultural use).43 Additionally, we also include geographic Altitude
(in thousand meters above sea level) that might capture economic
incentives to emigrate similarly to the urban dummy. Furthermore,
we also control for the political remoteness of a village by the Distance
to district (kabupaten) office (in hundreds of kilometers).
We capture the effect of village-level poverty by the share of Poor
families in a village. This variable measures the share of village house-
holds that are considered to be poor according to a set of welfare
criteria established by the Indonesian National Family Planning
Agency (BKKBN).44 This agency categorizes Indonesian households
into five classes of welfare status: pre-prosperous families KPS
(Keluarga Pra-Sejahtera), and families of prosperity status KS I to IV
(Keluarga Sejahtera) (Perdana & Maxwell, 2004). Poor families are

defined by the census as households belonging to the two lowest
categories KPS and KS I.45 By this definition, on average around
55 percent of village populations are considered to be poor. Variations
in potential fertility are controlled for by the variable Fertile couples,
which measures the number of couples of reproductive age in the
village, and is also based on information collected for the BKKBN.
We include the Transport station variable as a rough proxy for costs
of migration by indicating the presence of a bus or train station, air-
port, or seaport in the village. The variable has been set to one also
in neighborhoods of larger cities where a station in another part of
the city is available. The security environment is controlled for by
the variable Police presence which is set to one if the nearest police
station is easy to reach (which applies to 65.8 percent of the villages),
and zero otherwise. The conflict mitigation role of the police, as
opposed to the influence of the military and paramilitary groups,
has been documented for several Indonesian conflicts (see Barron
et al., 2004). We expect that population outflow is larger if the near-
est police station is far away (especially, in the face of a conflict).
Additionally, we also include controls that should capture the
migration patterns specific to the Acehnese conflict. For instance,
we know that in the three years of conflict ethnic-Javanese have
been much more likely to leave Aceh altogether (see above). To cap-
ture this pattern of migration, we include in our regressions the
Share of Javanese as a further control. The variable is based on infor-
mation from the 2000 Census of Indonesia, and is only available at
the subdistrict (kecamatan) level.46 We also include a set of district
indicators for those Acehnese regions which we know were the most
affected by the conflict: districts with high conflict incidence
(Central, North, West, South and Southwest Aceh) and also neigh-
boring districts that experienced population inflows. The remain-
ing nine, less affected districts in Northern and Southern parts of
Aceh serve as a control group. Additionally, we include dummies for
the three largest cities, Banda Aceh, Langsa and Lhokseumawe, of
which the latter was the hardest-hit area. These large centers may
experience very different migration patterns from the rest of the
country. Finally, we control for nonlinear effects of the village pop-
ulation size by including a fourth grade polynomial of population
size in all regressions.
4.4.3 Estimation results

Table 4.2 reports the estimates of two probit models explaining
conflict occurrences that differ only with respect to the conflict
variable; the dependent variable in column (1) is reported conflict,
in column (2) the presence of a conflict cluster in the subdistrict
(kecamatan). The likelihood that a conflict arises might be influenced
Table 4.2 Probit estimation: conflict incidence
Dependent variable (1) Conflict (2) Conflict cluster
Marg. eff. t-stat. Marg. eff. t-stat. Mean
Urban –0.020 –0.57 –0.027 –0.49 0.080

Altitude –0.029 0.36 –0.203 –1.09 0.178
Distance (dist office) 0.001 0.03 –0.062 –0.66 0.078
Poor families –0.015 –0.21 0.015 0.11 0.546
Fertile couples 0.019 1.11 0.042 1.29 0.986
Transport station –0.079** –2.21 –0.133*** –2.64 0.061
Police present 0.031 1.02 0.042 0.83 0.342
Share of Javanese 0.093 0.52 0.067 0.25 0.087
Central Aceh 0.291* 1.89 0.605*** 2.63 0.037
West Aceh 0.044 0.44 0.183 0.72 0.051
Nagan Raya 0.009 0.09 0.287 1.09 0.040
Southwest Aceh 0.210*** 3.10 0.546** 2.46 0.024
South Aceh 0.434*** 3.61 0.454*** 2.61 0.046
North Aceh 0.729*** 8.15 0.735*** 5.70 0.157
East Aceh 0.513*** 4.68 0.689*** 4.30 0.089
Aceh Tamiang –0.043 –0.44 0.040
Langsa (city) 0.274* 1.83 0.570** 2.09 0.010
Lhokseumawe (city) 0.500*** 2.73 0.563* 1.95 0.013
Banda Aceh (city) 0.205 1.39 0.354* 1.77 0.016
Population
polynomial Yes Yes
Number of
observations 5211 5211
Pseudo R2 0.324 0.342
Observed/predicted
conflict 0.238 0.172 0.347 0.304
Notes: The marginal effects are based on a probit regression, and are evaluated at the
sample mean. The regressions include a fourth order polynomial in population size and
a constant. The reported values of t-statistics are based on robust standard errors that
are clustered at subdistrict level (202 clusters). (*), (**), (***) denote values significant at
levels of 10, 5, and 1 percent.
by political, civil, or socioeconomic variables. However, in our case

both conflict variables are mainly explained by geographic loca-
tion. We do not find robust evidence for the role of ethnicity in
the Acehnese conflict occurrence either, as we find that the
share of ethnic-Javanese (the most populous and economically
most important minority) in a subdistrict is not related to conflict
occurrence.
The distribution of observed conflict over the districts supports
our trust in the reliability of the conflict variables: reported conflict
occurrence is much more likely in those districts that were actually
mostly affected in this period (North, East, Central, Southwest and
South Aceh, and the large cities).
According to the previous theoretical displacement model, we
jointly address the influence of socioeconomic determinants and
conflict variables for explaining net population change. We use both
OLS and quantile regressions as two alternative estimation tech-
niques, reporting the respective results in Tables 4.3 to 4.5.
In Table 4.3, the two specifications differ only with respect to the
definition of the conflict variable as measuring individual Conflict
or the presence of a Conflict cluster. Surprisingly, we do not find that
reported conflict has a robustly significant negative effect on net
population change. However, the conflict cluster variable has a sig-
nificantly negative effect on net population change in most of our
specifications. This may be due to the fact that the clustering of con-
flicts better reflects the relevant push factors for forced migration
over the whole three-year period than the reported conflict occur-
rence in the last year in any given village. From column (2) in
Table 4.3, we see that the presence of a conflict cluster in a subdistrict
(20 percent or more of the villages reporting conflict) reduced village
population on average by about 25 persons ceteris paribus. From the
geographic distribution of migration flows, we also see that the sin-
gle most affected district, Central Aceh, experienced also the highest
net population outflow of, ceteris paribus, almost 200 persons per vil-
lage on average.
The largest forced migration flows (within the province) have been
going to the Northeastern coast of Aceh: the city of Langsa, and the
neighbouring Aceh Tamiang district (c.f. Figure 4.1). The data also
supports the evidence that the Javanese population has been more
likely to migrate: net outward migration from a village increased by
Table 4.3 Determinants of change in village population (OLS)
Dependent variable Change in village population (in .00)
Coeff. (1) t-stat Coeff. (2) t-stat.
Conflict –0.120 –0.80

Conflict cluster –0.248* –1.79
Urban 0.902** 2.53 0.902** 2.53
Altitude –0.303 –0.86 –0.324 –0.93
Distance (distr office) 0.039 0.24 0.027 0.17
Poor families –0.252 –1.46 –0.250 –1.45
Fertile couples 0.652*** 3.11 0.658*** 3.17
Transport station 0.943** 1.97 0.925* 1.93
Police present 0.249*** 2.77 0.245*** 2.74
Share of Javanese –1.272** –2.15 –1.223** –2.10
Central Aceh –1.966** –2.55 –1.879** –2.50
West Aceh 0.283 0.88 0.308 0.98
Nagan Raya –0.403 –1.15 –0.369 –1.13
Southwest Aceh 0.222 0.63 0.331 0.92
South Aceh –0.172 –0.68 –0.115 –0.43
North Aceh 0.253* 1.81 0.354** 2.46
East Aceh 0.071 0.35 0.185 0.83
Aceh Tamiang 1.223*** 2.81 1.180*** 2.72
Langsa (city) 2.752** 2.53 2.838** 2.48
Lhokseumawe (city) 0.822 0.69 0.904 0.77
Banda Aceh (city) 0.382 0.41 0.423 0.45
Population polynomial Yes Yes
Number of observations 5,211 5,211

R2 0.133 0.134
Notes: Regressions are performed by OLS, and include a constant and fourth order poly-
nomial of population size. The reported values of t-statistics are based in robust stand-
ard errors that are clustered on subdistrict level (202 clusters). (*), (**), (***) denote
values significant at levels of 10, 5, and 1 percent.
about 12 persons on average with a 10 percentage point increase in

the share of Javanese population within the subdistrict.
By applying a quantile regression technique, the overall effect of
conflict-induced displacement provided by the OLS estimation
results can be further decomposed. The two alternative model speci-
fications (A and B) of the quantile estimations (Table 4.4) differ only
in the inclusion of district dummies: whereas in model B the conflict
cluster variable reduces both population outflow and inflow, the
Table 4.4 Quantile regressions of population change
Dependent Test
variable Q25 =
Change in village population (in .00) Q75
Q25 t-stat. Q50 t-stat. Q75 t-stat. p-val.
Model A
Conflict cluster –0.008 –0.58 –0.028** –2.31 –0.075** –2.37 0.030
Urban 0.030 0.61 0.059 1.52 0.043 0.35 0.920
Altitude –0.179*** –2.99 –0.091*** –2.88 –0.155** –2.14 0.751
Distance (dist
office) –0.024 –0.87 0.078*** 3.43 0.246*** 6.18 0.000
Poor families –0.003 –0.13 0.069*** 4.45 0.084* 1.79 0.047
Fertile couples 0.273*** 3.19 0.136*** 3.62 0.397*** 5.40 0.173
Transport
station 0.110 0.79 0.347*** 2.64 0.549*** 2.58 0.043
Police present 0.039*** 3.03 0.029*** 3.36 0.054** 2.22 0.512
Share of
Javanese –0.526*** –2.67 –0.164*** –4.16 –0.123 –1.31 0.037
Central Aceh –3.631*** –5.81 –0.683*** –2.35 0.460*** 2.57 0.000
West Aceh –0.171*** –4.31 0.026 0.94 0.221*** 2.68 0.000
Nagan Raya –0.177** –2.10 –0.052*** –2.75 –0.100* –1.68 0.408
Southwest
Aceh –0.242*** –3.83 –0.065 –0.74 0.071 0.86 0.000
South Aceh –0.104** –2.20 0.046 1.10 0.075 1.28 0.006
North Aceh 0.001 0.04 0.001 0.06 0.003 0.09 0.946
East Aceh –0.099 –1.07 0.011 0.39 0.042 0.61 0.157
Aceh Tamiang 0.218** 2.37 0.172*** 3.31 0.301** 2.45 0.540
Langsa (city) 0.087 0.20 1.142 1.27 3.141*** 2.86 0.001
Lhokseumawe
(city) –0.061 –0.41 –0.394*** –2.90 –0.649** –2.22 0.052
Banda Aceh
(city) 0.641* 1.86 0.428 1.50 1.446*** 2.89 0.112
Population poly Yes Yes Yes
Pseudo R2 0.076 0.020 0.085
Model B
Conflict cluster –0.040** –2.48 –0.043*** –5.10 –0.070*** –2.97 0.231
Urban 0.072 1.39 0.078** 2.15 0.233 1.61 0.235
Further
controls Yes Yes Yes
District
dummies No No No
Pseudo R2 0.043 0.012 0.069
Quartile range of population change

Q0 − Q25 Q25 − Q50 Q50 − Q75 Q75 − Q100
[−3,747;−6] [−5;9] [10;37] [38;2,116]
Notes: Observations N = 5,211. Reported values of t-statistics are based on bootstrap

standard errors (with 1000 replications). All regressions include a constant and a fourth
order polynomial in population size, the coefficients on which are not reported.
Model B includes the same set of explanatory variables as Model A except for the dis-
trict dummies; full results are available on request. (*), (**), (***) denote values signifi-
cant at levels of 10, 5, and 1 percent.
Table 4.5 Determinants of population change by conflict occurrence
Dependent variable Change in population in villages Test
(1) (1)
(1) w/o conflict (2) with conflict (2) (2)
Coeff. t-stat. Coeff. t-stat. p-val. p-val.
Urban 0.473 1.53 1.756** 2.08 0.070

Altitude –0.618* –1.78 0.822 0.76 0.101
Distance (dist office) –0.095 –0.63 0.686 1.21 0.086
Poor families –0.297* –1.69 –0.250 –0.54 0.461
Fertile couples 0.721** 2.47 0.652** 2.08 0.432
Transport station 0.863* 1.69 0.377 0.46 0.302
Police present 0.225** 2.52 0.417** 2.36 0.150
Share of Javanese –0.318 –0.46 –2.023** –2.32 0.052
Central Aceh –1.363* –1.95 –3.893** –2.56 0.053
West Aceh 0.255 0.82 –0.026 –0.09 0.241
Nagan Raya –0.155 –0.47 –3.211*** –6.29 0.000
Southwest Aceh 0.001 0.00 1.146 1.09 0.116
South Aceh –0.370 –1.10 0.297 1.21 0.065
North Aceh 0.094 0.72 0.422* 1.85 0.103
East Aceh 0.219 0.97 –0.635 –1.58 0.036
Aceh Tamiang 0.884** 2.09 0.597 0.51 0.401
Langsa (city) 1.498 1.49 6.665*** 2.73 0.043
Lhokseumawe (city) –0.065 –0.08 3.273* 1.75 0.039
Banda Aceh (city) 0.524 0.49 0.540 0.51 0.495
Population
polynomial Yes Yes
Number of
observations 3,972 1,239
R2 0.101 0.242
Notes: Regressions are performed by OLS, and include a constant and a fourth order
polynomial of population size, the coefficients on which are not reported. The reported
values of t-statistics are based on robust standard errors that are clustered on subdistrict
level (188/116 clusters). (*), (**), (***) denote values significant at levels of 10, 5, and
1 percent. The last two columns present p-values of pairwise hypotheses tests on the H0
hypothesis of inequality of coefficients in (1) and (2).
effect of conflict in outflow villages is directly captured through the

district dummies in model A.
This may be due to clustering of conflict affected subdistricts. The
impact of conflict on population outflow (near to the first quantile,
Q = 0.25) is smaller than its impact on inflow (near to the Q = 0.75
quantile). Furthermore, the quantile regressions also draw a more
differentiated picture of the geographic distributions of population
movements. In the districts of Central Aceh, West Aceh, and Nagan

Raya, we observe both villages with larger population outflows and
villages with larger inflows; this may reflect the presence of inner-
district migration.47 Beyond that, stronger migration flows can be
registered towards the coastal cities of Langsa and Banda Aceh, while
Lhokseumawe experienced less population inflow. This may be due
to the fact that neighborhoods in the industrial area of Lhokseumawe,
home to many Indonesian, foreign and local businesses, have been
particularly affected by illegal GAM activities such as the ‘village tax’
(Schulze, 2004).
The quantile regressions show that the decrease in village popu-
lation was larger in subdistricts with a relatively high share of
ethnic-Javanese, while ethnic composition does not explain differ-
ences in net village population inflow. Thus, although the pres-
ence of ethnic-Javanese is not related to the outbreak of conflict
(Table 4.2), it explains strongly the outward migration flows. These
results are reinforced when distinguishing between conflict and
non- conflict villages (see Table 4.5): Population is reduced with
the share of ethnic-Javanese, but only in conflict-affected villages.
Thus, corroborated by other sources (for example, Schulze, 2004),
we can state that although GAM officially denied targeting ethnic-
Javanese, they were nevertheless intimidated into leaving their
homes.
Beyond the evidence of conflict-induced displacement, we also
find a significant role for other socioeconomic determinants of
aggregate displacement movements. The Urban variable, reflecting
the concentration of economic activity, shows a clear pattern of
explanation: urban villages experience a larger increase in net popu-
lation (Table 4.3). In the quantile regressions (Table 4.4) the Urban
dummy loses significance, but geographic Altitude is significantly
related to larger decreases as well as smaller increases in population.
This result indicates that migration runs from rural and mountain-
ous central areas to the more urbanized Northern and Southern
coastal areas of the Aceh province. These two measures of economic
remoteness/urbanization show strong evidence for a rural–urban
migration pattern. The same pattern can be retraced in the case of
inward migration to the large cities, particularly Langsa. When dis-
tinguishing between villages that reported conflict and those that
did not (Table 4.5), it is apparent that the mere presence of conflict
in large cities or urban regions does not hinder an ongoing urbaniza-

tion process.
Our control for political remoteness, the Distance to the district
office turns out to be insignificant in most specifications, although it
does lead to larger population inflows among the inflow villages in
the quantile regressions. The indicator of availability of any trans-
port station in the community, which we use as a proxy for migra-
tion costs, is mostly positively related to population change. However,
it is most likely that this variable also captures pull factors of rural–
urban migration. Based on the evidence from quantile regressions
(Table 4/4), direct availability of transport opportunities seems to
play only a minor role in displacement decisions in the case of pop-
ulation outflow, but might act as a significant pull factor for popula-
tion inflows. However, since actual migration costs include a broad
range of expenses, they may not be well captured by a transport sta-
tion dummy.
Displacements are related to a lack of police presence and thus we
can infer that they are also due to institutional weaknesses of the
state. The presence of a police station has a significant and overall
robust positive effect on population change, reducing outflow and
increasing inflow (Table 4.4) in both conflict and non-conflict vil-
lages (Table 4.5). Thus, public institutions still play a significant role
in times of large population displacement.
The effect of poverty on net population change in a village acts
generally through several channels. The share of poor families in a
village has no robustly negative effect on population change in OLS
regressions (Table 4.3), although the negative effect becomes signifi-
cant in villages that did not report conflict (Table 4.5).
In quantile regressions the share of the poor in a village even
increases population inflows (Table 4.4). Thus, in the face of conflict,
poverty as a push factor seems to be less relevant; the results confirm
the theory of a ‘migration hump’ that is mainly produced by unavail-
able resources (and information) for migration of the lower income
groups. Furthermore, the increased pull effect of villages with a high
share of poor families may also be indicative of a policy bias, since
the composition of this variable is based on the registration for social
support programs for the poor. Poor families tend to migrate to des-
tination sites where governmental support is more likely, and thus,
registered poverty may also be related to a larger inward migration.
Our control variable for fertility, the number of fertile age couples,
is positive and significant in all regressions. Obviously, population
change is also driven by population growth. In all regressions we
also include a fourth grade polynomial for population size that indi-
cates a highly significant nonlinear influence of village size on pop-
ulation changes.
Finally, these results demonstrate that internal displacements in
conflict situations, at least in the context of Aceh, are not unidimen-
sionally caused by conflict variables. The factors that drive common
rural–urban migration movements are still relevant in a conflict sit-
uation. Thus, traditional push and pull factors are not suspended in
times of conflict, though it is obvious that large displacements are
primarily initiated by the conflict. However, without other eco-
nomic, political, social and institutional factors at work, such con-
flict-induced population movements would certainly be different in
their numbers and magnitudes.
The intention of this study is to contribute to an improved under-

standing of the determinants of civil conflicts and their related
forced migration movements. Our empirical analysis is based on vil-
lage-level data for the province of Aceh, collected by a regular village
census taken throughout Indonesia. Since we can only observe net
changes in village population, we are unable to model individual or
household migration behavior. Even though we cannot learn from
our analysis who migrates in the face of a conflict, we are able to
identify which villages are more prone to population outflows and
inflows. This meso-level approach is able to indicate the most rele-
vant push and pull factors at the village level. This is an innovation
in forced migration research.
In Aceh, conflict occurrence was very strongly related to the geog-
raphy of conflict clusters. The presence of such conflict clusters
contributed significantly to the displacements of the Acehnese pop-
ulation during our period of observation. Relatively large displace-
ment flows could be observed from the most conflict-affected regions
in central Aceh towards the more stable and prosperous urban areas
at the coast. However, there were also large population rearrange-
ments between villages within the most conflict-affected districts.
Although often categorized as a political conflict, the intimidation

of ethnic-Javanese had a substantial effect on outward migration,
particularly from conflict-affected villages with a relatively large
Javanese population. Furthermore, security, specifically police pres-
ence, has had a significant influence on migration flows.
Finally, while we know that the clashes between the GAM, militias
and military forces were the major driving forces behind the large
displacement of the Acehnese population, we also find convincing
evidence for the relevance of conventional socioeconomic migration
determinants. These results indicate that economic opportunities
and rural–urban migration incentives play an important role even in
the context of civil conflict. Traditional migration considerations are
still effective, even during periods of severe clashes. Potentially, civil
conflict might even work as an accelerator of an ongoing urbaniza-
tion process.
5
The Political Economy of
Refugee Migration
Abstract48
This chapter examines the driving forces of the magnitude, composi-

tion and duration of refugee movements caused by conflict and per-
secution. The decision to seek temporary or permanent refuge in the
region of origin or in a more distant asylum destination is based on
intertemporal optimization. We find that asylum seeking in Western
countries could rather be a phenomenon of comparatively less perse-
cuted people. In an attempt to reduce their respective asylum
burdens, Western countries and host countries in the region of origin
are likely to end up in a race to the bottom of restrictive asylum poli-
cies. As an alternative, this study shows that proactive refugee-related
aid transfers are, under certain circumstances, an effective instru-
ment to relieve Western countries from asylum pressure.
5.1 Introduction
Refugee situations are one of the most neglected or even forgotten

human sufferings. An involuntarily displaced mass of people often
lives for a significant period of time under appalling conditions in a
state of limbo, mostly in a country that is not their own. Their
dilemma is mostly multiple: going back to their home country is
dangerous, primarily because of ongoing insecurity and persecution;
in addition, they are often not allowed to settle freely and to be
economically self-reliant in the first asylum country, and finally,
they are often unable to seek asylum in a Western country, either
67
because of a lack of resources or the low prospects of receiving

asylum status and permanent residence rights. As a result, in 2005,
there were more than 9.2 million cross-border refugees, of which
5.5 million individuals were caught in a protracted or long-lasting
situation (UNHCR, 2006c). The number, as well as the duration, of
these situations has vastly increased during the past decade. In 1993,
27 protracted refugee situations existed and it is estimated that the
average duration of major refugee situations will have increased from
nine years in 1993 to an average duration of 17 years by the end of
2003 (UNHCR, 2004c, 2006c).
It is generally acknowledged that the major causes of long-lasting
refugee situations are political in nature. The United Nations High
Commissioner for Refugees (UNHCR) states that
Protracted refugee situations stem from political impasses. They

are not inevitable, but are rather the result of political action and
inaction, both in the country of origin (the persecution or vio-
lence that led to flight) and in the country of asylum. They endure
because of ongoing problems in the countries of origin, and
stagnate and become protracted as a result of responses to
refugee inflows, typically involving restrictions on refugee move-
ment and employment possibilities, and confinement to camps.
(UNHCR, 2004c)
This chapter provides the theoretical background for an analysis of

refugee migration and the role of alternative asylum policies in
explaining this phenomenon. A review of the economic migration
literature reveals a severe lack of research dealing with refugee move-
ments. Most of the early migration literature is based upon static
rural–urban or international wage differentials (Sjaastadt, 1962;
Todaro, 1969; Harris & Todaro, 1970). According to this, regional or
international wage differentials are the driving incentives for volun-
tary and permanent migration. This does not reflect the reality of
refugee movements. By definition, refugees do not flee for purely
economic reasons, at least initially; and, refugee migration is not
necessarily a one-way movement, and thus, not a permanent phe-
nomenon. For realizing the latter, temporary migration models
introduce a special loyalty held by individuals for their home coun-
try. Voluntarily return migration is then induced by the migrant’s
special preferences for consumption at home being higher than
The Political Economy of Refugee Migration 69
abroad (Djajic & Milbourne, 1988; Dustmann, 2003; Brücker &

Schröder, 2006).
However, this special loyalty for the home country is ‘reduced’ for
individuals that are victims of conflict and persecution. Refugees,
unlike economic migrants, leave their home country even if the eco-
nomic situation abroad is inferior. But, even if consumption at home
might be foiled by insecurity and danger, there is some empirical
evidence that refugees living in encampments under very poor con-
ditions just across the border often return home despite ongoing
repressions and substantial risks of being affected by civil strife
(Rogge, 1994). These effects are also not captured by existing migra-
tion models.
Furthermore, refugees also have the option to seek asylum in a
Western country that promises better economic perspectives,
although the incentive to move on to a Western country is cushioned
by the uncertainty about the asylum application’s outcome. Thus,
asylum policies of Western countries, as well as that of the conflict-
neighboring first asylum countries, need to be taken into account
when analyzing the migration decision of persecuted individuals.
Consequently, a satisfactory model for understanding the mecha-
nisms of refugee movements and their protraction must contain fea-
tures of (i) temporary and permanent migration, (ii) economic and
non-economic push and pull factors, (iii) uncertainty about the
success of an asylum-seeking process, and (iv) a non-static frame-
work for analyzing intertemporal migration patterns. Therefore, a
dynamic model is required to approach alternative migration options
in a long-term time horizon.
The innovation of this study is to model the simultaneous decision
on seeking asylum in a Western country, moving temporarily or per-
manently across the border to a neighboring first asylum country, or
staying at home. This framework will enable the analysis of some
alternative policies available to both neighboring first asylum and
Western destination countries that we indentify as decisive in
explaining the pattern and duration of international refugee migra-
tion movements.
The remainder of this chapter is structured as follows. The next
section outlines the migration model and derives some core
propositions. Section 5.3 discusses the implications of reactive and
proactive counter-asylum policies for the refugee migration pattern.
The fourth section concludes.
5.2 The model
This basic analytical framework models the intertemporal migration

decision of N heterogeneous individuals living in a country of origin
O before a conflict breaks out. Heterogeneity of individuals refers to
the extent each individual is affected by this conflict, that is individ-
ual persecution level p in the home country O is uniformly distrib-
uted with density N between [0,1] with p equal to unity reflects full
security. According to the respective persecution level, the individu-
als decide upon several migration options to realize for the rest of
their lifetime Ti, which is normalized to unity.49 The three migration
options are: first, staying in the conflict affected country of origin O
despite the threat of persecution, second, emigrating (with negligible
migration costs) for the period 0 < ti # 1 = Ti, that is temporarily or
permanently, to a neighboring first asylum country S where security
is guaranteed but where confinement to a camp or limitations in
self-reliance impose worse economic conditions, or third, the oppor-
tunity to migrate (with migration costs M W) to a Western industrial-
ized country W with better economic prospects (and security) to
seek asylum, but with the risk of being rejected and deported to the
home country O.50
Conditions are such that each individual could receive an income
level of yO at home, yS in the first asylum country, and yW in the
Western industrialized country, assuming that yW is by far larger
than the income levels of the other two developing countries, yO
and yS .51 However, income generation is hindered by the first asylum
country, that is the level of economic self-reliance s that a refugee
enjoys is a policy parameter with 0 < s < 1, indicating the proportion
of the standard income level yS that is actually available for refugees.
The available income to the refugee in the first asylum country is
then given by syS .52
The utility u(⭈) an individual derives from consumption either in
the country of origin (cO S
i ), the (neighboring) first asylum country (c i ),
W
or in the Western country (ci ), respectively, are given by:
␣
uO (c O
i ) p1i ␣cO
i , (5.1)
␣
u S (c Si ) c Si , (5.2)
␣
uW (c iW ) c iW . (5.3)
Generally, the migration decision and its realization takes place at

t = 0 for all N individuals.53
5.2.1 The cross-border option

For the first migration option, moving to the (neighboring) first asy-
lum country S, the individual must choose the duration for the stay,
ti, and the consumption levels abroad and at home (if migration is
temporary). Thus, each individual maximizes its lifetime utility ViS
with respect to duration and consumption, that is
ViS ␶ i[c iS ]␣ (1 ␶ i )p1-␣ O ␣

i [c i ] (5.4)
subject to the individual’s budget constraint:
␶ i sy S (1 ␶ i )y O ␶ ic Si (1 ␶ i )c O
i t 0. (5.5)
The first-order conditions of the corresponding Lagrange function

Li (with l as Lagrangian parameter) are:
wL !
␣ p1i ␣[cO i ]
␣ 1
␭ 0
wc i
O
(5.6)
wL !
␣[c Si ]␣ 1 ␭ 0
wci
S (5.7)
wL !
[c Si ]␣ p1i ␣[c O ␣
i ] ␭( sy ␥ c i c i ) 0
S O O S
w␭ (5.8)
wL !
␶ i sy S (1 ␶ i )y O ␶ ic Si (1 ␶ i )c O
i 0
w␭ (5.9)
The Lagrangian parameter l, that is the shadow value of income, is

also time-invariant. Eqs (5.6) and (5.7) imply that
pi c Si c Oi . (5.10)
This condition indicates that consumption at home is a linear

function of consumption abroad; that is the marginal utility of con-
sumption at home and abroad must be equal. This condition (5.10)
generates together with eqs (5.8) and (5.6) explicit consumption lev-
els for at home and abroad:
␣( yO sy S ) c Oi
cSi .
(1 p i )(1 ␣) pi (5.11)
Thus, for temporary migrants, consumption at home and abroad is

related to the bilateral income differential between the country of ori-
gin and the first asylum country; consumption abroad exceeds their
disposable income abroad (that is negative savings). Thus, in order to
not exceed the intertemporal budget, the refugee returns home.
Substituting eqs (5.10) and (5.11) into eq. (5.9), the optimal duration of
the stay abroad is determined by the other model parameters:
y O (1 ␣)
°0, if p ! O .
° y ␣sy S
°° ␣p i y O (1 ␣)
␶ i (p i ) ® S , (5.12)
° (p i 1) sy y O
° sy S ␣y O
°1, if p .
°̄ (1 ␣ )sy S
w␶ i ␣ w2␶ i 2␣
with wpi
0, and 0.
( p 1)2 wpi2 ( p 1)3
Figure 5.1 displays the duration function ti(pi) of the cross-the-
border option. The higher the persecution level (that is the lower pi),
the longer the refugee stays abroad. For individuals with a persecu-
tion level pi < p, the optimal value of ti is larger unity (that is the total
–
lifetime), which means that the individual migrates permanently
into the first asylum country; less persecuted individuals with pi > p –
stay at home for their lifetime. All individuals with medium persecu-
tion levels stay only temporarily in the first asylum country:
sy S ␣ y O y O (1 ␣ )
p pitemp p . (5.13)
(1 ␣ )sy S y O ␣ sy S
Proposition 5.1: The following constraints describe the cross-the-

border migration option:
(i) There are individuals that do not leave the home country, or only leave
temporarily, if there exists a negative income differential, that is, 0 #
p– # 1, if 0 # syS # yO.
(ii) There are individuals that leave the home country permanently, if the
self-reliance level in the first asylum country is not too strict, that is,
p $ 0, if s t ␣yS .
O
– y
(iii) There is a total permanent outflow of all N individuals, if the eco-
nomic situation abroad is definitely superior to the situation at home,
that is, p , p– > 1, if s ! y S .
O
– y
τi (pi )
0 p p 1 pi
Permanent Temporary Staying

in S in S in O
Figure 5.1 Cross-border migration duration of persecuted individuals
We assume the parameter values to be such that cases (i) and (ii)
apply for the subsequent analysis.
Finally, the optimized utility function with respect to the indi-
vidual persecution level is then given by substituting eqs (5.12),
(5.11), and (5.10) into (5.4), taking into account the constraints
of (5.13):
[ sy S ]␣ ,
° 0 d pi d p ,
° ª y O pi sy S º °
® (1 ␣ ) Z « O » , if ® p pi p ,
S
V ( pi ) (5.14)
¬ y sy
S
° ¼ °
° ¯ p d pi d 1,
¯ pi1␣[ y O ]␣ ,
0 0,
° 0 d pi d p ,
wV S
° (1␣ )(␣y sy (1 p(␣ 1)))
O S °
with ®Z ( y O sy S )(1 p )
! 0, if ® p pi p ,
wp ° °
°¯ (1 ␣ ) pi␣[ y O ]␣ ! 0, ¯ p d pi d 1,
0 0,
° 0 d pi d p,
wV
2 S
° ␣(1␣ )((1␣ )y sy ( p(1␣ )2))
O S °
and ®Z ! 0, if ® p pi p ,
wp 2 ° ( y O sy S )(1 p )2 °
° 1␣ O ␣ ¯ p d pi d 1,
¯ ␣(␣ 1) pi [ y ] ! 0,
␣
ª ␣( y O sy S ) º
and Z { « » ! 0.
¬ (1 pi )(1 ␣ ) ¼
5.2.2 The asylum-seeking option

As an alternative to the cross-border option, the refugee may move to
a third country to seek asylum. A generally preferred destination for
that movement is a Western country W that promises better eco-
nomic prospects. However, low asylum recognition rates that signal
low chances of being accepted may discourage asylum migration.54
Uncertainty about the asylum admission process exposes potential
asylum seekers to the risk of being deported to the country of ori-
gin.55 In this model, an asylum seeker is accepted with probability r,
which is the asylum recognition rate, or with probability (1 – r) she
or he is rejected and deported back to the home country where the
individual stays for the rest of lifetime Ti = 1.56 To realize this migra-
tion option, the asylum seeker must invest migration costs M W. These
costs must not be larger than the alternative lifetime income at home,
that is M W # yO, since in the case of deportation, the intertemporal
budget constraint must still hold.57 The recognition rate is the policy
parameter that indicates whether the Western asylum country is
rather liberal (r is large) or restrictive (r is small) in its asylum policy.
With respect to this asylum option, the expected utility for a (risk
neutral) individual is given by:
␣ ␣
ViW ric iW (1 ri )p1i ␣c 0i (5.15)
subject to the respective consumption levels abroad and at home:
cO = yO – M W, (5.16)
cW = yW – M W, (5.17)
Substituting eqs (5.16) and (5.17) into (5.15) delivers the expected
utility level of an individual:
ViW ri ( y W M W )␣ (1 ri )p1i␣ ( y O M W )␣ (5.18)
wV W w 2V W
with ! 0, and 0.
wp wp 2
Figure 5.2 displays the utility curves of the two alternative migra-
tion options, ViW (pi) and ViS (pi). Individuals choose the asylum-
seeking option if, and only if, ViW (pi) > ViS (pi). Depending on the
exogenous parameters, the cross-the-border utility function and
Vs, Vw
Vs(s, yo)
Vw(Mw,r, yo)
0 p* p** 1 pi
τi (pi )
0 p p 1 pi
Permanent Asylum Temporary Staying

in S in W in S in O
Figure 5.2 The refugee migration decision: who, where, and how long
concave-shaped asylum-seeking option can have a maximum of two

cut-off values. The two possible cut-off persecution levels p* and p**
(with p** $ p*) for the decision between the two migration options,
that is asylum seeking in the Western asylum country or cross-border
emigration, are determined by the equality of the two utility
functions:58
!
wV S ( p *) wV W ( p*)
V S ( p*) V W ( p *) 0 with d 0.
wp wp (5.19)
wV S ( p**) wV W ( p **)
!
V ( p**) V ( p**) 0 with
S W
t 0.59
wp wp
According to this condition, all individuals become asylum seekers

in the Western country, if
V W (p 0) ! V S (p 0), and V W (p 1) ! V S (p 1),
␣
(y O )␣ (y O MW )␣ § sy S ·
that is, if W ␣ W ␣
r !¨ W W ¸
.
(y M ) (y M )
W O
© y M ¹
This holds if the value of the asylum option is very high, that is if
the migration costs M W are very low, the income gap (yW – yO) is very
high, and/or the asylum policy of the Western country is very liberal
(high r).
There is only one intersection point p**, if
V W (p 0) ! V S (p 0), and V W (p 1) V S (p 1),

␣
(y O )␣ (y O MW )␣ § sy S ·
␣ ␣
r¨ W W ¸
.
that is, if (y M ) (y M )
W W O W
©y M ¹
The singular cut-off value p**, with 0 < p** < 1, is caused by low
migration costs M W or a ‘moderate’ Western asylum policy r with
individuals migrating either to the Western asylum country (those
with 0 < pi < p**) and those staying at home (or fleeing temporarily
to the neighboring country S). Furthermore, there is only one inter-
section point p*, if
V W (p 0) ViS (p 0), and V W (p 1) ! V S (p 1),

␣
that is, if §¨ sy S · (y O )␣ (y O MW )␣
W ¸
r W .
©y M ¹
W
(y MW )␣ (y O MW )␣
In this case, highly persecuted people (that is, pi < p*) leave the
home country and enter the neighboring country across the border,
while less persecuted individuals choose the long-distance asylum
option in a Western country. In this case, highly persecuted and risk-
averse individuals that face the risk (1 – r) of involuntary repatriation
(deportation) to the home country have a stronger incentive to avoid
this risk than less endangered individuals. Therefore, better eco-
nomic prospects in the Western asylum country have a relatively
strong impact for those individuals that have lower opportunity
costs arising from the option of living in persecution at home. Thus,
those who seek asylum in the Western country might be rather the
less persecuted refugees compared to those fleeing to a neighboring

first asylum country.
Two cut-off values (as in Figure 5.2) are possible, if
V W (p 0) V S (p 0), and V W (p 1) V S (p 1),

␣
§ sy S · (y O )␣ (y O MW )␣
that is, if ¨ W W ¸
!r W .
©y M ¹ (y M W )␣ (y O M W )␣
In this case, the Western asylum policy is rather strict. However, if
it is too strict, the asylum option is inferior for all N individuals and
no asylum migration to the Western country takes place (that is, no
intersection with V W (p) = VS (p)).
Proposition 5.2:
(i) There are individuals that do not leave the home country, if the income
differential between it and the Western country is not too large or
migration costs are rather high, that is if V W (p – 1) < VS (p – 1), that
is, if yW – M W < yO. (Note: Proposition 5.1(i) must still hold, that is,
syS < yO.)
(ii) There is no permanent emigration to the first asylum country, but
instead individuals migrate to the Western asylum country, if the asy-
lum recognition rate is not too low, that is if V W(p = 0) > VS (p = 0), that
␣
§ sy S ·
is, if r ! ¨ W ¸ .
©y M ¹
W
Since the persecution levels of the N individuals of the conflict-

affected country of origin are assumed to be uniformly distributed
between [0,1], the number of asylum seekers to the Western asylum
country ASW is then given by:60
ASW = N(p** – p*). (5.20)
The subsequent analysis of policy implications is based on compara-

tive statics based on the total differential of eqs (5.19) and the effects
on the number of asylum seekers to the Western asylum country ASW.
5.3 Counter-asylum policies
This section analyzes the effects of exogenous changes to some pol-

icy parameters on the pattern of refugee migration movements.
5.3.1 Defensive asylum policy: asylum restriction

5.3.1.1 Decline in self-reliance in the cross-the-border country
If the level of self-reliance s decreases, the consumption level of the
permanent cross-border refugees decreases, while that of the tempo-
rary refugees increases, and that of the individuals that stay at home
remains unchanged. Correspondingly, the average duration spent
abroad increases for the whole conflict-affected population:
yS ! 0
° 0 d pi d p,
wc S ° ␣ y S °
® 0 if ® p pi p , and (5.21)
ws ° (1 p )(1 ␣ ) °
° 0 ¯ p d pi d 1,
¯
0
° S O 0 d pi d p.
w␶ ° y y (1 ␣ ) °
® ! 0 if ® p pi p. (5.22)
ws ° ( sy y )
S O 2
°
°̄ 0 ¯ p d pi d 1.
The upper and lower cut-off values for temporary migration change
accordingly:
wp ␣(1 ␣ )y S y O ␣(1 ␣ )y S y O wp
! ! 0. (5.23)
ws ((1 ␣ )sy S )2 ( y O ␣ sy S )2 ws
This implies that if the first asylum country becomes more liberal
in granting higher levels of self-reliance s, then ceteris paribus the
number of hosted permanent refugees increases and the number of
returnees (that is, temporary refugees) decreases.
Graphically, a liberalization of self-reliance activities for cross-
border refugees shifts the VS (pi)-curve upwards for persecution levels
pi < p–, but remains unchanged for the less persecuted population
pi > p– (see Figure 5.3). Consequently, this positive shift in the value
of the cross-the-border option has a reducing effect on asylum
migration:
wV W wV S
dp *(**) t 0 for p*,
ws ws (5.24)
ds wV S wV W d 0 for p**,

wp wp
VS,VW
VS(s, yo)
syS
VW(MW, r, yo)
0 p* p*′ p**′p** 1 pi
Figure 5.3 Liberalization of self-reliance in the first asylum country
wV W wV S
resulting from eqs (5.19), 0 and t 0. 61 Thus, it follows
ws ws
that the number of asylum seekers decreases if the first asylum coun-
try becomes more liberal in its self-reliance policy:
d( p** p*)
N d 0. (5.25)
ds
Equations (5.22), (5.23), and (5.25) imply that if the economic sit-
uation for refugees in the first asylum country improves by granting
extended economic self-reliance, then (i) the number of permanent
cross-border refugees does not decrease, (ii) the number of asylum
seekers in the Western country does not increase, (iii) the total stock
of emigrants does not decrease, and (iv) the average duration of a ref-
ugee situation in the first asylum country does not decrease. Thus,
Western asylum countries have a reasonable self-interest in a more
liberal refugee policy in cross-the-border countries. This holds par-
ticularly when small geographical distances make the migration
costs M W for migrating to a Western country relatively affordable to
asylum seekers.
Proposition 5.3: An increasing degree of self-reliance in the first asylum

country in terms of a liberalizing refugee and integration policy (for exam-
ple, including work permissions, use of local services, and so on) increases
ceteris paribus the value of the cross-the-border option and protracts a ref-
ugee situation.
5.3.1.2 Decline in recognition rates in the Western asylum country

With respect to the restrictions on asylum immigration in the
Western country, asylum admission policy is an effective instrument
for regulating the stock of asylum seekers. As shown in Figure 5.4,
the value of the asylum option decreases for all N individuals, and
consequently, the number of individuals seeking asylum in the
Western country diminishes, increasing the number who choose to
stay in the region of origin:
wV W wV S
dp *(**) 0 for p*,
wr wr (5.26)
dr wV S
wV W ! 0 for p **,

wp wp
wV S wV W
by taking into account eqs (5.19), 0, and ! 0. 62 The same
wr wr
asylum reducing effect holds for increasing migration costs MW.63
Proposition 5.4: A more liberal asylum policy in the Western destina-
tion alleviates the refugee situation in the cross-the-border country, while
rising migration costs for realizing the asylum option to the Western asy-
lum country deflects refugee flows towards the first asylum country.
As a consequence of a more restrictive asylum policy in the Western

country, the inflow of a large refugee population aggravates the
public perception within the first asylum country of refugees as an
economic and political burden and a threat for the internal security
VS,VW
VS(syS,yO)
r
VW(MW, r, yO)
0 p* p*′ p**′ p** 1 pi
Figure 5.4 Declining asylum recognition rate in the Western asylum

country
of the civil society. Consequently, the first asylum country might

intensify encampment and reduce economic self-reliance, worsening
the refugees’ situation (Loescher & Milner, 2005). This policy
response of the first asylum country is driven by the expectation that
a more liberal encampment policy with enhanced opportunities for
the economic self-reliance of the refugees would rather protract the
refugee situation within its territory. Potential refugees expecting to
live under appalling encampment conditions are then ceteris paribus
more likely to repatriate earlier or to choose immediately the asylum
option in a Western country. As a consequence, the Western country
is likely to respond in an analogous manner by restricting asylum
conditions (that is, reducing recognition rates). Finally, a race to the
bottom is established with highly restrictive asylum policies in the
Western world and appalling refugee and encampment conditions in
the conflict-affected developing world.
The challenge of international refugee politics is to solve this
apparent dilemma of restrictive asylum policies, which is an ineffi-
cient equilibrium since investment into migration is only rewarded
for a subgroup of individuals. Deterrence, deflection, detention and
deportation of refugees are counterproductive measures for resolving
the appalling conditions for refugees, including the waste of resources
for an unsuccessful, long-distance asylum migration endeavor. But
beyond this, without dealing with the root causes of the refugee
movements, these measures are also costly to the respective asylum
countries, for example, by increasing costs for border control, admin-
istration and maintenance due to the increasing propensity of refu-
gees and asylum seekers to stay irregularly in the respective asylum
country and so on. An alternative to this bilateral asylum restriction
policy might be a more proactive approach, particularly on the side
of Western countries, that might tackle the underlying causes of
asylum seeking in both the country of origin and the first asylum
country.
5.3.2 Proactive asylum policy: migration-preventive

aid transfers
The Western country may invest resources into proactive measures
for tackling the root causes in the conflict-ridden country of origin
or sharing the refugee burden in the first asylum country, assisting
local integration that might also reduce the asylum migration
pressure to the Western country. Although there are various proac-

tive policies available, the focus in this analysis is on migration-
preventive aid transfers from the Western asylum country to the
country of origin or the first asylum country. The crucial question of
any proactive asylum policy in terms of migration-preventive aid is
whether aid can indeed reduce asylum migration flows. Or, in terms
of the present model, does an aid-induced increase of income levels
in the country of origin or the first asylum country relieve the asy-
lum burden in the Western country?
The following analysis presumes that aid is to some extent income-
effective. A necessary condition for this to be true is that aid, when
transferred from the donor to the respective recipient country, is
indeed channeled within the recipient country to the refugee popu-
lation where aid can develop the income-generating effect. This is a
necessary condition underlying the subsequent discussion.
5.3.2.1 Aid to the first asylum country

First asylum countries are often overstressed and unable to tackle the
political and economic challenges that large refugee inflows provoke.
A common reaction of these cross-the-border countries is to confine
refugees within camps, denying them freedom of movement, access
to social services or economic self-reliance. Aid targeted at refugees
in first asylum countries will promote the living standards of the
refugees. But, according to the previous model, an aid-induced
increase in the refugees’ income level implies that ceteris paribus the
refugee inflow into the first asylum country increases:
wpi ␣y O ␣s(1 ␣ )y O wpi
! ! 0, (5.27)
wy S s(1 ␣ )[ y S ]2 ( y O ␣sy S )2 wy S
However, the net outcome for the first asylum country depends on
the effect of asylum migration to the Western country:
wV W wV S

dp *(**)
wy S wy S t 0 for p*, (5.28)
dy S wV S
wV W d 0 for p**,

wp wp
wV W
with eqs (5.19) and by taking into account that 0,
wy S
wV S 64
and t 0.
wy S
Income-increasing aid towards the first asylum country has an

unambiguous reducing effect on the number of asylum seekers in
the Western country. Thus, if aid is channeled to the respective refu-
gee population (and if it is effective in increasing the refugees’ income
level), the asylum pressure to the Western country does not increase
(see also Figure 5.3):
d( p** p*)
N d 0. (5.29)
dy S
While this policy is beneficial to the Western country, it leaves the
first asylum country with a higher number of refugees. Interestingly,
although this policy might work for the interests of Western donor
countries, they are nevertheless reluctant to provide more resources
for that purpose (UNHCR, 2006b; Czaika & Mayer, 2008; see
Chapter 9). Why? Possibly, Western countries do not consider aid to
be as effective in generating additional income for the refugees or
they take the negative consequences for the first asylum countries
into account. Obviously, first asylum countries might be even more
reluctant for Western countries to adopt this policy, because it shifts
and consolidates the refugee burden onto their territory. As a conse-
quence, the first asylum country would respond by imposing further
restrictions on economic self-reliance and encampment conditions
for refugees. Beyond, governments of refugee-hosting developing
countries might oppose a likely diversion of aid amounts taken away
from the needs of the native population and reallocated to the refu-
gees within their territory.
5.3.2.2 Aid to the country of origin

Alternatively, the Western asylum country could transfer aid towards
the country of origin in order to reduce refugee outflow and facilitate
a sustainable voluntary repatriation. Pre-conditioned by the fact that
countries of origin do not hinder a voluntary repatriation of their
population after mass emigration, rapid post-conflict reconstruction
with appropriate levels of civil security, basic social services and eco-
nomic perspectives for returnees might be a prior rationale for target-
ing foreign aid towards the country of origin (UNHCR, 2006b).
Obviously, cross-the-border asylum countries are also interested in
sustainable return solutions with voluntary repatriation and reinte-
gration of refugees that would otherwise stay within their territory.
Thus, aid to the country of origin that ensures rising income levels
for the conflict-affected population is also in the interest of first asy-
lum countries, since – without considering asylum migration to the
Western country – the total stock of refugees into the first asylum
country decreases, with the less persecuted people going home first:
wpi ␣ ␣(␣ 1)sy S wpi

0. (5.30)
wy O
(␣ 1)sy S
␣sy S y O wy O
However, the consequences for the Western asylum country are

wV W wV S
less distinct. Taking into account eqs (5.19), t 0, and t 0,
wy O
wy O
65
the overall effect on the Western country is ambiguous:
wV W wV S
O
dp *(**)
wy O wy !
0 for p * and p**. (5.31)
dy O wV S wV W

wp wp
d( p ** p *) !
and thus, N 0. (5.32)
dy O
Figure 5.5 shows that in this case both curves shift upwards so that
the total effect on Western asylum migration becomes unclear. Aid
transfers to the country of origin are unambiguously beneficial for the
first asylum country, but not necessarily for the Western aid donor.
VS,VW
VS(s, yO)
yO
VW(MW, r, yO)
0 p*′ p* p**′ p** 1 pi
Figure 5.5 Increasing income level in the country of origin

A numerical simulation of such a policy, shown in Table 5.1, indi-

cates that the asylum pressure exerted on the Western asylum coun-
try is generally lower the more liberal the first asylum country is in
its self-reliance policy. However, the effect of income-generating aid
yO (AO) on the change in the number of asylum-seeking individuals
N . d(p** – p*) is not as clear. It is possible that for rather high values
of the parameter s, that is when economic self-reliance is widely
granted to refugees in the first asylum country, aid to the conflict-
affected country of origin could even increase the number of indi-
viduals that choose the asylum option in the Western country. Why?
Simply stated, the better the economic conditions in the country of
origin, the lower the value of the permanent cross-border option rel-
ative to the asylum option in the Western country.
This occurs when economic self-reliance in the first asylum coun-
try is generous (s is high) and therefore permanent cross-border
migration is relatively extensive. This would also imply that less per-
secuted people with medium to high levels of p choose, ceteris pari-
bus, this option. However, for these less persecuted individuals, the
risk of deportation (if they would choose the asylum option in the
Western country instead) is associated with a lower disutility of liv-
ing at home than to the highly persecuted individuals. Thus, in this
Table 5.1 Simulation: aid transfers and refugee migration66
Income Self-reliance Asylum Permanent Temorary

in O in S in W in S in S
yO s p** = p* p* p*
(AO) (for t(p*) = 1) (for t(p*) < 1)
100 0.95 0.162 0.693 0

110 0.95 0.542 0.219 0
120 0.95 0.580 0 0.093
130 0.95 0.530 0 0.063
100 0.90 0.609 0.250 0
110 0.90 0.682 0.079 0
120 0.90 0.633 0 0.040
130 0.90 0.558 0 0.035
100 0.85 0.788 0.068 0
110 0.85 0.731 0 0.024
120 0.85 0.655 0 0.018
130 0.85 0.573 0 0.020
particular case, aid to the country of origin reduces permanent

migration to the first asylum country but increases asylum migra-
tion to the Western country, because the less persecuted individuals
choose this option instead. Consequently, aid to the country of ori-
gin is only effective for the Western country if the cross-border
option is devaluated by a strict refugee policy of the first asylum
country. This ambiguous aid impact might also contribute to the fact
that Western donor countries are rather hesitant in transferring aid
to countries of origin for migration-preventive purposes (Czaika &
Mayer, 2008; see Chapter 9).
5.3.2.3 Self-financing proactive asylum policy

In light of these potentially unwanted effects of proactive measures
on refugee migration, we should explore whether Western asylum
countries would have any incentive to invest in such proactive mea-
sures. Given the efficacy of aid as an instrument for increasing the
respective income level of conflict-affected individuals living either
permanently or temporarily in the country of origin or in the first
asylum country, aid transfers require that there must be benefits, for
example, in terms of decreasing asylum costs, spilling over to the
Western (aid-giving) asylum country. Thus, this postulate presumes
double effectiveness: aid is effective in increasing respective income,
and consequently, it is effective in reducing asylum migration flows
to the Western country. If this double effectiveness holds, a self-
interested Western donor country allocates a positive amount of aid
A = AS + AO to the first asylum country (AS) or to the country of origin
(AO) according to the following marginality condition:
w( p** p*) wy S w( p** p*) wy O

S O d 0. (5.33)
wy S wA wy O wA
However, this allocation rule is not sufficient. Aid is only trans-
ferred if the aid-induced reduction in asylum-related costs DC(A) is
larger than the transferred amount of aid A.67
Figure 5.6 illustrates this relationship between the aid-reduced
asylum-related costs C(A) and the transferred amount of aid A.
According to this, migration-preventive aid is cost-effective (‘self-
financing’) if total costs TC(A) = C(A) + A are lower for a positive
amount of aid, that is if TC(A) # C(0). Consequently, if aid is
migration-effective, aid for refugees is in general a reasonable policy
option for Western asylum countries.
TC = C + A
45°
TC
C(0)
C(A)
A* A** Aid
Figure 5.6 Self-financing proactive asylum policy
This study contributes to the economic analysis of refugee migra-

tion. Utility-maximizing individuals that are in danger of their life,
leave their home countries not primarily out of economic reasons
but to avoid the disutility of persecution. However, refugees also
respond to economic incentives. Equivalent economic conditions in
a cross-the-border country or the expectation of asylum status in a
Western country with better economic perspectives induce long-
lasting absence from the country of origin. Consequently, refugee
situations that are generated by conflict and human rights violations
at home and are sustained in neighboring first asylum countries for
a long period of time go in hand with a twofold asylum policy fail-
ure. First, two-sided restrictive asylum policies force refugees to stay
in desperate encampment situations in remote areas. And second, a
lack of proactive engagement of Western countries either in the
country of origin or the first asylum country hinders the implemen-
tation of a durable solution in terms of voluntary repatriation and
reintegration at home or of local integration abroad.
6
Asylum Cooperation among
Asymmetric Countries
Abstract68
This chapter argues that asylum cooperation for tackling the conse-
quences of large asylum migration inflows is only possible among
rather symmetric countries. Highly asymmetric countries have no
incentives to join and remain in a stable coalition. The distinction of
cost and spillover asymmetries shows that financial transfers may
only release participation constraints, and thus, make asylum cooper-
ation feasible, if they are focused on tackling this asymmetry. This
result becomes relevant when applied to the context of the enlarged
European Union. I argue that there is potential for a future coopera-
tive burden-sharing regime for asylum, and particularly, if unanim-
ity is replaced by the double majority principle in votes of the
European Council, as the EU reform treaty suggests.
6.1 Introduction
In the recent past, asylum and immigration affairs have gained

increasing importance on the political agenda of the European Union
and its member states. There are two major reasons for this: first, a
strong pressure of (illegal) immigration and asylum seeking, and sec-
ond, an unequal (cross-country) distribution of asylum seekers and
refugees. Although the number of asylum claims in industrialized
countries has constantly dropped from its peak in 1992, it is still
twice as high as in 1980, with nearly 340,000 applications per year
in the industrialized Western countries by the end of 2005 (UNHCR,
88
Asylum Cooperation among Asymmetric Countries 89
2006b). This increased immigration pressure, particularly from

developing countries, has intensified the political debate and led to
various proposals for improving cooperation at the supranational
level. Furthermore, in addition to managing the overall inflow, the
uneven distribution of asylum seekers and refugees among European
host countries demands adequate burden-sharing policies.69
Before the end of the Cold War, refugees coming from the East
were welcomed in the Western world and reception countries derived
strategic benefit from the provision of refugee protection (Loescher,
1989). Today, without the strategic interests of the Cold War era, ref-
ugees are considered to be an ever-growing burden for economies
and societies and a threat to internal stability. The upward trend in
the number of asylum applications during the 1990s has made states
reluctant to continue providing generous protection to refugees
(Hatton, 2005).70
Especially among the European Union member states, policymak-
ers have adopted more restrictive legislative and administrative
measures to control the number of asylum seekers in order to reduce
the political, social, and economic costs of refugee protection
(Thielemann, 2005). In the supranational EU arena, support is grow-
ing for a Common European Asylum System (CEAS) that would help
harmonize national asylum policies and share the responsibility for
asylum seekers Union-wide. A central pillar of a future CEAS is the
implementation of an EU-wide asylum cooperation regime with a
common burden-sharing rule for refugee admission. However, since
decisions on a cooperative asylum regime would require the rarely
attainable unanimity of the European Council, member states still
act non-cooperatively and unilaterally in their decisions on national
admission levels.
In the following, we argue that the reason for non-cooperation on
refugee admission is the relative asymmetry of EU member states;
yet, in contrast to efforts for a global cooperation regime, asylum
burden-sharing among EU member states is nevertheless feasible. In
the theoretical part of this paper, we analyze two alternative cooper-
ative asylum burden-sharing regimes: at first, we model a simple
cooperative refugee burden-sharing regime, and in a second step, we
extend this with a monetary transfer scheme with cross-country
payments compensating for asylum provision abroad. This study
shows that under some particular circumstances, depending on the
type of heterogeneity (costs vs. spillovers), monetary transfers can

enhance national and collective welfare by widening participation
in a cooperative asylum regime.
Finally, we apply the propositions derived from the analytical frame-
work to the multi-country context of the enlarged European Union.
We are rather hesitant to argue that the status quo of non-cooperation
could be overcome solely through the implementation of an explicit
monetary burden-sharing regime such as the European Refugee Fund.
This pessimism is nurtured by the theoretical assumptions. It shows
that monetary transfers do not make asylum cooperation profitable if
the countries are characterized by cost asymmetries instead of spillo-
ver asymmetries. However, cost asymmetries are quite reasonable to
assume for the EU member states, particularly after the recent EU
enlargement. We are much more optimistic about the introduction of
a supranational EU asylum institution which implements a cooper-
ative asylum regime. This optimism is based upon the prospect of the
EU reform treaty replacing the unanimity rule by a double majority
principle (of at least 15 member states representing 65 percent of EU
population minimum) for votes in the EU Council. By simulation, we
find that when decisions are made according to this modified voting
rule, a ‘double majority’ in the EU Council is plausible, and thus a
cooperative asylum burden-sharing regime is feasible.
The remainder of this chapter is as follows. Section 6.2 briefly
reviews the relevant literature on refugee protection and monetary
transfers. Section 6.3 compares analytically two alternative cooper-
ative refugee burden-sharing regimes, with and without monetary
transfers. Section 6.4 applies these implications numerically on the
context of the enlarged European Union. Section 6.5 concludes the
chapter.
6.2 Literature review
By its very nature, asylum migration movements impact relations

between the various destination countries, as unilateral policy
responses by one country tend to generate externalities for other
countries. Thus, asylum provision and refugee protection has public
good characteristics with free-riding opportunities undermining
cross-country cooperation, leading to an under-provision of asylum
capacities. Suhrke (1998) discusses this problem of collective action
failure in international refugee matters and stresses the relevance of

state-specific interests in the establishment of an international refu-
gee burden-sharing system. Furthermore, she argues that the estab-
lishment of a collective action system in refugee affairs is generally
more difficult to achieve than in other international policy fields,
due to the unpredictability of future refugee flows, and thus, uncer-
tainty about future costs. Betts (2003) presumes that refugee admis-
sion generates both international (public) benefit spillovers as well as
country-specific (private) benefits, stemming both from human
rights considerations and norms of solidarity. Betts (2006) postulates
cost-benefit perceptions of refugee protection to be country-specific
and also dynamic. Hatton & Williamson (2005) and Hatton (2005)
refer to this cost-benefit approach for modeling international cooper-
ation in the field of refugee protection. Facchini et al. (2006) extend
this cooperation model by taking into account the democratic proc-
ess through which national asylum policy is determined. All of these
approaches are based on the premise that positive cross-border exter-
nalities from refugee protection exist and they demonstrate that
cooperation leads to a welfare improvement compared to a non-
cooperative situation.
Thielemann (2004) and Thielemann & Dewan (2006) analyze ref-
ugee burden-sharing within the EU and therewith focus on mecha-
nisms to overcome the very unequal distribution of refugees across
the EU. They argue that the unequal distribution of refugees can
basically be attributed to structural pull factors that are beyond the
control of states, since refugees often base their destination country
choice on non-political criteria (Holzer & Schneider, 2002; Neumayer,
2004). In this regard, most initiatives addressing the unequal distri-
bution of asylum seekers and refugees are insufficient, because they
neglect the root causes of unequal burdens. Thus, an optimization of
refugee policy must not exclusively rely on cost-benefit schemes, but
must also refer to the norms and institutional structures of the EU
member states.71 Crisp (2005) argues that the norms and instruments
of the refugee system will need to change in a variety of ways in
order to cope with the rising number of asylum applications and the
increasing irregular (‘illegal’) immigration pressure. He strongly
emphasizes the necessity for protection in the regions of origin (for
example, by promoting both adequate human rights standards and
economic progress) in order to avoid mass influxes into industrialized
countries. The empirical evidence on this issue is rather ambiguous.

While Neumayer (2005) provides evidence that improving economic
and political conditions in the country of origin lower the asylum
pressure on the destination countries, Rotte et al. (1997) find that
support of democracy and economic development will not necessar-
ily reduce migration, at least not in the short run, and thus, restric-
tive legal measures seem to be a more efficient instrument for
reducing asylum pressure. Hatton (2004) supports this argument,
but he further investigates the criteria that influence national refu-
gee policies, as well as the motivations for seeking asylum in Europe.
He finds that despite more restrictive asylum policies in the early
1990s, there was an overall increase in applications across EU mem-
ber states. Hatton concludes that the impact of restrictive asylum
policies has been superposed by rising worldwide political oppres-
sion and increasingly numerous conflicts. In summary, the literature
establishes that burden-sharing and a more equal distribution of
refugees among EU member states needs to take into account coun-
try-specific conditions of asylum countries (Neumayer, 2004).
Furthermore, reactive asylum burden-sharing is surely not sufficient
but needs to be paralleled by efforts tackling the root causes and
measures to manage the overall influx such as migration-preventive
development aid policies or peace-keeping missions.
In order to analyze whether a financial burden-sharing institution
is able to establish more cooperative asylum outcomes among
European countries, we briefly refer to the literature on the role of
side payments in cooperation regimes. The conditions for self-
enforcing agreements, that is agreements that are not based on the
commitment to cooperate, are extensively discussed in the environ-
mental cooperation literature (see, for example, Carraro & Siniscalco,
1993; Barrett, 1994). For establishing such an agreement, countries
first decide non-cooperatively whether or not to join an agreement
given the burden-sharing rule the agreement adopts, and afterwards
they decide upon provision levels of the public good. The problem is
that some countries may join such an agreement and others not, so
that a coalition is only formed by a subset of all countries. Non-
participating countries prefer to free-ride on the provision levels of
the cooperating members. Barrett (1999, 2001, 2002) extensively
analyzes country-specific cooperation incentives for the provision of
global (environmental) public goods. He emphasizes that the
characteristics of a good determine whether enforced international

cooperation or rather unilateral provision is preferable. Furthermore,
he states that for a high number of cooperating countries, a self-
enforcing agreement can only sustain a superior welfare outcome if
the global net benefit between the cooperative and the non-
cooperative outcome is small. In this context, Carraro & Siniscalco
(1993) show that stable coalitions only exist among a small number
of countries. However, they show that through cooperation inducing
monetary transfers, coalitions could be stabilized and expanded.
Harstad (2007) analyzes different types of agreements for the inter-
nalization of externalities in political cooperation. He derives some
conditions under which side payment schemes are efficient forms of
cooperation. For instance, the efficiency of side payments depends
on the heterogeneity of the cooperating partners, the amount of
externalities and the value of the international agreement. If bar-
gaining incentives are too high for some countries, a side payment
scheme can also produce a suboptimal outcome, since a growing
conflict of interests might delay the achievement of an agreement.
Side payments are more likely to increase efficiency if heterogeneity
is low and externalities are large. Jackson & Wilkie (2005) demon-
strate that side payments do not always lead to an increase in effi-
ciency, even in the case of complete information, binding contracts
and absence of transaction costs. They argue that efficiency is only
improved by side payments as long as transfers are used to ensure
that other participants internalize their externalities. If there are
incentives to misuse transfer payments to manipulate the behavior
of the cooperation partner to a country’s own advantage and not
necessarily to the socially desirable outcome, side payments can
deteriorate the combined welfare situation relative to a situation
without side payments. With respect to asylum policy, Facchini et al.
(2006) show that cross-country financial transfers as incentives to
internalize positive asylum spillover effects may result in a welfare
inferior outcome. This stems from strategic delegation that might
encourage countries to accept a suboptimal number of refugees.
Beyond this, the literature is rather limited regarding the welfare
aspects of financial burden-sharing instruments in asylum policy
issues, despite the fact that (earmarked) aid is already transferred
either via bilateral or multilateral channels (for example, by UNHCR)
to share asylum-burdens in some refugee-hosting countries.
Thielemann (2006) distinguishes several concepts of financial bur-

den-sharing systems beyond the reactive ‘physical’ burden-sharing
mechanism and the harmonization of asylum laws. Hereby, retro-
spective monetary compensations to the highly burdened asylum
countries are one possible approach. The European Refugee Fund
(ERF), established in 2000, is partly based on this concept. This fund
aims to balance the costs among member states for hosting and inte-
grating refugees. Thielemann (2005) critically analyzes the transfer
payment mechanism of the ERF and its effectiveness in promoting
burden-sharing and in enhancing the protection of refugees. He crit-
icizes the ERF for not taking country-specific conditions into account,
and emphasizes the importance of considering the member states’
heterogeneity when building a cooperative asylum regime. He con-
cludes that the ERF transfer payment logic is rather limited in its
redistributive impact and therewith in its potential as a burden-
sharing instrument. Additionally, Noll (2003) finds fault with the
financial volume of the ERF stating that given the total costs of
reception and the share channeled to each member, redistributive
effects are not more than a drop in the ocean.
Furthermore, a financial burden-sharing regime may also refer to a
market-based allocation mechanism where a traditional quota system
is amended by the possibility to trade the quota (see Schuck, 1997).
However, this concept is often criticized because of ethical reasons
(‘trading refugees’), and because of its narrow perspective, focusing
only on one aspect of refugee-related financial compensations
(Thielemann, 2006). Therefore, Thielemann & Dewan (2006) suggest
a more implicit trading of an asylum-related public good bundle, for
example, refugee protection, development aid, peace-keeping/mak-
ing efforts and so on, where countries specialize according to the
comparative advantage of their contribution (see Boyer, 1989).
Next, we refer in a welfare analysis to the question of whether asy-
lum cooperation among heterogeneous (asymmetric) countries is
generally possible, and then, whether financial transfers pose to be
an effective tool for enhancing cooperative solutions.
6.3 The analytical framework
Based on the overall implications of the literature, cooperation in

refugee admission can be welfare enhancing for host countries.
This is due to the internalization of cross-border externalities, stem-

ming from refugee protection in other countries than the own. We
reproduce this result in a static two-country model where both
countries aim to optimize their welfare with respect to the trade-
off between the costs and benefits accruing from refugee protec-
tion. Even though it would be interesting to analyze the formation
of cooperative asylum regimes with more than two (asymmetric)
countries, it is sufficient to analyze the theoretical implications in
a two-country model. However, in the empirical application, we
use a numerical simulation for analyzing the multi-country case of
the enlarged EU.
6.3.1 Asylum cooperation without monetary transfers

The model assumes an exogenous stock of asylum seekers. A certain
number (share) r of asylum seekers of this stock is offered legal refu-
gee status in country i (ri), or country i* (ri*). The respective admission
levels depend on national asylum policies. Each country decides uni-
laterally on the number (or, share) of refugees to be accepted. The
model assumes that asylum seekers are not strategic actors, but are –
if necessary – resettled between the two countries.72
Both countries benefit from refugee protection either domesti-
cally or abroad. Country-specific benefits are generally assumed to
be generated by economic (labor market) or political (international
prestige, regional stabilization) causes. Moreover, humanistic atti-
tudes imply that (altruistic) individuals are likely to gain from the
knowledge that refugees find protection in the own country or
somewhere else (Hatton & Williamson, 2005). Thus, each country
generates country-specific (private) benefits as well as positive
(public) externalities from refugee protection spilling over to the
other country.
Costs for refugee protection are assumed to be increasing with the
number of admitted refugees and are assumed to capture economic
and political factors. Economic costs might reflect administrative
costs for asylum processing, juridical support, maintenance costs,
nutrition, social security, education and so on, whereas political
costs are rather caused by integration endeavors and the promotion
of public xenophilia. Finally, we assume that both countries are
completely informed about all relevant parameters. The relationship
between costs and benefits caused by the admission of asylum
seekers can be illustrated by the following welfare functions for each

country i:
Wi ri i ri * ci ri2 , (6.1)
where Wi represents the country-specific welfare function, ri denotes

the number (or, share of an exogenously given asylum stock) of
admitted refugees in country i, and ci is the country-specific refugee
admission cost parameter.
We assume increasing marginal costs of hosting refugees: the
higher the number of refugees already admitted, the higher the mar-
ginal costs are for admitting another refugee. Parameter bi > 0 meas-
ures the amount of positive externalities spilling over from abroad.
That is, a country’s welfare level increases with the number of admit-
ted refugees abroad. Both countries maximize their respective wel-
fare functions simultaneously by taking the asylum choice of the
other country as given. Then, welfare maximization with respect to
the number of admitted refugee results in the following non-
cooperative outcome:
1
riu = , (6.2)
2ci
The optimal number of refugees increases with decreasing costs for

hosting refugees. According to the optimal number of admitted refu-
gees, country-specific welfare levels are given by:
1 1
Wiu i , (6.3)
4ci 2ci*
Next, we determine the optimal number of refugees when the

authority for national asylum policies is delegated to a supranational
institution. Such an institution maximizes collective welfare V = Wi +
Wi* by assessing the socially optimal number of refugees ri to be admit-
ted by country i as a cooperative outcome:
1 i *
ric riu , (6.4)
2ci
Cooperation leads to the internalization of positive externalities

generated by each country. This amounts in higher numbers of
admitted refugees in the cooperative scenario compared to
non-cooperation. As before, the optimal number of admitted refu-

gees decreases with rising costs for hosted refugees. Obviously, inter-
nalizing the positive spillovers of refugee hosting, asylum cooperation
increases national welfare levels of both countries:
1 i2* 1 i
Wic i . (6.5)
4ci 2ci*
As V c Wic Wic* maximizes the collective welfare due to the higher

overall number of admitted refugees, we know that unilateral actions
are jointly inferior, that is V c Wiu Wiu* . Thus, V c describes the wel-
fare benchmark for any international asylum regime.
Despite this, it might not be unilaterally optimal to make a cooper-
ation commitment. Each country is only willing to participate in an
international asylum regime if cooperation is beneficial for the
country. Although the maximization of collective welfare is the
social optimal solution, participation is not necessarily incentive-
compatible to each country. The profitability of an agreement for all
participating countries is a functional prerequisite for the cooper-
ation scheme to be implemented. Therefore, collective welfare
maximization must be amended by the following participation
constraint:
max V c s.t . Wic Wiu,

ri
and the necessary condition that ensures that both countries par-
ticipate in a cooperative asylum regime is then given by:
ci* 2ci*
i* i i* . (6.6)
2ci ci
The fulfilment of this participation constraint requires that both

countries are not too heterogeneous in the relevant parameters, that
is, the maximum spread in spillovers bi is determined by the respec-
tive cost asymmetries. Figure 6.1 indicates by the relatively narrow
sector I that for spillover asymmetries (and equal cost parameters)
asylum cooperation is only feasible when countries are rather similar
in their spillover parameters. Combinations outside this sector are
unilaterally inferior and asylum cooperation is not a Pareto-optimal
outcome. For countries with asymmetric costs, Figure 6.2 displays
βi
I
III II
II
(Ti −>i * > 0) I
1 III
(Ti *−>i > 0)
1 βi*
Figure 6.1 Asylum policy outcomes among countries with asymmetric

spillovers
Notes: Sector I represents the participation constraint (6.6) in a regime without mone-
tary transfers, sectors II and III, respectively, indicate participation constraints (6.15)
and (6.16) for a financial burden-sharing regime. Cost parameters ci are held constant
and are assumed to be equal for both countries. Cost asymmetries, for example, ci > ci*,
would turn sectors I and II clockwise.
ci I
II
II
1 ci*
Figure 6.2 Asylum policy outcomes among countries with asymmetric

costs
Notes: Spillover parameters bi are held constant and are assumed to be equal for both coun-
tries. Spillover asymmetries, for example bi > bi*, would turn sectors I and II clockwise.
sector I in which asylum cooperation is profitable for both coun-

tries.73 Subsequently, we analyze whether monetary transfers are able
to extend these rather narrow sector I’s to make asylum cooperation
more profitable also for more asymmetric countries characterized
either by spillover or cost heterogeneity.
6.3.2 Asylum cooperation with monetary transfers

A supranational institution with the authority to set adequate mon-
etary incentives can exert a decisive influence on the (autonomous)
implementation of cooperative asylum policies of asymmetric coun-
tries (Barrett, 2002). Cross-country transfers, provided unilaterally
but set collectively, can promote efficiency in political cooperation
by increasing the incentives to participate in an agreement. This par-
ticularly holds if cooperation has low opportunity costs for the
respective countries. Monetary transfers may enhance asylum provi-
sion levels and make cooperation a more attractive policy option
(particularly for countries that would actually lose from cooperation)
by internalizing the positive external effect of their actions. With
respect to the implementation of a financial transfer regime, Schuck
(1997) proposes a transfer system that consists of a centrally set quota
for each state and a subsequent possibility to trade the quota obliga-
tion with other states either in terms of financial transfers or other
resources (for example, package deals).74
We extend our model by including a monetary transfer scheme
and consider the following two-stage game: in the first stage, the two
countries simultaneously offer a monetary transfer t for each refugee
that the other country hosts; in stage two, each country decides uni-
laterally upon its optimal level of refugees to be admitted, taking the
transfer offers of the first stage into account.
As before, countries benefit from the admission of refugees to the
other country and likewise generate positive spillovers. This raises
the unilateral incentive to compensate the other country financially
in order to accept more refugees so that unilateral as well as collec-
tive welfare improve.
In order to derive the unilateral transfer schedule, eq. (6.1) is
adjusted as follows:
Wit ri i ri* ci ri2 ti oi* ri* t i * oi ri , (6.7)

Wit represents the unilateral welfare of country i including both con-

tributed and received transfers. ti* oi (ti oi* ) reflects the transfer con-
tribution schedule of country i* (i) towards country i (i*). Obviously,
unilateral welfare levels are directly increased by received transfers
and diminished by contributed transfers. Monetary contributions
also have an indirect positive impact on national welfare via spillo-
ver gains induced by a higher refugee protection level in the other
country.
For solving this two-stage game backwards, both countries first
anticipate the number of refugees the other country is going to accept
in the second stage depending on the transfers offered in the first
stage. For this, country i maximizes its welfare function Wit with
respect to ri in order to get:
1 t i* oi
riu . (6.8)
2ci
Equation (6.8) reflects the optimal number of refugees that coun-
try i admits, depending on the transfers it receives from the other
country. Obviously, if country i offers country i* a monetary com-
pensation per admitted refugee, country i* is encouraged to increase
its admission level above the optimal level without transfers.
Anticipating the maximization calculation of the other country,
the optimal unilateral monetary transfer offered by country i is
derived by inserting eq. (6.8) into eq. (6.7). The subsequent first order
condition of eq. (6.7) is then:
i 1
tu . (6.9)
i oi * 2
t u is the optimal amount of transfers that a country is willing to

pay as compensation for its received cross-country spillovers. The
optimal number of admitted refugees results from substituting
eq. (6.9) into eq. (6.8), respectively:
1 i*
riu . (6.10)
4ci
It can be seen that, for any spillover parameter bi > 0, the non-
cooperative admission level with transfer payments is always half
of that of the cooperative outcome without monetary transfers.
However, when comparing the two alternative non-cooperation
scenarios, it becomes obvious that admission levels with transfer

payments are only higher if the respective spillover parameter bi is
larger than unity. That is, non-cooperative refugee admission levels
are higher in the regime with monetary transfers only if refugee pro-
tection abroad generates comparatively higher benefits than domes-
tic protection. Accordingly, the amount of monetary compensation
country i offers to the other country increases with its externality
parameter bi. Net amounts of transfer payments T of country i (that
is received minus offered transfers) depend on their respective spill-
over and cost parameters and are given by
i2 1 i2 1
Ti oi* Ti* oi . (6.11)
4ci* 4ci
Equation (6.11) shows that for countries with asymmetric spillo-

vers, net transfers are always non-zero. Eq. (6.9) indicates that a coun-
try only offers non-negative gross transfers t u if bi is larger than unity.
For parameter values smaller than unity, countries do not contribute.
Thus, if bi > 1, the respective non-cooperative welfare level Wit , u for
country i is given by:
(i* 1)2 (i 1)2

Wit ,u . (6.12)
16ci 8ci*
Figure 6.1 shows that, outside the square area, at least one country
compensates the other country non-cooperatively for the spillover
gains; consequently, the country with the larger spillovers parameter
is the net contributor of financial transfers.
Next, it becomes evident that, under certain circumstances,
cooperatively set transfer payments (implemented, for example, by a
supranational authority) are a valuable instrument to attain the
cooperative benchmark solution without transfers. Under monetary
cooperation, payments are set according to equal refugee admission
levels with those of the cooperative (benchmark) solution without
transfers. Eq. (6.4) is set equal to eq. (6.8) and solve for the respective
financial transfers. Thus, by setting unilateral transfers according to
tc i , (6.13)
i oi *
the collective welfare V c Wit ,c Wit*,c is maximized.

By inserting eqs (6.4) and (6.13) into eq. (6.7), the cooperative wel-
fare level for country i is given by:
(1 i*) 2
Wit ,c , (6.14)
4ci
A country is only willing to participate in an international asylum
cooperation regime if cooperation is unilaterally beneficial. Although
the maximization of collective welfare delivers the social optimal
solution, participation might not be incentive-compatible for each
country. Even if collective welfare is maximized under the monetary
transfer regime, the outcome might be inferior for at least one coun-
try. In this case, not joining the cooperative financial transfer regime
is the dominant strategy.
For guaranteeing that the monetary cooperation regime is profit-
able for both countries the following conditions must hold:
Wit ,c Wit ,u and Wit ,c Wiu .
They are functional prerequisites for the implementation of the

monetary cooperation regime. Thus, a cooperative asylum regime is
only feasible if the following participation constraints are satisfied:75
2ci* (i* 1)2 3ci* (i* 1)2

1 i 1, (6.15)
3ci 2ci
2ci* i* c * * (i* 2)

1 1 i i i . (6.16)
ci 2ci
Again, these participation constraints are fulfilled as long as both
countries are not too heterogeneous. Constraint (6.15) ensures that
for both countries, cooperation within a monetary burden-sharing
regime (by offering higher transfers t c > t u) is superior to non-
cooperation. Constraint (6.16) also ensures that for spillover param-
eters bi < 1 , monetary cooperation is beneficial for both countries.
Figure 6.1 displays the sectors II (relevant for bi 1) and III (relevant
for bi < 1), the feasible combinations for which asymmetric countries
voluntarily access a cooperative asylum regime with monetary trans-
fers. Obviously, monetary transfers widen the rather strict symmetry
requirement of the narrow sector I which represents the cooperative
asylum regime without transfers. However, this holds only for low to
moderate spillovers; if both countries gain from mutually large spill-

overs, the participation constraint is stricter for the monetary cooper-
ation regime.
However, when analyzing the respective constraints for cost rather
than spillover asymmetry, as displayed in Figure 6.2, we see that the
participation constraints for the monetary cooperation regime are
entirely more strict than the asylum cooperation regime without
transfers – at least if spillovers are rather equal and mutually low. This
is shown in Figure 6.2, where sector I is strictly larger than sector II.
Thus, monetary transfers cannot promote cooperation if countries
are characterized rather by cost asymmetries than by spillover asym-
metries. Instead, a cost-sharing regime would be more appropriate.
6.3.3 Summary and interpretation

Asylum cooperation with or without monetary transfers is generally
not attainable among highly asymmetric countries. It is also shown
that non-cooperative transfers are only contributed when refugee
admission abroad generates higher benefits than domestic hosting,
that is if bi > 1. The reason that this does not result in a corner solu-
tion where only one country provides asylum is that refugee hosting
causes (private) costs for this country. Generally, if (public) spillover
benefits exceed (private) domestic benefits, countries are willing to
perform positive gross compensations to the other country, although
clearing gross transfers generates the result that, in the end, only the
country with the larger spillover gains is the net contributor. Since,
consequently, asylum cooperation is only feasible among rather sym-
metric countries (that is, solutions within sectors I and II, respec-
tively), it is actually required that countries are simultaneously
homogeneous with respect to both (public) spillovers and (private)
costs. For instance, for countries with relatively high costs or low
spillover gains, cooperation is not profitable and free-riding becomes
the superior strategy. In this case, non-cooperative unilateral asylum
policies dominate international asylum cooperation, even if it does
result in a collectively lower welfare level (Suhrke, 1998).
The impact of a monetary transfer regime on the feasibility of
cooperative asylum outcomes is ambiguous. The distinction between
cost and spillover asymmetries shows that, ceteris paribus, a monetary
burden-sharing regime is only able to extend the set of cooperative
solutions in the case of spillover heterogeneity; however, this only
holds among countries with mutually low spillover parameters (see

Figure 6.1). For situations where countries are rather characterized by
cost heterogeneity, a monetary transfer regime does not widen the
set of cooperative solutions. The reason for this result is that transfer
schedules do not depend on country-specific costs, but on spillover
benefits. Thus, in a multi-country context with relatively high cost
asymmetries, a cost-sharing regime is the more adequate tool for
generating an asylum cooperation solution.
Consequently, asylum cooperation among developing and devel-
oped countries seems to be rather unlikely because of the large cross-
country asymmetries with respect to both costs and spillovers. Thus,
our analysis implies that asylum cooperation is more likely on a
regional level among rather similar countries. Therefore, we will now
discuss these implications in a multi-country context by analyzing
the asylum cooperation potential among the member states of the
European Union.
6.4 Asylum cooperation in the European Union
The purpose of this section is to answer the question why – to date –

we have not seen a cooperative asylum regime within the EU. Olson
(1965) argues that in a multi-country context, in general, the more
actors involved, the more difficult it is to overcome collective action
failure. On the other hand, Jackson and Wilkie (2005) point out that
for an (endogenous) transfer scheme which internalizes two-sided
externalities, the more actors that are involved, the easier it is to
achieve efficient strategies, because strategic bargaining is limited
due to possible sanctions and reputation losses. Beyond these con-
trary arguments, we would expect, according to our model implica-
tions, that the EU is the regional context for which a cooperative
asylum burden-sharing regime may seem possible; though, the
degree of heterogeneity has increased significantly among the EU
member states since its recent enlargement. However, the evidence of
the past is different, and therefore, the focus here is to analyze the
reasons for this collective action failure in European asylum
cooperation.
Under the current status quo of the 27 member states of the
enlarged European Union, countries act unilaterally and non-
cooperatively to decide on national asylum recognition rates by
maximizing their (hypothetical) country-specific objective func-

tion. To simulate the non-cooperative outcome of the status quo, we
define for each EU-27 country i an objective function of a similar
type to the welfare function used earlier. The (non-cooperative) wel-
fare in the status quo for each member state is then given by
Wju ( j
j ) j j
j j ¦ ( i
i ) ( j
j ) j , ji 1,! ,27 (6.17)
i\ j
with rj as the country-specific asylum recognition rate, uj as the

number of asylum applications, aj as the parameter for the country-
specific attitude towards immigration, bj as the country-specific
spillover (altruism) parameter, and gj as the country-specific cost
exponent. As before, the function Wju captures the three elements:
country-specific (private) benefits, (public) cross-country spillovers,
and (private) costs for hosting refugees. We assume that member
states interact strategically and non-cooperatively, since country-
specific welfare depends on the asylum provision of all the other
member states. Furthermore, the member states’ choice parameter is
the asylum recognition rate as asylum application numbers are often
beyond the scope of national policies. Consequently, asylum recog-
nition rates are endogenous and the respective reaction functions
depend on refugee admission levels of the other member states.
As an alternative to this (non-cooperative) status quo outcome, we
estimate the potential for an EU-wide (cooperative) burden-sharing
rule. Any binding rule requires the support of all 27 member states
for its implementation (and enforcement). Thus, and different to the
previous analysis, we do not look for the cooperative refugee admis-
sion numbers that maximize collective EU-wide welfare, but for the
cooperative burden-sharing rule which is supported by most (at its
best, by all) member states. A cooperative burden-sharing rule modi-
fies welfare function (6.17) according to:
Wjc ( : j ) j : j j ¦ :i ( : j ) j , ji 1,! ,27 (6.18)

i\ j
where Vj indicates the country-specific number of refugees to be

admitted under a specific burden-sharing rule. We consider two
alternative burden-sharing rules, which are (i) an equal population
quota in terms of admitted asylum seekers per inhabitant, and (ii) an
equal asylum burden with respect to the member states’ economic
performance (c.f. Chapter 3). A cooperative EU-wide burden-sharing

rule becomes authoritative for all member states if the respective
supranational authority, the Council of the European Union, achieves
the required majority under its legislative statutes. In this analysis,
we refer to two voting alternatives in the EU Council: the unanimity
rule under the present status quo, and according to the design of the
EU reform treaty, the double majority voting principle.
Until now, decisions of the EU Council on asylum policy issues
have required unanimity, that is, each member state retains its sov-
ereignty to block any cooperative asylum burden-sharing scheme
that is not unilaterally beneficial. Obviously, this unanimity voting
rule is the core obstacle for more cooperative asylum solutions in the
EU; and even more so since the recent enlargement.76 However, the
design of the prospective EU reform treaty tackles this issue and
envisages the implementation of the double majority principle,
which would then also be applicable for asylum policy affairs. The
double majority principle requires the backing of at least 15 member
states, representing a minimum of 65 percent of EU population. This
modification facilitates the potential for effectual coalitions for a
cooperative asylum regime below unanimity. We apply this modi-
fied voting principle for each of the proposed burden-sharing rules
and we ‘forecast’ their expected outcome in a stylized vote of the EU
Council.
The data we use for this numerical simulation are provided in
Table A.4. Data on asylum applications (ui) and recognition rates (ri)
for the period 1993–2003 are provided by UNHCR (2004), popula-
tion data are from the World Bank (2004), and data on country-
specific attitudes on asylum and immigration (ai) and humanistic
attitudes in general (bi) are taken from the World Values Survey
(WVS, 2000). In its 1999 survey, conducted in 26 of today’s 27 EU
member states (the exception being Cyprus for which we take the
average of the other 26 countries), the WVS asks interviewees about
their attitude towards asylum and immigration and about their gen-
eral concern about human beings. Country coverage and the rele-
vance of content convinced us to use these country-specific data as
indicators for domestic benefit and cross-country spillovers.
Additionally, we use unemployment rates of foreign-born population
taken from OECD (2006) for adjusting the indicator for country-
specific asylum and immigration attitudes.
First, we determine the country-specific welfare functions Wju of

the status quo. Given the fact that all countries act non-cooperatively
in their asylum policies, we assume that during the observation
period the average number of admitted refugees (rj uj) maximized the
country-specific welfare of each member state. This presumes that
EU member states were (and still are) in a stable non-cooperative
Nash equilibrium concerning refugee admission. This enables us to
calculate the country-specific cost parameters gj (see Column VI,
Table A.4) for each country. Both welfare functions are now fully
specified, and we are able compare the respective welfare outcomes
of the two alternative cooperative burden-sharing rules with the out-
come of the status quo. For both burden-sharing rules, we run the
numerical simulation for four different levels. For the population
quota, we display levels between 0.1 and 0.4 refugees per thousand
inhabitants, and for the economic burden-sharing rule levels between
0.005 and 0.012 refugees per million US dollar GDP with the corre-
sponding country-specific number of refugees to be admitted. For
c
each country j, we calculate the numerical welfare benefit Wj (Vj) 2
u
Wj (rj uj) and indicate the hypothetical voting outcome for both bur-
den-sharing rules.
The results for each burden-sharing alternative are shown in
Tables 6.1 and 6.2. Each table displays the net benefit of the
non-cooperative outcome under the status quo (Column I) and the
outcomes for the different levels of the respective cooperative burden-
sharing rule. It becomes evident that under each burden-sharing
rule, unanimity of all EU-27 member states is not reached. Although
only a few countries (Italy, Spain and Portugal) would oppose an
equal EU-wide economic burden-sharing rule (refugees per GDP,
Table 6.2), without any financial transfers or other package deals, a
cooperative asylum burden-sharing solution is not implementable
under the current status quo. However, under the modified voting
constitution of the proposed EU reform treaty, asylum cooperation is
profitable for a ‘double majority’, that is, for at least 15 member states
representing at least 65 percent of the EU population. Then, voting
on this economic burden-sharing rule would have the potential for
the rule to be implemented. The prospects for the alternative burden-
sharing rule, the population quota, are less promising.
Table 6.1 shows that applying this rule (that is an equal EU-wide
population quota) does not even satisfy the double majority
Table 6.1 Asylum cooperation: equal population quota, EU-27
Population status
quota quo 0.1 0.2 0.3 0.4

EU-27 benefit Vote Refugee benefit Vote Refugee benefit Vote Refugee benefit Vote Refugee benefit
Australia 0.421 0 805 0.196 1 1,610 0.017 1 2,414 0.222 1 3,219 0.42
Belgium 0.606 0 1,033 0.282 1 2,067 0.035 1 3,100 0.347 1 4,133 0.66
Bulgaria 0.853 0 797 0.406 1 1,593 0.026 1 2,390 0.449 1 3,186 0.86
Cyprus 0.705 0 77 0.335 1 153 0.032 1 230 0.399 1 306 0.76
Czech Republic 0.750 0 1,021 0.368 0 2,042 0.013 1 3,063 0.326 1 4,084 0.65
Denmark 1.148 0 537 0.530 1 1,075 0.085 1 1,612 0.698 1 2,150 1.31
Estonia 0.287 0 136 0.150 0 272 0.051 1 407 0.012 1 543 0.04
Finland 1.074 0 520 0.508 1 1,040 0.049 1 1,560 0.603 1 2,080 1.15
France 0.610 0 5,949 0.278 0 11,897 0.027 1 17,846 0.178 1 23,794 0.35
Germany 0.469 0 8,250 0.193 0 16,499 0.012 1 24,749 0.117 1 32,998 0.21
Greece 1.637 0 1,063 0.809 0 2,126 0.036 1 3,189 0.700 1 4,252 1.40
Hungary 0.560 0 1,016 0.267 1 2,032 0.022 1 3,048 0.309 1 4,064 0.59
Ireland 1.408 0 392 0.668 1 784 0.062 1 1176 0.787 1 1,568 1.51
Italy 0.572 0 5,769 0.391 0 11,538 0.499 0 17,307 0.790 0 23,076 1.23
Latvia 0.361 0 234 0.220 0 468 0.189 0 701 0.267 0 935 0.45
Lithuania 0.186 0 347 0.121 0 694 0.124 0 1,041 0.187 0 1,388 0.30
Luxembourg 0.530 0 44 0.253 1 89 0.024 1 133 0.300 1 178 0.58
Malta 0.685 0 40 0.327 1 79 0.032 1 119 0.389 1 159 0.75
Netherlands 0.708 0 1,614 0.325 1 3,229 0.048 1 4,843 0.416 1 6,458 0.78
Poland 0.196 0 3,863 0.202 0 7,725 0.347 0 11,588 0.582 0 15,450 0.89
Portugal 1.139 0 1,018 1.012 0 2,035 1.798 0 3,053 3.428 0 4,071 5.87
Romania 0.419 0 2,230 0.517 0 4,460 1.129 0 6,690 2.136 0 8,920 3.49
Slovakia 0.593 0 538 0.281 1 1,076 0.027 1 1,614 0.332 1 2,152 0.63
Slovenia 0.770 0 196 0.367 1 393 0.026 1 589 0.413 1 786 0.79
Spain 0.765 0 4,092 0.493 0 8,183 0.535 0 12,275 0.790 0 16,367 1.22
Sweden 0.869 0 892 0.425 1 1,785 0.009 1 2,677 0.436 1 3,570 0.86
United Kingdom 0.937 0 5,923 0.373 1 11,846 0.142 1 17,769 0.630 1 23,692 1.10
Total 19.256 48,394 10.297 96,788 4.123 14,5182 0.117 193,576 1.95
EU-27 consent (no transfers) 0/0.00 15/0.30 20/0.64 20/0.64
Notes: Population quota in refugees per thousand inhabitant; ‘Status quo’ refers to year 2004. EU-27 consent: number of countries/EU population
share.
Table 6.2 Asylum cooperation: equal economic burdens, EU-27
GDP status
shares quo 0.005 0.008 0.010 0.012

EU-27 benefit Vote Refugee benefit Vote Refugee benefit Vote Refugee benefit Vote Refugee benefit
Australia 0.421 0 1,299 0.148 1 2,078 0.004 1 2,598 0.10 1 3,117 0.196
Belgium 0.606 0 1,616 0.215 1 2,586 0.013 1 3,232 0.16 1 3,879 0.313
Bulgaria 0.853 0 110 0.315 1 176 0.007 1 220 0.22 1 264 0.436
Cyprus 0.705 0 58 0.261 1 92 0.005 1 116 0.18 1 139 0.359
Czech Republic 0.750 0 468 0.278 1 749 0.000 1 936 0.19 1 1,124 0.369
Denmark 1.148 0 1,095 0.405 1 1,752 0.037 1 2,190 0.33 1 2,628 0.624
Estonia 0.287 0 48 0.107 0 77 0.001 1 96 0.07 1 115 0.137
Finland 1.074 0 855 0.396 1 1,368 0.003 1 1,709 0.27 1 2,051 0.529
France 0.610 0 9,033 0.239 0 14,452 0.091 0 18,066 0.01 1 21,679 0.056
Germany 0.469 0 12,659 0.162 0 20,254 0.066 0 25,318 0.02 1 30,381 0.005
Greece 1.637 0 889 0.626 0 1,423 0.040 1 1,779 0.34 1 2,134 0.724
Hungary 0.560 0 425 0.205 1 680 0.008 1 850 0.15 1 1,020 0.291
Ireland 1.408 0 672 0.522 1 1076 0.000 1 1,345 0.34 1 1,614 0.688
Italy 0.572 0 7,580 0.434 0 12129 0.572 0 15,161 0.73 0 18,193 0.926
Latvia 0.361 0 64 0.136 0 102 0.005 1 128 0.08 1 153 0.165
Lithuania 0.186 0 101 0.070 0 162 0.005 1 203 0.04 1 243 0.079
Luxembourg 0.530 0 125 0.195 1 200 0.003 1 250 0.13 1 300 0.266
Malta 0.685 0 24 0.254 1 38 0.005 1 48 0.18 1 58 0.350
Netherlands 0.708 0 2,594 0.242 1 4150 0.028 1 5,187 0.20 1 6,224 0.381
Poland 0.196 0 1,186 0.083 0 1897 0.027 1 2,372 0.01 1 2,846 0.037
Portugal 1.139 0 724 0.657 0 1158 0.599 0 1,447 0.65 0 1,737 0.774
Romania 0.419 0 329 0.158 0 526 0.012 1 658 0.08 1 789 0.174
Slovakia 0.593 0 174 0.218 1 279 0.006 1 349 0.16 1 418 0.304
Slovenia 0.770 0 146 0.285 1 233 0.003 1 291 0.19 1 349 0.385
Spain 0.765 0 4,405 0.437 0 7048 0.421 0 8,809 0.46 0 10,571 0.542
Sweden 0.869 0 1,599 0.323 0 2559 0.008 1 3,198 0.20 1 3,838 0.404
United Kingdom 0.937 0 9,959 0.266 1 15935 0.085 1 19,919 0.31 1 23,902 0.519
Total 19.256 58,237 7.636 93179 1.638 116,474 2.06 139,768 5.549
EU-27 consent (no transfers) 0/0.00 15/0.30 22/0.48 24/0.78
Notes: GDP shares in refugees per million US $ GDP; ‘Status quo’ refers to year 2004. EU-27 consent: number of countries/EU population share.
requirement. In this case, a common population quota could only be

feasible if it is amended by a suitable monetary compensation regime.
Thus, asylum cooperation was not realized in the past because equal
burden-sharing is not profitable to all EU member states.
The design of the supplementary monetary compensation regime
depends on the characteristics of the heterogeneity among the EU
countries. As stated above, monetary transfers that compensate for
spillover effects (b) might enforce a cooperative outcome of a par-
ticular burden-sharing rule only among countries with similar asy-
lum costs. However, since this presumption of cost symmetry among
member states was weakened through the recent EU enlargement,
such monetary transfers seem to be a rather inefficient instrument.
Instead, a monetary regime that compensates for cost asymmetries
among EU member states would be a more appropriate approach to
increase the incentives for a cooperative EU-wide asylum solution.
This chapter argues that refugee protection levels are insufficient

and suboptimal because of the absence of any institutionalized and
efficient refugee cooperation regime. It has been shown that because
of the public good character of asylum provision, with its externality
spilling over across other countries, non-cooperative deflection and
free-riding become collectively welfare diminishing. Asylum cooper-
ation is desirable, but it is less attainable the more heterogeneous the
affected countries are. Depending on the type of heterogeneity – cost
versus spillover asymmetries – monetary transfers may be able to
increase country-specific welfare and stabilize a cooperative asylum
regime. In general, this only holds among countries with minor
asymmetries. We argue that this condition for asylum cooperation is
rather satisfied on a regional level, for instance, among the member
states of the European Union. Since the implementation of a multi-
lateral asylum (burden-sharing) regime requires a supranational
institution which has the competence (i) to guarantee the steady
commitment of the participating countries, (ii) to coordinate the
allocation of refugee admission numbers, and (iii) to implement, if
efficient, an adequate monetary compensation scheme across
participating countries, the European Union is even more appropri-
ate because of its already established institutional authority. However,
the status quo requires the unanimity of all member states to enact
binding decisions of the EU Council on asylum affairs. This has not
been attained so far, resulting in the non-cooperation of the EU
member states on refugee admission issues. Since the European
Refugee Fund (ERF) does not take into account the heterogeneities in
asylum costs among the EU member states, which have further
increased since the recent enlargement, it is not adequately designed
to fundamentally change this non-cooperative outcome. Consequen-
tly, based on the implications of this paper, we can predict that the
ERF will have little impact on a more cooperative asylum outcome
within the EU, even if its financial volume is multiplied. For the
future, however, we are optimistic that cooperative asylum in the EU
is possible. The authority of the upcoming EU reform treaty and the
introduction of the double majority principle in the EU Council are
good reasons that cooperative asylum outcomes can become possible
in the EU – at least by means of an appropriate financial transfer
regime.
7
Refugee Movements and
Aid Politics
Abstract
This chapter links the topic of refugee migration to the issue of for-
eign aid. It provides an overview on aid provision and allocation
relevant for the subsequent chapters that are focused on the influ-
ence of refugee migration movements on the aid allocation decision-
making process of bilateral donors.
7.1 Introduction
The motives of foreign aid provision vary from donor to donor and
also over time. The spectrum of motives for giving official aid is
broad, ranging from ethical and humanitarian motives, to politi-
cal and security interests, and to economic and commercial con-
siderations. Former (special) interests, for example those of colonial
powers to give foreign aid to their former colonies, or influencing
voting patterns in the UN, have declined in relevancy, while new
objectives like the promotion of global public goods (for example,
sustaining and improving the global environment, reducing the
risks of epidemics such as HIV/AIDS or Malaria, hindering the flow
of narcotics or fighting international terrorism) have risen. In this
context, one of the new aid allocation objectives of the post–Cold
War era is the management of international asylum and refugee
movements, which is of particular interest of the remaining
chapters.
114
Refugee Movements and Aid Politics 115
7.2 Aid and refugee migration policies
Since the beginning of the 1990s, which saw an arising ‘asylum cri-
sis’ in Europe, the international debate concerning the interactions
between migration and aid policies has advanced, particularly in
donor countries. Western policymakers across the political spectrum
now discuss the use of development policy as an instrument for the
management of an increasing migration and asylum pressure.
Before the asylum crisis of the 1990s, donor governments were
rather reserved in the proactive use of short-term humanitarian aid
or long-term development assistance in refugee situations. Previously,
aid was provided for tackling large displacements in only a few spe-
cific cases and then for limited periods (Klingebiel, 1994). In the
meantime, however, this reluctance has changed. Aid interventions
are often considered as important instruments for tackling the causes
and consequences of refugee movements, and to assist the (re)inte-
gration and rehabilitation processes of refugees and returnees. For
instance, the German Ministry for Economic Development and
Cooperation (BMZ) views refugee-related development assistance as
an instrument for avoiding the causes and negative consequences of
large refugee movements (BMZ, 2006a; BMZ, 2006b, p. 292).
Although the effectiveness of aid for the prevention or cessation of
conflicts is still ambiguous (Collier, 2007), there are nevertheless high
expectations of the potency of development aid as a central policy
tool in refugee migration issues. It becomes increasingly important to
condition aid provision on good governance in terms of compelling
recipient countries to guarantee, for example, respect for human
rights, participation of all social and ethnic groups in political
decision-making, the rule of law and provision of legal security, or
progress in establishing a market-based and development-oriented
economic and social order. In this context, the reliability and the
time-consistent behavior of donor governments is most important for
improving the credibility of the announced policy measures of aid
donors (Collier, 2007).
Since the causal link between aid provision and good governance
reform is often not robust and ex ante conditionality is also often inef-
fective (Collier, 2007), Western donor governments are giving more
weight to mitigating the consequences of large refugee movements,
and thus, increasingly consider local integration and reintegration as

durable solutions.77 Basically, the underlying intention of Western
donor states, beside altruism, is primarily to limit mass refugee move-
ments to the region of origin before they move towards Western bor-
ders. Hence, humanitarian aid and development assistance contributes
to a stabilization of fragile regions and conflict areas (Zolberg et al.,
1989).
A sustainable mitigation of the consequences of mass flight move-
ments includes the repatriation, reintegration, and rehabilitation of
refugees and other displaced persons. Most generally, priority is given
to return options rather than a prolonged support in the first asylum
country. Obviously, repatriation and (post-conflict) reconstruction
requires that returnees take an active part; however, as a durable solu-
tion it is only attractive if international donors assist in creating hous-
ing, workplaces and basic infrastructure (Black & Koser, 1999).
For acting more proactively, it becomes increasingly important
for Western donors to link short-term relief aid, mid-term (re)inte-
gration efforts, and long-term (post-conflict) reconstruction and
economic promotion for closing the time gaps between the differ-
ent forms of aid provision (Crisp, 2001). In addition, new activities
to create conditions for a sustainable return movement of refugees,
but which do not correspond to the traditional conception of devel-
opment assistance, are required of international donors. These
activities include removing landmines, integrating demobilized
soldiers in aid programs and the enforcement of housing and
property rights (Black & Koser, 1999). For that purpose, aid inter-
ventions in (post-)conflict situations often go hand in hand with
political siding in favor of the target group. In order to become
active, compromises with state armed forces – and also with para-
military groups or even clan militia – are often unavoidable (de
Mello, 1995).
Secondary movements to and asylum seeking in Western states
are caused by a decelerated or failed (post-conflict) reconstruction
process in the country of origin and desperate situations in first
asylum countries. Aid interventions, inter alia, will reduce these
unwanted migration flows (see Chapter 5). However, with respect to
the high complexity of refugee migration movements in general
and secondary movements in particular, some evidence shows that
aid assistance is often rather ineffective in meeting secondary
migration prevention objectives, at least in the short run (Rotte

et al., 1997). Therefore, short-term interventions for fighting struc-
tural migration causes are very likely to fail. This applies to various
emigration causes such as ecology (environmental degradation),
economy (poverty), demography (population growth), or politics
(civil wars, human rights’ violations and so on). Even if donor states
take the task of mitigating the causes of refugee and secondary
migration movements seriously, they might only become effective
in the long term (Rotte et al., 1997; Burgdorff, 2002).
Nevertheless, the interest of Western governments in using devel-
opment policy for migration prevention objectives is undiminished.
Particularly the EU member states have emphasized the need to use
foreign aid to reduce migration and refugee flows from different
parts of the world into the EU or, as an alternative approach, to
reward other states (either within or outside the EU) that accept refu-
gees from other sending or first asylum countries (Castles, 2003). The
ambition to manage unwanted migration movements by means of
various forms of aid is present in both public debates and national
policies.
For instance, this is shown by a recent public poll run by the
European Commission in spring 2007, where EU citizens were asked
what they consider to be the major motivations for providing devel-
opment aid to poor countries (EC, 2007). In most European coun-
tries, citizens state that the main rationale for development aid is
based on the self-interest of the donor countries, such as improving
(bilateral) trade relationships, contributing to global and regional
stability or promoting good governance in the recipient countries.
Avoidance of emigration movements from development countries to
industrialized countries ranks fourth, while only 11 percent of EU
citizens believe that the main motivation for giving development aid
is to help people in need (Figure 7.1).
This EU-wide poll shows that particularly in the large European
migration destinations, France and Germany, using aid for migra-
tion prevention purposes is highly ranked. But also in the other EU
countries, avoiding large migration movements from poor devel-
oping countries to industrialized countries seems to be a major
motivation of donors for allocating aid for refugee-related pur-
poses. We will come back to this implication in the subsequent
chapters.
45
40
35
30
25
%
20
15
10
5
0
EU 27 Germany France Italy Sweden UK
Economic self-interest 28% 23% 31% 18% 41% 35%
Global stability 28% 38% 25% 26% 33% 28%
Democracy/governance 22% 29% 16% 32% 30% 28%
Avoid emigration 20% 24% 27% 17% 10% 10%
Prevent terrorism 19% 24% 14% 28% 17% 17%
Gain political allies 15% 13% 22% 6% 10% 16%
Clear conscience’ 12% 12% 19% 9% 16% 10%
Poverty reduction 11% 5% 9% 19% 13% 1%
Don’t know 9% 4% 7% 7% 6% 13%
Figure 7.1 Donor motivations of ODA provision, EU public poll,78 2007

Note: Data are based on EC (2007).
7.3 Aid allocation patterns and statistics
The international aid business is complex in a sense that participat-

ing actors often have diverse interests and motivations, unequal bar-
gaining power, and very different means and instruments. The
central actors in development cooperation are national states, multi-
lateral institutions, and non-governmental organizations. On the
donor side, the subsequent analysis primarily focuses on official (or,
governmental) bilateral aid organizations and the donor govern-
ments to investigate the underlying motivations for the allocation of
Official Development Aid (ODA) or humanitarian aid. On the recip-
ient side, national states are generally the main cooperation partners
of donor states. This means that the whole analysis is state-centered
and does not take into consideration the aid allocation procedures
on sub-national levels.
120
100
80
Total ODA
60 A. Bilateral ODA
A.1. Grants
40 A.2. Non Grants
B. Multilateral ODA
20
−20
1980
1982
1984
1986
1988
1990
1992
1994
1996
1998
2000
2002
2004
2006
Figure 7.2 Net disbursement of ODA, OECD/DAC countries, current US$ (in bn)
Note: Figure is based on data provided by OECD (2007a).
Based on data provided by OECD (2007a), Figure 7.2 displays total

amounts of Official Development Assistance contributed since 1980.
It is remarkable that net disbursements of bilateral and multilateral
aid from OECD/DAC countries to developing countries have almost
doubled since the end of the Cold War. Total official development
aid in 1990 was about US$ 54.3 billion; in 1997, total aid distribution
reached its lowest level during the 1990s of around US$ 48.5 billion
(in current US$). Since then, total aid levels have increased to more
than US$ 104 billion in 2006. Furthermore, it is shown that most
ODA is channeled through bilateral grants; multilateral aid makes up
only around a third of the total ODA amount.
This preference of donor states for bilateral aid allocation is also
evident in Figure 7.3. In 2006, the top ten donor states allocated
between 55 percent (Italy) and 90 percent (USA) of their total ODA
transfers through bilateral channels. The figure also shows the
importance of the five largest bilateral donors. The United States, the
United Kingdom, Japan, France and Germany contribute more than
two thirds of the total global ODA volume. In absolute terms for
2006, these five countries transferred around US$ 68.2 billion to the
developing world. However, in relative terms, the picture is less admi-
rable. Comparing the absolute aid volumes with the respective
economic performance reveals that these five, but also most other
DAC donor states, are far from realizing the target of 0.7 percent as
Other DAC
Norway:
donors: 16.5
3.0 (74%)
Italy: United States:
3.6 (55%) 23.5 (90%)
Canada:
3.7 (69%)
United Kingdom:
Sweden: 12.5 (70%)
4.0 (72%)
Netherlands:
5.5 (79%) Germany: Japan:
10.4 (67%) 11.2 (65%)
France:
10.6 (75%)
Figure 7.3 Top bilateral donor states in 2006, in current US$ (in bn; share of
bilateral ODA)
Note: Figure is based on data from OECD (2007a).
ODA–GNP ratio, which was passed by the UN General Assembly in

1970 as the goal for aid budgets of Western donor states.
The regional distribution of global aid flows, shown in Figure 7.4,
indicates that over the past decade there has been a strong trend
for reallocating aid volumes to the African continent. While, in
1996, sub-Saharan Africa received around a third of the global aid
budget, ten years later it had increased to around 45 percent of the
increased global ODA volume. This reallocation of ODA in favor of
sub-Saharan Africa was mainly at the expense of Eastern Asia, and
Central and South America. These two regions lost between 6 and
7 percent of total ODA, while other regions’ shares largely remained
constant. However, the geographic allocation of development aid
varies considerably among donor states. While ODA from former
European colonial powers is relatively concentrated to ex-colonial
states, the United States primarily allocates on the basis of national
and international security considerations, and Japan primarily allo-
cates to Asian recipients to further commercial interests (Degnbol-
Martinussen & Engberg-Pedersen, 2003).
Generally, in order to qualify for Official Development Assistance
(ODA), aid transfers from donor countries to recipients in the
Oceania Eastern Europe

4% 4% Central &
South America
15%
Sub-Sahara
1996 Africa
Middle East &
33%
North Africa
17%
South &
Central Asia East Asia
13% 14%
Eastern
Oceania Europe
1% 6% Central &
South America
8%
Middle East &

Sub-Sahara
2006 North Africa
Africa
19%
45%
East Asia
South & 8%
Central Asia
13%
Figure 7.4 Regional distribution of ODA, 1996 and 2006, current US$
Notes: Chart sizes are proportional (that is, in 2006 ODA was around double that of
1996) to the respective total ODA volumes of all DAC countries in that year. Figures are
based on data provided by OECD (2006a).
developing world or to multilateral agencies must fulfill three basic

conditions: first, they must be provided by the official sector, second,
its main objective needs to be economic development and welfare
promotion, and third, they are generally provided on concessional
terms, that is with at least a 25 percent grant element (OECD, 2007b).
Besides financial transfers, technical assistance as well as long-term
humanitarian aid is also included in official aid statistics. There have
been several modifications in the definition of ODA by the DAC. For
instance, already since the early 1990s, financial and technical assis-
tance to refugees in developing countries as well as temporary
expenses to refugees from developing countries during the first
12 months of stay in the donor country, and all costs associated with
eventual repatriation to a developing country, are reportable by the
donor countries as ODA (OECD, 2007b).
However, it is still rare that donor states report their respective
figures of refugee-related aid, only some rough estimates are available
indicating a share of about two percent (Table 7.1). Probably because
Table 7.1 Sector composition of total ODA grants, 3-year average, current
(bn) US$
25
20
15
10
0
1995–97 1998–00 2001–03 2004–06
Technical Co-operation 13.8 13.0 15.8 20.6
Project & Programme Aid 8.5 7.5 8.6 17.0
Debt Forgiveness 3.4 2.4 5.1 16.9
Humanitarian Aid 1.9 2.6 3.0 6.4
Administrative Costs 2.8 3.0 3.2 4.1
Refugees in Donor Country 0.7 1.0 1.4 2.0
Support to NGO’s 1.0 1.1 1.3 1.9
Food Aid 1.1 1.0 1.1 1.0
Note: Figure is based on data from OECD (2007a).

of better reporting by donor states, this figure has also increased tre-
mendously since the mid-1990s. In general, figures for the size of
official aid must be taken cautiously; they are sometimes highly
politicized and several donors have attempted to inflate the amount
of their aid assistance by including activities not strictly categorized
as ODA (Degnbol-Martinussen & Engberg-Pedersen, 2003).
Basically, the sectoral composition of ODA has also undergone sig-
nificant changes during the previous decade. Table 7.1 displays three-
year averages for the core sectors of ODA grants, indicating that
technical cooperation is still the most important sector in bilateral
development assistance. Aid devoted to this sector rose to more than
one fifth of total spending. Debt forgiveness, which is the major
cause of the tremendous increase of ODA volumes in recent years,
has more than tripled since 2000.
7.4 UN diplomacy and aid allocation
This section examines correlations between bilateral and multilat-

eral aid-giving patterns and the number of speeches country repre-
sentatives delivered in the two core bodies of the United Nations, the
UN Security Council and the UN General Assembly, respectively.79
For both developing and developed countries, the United Nations
bodies play an integral role as a forum of international communica-
tion and diplomacy, as well as arranging international partnerships
serving country-specific concerns. Particularly, the UN Security
Council and the UN General Assembly are the core arenas where
both developing and developed countries can receive international
attention for their issues. These two bodies offer an international
audience for country representatives and other delegates to promote
their country-specific political affairs and economic urgencies. Once
on the agenda, a country-specific concern has the diplomatic forum
to be promoted to its special needs. However, do these diplomatic
activities actually correspond to the political, economic or humani-
tarian efforts of bilateral and multilateral donors of aid?
This section seeks to analyze this issue by focusing on the question
whether UN diplomacy is in line with the transfer pattern of devel-
opment aid for countries in need. Hereunto, we investigate the cor-
relations between the number of speeches in the UN Security Council
and the UN General Assembly in the 1990s and the bilaterally and
multilaterally allocated development aid funds for those countries.

We presume that the more often delegates get the chance to speak
either in the UN Security Council or in the UN General Assembly for
country-specific concerns, the more these countries are also privi-
leged by financial and technical support in terms of bilateral or mul-
tilateral Official Development Assistance (ODA).
The literature on aid allocation determinants has become increas-
ingly voluminous in recent decades. A common approach in the
empirical literature differentiates between donor self-interests and
recipient needs as the main explanatory factors for bilateral and mul-
tilateral aid patterns. In particular, McKinlay & Little (1977, 1978)
and Maizels & Nissanke (1984) represent this concept, reflecting the
idea that aid-giving motivations of donor countries might be neither
entirely altruistic nor entirely selfish, which is why aid allocations
can be expected to be both promoting donor economic and political
interests as well as the economic and technical needs of recipient
countries. In this context, two well-known biases are extensively dis-
cussed. First, the population bias reflects a discrimination of more
populous countries in aid allocation patterns, and second, the
middle-income bias refers to a concave relationship between aid
and income per capita of a recipient country, especially disfavoring
the least developed countries.80 More recent studies have proposed
some additional aid allocation determinants such as good gover-
nance and human rights rewarding effects (for example, Svensson,
2000; Alesina & Weder, 2002; Neumayer, 2003), or expressing the
importance of socio-political stability (Chauvet, 2003). Collier &
Hoeffler (2004) provide some evidence for the discrimination of
post-conflict societies, whereas Czaika (2005a; see Chapter 8) dis-
cusses the influx of asylum seekers and refugees as an influential
determinant in bilateral aid decisions.
In the following, we take some of these previously tested variables
into consideration and analyze the bilateral aid allocation patterns
of 18 donor countries to 148 recipient countries, as well as the total
multilateral aid allocation between 1991 and 1999, and introduce
the number of speeches held in both the UN Security Council and
the UN General Assembly, respectively, as the main variables of inter-
est. Furthermore, we control (i) for some donor self-interest variables
like trade promotion, colonial ties, geographical proximity, bilateral
development cooperation duration and country size in terms of
population, (ii) for some recipient need variables like income per cap-
ita, human development level and intensity of internal conflicts dur-
ing that time period, and (iii) for some good governance variables
like the respect for political freedom or the quality of institutional
capacities. We expect a positive correlation of our two main variables
of interest, that is the number of speeches in the UN Security Council
and the UN General Assembly, and the amount of aid developing
countries received during the 1990s. Since we consider the UN
Security Council to be the more powerful arena, we anticipate a
stronger correlation with those concerns raised in this arena com-
pared to the UN General Assembly, which receives in general less
global publicity. Hence, we expect that UN Security Council speeches
are more important in reflecting the international development
assistance pattern than its UN General Assembly counterpart.
7.4.1 Data
All data on ODA flows are taken from the OECD Geographical
Distribution of Financial Flows to Aid Recipients database (OECD,
2004).81 We use the nine-year average of the total amounts allocated
on a bilateral basis from 18 donors to 148 recipient countries and
alternatively the total amount of ODA disbursed multilaterally. UN
Security Council (Scouncil) and General Assembly (Gassembly)
speeches made by country representatives or delegates are indexed in
the UN Bibliographic Information System (United Nations, 2005b).
Figures for Income per capita and Population were taken from World
Development Indicators (World Bank, 2004). We use the physical
quality of life index (PQLI) as a control variable for human develop-
ment needs, which is calculated, based on data from World Bank
(2004), according the formula provided by Neumayer (2003, p. 51).
Possible good governance rewarding effects are captured twofold.
First, we control for political freedom allowance, represented by the
Democracy variable, which is measured as the equally weighted sum
of the political rights and civil liberties index from FreedomHouse
(2005).82 Second, good Governance reflected in the country-specific
politico-institutional quality is calculated as an equally weighted
composite index based on governance data from Kaufmann et al.
(2003).83 In addition, we take into account the severity of past or run-
ning internal clashes by Conflict intensity data extracted from the
Armed Conflict dataset of PRIO (2005). In addition, we control for
some bilateral factors like the development cooperation history (ODA

history) between donor and recipient countries before 1990 in num-
ber of years and for former colonial ties (Colony). A geographical
proximity (Distance) variable controls for close political and eco-
nomic interest spheres, and lastly, bilateral trade volumes (Trade) are
used in regressions to represent the special commercial interests of
donor countries. Data on geographical distances are taken from
CEPII (2005), whereas export data for the 18 donor countries are
taken from the Bilateral Trade Database of OECD (2005).8
We run the regressions with a pooled least squares fixed effects
model with heteroscedasticity robust standard errors estimating the
following regression equation:
ODAij a0 f(Scouncil, Gassembly; b) f(Xij, Qj; g) «ij (7.1)
where ODAij denotes the total amount of aid given by country i to

country j, the Scouncilj and Gassemblyj variables express the number
of speeches held in the respective arena in favor of country j, Xij is a
vector of bilateral control variables and Qj is a vector reflecting uni-
lateral control variables. b and g are estimation coefficient vectors,
and «ij stands for the error term. Since the number of speeches and
the amounts of aid might have similar underlying determinants, we
have possibly a problem of endogeneity; thus, we cannot identify
causality, but only correlation between the ODA allocation pattern
and the pattern of speeches held in these two UN bodies.
7.4.2 Results
Table 7.2 reports the regression results for both the bilateral aid alloca-
tion decisions (Regressions I–III) and the multilateral aid allocation
patterns (Regression IV). The regressions for the bilateral allocations
display throughout not unambiguous effects of speeches in UN bodies
in reflecting international aid support for countries in need. Speeches
in the UN Security Council correlate significantly with both bilateral
aid and multilateral aid allocations. On the other hand, the UN General
Assembly variable exhibits insignificant (negative) coefficients. During
the 1990s, about 30,000 UN General Assembly speeches on concerns
of the 148 recipient countries in our dataset (approximately 200 on
average) had no measurable positive correlation, neither with bilateral
nor with multilateral development aid transfer patterns.
Table 7.2 Bilateral and multilateral aid allocation. Dependent variable:

log of ODA
I II III IV
Constant 2.363 1.573 1.308 1.945

(0.897) (0.580) (0.492) (0.681)
Scouncil 0.116* 0.091* 0.138**
(2.540) (1.973) (2.889)
Gassembly 0.182 0.111 0.076
(1.663) (1.038) (0.616)
Income 1.614** 1.587** 1.673** 1.895**
(3.052) (2.972) (3.271) (3.300)
Income (sq.) 0.154** 0.154** 0.155** 0.184**
(4.474) (4.419) (4.699) (4.561)
Population 0.290 0.162 0.226 0.430
(1.229) (0.681) (0.951) (1.712)
Population (sq.) 0.005 0.002 0.001 0.007
(0.406) (0.151) (0.099) (0.539)
PQLI 0.218 0.301 0.463 0.795
(0.452) (0.599) (0.936) (1.765)
Democracy 0.335* 0.363* 0.328* 0.580**
(2.117) (2.304) (2.083) (3.132)
Governance 0.713* 0.594* 0.675* 0.671*
(2.330) (2.056) (2.283) (2.486)
Conflict 0.122 0.119 0.120 0.054
(1.846) (1.767) (1.734) (0.931)
ODA history 0.473** 0.459** 0.466** –
(8.803) (8.201) (8.373)
Trade 0.266** 0.264** 0.279** –
(5.650) (5.531) (5.827)
Distance 0.308** 0.311** 0.337** –
(3.124) (3.084) (3.339)
Colony 1.372** 1.365** 1.344** –
(9.878) (9.648) (9.688)
R2 adjusted 0.663 0.661 0.659 0.727
Number of 2592/144 2592/144 2592/144 148
observations
(pooled)
Notes: Regression I–III results are for 18 donor and 148 recipient countries, Regression IV
results for total multilateral aid to 148 recipient countries. t-values are shown in paren-
theses. (*) and (**) denote statistical significance at the 5 percent and 1 percent level,
respectively. Estimates for country dummies are available on request.
Table 7.3 lists donor-specific regression estimates for UN Security

Council speeches, which shows significant results for some bilateral
donors. It is striking that explicitly those donor countries that were
non-permanent UN Security Council members during the observed
time period have positive and significant estimates. We find for the
non-permanent UN Security Council members Sweden (member in
1997/1998), Italy (1995/1996), Spain (1993/1994, not significant),
Austria (1991/1992) and Canada (1989/1990, that is shortly before
the observation period) as well as for the permanent members France,
UK and USA (not significant) rather strong correlation, whereas for
the non-members Australia, Switzerland and Denmark (both even
with negative signs), Finland, Ireland, Netherlands and Norway esti-
mates are rather low and not significant.85
Table 7.3 Bilateral donor-specific coefficients. Dependent

variable: log of ODA
Donor Scouncil
Australia 0.032 (0.423)

Austria 0.182* (2.263)
Belgium 0.067 (0.995)
Canada 0.188** (2.599)
Switzerland 0.009 (0.133)
Denmark 0.039 (0.460)
Spain 0.131 (1.462)
Finland 0.121 (1.620)
France 0.186** (2.763)
UK 0.158 (1.887)
Germany 0.208** (2.944)
Ireland 0.088 (1.495)
Italy 0.214* (2.543)
Japan 0.027 (0.317)
Netherlands 0.091 (1.095)
Norway 0.091 (1.145)
Sweden 0.208* (2.444)
USA 0.164 (1.462)
R2 adj 0.670
Number of observations 2592/144
Notes: Estimates of the other explanatory variables are the same

as those given in Table 7.2. t-values are shown in parentheses.
(*) and (**) denote statistical significance at the 5 and 1 percent
level, respectively.
For the other control variables, we can confirm the middle-income

bias for both bilateral and multilateral aid, whereas a large country
bias is not evident in our regressions. The recipient need variables
show no significant signs indicating an inferior importance in the
aid decision-making process. However, we find some evidence for
the good governance and democracy rewarding effects, respectively,
for both aid allocation channels. For bilateral donors, we can show
some strong evidence for distinctive self-interest, reflected in higher
aid amounts where long-lasting development cooperation relation-
ships exist. Commercial, political and cultural ties in terms of spe-
cial preferences for trade partners, geographically close countries,
and former colonies are other factors of aid determination.
Thus, elaborating on the correlation of UN speeches delivered by
representatives and delegates of developing and transition countries
with the international allocation pattern of bilateral and multilateral
development assistance provides some indication for an underlying
relationship. The evidence for the hypothesis that the allocation of
development assistance follows the pattern of international diplo-
macy activities is rather weak – at least for the two main arenas of the
United Nations. However, if this hypothesis is valid at all, then it is
for the UN Security Council, where a positive relationship was found
for several bilateral donors as well as for total multilateral aid.
Correlation with diplomatic activities was particularly found for the
aid allocations of – at least temporary – UN Security Council mem-
bers. In contrast, diplomatic activities in the UN General Assembly
are a rather poor indicator for aid allocation decisions. Both bilateral
and multilateral aid allocations are not in line with the appeals put
forward through this body. This result indicates the superior repute
of the UN Security Council as a ‘donor institution’ compared to the
UN General Assembly. However, this rather weak result might also be
interpreted as an indicator of the different time perspectives of UN
diplomacy and international development assistance. While, con-
trary to diplomatic activities that might have been determined by
political urgencies, international aid provision seems rather influ-
enced by long-term considerations.
‘More aid’ is a permanent calling of development aid organizations

and other stakeholders, which has been repeated time and time again
at various UN bodies and international summits.86 More aid for first

asylum countries and countries of origin seems to be only a refine-
ment of this call.
For instance, the UNHCR Convention Plus Initiative promotes an
enhanced burden-sharing with refugee host countries and countries
of origin of refugee movements (UNHCR, 2004a). With respect to
countries of origin, facilitating a sustainable voluntary repatriation
by alleviating reintegration, rehabilitation and (post-conflict) recon-
struction is the primarily rationale for increased aid volumes. More
aid for countries of origin will ensure rapid post-conflict reconstruc-
tion with appropriate levels of civil security, basic social services and
better economic potential for those refugees willing to return
(UNHCR, 2007).
In fact, voluntary repatriation is often not possible for years, and
thus, protracted refugee situations are a major consequence. Refugees
are confined to camps and closed settlements mostly for economic
and security reasons. First asylum countries lack assistance (and
financial incentives) to overcome these political and economic
impediments to allow refugees greater individual freedom, such as
access to basic social support or the opportunity to unfold produc-
tive capacities and self-reliance in the host country (UNHCR, 2007).
Western donor states are generally highly interested in the process
of political stabilization, economic reconstruction and peace-
building in a (post-)conflict context. However, for various reasons,
such as free-riding interests, they are rather reluctant to increase
their aid funds for (refugee-related) challenges in the developing
world. Therefore, the remaining parts of this book discuss the main
determinants of bilateral aid allocation decision, and whether donors
respond to different types of refugee movements differently.
8
Aid Allocation and
Asylum Migration
Abstract87
In this chapter, the influence of immigrants and asylum seekers on

the aid allocation pattern of bilateral donors is investigated.
Immigrants constitute a respected interest group in the aid alloca-
tion process in the donor country and seek to influence bilateral aid
flows in favor of their country of origin. We distinguish between
immigrants (with foreign citizenship) living permanently in the
donor country and asylum applicants who are non-permanent in
their status, but who play an important role as information and
‘urgency indicators’. For Germany, some empirical evidence can be
found indicating that permanent immigrants as well as asylum seek-
ers influence the bilateral aid allocation pattern.
8.1 Introduction
Since the end of the Cold War and the demise of the Soviet Union,
transnational migration has become a major concern of Western
European governments. In particular, the large increase of asylum
seekers from the various developing and transition countries has had
a noticeable impact on European societies and politics. The policy
response of European governments to the growing influx of unwanted
immigrants from the South and East was twofold:
First, there were attempts to reduce immigration by legal means,

tightening asylum laws and cracking down on illegal migrants.
131
Second, the migration issue led to a reorientation of development

policies toward a broader perspective of development, including
economic, social and political elements. (Rotte et al., 1997)
This chapter focuses on the influence of immigration (permanent

and asylum) with respect to the aid allocation policy of a donor
country that is at the same time a migrants-receiving country.88 For
many donor countries, immigration in general and asylum migra-
tion in particular constitute very relevant determinants of bilateral
aid allocation that have never been specifically studied.89 Immigrants
(permanent and asylum) form a respected group in many societies,
especially in classical immigration countries like the US, Canada,
Australia, or New Zealand, but also in the non-classical immigration
countries of Western Europe. Worldwide, two-thirds of all asylum
applications were lodged in the EU-15 countries, and the absolute
number of applications confronting Western Europe has risen dra-
matically since the end of the Cold War (UNHCR, 2003).90 Among
the EU member countries, Germany plays a dominant role as a main
migration-receiving country. During the 1990s, Germany was con-
fronted with about 50 percent of all European asylum applications.
The total number of asylum seekers in Germany rose dramatically in
the early 1990s and then stabilized at a relatively high level
(Figure 8.1).
Germany, as one of the main havens for immigrants in Europe in
the 1990s, is thus an interesting case for studying the influence of
immigrants and asylum seekers on bilateral development aid alloca-
tion policy. We presume that Germany, as a major host country for
asylum seekers and other immigrants, concentrates its development
aid efforts, along with other criteria, on recipient countries that are
the main sending countries of asylum-seeking people to Germany.
In the aid allocation literature, the question of whether asylum seek-
ers have an impact on the aid allocation decision of their host coun-
try is a rather neglected one. This chapter seeks to fill this gap.
8.2 Aid allocation literature review
The aid allocation literature of the last thirty years offers numerous
determinants for explaining aid allocation patterns, with varying
degrees of relevancy. A cornerstone for explaining donors’ aid
Aid Allocation and Asylum Migration 133
500
450
400
350
300
250
200
150
100
50
0
1982
1984
1986
1988
1990
1992
1994
1996
1998
2000
1980
Germany France UK Italy Sweden
Figure 8.1 Asylum applications in EU countries (in 1000)

Source: Data provided by MPI (2005), http://www.migrationinformation.org/
allocation motivation is the recipient need and donor self-interest

approach (for example, Maizels & Nissanke, 1984). This approach
postulates that donors care about the needs of the recipients, but at
the same time also use aid to advance their self-interest. Based on
various empirical studies, some evidence exists that donor govern-
ments mainly pursue their political, economic and strategic self-
interests in the way they allocate aid. Objectives of poverty
reduction and development promotion are less important, especially
in the bilateral aid allocation of the large donors. The dominance of
self-interest in the motivation of donors partially excludes the ‘bet-
ter’ motivation of focusing on the recipients’ needs and constrains
poverty-oriented aid allocation (World Bank, 1998; Collier & Dollar,
2002). However, Trumbull & Wall (1994), Alesina & Dollar (2000),
McGillivray (2003b), and others show that even larger bilateral
donors favor developmental criteria and, especially since the end of
the Cold War, have refocused their aid support away from non-
developmental concerns to more poverty and need-oriented aspects
of the recipient countries (McGillivray, 2003a).
Beside these indigence factors, both economic factors – like trade
promotion in terms of tied aid (McGillivray, 2003a) – and political
factors might have a major influence on aid allocation patterns

(Neumayer, 2003).91 Svensson (2000) and Alesina & Weder (2002)
find some evidence that good governance structures are rewarded by
higher aid amounts as in cases where well-established civil and polit-
ical rights or low corruption levels are observed. Arvin & Drewes
(1998, 2001), studying Canadian and German aid allocations, pro-
vide evidence for a large country bias, which indicates a negative
relationship between per capita assistance and the population of aid-
receiving countries. While Chauvet (2003) expresses the importance
of socio-political stability on the aid allocation of donors, Collier
(2000) and Collier & Hoeffler (2004) provide some evidence that
donors do not respond appropriately to (post-)conflict societies in
their allocation of bilateral aid. In this context, Neumayer (2003)
accentuates the relevance of human rights aspects in the recipient
country for the donors’ aid decision.
In this study, we attempt to extend this set of aid allocation deter-
minants. We propose a novel hypothesis and extend the standard
empirical framework used for explaining bilateral aid allocation pat-
terns by adding the hitherto neglected impact of immigrants on the
donors’ aid allocation decision. Generally, donor governments are
exposed to the influence of immigrant groups interested in the wel-
fare of their country of origin. We differentiate between immigrants
with permenant residence permits in the donor country and immi-
grants without permanent residence permits (such as asylum seek-
ers). The driving force of allocating aid in favor of a home country is
the collusion of these two groups, both of whom care for their coun-
try of origin. Although asylum seekers, as well as permanently resid-
ing ethnic groups in the host country, might be skeptical about the
government and the political system of their country of origin, it
seems reasonable that they would still have strong links to and inter-
ests for their compatriots at home. Despite an aversion to the ruling
governmental system at home, immigrants encourage aid and sup-
port in favor of their home country, especially when it is transferred
on a sub-state level, for example, by non-governmental organiza-
tions (NGOs). Beyond this, asylum seekers play an important role as
‘first hand’ information providers and ‘urgency indicators’ for the
donor country. When people are forced to leave their home country,
donor governments acknowledge the special need for assistance in
these countries.
In the next section, we theoretically set out these ideas in a politi-

cal economic framework to illustrate the core determinants of a bilat-
eral aid allocation decision. From this model, we formally derive the
hypotheses to be tested in Section 8.5 for the case of German aid
allocation in the 1990s. Section 8.6 concludes this chapter.
8.3 The model
The focus of this section is to elaborate on the relationship between

official bilateral aid, the stock of legally immigrated foreigners, and
the influx of asylum seekers. To do this, we use a multi-household
three-country model. Within this model, we assume five types of
households: three types in the donor country, including two immi-
grant households (types 1 and 2) and one native household (labeled
type 3), and two types of households in two recipient countries,
labeled a and b. The households differ in incomes and utilities. All
three domestic household types, as well as the households in the
–
recipient countries, each receive exogenous incomes Y i (with i = 1, 2,
– 92
3) and Y j (with j = a, b). Li (with i = 1, 2, 3) stands for the number of
households of type i in the donor country and Lj (with j = a, b) for the
number of households in recipient country j. The donor government
levies a proportionate and exogenous income tax t on all domestic
(native and permanent immigrant) households. The tax revenue
T t ( L1 Y1 L2 Y2 L3 Y3 ) is transferred abroad as Official Deve-
lopment Assistance (ODA).
The (indirect) household utilities ui (with i = 1, 2, 3) are expressed
by the following functions:
u1 V[(1 t)Y1 ] ␭(A␣ )L␣ u␣ , (8.1)

u2 V[(1 t)Y2 ] ␭(A ␤ )L ␤ u ␤ , (8.2)
u3 V[(1 t)Y3 ]. (8.3)
The immigrant households are altruistic towards their country of

origin, and the altruism depends on the number of asylum seekers
from their home country residing in the donor country (that is, the
presence of more asylum seekers enhances the altruism of the per-
manent immigrants for their country of origin).93 We assume, there-
fore, that l(Aj) > 0 and l9(Aj) > 0, with j = a, b. This functional
relationship reflects the compassion of the permanent immigrants
for their compatriots.94 This ‘concern effect’ increases with the

number of new immigrants. The immigrant households’ utility func-
tion incorporates an additional term reflecting their altruism for
their compatriots in the home country.
The ODA is allocated as a share µ of the total ODA volume to the
two recipient countries, according to the marginal utility of income
the immigrant households receive through their altruism for their
countries of origin. In the recipient country, the transfer is allo-
cated to the households using a lump sum approach, that is, each
household receives the same amount additional to their exogenous
income. The efficiency of the aid transfer process is expressed by
the parameter dj (with j = a, b).95 The indirect utility functions uj
(with j = a, b) of the households in the two recipient countries are
given as:
(1 ␮ )T ␦␣ º
u␣ V ª« Y␣ »¼ ,
¬ L␣ (8.4)
ª ␮T ␦␤ º
u␤ V « Y␤ .
¬ L ␤ »¼ (8.5)
Finally, we assume positive and diminishing marginal utility of

income for all households. Hence, by assuming Ya > Yb, we get V’(Yb) >
V’(Ya) with V’’(.) < 0.
The growing concern caused by the influx of foreign nationals is
reflected in the social welfare function the donor government seeks
to maximize. The objective function of the donor government is an
additive separable welfare function maximizing total welfare W of
all households in the donor country by determining the allocation
parameter µ:
max W L1u1 L 2 u 2 L 3u3

␮ (8.6)
The government maximizes total welfare by choosing the optimal

allocation parameter µ, which determines the share the two recipient
countries receive. Assuming an inner solution, we get the optimal m*
from eq. (8.6):
wW !
L 2␭(A ␤ )V'(Y␤ )␦␤ L1␭(A ␣ )V'(Y␣ )␦␣ 0 (8.7)
w␮
Equation (8.7) determines the conditions for a corner solution, in

which the poorer country b receives all the ODA (m* = 1), if
• the number of permanent immigrant households from the poorer

country (L2) is relatively large,
• the number of asylum seekers from the poorer country b is rela-
tively large generating relatively strong altruism of the compatri-
ots living permanently in the donor country,
• the efficiency parameter dj of the poorer country is relatively high.
The corner solution is of some practical interest because some

countries receive either no aid or only a negligible amount from var-
ious donor countries.96 In order to derive the core hypotheses that we
are going to test, the interior solution of the allocation parameter
must before be analyzed.
8.4 Comparative statics
The main focus of this model is the influence of immigrant house-

holds and asylum seekers on the outcome of the social welfare func-
tion, in terms of the equilibrium value of the allocation parameter µ.
We begin with the impact of asylum migration on the aid allocation
parameter. By enhancing the concern of the permanent immigrants,
asylum seekers can influence the aid allocation decision in favor of
their home country. Thus, we derive the following core hypothesis
from eq. (8.7):
Hypothesis 1: Refugee-sending countries receive a greater share of

bilateral aid:
d␮ L 2␭ '(A ␤ )V'(Y␤ )␦␤
! 0,
dA ␤ D (8.8)
␦␤2 ␦␣2
where D L2␭(A ␤ )V''(Y␤ ) T L1␭(A␣ )V␣''(Y␣ )
T 0.
L␤ L␣
Before testing this hypothesis empirically, we can derive some fur-
ther testable hypotheses from eq. (8.7), namely:
Hypothesis 2: Countries with a large immigrant base in donor coun-

tries receive a greater share of bilateral aid:
d␮ ␭(A␣ )V'(Y␣ )␦␣

! 0. (8.9)
dL1 D
Hypothesis 3: More-populous countries receive a greater share of

bilateral aid:
d␮
L2 ␭(A ␤ )V''(Y␤ )␦␤2 ! 0.
␮T
L2␤
(8.10)
dL␤ D
Hypothesis 4: The good governance rewarding effect exists only for

countries with an initially low efficiency parameter. If such countries
improve their governance structure, they can expect a larger share of aid,
simultaneously reducing the share of bilateral aid received by countries
with a high governance level:
d␮ L1␭(A␣ )V'(Y␣ ) + L1␭(A␣ )␦␣V''(Y␣ ) (1L␮ )T !

␣
0. (8.11)
d␦␣ D
Hypothesis 5: Countries with a low per capita income receive ceteris

paribus more aid:
d␮ L1␭(A␣ )V''(Y␣ )␦␣

! 0. (8.12)
dY␣ D
We can now summarize the marginal impacts of major determi-

nants of interest on bilateral Official Development Assistance (ODA)
allocations:
( ) () () r) ()
§P P P (P P·
ODAij ODAij ¨ Lji , Aji , Yj , ␦ j , Lj ¸ .
¨ ¸
© ¹
8.5 Empirical evidence
This section empirically tests the hypotheses formally derived above.

The objective is to explain the behavior of Germany as a bilateral
donor of foreign aid. Therefore, we focus on its bilateral aid alloca-
tion pattern in terms of amounts of aid given to recipient countries.
The following variables are controlled for: First, we assume that the
longer the development assistance relationship between Germany
and a recipient country has lasted, the stronger the bilateral ties and
the more aid this country receives. Development cooperation his-
tory, therefore, should play an important role in current aid alloca-
tion decisions. Furthermore, we expect that geographical distances
between donor and recipient country should influence the aid pat-
tern. The geographical distance variable, which is well known from
gravity models in international trade, suggests that countries that
are situated geographically closer to the donor country receive more
aid (see Chapter 7.4). In addition, bilateral trade volumes should be
taken into account as a variable representing the commercial interest
of the donor country. We assume that Germany allocates more ODA
to trade partners according to bilateral trade volumes.97 Lastly, we
control for the strategic interest of Germany in Eastern Europe, where
most of the countries were treated as potential EU-accessing coun-
tries. Thus, we expect that Germany allocates relatively more aid to
the Eastern European transition countries.
8.5.1 Data and methodology

The data used for the dependent variable represent the annually dis-
bursed bilateral Official Development Assistance (ODA) amounts
from Germany to 148 developing countries and countries in transi-
tion during the 1990s (from 1991 until 1999). In order to balance out
the large variations in the annual volumes, we take nine-year aver-
ages. The ODA data are obtained from the Development Assistance
Committee (DAC) of the OECD.98 We include both Part I and Part II
countries in our sample.99
For the purpose of controlling the significance of the explanatory
variables on the allocation of German ODA in the 1990s we use
OLS estimation. It is remarkable that Germany (as with other major
donors, such as Japan) gives at least some aid to almost every
developing and transition country.100 Although the Development
Assistance Committee (DAC) of the OECD (2004) provides ODA data
for more than 150 countries, in the case of aid flows from Germany
in the 1990s the sample of countries is smaller because data for some
explanatory variables are not available. If not constant, all variables
represent nine-year averages from 1991 to 1999.
We estimate the following regression equation:
ODAj ␤0 f ( asyl j , forpop j ; ␤ ) ␥ X j ⑀ j ,
where j indexes aid-receiving countries and immigrant-origin coun-
tries, and b and g are estimation coefficients. We focus our regression
on the impact of the asylum seeker and foreign population variables,
and the interaction of the two. The asyl variable expresses the
logarithm of the average number of asylum applications from citi-

zens of country j in Germany from 1991–1999, whereas the forpop
variable denotes the logarithm of the average stock of population
with citizenship of country j living permanently in Germany.101
With the vector Xj we capture other exogenous control variables that
might affect the allocation of aid:102
• Population: the logarithm of the average stock of population of

recipient country j.103
• Income: the logarithm of the average per capita income in con-
stant 1995 US dollars in recipient country j.104
• PQLI (Human well-being): we use the logarithm of the physical
quality of life index (PQLI) for capturing human needs in the
recipient country j. The PQLI is an aggregate measure of life expec-
tancy, infant mortality and literacy, all equally weighted on a 0 to
100 scale.105
• Distance: the logarithm of the geographical distance of recipient
country j from Germany in kilometers.106
• Democracy: the logarithm of the democracy variable reflects the
average of both the political rights and civil liberties index for
country j.107
• Trade: the logarithm of average export volumes in goods from
Germany to recipient country j.108
• Aid ties: the logarithm of the number of years of development
cooperation between Germany and recipient country j before
1990.109
• Eastern Europe: dummy controlling for Germany’s strategic inter-
est in the region of Eastern European transition countries.
Table 8.1 provides Pearson correlation coefficients for the depen-

dent variable and all explanatory variables. Most of the correlations
are not very high, except for the correlation between foreign popula-
tion and asylum seekers, which reveals some evidence for a network
and size effect, respectively. To avoid multicollinearity, we exclude
one of these variables in the regression specifications I, II and III
(Table 8.2). Several studies (for example, Neumayer, 2003) point out
a nonlinearity in population size and demonstrate a bias towards less
populous countries that is reversed for very populous countries.
Furthermore, we test for nonlinearity in the asylum seeker variable.
Table 8.1 Bivariate correlates of German ODA and explanatory variables
Asyl Forpop Pop Income PQLI Distance Dem Trade Aid ties
ODA 0.15 0.19* 0.25** –0.01 0.14 –0.25** –0.11 0.70** –0.18
(1.78) (2.31) (3.23) (–0.08) (1.74) (–3.08) (–1.38) (11.78) (–2.21)
Asyl 0.62** 0.05 –0.07 0.09 –0.27** 0.10 0.14 0.04
(9.58) (0.57) (–0.86) (1.10) (–3.35) (1.26) (1.74) (0.48)
Forpop 0.04 0.00 0.09 –0.21* 0.03 0.34** 0.06
(0.43) (0.002) (1.14) (–2.57) (0.34) (4.35) (0.72)
Population –0.08** 0.03 0.04 0.10 0.36** 0.06
(–0.96) (0.31) (0.47) (1.20) (4.72) (0.78)
Income 0.52** –0.03 –0.20* 0.29** –0.05
(7.40) (–0.30) (–2.52) (3.63) (–0.64)
PQLI –0.02 –0.42** 0.32** –0.35**
(–0.22) (–5.64) (4.15) (–4.50)
Distance –0.19* –0.22** 0.26**
(–2.34) (–2.78) (3.30)
Democracy –0.17* 0.21*
(–2.07) (2.60)
Trade –0.14
(–1.76)
Notes: t-statistics in parentheses. Critical t-values at 1 percent (5 percent) level: 2.61 (1.98). (*), (**) statistically significant at the
5 percent (1 percent) level.
Table 8.2 German aid allocation. Dependent variable: log of aid, 1991–1999
I II III IV
log(Population) 1.39*** 1.32*** 1.36*** 1.47***

(3.44) (3.27) (3.42) (3.63)
[log(Population)] (sq.) –0.04** –0.04* –0.04** –0.05**
(–2.10) (–1.93) (–2.04) (–2.26)
log(Income) –0.40** –0.40** –0.45*** –0.42**
(–2.23) (–2.24) (–2.63) (–2.39)
log(PQLI) –0.49 –0.35 –0.65 –0.60
(–0.62) (–0.43) (–0.80) (–0.76)
log(Asyl) 0.08** 0.16* – –
(2.29) (1.69)
[log(Asyl)] (sq.) – –0.007 – –
(–0.91)
log(Forpop) – – 0.11* –
(1.83)
log(Asyl) ∙ log(Forpop) – – – 0.007**
(2.31)
log(Aid ties) 0.35*** 0.37*** 0.31*** 0.33***
(3.57) (3.54) (2.99) (3.23)
log(Distance) 0.18 0.20 –0.03 0.13
(–0.68) (0.73) (0.13) (0.49)
Democracy –1.26*** –1.24*** –1.18*** –1.24***
(–4.20) (–4.10) (–4.07) (–4.13)
log(Trade) 0.13 0.13 0.10 0.12
(1.18) (1.20) (0.91) (1.13)
Eastern Europe 2.62*** 2.70*** 2.44*** 2.50***
(5.28) (5.53) (4.93) (5.03)
R2 adjusted 0.70 0.69 0.70 0.70
Number of observations 148 148 148 148
Notes: Standard errors tested for heteroscedasticity (White test). Absolute t-values in
parentheses. A parameter estimate indicated with (*), (**), (***) is statistically significant
at 10, 5, or 1 percent level.
We capture nonlinearity by adding squared terms for these two

variables.
8.5.2 Empirical results

Table 8.2 provides the OLS estimation results for the average bilateral
aid flows of Germany. The four regression models differ in the spec-
ification of the immigration variables, foreign population and asy-
lum seekers. Besides these two variables, all other control variables
are included in all four specifications identically. Obviously, the vari-

ables of special interest concern the influence of the immigration
determinants. In Regression I and II we test for the influence of asy-
lum seekers on the ODA allocation pattern. This reveals a positive
and significant impact of the linear asylum variable, whereas the
squared term in specification II is insignificant, rejecting the hypoth-
esis of a diminishing marginal impact of asylum immigration. In
Regression III, we test for the influence of the population with for-
eign citizenship and durable residence permits on the aid allocation
decision. As Lahiri & Raimondos-Møller (2000) theoretically point
out, ethnic groups (that is, immigrants) influence aid allocations in
favor of their countries of origin through lobbying activities.
We find some empirical evidence, indicated by a positive and sig-
nificant elasticity of about 0.11, suggesting that immigrant groups
play a non-negligible role in the aid allocation process in Germany:
the average ODA volume Germany allocates to a recipient country
increases by around 11 percent if the stock of immigrants from that
country doubles. Specification IV includes an interaction term of
asylum applicants and the foreign population, which confirms the
results of specifications I, II and III: the marginal impact of asylum
migration on allocated aid is positive and increasing with the stock
of foreign population from the country of origin. Therefore, we can
reject the null hypothesis that asylum migration has no influence on
the bilateral aid allocation pattern of a donor country.
Concerning the other control variables, we can confirm the exis-
tence of a large country bias, which states that large, populous coun-
tries receive a relatively smaller amount of aid compared to less
populous countries.110 As for the influence of the recipients’ needs,
we find that only the income variable indicates a negative and signif-
icant effect in all regressions (with elasticity between –0.39 and
–0.45), whereas the physical quality of life index (PQLI) variable is
not significant at all. However, to conclude that per capita income as
a monetary parameter plays a more important role in the allocation
decision than human well-being is rather vague since both variables
are significantly correlated. The variables expressing trade ties
between Germany and the recipient country, as well as the geograph-
ical proximity, display the expected signs but are not significant.
However, our regressions do confirm that the good governance
rewarding effect reported, for example, by Alesina & Dollar (2000) or
Neumayer (2003), is found at least partially in German ODA alloca-

tions. The democracy variable, as a quality parameter for the exis-
tence of democratic institutions, delivers positive and significant
coefficients in all specifications. Finally, two control variables show
remarkable explanatory power in explaining Germany’s bilateral aid
allocation. First, development cooperation history between Germany
and recipient countries seems to be rather important, meaning that
the longer a country receives aid, the more aid that country receives.
And second, our regression shows that Germany has a strong strate-
gic interest in Eastern Europe, manifested by the huge amount of
official aid (OA) allocated to this region.
This chapter focuses on the hitherto-unexplored question of whether

the bilateral aid allocation pattern of a donor country is determined
by the influence of immigrant groups. We differentiate between the
foreign population that lives permanently in the donor country, but
still has foreign citizenship, and immigrants seeking asylum who do
not have permanent residency status. We assume that the interests of
both types of immigrants are symbiotic, but the asylum seekers have
a weaker status and less means to influence the donor in favor of
their country of origin. Therefore, asylum seekers’ ‘agents’ are their
compatriots already living in the donor country. Our results suggest
that the development policy of a donor country is significantly influ-
enced by permanent immigrants and asylum applicants. In the case
of Germany, we can assert that people with foreign citizenship (as
the core interest group) as well as asylum seekers (as transmitters of
information about human needs and urgencies) play a decisive role
in forming public opinion, and through this, in shaping the political
aid allocation decision-making process in the donor country. This
applies in particular to donor countries which are – like Germany –
major host countries for asylum seekers. This result can be viewed in
two ways. On the one hand, bilateral aid allocation decisions could
simply be seen as influenced (‘immigration bias’) by the different
immigrant groups in the donor country. On the other hand, the
influx of asylum seekers could be seen as helping to overcome infor-
mation problems of donor governments about the severity of a spe-
cific political and economic situation in the migrants’ sending
country. Development policymakers in the donor countries should

take this into account and scrutinize the relative neediness between
countries that send and do not send asylum seekers. In this context,
donor governments should take also into consideration the potential
for a rather perverse incentive to force emigration in less-developed
recipient countries. If countries receive more aid when they ‘send’
asylum seekers to donor countries, development aid can itself have a
counterproductive effect, which is definitely not intended by the
donor country.
However, in order to fully corroborate theory and empirical evi-
dence, aid allocation patterns of other major donor countries – which
are also major host countries for asylum seekers – should also be
examined. This is the task of the next chapter.
9
Refugee Movements and
Aid Responsiveness
Abstract111
This chapter analyzes the impact of refugee migration movements

on the long-term and short-term aid allocation decisions of bilateral
donors. We distinguish between different types of forced migrants:
internally displaced persons (IDPs) that stay in their country of ori-
gin, cross-border refugees that flee to neighboring countries, and
asylum seekers in Western donor states. For the period 1992 to 2003,
empirical evidence on 18 donor and 148 recipient countries suggests
that short-term emergency aid is given to all types of refugee situa-
tions, but is predominantly directed towards the countries of origin.
For the allocation of long-term development aid, donor states focus
even more on the sending countries of forced migrants; they increase
aid volumes only for the home countries of refugees, not for the host-
ing countries. This preference for the countries of origin is even
stronger when they are sending countries of asylum seekers to the
Western aid-giving states.
9.1 Introduction
Foreign aid serves a multitude of objectives. For some donor states,

the allocation and type of aid is largely shaped by concerns for the
development needs of recipient countries, while other states use aid
rather as an instrument of foreign and commercial policy interests.
Since the early 1990s, the criteria for bilateral aid allocation decisions
have shifted towards some new objectives (Hjertholm & White,
146
Refugee Movements and Aid Responsiveness 147
2000). One of these new goals of development policy is mitigating

the root causes of the heightened migration pressure from refugees
and other migrants coming from developing countries to Western
industrialized countries. The prevention of refugee movements and
the cessation of long-lasting refugee situations have gained priority
in international politics as primary development and foreign policy
objectives, although bilateral donor governments have, as yet, been
rather slow in implementing these objectives (UNHCR, 2006b).
Western countries that are preferred destinations for asylum seek-
ers and refugees are, in principle, more sensitive to addressing migra-
tion and refugee issues in the formulation of their development aid
strategies. For instance, the German government frequently expresses
its concerns regarding large refugee movements (for example, BMZ,
1994; BMZ, 2006b). During the height of the asylum crisis in the
early 1990s, the German Federal Ministry for Economic Cooperation
and Development (BMZ) phrased a concept paper for their develop-
ment assistance policy, which explicitly aims to (i) achieve the elim-
ination of the reasons that cause flight and unwanted emigration,
(ii) provide support for neighboring first asylum countries in coping
with refugee crises and the alleviation of their burdens, and (iii) pro-
vide aid for reintegrating returning refugees into their country of
origin (BMZ, 1994). Other Western asylum destinations have similarly
rephrased their development policies to focus more on migration-
related issues.112
This chapter examines the question of whether Western donor
states indeed responded to forced migration flows, and if yes, how
their aid allocation policies have changed in face of this ‘age of
migration’ (Castles & Miller, 2003) since the end of the Cold War.
In particular, we investigate how different types of refugee popula-
tions influenced the respective cross-country aid allocation decisions
of bilateral donors. We analyze whether internal displacements,
cross-border refugee migration movements or flows of asylum-
seeking people towards Western countries have had any significant
influence on the bilateral aid-giving patterns of the affected donor
states. Aid allocation determinants and differences between donor
states are tested using a large panel dataset covering the post–Cold
War period from 1992 until 2003. Three relatively robust results
are discovered: first, donors differentiate between internal and
external displacements, that is, cross-border refugee outflows create
significantly greater donor responsiveness than displacements that

are confined within the country of origin. Second, cross-border ref-
ugee movements attract significantly more aid to the respective
countries of origin than to the (neighboring) first asylum countries
that host the bulk of the global refugee stock. And third, refugee
movements that reach the Western donor countries ‘physically’ as
asylum seekers provoke an even larger aid response.
The remainder of this chapter is structured as follows: the next
section provides a brief review of aid allocation and forced migration
literature. In Section 9.3, a description of the empirical framework is
outlined with a short discussion of the set of explanatory variables
that is used to explore and test bilateral aid-migration patterns.
Finally, the regression results for both the allocation of Official
Development Assistance (ODA) and emergency aid are discussed.
Section 9.4 concludes and discusses policy implications.
9.2 Refugee movements: some policy options
The 1951 ‘UN Convention on the Status of Refugees’, amended by

the Protocol of 1967, defines a refugee as an individual, who owing
to a
well-founded fear of being persecuted for reasons of race, religion,

nationality, membership of a particular social group or political
opinion, is outside the country of his nationality and is unable or,
owing to such fear, [ ... ] is unwilling to return to it. (Article 1, The
1951 Convention Relating to the Status of Refugees)
The interpretation and implementation of this convention is

contingent on the country where an asylum application is placed.
But the principle of ‘non-refoulement’ denies a signing country the
right to refuse protection to asylum seekers, even when refugee sta-
tus is not granted.113 Although the basic classification of refugee sta-
tus has not changed, the number of persons seeking protection
abroad has multiplied in recent decades (Widgren, 1989; Schuck,
1997; Hatton, 2004). Additional to the increasing number of cross-
border refugees, the global population of internally displaced per-
sons (IDPs) has also increased tremendously during the same period
(UNHCR, 2004b). However, IDPs and cross-border refugee movements
have predominantly been a phenomenon of the developing world,

particularly since the end of the Cold War. Although these refugee
movements have become increasingly intercontinental, the majority
of displaced populations remain in the region of origin (Hatton &
Williamson, 2005). Only a small, but increasing, share of the total
population of these forced migrants has the resources to seek asylum
in Western industrialized countries.114 There, particularly since the
growing numbers of asylum seekers in the early and mid-1990s,
domestic pressure to restrict asylum immigration continues to
increase due to the perception that socioeconomic costs of integrat-
ing people with completely different ethno-cultural background are
too high to accommodate the rising influence of refugees (Stein,
1986; Boswell, 2003).
Consequently, Western industrialized countries are confronted
with a balancing act between humanitarian aspects and their ‘self-
protection interests’: on the one hand, concerned with the preven-
tion of large inflows of asylum seekers towards the own country,
they are also impelled in upholding solidarity with these highly vul-
nerable individuals. On the other hand, many (mostly African)
developing countries perceive that they shoulder a disproportionate
burden of hosting refugees in comparison to the international com-
munity, since the largest part of the global refugee population is
hosted in the region of origin, mostly in conflict-neighboring coun-
tries, and only a small share finds refuge in Western industrialized
countries (UNHCR, 2004b; Betts, 2004). Many donor governments
address this problem and express their intention to share this refugee
burden of developing countries either financially or ‘physically’. In
fact, the attempt to prevent asylum migration to Western countries
by stabilizing the region of origin is the overarching interest of
Western donors.
What drives the international refugee migration pattern? Rotte
et al. (1997) find evidence that the emigration pressure of refugees is
reduced by both a decreasing per-capita income gap between the
country of origin and the intended country of asylum and an increase
in the absolute level of per-capita income. Furthermore, the migra-
tion propensities of refugees are lessened by diminishing levels of
economic hardship, political oppression, violence, and human rights
abuses in the country of origin and increasing restrictiveness of asy-
lum policies in the host country (Hatton, 2004; Neumayer, 2005b).115
In contrast, Gosh (1992) provides evidence that improving economic

conditions along with political reforms may initially lead to a rise in
emigration, while in the long run, at a higher level of income, the
desire to leave the home country decreases. However, neither Hatton
(2004) nor Neumayer (2005b) find empirical evidence for this ‘refu-
gee hump’ effect.
Besides this ambiguity on the relative importance of these root
causes on emigration, there is also a controversy regarding the effec-
tiveness of development aid in tackling and relieving these causes of
refugee movements. On the one hand, it is highlighted that the
influence of aid intervention on the root causes of emigration and
the success of altering deeply rooted political and economic griev-
ances in the country of origin are rather limited (for example, Byrne,
2003). On the other hand, development assistance is, under certain
conditions, effective in reducing conflict risk by promoting eco-
nomic progress. That is, even if aid does not have a significant (posi-
tive) impact on the quality of policy in a recipient country, it
nevertheless might enhance economic growth, which might reduce
conflict risk, and as a consequence, lower the propensity to emigrate
(for example, Collier & Hoeffler, 2002). Rotte et al. (1997) and
Neumayer (2005b) provide some evidence that aid is effective for the
support of democracy or economic development to reduce refugee
migration flows, at least in the long run. Consequently, donors with
a long-term perspective might use the aid instrument for their
migration-preventive interests.
Beyond these questions of aid effectiveness, the facts about the
determinants of aid allocation policies of donor governments are
much more stylized. Generally, and throughout the empirical litera-
ture of the last decades, convincing evidence has been put forward
that donor self-interest variables like trade relations, political similar-
ity, geographical distance, military expenditures or colonial ties
are the most important determinants of bilateral aid allocations.116
Beside these ‘hard’ factors of economic and political self-interest, the
influence of recipient need on bilateral aid allocations is less distinc-
tive. The evidence shows that donor preference for poverty allevia-
tion or other human development factors, like the promotion of
democracy and human rights, are rather volatile with respect to time
and differ largely between donor states (Alesina & Dollar, 2000;
Neumayer, 2003; Nunnenkamp & Thiele, 2006).
We contribute a new aspect to this literature on aid allocation

determinants with its particular focus on the influence of refugee
movements. There are basically two alternative explanations for the
underlying motives for donors to respond to refugee migration.
Western donor states might either respond to various types of refu-
gee migration by transferring different forms of aid (that is short-
term emergency versus long-term development aid) for reasons of
‘self-protection’ against large emigration movements, or because of
altruistic (humanitarian) burden-sharing motives. Thus, the under-
lying argument of the subsequent analysis presumes that the more a
donor state perceives itself as (directly or indirectly) ‘affected’ by ref-
ugee movements, the more it reacts by transfering aid, either to the
country of origin or to any other first asylum country for migration
prevention and/or burden-sharing purposes. We presume that donor
countries perceive themselves to be more affected by refugee migra-
tion when refugees travel farther away from their homes into other
developing countries, since their propensity for secondary move-
ments towards Western countries is larger. According to this assump-
tion, donor countries should, ceteris paribus, transfer less aid to
countries with internally displaced persons and more to countries
with a large number of cross-border refugees, and even more to coun-
tries that already send a significant number of asylum seekers to
Western countries (see also Chapter 8).
9.3 Empirical analysis
The following empirical analysis tests whether bilateral aid alloca-

tion decisions are influenced by any type of refugee migration. Based
on the core hypothesis that donor states respond more generously if
they perceive themselves as affected by any type of refugee move-
ment, we will test whether ceteris paribus countries with internally
displaced persons (IDPs) receive less bilateral aid than countries that
receive cross-border refugees. With respect to cross-border refugee
populations, we distinguish between aid that is received by the coun-
tries of origin and aid that is received by the first asylum countries.
Although the two underlying motives of migration prevention and
burden-sharing might overlap, we suggest that aid to the first asy-
lum countries tend to be motivated by altruistic burden-sharing
intentions, while aid to the country of origin is rather motivated by
migration prevention interests in terms of reducing emigration pres-

sure and inducing voluntary repatriation of cross-border refugees
and asylum seekers.117 In this context, we also test whether asylum
seekers that enter the Western (donor) countries provoke a particular
donor response to the respective sending countries.
9.3.1 Methodology
For testing these outlined hypotheses, three different estimation
techniques are applied to check for robustness of the results: A stan-
dard two-part model, a Heckman sample selection model and a lin-
ear fixed effects model. For a discussion of these alternative estimation
methods in the context of foreign aid allocation decisions see, for
example, Neumayer (2003) and Berthélemy (2006).
The reason that the robustness of the results is controlled for by
these alternative estimation models is that the truncated dependent
aid variable is not fully continuous with a positive probability mass
at the value zero (given that aid disbursements are non-negative).
The literature provides different estimation techniques to account
for nonlinearity and the existence of a censored dependent variable
(Apodaca & Stohl, 1999; Neumayer, 2003; Berthélemy & Tichit,
2004).118
The problem can be formally described as follows:
Pr (Yijt ! 0) F (␣ ␥ Xijt ⑀ijt ), (9.1)

Yijt ␤ ␦ Zijt ␮ijt . (9.2)
Bilateral aid transfer Y is estimated in the first stage as a binary

dependent variable, while in the second stage, the aid allocation
equation is only based on the sub-sample with positive observations.
The set of independent variables is given by X and Z, a and b are
constants, g and d are vectors of coefficients, and e and m are inde-
pendent and normally distributed error terms. F(.) represents the
cumulative distribution function. Indices i and j specify respective
donor and recipient countries, and t indicates the time period.
A first, and common, approach is to estimate the binary estima-
tion model and the level model separately, that is the binary first
stage of aid eligibility applies a logit (or alternatively a probit) model,
while on the second-level stage a linear model is run with the reduced
dataset. Here, the crucial assumption is that the choice of an aid
recipient is independent from the respective aid level. However, the

error terms are not necessarily independent, and thus, it is likely that
the second-stage regression on aid levels is biased. Nevertheless, this
model is used as a benchmark; its results are checked by the follow-
ing two alternative approaches.
A Heckman sample selection model estimates the two stages simul-
taneously, assuming that the error terms e and m of the eligibility
stage and the level stage regressions are not independent, which is a
plausible assumption. Therefore, the binary stage is still run with a
logit estimation, while the level stage includes additionally the
inverse Mill’s ratio from the first stage in order to correct for selec-
tion bias. Aid allocation eq. (9.2) is then modified to be:
Yijt ␤ ␦ Zijt ␳␴ f (␣ ␥ Xijt ⑀ijt ) / F(␣ ␥ Xijt ⑀ijt ) ␮ijt , (9.3)
with r representing Cov(e, m), s denoting Var(m), and f(.)/F(.) captur-

ing the inverse Mill’s ratio.119
Since the results of the Heckman model are qualitatively not very
different from those obtained from the linear regression on the
reduced dataset (eq. 9.2), the properties of the selection equation and
of its error term are not particularly important when estimating
the regression on aid levels.120 This result is similar to those of, for
example, Alesina & Dollar (2000) or Berthélemy (2006), who all
conclude that a linear regression was basically as good as Heckman
estimation.
Third, a fixed effects model is applied for the aid level equation to
control for donor-recipient-specific heterogeneity, which might not
be captured by the set of explanatory variables. The inclusion of
donor-recipient fixed effects may also correct for the problem of
unobserved as well as missing variables. Compared to the other two
approaches, some results for the fixed effects regression are signifi-
cantly different, indicating that donor-recipient heterogeneity matter
and should not be neglected, particularly not in a large dataset.121
Furthermore, as donor decisions on aid allocations and the actual
aid transfers are predominantly not settled in the same year, most
variables enter with a one-year time lag. This also reduces potential
simultaneity problems. Estimation results are reported with standard
errors which are robust in terms of arbitrary heteroscedasticity and
serial correlation.
9.3.2 Data
The sample contains 18 donor countries and 148 recipient countries,
covering the years from 1992 to 2003. Data on both bilateral (long-
term) development aid and (short-term) emergency aid are provided
by the Development Assistance Committee of the OECD (2006a).
Both dependent variables are defined in per capita terms. ODA is
characterized by its composition of grants and highly concessional
loans, the objective of promoting economic development and
welfare, and its distribution by the official sector.122 We use ODA
gross disbursements instead of commitments, since they reflect the
aid volume actually transferred to the recipient country.123 For emer-
gency aid, we use disbursement data provided by the Creditor
Reporting System (CRS) of OECD (2006a). For the dependent vari-
ables (only for aid levels) and most explanatory variables, we use
logarithms for capturing possible nonlinear effects and to interpret
estimates as elasticity (Neumayer, 2003).
As the core explanatory variables, four respective categories of ref-
ugee migration are distinguished: (i) internally displaced persons
(IDP), that is refugees who did not (yet) cross borders, (ii) total stock
of refugee outflow (Ref origin), that is number of cross-border refugees
by country of origin, (iii) total stock of refugees living in a recipient
country (Ref asylum), that is the number of hosted refugees, and
(iv) bilateral asylum applicants in the Western donor country (Asylum
seeker), that is, persons applying for asylum status according to UN
convention in the donor country stemming from the respective aid
recipient. For all categories, the respective variables are defined in
absolute numbers and all data are provided by UNHCR (2006c).124
It is presumed that donor states respond differently towards each
one of the four refugee categories. According to the previous discus-
sion, basically for all categories of refugees both motivations for aid
provision can underlie, that is migration prevention interests and
altruistic (humanitarian) burden-sharing, although burden-sharing
with first asylum countries and migration prevention towards the
countries of origin seems more plausible. The other control variables
reflect categories of donor self-interest, recipient need and good gov-
ernance.125 The Bilateral trade variable captures the major commer-
cial interests of the donor. It is defined by goods and services exported
to the recipient country as share of total exports. Hereby, we assume
that larger trade volumes result in larger aid amounts. Although a
reversal effect may exist due to aid-tying, Lloyd et al. (2001) find
no evidence that tied aid increases bilateral trade volumes signifi-
cantly.126 Taking into account the economic potential, total Popula-
tion of the recipient country is included, assuming that larger
countries receive more aid. The External debt variable measures the
total debt of the recipient country as percentage of GDP.127 The most
common recipient need variable is Income per capita, which approxi-
mates the economic needs of the recipient country’s population: the
worse the economic hardship, the more aid is allocated. However,
since the income per capita is evidently only an incomplete measure
for basic human needs, the Physical Quality of Life Index PQLI is
additionally included as a proxy for human well-being in a recipient
country (Morris, 1979; Moon, 1991).128
The number of fatalities caused by natural disasters (NatDis deaths)
is used as a proxy for the immediate basic needs after a natural catas-
trophe. It measures deaths by natural disasters such as droughts,
famines or earthquakes. Obviously, the allocation of aid, particularly
short-term emergency aid, is expected to correspond positively to the
extent of a calamity.
Good governance and democratic structures are controlled for
by the CPRI variable, which is generated by the civil and political
rights indices from FreedomHouse (2006), indicating for example
the right to participate freely in the political process, or the freedom
of expression and belief without interference by the state.129 According
to good governance rewarding policies, Western donor states are
committed to promoting better democratic institutions in recipient
countries by transferring larger amounts of aid to countries that
adopt these institutions. A proxy for the political fragility and insta-
bility of the recipient country is taken into account by the number of
Battle deaths caused by armed combats. Armed conflicts destabilize
the political and economic conditions, which results in major fail-
ures of states to insure the basic needs of the population. Donor states
are often rather reluctant to support politically fragile states since the
allocation of development assistance is more delicate in conflicting
countries (McGillivray, 2003b). We nevertheless assume that, ceteris
paribus, peace-building interests (as a precondition to sustainable
migration prevention) of donor states prevail, so that we expect that
development and emergency aid amounts increase with the number
of conflict fatalities.
9.3.3 Results
9.3.3.1 Aggregate aid allocation
Table 9.1 provides the estimation results for the allocation equation
for both long-term ODA as well as short-term emergency aid.130 It is
shown that there is not much correlation between the error terms of
the aid eligibility and aid level stages; thus, the results of the OLS and
the Heckman two-stage procedure are basically consistent and robust
in the use of these two alternative estimation methods.131
However, the introduction of donor-recipient fixed effects into the
panel regression significantly changes some of the results. The out-
come for the fixed effects model suggests that we have unobserved
effects that are not captured by the set of the explanatory variables.
Consequently, in the subsequent discussion we refer primarily to
these regression results. The coefficients for the four refugee catego-
ries show a clear-cut pattern of explanation. With respect to both aid
types, the strongest effects refer to the Asylum seeker variable, which
indicates that donor states significantly increase their aid volumes
for the respective sending countries of their asylum applicants. In
fact, the effect for short-term emergency aid is even threefold larger
than for long-term aid.132
This result, compared with the outcome for the Ref origin variable,
shows that donors have a stronger reaction towards migration move-
ments if they are directly affected by the refugee outflows.
Accordingly, donor states do not respond to the same extent if refu-
gees are only displaced internally. In this case, donors respond – if
they respond at all – only with short-term emergency aid without
regarding the long-term need to relieve protracted internal displace-
ment situations.
This pattern of interest of donor states becomes even clearer when
comparing the assistance directed towards the home countries of
cross-border refugees with that for (neighboring) refugee-hosting
countries (Ref asylum). This latter variable, which reflects the refugee
burden of first-asylum countries, is only positive and significant in the
fixed effects regression on emergency aid allocation. This means that
conflict-neighboring host countries are scarcely ‘compensated’ by the
international community for providing (often long-lasting) protection
of the displaced population. Development aid allocations are – on
average – not affected by these cross-border refugee populations,
Table 9.1 Aid allocation determinants, 1992–2003, 18 donors, 148 recipients
Dependent ODA Emergency aid

variable
OLS Heckman Fixed eff. OLS Heckman Fixed eff.
Ln(Asylum 0.054** 0.053** 0.054** 0.129** 0.143** 0.178**

seeker) (6.52) (5.47) (6.93) (10.88) (9.79) (7.58)
Ln(Ref asylum) –0.030** –0.030** 0.001 –0.014 –0.017* 0.065**
(5.64) (5.47) (0.16) (1.59) (1.95) (3.31)
Ln(Ref origin) 0.020** 0.018* 0.014* 0.108** 0.122** 0.086**
(2.80) (1.98) (2.21) (9.13) (8.31) (3.95)
Ln(IDP) –0.001 –0.001 0.000 0.020** 0.022** 0.025**
(0.15) (0.19) (0.10) (2.84) (3.17) (2.77)
Ln(Population) –1.068** –1.070** –1.373** –1.017** –0.994** –2.102**
(58.86) (52.23) (8.59) (34.14) (30.32) (4.00)
Ln(Income pc)° –0.807** –0.787** –0.167** –0.387** –0.578** –0.321
(22.26) (9.60) (2.93) (6.61) (4.43) (1.57)
Ln(Bilateral 0.403** 0.396** 0.059** 0.002 0.072 –0.022
trade)° (35.53) (13.78) (4.82) (0.10) (1.54) (0.58)
CPRI –0.077** –0.073** –0.034* –0.025 –0.059* –0.049
(5.93) (3.84) (2.20) (1.14) (1.96) (1.06)
External Debt –0.024** –0.023** –0.014 0.053* 0.045 –0.029
(4.16) (3.48) (1.04) (2.20) (1.83) (0.38)
PQLI –0.003* –0.003* –0.001 0.000 –0.002 0.011
(2.34) (2.02) (0.25) (0.07) (0.73) (0.94)
Ln(Battle 0.019** 0.020* 0.015** 0.082** 0.075** 0.040**
deaths)° (2.67) (2.57) (3.07) (8.77) (7.09) (3.32)
Ln(NatDis 0.075** 0.074** 0.015** 0.047** 0.050** 0.037**
deaths)° (9.27) (9.24) (3.60) (3.92) (4.11) (3.16)
Intercept 25.65** 25.50** 22.21** 14.221** 15.63** 30.45**
(53.85) (35.45) (9.37) (17.94) (13.35) (3.77)
Imills –0.141 1.526
(0.27) (1.64)
Number of
observations 16,990 21,520 16,990 21,520
Uncensored
observations 16,990 5,103
R2 adj. 0.250 0.251 0.854 0.406 0.406 0.653
Hausman test 272.79** 33.17**
Notes: Lagged variables are signified by (°). The Heckman estimation is run with two-stage
estimation. OLS without sample selection correction. Coefficients for fixed effects are not
reported. t-values are in parentheses. (*), (**) significant at the 5 (1) percent level.
although promotion of local integration in the first-asylum country is

often considered as a durable solution, and thus, also as a possible
migration prevention strategy for Western countries.
Aid assistance is shown to be primarily focused on the source
countries of refugee movements and not so much on those countries
bearing the burden of hosting refugees. This may suggest that donor
governments rather support voluntary repatriation of cross-border
refugees to their home country instead of local integration in the
first asylum country. Source countries receive larger amounts of
emergency and development aid for larger refugee exoduses. However,
aid responsiveness of Western donor states towards the source coun-
tries of refugees is even stronger if parts of the population spill over
to their own territory as asylum seekers, indicating motives of self-
protection. This pattern of explanation actually holds for both short-
term emergency aid and long-term development assistance. However,
if the displaced population does not leave the home country, but
instead remains within its own borders, additional aid is – on
average – only provided in the form of short-term emergency aid, but
not as long-term development aid. In the next section, however, dis-
aggregation of these results will show that the willingness to respond
to refugee movements by providing additional aid varies tremen-
dously among donor states.
For the non-migration variables, the results are consistent with
many other cross-country studies on foreign aid allocation. The find-
ings assert that poorer countries receive more aid per capita through-
out the different regressions. However, this result holds rather for
income poverty, while the influence of the composite (non-pecuniary)
human development indicator PQLI is rather weak. That is, while the
Income variable shows high significance for almost all regressions,
the PQLI coefficients do not exert a significant and robust influence
on the donor’s decision to allocate ODA or emergency aid.
Furthermore, countries with better developed democratic institu-
tions are likely to receive more development aid than those with
rather autocratic tendencies. This reflects the interest of bilateral
donors in rewarding good governance structures through enhanced
ODA transfers. Bilateral trade volume has a predominantly positive
and significant influence on the allocation of ODA, but not on the
allocation of emergency aid. Thus, ODA per capita is a robust correlate
to the donors’ share of exports to the recipient country. The influence
of external debt, which plays a significant role in determining

whether a country is eligible to receive aid (not reported), displays
only a weak influence at the allocation stage.
Furthermore, with respect to population size, although large coun-
tries receive more aid in terms of total volume (see, for example,
Neumayer, 2003), they receive less in terms of aid per capita: coun-
tries that are one percent larger than others receive on average around
one percent less ODA per capita. This indicates the existance of a
large-country bias.
Finally, humanitarian catastrophes generally initiate solidarity
with the affected country, regardless of whether natural disaster or
violent conflict is the root cause. However, donors seem to distin-
guish between victims of natural disasters and those of violent
clashes. While donor countries tend to allocate more ODA in the
aftermath of natural catastrophes than in violent conflict situations,
the pattern for emergency assistance is the other way around:133
countries with a high number of conflict casualties attract signifi-
cantly more attention from bilateral donors in terms of emergency
aid assistance than those with an equivalent number of natural
disaster victims. Hence, human rights violations and armed conflicts
are generally not disciplined by aid cuts, that is, compassion seems to
dominate reprehension.
9.3.3.2 Donor-specific aid allocation

For the purpose of testing whether there are significant differences
among donor countries with respect to aid responsiveness towards
refugee movements, we modify the previous fixed effects model by
introducing interaction terms between the set of explanatory vari-
ables and a donor-specific dummy Da:134
Yijt ␤ ␦ ' D␣ Zijt ␦ Zijt ␮ijt (9.4)
with Da (i) = 1 if i = a, and Da (i) = 0 otherwise.

Based on this extended fixed effects regression, Table 9.2 reports
the coefficients (as elasticities) of these interaction terms for the
Asylum seeker variable as well as the IDP variable for both ODA and
emergency aid allocations.
The overall picture of this exercise is that donors respond very
differently to these two types of refugee migration. For instance,
Table 9.2 Categorization of aid donors
Donor Elasticity of aid to number of Aid

Responsiveness
Asylum seeker IDPs
Emerg Emerg Emerg

ODA Aid ODA Aid ODA Aid
Norway 0.177*** 0.121* 0.048*** 0.073*** ++ ++

(5.32) (1.74) (2.80) (2.68)
Austria 0.175*** 0.533*** –0.039** 0.017 ++ +
(5.77) (5.43) (2.41) (0.40)
Italy 0.110*** 0.076 0.019 0.103*** + +
(3.15) (0.75) (1.00) (2.74)
Switzerland 0.078*** 0.218*** –0.012 0.005 + +
(2.88) (4.14) (0.76) (0.19)
USA 0.059*** 0.245** 0.032* 0.042 ++ +
(2.74) (2.35) (1.95) (0.89)
Australia 0.027 0.209** 0.055** –0.072 + +
(0.85) (2.06) (2.56) (1.01)
Sweden 0.146*** 0.082 0.025 0.008 + 0
(4.75) (0.90) (1.54) (0.26)
Netherlands 0.127*** 0.098 –0.006 0.012 + 0
(3.14) (0.82) (0.39) (0.36)
Ireland 0.026 0.311 0.055*** 0.083 + 0
(0.68) (0.49) (2.92) (0.53)
Germany 0.057 0.262** –0.031* 0.063** – ++
(1.21) (2.30) (1.83) (2.10)
UK –0.017 0.193*** 0.000 –0.014 0 +
(0.64) (2.66) (0.00) (0.37)
Spain 0.071* 0.136 –0.005 0.055 + 0
(1.79) (0.64) (0.25) (0.77)
Japan 0.067 –0.054 –0.004 –0.039 0 0
(0.84) (0.20) (0.22) (0.29)
Finland 0.056 0.158 –0.017 0.013 0 0
(1.11) (1.06) (0.91) (0.39)
Canada 0.025 0.234 0.000 –0.007 0 0
(0.59) (1.64) (0.00) (0.21)
France 0.023 0.228 0.009 0.064 0 0
(0.59) (1.43) (0.57) (1.09)
Belgium –0.017 –0.182 –0.034* 0.032 – 0
(0.39) (0.87) (1.87) (0.78)
Denmark –0.105* – –0.056** – ––
(1.71) – (2.32) –
Notes: Elasticity estimates are based on fixed-effect regression with additional donor-specific
parameters that are estimated together. t-values are in parentheses. (*), (**), (***) significant at
the 10, 5, and 1 percent level. The results for Denmark are based on a rather small number of
observations.
countries such as Norway, Austria, or the US stand out as countries

with a strong migration-sensitive aid allocation. These and other
havens for asylum applicants react very strongly to increasing num-
bers of asylum seekers by boosting respective foreign aid budgets
towards the sending countries. At least eleven donor states increase
their aid budgets significantly (either long-term development assis-
tance or short-term emergency aid) for the countries of origin of
their respective asylum applicants. Other donor states like Japan,
France, Canada and Finland do not respond at all to any of these dif-
ferent categories of refugee migration, while Denmark may even
partially reduce its bilateral aid budget towards refugee-sending
countries.
With regard to internal displacement, aid responsiveness is much
weaker and not always in favor of the respective recipient country.
Only a third of all observed donor states react positively to internal
displacements by allocating any additional aid towards these coun-
tries, while four countries even significantly reduce their correspond-
ing short-term or long-term aid budgets. In particular, Italy and
Norway become more generous towards countries that experience
large internal displacements. As violent conflicts as well as natural
catastrophes are controlled for as possible root causes for internal
displacements in these countries, it is recognizable that these donors
have a special interest in supporting recipient countries with a sig-
nificant number of IDPs.
This behavior might originate from altruism towards these people,
who are generally poorly assisted by the recipient country’s govern-
ment or any other external donors.
However, these donors may also have farsighted migration-
prevention interests since protracted and unassisted internal dis-
placements could further destabilize the country or the whole region,
which would force parts of the population to cross borders and
become refugees in the region of origin or even asylum seekers to the
Western countries.
In general, countries of origin tend to receive more aid when their
citizens leave the country as refugees and asylum seekers. As stated
in Chapter 8, this might induce counterproductive incentives, as
sending countries could be tempted to enhance refugee flights (or, at
least, not to prevent them) to attract more short-term and long-term
aid. Countries of origin receive even more foreign aid when refugees
seek asylum in the donor states and less when they stay in the region
of origin. At least two explanations for this result are possible: first,
urgency of a conflict situation is transmitted to the donor countries
by the inflow of asylum seekers. Aid might then be an instrument for
stabilizing the crisis situation in the country of origin, so that emi-
gration might become a less attractive option and voluntary repatria-
tion of refugees is supported. A second explanation is that asylum
seekers, together with other foreign-born compatriots living in the
donor country, proactively lobby the aid allocation decision-making
process for the sake of their compatriots at home (Lahiri & Raimondos-
Møller, 2000; Anwar & Michaelowa, 2006; Chapter 8).135
This chapter focuses on the influence of refugee movements on the

allocation of long-term development aid and short-term emergency
aid by Western donor countries. Until now, the aid allocation litera-
ture has not addressed this issue. Increasing refugee flows, particu-
larly within the most fragile regions of the developing world,
highlight the need of both the sending and the hosting countries for
international support. Two underlying motives for an increased aid
support towards conflict-affected countries are often discussed:
altruistic burden-sharing towards first asylum countries and coun-
tries of origin (in case of IDPs), or migration management interests
by preventing further de-stabilization of the countries of origin and
reducing the emigration pressure towards the Western industrialized
countries. This study shows that bilateral aid allocation policies are
primarily focused on countries of origin. This indicates that the
underlying interests of donor states are rather focused on migration
prevention instead of on altruistic burden-sharing motives, since
first asylum host countries, as well as countries with a high number
of IDPs, are widely neglected.
Beyond this, what do these results imply? In fact, some donor gov-
ernments seem to be motivated to act proactively, seeking to allevi-
ate the root causes of cross-border refugee movements by transferring
aid towards refugee-sending countries. They probably do this with
an intention of decreasing further refugee outflows and to increase
repatriation willingness of refugees living abroad (preferably those
living in the donor country itself). However, this analysis shows that
this policy approach is not common to all donor countries, since

they are actually very heterogeneous in their migration-related aid
allocation pattern. There might be several reasons for this heteroge-
neity among donors. First, if donor states are not affected or insignif-
icantly affected by refugee movements and asylum seekers (like
Japan), then there is obviously no need for any migration-preventive
aid policy. Second, even if a donor state is affected by refugee move-
ments (for example, by asylum seekers), it does not necessarily trust
in the effectiveness of aid for migration prevention purposes. And
third, even if a refugee-affected donor state trusts in the effectiveness
of aid-induced migration prevention, it has an incentive to free-ride
on aid provisions of other donor states that are more willing to chan-
nel aid towards refugee hot spots.
10
Conclusion
10.1 Summary
This book addresses the causes and incidences of refugee movements

and the policy responses of Western asylum and aid-giving coun-
tries. In particular, we focus on the positive and normative aspects of
refugee migration and its management by Western industrialized
countries. Thereby, asylum policy and aid policy are the two main
instruments of interest.
The analysis began by analyzing the worldwide incidence of refu-
gee movements. We develop a so-called Refugee Burden Index that
reveals the heterogeneity of both the global and intra-regional distri-
bution of the refugee population, and thus, unequal burdens across
host countries. The causes of refugee migration are explored using a
case study on the civil conflict in Aceh, Indonesia, identifying the
driving forces of forced migration. This shows that the main deter-
minants of displacement are not only the immediate conflict factors
like violence and fear of death, but also other traditional migration
causes like socioeconomic factors.
The focus of the theoretical part of the book is twofold. First, we
elaborate the driving forces of the magnitude, composition and dura-
tion of refugee movements, as well as its implications for the Western
asylum countries and the conflict-neighboring first asylum coun-
tries. Since both types of asylum countries are interested in reducing
their respective burdens, it is likely that they end up in a ‘race to the
bottom’ of restrictive asylum policies that are intended to divert ref-
ugees and asylum seekers from the own territory. However, without
164
Conclusion 165
tackling the root causes of emigration, refugees and other migrants

switch to alternative, mostly illegal migration routes. Therefore, it is
argued that more proactive policy alternatives like aid transfers
should be applied to overcome this race to the bottom. Furthermore,
it is shown that aid transfers towards the conflict-neighboring coun-
tries are clearly beneficial to the Western donor and (potential) asy-
lum country. However, this approach shifts the burden of hosting
refugees even more to the first asylum countries since the magnitude
as well as the average duration of the refugee situation increase as a
consequence of this policy. Instead, aid towards the country of ori-
gin would release both neighboring first asylum countries and the
Western donor country from their refugee burdens, although under
some certain circumstances the asylum pressure towards the Western
asylum country might also increase.
Next, we analyze an international refugee regime with and with-
out financial cooperation schemes among heterogeneous asylum
countries. The analysis shows that asylum cooperation is rather
attainable among less heterogeneous countries with respect to their
humanitarian attitude and their asylum cost schemes. Financial
transfers are able to stabilize a cooperative international asylum
regime if they compensate the respective asymmetries in cross-
border externalities or country-specific costs. This means, if coun-
tries are characterized by not too heterogeneous asylum costs, an
international cost-sharing regime is able to implement a cooperative
asylum solution.
The empirical section identifies the influence of refugee migration
movements on the aid allocation decisions of bilateral donors. First,
it has been stated that both permanently settled foreign-born
immigrants and asylum seekers influence aid allocation decisions in
favor to their home country. The analysis shows that it is important
to acknowledge influence of foreign population on domestic aid pol-
icies. This discussion is continued by analyzing the general influence
of refugee movements on long-term development aid and short-term
humanitarian aid allocation decisions of bilateral donors. It is shown
that donor states respond to all types of forced migration (that is
internally displaced persons, cross-border refugees or international
migrants seeking for asylum in donor states) by increasing their
humanitarian aid efforts, while the allocation of long-term develop-
ment assistance is rather targeted only to the countries of origin,
particularly when these countries are major sending countries of

bilateral asylum applicants in the donor country.
10.2 Policy implications
Although the complexity of the linkages between conflict, refugee

migration and aid permits only modest statements, it becomes clear
that there is an unambiguous need for alternatives to restrictive asy-
lum and immigration policies, given the increasing trend of reliance
on restrictive policies. Proactive international root-cause strategies
and burden-sharing initiatives towards conflict-affected and destabi-
lized regions are thus unavoidable.
Despite the relative openness of their societies and economies,
asylum seeking and other low-skilled immigration are increasingly
unpopular in the public perception in Western countries. This is
related to the apparently negative economic, political and social
effects. Through the establishment and consolidation of transna-
tional migration networks, unwanted migration movements have
become increasingly difficult for governments to avoid or to control.
Many Western governments still focus only on one-dimensional
restrictive asylum and immigration policies that are, however, severely
insufficient and overstrained since they often neglect to address the
root causes of such migration movements. Instead of curbing emigra-
tion, restrictive asylum policies tend to stimulate irregular immigra-
tion, human trafficking and smuggling, or alternatively but often not
better, force refugees to live for a protracted period of time in remote
areas and desperate situations in the regions of origin.
Therefore, the author of this book pleads generally for a more
comprehensive asylum and refugee policy approach of Western
countries. Such an approach would combine the following non-
substitutable policies. A stepwise liberalization of asylum and migra-
tion policies, including the expansion of legal access opportunities
for low-skilled migrants, should reduce both the pressure on asylum
systems and the number of irregular (undocumented) immigrants.
This would diminish inefficiencies in the allocation of economic
(and human) resources, mitigate the pressure on asylum systems in
Western asylum countries, and avoid repatriation (deportation) costs
when more asylum seekers are allowed to stay. However, it is
questionable whether Western governments are willing to follow
Conclusion 167
this path, since migration movements are not finite and a significant
decrease in South–North migration pressure is hardly to be expected
in the near future.
Since a liberalized asylum policy alone is hardly implementable
and sustainable, it needs to be complemented by some further pro-
active initiatives such as promoting refugee protection in the region
of origin. This would require that Western countries extend the nec-
essary financial and technical assistance to share the burden of host
countries in the conflict-affected regions. Acknowledging their co-
responsibility for refugees in poor and overburdened first asylum
countries is not only an ethical proviso for Western countries, but
also a long-term investment in conflict prevention and regional sta-
bility, and thus, in the prevention of future asylum migration flows
that might reach the own territory. Therefore, directing significant
financial and technical support to highly burdened first asylum
countries for promoting their economic and human development,
stabilizing their political and social coherence, and assisting local
integration and self-reliance of hosted refugees, needs to be another
integral part of a comprehensive refugee and asylum policy. However,
these efforts are not sufficient since they tend to shift refugee-
hosting burdens even more to the first asylum countries, which there
may increase social, political and ethnic tensions and overstrain
local institutional capacities.
Consequently, these efforts need to be complemented by activities
that aim to address the root causes of conflict, persecution and eco-
nomic despair in countries of origin. A targeted and country-specific
use of appropriate aid interventions, that is, long-term development
assistance and short-term emergency aid – beside other policies such
as peace-building (diplomatic or military) interventions for human
rights protection and conflict prevention and resolution as well as
more market-based instruments such as FDI and trade promotion –
might be effective for tackling the root causes of emigration. Ideally,
aid interventions should be extended before large emigrations are
expected, but latest, once mass emigration outflux has begun.
Concerted aid interventions must be implemented to reduce further
outflows and encourage voluntary return of refugees back to their
countries of origin.
The crucial and most controversial aspect of such a policy approach
is the migration-preventive function of aid. Aid effectiveness is often
considered as restricted since the scope and duration of aid programs

are too limited to have any significant effect (de Haas, 2006). Aid for
the improvement of living conditions, that is the root causes of
refugee movements, may only be migration-preventive if a relatively
small geographical area with a relatively small number of individuals
is assisted with relatively large amounts of aid. However, if only mod-
est aid amounts are available for highly fragile conflict-ridden and
post-conflict countries, the migration-preventive potential of aid is
rather limited to improve the local living conditions in such a way
that substantial outflows could be avoided.
Furthermore, the measurement of the migration-preventive effects
of aid might be one reason why donor states are so far rather hesitant
to increase aid amounts for that purpose. Since preventive policies
aim to reduce an unwanted incidence, aid must prove to be effective
in reducing migration flows. The evaluation of migration-preventive
policies, however, is only possible with an unambiguous predefined
benchmark. Additionally, since refugee migration is a complex phe-
nomenon that is determined by a multiplicity of factors, it is difficult
to evaluate ex post to what extent aid assistance has contributed to a
reduction in refugee migration movements. That is, even if emigra-
tion numbers from a region or a country decrease during an ‘aid
cure’, it cannot necessarily be concluded that it has taken place due
to the provision of aid. Moreover, as has been shown in Chapter 5,
aid towards the countries of origin could also increase asylum migra-
tion movements towards the Western countries, which would run
contrary to the donor interests. Thus, uncertainty about the impact
of aid on migration propensities makes donors still rather passive to
tackle this issue by foreign aid. Thus, more research needs to be
done on the migration-preventive effectiveness of foreign aid for
improving knowledge on its efficacy and reducing the uncertainty
and unwillingness of Western countries to implement this policy
option.
Finally, asylum migration movements often happen in the context
of transnational, social networks. These cannot be changed by one
(affected) country alone, and thus, international cooperation is
essential. Generally, it is obvious that the control of transnational
migration movements and the mitigation of their consequences
must take place in an environment of intergovernmental and inter-
national cooperation, integrated with non-governmental groups and
Conclusion 169
civil society. It is particularly important that instruments for the

control of migration and asylum movements are improved, that is
clear and transparent immigration regulations and their intergov-
ernmental coordination and harmonization. For monitoring and
coordinating forced migration movements, an international refugee
regime is necessary, which specifies the obligations sending, transit
and asylum countries have with respect to admission, repatriation
and reintegration. Since, as laid out in Chapter 6, the willingness of
highly heterogeneous host states to set up a cooperative regime is
rather low, intra-regional cooperation among less asymmetric coun-
tries seems more feasible in this respect.
For Western industrialized countries, a comprehensive asylum and
migration policy approach that also contains proactive (aid) policy
elements is definitely more promising than the status quo of one-
dimensional restrictiveness of asylum and immigration regulations.
Although this is not a cure-all of the deficiencies of the international
refugee regime, it has the potential (i) to improve the individual pro-
tection and living conditions of those that flee from conflict, vio-
lence and persecution, (ii) to reduce the pressure on (and abuse of)
asylum systems in Western countries, and (iii) to alleviate the bur-
dens of overstrained first asylum countries in the developing world.
Appendix
Table A.1 Specifications of the Refugee Capacity Index
Specification Description and Source
log yi log y min Income per capita in constant

GDPI i US$, purchasing power parity,
log y max log y min
2002. Goalposts: 100 US$ (min)
and 40,000 US$ (max); Data:
UNDP (2004): Human
Development Report
log yi log y min Population in mill. in 2003 per
GDPI i square kilometer of arable land.
log y max log y min
Maximum goalpost: 1 mill. per
km2; Data: World Bank (2004):
World Development Indicators
1 § · Shares of major ethnic, linguistic

3 ¨© ¦ ¦ li2,n ¦ ri2,o ¸¹
ELRI i ei2,m and religious groups in country i;
m n o Data: Alesina et al. (2003)
di Average score of civil liberties
PFI i index and political rights index,
7
redefined on a range from 0 to 1;
Data: FreedomHouse (2005):
Country ratings
Original score is re-scaled on a
gi
PSI i range from 0 to 1; Data:
5 Kaufmann et al. (2003)
170
Appendix 171
Table A.2 Descriptive statistics, Aceh sample
Variable Definitions Mean St Dev Min Max
Population Change in village population 0.004 2.523 –37.47 21.16

change (in hundreds) between 1999
and 2002
Conflict Dummy variable set to one if a 0.238 – 0 1
conflict with or without
casualties has been reported
(within the previous year,
reported in 2002)
Conflict Dummy variable set to one if 0.347 – 0 1
cluster at least 20% of the villages
within the same subdistrict
(kecamatan) reported conflict
(within the previous year,
reported in 2002)
Urban Dummy variable set to one if 0.080 – 0 1
municipality was officially
considered as urban in 1999
(classification is based on
economic structure and access
to services)
Altitude Altitude in thousand meter 0.178 0.311 0 2
above sea level
Distance (dist Distance to the district 0.463 0.345 0 4.85
office) (Kabupaten) office measured in
hundreds of kilometers in 1999
Poor families Share of families in the village 0.546 0.268 0 1
who were officially registered
as poor in 1999
Fertile couples Number of fertile age couples in 0.986 1.179 0 15.1
1999 in the villages (in
hundreds)
Transport Dummy variable set to one if 0.061 – 0 1
station transport station (bus, train,
airport, seaport) is available in
the village in 2002 (for large
towns is also set to 1 if other
parts of the town have a
station)
Continued
172 Appendix
Table A.2 Continued
Variable Definitions Mean St Dev Min Max
Police Dummy variable set to one if 0.658 – 0 1

presence police station was not far/very
far to reach in 1999
Share of Share of Javanese population 0.087 0.140 0 0.689

Javanese within the subdistrict
(kecamatan), estimations are
based on Indonesian Census
2000
Population Village population in 1999 (in 0.707 0.801 0.047 9.681
thousands)
Central Aceh District dummies 0.037 – 0 1
West Aceh District dummies 0.051 – 0 1
Nagan Raya District dummies 0.040 – 0 1
Southwest District dummies 0.024 – 0 1
Aceh
South Aceh District dummies 0.046 – 0 1
North Aceh District dummies 0.157 – 0 1
East Aceh District dummies 0.089 – 0 1
Aceh Tamiang District dummies 0.040 – 0 1
Langsa (city) District dummies 0.010 – 0 1
Lhokseumawe District dummies 0.013 – 0 1
(city)
Banda Aceh District dummies 0.016 – 0 1
(city)
Notes: All descriptive statistics pertain to the N = 5,211 villages in our sample.
Appendix 173
Table A.3 Derivations of the refugee migration model
wV W
(1) (1 ␣ )(1 r )( y O M W )␣ p ␣ ! 0.
wp
w 2V W
(2) ␣(1 ␣ )(1 r )( y O M W )␣ p ␣ 1 0.
wp 2
1
§ ( y O M W )␣ r ( y M M W )␣ sy S · 1 ␣
(3) p* ¨ ¸ if 0 d p* d p.
© r 1 ¹
1
§ r ( y W M W )␣ · 1 ␣
(4) p** ¨ O ␣ W ␣ ¸
if p d p** d 1.
© [ y ] (1 r )( y M ) ¹
O
␣ y S ( sy S )␣ 1
° ! 0, 0 d pi d p.
wV S ° (1 ␣ )y S ( p ␣ 1)y O ␣ psy S °
(5) ®Z ! 0, if ® p pi p.
ws ° ( y O
sy S 2
) °
° 0
0, ¯ p d pi d 1.
¯
wV S ( p ) wV S ( p )
Proof: ␣ y S ( sy S )␣ 1 ! 0, and 0.
ws ws
wV W
(6) ( y W M W )␣ p1␣ ( y O M W )␣ ! 0.
wr
0
° 0, 0 d pi d p.
wV S
° (1 ␣ )(␣ y sy S (1 p(␣ 1)))
O
°
(7) ®Z ! 0, if ® p pi p.
wy O ° ( y O
sy S 2
) °
! 0, ¯ p d pi d 1.
° ␣ p1␣[ y O ]␣ 1
¯
wV W
(8) ␣(1 r ) p1␣ ( y O M W )␣ 1 ! 0.
wy O
␣ 1
wV W § r ( yW y O ) y O M W ·
(9) ¨ ¸ ( y W y O ) ! 0.
wr © (1 r ) p r ¹
␣
r ( yW M W )␣
␣r ( yW M W )␣ §¨ ·1 ␣ ([ y O ]␣ 1 ( r 1)( y O M W )␣ 1 )
¸
(10) wp © [ y O ]␣ ( r 1)( y O MW )␣ ¹
**
0.
wy O (1 ␣ )([ y O ]␣ ( r 1)( y O M W )␣ )2
Proof: [ y O ]␣ 1 ( r 1)( y O M W )␣ 1 ! 0, since y O M W t 0.

␣
§ ( y O MW )␣ ( [ sy S ]␣ ) r ( yW M W )␣ ·1 ␣
␣( y O M W )1 ␣ ( [ sy S ] r ( yW M W )␣ ) ¨ ¸
(11) ␦ p * © ( r 1) ¹
0.
␦y O (1 ␣ ) ( r 1)
174 Appendix
Table A.4 Descriptive statistics, EU-27
EU-27 I II III IV V VI VIIa VIIb VIII
Austria 15.0 17,874 8.6 8,048 0.017 1.45 4.8 51.6 4.2
Belgium 31.7 21,072 17.6 10,333 0.021 1.36 6.1 31.2 6.4
Bulgaria 63.4 1,259 14.3 7,965 0.016 1.47 8.8 30.8 9.2
Cyprus 14.1 1,011 4.6 0,765 0.002 1.66 7.2 37.7 7.6
Czech Rep. 6.2 6,499 13.0 10,210 0.021 1.50 2.9 30.4 8.1
Denmark 63.6 7,981 14.1 5,374 0.011 1.35 7.4 23.8 12.7
Estonia 16.7 14 10.3 1,358 0.003 3.75 6.9 36.7 3.1
Finland 54.6 1,924 23.3 5,199 0.011 1.46 10.4 34.7 11.6
France 19.2 35,028 16.4 59,485 0.123 1.36 5.6 33.7 6.4
Germany 12.3 95,817 14.6 82,495 0.170 1.34 4.8 32.5 4.6
Greece 9.8 3,554 15.4 10,631 0.022 1.57 6.2 44.9 17.7
Hungary 32.5 4,546 4.2 10,159 0.021 1.32 2.1 12.1 6.1
Ireland 11.1 5,915 11.1 3,920 0.008 1.54 8.3 46.7 15.2
Italy 15.0 10,519 14.1 57,69 0.119 1.48 9.7 47.4 6.0
Latvia 8.8 22 7.4 2,338 0.005 4.30 7.0 26.9 3.9
Lithuania 16.9 221 11.1 3,469 0.007 1.83 4.4 36.5 2.0
Luxembourg 13.8 1,069 7.0 0,444 0.001 1.70 7.4 51.9 5.7
Malta 48.4 183 11.1 0,397 0.001 1.63 1.5 30.0 7.4
Netherlands 49.4 33,494 12.2 16,144 0.033 1.34 3.9 35.5 7.4
Poland 9.0 3,568 15.8 38,626 0.080 1.47 6.2 18.5 2.1
Portugal 16.3 325 7.2 10,177 0.021 1.90 11.6 61.4 12.3
Romania 13.3 1,223 7.0 22,300 0.046 1.68 17.8 38.8 4.5
Slovakia 31.7 3,315 24.2 5,379 0.011 1.40 2.6 27.3 6.4
Slovenia 9.7 1,403 7.5 1,964 0.004 1.69 4.6 48.1 8.3
Spain 15.0 7,208 19.4 40,917 0.085 1.54 16.5 59.6 8.1
Sweden 48.3 17,136 15.5 8,924 0.018 1.42 16.3 54.5 8.5
UK 37.1 68,003 9.5 59,229 0.122 1.32 4.4 34.0 10.7
Notes: (I) Average Asylum Recognition Rates, 1993–2002, in percent, (UNHCR, 2004b),
(II) Average Asylum Applications, 1994–2003, (UNHCR, 2004b), (III) Unemployment
rate, foreign-born population, average 1995–2005, (OECD, 2006b), (IV) Population,
2003, in million, (World Bank, 2004), (V) EU population share, 2003, in percent, own
calculation, (VI) Country-specific cost exponents, own calculation, (VII) Immigration
policy attitude: ‘How about people from other countries coming here. What do you
think the government should do?’ (VIIa) Response (in percent): ‘Let anyone come’
(VIIb) Response (in percent): ‘As long as jobs available’. (WVS, 2000), (VIII) General
altruism: ‘To what extent do you feel concerned about the living conditions of human
kind’, Response (in percent): ‘very much’, (WVS, 2000).
Appendix 175
Table A.5 Descriptive statistics, UN speeches sample
Mean Median Max Min St Dev
ODA bilateral 136.74 11.80 14,795.60 –0.60 644.58

ODA multilateral 1,054.49 586.55 9,704.60 2.80 1463.28
Scouncil 26.16 8.50 298.00 0.00 43.89
Gassembly 198.32 135.00 922.00 2.00 195.41
Income 2,230.12 1,046.39 23,023.33 100.56 3,249.91
Population 32,387.53 6,750.17 1,203,797.00 72.33 126,531.60
PQLI 69.47 74.41 87.51 33.72 13.67
Democracy 4.20 4.11 7.00 1.00 1.73
Governance 2.20 2.18 4.18 0.50 0.63
Conflict 5.03 0.00 85.00 0.00 10.47
ODA history 9.76 7.00 30.00 0.00 10.08
Trade 268,819 17,367 55,002,671 2 1,533,846
Distance 7,247.12 7,074.27 18,587.08 59.62 3873.05
Colony 0.04 0 1 0 0.21
Table A.6 Data description and sources, UN speeches sample
Description Source
ODA Log of sum of bilateral ODA OECD (2004): International

bilateral disbursements (in mill. Development Statistics
constant US$ of 2002)
ODA Log of sum of multilateral ODA OECD (2004): International
multilateral disbursements (in mill. Development Statistics
constant US$ of 2002)
Scouncil Number of Speeches in the UN United Nations (2005b):
Security Council Bibliographic Information
System
Gassembly Number of Speeches in the UN United Nations (2005b):
General Assembly Bibliographic Information
System
Income Log of GDP per capita (in World Bank (2004): World
constant US$ of 1995) Development Indicators
Population Log of average number of World Bank (2004): World
inhabitants (in thousands) Development Indicators
PQLI Average of literacy rate, adjusted Own calculation based on
mortality rate and adjusted life data from World Bank
expectancy (scaling from 0 (2004): World Development
(worst) to 100 (best)). Indicators
Continued
176 Appendix
Table A.6 Continued
Description Source
Democracy Average of Civil Rights Index FreedomHouse (2005):

and Political Rights Index Freedom in the World
ranging from 1 (best) to 7
(worse)
Governance Average of good governance Kaufmann et al. (2003):
indices on voice and Governance matters III:
accountability, political stability, governance indicators for
government effectiveness, 1996–2002
regulatory quality, rule of law,
and corruption control
Conflict Number of conflicts times PRIO (2005): Armed Conflict
conflict intensity (1=low, Database
2=medium, 3=high)
ODA history Number of bilateral Own calculation based on
development cooperation OECD (2004): International
between 1960 and 1989 Development Statistics
Trade Bilateral trade volume (in OECD (2005): Bilateral Trade
thousand US$) Database
Distance Bilateral geographical distance CEPII (2005): Geographical
between countries, that is their Distances Dataset
main cities or agglomerations
(in kilometer)
Colony Dummy variable on bilateral CIA (2004): World Factbook
colonial history
Table A.7 Descriptive statistics, German aid allocation sample
ODA 388.67 149.45 7,019.60 0.30 819.88

Population 32,387.53 6,750.17 1,203,797.00 72.33 126,531.60
Income 2,230.12 1,046.39 23,023.33 100.56 3,249.91
PQLI 69.47 74.41 87.51 33.72 13.67
Asylum seeker 12,047.56 533.00 450,775.00 0.00 45,739.65
Foreign pop 34,292.29 2,497.50 2,066,917.00 2.40 182,900.40
ODA ties 18.35 25.00 30.00 0.00 11.52
Distance 6,293.63 5,941.79 16,654.11 546.00 3,430.66
Democracy 4.20 4.11 7.00 1.00 1.73
Trade 681,599 68,085 8,836,294 225 1,498,291
Eastern Europe 0.13 0 1 0 0.34
Appendix 177
Table A.8 Data description and sources, German aid allocation sample
Description Source
Log(ODA) Log of sum of German ODA OECD (2004):

disbursements (in mill. International
constant US$ of 2002) Development Statistics
Log(Population) Log of average number of World Bank (2004):
inhabitants (in thousands) World Development
Indicators
Log(Income) Log of GDP per capita (in World Bank (2004):
constant US$ of 1995) World Development
Indicators
Log(PQLI) Log of average of literacy Own calculation based
rate, adjusted mortality rate on data from World Bank
and adjusted life expectancy (2004): World
scaling from 0 (worst) to 100 Development Indicators
(best)
Log(Asylum Log of average number of UNHCR (2003):
seeker) asylum applications (also Statistical Yearbook,
pending cases) Database
Log(Foreign Log of average stock of MPI (2005): Migration
pop) foreign-born population Information Source
Log(ODA ties) Log of number of years Own calculation based
German bilateral on OECD (2006a):
development cooperation International
between 1960 and 1989 Development Statistics
Log(Distance) Log of bilateral geographical CEPII (2005):
distance between countries, Geographical Distances
that is the main cities or Dataset
agglomerations (in
kilometers)
Democracy Average of Civil Rights FreedomHouse (2005):
Index and Political Rights Freedom in the World
Index ranging from 1 (best)
to 7 (worse)
Log(Trade) Log of average German OECD (2005): Bilateral
export volumes (in Trade Database
thousand current US$)
Eastern Europe Dummy variable for Eastern
European countries
178 Appendix
Table A.9 Descriptive statistics, aid panel dataset
ODApc 2.34 0.07 883.09 0 16.00

EMApc 0.06 0.00 44.65 0 0.57
Asylum
seeker 175 1 115,395 0 1,596
Ref asylum 70,291 1,987 4,150,723 0 242,891
Ref origin 72,484 857 4,552,153 0 288,685
IDP 32,081 0 1,290′ 0 133,514
Population 33.60 7,265′ 1,290′′ 71,079 130
GDPpc 4,684 3,446 25,168 440 4265
Bilateral
trade 0.0014 0.0001 0.37 0 0.0063
CPRI 4.12 4.00 7.00 1 1.82
External
debt 1.13 0.59 47.09 0.00 3.36
PQLI 69 77 98 3 22.5
Battle
deaths 484 0 48,034 0 2,858.3
Natdis
deaths 328 0 30,005 0 2,245.2
Table A.10 Correlation matrix, aid panel dataset
(1) (2) (3) (4) (5) (6) (7) (8) (9) (10) (11) (12) (13) (14)
(1) ODApc 1.00 0.08 0.02 –0.03 –0.03 –0.02 –0.03 0.06 –0.01 –0.07 –0.01 0.04 –0.02 –0.02
(2) EMApc 0.08 1.00 0.02 0.12 –0.01 0.04 –0.02 –0.06 –0.02 0.03 –0.00 –0.04 –0.00 0.01
(3) Asylum seeker 0.02 0.02 1.00 0.03 0.06 0.03 0.08 –0.00 0.11 0.04 –0.02 0.02 0.02 0.04
(4) Ref origin –0.03 0.12 0.03 1.00 0.10 0.31 0.03 –0.22 –0.01 0.34 –0.00 –0.25 0.01 0.16
(5) Ref asylum –0.03 –0.01 0.06 0.10 1.00 –0.01 0.15 –0.12 0.03 0.25 –0.03 –0.12 0.06 0.05
(6) IDP –0.02 0.04 0.03 0.31 –0.01 1.00 –0.02 –0.08 –0.01 0.14 –0.02 –0.06 –0.03 0.16
(7) Population –0.03 –0.02 0.08 0.03 0.15 –0.02 1.00 –0.04 0.31 0.12 –0.04 0.04 0.40 0.05
(8) GDPpc 0.06 –0.06 –0.00 –0.22 –0.12 –0.08 –0.04 1.00 0.17 –0.41 0.18 0.59 –0.06 –0.06
(9) Bilateral trade –0.01 –0.02 0.11 –0.01 0.03 –0.01 0.31 0.17 1.00 –0.02 0.05 0.15 0.07 0.03
(10) CPRI –0.07 0.03 0.04 0.34 0.25 0.14 0.12 –0.41 –0.02 1.00 –0.11 –0.42 0.02 0.16
(11) External debt –0.01 –0.00 –0.02 –0.00 –0.03 –0.02 –0.04 0.18 0.05 –0.11 1.00 0.01 –0.02 0.00
(12) PQLI 0.04 –0.04 0.02 –0.25 –0.12 –0.06 0.04 0.59 0.15 –0.42 0.01 1.00 –0.03 –0.09
(13) Natdis deaths –0.02 –0.00 0.02 0.01 0.06 –0.03 0.40 –0.06 0.07 0.02 –0.02 –0.03 1.00 0.04
(14) Battle deaths –0.02 0.01 0.04 0.16 0.05 0.16 0.05 –0.06 0.03 0.16 0.00 –0.09 0.04 1.00
180 Appendix
Table A.11 Data definitions and sources, aid panel dataset
Variable Definition Source
Ln(ODApc) Log of Official Development OECD (2006a):

Assistance per capita (in 2002 International
US$) Development Statistics
Ln(EMApc) Log of Emergency aid per OECD (2006a): Creditor
capita (in 2002 US$) Reporting System
Ln(Asylum Log of number of bilateral UNHCR (2006c):
seeker) asylum applications registered Statistical Yearbook,
in donor country CD-ROM
Ln(Ref Asylum) Log of number of refugees UNHCR (2006c):
registered in recipient country Statistical Yearbook,
CD-ROM
Ln(Ref Origin) Log of number of refugees UNHCR (2006c):
originating from recipient Statistical Yearbook,
country CD-ROM
Ln(IDP) Log of number of internally UNHCR (2006c):
displaced persons in recipient Statistical Yearbook,
country CD-ROM
Ln(Population) Log of number of inhabitants World Bank (2006): World
(in mill.) Development Indicators
Ln(GDPpc) Log of GDP per capita (in World Bank (2006): World
current US$, PPP) Development Indicators
Ln(Bil. trade) Log of bilateral trade volume OECD (2005):
(exports in current US$) as International Trade by
share of total exports Commodity Database
CPRI Average of Civil Rights Index FreedomHouse (2006):
and Political Rights Index Freedom in the World
ranging from 1 (best) to 7
(worse)
External debt Ratio of External debt (in mill. JEDH (2007): Joint
of current US$) to GDP (in External Debt Hub
mill. of current US$)
PQLI Average of literacy rate, Own calculations based
adjusted mortality rate and on data from World Bank
adjusted life expectancy, (2006): World
scaling from 0 (worst) to 100 Development Indicators
(best)
Ln(Battle Log of number of annual PRIO (2005): Armed
deaths) battle fatalities Conflict Database
Ln(NatDis Log of number of casualties CRED (2006): EM-DAT:
deaths) caused by natural disasters Emergency Event Database
Notes
1. The restrictions that have been introduced are mostly concerned with
the reception conditions of asylum seekers for the purpose of reducing
costs and to prevent the abuse of the asylum system by so-called ‘bogus
refugees’ (Neumayer, 2005b). For instance, restrictive measures related to
housing, detention, dispersion, deportation, reduction of welfare bene-
fits, access to the labor market, re-definition of the term ‘refugee’, intro-
duction of ‘temporary protection’ and ‘safe third country’ principles,
intensified border and internal controls, and so on (Boswell & Crisp,
2004).
2. Each chapter, except Chapters 2 and 7, is based on an article already pub-
lished or currently under review in a peer-reviewed economics or political
science journal. Therefore, it has to be acknowledged that each of these
chapters can (and should) be read and understood as self-contained
units.
3. Political and social costs, such as anxieties on the influence of large-scale
immigration on identity or security, are much too abstract to be estimated
in monetary terms. Only the monetary costs for the reception and appli-
cation processing of asylum seekers have roughly been estimated for some
Western asylum countries by Martin et al. (2005).
4. Despite the fact that some authoritarian states also constrain emigration,
they are not included in this analysis.
5. For instance, in Germany, the Residence Act (Section 29) defines the gen-
eral conditions for the subsequent immigration of family members
accordingly by (a) the foreign family member already living in Germany
has a residence or settlement permit, (b) sufficient living space is avail-
able, (c) the family member has sufficient means to support him or herself
without recourse to public funds, and (d) no grounds for expulsion exist.
6. For instance, the German green card program for IT-specialists has been
in place since 2000, or the proposed pan-European program of the
European Commission that is targeted at highly skilled workers.
7. This number covers persons who were recognized according to the Geneva
Refugee Convention or the Refugee Convention of the Organization for
African Unity (OAU), as well as on persons with a temporary residence
right for humanitarian reasons.
8. Irregular migration is another type of migration, which probably consists
of the fastest growing group of migrants. However, since they are statis-
tically not recorded, the number can only be roughly estimated. It is
estimated that there are between 15 to 30 million illegal or irregular
immigrants worldwide (ILO, 2004).
9. According to the 1951 Convention and the 1967 Protocol relating to the
Status of Refugees, a refugee is someone who ‘(i) has a well-founded fear of
181
182 Notes
persecution because of his/her race, religion, nationality, membership in a

particular social group, or political opinion; (ii) is outside his/her country of
origin; and (iii) is unable or unwilling to avail him/herself of the protection of
that country, or to return there, for fear of persecution.’ In 1969, the
Convention Governing the Specific Aspects of Refugee Problems in
Africa of the Organization of African Unity (OAU), added to this defini-
tion of a ‘refugee’ of the UN Refugee Convention a broader considera-
tion of any person ‘compelled to leave his/her country owing to external
aggression, occupation, foreign domination or events seriously disturbing pub-
lic order in either part or the whole of his country of origin or nationality.’ In
1984, the Cartagena Declaration adds a more objectively based consider-
ation to the 1951 UN Convention by including persons who flee their
countries ‘because their lives, safety or freedom have been threatened by gen-
eralized violence, foreign aggression, internal conflicts, massive violation of
human rights or other circumstances which have seriously disturbed public
order.’ (UNHCR, 2001, p. 9).
10. Although the estimated number of environmental refugees is expected
to double between 1995 and 2010 (Myers, 2002), this type of refugee is
not included in this study, since they are not captured by the official UN
definitions and therefore not recognized by national states as eligible for
asylum.
11. For this reason, the UN Security Council authorized humanitarian inter-
ventions, for example, in Northern Iraq, in Somalia, in ex-Yugoslavia
and in Haiti, in order to prevent mass flight movements.
12. For instance, several times military support was engaged in order to be
able to manage the immense logistical challenges of caring for large
numbers of refugees (de Mello, 1995).
13. See, for example, the overview by Angenendt (2002).
14. Several indicators for this development are identifiable, for example sev-
eral millions waiting for legal immigration to the US (Smith & Edmonston,
1997), and the continuous flow of irregular immigration into Western
countries (estimated at 300,000 per year in the US, and 400,000–500,000
per year in the EU) (see Hatton & Williamson, 2002).
15. In 2005, after the dramatic refugee situation in Ceuta and Melilla,
José Manuel Barroso, president of the European Commission, stated that
‘The problem of immigration, the dramatic consequences of which we are wit-
nessing, can only be addressed effectively [ ... ] through an ambitious and coor-
dinated development [plan] to fight its root causes’ (cited from de Haas,
2006).
16. UNHCR and ILO started a project in 1992 with the working title ‘Aid in
place of migration?’ to investigate whether more or different develop-
ment aid could diminish unwanted emigration (see Böhning & Schloeter-
Paredes, 1994).
17. This chapter is based upon Czaika (2005b).
18. See, for example, the issue of Journal of Refugee Studies, Vol. 16, No. 3,
2003.
Notes 183
19. See, for example, UNHCR’s Convention Plus Initiative, founded in 2003,
or the debate in the EU about asylum reception centers in North Africa.
20. For instance, UNHCR’s recent statistical yearbooks contain three differ-
ent indicators of host country capacity and contributions (for a discus-
sion see UNHCR, 2002) that give a good first hint as to a country’s borne
refugee burden.
21. See, for example, Bach & Carroll-Seguin (1986) and Wooden (1991) for
South-East Asian refugees in the US and in Australia, or Hauff & Vaglum
(1993) for Vietnamese refugees in Norway.
22. UNHCR defines a protracted refugee situation as a refugee population of
25,000 or more persons living in exile for five or more years in the host
country. At the end of 2003, there were 38 major protracted situations
affecting about 6.2 million refugees (UNHCR, 2003).
23. The author is aware of the economic and sociocultural contributions of
refugees for a host country. Particularly young immigrants can produce
a net benefit for the host country notably if they stay in that country
throughout their lifetime (Borjas, 1994).
24. See Table A.1 in the Appendix for further information about the RCI
component specifications and the data sources.
25. See Table A.1 in the Appendix for the methodology of the ELRI.
26. For instance, political rights refer to fairness in elections with a real pos-
sibility to overtake power via elections, freedom for organizing in par-
ties, the existence of party competition and allowance for opposition,
and so on. Civil liberties reflect the freedom of assembly, freedom for
any religious activities, freedom of the media, protection from political
suppression, and so on.
27. See Table A.1 in the Appendix.
28. We took this indicator from the governance indicators data set of
Kaufmann et al. (2003). This indicator is based on several different
sources, including polls of experts and surveys of residents and entrepre-
neurs of a specific country. A linear unobserved components model is
used to aggregate these various sources into one aggregate indicator. It is
then normalized such that it ranges from –2.5 to +2.5 and has a mean of
zero with a standard deviation of one. A higher value signals a higher
degree of political stability.
29. Sensitivity analysis has shown that changes in the weights of the RCI
components have little effect on the country rankings (that is Spearman
rank correlations of alternative weight combinations are always positive
and strong significant).
30. It is to be noted that a small amount of data that was missing from the
data sets for some sub-indices has been sourced from the CIA World
Factbook (CIA, 2004).
31. Countries at the bottom of our RBI 03 ranking (not reported) were either
actual refugee-producing countries like Sri Lanka (RBI 03 score of about
–1.00) or countries that are located more in a regional periphery which is
the case for most island countries.
184 Notes
32. The complete output of RBI scores for all 174 countries is not reported,
but is available on request.
33. This regional classification of the countries reflects UNHCR’s arrange-
ments of regional bureaus and operations. This classification takes into
account cultural similarities and geographical coherence.
34. The complete results for RBI 03 on a regional basis are available from the
author on request.
35. ‘Caswaname’ describes the region of Central Asia, South-West Asia,
North Africa and the Middle East.
36. This chapter is based upon Czaika & Kis-Katos (2009).
37. In what follows, we use the term village for both villages and urban
neighborhoods.
38. Villages with conflict constitute nearly the same proportion of Aceh as
a whole and in our matched 91.3 percent sample. As a comparison,
PODES 2003 reports a share of conflict of 23.8 percent in the whole Aceh
province.
39. The variable ‘population change’ is corrected by the numbers of deaths
because of conflicts or epidemics.
40. Another possibility to examine the push and pull factors of migration
would be to define indicator variables of large population outflow (or
inflow), and estimate probit or ordered probit models. However, by
applying this procedure we would lose information on the intensity of
population change. Additionally, as our population change measure is
relatively noisy, we find it less obvious to identify a discrete regime
switch between outflow and inflow villages. This is the main reason why
we prefer the use of quantile regressions, which allows us a smoother
identification of inflow and outflow villages.
41. The number of deaths by conflict remains far below the population
flows due to displacement. However, changes in fertility behavior might
have played a significant role as well, since fertility might have been
strongly reduced in conflict-ridden areas. For village-specific mortality
factors we partly correct by subtracting from population decrease the
deaths by conflict and by epidemics over the last period.
42. Descriptive statistics and definitions of variables are presented in
Table A.2 in the Appendix.
43. We experimented with these alternative measures and all yielded very
similar results.
44. These welfare criteria include food consumption habits, access to health
care, the possession of alternative sets of clothing, information on the
floor material of the dwelling, and on the household members’ ability to
practice their religion.
45. These families were the main targeted beneficiaries of the Social Safety
Net Program of the Indonesian government, which addressed rising
poverty during and after the economic crisis. Thus, the variable might
also reflect a certain policy bias; villages with a larger share of poor fam-
ilies might also be those with better social safety coverage.
Notes 185
46. For a few subdistricts with missing data, district-level averages or aver-
ages of the neighboring districts have been imputed.
47. This pattern might be amplified by measurement errors in village popu-
lation if they were larger in these most affected regions.
48. This chapter is based upon Czaika (2008b).
49. This assumption does not change the qualitative results of the subse-
quent model.
50. For simplicity reasons, asylum recognition rates are not specific to indi-
viduals but to the whole population of asylum seekers.
51. We further assume that before the conflict broke out at time t = 0 – e,
only economic migrants that could afford the high migration costs M W
left the country of origin towards the Western country. For the remain-
ing N individuals staying permanently at home is the superior solution
in times of peace.
52. For instance, economic self-reliance might be restricted by confinement
to a refugee camp or a closed settlement, constrained work permission,
limited land ownership, and so on, all of which lessens human capital
productivity and the potential to generate income in the first asylum
country.
53. At this point of time, both asylum countries have already decided on
their respective asylum policies.
54. This implication has been confirmed in several empirical studies provid-
ing evidence for the negative impact of low recognition rates on asylum
applications in EU member states (Vink & Meijerink, 2003; Neumayer,
2004; Neumayer, 2005a).
55. In this model, we consider only the case in which rejected asylum seek-
ers are deported to their country of origin, acknowledging that this is
often not accomplished by asylum states because of the non-refoulement
proviso of the 1951/67 UN Refugee Convention. In fact, rejected asylum
seekers are often sent back to a safe third country.
56. For reasons of simplicity we assume that the migration movement itself
and the asylum procedure are not time-consuming.
57. In fact, we exclude the case that a rejected and deported asylum seeker
emigrates again to the first asylum country. This model variation is pos-
sible, but makes it less tractable.
58. See Table A.3 in the Appendix for some explicit calculations of p*
and p**. wv S ( p ) wvW ( p )
59. If the two curves osculate, that is, if wp wp , it follows p* = p**,
indicating no asylum migration to the Western country.
60. Obviously, the number of admitted asylum seekers is ASW* = r N(p** – p*).
61. See derivation of the latter in Table A.3 in the Appendix.
63. If the Western asylum country held no bias for asylum seekers from a
specific background, that is if V W (p) is equal for asylum seekers from
different countries, then asylum recognition rates tend ceteris paribus
to be lower for asylum seekers from countries close to the Western
186 Notes
destination, since migration costs for them are presumably more moder-
ate (Proof: drW > 0 for dV W = 0).
dM
65. See derivations in Table A.3 in the Appendix.
66. The simulation is run with the following parameter values:
a = 0.8; r = 0.38; yS = 100; yW = 300; M W = 70.
67. These costs might capture all type of costs that the Western asylum
country accrues for hosting asylum seekers (for example, administration,
maintenance, repatriation and deportation and so on).
68. This chapter is based upon Czaika (2009).
69. Countries like Norway or Sweden received around 15 refugees per 1000
inhabitants during the period 2000 to 2004, while the European average
has been about 2.7 refugees per 1,000 inhabitants, with France (3.9) and
Germany (4.7) as the main havens in Europe in total numbers (UNHCR,
2006a).
70. During the last decade, the EU countries introduced a wide range of reg-
ulations towards asylum seekers. For instance, measures to tighten exter-
nal border controls based on the Schengen Convention (1990) and the
Maastricht Treaty (1993), universal carrier sanctions, extension of visa
restrictions, the safe third country principle, and asylum procedures
only to take place in the country of first entry according to the Dublin
Convention in 1990 (Hatton, 2005). The Amsterdam Treaty (1997) was a
first step to delegate asylum decision and enforcement power on the
supranational level. In 2000, the European Refugee Fund (ERF) as a first
step to an institutional and financial cooperation regime has been
established.
71. Noll (1997, 2003) analyzes the problem of collective action failure in the
EU and discusses the normative aspects of refugee protection. He points
out the conflicting aspects between international, EU, and national asy-
lum laws, and emphasizes the need for further harmonization of asylum
legislations.
72. In 2001, an EU directive (Council Directive 2001/55/EC, OJ L 212)
referred to the practical and ethical problem of ex post resettlement of
asylum seekers by establishing the non-binding mechanisms based on
‘double voluntarism’ (cf. Thielemann, 2006). This means that both asy-
lum seekers and recipient states are required to agree before an asylum
seeker can be moved between countries. Although this directive has not
yet been applied in the EU, this mechanism may underlie the present
model to implement a burden-sharing rule.
73. Sector lines I in Figure 6.2 are derived from rearranging participation
constraint (6) according to: i*2 c i * c i 2 2i* c i * .
2 2
2 i i
74. This aspect of concrete implementation of a financial burden-sharing
regime is beyond this general welfare analysis. For a discussion of narrow
asylum trading system, see, for example, Schuck (1997) whereas a more
comprehensive transfer system based on the comparative advantages of
states in providing public goods (cf. Boyer, 1989) is analyzed by
Notes 187
Thielemann & Dewan (2006). The latter approach may be relevant in the
context of the EU, where implicit package deals and intertemporal com-
pensations are often used to find consensus.
75. Corresponding constraints for costs asymmetries are derived from rear-
ranging participation constraints (6.15) and (6.16), respectively:
2c i * ( i * 1)2 3c i * ( i * 1)2 2 i *c i * c i *i * (i * 2)
ci , and i (i 2)
ci 2 i
.
3 (i 1)2 2 ( i 1)2
76. Note that the so-called ‘Nizza rules’ on voting in the EU Council, which
release the tight corset of unanimity, are still not yet applicable on asy-
lum affairs.
77. See, for example, the Convention Plus initiative initiated and coordi-
nated by the United Nations High Commissioner for Refugees (UNHCR).
This initiative pursues to generate multilateral agreements for (i) a strate-
gic use of resettlement as a tool of protection, (ii) an effective targeting
of humanitarian and development assistance, and (iii) a clarification
of responsibilities among nation states in case of irregular secondary
movements.
78. The corresponding question was: ‘Development aid means giving grants
or loans to developing countries which aim to promote economic devel-
opment and human welfare. We are not talking here about humanitar-
ian aid (that is assistance provided in emergency situations like war,
natural disaster, famine, and so on), but about development aid. What in
your opinion are the two main motivations for richer countries to pro-
vide development aid to poor countries? (max. 2 answers).’
79. This section refers to the article by Czaika (2008a).
80. See, for example, Isenman (1976), Dowling & Hiemenz (1985), Trumbull &
Wall (1994), Arvin & Drewes, (1998, 2001).
81. Official Development Assistance (ODA) to developing countries as well
as Official Assistance (OA) to transition and newly industrialized coun-
tries is comprised of net amounts of both grants and concessional loans
minus amortization, converted to constant 2002 US$.
82. Political rights refer, for example, to fairness in elections with real feasi-
bility to overtake power, freedom for founding a party, party competi-
tion and allowance of political opposition. Civil liberties relate to
freedom to assemble, religious freedom, freedom of the media, or protec-
tion from political suppression.
83. The governance variable composites data for voice and accountability,
political stability, government effectiveness, regulatory quality, rule of law
and corruption control. These data are originally normalized so that they
range from –2.5 to +2.5 with mean of zero and a standard deviation
of one.
84. See Tables A.5 and A.6 for data description, sources, and descriptive
statistics.
85. However, Belgium (member in 1991/1992) and Japan (member in
1992/1993), which have no significant estimates for the Scouncil variable,
weaken the idea that UN Security Council membership makes donor
countries more sensitive to appeals from developing countries.
188 Notes
86. For instance, at the Financing for Development Summit in Monterrey

(Mexico) in 2002, or at the G-8 Summit in Gleneagles (Scotland) in 2005,
where G-8 countries pledged to increase the overall aid to developing
countries by US$ 50 billion by 2010.
87. This chapter is based upon Czaika (2005a).
88. This chapter neglects legal policy issues of immigration management,
like the tightening of asylum laws, speeding up application and decision
procedures, or the introduction of the safe third country principle.
89. The theoretical paper of Lahiri & Raimondos-Møller (2000) is an excep-
tion. Their model illustrates the impact of ethnic groups (that is, perma-
nent immigrants by definition) on the political process in donor
countries through their lobbying activities, which influence the aid allo-
cation decision in favor of their country of origin.
90. Asylum seekers partially emigrate because of economic reasons (‘bogus
refugees’, see Neumayer, 2005b), however, most asylum applicants are
victims of war, oppression, natural disasters and so on, that is, they are
forced to leave and are therefore non-voluntary emigrants. As such,
we can recognize asylum seekers as worldwide transmitters of informa-
tion about urgent physical needs and human atrocities.
91. In several empirical studies, political factors are very significant for
both aid allocation decisions and aid effectiveness (see, for example,
Boone, 1996; Burnside & Dollar, 2000; Svensson, 1999; Collier & Dollar,
2002).
92. We assume all markets to be perfectly competitive and all countries to be
small, open economies with exogenously determined commodity prices.
Furthermore, all factors are inelastically supplied, which assures exoge-
nously given gross factor incomes.
93. Altruism, in this context, indicates the concern of immigrant house-
holds for their country of origin. As more natives are forced to leave the
home country to seek asylum in the new, common host country, the
immigrant households become increasingly concerned about their orig-
inal home country.
94. This altruism parameter may also contain self-interested motivations of
the immigrant groups in strengthening their status as a special immi-
grant group. However, this interpretation is rather secondary in this
study.
95. This parameter represents the political, economic and civil environment
in terms of income-enhancing factors, like a low level of corruption,
democracy, political freedom, civil rights and so on. Or, from another
angle, this parameter stands for the efficiency of a recipient country in
transferring aid to its households.
96. In the 1990s, Germany actually gave a positive amount of aid to all devel-
oping countries, and therefore the corner solution has no explanatory
relevance for the German aid allocation pattern. Other donors, like the
US or the Nordic countries, are more selective.
97. We do not take into account the fact that at least a small part of ODA
volumes are tied to imports from the donor country.
Notes 189
98. The OECD/DAC defines ODA according to three criteria: first, aid must
be given by official national or international (development) agencies;
second, aid must have the objective of promoting economic and social
development in the recipient country; and third, ODA must have a
minimum grant element of 25 percent (Cassen, 1994).
99. The DAC separates recipient countries into two parts. Part I countries
can be distinguished into least developed countries (LDCs), low income
countries (LICs), low-middle income countries (LMICs), upper-middle
income countries (UIMCs) and high income countries. Part II countries
are countries in transition (all Eastern European countries and coun-
tries of the former Soviet Union) and other advanced developing
countries. Bilateral and multilateral aid to Part I countries is called offi-
cial development assistance (ODA), whereas Part II countries receive
official aid (OA).
100. In the case of a small number of recipients (and therefore many zeros in
the dependent variable) a tobit or probit estimation model testing for
aid eligibility could be appropriate.
101. The data source for these two immigrant variables are the Migration
Information Source (http://www.migrationinformation.org/) and the
Federal Statistical Office of Germany (http://www.destatis.de/).
102. See Tables A.7 and A.8 for data description, sources, and descriptive
statistics of the sample.
103. Data from World Bank (2004).
104. Ibid.
105. The PQLI is computed from data of World Development Indicators
(World Bank, 2004) according to the formula provided by Neumayer
(2003, p. 51).
106. Data from Centre d’études prospectives et d’informations internation-
ales (CEPII, 2005).
107. Political rights refer to fairness in elections with a real possibility to
overtake power via elections, freedom for organizing parties, the exis-
tence of party competition and allowance for opposition, and so on.
Civil liberties reflect the freedom of assembly, freedom for any religious
activities, freedom of the media, protection from political suppression,
and so on (see FreedomHouse, 2005).
108. Data from OECD (2005).
109. Data from OECD (2004).
110. The calculated peak of transferred aid amounts is at a population level
of about 40 million people.
111. This chapter is based upon Czaika & Mayer (2008).
112. In 2002, Prime Ministers Tony Blair of the United Kingdom and Jose
María Aznar of Spain announced a plan to link aid provision to migrants
sending countries to their willingness to cooperate in containing illegal
immigration (see Bhagwati, 2003). Furthermore, French Foreign
Minister, Bernard Kouchner, on his first visit to Africa in June, 2007,
called for more development aid to help curb emigration (Reuters,
2007).
190 Notes
113. Many individuals who are not granted refugee status according to
Article 1 may still not be deterred as the Conventions’ Article 33 declares
that ‘no contracting state shall expel or return (“refouler”) a refugee in any
manner whatsoever to the frontiers of territories where his life or freedom
would be threatened on account of his race, religion, nationality, membership
of a particular social group or political opinion.’
114. As there tend to be multiple reasons for emigration, the distinction
between refugees and economic migrants becomes more and more del-
icate. A common but not unproblematic way to separate voluntary from
forced migrants refers to the original motivation of the migrant to leave
the place of origin. According to this, the underlying root cause of vol-
untary migration is economic, while either internal or cross-border ref-
ugee movements are rather caused by non-economic factors as
mentioned in the Geneva Convention (see above).
115. Czaika (2008b) models the influence of economic incentives and indi-
vidual persecution on the migration decision-making of refugees (see
Chapter 5).
116. See, for example, Dudley & Montmarquette (1976), McKinlay & Little
(1977, 1978), Maizels & Nissanke (1984), McGillivray (1989), Trumbull &
Wall (1994), Alesina & Dollar (2000), Neumayer (2003), Berthélemy &
Tichit (2004), or Berthélemy (2006).
117. An alternative argumentation, which states that aid towards first asy-
lum countries serves rather migration-prevention interests, while aid
for countries of origin is rather intended by burden-sharing motiva-
tions, is basically possible, but seems less plausible to the authors.
118. For a discussion on different approaches for tackling sample-selection
biases, see, for example, Heckman (1979), Manning et al. (1987),
Leung & Yu (1996) or Puhani (2000).
119. See, for example, Neumayer (2003) for a discussion of some caveats of
this approach in the context of aid allocation.
120. The coefficient of the inverse Mill’s ratio is insignificant, indicating
that selection bias does not seem to be a problem in our sample.
121. We tested the fixed-effect model against random-effects, but the
Hausman test rejected the latter.
122. Chang et al. (1999) discuss the many shortcomings of the official devel-
opment aid data from the OECD that include the underestimation of
the aid content, the over-representation of loans with high concession-
ality, as well as the constant interest rate of 10 percent used to calculate
the grant element of the highly concessional loans. They developed the
aid measure EDA (effective development assistance) to eliminate most
of the failures of ODA. However, as Ovaska (2003) proves that both
concepts yield essentially the same results, we use the standard ODA
measures.
123. Neumayer (2003) states that gross data are non-negative and conceptu-
ally closer to commitments than net ODA disbursements, but the prob-
lem with gross data is that parts of the amounts disbursed are not at the
Notes 191
country’s unrestricted disposal, as they are used to repay current

loans.
124. Table A.10 in the Appendix displays the correlation matrix for this set
of explanatory variables. It indicates that most of the cross-correlations
are not significant. Therefore, we should not have multicollinearity
problems between the different refugee variables.
125. Descriptive statistics and information on the composition and sources
of the dataset are provided in the Tables A.9 and A.11 of the Appendix.
126. Since the correlation between the two variables is negligible (Table A.10),
potential aid and trade simultaneity should not exert too much influ-
ence on the estimation results.
127. In former studies, including external debt was problematic due the poor
data availability. However, this has improved significantly, since the
Bank for International Settlements (BIS), the International Monetary
Fund (IMF), the Organization for Economic Co-operation and
Development (OECD) and the World Bank (WB) operate a new joint
database on external debt.
128. Morris (1979) develops the PQLI as an aggregate measure of infant mor-
tality, literacy and life expectancy. All three variables are transformed
to a scale from 0 (worst) to 100 (best), from which the average is taken.
We prefer the PQLI to the Human Development Index (HDI), since the
latter is available for fewer countries, and additionally, income per cap-
ita is not a component of the PQLI. Thus, collinearity between the two
variables is largely avoided.
129. See http://www.freedomhouse.org/ for an explanation of the index
methodology.
130. Results for the first stage logit regression are available on request.
131. The inverse Mill’s ratio is in both specifications not significant. This
indicates similarity of the OLS regression and the Heckman approach.
132. This result provides additional evidence to earlier results reported in
Chapter 8, where a similar result for German aid allocation pattern is
found.
133. At least when referring to the OLS and Heckman regressions.
134. This procedure refers to the methodology described and used by
Berthélemy (2006). He applies a similar exercise for the categorization
of donor countries with respect to the influence of trade volumes on
bilateral aid transfers.
135. This argument is based on the theory by Lahiri & Raimondos-Møller
(2000). They argue that political support is maximized by politicians
(or, the government) by considering the lobbying activities of ethnic
groups who may carry out propaganda or financial contributions in
exchange for preferred aid policies (see also Anwar & Michaelowa,
2006). Their theoretical framework refers to the approach of Grossman &
Helpman (1994).
Bibliography
Acharya, A., & Dewitt, D. (1997). Financial Burden-Sharing. In J. C. Hathaway

(ed.), Reconceiving International Refugee Law (pp. 111–144). Boston:
Kluwer.
AI (2004). New Military Operations, Old Patterns of Human Rights Abuses in
Aceh. Jakarta: Amnesty International. Retrieved from http://web.amnesty.
org/library/index/engasa210332004 (accessed on 3 January 2007).
Alesina, A., Devleeschauwer, A., Easterly, W., Kurlat, S., & Wacziarg, R. (2003).
Fractionalization. Journal of Economic Growth, 8(2), 155–194.
Alesina, A., & Dollar, D. (2000). Who Gives Foreign Aid to Whom and Why?
Journal of Economic Growth, 5(1), 33–63.
Alesina, A., & Weder, B. (2002). Do Corrupt Governments Receive Less
Foreign Aid? American Economic Review, 92(4), 1126–1137.
Angenendt, S. (2002). Das Weltflüchtlingsproblem und die Vereinten
Nationen. Aus Politik und Zeitgechichte, B27–B28.
Annett, A. (2000). Social fractionalization, Political Instability, and the Size
of Government. IMF Working Paper, No. 82, Washington, DC: IMF.
Anwar, M., & Michaelowa, K. (2006). The Political Economy of US Aid to
Pakistan. Review of Development Economics, 10(2), 195–209.
Apodaca, C., & Stohl, M. (1999). United States Human Rights Policy and
Foreign Assistance. International Studies Quarterly, 43(1), 185–198.
Arvin, B. M., & Drewes, T. (1998). Biases in the Allocation of Canadian
Official Development Assistance. Applied Economics Letters, 5(12), 773–775.
Arvin, B. M., & Drewes, T. (2001). Are There Biases in German Bilateral Aid
Allocations? Applied Economics Letters, 8(3), 173–177.
Bach, R., & Carroll-Seguin, R. (1986). Labour Force Participation Household
Composition and Sponsorship Among Southeast Asian Refugees.
International Migration Review, 20(2), 381–404.
Barrett, S. (1994). Self-Enforcing International Environmental Agreements.
Oxford Economic Papers, 46, 878–894.
Barrett, S. (1999). A Theory of Full International Cooperation. Journal of
Theoretical Politics, 11(4), 519–541.
Barrett, S. (2001). International Cooperation for Sale. European Economic
Review, 45(10), 1835–1850.
Barrett, S. (2002). Consensus Treaties. Journal of Institutional and Theoretical
Economics, 127(4), 529–547.
Barron, P., Kaiser, K., & Pradhan, M. (2004). Local conflict in Indonesia:
Measuring Incidence and Identifying Patterns. World Bank Policy Research
Working Paper, No. 3384.
Berthélemy, J. -C. (2006). Bilateral Donors’ Interest vs. Recipients’ Deve-
lopment Motives in Aid Allocation: Do All Donors Behave the Same? Review
of Development Economics, 10(2), 179–194.
192
Bibliography 193
Berthélemy, J. -C., & Tichit, A. (2004). Bilateral Donors’ Aid Allocation

Decisions – A Three-dimensional Panel Analysis. International Review of
Economics & Finance, 13(3), 253–274.
Betts, A. (2003). Public Goods Theory and the Provision of Refugee Protection:
The Role of the Joint Product Model in Burden-Sharing Theory. Journal of
Refugee Studies, 16(3), 274–295.
Betts, A. (2004). International Cooperation and Targeting Development
Assistance for Refugee Solutions: Lessons from the 1980s. New Issues in
Refugee Research, UNHCR Working Paper, No. 107. Geneva: UNHCR.
Betts, A. (2006). What Does Efficiency Mean in the Context of the Global
Refugee Regime? British Journal of Politics and International Relations, 8(2),
148–173.
Bhagwati, J. N. (2003). Borders Beyond Control. Foreign Affairs, 82(1),
98–104.
Bhagwati, J. N., Schatz, K. -W., & Wong, K. -y. (1984). The West German
Gastarbeiter System of Immigration. European Economic Review, 26(3),
277–294.
Black, R., & Koser, K. (eds) (1999). The End of the Refugee Cycle? Refugee
Repatriation and Reconstruction. New York: Berghahn Books.
BMZ (1994). Konzept Flüchtlingspolitik im Rahmen der Entwicklung-
szusammenarbeit, BMZ aktuell, No. 040. Bonn: Bundesministerium für
Wirtschaftliche Zusammenarbeit und Entwicklung.
BMZ (2006a). Die Zusammenarbeit des BMZ mit dem Amt des Hohen
Flüchtlingskommissars der Vereinten Nationen (UNHCR). BMZ spezial,
No. 138. Bonn: Bundesministerium für Wirtschaftliche Zusammenarbeit
und Entwicklung.
BMZ (2006b). Medienhandbuch: Entwicklungspolitik 2006/2007. Berlin: Bun-
desministerium für Wirtschaftliche Zusammenarbeit und Entwicklung.
Böhning, W., & Schloeter-Paredes, M. -L. (eds) (1994). Aid in Place of Migration?
Selected Contributions to an ILO-UNHCR Meeting. Geneva: International
Labour Office.
Boone, P. (1996). Politics and the Effectiveness of Foreign Aid. European
Economic Review, 40(2), 289–329.
Borjas, G. J. (1982). The Earnings of Male Hispanic Immigrants in the United
States. Industrial and Labour Relations Review, 35(3), 343–353.
Borjas, G. J. (1994). The Economics of Immigration. Journal of Economic
Literature, 32(4), 1667–1717.
Borjas, G. J. (1995). The Economic Benefits from Immigration. Journal of
Economic Perspectives, 9(2), 3–22.
Boswell, C. (2003). Burden-Sharing in the European Union: Lessons from the
German and UK Experience. Journal of Refugee Studies, 16(3), 316–335.
Boswell, C., & Crisp, J. (2004). Poverty, International Migration and Asylum.
UNU WIDER Policy Brief, No. 8. Helsinki: UNU WIDER.
Boyer, M. A. (1989). Burden-sharing and Comparative Advantage in the
Western Alliance System. Journal of Conflict Resolution, 33(4), 700–727.
BPS (n.d.). Population Growth by Province. Retrieved from www.bps.go.id/
(accessed on 12 January 2007).
194 Bibliography
BPS (2000). Potensi Desa Sensus Ekonomi. Jakarta: BPS Statistics Indonesia.
BPS (2003). Potensi Desa Sensus Ekonomi. Jakarta: BPS Statistics Indonesia.
Brücker, H., & Schröder, P. J. H. (2006). International Migration with
Heterogeneous Agents: Theory and Evidence. DIW Discussion Paper,
No. 2049. Berlin: Deutsches Institut für Wirtschaftsforschung.
Burgdorff, T. (2002). Der Beitrag der deutschen Entwicklungszusammenarbeit zur
Entschärfung der Flüchtlings- und Migrationsproblematik: Die Fallstudie
Guatemala. Dissertation, Universität Bonn.
Burnside, C., & Dollar, D. (2000). Aid, Policies, and Growth. American
Byrne, R. (2003). Harmonization and Burden Redistribution in the Two
Europes. Journal of Refugee Studies, 16(3), 336–358.
Carraro, C., & Siniscalco, D. (1993). Strategies for the International Protection
of the Environment. Journal of Public Economics, 52(3), 309–328.
Carrington, W. J., Detragiache, E., & Vishwanath, T. (1996). Migration with
Endogenous Moving Costs. American Economic Review, 86(4), 909–930.
Cassen, R. (1994). Does Aid Work? Oxford: Oxford University Press.
Castles, S. (2003). Why Migration Policies Fail. Ethnic and Racial Studies, 27(2),
205–227.
Castles, S., & Miller, M. J. (2003). The Age of Migration: International Population
Movements in the Modern World. New York: Palgrave Macmillan.
CEPII (2005). Geographical Distances Dataset. Paris: Centre d’Etudes Pro-
spectives et d’Informations Internationales. Retrieved from www.cepii.fr/
anglaisgraph/bdd/distances.htm (accessed on March 2005).
Chang, C. C., Fernandez-Arias, E., & Serven, L. (1999). Measuring Aid Flows:
A New Approach. World Bank Policy Research Paper, No. 2050.
Chauvet, L. (2003). Socio-political Instability and the Allocation of Inter-
national Aid by Donors. European Journal of Political Economy, 19(1), 33–59.
Chiswick, B. R., & Hatton, T. J. (2002). International Migration and the Inte-
gration of Labor Markets. In M. D. Bordo, A. M. Taylor & J. G. Williamson
(eds), Globalization in Historical Perspective (pp. 65–117). Chicago: University
of Chicago Press.
CIA (2004). The CIA World Factbook. Washington, DC: CIA Press.
Collier, P. (2000). Economic Causes of Civil Conflict and Their Implications for
Policy. Washington, DC: World Bank.
Collier, P. (2007). The Bottom Billion: Why the Poorest Countries are Failing and
What Can Be Done About It. Oxford: Oxford University Press.
Collier, P., & Dollar, D. (2002). Aid Allocation and Poverty Reduction.
European Economic Review, 46(8), 1475–1500.
Collier, P., & Hoeffler, A. (2002). Aid, Policy and Peace: Reducing the Risks of
Civil Conflict. Defence and Peace Economics, 13(6), 435–450.
Collier, P., & Hoeffler, A. (2004). Aid, Policy and Growth in Post-conflict
Societies. European Economic Review, 48(5), 1125–1145.
Collier, P., Elliott, V., Hegre, H., Hoeffler, A., Reynal-Querol, M., &
Sambanis, N. (2003). Breaking the Conflict Trap: Civil War and Development
Policy. Washington, DC: World Bank and Oxford University Press.
Bibliography 195
CRED (2006). EM-DAT – Emergency Event Database. Université de Louvain:

Centre for Research on the Epidemiology of Disasters. Retrieved from
www.emdat.be/Database/ (accessed on May 2006).
Crisp, J. (1999). Who has Counted the Refugees? UNHCR and the Politics of
Numbers. New Issues in Refugee Research, UNHCR Working Paper, No. 12.
Geneva: UNHCR
Crisp, J. (2001). Mind the Gap! UNHCR, Humanitarian Assistance and the
Development Process. International Migration Review, 35(1), 168–191.
Crisp, J. (2005). A New Asylum Paradigm? Globalization, Migration and the
Uncertain Future of the International Refugee Regime. STAIR, 1(1), 39–53.
Czaika, M. (2005a). Aid Allocation and Asylum Migration in the 1990s.
Applied Economics Quarterly, 51(3), 289–303.
Czaika, M. (2005b). A Refugee Burden Index: Methodology and its Applica-
tion. Migration Letters, 2(2), 101–125.
Czaika, M. (2008a). Cheap Talk in the UN Arenas: Some Evidence on the
Impact of UN Speeches on Aid Allocation Decisions. Applied Economics
Letters, 15(3), 187–191.
Czaika, M. (2008b). The Political Economy of Refugee Migration. IEP
Discussion Paper, No. 7, University of Freiburg.
Czaika, M. (2009). Asylum Cooperation among Asymmetric Countries – the
case of the European Union. European Union Politics, 10(1), 89–113.
Czaika, M. & Kis-Katos, K. (2009). Civil Conflict and Displacement – Village-
level Determinants of Forced Migration in Aceh. Journal of Peace Research,
46(3), 399–418.
Czaika, M., & Mayer, A. (2008). Refugee Movements and Aid Responsiveness
of Bilateral Donors. IEP Discussion Paper, No. 6, University of Freiburg.
de Haas, H. (2006). Turning the Tide? Why ‘Development Instead of Migration
Policies’ are Bound to Fail. IMI Working Paper, No. 2. University Oxford:
International Migration Institute.
de Mello, S. Vieira (1995). Toward Cooperation Between Humanitarian Aid
and Military Forces. In G. A. Joulwan & R. Weissinger-Baylon (eds),
Stability, Democracy & Peace Through Cooperation. The New NATO. Brussels:
NATO.
Degnbol-Martinussen, J., & Engberg-Pedersen, P. (2003). AID: Understanding
International Development Cooperation. London, UK: Zed Books Ltd.
Djajic, S., & Milbourne, R. (1988). A General Equilibrium Model of Guest-
worker Migration: The Source-country Perspective. Journal of International
Economics, 25(3–4), 335–351.
Dowling, J. M., & Hiemenz, U. (1985). Biases in the Allocation of Foreign Aid:
Some New Evidence. World Development, 13(4), 535–541.
Dudley, L., & Montmarquette, C. (1976). A Model of the Supply of Bilateral
Foreign Aid. American Economic Review, 66(1), 132–142.
Dustmann, C. (2003). Return Migration, Wage Differentials, and the Optimal
Migration duration. European Economic Review, 47(2), 353–369.
EC (2007). Europeans and Development Aid, Special Eurobarometer No. 280/
Wave 67.1 TNS Opinion & Social. Brussels: European Commission.
196 Bibliography
Engel, S., & Ibánez, A. María. (2007). Displacement Due to Violence in

Colombia: A Household-Level Analysis. Economic Development and Cultural
Change, 55(2), 335–365.
Esman, M., & Herring, R. (2001). Carrots, Sticks, and Ethnic Conflict: Rethinking
Development Assistance. Ann Arbor, MI: University of Michigan Press.
Facchini, G., Lorz, O., & Willmann, G. (2006). Asylum Seekers in Europe:
The Warm Glow of a Hot Potato. Journal of Population Economics, 19(2),
411–430.
Faini, R. (1996). Increasing Returns, Migrations and Convergence. Journal of
Development Economics, 49(1), 121–136.
Fischer, P. A., Malmberg, G., Holm, E., & Straubhaar, T. (1998). Why do People
Stay? The Insider Advantages Approach: Empirical Evidence from Swedish
Labor Markets. CEPR Working Paper, No. 1952.
FreedomHouse (2005). Freedom of the World: Database, Washington, DC.
Retrieved from www.freedomhouse.org/ (accessed on April 2005).
FreedomHouse (2006). Freedom of the World: Database, Washington, DC.
Retrieved from www.freedomhouse.org/ (accessed on May 2005).
Ghatak, S., Levine, P., & Price, S. W. (1996). Migration Theories and Evidence:
An Assessment. Journal of Economic Surveys, 10(2), 159–198.
Gosh, B. (1992). Migration-development Linkages: Some Specific Issues and
Practical Policy Measures. International Migration, 30(3–4), 423–456.
Grossman, G. M., & Helpman, E. (1994). Protection for Sale. American
Harris, J. R., & Todaro, M. (1970). Migration, Unemployment and Develop-
ment: A Two-Sector Analysis. American Economic Review, 60(1), 126–142.
Harstad, B. (2007). Harmonization and Side Payments in Political Coopera-
tion. American Economic Review, 97(3), 871–889.
Hathaway, J. C. (1990). A Reconsideration of the Underlying Premise of
Refugee Law. Harvard International Law Journal, 31(1), 129–147.
Hatton, T. J. (2004). Seeking Asylum in Europe. Economic Policy, 19(38), 5–62.
Hatton, T. J. (2005). European Asylum Policy. National Institute Economic
Review, 194(1), 106–119.
Hatton, T. J., & Williamson, J. G. (2002). What Fundamentals Drive World
Migration? NBER Working Paper, No. 9159.
Hatton, T. J., & Williamson, J. G. (2005). Refugees, Asylum Seekers and Policy
in Europe. CEPR Working Paper, No. 5058.
Hauff, E., & P. Vaglum. (1993). Integration of Vietnamese Refugees into
Norwegian Labor Market: The Impact of War Trauma. International
Migration Review, 27(2), 388–405.
Heckman, J. J. (1979). Sample Selection Bias as a Specification Error.
Econometrica, 47(1), 153–161.
Hedman, E. -L. E. (2005). A State of Emergency, A Strategy of War: Internal
Displacement, Forced Relocation, and Involuntary Return in Aceh. In
E. -L. E. Hedman (ed.), Aceh under Martial Law: Conflict, Violence and
Displacement (pp. 7–11). Oxford: Refugee Studies Centre.
Hjertholm, P., & White, H. (2000). Foreign Aid in Historical Perspective:
Background and Trends. In F. Tarp & P. Hjertholm (eds), Foreign Aid and
Bibliography 197
Development: Lessons Learnt and Directions for the Future (pp. 80–102).
London, UK: Routledge.
Holzer, T., & Schneider, G. (2002). Asylpolitik auf Abwegen: Nationalstaatliche
und europäische Reaktionen auf die Globalisierung der Flüchtlingsströme.
Opladen: Leske + Budrich.
Hovy, B. (2001). Statistically Correct Asylum Data: Prospects and Limitations.
New Issues in Refugee Research, UNHCR Working Paper, No. 12. Geneva:
UNHCR.
HRW (2005). Aceh’s Forgotten Victims. Jakarta: Human Rights Watch.
ICG (2002). A slim Chance for Peace. Brussels: International Crisis Group.
IDMC (2006a). Internal Displacement, Global Review on Trends and Developments
in 2005. Geneva: Internal Displacement Monitoring Centre.
IDMC (2006b). Support Needed for Return and Re-integration of Displaced
Acehnese Following Peace Agreement. Geneva: Internal Displacement Moni-
toring Center. Retrieved from www.internal-displacement.org/ (accessed
on 3 January 2007).
ILO (2004). A Fair Globalization: Creating Opportunities For All. Geneva:
International Labor Organization.
Isenman, P. (1976). Biases in Aid Allocations against Poorer and Larger
Countries. World Development, 4(8), 631–641.
Jackson, M. O., & Wilkie, S. (2005). Endogenous Games and Mechanisms:
Side Payments among Players. Review of Economic Studies, 72(2), 543–566.
Jacobsen, K. (1997). Refugees’ Environmental Impact: The Effect of Patterns
of Settlement. Journal of Refugee Studies, 10(1), 19–36.
Jandl, M. (1995). Structure and Costs of the Asylum Systems in Seven European
Countries. Vienna: International Centre for Migration Policy Development.
JEDH (2007). Joint External Debt Hub of BIS, IMF, OECD and World
Bank. Washington, DC: World Bank. Retrieved from www.jedh.org/
(accessed on January 2007).
Kaufmann, D., Kraay, A., & Mastruzzi, M. (2003). Governance Matters III:
Governance Indicators for 1996–2002. World Bank Policy Research Paper,
No. 3106. Washington, DC: The World Bank.
Kibreab, G. (1997). Environmental Causes and Impact of Refugee Movements:
A Critique of the Current Debate. Disasters, 21(1), 20–38.
Klingebiel, S. (1994). Entwicklungszusammenarbeit und die Flüchtlings- und
Migrationsproblematik. Aus Politik und Zeitgeschichte, Das Parlament,
B 20/94, 18–25.
Lahiri, S., & Raimondos-Møller, P. (2000). Lobbying by Ethnic Groups and
Aid Allocation. Economic Journal, 110(462), C62–C79.
Leung, S. Fai, & Yu, S. (1996). On the Choice between Sample Selection and
Two-Part Models. Journal of Econometrics, 72(1–2), 197–229.
Lloyd, T., Morrissey, O., & Osei, R. (2001). Problems with Pooling in Panel
Data Analysis for Developing Countries: The Case of Aid and Trade
Relationships. CREDIT Research Paper, No. 01/14. Nottingham, UK: Center
for Research in Economic Development and International Trade.
Loescher, G. (1989). The European Community and Refugees. International
Affairs, 65(4), 617–636.
198 Bibliography
Loescher, G. (2002). The UNHCR and World Politics: A Perilous Path. Oxford:
Oxford University Press.
Loescher, G., & Milner, J. (2005). Protracted Refugee Situations: Domestic and
International Security Implications (Adelphi Papers). London: Routledge.
Maizels, A., & Nissanke, M. (1984). Motivations for Aid to Developing
Countries. World Development, 12(8), 879–900.
Manning, W. G., Duan, N., & Rogers, W. H. (1987). Monte Carlo Evidence on
the Choice between Sample Selection and Two-Part Models. Journal of
Econometrics, 35(1), 59–82.
Martin, S., Schoenholtz, A. J., & Fisher, D. (2005). The Impact of Asylum on
Receiving Countries. In G. J. Borjas & Jeff Crisp (eds), Poverty, International
Migration and Asylum (pp. 99–120). New York: Palgrave Macmillan.
Massey, D. S., Arango, J., Hugo, G., Kouaouci, A., Pellegrino, A., & Taylor, J.
Edward. (1998). World in Motion: Understanding International Migration at the
End of the Millennium. New York: Oxford University Press.
McGillivray, M. (1989). The Allocation of Aid Among Developing Countries:
A Multi-Donor Analysis Using a Per Capita Aid Index. World Development,
17(4), 561–568.
McGillivray, M. (2003a). Aid Effectiveness and Selectivity: Integrating
Multiple Objectives into Aid Allocation. DAC Journal, 4(3), 23–36.
McGillivray, M. (2003b). Modelling Foreign Aid Allocation: Issues, Approaches
and Results. Journal of Economic Development, 28(1), 171–188.
McKinlay, R. D., & Little, R. (1977). A Foreign Policy Model of US Bilateral Aid
Allocation. World Politics, 30(1), 58–86.
McKinlay, R. D., & Little, R. (1978). A Foreign-Policy Model of Distribution of
British Bilateral Aid, 1960–70. British Journal of Political Science, 8(3),
313–332.
Mertus, J. (1998). The State and the Post-Cold War Refugee Regime: New
Models, New Questions. International Journal of Refugee Law, 10(3), 321–348.
Moon, B. (1991). The Political Economy of Basic Humanitarian Needs. Ithaca:
Cornell University Press.
Moore, W. H., & Shellman, S. M. (2004). Fear of Persecution: Forced Migration,
1952–1995. Journal of Conflict Resolution, 40(5), 723–745.
Morris, D. (1979). Measuring the Condition of the World’s Poor – The Physical
Quality of Life Index. New York: Pergamon Press.
MPI (2005). Migration Information Source. Washington, DC: Migration Policy
Institute. Retrieved from www.migrationinformation.org/ (accessed on
February 2005).
Münz, R., & Weiner, M. (eds) (1997). Migrants, Refugees, and Foreign Policy: U.S.
and German Policies toward Countries of Origin. Oxford: Berghahn Books.
Myers, N. (2002). Environmental Refugees: A Growing Phenomenon of the
21st Century. Philosophical Transactions of Biological Sciences, 357(1420),
609–613.
Neumayer, E. (2003). The Pattern of Aid Giving: The Impact of Good Governance
on Development Assistance. New York: Routledge.
Neumayer, E. (2004). Asylum Destination Choice: What Makes Some
European Countries More Attractive Than Others? European Union Politics,
5(2), 155–180.
Bibliography 199
Neumayer, E. (2005a). Asylum Recognition Rates in Western Europe – Their

Determinants, Variation and Lack of Convergence. Journal of Conflict
Resolution, 49(1), 43–66.
Neumayer, E. (2005b). Bogus Refugees? The Determinants of Asylum
Migration to Western Europe. International Studies Quarterly, 49(4),
389–409.
Noll, G. (1997). Prisoners’ Dilemma in Fortress Europe. On the Prospects of
Burden Sharing in the European Union. German Yearbook of International
Law, 40, 405–437.
Noll, G. (2003). Risky Games? A Theoretical Approach to Burden Sharing in
the Asylum Field. Journal of Refugee Studies, 16(3), 236–252.
Nunnenkamp, P., & Thiele, R. (2006). Targeting Aid to the Needy and
Deserving: Nothing but Promises? World Economy, 29(9), 1177–1201.
Nyberg-Sørensen, N., van Hear, N., & Engberg-Pedersen, P. (2003). Migration,
Development and Conflict: State of the Art Overview. In N. Nyberg
Sørensen & N. van Hear (eds), The Migration-Development Nexus (pp. 5–50).
Geneva: International Organisation for Migration.
OECD (2004). International Development Statistics: Aid statistics. Paris: OECD/
DAC.
OECD (2005). Bilateral Trade Database. Paris: OECD.
OECD (2006a). International Development Statistics. Paris: OECD.
OECD (2006b). International Migration Outlook: SOPEMI. Paris: OECD.
OECD (2007a). International Development Statistics. Paris: OECD.
OECD (2007b). Is it ODA? Factsheet. Paris: OECD/DAC.
Olson, M. (1965). The Logic of Collective Action. Cambridge, MA: Harvard
University Press.
Ovaska, T. (2003). The Failure of Development Aid. Cato Journal, 23(2),
175–188.
Perdana, A. A., & Maxwell, J. (2004). Poverty Targeting in Indonesia: Pro-
grams, Problems and Lessons Learned. CSIS Economics Working Paper,
No. 083. Jakarta, Indonesia: Centre for Strategic and International Studies.
PRIO (2005). Armed Conflict Database. Oslo: Peace Research Institute.
Puhani, P. A. (2000). The Heckman Correction for Sample Selection and its
Critique. Journal of Economic Surveys, 14(1), 53–68.
Ramly, A. Aulia. (2005). Modes of Displacement during Martial Law. In
E. -L. E. Hedman (ed.), Aceh under Martial Law: Conflict, Violence and
Displacement (pp. 13–20). Oxford: Refugee Studies Centre.
Reuters (2007). Kouchner says aid can curb migration on Africa tour. Retrieved
from www.reuters.com/article/latestCrisis/idUSL08677457 (accessed on
19 March 2008).
Rogge, J. R. (1994). Repatriation of Refugees: Not So Simple ‘Optimum’
Solution. In T. Allen & H. Morsink (eds), When Refugees Go Home: African
Experiences (pp. 14–49). Trenton, NJ: Africa World Press.
Rotte, R., Vogler, M., & Zimmermann, K. F. (1997). South-North Refugee
Migration: Lessons for Development Cooperation. Review of Development
Economics, 1(1), 99–115.
Sandler, T. (1992). Collective Action: Theory and Applications. Ann Arbor, MI:
University of Michigan Press.
200 Bibliography
Sassen, S. (1998). Beyond Sovereignty: Immigration Policy Making Today. In

S. Jonas & S. D. Thomas (eds), Immigration: A Civil Rights Issue for the Americas
(pp. 15–26). Wilmington, DE: Scholarly Resources Books.
Schmeidl, S. (1997). Exploring the Causes of Forced Migration: A Pooled
Time-series Analysis, 1971–1990. Social Science Quarterly, 78(2), 284–208.
Schmeidl, S. (2001). Conflict and Forced Migration: A Quantitative Review.
In A. R. Zolberg & P. M. Benda (eds), Global Migrants, Global Refugees: Pro-
blems and Solutions (pp. 62–94). New York and Oxford: Berghahn Books.
Schuck, P. H. (1997). Refugee Burden-Sharing: A Modest Proposal. Yale Journal
of International Law, 22, 243–297.
Schulze, K. E. (2004). The Free Aceh Movement (GAM): Anatomy of a
Separatist Organization. Policy Studies, No. 2. Washington, DC: East-West
Center.
Sidel, J. T. (1999). Indonesia Update: Trends toward Consolidation, Threats of
Disintegration. WriteNet, No. 18. Retrieved from http://web.archive.org/
(accessed on January 2007).
Sjaastadt, L. (1962). The Costs and Returns of Human Migration. Journal of
Political Economy, 70(supplement), 80–93.
Smith, J. P., & Edmonston, B. (1997). The New Americans: Economic Demographic
and Fiscal Effects of Immigration. Washington, DC: National Academy
Press.
Stark, O. (1991). The Migration of Labor. Oxford: Blackwell.
Stein, B. N. (1986). Durable Solutions for Developing Country Refugees.
International Migration Review, 20(2), 264–282.
Straubhaar, T. (2000). International Migration: Gehen oder Bleiben: Wieso
gehen wenige und bleiben die meisten? HWWA Discussion Paper, No. 111.
Hamburg: HWWA.
Suhrke, A. (1998). Burden-Sharing during Refugee Emergencies: The Logic
of Collective versus National Action. Journal of Refugee Studies, 11(4),
396–415.
Svensson, J. (1999). Aid, Growth and Democracy. Economics and Politics, 11(3),
275–297.
Svensson, J. (2000). Foreign Aid and Rent-seeking. Journal of International
Economics, 51(2), 437–461.
Thielemann, E. R. (2003a). Between Interests and Norms: Explaining Burden-
Sharing in the European Union. Journal of Refugee Studies, 16(3), 253–273.
Thielemann, E. R. (2003b). Editorial Introduction. Journal of Refugee Studies,
16(3), 225–233.
Thielemann, E. R. (2004). Why Asylum Policy Harmonization Undermines
Refugee Burden-sharing. European Journal of Migration and Law, 6(1),
47–65.
Thielemann, E. R. (2005). Symbolic Politics or Effective Burden-Sharing?
Redistribution, Side-payments and the European Refugee Fund. Journal of
Common Market Studies, 43(4), 807–824.
Thielemann, E. R. (2006). Burden-sharing: The International Politics of
Refugee Protection. CCIS Working Paper, No. 134. San Diego: Center for
Comparative Immigration Studies, University of California.
Bibliography 201
Thielemann, E. R., & Dewan, T. (2006). The Myth of Free-Riding: Refugee

Protection and Implicit Burden-sharing. West European Politics, 29(2),
351–369.
Todaro, M. P. (1969). A Model for Labor Migration and Urban Unemployment
in Less Developed Countries. American Economic Review, 59(1), 138–148.
Todaro, M. P., & Smith, S. C. (2006). Economic Development: Harlow: Addison
Wesley.
Trumbull, W. N., & Wall, H. J. (1994). Estimating Aid-Allocation Criteria with
Panel Data. Economic Journal, 104(425), 876–882.
UNDP (2004). Human Development Report: Cultural Liberty in Today’s Diverse
World. New York: UNDP.
UNDP (2006). Consultant, IDP Registration, Survey and Support to Local
Authorities in Central Aceh. New York: UNDP.
UNFPA (2006). State of the World Population: A Passage to Hope. New York:
UNFPA.
UNHCR (2000). The State of The World’s Refugees: Fifty Years of Humanitarian
Action. Oxford: Oxford University Press.
UNHCR (2001). Refugee Protection: A Guide to International Refugee Law.
Geneva: UNHCR.
UNHCR (2002). Selected Indicators Measuring Capacity and Contributions of
Host Countries. Geneva: UNHCR.
UNHCR (2003). Trends in Displacement, Protection and Solution: Statistical
Yearbook. Geneva: UNHCR.
UNHCR (2004a). Convention Plus: Issues Paper on Targeting of Development
Assistance, Annex II. Geneva: UNHCR.
UNHCR (2004b). Global Refugee Trends. Geneva: UNHCR.
UNHCR (2004c). Protracted Refugee Situations: UN Document EC/54/SC/CRP.14.
Geneva: UNHCR.
UNHCR (2004d). Trends in Displacement, Protection and Solution: Statistical
UNHCR (2005). Statistical Yearbook. Geneva: UNHCR.
UNHCR (2006a). Asylum Levels and Trends in Industrialized Countries, 2005.
Geneva: UNHCR.
UNHCR (2006b). State of the World’s Refugees: Human Displacement in the New
Millennium. Oxford: Oxford University Press.
UNHCR (2006c). Trends in Displacement, Protection and Solution: Statistical
UNHCR (2007). Targeting Development Assistance, Including International
Cooperation for Finding Durable Solutions for Protracted Refugee
Situations. Refugee Studies Quarterly, 26(1), 177–182.
United Nations (2005a). 2004 World Survey on the Role of Women in Develop-
ment: Women and International Migration. New York: United Nations
Publications.
United Nations (2005b). Bibliographic Information System: Database. New York:
United Nations.
United Nations (2006a). International Migration 2006. New York: UN
Populations Divisions.
202 Bibliography
United Nations (2006b). Population Database: World Migrant Stock. Retrieved

from http://esa.un.org/migration/index.asp?panel=1 (accessed on 3 May
2007).
UNOCHA (2003). Mission Report North Sumatra Province. Jakarta: UNOCHA.
Vink, M., & Meijerink, F. (2003). Asylum Applications and Recognition
Rates in EU Member States 1982–2001: A Quantitative Analysis. Journal of
Refugee Studies, 16(3), 297–315.
Weiner, M. (1993). Security, Stability, and International Migration. Inter-
national Security, 17(3), 91–126.
Widgren, J. (1989). Asylum Seekers in Europe in the Context of South-North
Movements. International Migration Review, 23(3), 599–605.
Wooden, M. (1991). The Experience of Refugees in the Australian Labor
Market. International Migration Review, 25(3), 514–535.
World Bank (1998). Assessing Aid – What Works, What Doesn’t and Why.
Oxford: Oxford University Press.
World Bank (2004). World Development Indicators (CD-ROM). Washington,
DC: World Bank.
World Bank (2006). World Development Indicators (CD-ROM). Washington,
DC: World Bank.
WVS (2000). World Values Survey: Database. Retrieved from www.
worldvaluessurvey.org/ (accessed on November 2006).
Zolberg, A. R., Suhrke, A., & Aquayo, S. (1989). Escape from Violence: Conflict
and the Refugee Crisis in the Developing World. Oxford: Oxford University
Press.
Index
aid allocation seeker, 1–4, 8–9, 11–13, 18–22,

bilateral, 119, 124, 126, 130, 131–5, 25–32, 35–42, 74–7, 79–81, 83,
138, 143–4, 146, 150, 151, 162 88–91, 95, 105, 124, 131–4,
decision, 114, 126–30, 132–5, 135–44, 146–63, 164–6, 181,
137–8, 143–4, 146–7, 151–2, 162, 185–8
165, 188 standard, 5, 22
determinant, 7, 124, 134, 147, system, 5, 27, 89, 166, 169, 181
151, 157 asylum migration, 1, 4, 7, 8, 14, 19,
foreign, 5, 83, 114, 117, 138, 146, 24, 74, 77, 78, 81, 82, 84, 86,
152, 158, 161, 168 132, 137, 143, 149, 185
multilateral, 124, 126, 127, 129 flow, 82, 86, 88, 167
pattern, 118, 124, 126, 131–4, 138, movement, 5, 8, 90, 168
143, 144–5, 163, 188, 191
aid responsiveness, 146, 158, 159, benefit
160, 161 of immigration, 4
altruism, 29, 95, 105, 116, 124, net, 51, 52, 93, 107, 183
135–7, 151, 154, 161–2, 174, 188 spillover, 21, 101, 103, 104
asylum strategic, 21, 89, 103, 106
applicant, 8, 9, 131, 143, 144, 154, burden-sharing
156, 161, 166, 188 financial/monetary, 90, 94, 98,
application, 69, 89, 91, 105–6, 102, 186
132, 140, 148, 174, 177, 180 regime, 5, 20, 22, 88, 89, 90, 103,
country, 4, 67, 69, 70–87, 116, 151, 104, 112
158, 165, 185, 186 rule, 34, 89, 92, 105, 106, 107,
crisis, 115, 147 112, 186
law, 94, 131, 186, 188
policy, 1, 4, 5, 19, 20, 21, 74–7, conflict
78–80, 81–2, 86–7, 91, 93, 98, armed, 6, 15, 19, 44–7, 125, 155,
106, 164, 167 159, 176, 180
pressure, 7, 67, 83, 85, 92, 115, 165 civil, 13, 43, 45, 50, 53, 65, 66, 164
proactive asylum policy, 81–7 internal, 125, 182
procedure, 4, 185, 186 risk, 23, 150
provision, 7, 21, 22, 34, 89, 90, 99, victim, 47, 53, 69, 159, 171, 180, 188
105, 112 violent, 2, 3, 8, 13, 15, 23, 51,
recognition rate, 74, 77, 80, 104, 159, 161
105, 174, 185 cooperation
regime, 7, 22, 89, 90, 94, 95, 97, asylum, 7, 88, 89, 90, 94, 95–9,
102, 104, 106, 112, 165 102, 103, 104, 107, 112, 165
restrictive asylum policy, 4, 5, 7, 9, development, 23, 118, 124, 129,
67, 80, 81, 87, 92, 164, 166 140, 144
203
204 Index
cooperation – continued double majority, 88, 90, 106,

financial, 7, 165, 186 107, 113
history, 126, 138, 144 enlargement, 90, 95, 122
international, 11, 20–2, 91, 168 European Refugee Fund, 90, 94,
technical, 123 113, 186
cost member states, 89, 90, 91, 92, 104,
asylum, 28, 86, 112–13, 165 106, 107, 112, 113, 117, 185
economic, 4, 89, 95, 149 population, 90, 106, 107, 109,
heterogeneity, 6, 90, 97, 98, 99, 111, 174
103, 104, 112 reform treaty, 7, 88, 90, 106,
migration, 51, 52, 64, 70, 74, 76, 107, 113
77, 79, 80, 185, 186 unanimity, 88, 89, 90, 106, 107,
monetary, 6, 22, 181 113, 187
opportunity, 76, 99
parameter, 96, 97, 98, 101, 107 good governance, 115, 117, 124, 125,
-sharing, 103, 104, 165 129, 134, 154, 155, 158, 176
transaction, 93 institution, 23, 30, 32
country rewarding, 125, 138, 143, 155
destination, 14, 15, 52, 91
first asylum, 67, 69, 70, 71, 72, human rights, 18, 19, 30, 91, 115,
77–87, 116, 151, 158, 185 124, 134, 167
host, 6, 12, 15, 21, 25, 26, 28–31, violation, 3, 10, 15, 44, 47, 48, 87,
35, 130, 132, 134, 149, 183, 188 117, 149, 159, 182
of origin, 8, 14, 23, 68–77, 81–7,
92, 116, 130–5, 143–54, 162, immigration
165, 182, 185, 188 group, 28–9, 134, 143, 144, 188
Western asylum, 74, 75–7, 80–5, household, 135–7, 188
165, 185, 186 irregular, 20, 166, 181
permanent, 131, 135, 137, 144, 188
development income
assistance, 8, 23, 115–16, 123, differential, 68, 72, 76, 77, 149
125, 129, 138, 147, 150, 155–8, middle-income bias, 124, 129
161, 165–7, 187, 190 insecurity, 2, 3, 11, 51, 67, 69
economic, 18, 23, 92, 122, 150, integration
154, 187 local, 42, 81, 87, 116, 158, 167
human, 125, 150, 158, 167 reintegration, 16, 83, 87, 116,
social, 5, 189 130, 169
donor internally displaced person, 3, 8, 11,
country, 8–9, 86, 122, 131–4, 13, 15, 16, 17, 36, 44, 48, 146,
135–7, 139, 143–5, 154, 162, 148, 151, 154, 165, 180
165–6
government, 24, 115, 118, 133–6, labor market, 4, 28, 95, 181
144–5, 147, 149–50, 158, 162
self-interest, 124, 133, 150, 154 migration
flow, 5, 10, 44, 56, 59, 63, 66, 116,
European Union 147, 150, 166–8
asylum policy, 89, 104, 132 forced, 1, 2, 6, 8, 17, 44, 50, 56, 59,
Council, 90, 106, 113, 187 65, 147, 164, 165, 169
Index 205
migration – continued need, 124, 125, 129, 133, 143, 150,

management, 24, 162, 188 154, 155
prevention, 5, 20, 22, 23, 117, 151, refugee
152, 154, 155, 158, 162, 163, 190 environmental, 16, 182
policy, 5, 21, 169, 174 movement, 1, 5, 6–8, 17, 23, 44,
return, 68 67–9, 81, 114–16, 130, 146–51,
unwanted, 116, 117, 166 158, 162–5, 168
protection, 1, 7, 16–22, 89, 90, 91,
network 94–5, 100, 101, 112, 167, 186
social, 2, 5, 14, 15, 22, 51, 52, 140, repatriation, 16, 18, 116, 122, 162,
166, 168 166, 169, 186
voluntary, 42, 83, 87, 130, 152, 162
Official Development Assistance,
118, 119–25, 138, 139, 148, 180, self-reliance, 18, 72, 78–9, 85,
187, 189 130, 167
volume, 119–23, 125, 136, 143, 188 economic, 70, 79, 81–5, 185
spillover
package deal, 99, 107, 187 asymmetry, 88, 90, 97, 98,
persecution, 2, 3, 7, 10, 13, 15, 51, 103, 112
67, 68–78, 87, 167, 169, 182, 190 parameter, 93, 97, 98, 100, 101,
fear of, 13, 51, 182 102, 104, 105
population
conflict-affected, 17, 78, 84 transfers
foreign-born, 2, 106, 174, 177 aid, 7–8, 23–4, 67, 81–6, 120,
poor, 1, 2, 14, 23, 56, 57, 64, 117, 153, 165
137, 158, 167, 171, 184 financial, 5, 7, 42, 88, 93, 94, 99,
refugee, 4, 6, 26, 28, 30–40, 42, 101, 107, 122, 165
44, 80–3, 147, 149, 151, 156, monetary, 90, 93, 95, 98–103, 112
164, 183
post-conflict welfare
reconstruction, 83, 116, 130 analysis, 94, 186
situation, 23, 124, 168 collective, 90, 96, 97, 99, 101, 102
criteria, 56, 184, 187
recipient function, 96, 100, 105, 107,
country, 8, 82, 115, 117, 124–7, 136, 137
132–45, 146–50, 152–5, 158–61, maximization, 96, 97
180, 188, 189 system, 4, 22

Thepoliticaleconomyofrefugeemigrationandforeignaid PDF

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Thepoliticaleconomyofrefugeemigrationandforeignaid PDF

Uploaded by

Copyright:

Available Formats

The Political Economy of Refugee

Migration and Foreign Aid

2 Forced Migration Patterns and Politics 10

3 International Refugee Burdens 25

4 Determinants of Forced Displacement:

5 The Political Economy of Refugee Migration 67

6 Asylum Cooperation among Asymmetric Countries 88

7 Refugee Movements and Aid Politics 114

9 Refugee Movements and Aid Responsiveness 146

2.1 International migration stocks (2005) and

A.7 Descriptive statistics, German aid allocation sample 176

1.1 Global distribution of refugees, in millions 3

BIS Bank for International Settlements

PDI Population Density Index

I would like to express my sincere gratitude to my supervisor, Prof.

If you don’t help the third world,

This book addresses the causes and incidences of refugee movements

1.1 Research objective

The relevance of this work is self-evident. In 2005, the United Nations

Figure 1.1 Global distribution of refugees, in millions

Particularly during the 1990s, the rise in international refugee

asylum flows to other countries, and consequently, to a race to

1.2 Research outline

The remainder of the book proceeds as follows.2 Chapters 2 and 7

In Chapter 3, the incidence and distribution of the current global

conflict and persecution. We presume that the migration decision is

with the period of increasing asylum migration movements towards

asylum applicants in the donor country itself. Further disaggregation

This chapter provides a general overview on migration characteris-

Migration is a global phenomenon. There is no country in the world

migration continues to increase, and policymakers in many states

2.2 Migration characteristics

The global migration pattern is difficult to describe and can only

Table 2.1 International migration stocks (2005) and flows (2000–05), in

Population Migrant stock Net annual inflow

Developed countries 1,211,265 115,397 2,622

Source: United Nations (2006b): Population Division.

world, which is around 0.2 percent of the developed countries’

2.2.1 Migration types

The second most important migration form is labor migration,

2.2.2 Migration causes

distinction of political and non-political migration causes (Hathaway,

Environmental factors have increasingly become a cause for emi-

2.3 International actors for refugee protection

2.3.1 UNHCR and its historical background

possibility to limit their obligations to only European refugees, while

2.3.2 Other international organizations

(ICRC), which ensures humanitarian protection of the civilian pop-

2.3.3 National states

2.4 Refugee politics

As the international economic integration of markets proceeds,

migration movements will continue to increase, and individual

One is to seek better instruments to control, to establish and

Subsequent chapters are primarily concerned with all of these pol-

2.4.1 Burden-sharing and international cooperation

have substantial problems designing adequate immigration and

the relative attractiveness of potential host states for asylum seekers

2.4.2 Migration prevention and aid assistance

and requires country-specific analysis. One policy instrument for

aid for emigration-prevention purposes is still lacking. Nevertheless,

Many developing, as well as developed, countries provide temporary

There is an ongoing debate in academia18 as well as in politics19 focus-

world and it is usually their neighboring countries (first asylum

ref iadj,fair siadj,fair ¦ ref idf (3.2)

sifair RCI i ¦ i sifair RCI i

ref iadj,df siadj,df ¦ ref idf (3.3)

and sidf ref idf ¦ i ref idf .