University of Southern California Dissertations and Theses

The allocation and accessibility of aid: the role of recipient government agency in dictating donor decisions

(USC Thesis Other)

The allocation and accessibility of aid: the role of recipient government agency in dictating donor decisions

PDF

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content i
THE ALLOCATION AND ACCESSIBILITY OF AID
THE ROLE OF RECIPIENT GOVERNMENT AGENCY IN DICTATING DONOR DECISIONS
BY
ALIX REBECCA ZIFF
A DISSERTATION PRESENTED TO THE
FACULTY OF THE USC GRADUATE SCHOOL
UNIVERSITY OF SOUTHERN CALIFORNIA
IN PARTIAL FULFILLMENT OF THE
REQUIREMENTS FOR THE DEGREE
DOCTOR OF PHILOSOPHY
(POLITICAL SCIENCE AND INTERNATIONAL RELATIONS)
AUGUST 2024

ii
DEDICATION
I would like to express the extent to which this dissertation was made possible by the
unconditional support of my family. I extend my sincerest thanks to my husband for his
patience, my advisor for his unwavering belief in my capabilities, and USC’s Center for
International Studies for enabling field work to underrepresented regions. I would like to
dedicate this dissertation to my newborn: may they grow with an appreciation for all they
have and a determination to improve the lives of others.

iii
TABLE OF CONTENTS
DEDICATION ................................................................................................................................................ ii
LIST OF TABLES........................................................................................................................................... iv
LIST OF FIGURES............................................................................................................................................v
ABBREVIATIONS............................................................................................................................................vi
ABSTRACT ..................................................................................................................................................vii
ARTICLE 1: AID ACCESSIBILITY: EDUCATION & EQUITY IN KENYA 1
............................................................................................................................. BIBLIOGRAPHY
............................................................................................................................. APPENDIX A
............................................................................................................................. APPENDIX B
............................................................................................................................. APPENDIX C
ARTICLE 2: AID ALLOCATION TO GEO-STRATEGIC STATES: AID COMPOSITION & DELIVERY IN DJIBOUTI 26
............................................................................................................................. BIBLIOGRAPHY
............................................................................................................................. APPENDIX A
ARTICLE 3: CAMEROON IN CONFLICT: ASSESSING SUBNATIONAL ALLOCATION OF HUMANITARIAN AID 47
............................................................................................................................. BIBLIOGRAPHY
............................................................................................................................. APPENDIX A

iv
LIST OF TABLES
ARTICLE 1: AID ACCESSIBILITY: EDUCATION & EQUITY IN KENYA
Table 1 Need-Driven & Donor-Driven Theories…………………………………………………4
Table 2: GLM Regression Analyses of the Relationship between educational aid and need…….8
Table A.1: Regressions on UN and NGO Aid Projects and Education Indicators ………………23
Table A.2: Regressions on UN and NGO Aid Projects and Supply-Side Indicators…………….24
ARTICLE 2: AID ALLOCATION TO GEO-STRATEGIC STATES: AID COMPOSITION & DELIVERY IN DJIBOUTI
Table A.1: Regression Analysis on Politically-Driven Variables and ODA…………………….46
Table A.2: Regression Analysis on Need Variables and ODA………………………………….46
ARTICLE 3: CAMEROON IN CONFLICT: ASSESSING SUBNATIONAL ALLOCATION OF HUMANITARIAN AID
Table 1: Single-Variate Regression Models……………………………………………………..57
Table A.1: Regressions on Need Indicators and Logged Amount of Aid……………………….68

v
LIST OF FIGURES
ARTICLE 1: AID ACCESSIBILITY: EDUCATION & EQUITY IN KENYA
Figure 1: Kenyan Enrollment Rates by County Over Time……………………………………….6
Figure 2: Breakdown of Donor Organizations by Type…………………………………………...7
Figure 3: Deviations from Need-Driven Aid across counties in Kenya…………………………...9
Figure A.1 Pie Chart on the Average Numbers of Terrorist Attacks Per Region 2012-2021……24
Figure A.2: Pie Chart on the Average Numbers Conflicts Per Region 2012-2021………………24
ARTICLE 2: AID ALLOCATION TO GEO-STRATEGIC STATES: AID COMPOSITION & DELIVERY IN DJIBOUTI
Figure 1: Map of Djibouti’s Geo-Strategic Location……………………………………………30
Figure 2: The relationship between geo-strategic value and aid…………………………………31
Figure 3: Timeline of ODA and major events in Djibouti……………………………………….32
Figure 4: Distribution of Type, Sector, and mode of Delivery of Aid from OECD Data………..35
Figure 5: Map of Djibouti’s Foreign Presence…………………………………………………...37
ARTICLE 3: CAMEROON IN CONFLICT: ASSESSING SUBNATIONAL ALLOCATION OF HUMANITARIAN AID
Figure 1: Amount of Humanitarian Aid per capita per Country in 2023…………………………51
Figure 2: Map of Regional Disparities Cameroon, 2023…………………………………………53
Figure 3: Amount of Aid Allocated to Conflict-Affected Areas…………………………………54
Figure 4: Regional Variation of Need Variables…………………………………………………56
Figure 5: Amount of Aid Allocated and the Number of Conflict Events per Year……………...58
Figure 6: Amount of Aid Allocated by Bilateral Donors to Conflict-Affected Areas…………..61

vi
ABBREVIATIONS
ACLED—Armed Conflict Location and Event Data Project
ASALs—Kenya’s Arid and Semi-Arid Lands
DFID—United Kingdom Department for International Development
EFA–Education for AllEU—European Union
HDI—Human Development Index
IO—International Organization
ISWAP—Islamic State West Africa Province
NATO—North Atlantic Treaty Organization
NGO—Non-Governmental Organization
NRC—Norwegian Refugee Council
ODA—Official Development Assistance
OECD—Organization for Economic Cooperation and Development
UIS—UNESCO Institute of Statistics
UN—United Nations
UNESCO—United Nations Educational, Scientific and Cultural Organization
UNGA—United Nations General Assembly
UNOCHA—UN Office for Coordination of Humanitarian Affairs
UPE—Universal Primary Education
USAID—United States Agency for International Development
WIDE—World Inequalities in Education Database

vii
ABSTRACT
This three-article dissertation project analyzes the subnational allocation of foreign aid. I
empirically test whether aid is need-driven and allocated to the places and populations with the
highest intensity of need, or whether aid is politically-driven and allocated according to the
political agendas of donors and recipient governments. In finding that all aid appears to be
plagued by politics, I demonstrate how recipient government agency can influence donor
decisions, resulting in either the mitigation or exacerbation of existing inequities. These studies
collectively contribute to the literature by incorporating recipient government agency, an
increasingly important and notably neglected determinant of aid into an area of research
dominated by the perspectives of Western donors. If altruistic actors are able to appropriately
identify, acknowledge, and assess recipient government power and preference, they can
overcome obstacles to extending equitable access to aid amongst politically disfavored
populations.

1
AID ACCESSIBILITY: EDUCATION & EQUITY IN KENYA
ABSTRACT
The allocation of foreign aid reflects both supply-side factors (the interests of donors) and needside factors (the needs and interests of potential recipients). Substantial evidence suggests that
supply-side factors dominate aid delivery, as both donor and recipient governments allocate aid
according to their own domestic political agendas. Are there ever conditions under which need
drives allocation and aid flows primarily to those that need it most? The distribution of education
aid across regions in Kenya offers a most-likely case for such equitable, need-driven aid
allocation. Kenya’s long history of collaboration with aid organizations and political stability
with still high levels of need create favorable conditions for aid delivery to occur equitably. Yet,
statistical analysis suggests that even there, the highest volume of aid does not flow to the
neediest areas. The introduction of new subnational data on aid location maps the concentration
of aid allocation according to need in some parts of the country. It also reveals that aid is nearly
absent in some Somali-majority areas which have the highest levels of need but the fewest
interventions. Such inequitable distribution of both education and assistance in its provision
demonstrates how the political preferences of donors and recipient governments can divert aid
away from the populations and places that need it most, reducing its effectiveness.
INTRODUCTION
Optimists, altruists, and the naïve expect that humanitarian foreign aid is allocated to
recipients based on the intensity of their need. However, decades of research demonstrates that
other factors often intervene. Donor countries allocate aid to further their own geo-strategic
interests (McKinlay and Little 1977, 1978; Alesina & Dollar, 2000); recipient-country
governments channel foreign aid to match their own domestic political agenda (Bauer, 2000;
Easterly, 2007); and aid flows to areas that are safe, inexpensive, and convenient for aid
organizations to work (Psacharopoulos, 1994; Heinrich, 2013; Liston, 2008). This paper explores
whether there are any conditions under which need for aid dominates these supply-side factors,
and under which aid is equitably distributed across regions.
Conventional wisdom holds that need-driven factors dominate when aid is ‘altruistic’,
with donors targeting places in crisis and underserved populations (Sachs, 2006; 2005;
Lancaster, 2007; Keck & Sikkink, 1998; Büthe, Major, & De Souza, 2012). Supply-side factors
dominate when aid allocation reflects the political, economic, and security priorities of the donor
and the recipient government (Boschini and Olofsgard 2007; Hook 1995; Schraeder et al. 1998).
Scholarly analysis of the relative weighting of supply-side and need-side factors has primarily
consisted of large-N cross-national studies. These, even when aggregated, are only able to
suggest that aid is allocated differently depending on the interests of the donor and characteristics
of, or conditions within, the recipient country. With more granular subnational data, we can
identify whether and when aid is actually allocated to the communities most in need.
The within-country research that does exist, has mixed results. Some studies analyzing
recipient outcomes have found generalizable evidence supporting subnational targeting of aid,
showing how proximity to aid projects has a positive impact on health indicators and poverty
reduction (Wayoro & Ndikumana, 2020; Odokonyero, Marty, Muhumuza et al. 2018). Other
studies, in the donor motivations literature, suggest that subnational aid allocation is still political
(Song, Brazys, & Vadlamannati, 2020; Briggs, 2017, 2018, 2021; Jablonski, 2014; Öhler et

2
al.,2019). The present project contributes to this literature using a most-likely case for aid to be
altruistic and allocated according to need.
Kenya’s education sector can be considered as a most-likely case to test whether need can
dominate as a determinate of aid allocation. Most-likely case studies evaluate a hypothesis in a
context where “conditions seem to make the case unusually favorable for the theory” (Odell,
166, 2001). If a theory is invalid “even in a most-likely case, this evidence would provide strong
support for the expectation that it will fail even more clearly in less hospitable circumstances,”
(ibid, 2001). Kenya is a most-likely country case with a high density of need, few obstacles to
operation, and generally favorable political conditions of aid allocation as demonstrated by a
robust aid network. Education is a most-likely sector because recipient outcomes are wellmeasured, the sector is well-supported by donors, and education is a global focus for
development communities (Song, Brazys, & Vadlamannati, 2020). Kenya’s education sector
specifically can be considered a most-likely case for educational aid to meet educational need
because of the high levels of government collaboration with aid organizations and the parallel
efforts to extend educational access and attainment. The Kenyan government has mirrored
international development agendas introducing the Free Primary Education initiative in 2002 as a
response to the Millennium Development Goal of Universal Primary Education, and equity
initiatives in 2015 as a response to the Sustainable Development Goals’ focus on equality in
education. With both the international development community and Kenyan government
prioritizing extending access to education, if foreign aid is ever able to flow efficiently to those
most in need, it should be able to do so in this context.
In order to test the relationship between educational aid and educational need in Kenya, a
new dataset mapping the subnational distribution of sector-specific aid is introduced and
compared to data on need indicators from a range of government entities and development
organizations. The new data include the source and locations of operation for 652 unique aid
projects with additional information on the donors, government collaboration, and target
populations. These data are both more inclusive and specific than datasets used in similar
subnational analyses1
and enable statistical analysis that identifies places where aid is
concentrated in the country and where it is absent. The data are inclusive of interventions from
NGOs, IOs, bilateral aid agencies, as well as smaller-scale non-profits. They are geo-coded at the
county level from 2012-2022 and are further detailed in Appendix A. In the statistical analyses, a
count of the number of projects from these data is used to test whether the highest concentration
of aid matches the highest concentration of need.2
The analysis suggests that there is a positive
correlation between educational aid and educational need. It also identifies an uneven
distribution of aid interventions across Kenya. Most notably, absence of aid in two Somalimajority counties has interesting implications, since they also have the highest intensity of need,
both in primary education and other sectors.
This article supports the findings of recent subnational studies and demonstrates how,
even in a most-likely case for need-driven equitable allocation, aid can still be systematically
diverted away from the places and populations that need it most. While many studies suggest
that recipient governments are strategic actors in the aid nexus, their role is usually discussed
through their trying to pull in more aid or to push aid to politically favored populations.
1 For example, Brass (2010,2012) uses only registered NGOs to explain where donors go and Song, Brazys, &
Vadlamannati (2020) use only World Bank projects. This study separately and collectively considered NGO
projects, UN projects, and those implemented by smaller scale non-profits.
2
In the appendices, this sector-specific dataset is considered alongside a more general measure of aid organization
presence that combines NGO data reported by the government with data from international organizations reported
by UNOCHA.

3
However, in a most-likely case context, the recipient government acts differently, and seems to
pull aid away from a politically disfavored population. Such a finding calls for more similar
studies to be done using more granular data that considers the role aid can play in the mitigation
or exacerbation of inequity especially amongst politically disfavored and marginalized groups.
I first outline the education sector in Kenya as a most-likely case for equitable aid
allocation and build upon existing evidence that explains how supply-side and need-side factors
influence donor decisions as to where they will channel aid. I subsequently introduce subnational
data and analyze them to identify where aid is absent. Then, I outline a brief history of
discrimination against Somalis in Kenya, setting the stage for additional studies to test the
inequitable allocation of aid. In this way, using detailed subnational data to map the
concentrations and absences of aid can improve our understanding of aid allocation and the
accessibility of aid amidst inequity.
The identification of the uneven subnational distribution of aid, even in a most-likely
case, suggests that aid will almost never, be allocated efficiently to those in greatest need unless
donors make a high priority of enforcing equity-based allocation within recipient countries.
Developing, revising, and implementing strategies to achieve need-based allocation will require
more careful subnational studies that analyze the motivations of donors, the actions of recipient
governments, and implications for populations in need. By highlighting inequities in places with
high intensity of need and where equitable aid distribution is expected, this article advocates for
increased attention from scholars, practitioners, and policy-makers to the accessibility of aid for
those underserved and marginalized groups.
THE ALLOCATION OF EDUCATION AID: THREE THEORIES
There is academic, practitioner, and policy consensus on the positive effect that education
has on social, political, and economic outcomes in developing country contexts. At individual,
local, and national levels, an increase in educational attainment has been shown to lead to
increased economic status (Krueger and Lindahl 2001; Barro, 2001; Hanushek, 2016), political
development (Przeworski et al., 2000; Barro, 1999), civic engagement (Milligan et al., 2004;
Kam & Palmer, 2008; Berinsky & Lenz, 2011), and decreased social inequality (Gregorio & Lee,
2002; Walsemann, Gee, & Ro, 2013). The global consensus on the importance of education, the
high number of aid organizations involved, and the focus of the development community on
education makes it a most-likely sector for equitable aid allocation.
High-level international forums and large-scale collaborative initiatives continue to aim
to extend educational access and achieve Universal Primary Education (UPE). These Education
for All (EFA) initiatives have been well-supported by governments as well as bilateral,
multilateral, and private donors. Through such collective campaigns, achieving UPE by 2015
seemed feasible with the fiscal, administrative, and operational support of recipient governments
by so many donors. Yet, in 2019 prior to the global pandemic, over 265 million children were
out of school (GPE, 2019). In Sub-Saharan Africa, 22% of children have never been to school
and 40% will not complete primary school (WIDE, 2019). These figures have only increased
with widespread interruptions in education resulting from COVID-19, and systematically
disadvantaged places and populations will continue to bear the brunt of this struggle as the world
recovers. We know that despite these staggering statistics, there are thousands of organizations
and hundreds of millions of dollars mandated to supply aid and assistance in the provision of
education. What we don’t know is why this aid is often diverted away from the populations who
need it most.
The naïve expectation with respect to foreign aid targeting is need-driven and posits that
organizations will channel aid into country and community contexts where the need for aid is

4
highest (Sachs, 2006; Drury et al., 2005; Lancaster, 2007; Keck & Sikkink, 1998; Büthe, Major,
& De Souza, 2012). Here, the term ‘organizations’ includes bilateral, multilateral, and private
donors. Within the education sector, I expect aid to flow to communities with the most children
out of school and lowest enrollment rates.
H₁: Higher numbers of out of school children and lower enrollment rates (need) are
associated with higher numbers of aid interventions (supply)
However, substantial scholarship suggests that aid is not always allocated according to recipient
need (Easterly, 2003; Bueno de Mesquita & Smith 2009, 2015; Bearce & Tirone 2010). Instead,
aid may reflect political or practical donor-driven considerations. Donor-driven theories focus on
the supply-side of aid and hold an expectation that organizations choose places where they can
avoid logistical obstacles to operation and where they can achieve political aims. Political factors
include the geo-strategic economic, political, and security interests of the donor (Alesina and
Dollar 2000; Boschini and Olofsgard 2007; Hook 1995; McKinlay and Little 1977, 1978;
Schraeder et al., 1998; Bueno de Mesquita & Smith, 2018). In the subnational context, donors
driven by political factors allow recipient governments to allocate aid and are less concerned
with equitable distribution across the country (Dreher, Langlotz, & Marchesi, 2017; Song,
Brazys, & Vadlamannati, 2020). Therefore, aid may be disproportionately distributed to
populations favored by the government (Alesina, Baqir, & Easterly, 1999).
H₂: Presence of a population that is politically disfavored by the donor is associated with
fewer aid interventions
Practical/logistical supply-side factors are less explored in the literature and reflect the rate of
return for operation in a subnational location. The convenience and cost per capita for an aid
organization to work in a particular place will increase when there is limited existing
infrastructure, security concerns, and fewer people to serve. I expect that these locations will be
avoided in favor of places with more convenient working conditions that can reach more people
(Brass, 2010).
H₃: More infrastructure, security, and people in an area is associated with more aid
interventions
Theory: Need-Driven Practical Donor-Driven Political Donor-Driven
Donor
Motivation:
Need of recipient
population
Financial feasibility of
operation for donor
Political, economic, or
security interests of donor
and/or favor by recipient
government
Table 1 Need-Driven & Donor-Driven Theories
A review of a wide range relevant literature from different disciplines finds evidence for
all three theories, thereby demonstrating how testing each of these in a most-likely case scenario
could fill a gap in the scholarship.3 By focusing on need-driven theory, while also analyzing
political and logistical supply-side factors, it integrates consideration of potential omitted
variables that also affect aid allocation and are found in single theory studies. Using a mostlikely case framework avoids the contextual limitation of case studies to contribute to the larger
literature. In addition, because bilateral aid agencies, multilateral institutions, NGOs, and
3
I reference seminal works from political science, economics, education scholarship, development studies, in-depth
ethnographic research, and historical accounts. I also use policy papers and practitioner publications for further
background.

5
smaller-scale organizations have different determinants of aid (Berthélemy, 2006) I assess all
actors in the aid nexus separately and collectively to capture as many different motivations for
aid allocation as possible.
THE EDUCATION SECTOR IN KENYA AS A MOST-LIKELY CASE
In comparison to other countries in the region, Kenya has more favorable practical and
political factors for aid allocation. In addition to high levels of need, which support need-driven
explanations, Kenya also has an extensive aid network, history of government collaboration,
infrastructure, and available data. Better infrastructure with more electricity and internet
capability than most other countries on the continent (UIS, 2019) make for easier operation by
organizations. Such comparatively strong infrastructure may be why Kenya has one of the most
robust international development networks in the global south, with over fifty thousand formally
registered NGOs (National Council of NGOs, 2020). The country is dotted with ‘NGO Centers’,
many of which have existed for decades, that are supported by citizens, private funding flows,
bilateral, and multilateral donors alike (Brass, 2021). The government has, and continues to,
collaborate with aid organizations, thereby eliminating potential political conflict between donor
governments and recipient countries.4
These favorable supply-side factors attract additional
donors who will work where they can expect high returns to investment with a lower cost per
capita (Psacharopoulos, 1994). Therefore, with few practical obstacles and favorable political
conditions, Kenya can be seen as a most-likely case for equitable need-driven aid allocation.5
The education sector is an attractive target for aid allocation because of its prioritization
by the international development community, its visible and measurable outcomes, and its
defensibility as an altruistic contribution to recipients. It is also considered a policy priority by
the Kenyan government. Since independence, each administration has attempted to employ or
expand free and compulsory primary education (FPE). Numerous FPE initiatives attempted to
increase education availability through building schools, curricular development, and investment
in school system capacity. Given this consistency of government attention and the amount of aid
allocated to the education sector, it is reasonable to consider Kenya a most-like case for progress
in the expansion of education. Yet, even here, with existing education infrastructure and
assistance from aid organizations, UPE has yet to be achieved.
While some studies claim that Kenyan FPE initiatives have increased enrollment
significantly (Oketch & Sommerset, 2010; Ohba, 2009; Oketch, Mutisya, Ngware, & Ezeh,
2010), others suggest that they were “little more than cosmetic” (Colclough & Webb, 2010;
4 Government collaboration is an especially significant indicator for feasibility of interventions in the education
sector as a prerequisite of many organizations who prioritize sectoral improvement through systematic change in
collaboration with the government (Fhi360, 2017)
5 Kenya’s history of working with development organizations and the associated availability of data has also made it
a popular choice for case studies assessing aid (Brass, Longhofer, Robinson, & Schnable, 2018). I build upon Brass
(2010,2012) and Song, Bazys, & Vadlamannati (2020) by integrating a larger county level dataset which includes a
broader range of interventions from all types of donors (multilateral institutions, bilateral aid agencies, NGOs and
smaller-scale non-profits), not just the World Bank or NGOs registered with the Kenyan government. I also expand
on and test three theories of the determinants of aid. While my measures of need remain consistent with these two
studies and focus on the levels of need of populations, I incorporate both donor-based and recipient-based
interpretations of political supply-side factors. I also create a more inclusive theory of practical supply-side factors
that include measures of convenience, but also cost-per-capita and other variables that proxy as obstacles for
organization operations. As a result of these expansions, I find that donor-driven rather than need-driven
explanations better predict where aid is concentrated and where it is absent.

6
Tooly, 2009; Bold et al., 2012). In 2019, statistics from the Ministry of Education and UNESCO
claim that Kenya has achieved FPE with a 99.6% enrollment rate.6 Yet, when these data are
disaggregated at the county level, some areas have enrollment rates as low as 27%. In addition,
the Kenyan government reports that only 59% of education expenditures come from central
(55%) and local (4%) governments, leaving 33% of the financial burden on parents (KICD,
2020). Figure 1 maps this variation in enrollment rates over time, illustrating substantial
variation in the intensity of need for assistance in extending education and the extent to which
earlier efforts have succeeded in some contexts and failed in others.
Figure 1: Kenyan Enrollment Rates by County Over Time
In the 1980s and 1990s, austerity measures and limited capacity decreased quality and
increased the opportunity costs of school attendance (Sommerset, 2009; Nishimura & Yamano,
2012; Murphy, 2003). As evident by the regional disparity illustrated in Figure 1, FPE initiatives
thereafter were largely concentrated in urban areas amongst lower-middle class citizens.
Refugees, immigrants, girls, and rural communities were largely unreached by these
improvements (Sommerset, 2009). In their 2010 survey of the impact of the FPE measure on
four diverse districts across Kenya, Oketch & Sommerset found that the extension of equitable
access to education was uneven within their sample of Nyeri, Kisii, Kajiado, and Nairobi.
Particularly in urban areas, schools imposed additional barriers to access through amenities fees
for essential services like bussing, libraries, and lunches. The school system was oversaturated
and lacked the administrative and fiscal capacity to support this increase in enrollment (Lucas &
Mbiti, 2012; Nishimura & Yamato, 2013). There have also been consistent teacher shortages
attributed to the increase in enrollment since the implementation of FPE and the Kenyan
Ministry of Education claims to need over 100,000 teachers (NESSP, 2018) with projected
increases over the next five years.
Such demographic divisions are commonly cited in education literature as well as fields
of practice and scholarship for disaggregated subnational data and cultural competence in their
analysis. Inequities are especially evident with urban-rural divides (Sahn & Stifel, 2003;
UNESCO, 2018), gendered discrepancies (Burde, Kapit, & Wahl, et al. 2017)), and amongst
lower socioeconomic strata (UNESCO, 2018). Systemic discrimination against populations from
certain racial, ethnic, religious, or language groups is also evident in public service provision
inequities (Robinson, 2007). This varies depending on the specific country and often even
within-community context, but numerous case studies suggest that being a minority has inverse
6 Pre-pandemic data are considered in the study as a means to gain a general understanding of need for aid and
assistance in education and with the assumption that those with limited access prior to the pandemic, will have
comparatively similar or worse access thereafter its conclusion.

7
relationships with educational access, enrollment, and retention (Tackey, 2009; Machin &
McNally, 2006; Psacharopoulos, 1993; Rose, 2015; UNESCO, 2018).
These issues are especially acute amongst certain places and populations in Kenya (Julius
& Bawane, 2011). A qualitative assessment of education data across sources shows that the more
central counties have better quality education, fewer out of school students, and higher
attainment.7 Students living in urban areas closer to the capital, have a higher chance of entering
secondary school and remaining in school8
. Rural counties and desert areas (ASALs) have poorer
quality education, more inequity between populations, and more acute shortages in teachers and
resources (UNESCO, 2018; Coclough & Webb, 2012). These issues of availability and
accessibility are compounded for children from underrepresented groups such as refugees,
immigrants, girls, children with disabilities, and minority populations. Therefore, to understand
the distribution of education in Kenya, it is important to incorporate disaggregated measures of
need that reflect the demography of the country.
Assessing the Distribution of Aid
I introduce a new county-level dataset to measure the supply of aid extending availability of and
access to primary education.9 These data range from 2012 to 2022 and are geo-coded at the
county level and detail each of the 652 unique project’s location(s) of operation, target
population, mission, and affiliated organizations. The broad inclusion criteria aim to capture all
aid and assistance going towards the extension of primary education in Kenya regardless of scale
or size. The extensive data collection procedure used thousands of policy papers, datasets, budget
7 The lack of consistent data disaggregated at the same levels and over similar years makes overtime quantitative
analysis difficult. The qualitative assessment compared reports from 1990, 2003, 2004, 2008, 2010, 2014, and 2018
from the Kenyan Ministry of Education, Kenyan Bureau of Statistics (KNBS), World Education Inequalities
Database (WIDE), DHS surveys, and MICS surveys.
8 This supports the findings of Liston (2008) who also notes an urban bias with NGO concentrations in Kenya
across sectors.
9 Descriptive statistics and more detail in Appendix A

8
reports and advocacy publications as well as existing networks. The resulting snowball sample
represents a diverse and robust community of donors .10
Figure 2: Breakdown of Donor Organizations by Type
While it appears that majority of interventions are NGOs, Figure 2 reflects the implementation
of projects, not their funding source or level of associated resource dependency on other
organizations. For example, 35% of these donors collaborate directly with the Kenyan
government in implementation of their projects and 60% of projects involve more than two
donor organizations. Where prominent sources of education data are notorious for missingness
and underrepresentation of minority or marginalized populations, this dataset paints a detailed
picture of the aid landscape across Kenya’s education sector.
If aid organizations select into places with the most need, then we would see
concentrations of aid allocated to populations with the highest number of out of school children,
lowest enrollment rates, and poorest quality schools. The analysis below uses a negative
binomial model to individual test the relationship between education aid and several measures of
education need. It finds an overall positive correlation between the number of aid projects and
need for extended education.
GLM Regressions on Education Aid and Educational Need and Controlled for Population Size
Dependent variable: Number of Education-Targeted Aid Projects
(1) (2) (3) (4) (5)
Enrollment Rate -13.128**
10 However, because these data were accessed online, interventions associated with large, prominent organizations
will be systematically overrepresented and those smaller-scale organizations who work locally and lack online
presence will be underrepresented.

9
(5.490)
Number of Out of School Children -0.009***
(0.003)
Private Schools 0.00004***
(0.00001)
Student Teacher Ratio 0.149
(0.111)
Kenyan Government Projects Extending Education 7.802***
(2.893)
Constant 20.693*** 3.926 -1.736 8.186*** 5.065***
(5.642) (2.841) (2.588) (2.027) (1.652)
Observations 46 46 46 47 46
Log Likelihood -168.025 -169.689 -163.338 -173.937 -167.316
Akaike Inf. Crit. 340.051 343.379 330.675 351.873 338.633
Note: *p
**p
***p<0.01
Table 2: GLM Regression Analyses of the Relationship between educational aid and need
The above analysis shows that lower enrollment rates and more out of school children are
correlated with more aid projects in an area. A higher number of private schools in an area being
met with more aid projects, suggests that effort is being made to decrease financial barriers to
accessing education. Additionally, regions of focus for the Kenyan government’s efforts for
extending public education are similarly associated with more aid projects, further demonstrating
a close connection between donors and the recipient government. These results are encouraging
for the most part, suggesting that there are conditions under which aid allocation reflects
recipient need.
WHERE DOES AID SUPPLY DIVERGE FROM NEED?
Consistent with need-driven expectations and H₁, with a few exceptions, the counties
with higher measures of need also have more interventions. However, it is those exceptions
which have the most important substantive implications. These poorly predicted cases include
the four counties with the highest intensity of need. Wajir and Mandera counties have the lowest
levels of education access and attainment, and Garissa and Turkana counties have the largest
refugee populations. Without these counties considered, the parameters of the sample would
change significantly with the lowest enrollment rate going from only 28% in Mandera county to
a low of 65% enrollment in Mombasa. The fact that these exceptions are most closely connected
to need-driven theory and are also least associated with the relationship between intensity of
need and interventions suggests that further analysis is necessary to determine whether aid is
actually need-driven in Kenya’s education sector.
The counties whose aid supply is poorly predicted by their level of need can be grouped
into the three categories and are visualized in Figure 3. First, in Bomet and Machakos counties,
there are few out of school children and high enrollment rates (above 92%). Yet, these counties

10
have the most interventions; the need for aid is low, but the supply is high. Investigating each of
the projects’ history, target population, and associated organizations reveals several potential
reasons for this overallocation of aid. In both counties, there are favorable supply-side
conditions: they are close to the capital, function as agricultural hubs with urban centers, and
have histories of collaboration between the community and aid organizations. In Bomet county,
many aid organizations entered during and after periods of ethnic violence that erupted in the
Rift Valley province (Stewart, 2010). These surges of aid allocation in the 1990s, and again after
the Kenyan Crisis in 2007-2008, resulted in the establishment of an aid network that continues to
function. However, the nature of the current interventions has shifted away from primary
enrollment efforts of the 1990s to student retention, advancement, and quality instruction.
Interventions in Machakos county similarly reflect past FPE projects with new purposes, so
need-driven theory cannot be rejected here.
Figure 3: Deviations from Need-Driven Aid across counties in Kenya
In Garissa and Turkana counties, higher levels of need are associated with more
interventions. These are well predicted cases demonstrating instances where aid organizations
meet need left by gaps in public service provision. Both counties are rural, far from the capital,
and more significantly—home to the two largest refugee camps in Kenya: Dadaab and Kakuma.
Dadaab and Kakuma are home to over a million persons without Kenyan citizenship, most of
whom are registered refugees and account for 84% of the total Kenyan refugee population
(CARE, 2019; UNHCR, 2019). The Kenyan government lacks the capacity to provide for such a
large refugee population, and so, since their establishment in the 1990s, external entities have
supported provision of public services in the camps. The concentration of interventions in
Garissa and Turkana therefore supports the need-driven explanation of aid allocation, with
organizations operating in these locations with low enrollment rates, high numbers of out of
school children, and limited public service provision.
Wajir and Mandera counties, however, have similarly high levels of need but very few
interventions. Wajir, Mandera, and Garissa counties comprise the Northeastern Province of
Kenya. These three counties consistently have the highest level of need across all measures.
Therefore, provincial level data can be used to complement the measures already included. By a
large margin, the Northeastern Province has the highest proportion of children who have never
attended school (34%) and are out of school (44%) (WIDE, 2019). We know that with the

11
Dadaab refugee camp, Garissa County receives significant support from external entities. Yet,
neighboring counties Wajir and Mandera, are comparatively desolate. Ministry of Education data
shows that less than one third of primary-aged children of citizens living in these two counties
were enrolled in school in 2019. Since then, while all counties experienced significant increases
in the number of children out of school due to the pandemic, there were disproportionate
increases in these underserved areas (Ngwacho, 2020). The pandemic has exacerbated existing
inequities in access and availability of education to the Northeastern Province making it an ideal
candidate for need-driven aid allocation, yet Wajir and Mandera still have the fewest number of
interventions.
The state of public services in Wajir and Mandera reflect a long history of systematic
discrimination against the Somali-majority population of the region. The neglect of the
Northeastern Province is evident not only in education, but also across social, political, and
economic sectors. Regional favoritism in public service provision and development is a welldocumented phenomenon, especially in Sub-Saharan Africa (Bates, 1984; Hodler & Rashkey,
2014). Weitzberg (2017) details the systematic marginalization of Somalis in Kenya and
elsewhere over the last several decades and claims it to be a persistent and worsening issue
largely ignored by the international community. Perhaps then, these legacies of discrimination
can confirm H₂ and explain the absence of both basic education and aid allocation in Wajir and
Mandera.
Politically motivated diversion of public services and aid away from Somalis is not a farfetched theory considering the central role Kenya plays as a case study in the large bodies of
literature on how regional and ethnic favoritism can influence policy, elections, and the
distribution of public goods (Alesina, Baqir, & Easterly, 1999), especially in Kenya (Jablonski,
2014; Dickens, 2018; De Luca, Hodler, Raschky, & Valsecchi, 2018; Burgess et al., 2015). In
the context of education, Kramen and Posner (2016) find that throughout Kenya’s history,
having a co-ethnic serve as the president during one’s childhood results in an increase in
educational attainment and Li (2018) finds that it improves the likelihood of primary school
completion. Such evidence certainly supports the idea that public service provision is influenced
by political favor. However, like most of the literature on ethnic favoritism, both of these articles
focus on the connection between political office and ethnic favoritism. Despite the similar
application to education, the limited scope of ethnicity in relation to political outcome results in
the omission of an essential consideration in education and public service provision: the
treatment of politically disfavored and marginalized groups.
Disparities in education between the northeastern Somalia-majority counties, and the rest
of Kenya are evident even prior to independence. Turton’s (1974) historiography of early
education provision in the Somali-majority areas of Kenya demonstrates the persistence of
discrimination against Somalis from the colonial era through the 1970s. Prior to 1946, there was
not a single school in modern-day Wajir and Mandera counties, despite several complaints and
petitions recorded as early as 1930.11 Colonial administrations cast doubt on the value of
extending educational access to Somalis and dismissed complaints, ironically citing that they
would not favor the Somali community over any other (ibid, 1974).12 At the time of
independence, less than 1% of Somalis in Kenya had access to education (Wajir District Annual
Report, 1963).
11 In 1959, there were fewer than 5 primary schools in the area, and community organizations complained of 4-year
wait times for new schools to be permitted (ibid, 1974)
12 The lack of education was particularly surprising considering the successes of colonial administration
collaboration with Koranic schools in British Somaliland.

12
The government’s imposition of a 28-year state of emergency upon independence, further
suggests systematic disenfranchisement of Somali populations. The government diverted
development funds away from the region and used propaganda to ‘other’ Somalis and suppress
secessionist uprisings in 1963-1967 (Whittaker, 2012). Therefore, early FPE efforts excluded
Somali-majority regions. Then, in 1971, a presidential decree abolished school fees in northern
counties as part of an initiative to increase educational access (Government of Kenya, Gachathi
Report, 1976). However, the “failure to conform” by Muslim and pastoralist communities and
the insubstantial increases in enrollment led the government to scale down these education
expansion efforts in favor of decentralized cost-sharing policies in the 1980s and further
disadvantaged the Northeastern Province (Ruto, Ongwenyi, & Mugo, 2010). The 2003 FPE
policies and increases in government expenditures succeeded in some areas, but progress in
Wajir and Mandera still lagged, highlighting the limited administrative capacity and education
infrastructure in the region.
Today, the continued marginalization of Somalis by the Kenyan government can still be
seen through general exclusion from political processes, explicit prevention of migration by
Somali-Kenyans to other parts of the country, and ethnic violence both perpetuated by the state
and ignored by it (Feghali, Faria, & Jama, 2021; Weitzberg, 2017). Territorial disputes between
Kenya and Somalia have strained diplomatic ties between the two countries and resulted in
UNSC and ICJ involvement. The global war on terror and insurgence of Al-Shabaab in Somalia
and Kenya, has resulted in stereotyping of Somalis as ‘terrorists’ prompting targeted violence
and discrimination (Lochery, 2012; Human Rights Watch, 2013). This persistent ‘othering’ of
Somalis through the media is reminiscent of colonial and independence era propaganda and is
similarly successful at inhibiting social cohesion. It is possible that exacerbation of the existing
social inequities between the Northeastern Province and the rest of Kenya could explain why
need-driven theory does not explain aid allocated to these areas.
Supply-Side Explanations
It is much easier to see the connection between marginalization by a government and
limited access to public services than it is to theorize why international aid organizations are not
reaching underserved communities. Altruistic aid should target the places and populations with
the highest levels of perceived need, yet Wajir and Mandera consistently have the fewest number
of interventions. Recent academic and practitioner studies of aid allocation suggest that these
inequities could be a result of other factors that influence where aid organizations work.
The practical supply-side explanations for lack of aid allocation to Wajir and Mandera
suggest that organizations may not work there due to financial feasibility or other obstacles to
operation, confirming H₃.13 Limited existing infrastructure and aid networks in the north increase
barriers to entry (Heinrich, 2013). The distance from the capital or other urban hubs, makes for a
challenging commute (Liston, 2008). Inconsistent electricity, limited internet capability, lack of
lettable buildings, and resource shortages make it difficult for aid organizations to host offices or
run schools. In Northern Kenya, security remains a serious concern with the threat of AlShabaab, border disputes, and local conflict (Haider, 2020).14 With fewer people and more
obstacles than other areas, the cost per capita to deliver aid is higher in Wajir and Mandera as
13 The analyses support Brass (2010, 2012) who also uses a within-country analysis of Kenya to investigate donor
motivations and finds that need-driven (need) and logistical factors (convenience) dictate the location of NGO
operations. While this study uses slightly different interpretations and measurements of political and practical
factors, it confirms her findings that both need and convenience play a role in where aid organizations work.
14 As visualized in Appendix D, the number of terrorist attacks and conflicts are substantially higher in the
Northeastern Province. While this may provide evidence for supply-side explanations for the relative absence of aid,
such explanations are not independent of marginalization of the Somali community.

13
compared to other counties which have lower levels of need but more easily accessible
populations.
Therefore, organizations could, in theory, still be allocating aid altruistically. However,
rather than focusing on the places and populations that have the highest need, organizations may
choose to target places with a lower cost per capita and higher population density. That way, they
can extend education to a larger proportion of the population. While this leaves Wajir and
Mandera still without equitable access to aid or education, it also creates an actionable agenda. If
this is the case, then, by taking these barriers to entry and obstacles to operation into
consideration, aid organizations could adapt targeted assistance to Wajir and Mandera through
existing community-based channels, thereby lowering the cost per capita. In this way, integrating
framework for the practical supply-side considerations of can provide a foundation from which
scholars and practitioners can collaborate to develop more financially feasible and culturally
competent approaches to aid allocation.
However, in practice, it appears that interventions meant to address these practical/
logistical obstacles have not been entirely successful and political supply-side theory may better
explain the inequitable distribution of aid and education in Wajir and Mandera. Kenya’s National
Council for Nomadic Education (NACONEK), for example, is a governmental agency founded
for the exact purpose of extending education to these underserved regions. According to their
latest annual report, there are specific challenges to providing education in Wajir and Mandera.
They cite pastoral migratory patterns, inclimate weather, and “inter-clan violence” or “intra
community conflict” as particularly challenging (NACONEK, 2022). At first, it seems to support
the practical supply-side explanation, discussing how drought, flash floods, and migratory
patterns make transportation of learning materials or construction of school sites difficult.15

However, despite similar conditions with respect to weather and violence, Garissa County has a
large concentration of aid organizations and education-extending operations. A thirty-year
collaboration between donors and the Kenyan government in Dadaab refugee camp has resulted
in need-driven aid allocation in Garissa.16 Perhaps then refugee status is a necessary condition
for aid organizations to bypass the political preferences of a recipient government.
In that same 2022 NACONEK report, a section on Mandera states that “6 out of 10
children are out of school due to poverty, long distances to schools, violence, cross-border
activities, and the fact is that education is not a priority in these families” (ibid, 2022). While
need for education can be difficult to measure, there is substantial archival evidence with nearly
a century of petitions for the extension of educational access in these areas suggesting that
education is indeed a priority. Such a report demonstrates bias against Somalis within the very
government institution that is meant to address education inequities in those regions.
Similarly, in response to Al-Shabaab attacks on schools in these areas, the government
responded by requiring the relocation “non-native” teachers, exacerbating already existing the
teacher shortage and further limiting educational access. If marginalization by the Kenyan
government explains the higher measures of educational need in these Somali-majority areas,
then there is reason to believe that other marginalized groups would have similar outcomes. Yet,
the presence of populations classified as marginalized groups according to the Kenyan
constitution is actually associated with more government efforts to extend education and also
15 The analysis of practical and political factors such as road coverage, drought, flood, and bribery suggest that these
factors do not explain the variation in the distribution of public schools nor educational aid.
16 Even in Garissa with Dadaab refugee camp, there is evidence of Somali marginalization with several threatened
closures in 2015, 2016, and 2019 as well as the illegal repatriation of Somali refugees back to Somalia (Kituo cha
Sheria, 2019).

14
more projects from aid organizations. Thus, there is something systematically different about the
treatment of Somalis in Kenya.
These examples suggest that if aid organizations are not carefully considering subnational
political factors, they may be complicit in the marginalization of disfavored populations by
recipient governments. The findings of this study support those of recent scholarship which show
how a range of donors are unable to, or uninterested in, preventing governments from diverting
aid away from politically disfavored groups (Briggs, 2014; Jablonski, 2014; Masaki, 2018;
Anaxagorou et. al, 2020). Still, in other country contexts, many aid organizations have acted as
important public service providers for marginalized groups and agendas of inclusive
development have been adopted by aid agencies and governments. The growing literature on
addressing inequity, reaching the marginalized, and inclusive development, illustrate increased
attention to historically underserved groups. So why not with Somalis in Kenya?
While the testing of explanations for the absence of aid amongst Somali-majority areas
are outside of the scope of this study, inductive reasoning suggests several possible explanations
that pave way for the collection and testing of more granular data. For example, it is possible that
perceptions of Somalis by the international community could contribute to the absence of aid.
Ethnographic studies of the Somali diaspora note that discrimination and marginalization are not
unique to Kenya (Mutsune, 2006; Lochery, 2012; Wairuri, 2020; Weitzberg, 2015, 2017). The
Global War on Terror has led to a sharp rise in Islamophobic ideals and action (Kundnani, 2004;
Asultany, 2013; Sides & Gross, 2013). This, in conjunction with the piracy crisis off the Gulf of
Aden and the rise of Al-Shabaab, has compounded the effect of this conflation and stereotyping
for Somalis (Fellin, 2015). Public perception of recipient populations can play a role in aid
allocation (Heinrich et al., 2005, 2016; Joly, 2014; Paxton & Knack, 2012) as bilateral aid
agencies answer to the constituency of their country, multilateral institutions rely on their
membership, and NGOs rely on public and private donors (de Waal, 1997; Hancock,
1989).Therefore, the negative perceptions of Somalis by the public could influence aid
organizations’ absence in the Somali-majority counties.
In addition, many bilateral aid agencies advertise their use of foreign aid to advance
national interest (USAID, 2020; DFID, 2020; GOC, 2021) and generally allocate aid through
recipient governments. Since donor governments often give more to their allies (Bouton &
Carter, 2014; Alesina & Dollar, 2000; Schraeder, Hook, & Taylor, 1998)17 so the territorial and
diplomatic disputes between Kenya and Somalia may drive donors’ decisions to avoid allocating
aid to Somalis if they are allied with Kenya. Bilateral donor decisions may also be driven by
institutional agreements such as UNSC Resolution 2462 which prevents donor governments
from providing bilateral aid to groups associated with terrorist activity. The presence of some Al
Shabaab activity in the Northeastern Provinces and this resolution may explain the absence of
bilateral projects in these Somali-majority areas.
Finally, donors may not intend to be affected by political supply-side factors but still do
so through government regulation of aid or contingent collaboration. The Kenyan government
may be continuing a pattern of diverting resources away from Wajir and Mandera either with or
without the knowledge of donors. Fungibility literatures show how governments redirect aid
strategically toward specific areas and away from others (Feyzioglu et al., 1998; Rana & Koch,
2021; O’Brien-Udry, 2021). locations of operation are thus likely also determined by the central
government and can be channeled away from these needy counties. It is also possible that the
government intentionally incentivizes or advises aid agencies to work elsewhere. These political
17 Zhang (forthcoming) provides new counter evidence for this pattern by looking specifically at times of global
insecurity and focusing on small states.

15
supply-side explanations invite further research that test such explanations and consider how
recipient governments, the national interest of donor governments, and public perception may
influence aid allocation at the subnational level.
CONCLUSION & CONTRIBUTION
While we hope that aid allocation is need-driven and allocated to areas with high
intensity of need, persistent inequities within countries with concentrations of aid organizations
suggests that need rarely dominates donor’s decisions. A within-country study of Kenya provides
opportunity to assess need in a narrow domain, with its strong aid infrastructure, a long history of
collaboration with donors, and high levels of need. The education sector in Kenya provides a
‘most-likely case’ for equitable aid allocation with data availability, education-oriented policy
agendas, existing aid infrastructure, and state capacity. Quantitative analysis of the relationship
between need for aid and its supply shows some support for the need-driven hypothesis but also
identifies a few extreme cases with important implications.
An evaluation of the uneven distribution of aid in Kenya highlights that these poorly
predicted counties are significantly the places with the lowest enrollment rates and highest
numbers of children out of school. In Garissa and Turkana counties, the largest refugee camps
generate two of the largest levels of need. This need is met by the largest share of education
interventions, consistent with need-driven theory. However, in Wajir and Mandera there is a
clear deviation from need-driven aid allocation with few interventions in the two counties with
the highest levels of need. This shows that, even in a most-likely country case for need-side
factors, supply-side considerations can divert aid away from the populations and places that need
it most. I argue that political factors, especially disfavor of or discrimination against particular
populations, likely explain the diversion of aid away from the Somali-majority counties of Wajir
and Mandera.
The ability of supply-side factors to divert aid away from the neediest recipients, even in
the most favorable cases, has important implications for those who aspire to distribute foreign
aid efficiently and equitably. Among scholars, additional research is necessary to understand the
extent to which discrimination is driven by the preferences of the recipient government, the
biases of donor governments, the aid organizations themselves, or some other mechanism.
Among practitioners, it is necessary to recognize how profoundly challenging it is to achieve
need-driven aid allocation. Doing so requires overcoming both practical barriers associated with
reaching the most vulnerable individuals, as well as political barriers that can divert aid away
from marginalized groups. Bilateral aid and education aid are especially likely to flow through
the central government18, making attention paid to subnational politics particularly important if
aid is meant to reach underserved communities.
18 Education aid most often flows through the central government as a means to reform national curricula, to
standardize, monitor, and evaluate progress, and as a result of Western traditional education structures. Education
aid also often requires government collaboration for sustainable systematic change in the provision of this public
service. Therefore, channeling aid to the education sector may be conditional on the collaboration of the
government. This makes attention paid to subnational politics particularly important if aid is meant to reach
underserved communities.

16
BIBLIOGRAPHY
Aldashev, Gani, and Thierry Verdier. 2008. “Goodwill Bazaar: NGO Competition and Giving to Development ☆.”
Journal of Development Economics 91: 48–63.
Alesina, Alberto, and David Dollar. “Who Gives Foreign Aid to Whom and Why?” Journal of Economic Growth 5,
no. 1 (2000): 33–63.
Ali, Murad, Glenn Banks, and Nigel Parsons. 2015. “Why Donors Give Aid and to Whom ? A Critique of the
Historical and Contemporary Aid Allocation Regime.” The Dialogue 10(2): 116–31.
Apodaca, Clair. "Foreign aid as foreign policy tool." In Oxford research encyclopedia of politics. 2017.
Alsultany, E. (2013). Arabs and Muslims in the media after 9/11: Representational strategies for a" postrace"
era. American Quarterly, 65(1), 161-169.
Audette, Nicole et al. 2020. Personal Narratives Reduce Negative Attitudes toward Refugees and Immigrant
Outgroups: Evidence from Kenya.
Bapat, Navin A. "Transnational terrorism, US military aid, and the incentive to misrepresent." Journal of Peace
Research 48, no. 3 (2011): 303-318.
Barnett, Michael. "Humanitarianism transformed." Perspectives on politics 3, no. 4 (2005): 723-740.
Barro, Robert J. "Determinants of democracy." Journal of Political economy 107, no. S6 (1999): S158-S183.
Barro, Robert J. "Education and economic growth." The contribution of human and social capital to sustained
economic growth and well-being 79 (2001): 13-41.
Bauer, Peter Tamas. From Subsistence to Exchange and Other Essays. Princeton: Princeton University Press, 2009.
Bearce, David H., and Daniel C. Tirone. "Foreign aid effectiveness and the strategic goals of donor governments."
The Journal of Politics 72, no. 3 (2010): 837-851.
Berinsky, Adam J., and Gabriel S. Lenz. "Education and political participation: Exploring the causal link." Political
Behavior 33, no. 3 (2011): 357-373.
Bermeo, Sarah Blodgett, and David Leblang. "Migration and foreign aid." International Organization 69, no. 3
(2015): 627-657.
Bermeo, Sarah Blodgett. "Aid allocation and targeted development in an increasingly connected world."
International Organization 71, no. 4 (2017): 735-766.
Bermeo, Sarah Blodgett. "Foreign aid and regime change: A role for donor intent." World Development 39, no. 11
(2011): 2021-2031.
Berthélemy, Jean Claude. "Bilateral donors’ interest vs. recipients’ development motives in aid allocation: do all
donors behave the same?." Review of Development Economics 10, no. 2 (2006): 179-194.
Berthélemy, Jean-Claude, and Ariane Tichit. "Bilateral donors' aid allocation decisions—a three-dimensional panel
analysis." International Review of Economics & Finance 13, no. 3 (2004): 253-274.
Blodgett Bermeo, Sarah. 2017. “Aid Allocation and Targeted Development in an Increasingly Connected World.”
International Organization 71: 735–66.
Bold, Tessa, Mwangi Kimenyi, Germano Mwabu, Alice Ng’ang’a, and Justin Sandefur. "Interventions &
institutions experimental evidence on scaling up education reforms in Kenya." Preliminary draft. Available at
http://www. iies. su. se/polopoly_fs/1.101632 13481 (2012): 37980.
Boschini, Anne, and Anders Olofsgård. "Foreign aid: An instrument for fighting communism?." The Journal of
Development Studies 43, no. 4 (2007): 622-648.
Boutton, Andrew, and David B. Carter. “Fair-Weather Allies? Terrorism and the Allocation of US Foreign Aid.”
Journal of Conflict Resolution 58, no. 7 (October 2014): 1144–73.

17
Brass, Jennifer N. 2012. “Why Do NGOs Go Where They Go? Evidence from Kenya.” World Development 40(2):
387–401.
Brass, Jennifer N., Wesley Longhofer, Rachel S. Robinson, and Allison Schnable. 2018. “NGOs and International
Development: A Review of Thirty-Five Years of Scholarship.” World Development 112: 136–49.
Brass, Jennifer N., Wesley Longhofer, Rachel S. Robinson, and Allison Schnable. "NGOs and international
development: A review of thirty-five years of scholarship." World Development 112 (2018): 136-149.
Brass, Jennifer Naomi. Surrogates for government? NGOs and the state in Kenya. University of California,
Berkeley, 2010.
Briggs, Ryan C. 2018. “Leaving No One Behind? A New Test of Subnational Aid Targeting.” Journal of
International Development 30(5): 904–10.
Briggs, Ryan C. 2021. “Why Does Aid Not Target the Poorest?” International Studies Quarterly 65(3):
739–52.
Burnside, Craig, and David Dollar. "Aid, policies, and growth." American economic review 90, no. 4 (2000): 847-
868
Büthe, Tim, Solomon Major, and André Melloe De Souza. 2012. “The Politics of Private Foreign Aid: Humanitarian
Principles, Economic Development Objectives, and Organizational Interests in NGO Private Aid Allocation.”
International Organization 66(4): 571–607. https://about.jstor.org/terms (February 7, 2020).
By, Edited, Nabamita Dutta, and Claudia R Williamson. Lessons on Foreign Aid and Economic Development Micro
and Macro Perspectives.
Colclough, Christopher, and Andrew Webb. "A triumph of hope over reason? Aid accords and education policy in
Kenya." Comparative Education 48, no. 2 (2012): 263-280.
Cooley, Alexander, and James Ron. "The NGO scramble: Organizational insecurity and the political economy of
transnational action." International security 27, no. 1 (2002): 5-39.
De Mesquita, Bruce Bueno, and Alastair Smith. "A political economy of aid." International Organization 63, no. 2
(2009): 309-340.
De Mesquita, Bruce Bueno, and Alastair Smith. "Political loyalty and leader health." Quarterly Journal of Political
Science 13, no. 4 (2018): 333-361.
De Waal, Alexander, and Director Alex de Waal. Famine crimes: politics & the disaster relief industry in Africa.
Indiana University Press, 1997.
Dollar, David, and Victoria Levin. "The increasing selectivity of foreign aid, 1984–2003." World development 34,
no. 12 (2006): 2034-2046.
Dreher, Axel, Jan-Egbert Sturm, and James Raymond Vreeland. "Development aid and international politics: Does
membership on the UN Security Council influence World Bank decisions?." Journal of Development Economics 88,
no. 1 (2009): 1-18.
Dreher, Axel, Stephan Klasen, James Raymond Vreeland, and Eric Werker. "The costs of favoritism: is politically
driven aid less effective?." Economic Development and Cultural Change 62, no. 1 (2013): 157-191.
Drury, A. Cooper, Richard Stuart Olson, and Douglas A. Van Belle. "The politics of humanitarian aid: US foreign
disaster assistance, 1964–1995." The Journal of Politics 67, no. 2 (2005): 454-473.
Dryden-Peterson, Sarah. "Barriers to accessing primary education in conflict-affected fragile states." Save the
Children (2009).
Easterly, W. (2007). Was development assistance a mistake?. American Economic Review, 97(2), 328-332.
Easterly, William. "Can foreign aid buy growth?." Journal of economic Perspectives 17, no. 3 (2003):
Easterly, William. “Was Development Assistance a Mistake?” The American Economic Review 97, no. 2 (2007):
328–32.

18
Feghali, Sylvia, Caroline Faria, and Fardowsa Jama. "“Let us create space”: Reclaiming peace and security in a
Kenyan Somali community." Political geography 90 (2021): 102453.
Fellin, Melissa. "The impact of media representations on Somali youth's experiences in educational spaces."
Landscapes of Violence 3, no. 3 (2015): 5.
Feyzioglu, Tarhan, Vinaya Swaroop, and Min Zhu. "A panel data analysis of the fungibility of foreign aid." The
World Bank Economic Review 12, no. 1 (1998): 29-58.
Findley, Michael G, Helen V Milner, and Daniel L Nielson. 2017. “The Choice among Aid Donors: The Effects of
Multilateral vs. Bilateral Aid on Recipient Behavioral Support.” Review of International Organizations 12(2): 307–
34.
Fleck, Robert K., and Christopher Kilby. "How do political changes influence US bilateral aid allocations? Evidence
from panel data." Review of Development Economics 10, no. 2 (2006): 210-223.
Gartzke, Erik, and Megumi Naoi. "Multilateralism and democracy: A dissent regarding Keohane, Macedo, and
Moravcsik." International Organization 65, no. 3 (2011): 589-598.
Goujon, Anne et al. 2016. “A Harmonized Dataset on Global Educational Attainment between 1970 and 2060 - An
Analytical Window into Recent Trends and Future Prospects in Human Capital Development.” Journal of
Demographic Economics 82(3): 315–63.
Government of Kenya, Gachathi Report, 1976
Albright, A., & Bundy, D. A. (2018). The Global Partnership for Education: forging a stronger partnership between
health and education sectors to achieve the Sustainable Development Goals. LANCET CHILD & ADOLESCENT
HEALTH, 2(7), 473-474.
Gregorio, Jose De, and Jong–Wha Lee. "Education and income inequality: new evidence from cross-country data."
Review of income and wealth 48, no. 3 (2002): 395-416.
Haider, Huma. "Conflict analysis of North Eastern Kenya." (2020).
Hancock, Graham. Lords of powerty: the free-wheeling lifestyles, power, prestige and corruption of the multi-billion
dollar aid business. Macmillan, 1989.
Hanushek, Eric A. “Will More Higher Education Improve Economic Growth?” Oxford Review of Economic Policy
32, no. 4 (2016): 538–52.
Heinrich, Tobias, Yoshiharu Kobayashi, and Kristin A. Bryant. "Public opinion and foreign aid cuts in economic
crises." World Development 77 (2016): 66-79.
Heinrich, Tobias. "When is foreign aid selfish, when is it selfless?." The Journal of Politics 75, no. 2 (2013): 422-
435.
Hoeffler, Anke, and Verity Outram. "Need, merit, or self‐interest—what determines the allocation of aid?." Review
of Development Economics 15, no. 2 (2011): 237-250.
Hook, Steven W. "National interest and foreign aid." In National Interest and Foreign Aid. Lynne Rienner
Publishers, 1995.
Jablonski, Ryan S. 2014. “How Aid Targets Votes: The Impact of Electoral Incentives on Foreign Aid Distribution.”
World Politics 66(2): 293–330.
Joly, Jeroen. "Do the media influence foreign aid because or in spite of the bureaucracy? A case study of Belgian aid
determinants." Political Communication 31, no. 4 (2014): 584-603.
Kam, Cindy D., and Carl L. Palmer. "Reconsidering the effects of education on political participation." The Journal
of Politics 70, no. 3 (2008): 612-631.
Keck, Margaret E., and Kathryn Sikkink. Activists beyond borders: Advocacy networks in international politics.
Cornell University Press, 1998.

19
Kenny, L., Hassan, R., Bacchus, L. J., Smith, M., Shell-Duncan, B., Dagadu, N. A., ... & Hossain, M. (2021).
Reproductive health decision making among nomadic pastoralists in North Eastern Kenya: a qualitative social
network analysis. Reproductive Health, 18(1), 1-16.
Kenyan Institute of Curricular Development, 2020
Kenyan National Council of NGOs, 2020
Kilby, Christopher, and Axel Dreher. "The impact of aid on growth revisited: Do donor motives matter?."
Economics Letters 107, no. 3 (2010): 338-340.
Kilby, Christopher. "The political economy of project preparation: An empirical analysis of World Bank projects."
Journal of Development Economics 105 (2013): 211-225.
Krueger, Alan B., and Mikael Lindahl. "Education for growth: Why and for whom?." Journal of economic literature
39, no. 4 (2001): 1101-1136.
Lancaster, Carol. Foreign aid: Diplomacy, development, domestic politics. University of Chicago Press,
Liston, V. "NGOs and spatial dimensions of poverty in Kenya, paper prepared for the 2008 African Studies
Association UK (ASAUK) conference." Preston, Lancashire Sept 11 (2008).
Lochery, Emma. "Rendering difference visible: the Kenyan state and its Somali citizens." African Affairs 111, no.
445 (2012): 615-639.
Lucas, Adrienne M., and Isaac M. Mbiti. "Access, sorting, and achievement: The short-run effects of free primary
education in Kenya." American Economic Journal: Applied Economics 4, no. 4 (2012): 226-53.
Maizels, Alfred, and Machiko K. Nissanke. "Motivations for aid to developing countries." World development 12,
no. 9 (1984): 879-900.
Masaki, Takaaki. 2018. “The Political Economy of Aid Allocation in Africa: Evidence from Zambia.” African
Studies Review 61(1): 55–82.
McGillivray, Mark, and Farhad Noorbakhsh. 2004. “Aid , Conflict and Human Development.” Making Peace
World.
McKinlay, R. D., and R. Little. “A Foreign-Policy Model of the Distribution of British Bilateral Aid, 1960-70.”
British Journal of Political Science 8, no. 3 (1978): 313–31.
Mendenhall, Mary A. 2014. “Education Sustainability in the Relief-Development Transition: Challenges for
International Organizations Working in Countries Affected by Conflict.” International Journal of Educational
Development: 10.
Mesquita, Bruce Bueno de, and Alastair Smith. "Tanzania’s economic and political performance: A district-level
test of selectorate theory." In Politics in South Asia, pp. 31-48. Springer, Cham, 2015.
Milligan, Kevin, Enrico Moretti, and Philip Oreopoulos. "Does education improve citizenship? Evidence from the
United States and the United Kingdom." Journal of public Economics 88, no. 9-10 (2004): 1667-1695.
Morrissey, Oliver. "Conditionality and aid effectiveness re‐evaluated." World economy 27, no. 2 (2004): 153-171.
Mutsune, Emmanuel. Kenyan Somalis: Citizens or subjects. York University
Neumayer, E. (2003). The Pattern of Giving Aid: The Impact of Good Governance on Development Assistance.
Routledge Studies in Development Economics 34. London and New York.
Ngwacho, Areba George. "COVID-19 pandemic impact on Kenyan education sector: Learner challenges and
mitigations." Journal of Research Innovation and Implications in Education 4, no. 2 (2020): 128-139.
Nicolai, Susan, and Sébastien Hine. 2015. Investment for Education in Emergencies: A Review of Evidence.
Nishimura, Mikiko, and Takashi Yamano. "Emerging private education in Africa: Determinants of school choice in
rural Kenya." World Development 43 (2013): 266-275.

20
Nowak Lehmann, Felicitas, Axel Dreher, Dierk Herzer, Stephan Klasen, and Inmaculada Martínez Zarzoso. "Does
foreign aid really raise per capita income? A time series perspective." Canadian Journal of Economics/Revue
canadienne d'économique 45, no. 1 (2012): 288-313.
O’Brien-Udry, C. (2021). Aid, Blame, and Backlash.
Odell, John S. "Case study methods in international political economy." International studies 2, no. 2 (2001): 161-
176.
Ohba, Asayo. Does free secondary education enable the poor to gain access? A study from rural Kenya. Create,
2009.
Oketch M., & Sommerset, A. (2010). Free Primary Education and after in Kenya: enrollment impact, quality effects,
and the transitions to secondary school. CREATE PATHWAYS TO ACCESS. (Research Monograph No. 37).
London: Institute of Education/University of Sussex.
Oketch, Moses, Maurice Mutisya, Moses Ngware, and Alex C. Ezeh. "Why are there proportionately more poor
pupils enrolled in non-state schools in urban Kenya?." International Journal of Educational Development 30, no. 1
(2010): 23-32.
Østby, Gudrun, and Henrik Urdal. 2010. Education and Civil Conflict : A Review of the Quantitative, Empirical
Literature (Background Paper Prepared for the Education for All Global Monitoring Report 2011).
Paxton, Pamela, and Stephen Knack. "Individual and country-level factors affecting support for foreign aid."
International Political Science Review 33, no. 2 (2012): 171-192.
Peterson, Timothy M, and James M Scott. 2018. “The Democracy Aid Calculus: Regimes, Political Opponents, and
the Allocation of US Democracy Assistance, 1981–2009.” International Interactions 44(2): 268–93.
Przeworski, Adam, R. Michael Alvarez, Michael E. Alvarez, Jose Antonio Cheibub, Fernando Limongi, and
Fernando Papaterra
Limongi Neto. Democracy and development: Political institutions and well-being in the world, 1950-1990. No. 3.
Cambridge University Press, 2000.
Psacharopoulos, George, and Maureen Woodhall. Education for development. Oxford: Oxford University Press,
1993.
Psacharopoulos, George. "Returns to investment in education: global update." World development 22, no. 9 (1994):
1325-1343.
Purushothaman, Chithra. "The Evolving Bilateral Aid Architecture." In Emerging Powers, Development
Cooperation and South-South Relations, pp. 29-61. Palgrave Macmillan, Cham, 2021.
Rana, Zunera Ahmad, and Dirk Jan Koch. "What happens to aid fungibility when the recipient government takes
control? Effects of aid ownership in Rwanda." Development Policy Review (2021): e12604.
Rieff, David. "Humanitarianism in crisis." Foreign Affairs (2002): 111-121.
Robinson, Mark. "Does decentralisation improve equity and efficiency in public service delivery provision?."
(2007).
Rodrik, Dani. "Understanding policy reform." In Modern Political Economy and Latin America, pp. 59-70.
Routledge, 2018.
Ruto, Sara Jerop, Zipporah Nyamauncho Ongwenyi, and John Kabutha Mugo. "Educational marginalisation in
northern Kenya." Paper commissioned for the EFA Global Monitoring Report (2010).
Sachs, Jeffrey D. Investing in development: A practical plan to achieve the Millennium Development Goals. CRC
Press, 2005.
Sachs, Jeffrey D. The end of poverty: Economic possibilities for our time. Penguin, 2006.
Sahn, David E., and David C. Stifel. "Urban–rural inequality in living standards in Africa." Journal of African
Economies 12, no. 4 (2003): 564-597.

21
Sandefur, Justin, and Amanda Glassman. The political economy of bad data: evidence from African survey &
administrative studies-working paper 373. No. 373. 2014.
Savun, Burcu, and Daniel C. Tirone. "Foreign aid as a counterterrorism tool: more liberty, less terror?." Journal of
Conflict Resolution 62, no. 8 (2018): 1607-1635.
Schnable, Allison, Anthony DeMattee, Rachel Sullivan Robinson, and Jennifer N. Brass. "International development
buzzwords: Understanding their use among donors, NGOs, and academics." The Journal of Development Studies 57,
no. 1 (2021): 26-44.
Schraeder, Peter J., Steven W. Hook, and Bruce Taylor. "Clarifying the foreign aid puzzle: A comparison of
American, Japanese, French, and Swedish aid flows." World politics 50, no. 2 (1998): 294-323
Seybolt, Taylor B. 2009. “Harmonizing the Humanitarian Aid Network: Adaptive Change in a Complex System.”
International Studies Quarterly 53(4): 1027–50.
Shields, Robin, and Francine Menashy. 2019. “The Network of Bilateral Aid to Education 2005–2015.”
International Journal of Educational Development 64: 74–80.
Sides, John, and Kimberly Gross. "Stereotypes of Muslims and Support for the War on Terror." The Journal of
Politics 75, no. 3 (2013): 583-598.
Simson, Rebecca, and Elliott Green. 2019. “Ethnic Favouritism in Kenyan Education Reconsidered: When a Picture
Is Worth More than a Thousand Regressions International Development: 19–196.
Somerset, Anthony. "Universalising primary education in Kenya: the elusive goal." Comparative education 45,
(2009): 233-250.
Song, Tianyang, Samuel Brazys, and Krishna Chaitanya Vadlamannati. 2021. “Which Wheel Gets the Grease?
Constituent Agency and Sub-National World Bank Aid Allocation.” Journal of Development Studies 57(3): 519–
33.
Stewart, Frances. "Horizontal inequalities in Kenya and the political disturbances of 2008: some implications for aid
policy." Conflict, Security & Development 10, no. 1 (2010): 133-159.
Stone, Randall W. "The scope of IMF conditionality." International organization 62, no. 4 (2008): 589-620.
Swedlund, Haley J. 2017. “Can Foreign Aid Donors Credibly Threaten to Suspend Aid? Evidence from a CrossNational Survey of Donor Officials.” Review of International Political Economy 24(3): 454–96.
Zheng, H., & Li, C. (2022). Can money buy friendship?—Evidence from the US and China’s competition for
influence through foreign aid. The World Economy, 45(10), 3224-3245.
UIS 2019, 2018, 2017, 2022
UNESCO, 2011, 2015, 2018, 2019
Wairuri, Kamau. "Andrew Faull, Police Work and Identity: a South African ethnography. Abingdon: Routledge
(hb£ 115–978 1 138 23329 4). 2018, xxix+ 200 pp." Africa 90, no. 4 (2020): 799-800.
Wajir District Annual Report, 1963, 1965
Walsemann, Katrina M., Gilbert C. Gee, and Annie Ro. "Educational attainment in the context of social inequality:
new directions for research on education and health." American Behavioral Scientist 57, no. 8 (2013): 1082-1104.
Weitzberg, Keren. "The unaccountable census: Colonial enumeration and its implications for the Somali people of
Kenya." The Journal of African History 56, no. 3 (2015): 409-428.

22
ARTICLE 1: APPENDIX A
THE PROJECT-LEVEL EDUCATION AID DATASET
The project-level dataset introduced here includes 1,762 observations with 652 unique
interventions extending educational aid across Kenya. The data range from 2012 to 2021 with
200 completed projects and the remainder continuing in some capacity. The dataset details each
project’s duration, location(s) of operation, target population, and extent of government
collaboration. With less than 10% of projects including information on monetary flows, the
amount allocated were excluded from the data.
The data collection procedure used a snowball sample from an existing network within the aid
community. Then, using online information from organization websites, reports, and publications
additional partners and similar organizations were coded. The initial phase of data collection was
compiled into an organization list and reviewed by an anonymized Kenyan government
representative and contacts from three prominent aid organizations in Kenya. With their
additions, another phase of data collection used inter-coder reliability checks to see whether the
process of snowball sampling was subject to bias. Then, two years after data collection and
preliminary analyses were completed, an additional data collection phase occurred to capture
potential pandemic-related variation in the allocation of aid. In this way, the incorporation of
several phases of data collection, local actor review, and inter-coder reliability checks enabled
the development of a robust dataset.
While these data are primarily used in count form to map the distribution of aid projects aiming
to extend access to education, they also include information on the donors, target population, and
government collaboration. In conjunction with the analyses of the count data and need indicators,
the more detailed data were considered to see if there were patterns in how specific donors
allocated aid. 60% of these projects involve more than one organization, usually with a singular
implementing organization and several supporting donors. While there were no significant
patterns in aid allocation by donor, these details provided interesting insight into the role that
recipient and donor governments play in subnational aid. In general, bilateral aid agencies act as
donors rather than the implementing partners in a project. They also are more likely to allocate
aid to nation-wide projects than other donors. In addition, a quarter of the projects in the data
involve collaboration with the Kenyan government. However, no one donor type nor target
population include systematically more instances of government collaboration.
Education data, especially in developing country contexts, struggle with missingness
(Montejourdes, 2013). The drafted dataset included over 3,000 observations but through careful
review of these data, approximately 600 of these operations were different names or
organizations’ work on the same project and another 400 lacked temporal parameters or detailed
enough information to be included in the final data. The result was 1,700 observations and 652
unique aid projects. The larger number of observations reflects aid projects with multiple
locations or phases of operation with different target populations.

23
ARTICLE 1: APPENDIX B
EDUCATION INDICATORS AND INTERVENTIONS COMPARISON TO UN AND NGO DATA
In table A.1 below, two additional and alternative datasets are used to as robustness checks on
the new project-level dataset to see if the positive relationship between educational need and
education aid holds. While both measures have more missingness and lack sectoral specificity
they provide strong alternatives to the introduced data on education aid. Both of these alternative
datasets were collected and compiled into count data at the county level. The UN data includes
projects with UN organization partnerships, funding, or collaboration and was collected from
UNOCHA dashboards. The NGO data come from the Kenyan government’s registry of NGOs.
Education Indicators and Interventions
Dependent variables: UN and NGO Registry Aid Projects
(1) (2) (3) (4)
Enrollment Rate -89.669*** 77.103
(13.996) (50.013)
Number of Out of School Children 0.050*** 0.120***
(0.015) (0.042)
Population 0.00002**
-0.0001**
(0.00001) (0.00003)
Constant 126.833*** 56.368*** 72.623 100.645***
(14.436) (9.597) (51.588) (25.969)
Observations 47 47 47 47
Log Likelihood -215.965 -226.140 -275.821 -272.925
Akaike Inf. Crit. 435.930 458.280 555.642 551.849
Note: *p<0.1; **p<0.05; ***p<0.01
Table A.1: Regressions on UN and NGO Aid Projects and Education Indicators
This analysis supports the findings in the main text, that in general, aid is allocated to places with
more need. When controlled for population size, both the UN and NGO registry data show that
more out of school children are correlated with more aid projects. The UN data also support the
finding from the education aid analyses that lower enrollment results in more interventions.

24

25
ARTICLE 1: APPENDIX C
SECURITY AND SUPPLY-SIDE CONSTRAINTS IN NORTHEASTERN KENYA
As shown in Figure A.2 there are substantially more security concerns in the Northeastern
province of Kenya as compared to other regions. The supply-side theories and H₃ indicate that
these security concerns may deter aid organizations from working where there is higher cost per
capita to extent access to education.
Figures A.1 & A.2: Pie Charts on the Average Numbers of Terrorist Attacks and Conflicts Per Region 2012-2021
However, Table A.2 finds little support for these supply-side factors. Road coverage, a common
measure of convenience and infrastructure, is not correlated with the presence of aid. Another
measure of convenience and cost of operation for an organization is the level of local corruption.
Yet the average number of bribes is not significant, nor do Wajir and Mandera have higher than
average levels of corruption. The incidence of terrorism, however, is higher in the Northeastern
province. The correlation between the number of NGO projects and terrorist attacks while
significant, simply reflects the amount of terrorist activity in Garissa county. Garissa has
substantially more terrorist attacks than the average county as seen in Figure A.2 and also
substantially more aid projects since it is the location of Dadaab refugee camp. Therefore, there
is not enough evidence to suggest that supply-side factors explain the absence of aid in Wajir and
Mandera.
Supply-Side Indicators
Dependent variables: Number of NGO and UN Projects
(1) (2) (3) (4) (5) (6)
Average Annual Number of Bribes
Requested -0.463 -0.020
(0.358) (0.968)
Road Coverage 0.004 -0.009
(0.003) (0.007)

26
Average Annual Number of Terrorist
Attacks 0.634***
-0.314
(0.233) (0.665)
Population -0.00000 -0.00000 -0.00001 0.00001 0.00001 0.00001
(0.00001) (0.00001) (0.00001) (0.00002) (0.00002) (0.00002)
Constant 50.209*** 28.765*** 38.437*** 146.342*** 160.071*** 143.893***
(12.157) (10.371) (8.058) (32.857) (27.190) (23.021)
Observations 46 46 47 46 46 47
Log Likelihood -225.377 -225.347 -227.522 -271.112 -269.683 -276.862
Akaike Inf. Crit. 456.754 456.694 461.044 548.224 545.366 559.724
Note: *p<0.1; **p<0.05; ***p<0.01
Table A.2: Regressions on UN and NGO Aid Projects and Supply-Side Indicators

27
AID ALLOCATION TO GEO-STRATEGIC STATES:
AID COMPOSITION & DELIVERY IN DJIBOUTI
ABSTRACT
Djibouti has gained global recognition and geo-strategic significance in the twenty-first
century, becoming instrumental in cross-continental trade and a cornerstone for the fight against
extremism and piracy in the region. Optimists would suggest that associated increases in aid
allocation, involvement, and investment by global powers is likely to result in positive
development outcomes. However, Djibouti remains amongst the least developed countries in the
world and, for the seventh consecutive year, human development stagnates. Conventional
wisdom suggests that states like Djibouti with geo-strategic significance receive international aid
and attention not for development, but in exchange for policy concessions. This politicallydriven aid is usually allocated as unconditional bilateral development assistance. Yet, Djibouti
receives more humanitarian aid delivered through multilateral institutions. If Djibouti is indeed a
classic geo-strategic country context where aid is politically-driven, then why would donors
disguise aid as altruistic? Tracing the trajectory of aid, investment, and involvement by global
powers in Djibouti demonstrates competitive behavior between donors. As Djibouti’s autocratic
government recognizes their newly acquired status, they exercise agency and accept aid from
additional donors. If each donor expects to influence the recipient government, then they will
respond to contributions by rival donors with increases in aid, perpetuating a cycle of
competitive additional allocation. Therefore, donors may disguise politically-motivated aid to
obscure their agenda from rivals and avoid these costs of competition. Identifying these new
ways of allocating politically-motivated aid in geo-strategic states is important in order for
altruistic actors to be able to identify politically-driven vs. need-driven aid, and reach
populations in need.
INTRODUCTION
Djibouti, a small state off the Horn of Africa, has gained global recognition and
geostrategic significance in the twenty-first century. It has become instrumental in crosscontinental trade and a cornerstone for the fight against extremism and piracy in the region. With
significant aid allocation, involvement, and investment by global powers in recent years,
journalists, scholars, and policy-makers had suggested that Djibouti would be the Dubai or
Singapore of Africa (Dawaleh, 2015; Global Times, 2015; Gauche, 2016). Yet, this potential has
not been realized, and for the seventh consecutive year, human development stagnated in 2023.
Djibouti continues to be ranked amongst the least developed countries in the world with high
levels of inequality and unemployment, low levels of education, and extreme poverty. Why?
The strategic goals of donor governments, rather than the needs of recipient populations,
are widely considered to be the primary determinants of aid allocation globally. Aid can be a
means to project power (Van Der Veen, 2011) or enact foreign policy (Kilby & Dreher, 2010).
Donor governments give more aid to countries with potential for favorable economic and
security gains (Apodaca, 2017). They also allocate aid for specific foreign policy outcomes
including democratization (Bermeo, 2011), regional stabilization (Mosse, 2005), counterterrorism (Bouton & Carter, 2014), or to access particular markets (Chenery, 1965; Papanek,
1972), resources (Kapfer et al., 2007), and trade (Berthelmy & Tichit, 2004). While the policy
concessions prioritized differ depending on the donor-recipient dyad (Koch et al., 2017),
scholarly consensus suggests that recipient need does not drive aid allocation.

28
With its important role in regional economy and security, Djibouti has emerged as a
quintessential example of a geo-strategic recipient state. Therefore, if aid is politically-driven and
allocated strategically to Djibouti, it should reflect the foreign policy agendas of donors and be
delivered as bilateral development aid, without constraints. Such unrestricted development aid
given by one government to another provides the most value to the recipient government who is
being bribed, at the lowest cost to the donor. Yet, data from the OECD reporting of Official
Development Assistance and UNOCHA’s Financial Tracking Service demonstrate the opposite.
Aid is allocated to Djibouti in ways that usually signal need-driven rather than politically-driven
aid: through multilateral institutions as humanitarian assistance. Increases in humanitarian
assistance though has not resulted in a decrease in intensity of need, suggesting that these
allocations are not actually altruistic. If Djibouti is indeed a classic geo-strategic country context
where aid is politically-driven, then why would donors try to disguise aid as altruistic?
The unusual ways in which politically motivated aid is allocated to Djibouti may be
directed by the activities of other donors. Especially in the context of geo-strategic countries,
donors have been found to increase the amount of aid they allocate in response to actions of
other donors (Davies & Klasen, 2019). Therefore, as more donors enter Djibouti’s aid ‘market’
hoping to acquire policy concessions, they may have incentive to disguise politically-driven aid
in order to obscure their agenda from competitors. I expect that donor countries wish to provide
aid in exchange for concessions from Djibouti’s government without sparking competing
contributions from their geopolitical rivals. Recent studies suggest that humanitarian aid is not
immune to political influence (Gartzke & Naoi, 2011) and that decisions by multilateral
institutions can still be driven by bilateral donors (Eichenauer & Knack, 2015). Therefore,
though delivered in ways that appear to be altruistic, donors may still use humanitarian aid to
achieve policy goals.
Just as donors have incentive to compete for influence, recipient governments have
incentive to attract donors and aid. These interactions are rapidly evolving in the current geopolitical environment. With the emergence of new and more “non-traditional” donors, increased
competition will likely continue to affect the aid landscape (Calleja & Rowlands, 2015).
Distinguishing politically-driven versus need-driven aid in this new environment is thus essential
for altruistic actors to be able to accurately assess and access needy populations.
I first outline the expectations for how donors behave in the context of geo-strategic
states. Then, I trace the evolution of international interest, involvement, and investment in
Djibouti with special attention paid to the associated aid allocations by competing global powers.
Next, I analyze singular contributions from different donors to demonstrate how the introduction
of new actors increases aid allocated by existing donors. Finally, I conclude by investigating the
different incentives which donors have to disguise their aid allocations and associated policy
agendas from other donors.
GEOSTRATEGIC AID ALLOCATION
The debate surrounding the determinants of aid suggests that aid is either need-driven or
politically-driven. Altruistic aid is need-driven and allocated where it is most needed: to
populations experiencing crisis or poverty. Geo-strategic aid is politically-driven and allocated to
achieve foreign policy goals or to gain power in international politics. Substantial scholarship
supports politically-driven theory, showing how aid is rarely allocated altruistically, and instead
reflects the interests of aid suppliers (McKinley & Little; Shraeder, Hook, and Taylor; Bueno de
Mesquita & Smith, 2009). However, the determinants of these contributions may differ
depending on the donor (Berthélemy, 2003, 2005) and recipient (McGillivray, 2005).

29
In the literature, divergences in donor motivation have thus far been distinguished through
the type, delivery, composition, and effectiveness of aid to different recipients. While
classifications of aid are many, here, two types of aid are considered: humanitarian aid and
development aid. Humanitarian aid aims to provide immediate, impartial, and life-saving relief
to recipient populations. It usually is allocated in response to crises including conflict or natural
disaster. It is thus generally shorter-term and includes in-kind assistance with the provision of
life-saving goods or services directly to the population. In idea, humanitarian aid follows three
principles set forth by several international laws and agreements: impartiality, neutrality, and
independence (Geneva Conventions, 1948). Humanitarian assistance, therefore, does not require
cooperation with the government and is often conducted through multilateral institutions and
nongovernmental organizations. This makes the commitment and delivery of humanitarian aid
less susceptible to misappropriation and more likely to reach needy populations (Apodaca,
2017).
Alternatively, development aid aims to eradicate poverty and improve economic conditions
through systemic support. It usually involves longer-term projects and cooperation with the
government in order to build state capacity and ensure sustainable improvements. In practice,
this type of aid is often criticized for its fungibility, as large unrestricted funds can be more easily
misappropriated by corrupt recipient governments. Such politically-driven development aid is
often used by self-interested donors to buy the favor of or influence over the recipient
government without expectation of positive development outcomes amongst the population
(Bueno de Mesquita & Smith, 2016; Swedlund, 2016). In this way, if a donor is politically
motivated, then they would be expected to give more development aid to secure policy
concessions than humanitarian aid (Headey, 2008).
H₁: Djibouti will receive more development aid than humanitarian aid
One would also expect for politically-driven aid to be distributed as bilateral assistance
from donor governments rather than delivered through multilateral institutions. If aid were needdriven and focused on attaining positive recipient outcomes, donors would defer to multilateral
institutions for aid delivery because multilateral aid has been found to be more selective, more
targeted, less strategic, and less fungible than bilateral aid (Bueno de Mesquita & Smith, 2009,
2007; Dollar & Levin, 2006; Burnside & Dollar, 2000; Neumeyer, 2003; Maizels and Nissanke
1984; Martens et al. 2002; Rodrik 1996). In addition, multilateral delivery of aid can be directed
by the institution rather than by the bilateral donors’ contribution. This can obscure the intent of
donors and diffuse the impact on the recipient government. Politically motivated donors,
therefore, prefer to give aid bilaterally to maintain control over both the aid and the recipient
government (Riddell, 2014; Apodaca, 2017; Dietrich, 2013).
H₂: Djibouti will receive more bilateral aid than multilateral aid
Similar to multilateral delivery, targeted aid attempts to avoid misappropriation and
diversion of resources by directly allocating aid to specific sectors, locations, and populations in
need. Need-driven donors concerned with aid effectiveness or potential misappropriation by
recipient governments would therefore, use more conditional and directly managed aid (Knack &
Rahman, 2007; Winters, 2009). Whereas, politically-driven aid would be less constrained, have
less oversight, and less concern regarding recipient outcomes (Annen & Moers 2016; Bigsten &
Tengstam 2015; Bourguignon & Platteau 2015; Frot & Santiso 2009; Easterly & Williamson
2011; Fuchs et al. 2015; Knack & Rahman 2007; Acharya et al. 2006).
H₃: Djibouti will receive more unconstrained aid than targeted aid
For politically-driven donors, the success of strategic aid should be measured through
whether policy concessions are able to be achieved, rather than the effectiveness of aid. In fact,
even though aid has been found to stimulate economic growth in countries with strong

30
institutions and limited corruption (Burnside & Dollar, 2000, 2004), politically-driven donors do
not appear to reward good governance (Alesina & Weder, 2002; Easterly & Pfutze, 2008). Some
studies have even found that aid flows are actually positively correlated with corruption (Alesina
& Weder, 2002; Easterly & Pfutze, 2008; Easterly & Williamson, 2011; Isopi & Mattesini,
2008). If aid is politically-driven in a country susceptible to grand corruption, such as Djibouti,
then aid should not be expected to stimulate economic development.
H₄: Djibouti’s development trajectory is not correlated with the amount of aid received
The large body of literature on the determinants of aid reviewed above provide clear
characteristics of need-driven and politically-driven aid. Aid driven by need will focus on
recipient populations and is more likely to be humanitarian, multilateral, targeted, and effective.
Politically-driven aid is more likely to be focused on cooperation with the recipient government
to achieve foreign policy goals, and is more likely to be bilateral and allocated as broad,
unconstrained, development assistance. In this way, by analyzing the delivery, composition, and
effectiveness of aid, donor motivations should be identifiable. Distinction of donor motivation is
essential if altruistic actors (whether they be NGOs, multilateral institutions, or the rare dogooder donor government) are to reach the places and populations that need the most assistance.
If aid is easily identified as politically-driven, altruistic actors can complement allocations with
more effective measures that will reach needy recipients.
INTERNATIONAL INTEREST & INVOLVEMENT IN DJIBOUTI
Djibouti can be identified as a most-likely case for politically-driven aid because of its
location, domestic governance, and donor network. Djibouti is ranked in the bottom fifty
countries of the Corruption Perceptions Index and Global Freedom Scores. The president is
serving his fifth term, after repressed elections which declared his winning with 97% of the vote.
President Guellé not only runs a restrictive regime, but is also deeply involved in foreign policy
decisions, often personally handling negotiations for trade agreements, foreign military presence,
and foreign direct investment personally (Freedom House, 2023). While some donors are
disincentivized by poor governance (Bermeo, 2010; Dietrich, 2013), politically-motivated donors
are more likely to allocate aid where aid can be used in exchange for policy concessions
(Easterly & Williamson, 2011). President Guellé’s personalist autocracy thus creates ideal
governance conditions for politically-driven aid.
Djibouti’s geostrategic location in the Horn of Africa is an increasingly pivotal place for
economic exchange and securitization. Despite being a small country with only 9,000 sq/mi and
approximately 900,000 citizens, its location at the intersection of the Red Sea and Gulf of Aden
have given the government unusual agency in international trade (Cabestan, 2020). Trade has
been found as a common politically-driven determinant of aid allocation, whereby donors will
allocate aid in tandem with, or to incentivize, favorable trade treaties (Bethélemy, 2006;
Morrissey, 1993; Ali, Banks, & Parsons, 2015) and nearly 10% of global trade transits around
Djibouti’s coast. The small country has become a hub for oil export out of the Middle East and
the primary port from which all Ethiopia’s sea-borne exports travel. With such a strategic
location for cross-continental trade, Djibouti would be an optimal place for donors to give aid in
exchange for economic gain.

31
Figure 1: Map of Djibouti’s Geo-Strategic Location
Djibouti is also surrounded by countries in conflicts whose outcomes affect global power
dynamics in addition to regional stability. To the east, a 10 year civil war in Yemen has attracted
both aid and military intervention from Iran, Russia, the United States, Saudi Arabia and its
coalition of Gulf countries. To the south, Somalia continues to struggle with conflict, state
failure, and a terrorist insurgency that threatens nearby neighbors. Djibouti also connects to the
contentious border between Ethiopia and Eritrea, where a tenuous relationship has been made
more precarious with the outbreak of renewed fighting in Ethiopia’s Tigray region. As a result of
this important position in the promotion of regional stability, Djibouti has the highest
concentration of foreign military bases in the world. The United States, United Kingdom,
Germany, Spain, France, Saudi Arabia, United Arab Emirates, Japan, and China all maintain
troops or bases in Djibouti. In terms of security, its geostrategic location has made Djibouti a
cornerstone in regional stability, giving donors incentive to allocate aid as an act of goodwill to
facilitate foreign military presence (Kisangani & Pickering, 2015), as a means to influence
Djibouti as an ally (Apodaca, 2017), and as a way to combat terrorism (Bouton & Carter, 2014).
Politically-Driven Donors
We expect donors to be more likely to work within these conditions, where aid can be used to
indicate, influence, and implement foreign policy. However, prior to Djibouti’s ascension as an
important geo-strategic state, the majority of donors appeared to be acting altruistically. With a
few exceptions, these donors were also those traditionally classified as need-driven such as the
Scandinavian countries (Sweden, Norway, Denmark), Canada, Switzerland, and the Netherlands
(Berthèlemy, 2003; McGillivray, 2008; Ali, Banks, & Parsons, 2015). Their contributions were
small, amounting on average to $300,000 and allocated mostly during and after the civil war. At
the turn of the century, donors who have been classified as more politically-motivated like the
United States, France, Japan, and Italy increased their aid allocations and continued to do so,
especially when rival donors entered Djibouti.
The same geo-strategic factors which make Djibouti a favorable place to allocate aid also
incentivize donor competition. As new donors enter a recipient country, they compete for favors
from the domestic government (Powell & Bobba, 2006, 8). Thus, new donors are often
associated with increases in the amount of aid allocated to a recipient government (Davies &
Klasen, 2019). The recipient government therefore has more agency than they otherwise would
and are incentivized to continue to play donors off one another (Dunning, 2004). In response to
recipient-generated donor competition, donors could resist by reducing aid or coordinating with
other donors. If donors resist though, they risk losing their own policy concessions and also

32
increasing the influence of rival donors (Kim & Menninga, 2020). Therefore, donors have
incentive to compete rather than cooperate.
Figure 2: The relationship between geo-strategic value and aid
Tracing the trajectory of increases in geo-strategic value, aid allocation, and donor
competition over time reveals the strategies which donors employ to protect their interests and
influence in Djibouti. As the geo-political landscape evolves, so too do the motivations of aid
allocation for bilateral donors. As illustrated by Figure 2, increases in geo-strategic value in
Djibouti increase the amount of aid, the agency of the recipient government, and the number of
donors allocating aid. As more attention and interest in Djibouti is generated amongst the
international community, the government gains more agency and is able to increase the cost of
working within the country through negotiating aid alongside business deals or policy
concessions. Therefore, the donors which enter amidst increases in geo-strategic value are more
likely to have politically-motivated agendas. New donors thus pose a threat to existing
politically-motivated donors who are already using aid to exert influence in Djibouti. In
response, those existing actors choose to compete and increase their aid allocations, thereby
increasing the overall volume of aid. In this way, Djibouti’s significant geo-strategic value and
the prominence of a government willing to provide policy concessions create probable conditions
for both politically-driven aid allocation and donor competition.
Increases in International Involvement
Throughout the twentieth century, Djibouti was largely ignored by the international
community (Sun & Zoubir, 2016). Aside from the port, France (Djibouti’s former colonizer) and
Italy (regional colonizer) were the only foreign actors present. They were also the primary
donors in the 1980s and 1990s, complemented by occasional small contributions from other
donors throughout the civil war period. Aid could be identified as need-driven, delivered through
multilateral channels, targeted to conflict-affected areas, and resulted in positive outcomes for
the population. Despite high intensity of need, from 1980 to 2001, Official Development
Assistance (ODA) in Djibouti averaged $100,000 reaching a peak of only $500,000 at the end of
its 8 year civil war. The comparative absence of aid in this period can be largely attributed to
Cold War politics which dominated the geo-strategic aid nexus and created competition between
donors focused on larger geo-strategic states (Lee, 2022).
Djibouti became an important geo-strategic ally in 2002. The US’s Global War on Terror
made the Gulf of Aden an essential location for securitization against extremism on the Arabian
Peninsula and Horn of Africa. Within months of 9/11, the United States expanded an existing
French military base, creating several security coalitions. The allocation of American aid
followed the establishment of the base to the extent that between late 2001 and 2003, ODA
increased by 1,000%. In this way, politically-driven aid allocation was thus used by the United
States to facilitate foreign military presence (Kisangani & Pickering, 2015). After the U.S.

33
negotiated permanent military presence with complementary aid allocation, President Guellé
publicly pledged Djibouti’s support for the Global War on Terror and deemed the country a
location for global cooperation and securitization. In response, new donors like Germany
entered Djibouti to protect their shipping lanes, allocating aid as a means to gain access for their
troops. As the threat of terrorism spread in neighboring Somalia, Yemen, Sudan, and Egypt,
more international powers became interested in establishing troop presence in the region and
began negotiating with President Guellé. As Djibouti solidified its geo-strategic status at the turn
of the century, aid was thus directed by the recipient regime and allocated according to the
politically-driven expectations established in H₁, H₂, H₃, and H₄ as unconstrained, bilateral,
development aid which did not improve living conditions for the populations.
Figure 3: Timeline of ODA and major events in Djibouti
Substantial increases in piracy19 activities in 2008 also increased Djibouti’s geostrategic
value. With Eritrean and Djiboutian troops distracted by a contested coastal border and little
oversight along the Somali coast, incidents of hijacking and ransoming became commonplace,
interfering with international shipping lanes. In consideration of the significant volume of global
trade passing through the affected waters, UN member states appealed to NATO and global
powers for the protection of high seas commerce. The UN sanctions in place against Eritrea, and
Somalia’s status as a failed state made Djibouti the only feasible option as a base for anti-piracy
operations. President Guellé consequently was thus able to leverage economic interests of global
powers to acquire increased aid and investment in exchange for access to Djibouti’s coast and
ports.
As unprecedented amounts of money flowed in, Guellé continued negotiating the opening of
the airport, ports, and capital city as the center of counter-piracy operations. He increased the
cost of leases, enabling the United States, France, and Italy to expand existing bases that also
hosted NATO, Spanish, German, and EU naval troops. For the first time since the Second World
19 Though piracy lacks a cohesive academic definition, acts of piracy according to international maritime law
include: hijacking of vessels, kidnapping of crew, ransoming of persons, or other criminal behaviors committed at
sea (UNCLOS, 1982).

34
War, Japan established a foreign overseas base in Djibouti. Guellé also welcomed small counterpiracy interventions by Chinese naval troops, which were run out of a domestic port.
Significantly, with the increase in interest in Djibouti, the composition of aid shifted. While the
development of infrastructure was paid either directly by international actors or through
investment funds, aid allocations increased both through bilateral and multilateral channels. Yet
even with the increase in international attention and aid, the small population remained in
poverty and without access to quality public service provision.
In 2011, the number of donors and the amount of aid increased again. Politically-driven
theory would suggest that this influx was due to either an increase in incidence of piracy activity
or the entrance of China as a more engaged actor in Djibouti. Studies have shown that, especially
in the case of China, competition for influence using aid (Kilama, 2018), foreign investment
(Lake, 2018), trade flows (Kim, 2019), and foreign troop presence (Joyce & Blankenship, 2023),
increase with the entrance of rival power into a recipient country. In Djibouti, a new expansion
by China in terms of investment and trade resulted in an increase of aid the following year by
both the United States and Japan. This was again exemplified in 2014 when China allocated
significant investment alongside the acquisition of their own port. As a result, overall aid
increased to Djibouti, especially from rival donors.
While most recent aggregated aid data have yet to be released, early reports from 2024
suggest that negotiations with India in the past year for their own military base have similarly
resulted in significant increases in aid to Djibouti. Additionally, discussions between Russia and
Djibouti have generated new aid arrangements from rival donors. In this way, aid appears to not
only reflect the agency of the recipient government taking advantage of geo-strategic value to
acquire monetary concessions in exchange for access, but increases in aid also appear to reflect
competition between donors.
Policy concessions, recipient government agency over aid, and donor competition suggest
classic politically-driven aid allocation. However, the evolution of the composition and delivery
of aid to Djibouti over time, calls into question the scholarly consensus on how politically-driven
aid is allocated to geo-strategic states. Prior to 2001, the majority of aid was humanitarian and
allocated by mostly altruistic donors suggesting need-driven delivery. Then, aid became
politically-driven, with politically-motivated donors negotiating bilateral aid in exchange for
access to the country. Yet, as the geo-strategic status of Djibouti increased, the delivery and
composition of aid changed. The entry of rival donors surprisingly resulted in increases in
humanitarian contributions delivered through multilateral institutions. If aid to this geo-strategic
state is increasingly politically-driven, then why would the composition and delivery of aid
appear altruistic?
THE COMPOSITION & DELIVERY OF AID
Tracing the trajectory of geo-strategic value, international investment, and aid over time
would suggest that aid allocation to Djibouti is indeed politically-driven rather than associated
with the needs of the population. in an attempt to test these relationships between aid allocation,
geo-strategic value, and recipient need, data are extracted from the World Bank and OECD on
Official Development Assistance (ODA) and financial flows. The amount of aid measures are
taken directly from World Bank data and the number of donors are calculated from
disaggregated contribution-level OECD estimates. Geo-strategic value indicators include a
measure of trade openness from the World Bank’s World Integrated Trade Solutions dataset,
dyadic trade data from the Correlates of War dataset, as well as two dummy variables on the
opening of a new foreign port and new foreign military presence. Each of these are correlated

35
with an increase in logged ODA the following year, supporting politically-driven theories of aid
allocation. Development indicators for Djibouti’s population include indices on health,
education, and human development. Counter to H₄ and politically-driven intuition, aid also has a
positive effect on these development outcomes the year after allocation. These counter-intuitive
and somewhat conflicting results are available in Appendix A.
If aid was politically-driven, then the observed relationship with increases in geostrategic
value predicting increases in aid would be expected. However, politically-driven theory would
also predict no relationship between increased aid and development outcomes. Yet, analyses
show both geo-strategic value predicting aid and aid explaining development outcomes. The
presence of both relationships contradicts need-driven and politically-driven theories of aid
allocation suggesting that quantitative analyses alone do not explain how aid is allocated in
Djibouti.
Qualitative analyses which disaggregate the composition and delivery of aid can provide
additional detail and insight into whether aid is need-driven or politically-driven and expose how
aid allocation has evolved. The type, target, and delivery of aid appear to be altruistic. Figure 4
illustrates how the majority of aid given by donor governments in the twenty-first century was
humanitarian and delivered through multilateral institutions, negating both H₁ and H₂. In addition
and counter to H₃, aid allocations were also almost entirely targeted, not only sectorally-specific
but also earmarked for particular projects.
Figure 4: Distribution of Type, Sector, and mode of Delivery of Aid from OECD Data
Targeted humanitarian aid, delivered through multilateral institutions, traditionally indicates
that donor governments are more interested in recipient population outcomes than recipient
government concessions (Apodaca, 2017). Observing this in Djibouti is surprising given the
context of a classic geo-strategic country case, where international investment and aid are
negotiated by an autocratic president taking advantage of increases in geo-strategic value. On
one hand, looking at the effect of politically-driven indicators on ODA and tracing the geostrategic trajectory of Djibouti over time suggests that aid allocation is politically-driven. On the
other hand, looking in isolation at the analysis of aid data and need-driven indicators suggests
that contrary to existing scholarship and the present hypotheses, aid is allocated altruistically in
Djibouti. When these results are taken together, what can the behavior of donors in Djibouti
explain about how aid is allocated to geo-strategic states?
First, while quantitative analyses and supporting scholarship suggest that aid to Djibouti is
need-driven, the small population still lacks access to basic public services and remains
impoverished. Aid is most effective when it is sectorally specific and delivered through
specialized multilateral institutions (Easterly & Pfutze, 2008; Martens et al., 2002) as is seen
with the delivery of aid to Djibouti. Low levels of aid allocation also hinder development, while
large increases in the amount of aid given can reduce poverty (Stiglitz, 2002; Sachs, 2009).

36
Therefore, according to aid effectiveness literature, aid to Djibouti should be effective in
mitigating crises and combating poverty.
Aid effectiveness can, however, also be impacted by a lack of infrastructure, unwelcoming
government, or by having a large, dispersed, rural population (Lopez et al, 2020). Djibouti has
none of these unfavorable conditions for aid allocation. In fact, its small, easily accessible, urban
population should benefit from ease of access to aid. With optimal delivery by donors and
favorable conditions for the allocation of aid, it would be expected that Djibouti’s development
would improve alongside increases in aid and negate H₄. However, sectoral reports and current
development statistics do not align with the aforementioned results suggesting a positive
relationship between aid allocation and human development outcomes. For example, with annual
aid contributions to the education sector from 2011 to 2019, goals for improving primary
enrollment should have been attainable. Similar small countries receiving slightly less aid to the
sector in the same period resulted in significant improvements: Rwanda saw a 20% increase in
enrollment and Burundi experienced a 14% increase while Djibouti’s enrollment only improved
by 5% (UIS, 2023).
After initial improvements with the end of the civil war and inflow of aid in 2002, most
humanitarian and development outcomes have either stagnated or deteriorated. Despite the
amount of aid being correlated with need-driven indicators and favorable delivery by donors, the
population remains impoverished and ill-equipped to respond to or recover from crises.
Infrastructure has consistently improved according to World Bank figures (likely a necessity for
foreign military and port presence), but three quarters of the population remain in severe poverty,
30% of children and youth are out of school, and a fifth of the country are acutely food insecure
(WFP, 2023). With only one million citizens, and an average of $160 million dollars of aid
coming in each year since 1989, Djibouti’s population should not still be struggling to this extent
if aid was need-driven.
It is possible that these development outcomes can be explained by different donors having
different motives, the role which the recipient government plays, and the evolution of aid
allocation. The contributions from classically altruistic actors like the Scandinavian countries
remain small with increases in aid reflecting politically-motivated donors like the United States,
France, Japan, Italy, and China. Therefore, the amount of humanitarian aid that is need-driven is
likely significantly smaller than the inflow of funding would suggest. Perhaps then, humanitarian
aid allocated through multilateral institutions is still politically-driven. A line of literature
running counter to H₁ and H₂ support this idea, showing how multilateral institutions can be
influenced by powerful donors who direct humanitarian aid according to their own interests
(Gartzke & Naoi, 2011). This appears to be the case in Djibouti, where the vast majority of
multilateral aid consists of contributions by politically-motivated bilateral donors. In addition,
some scholars have suggested that donors may allocate humanitarian aid in similar ways to
strategic development aid, as part of a package to gain policy concessions or as a means to
mitigate crises and protect economic or security interests within the recipient country (Bayer &
Rupert, 2004; Blanton & Blanton, 2007; Fink & Radaelli, 2011; Narang, 2016; ). Therefore, if
humanitarian aid delivered through multilateral institutions can be influenced by the agendas of
politically-motivated donors, then aid to Djibouti is likely still politically-driven.
Finally, need-driven theories show the importance of good-governance in providing relief to
populations experiencing crises or poverty (Levin & Dollar, 2004; Easterly, 2007; Kalyvitis and
Vlachaki, 2010; Dietrich & Wright, 2015). Yet, between 2010 and 2011, immediately preceding
the largest increase in aid, Djibouti’s government systematically broke down democratic
institutions and violated civil liberties to secure President Guellé’s political power. In other
countries, human rights violations by a recipient government or declines in democracy have

37
resulted in decreases in aid20 (Payaslian,1996; Lebovic & Voeten, 2009). Yet in Djibouti, the
neglect of the population by a restrictive autocratic regime is wholly ignored without decreases
in the number of donors present nor the amount of aid allocated. This aligns with both
politically-driven theory and scholarly skepticism regarding the role that human rights and
governance play in donor decisions (Neilson, 2013). Therefore, despite the data on delivery and
composition of aid appearing to be need-driven, there is compelling evidence demonstrating how
aid allocation to Djibouti can still be politically-driven.
If donors are only investing in Djibouti to elicit policy concessions and protect their
interests, then why would they bother with humanitarian aid or multilateral delivery?
DISGUISING POLITICALLY-DRIVEN AID
Donors may desire to disguise politically-driven aid in order to be perceived as altruistic
and/or to obscure their agenda from other donors. In the past, donor governments have used aid
to create façades of altruism for themselves amongst the international community, amongst the
recipient country, or amongst their own population as a means to achieve political outcomes.
First, in the international community, aid has been used as a form of soft power (Piknerová,
2014; Kim, 2019; Krzymowski, 2022) where the ability to allocate aid altruistically can signal
wealth and power. For Gulf states, the projection of altruistic aid has also been noted as a way to
counter accusations of the support of terrorist insurgencies after 9/11(Momani and Ennis, 2012).
For non-DAC donors, the allocation of altruistic aid has symbolized attaining status as an
influential international actor and global power (Sen, 2018).
Second, the perception of aid as altruistic amongst recipient populations has also been found
to create affinity for a donor and their ideology (Nye, 2004; Dietrich, Mahmud & Winters, 2017;
Blair, Marty, and Roessler, 2022). Third, donor states may also maintain a façade of altruistic aid
to curry the favor of their own populations. In democratic donor states, research shows how
public perceptions of recipient need are necessary to justify aid (Nielsen ,2013; Dietrich and
Murdie, 2017; Allendoerfer, 2015) or are used to elicit votes prior to elections (Annen &
Strickland, 2017).
Finally, donors may disguise politically-driven aid as altruistic in order to gain more access
to a geo-strategic country without alarming rival donors. Politically-driven theories contend that
aid can not only be used to attain policy concessions in security and economy, but also to project
power and international influence (Lancaster, 2000, 81). Therefore, when rival actors allocate aid
to a new recipient country or increase aid to a recipient country, existing donors often increase
the amount of aid allocated (Davies & Klassen, 2011; Kilama, 2016; Frot & Santiso, 2011;
Steinwand, 2015; Fuchs et al., 2015; Gehring et al., 2017). The associated increasing cost of
donor competition can thus incentivize the disguising of politically-driven aid. If aid from a
donor appears to be allocated altruistically, for example through multilateral institutions as
humanitarian assistance, then perhaps rival donors will not respond. If rivals do not react to
altruistic-appearing increases in aid and the recipient government recognizes the contribution,
then the donor can attain additional influence without added costs of competition.
In the case of Djibouti, donor competition provides a compelling explanation for the nontraditional delivery and composition of politically-driven aid. As illustrated by Figure 5, several
global powers, some of which are in direct competition with one another, have leased land from
Djibouti for foreign ports and military bases. These leases make up a significant proportion of
20 Even traditionally politically-motivated aid allocators have given ‘aid sanctions’ in response to recipient human
rights violations. For example, the United States has withheld aid from Sudan (1988) and Myanmar (1990, 2018).

38
Djibouti’s income. The small state hosts the largest non-domestic military bases for France,
Italy, and Japan with other foreign bases for the United States, Saudi Arabia, and China as well
as negotiations for new bases for Russia and India. Troops from the United Kingdom, Germany,
Spain, and India are also present. Djibouti’s ports work with over a dozen countries, including
Brazil, the United Arab Emirates, Ethiopia.
Figure 5: Map of Djibouti’s Foreign Presence
In looking at the evolution of relationships between Djibouti and politically-motivated
donors, it is clear that aid is allocated as part of a package enabling access to the geo-strategic
state. Especially amongst global powers competing for influence, President Guellé leverages
dialogues and diplomatic relations with rivals to elicit more monetary support in the form of
lease payments and aid allocations. Such negotiations to gain influence and power by both the
donor and recipient governments are notoriously difficult to demonstrate empirically due to an
intentional lack of data (Trunkos, 2020). Therefore, it is easiest to illustrate the role that recipient
government agency and aid play by looking at the development of relationships between
Djibouti and donors individually.
China first negotiated military presence as part of counter-piracy operations in 2008 through
a logistics support base in Djibouti’s port. This coincided with the first significant inflow of
Chinese aid or investment into Djibouti.21 In discussing the development of a permanent Chinese
military presence, Chinese government officials noted Djibouti as “a strategic strong point” for
China’s regional interests (Kennedy, 2019). In 2015, the People’s Liberation Army Support Base
opened with an annual $20 million lease for the land accompanied by a $14 billion pledge for
“investment”- clear evidence of aid for access. At the same time Guellé negotiated the Chinese
deal, he took advantage of the Sino-American rivalry to renegotiate the United States’ lease of
their expansive base, increasing the cost from $30 million to $63 million and requiring additional
‘investment in development’ (NYT, 2014).
Another rival of China, Japan, has chosen to strategically cooperate with other powers
present within Djibouti. Japan’s military base shares airport access with American, French, and
Italian bases and even has allowed Indian troop presence as they negotiate their own base (Island
Society, 2017). This cooperation in an arena of great power competition has been attributed to
Japan’s interest in using Djibouti as a means to promote Japanese influence and counter Chinese
influence in the region (The Trumpet, 2018; Nobuhiro Kubo-Reuters, 2016). China has accused
21 Data on Chinese monetary support do not distinguish between investment and aid.

39
Japan of interfering with Chinese land exercises and the United States of interfering with
Chinese air operations and (SIPRI, 2019). As tensions within Djibouti increase between these
rivals, other countries have tried to expand their own influence.
While sharing air facilities, Italy has made clear that they operate separately from other
Western powers and EU presence. Italian involvement in Djibouti has been consistent since the
1980s, a legacy colonization in the Horn of Africa. Similar to Italy’s former colonies of Eritrea,
Somalia, and Ethiopia, postcolonial engagement in Djibouti was primarily economic with
consistent but limited aid allocation and trade relations (Brioni & Gulema, 2018). A shift in
Italian investment in Djibouti coincided with the establishment of the Italian military base
Amedeo Guillet in “an area destined to be more important and strategic than Suez and Gibraltar”
according to the inaugural speech given by Admiral Binelli Mantelli in 2012. While Prime
Minister Mario Monti negotiated with President Guellé in the year prior to Amedeo Guillet being
built, aid from Italy to Djibouti increased nearly 900% from $470,000 to $4.6 million dollars. In
his speech, the Admiral also highlighted how Djibouti has been used as a forum for great power
competition, stating that Italian “presence here is of fundamental importance” as “it is no
coincidence, moreover, that many countries, from Japan to France, to the United States, have
installed their military bases here. Now we are here too. And we will be there for many years," .
If Italy wants to maintain favor with Djibouti’s government without incurring costs of donor
competition, then perhaps their allocation of primarily humanitarian aid since 2016 is an attempt
to disguise politically-driven aid as altruistic.
The United States and France have similarly disguised aid as altruistic, accompanying large
lease payments for their military bases with annual allocations of humanitarian aid. It appears
that in years when new international actors like China (2010), India (2018), and most recently,
Russia (2023) begin negotiations with the government of Djibouti, the amount of aid allocated
annually by rivals like the United States, Japan, and France increase. Here, the role of the
recipient government is particularly interesting, as aid appears to accompany existing annual
costs of ensuring economic and security concessions in Djibouti. Recipient governments
exploiting donor competition to elicit more aid has been noted across Africa (Reisen and Stijns,
2011; Dreher, Fuchs, & Nunnenkamp, 2013). However, in this context, the growing geo-strategic
value of Djibouti has given the autocracy a unique amount of agency. Therefore, though the
balance is precarious 22, increased donor competition seems to increase the price of aid-forpolicy concessions. Therefore, in Djibouti donors not only increase aid in reaction to rival
donors, but also as a requisite for recipient government favor.
CONCLUSION
The intensity of international attention, involvement, investment, and aid in Djibouti
continues to increase as economic and security interests in the Gulf remain threatened with everevolving crises in the region. The geographic location of Djibouti and willingness of the
autocratic government to use aid for policy has made the country a most likely case for geostrategic aid. As Djibouti’s geo-strategic value increases, so too does the amount of aid allocated,
the number of donors, and the agency of the recipient government. As more donors enter
Djibouti, the government leverages their status, and rival donors feel their influence threatened.
As a result, donor competition emerges, increasing the cost of allocating aid in exchange for
access. Therefore, the delivery and composition of aid has evolved, with donors disguising aid as
22 As President Guellé continues to leverage Djibouti’s geo-strategic value, especially in the context of oil export
and Iranian-backed offensives in the Gulf of Aden, countries like the UAE, Saudi Arabia, and the United States have
expressed discontentment with Guellé’s “greed”.

40
altruistic in order to avoid alerting rivals and further increasing the cost of doing business in
Djibouti.
Understanding how the trajectory of geo-strategic value has caused changes in the ways aid
is allocated is particularly important. With negotiations nearly settled for India to build a military
base, and discussions with Russia currently under way to do the same, donor competition within
this geo-strategic state is only going to increase. Therefore, Djibouti should continue to be
monitored as an indication of how the allocation of aid is changing amidst increasing great
power competition. Djibouti provides an example of how politically-motivated donors can
evolve in response to recipient government agency and donor competition.
If the delivery and composition that enables identification of politically-driven aid has
changed, as the case of Djibouti suggests, then scholars and practitioners need to adjust. In order
for altruistic actors to reach populations in need, it is imperative that they can distinguish
between need-driven and politically-motivated donors. In demonstrating how donor-driven aid
can be disguised as altruistic, this paper calls for more studies and action to be taken to better
understand how traditionally need-driven channels of aid allocation can be usurped for political
purposes.

41
BIBLIOGRAPHY
Acharya, A., Dreher, A., & Fuchs, A. (2016). The relationship between aid and government procurement of
infrastructure. World Bank Economic Review, 30(2), 305-326. https://doi.org/10.1093/wber/lhv052
Adelman, I., & Chenery, H. B. (1966). The foreign aid and economic development: the case of Greece. The Review
of Economics and Statistics, 1-19.
Alesina, A., & Weder, B. (2002). Do corrupt governments receive less foreign aid? American Economic Review,
92(4), 1126-1137. https://doi.org/10.1257/00028280260344443
Ali, I., Banks, A., & Parsons, C. (2015). The geopolitics of African regionalism. Review of African Political
Economy, 42 (143), 163-181. https://doi.org/10.1080/03056244.2015.1013552
Allendoerfer, C. (2015). Public opinion and foreign aid: A review essay. International Interactions, 41(3), 604-615.
https://doi.org/10.1080/03050629.2015.1052353
Annen, K., & Moers, L. (2016). Donor competition for aid impact, and aid fragmentation. Journal of Development
Economics, 118, 127-146. https://doi.org/10.1016/j.jdeveco.2015.12.006
Annen, K., & Strickland, C. (2017). Elections, protests, and incumbent politics in African multiparty elections.
Comparative Political Studies, 50(3), 261-292. https://doi.org/10.1177/0010414016636595
Apodaca, C. (2017). Foreign Aid as Foreign Policy Tool. In Oxford Research Encyclopedia of Politics.
https://doi.org/10.1093/acrefore/9780190228637.013.332
Apodaca, C. (2017). The political economy of humanitarian aid: Understanding the impact of US foreign aid on
civil conflict. Foreign Policy Analysis, 13 (2), 290-309. https://doi.org/10.1093/fpa/orw047
Bayer, R. C., & Rupert, M. (2004). Peace through globalization: Successes and failures since World War II. Journal
of Conflict Resolution, 48 (6), 783-807. https://doi.org/10.1177/0022002704269353
Bermeo, S. B. (2011). Foreign Aid and Regime Change: A Role for Donor Intent. World Development, 39(11),
2021–2031. https://doi.org/10.1016/j.worlddev.2011.07.019
Berthélemy, J. C. (2003). Bilateral donors' interest vs. recipients' development motives in aid allocation: Do all
donors behave the same? Review of Development Economics, 7(2), 179-194. https://doi.org/10.1111/1467-
9361.00183
Berthélemy, J. C. (2005). Bilateral donors' aid allocation decisions: A three-dimensional panel analysis.
International Review of Economics & Finance, 14(3), 253-274. https://doi.org/10.1016/j.iref.2004.05.001
Berthélemy, J. C. (2006). Bilateral donors' aid allocation decisions: A three-dimensional panel analysis.
International Review of Economics & Finance, 14 (3), 253-274. https://doi.org/10.1016/j.iref.2004.05.001
Berthélemy, J.-C., & Tichit, A. (2002). Bilateral Donors’ Aid Allocation Decisions A Three-dimensional Panel
Analysis.
Bigsten, A., & Tengstam, S. (2015). International coordination and the effectiveness of aid. Journal of Development
Studies, 51(7), 741-757. https://doi.org/10.1080/00220388.2015.1019278
Blair, R., Marty, R., & Roessler, P. (2022). How altruistic is humanitarian aid? The Journal of Politics, 84(1), 89-
104. https://doi.org/10.1086/690235
Blanton, R. G., & Blanton, S. L. (2007). Humanitarianism with a moral face: The diplomacy of compassion.
International Studies Review, 9 (1), 35-65. https://doi.org/10.1111/j.1468-2486.2007.00647.x
Bourguignon, F., & Platteau, J. P. (2015). Aid effectiveness: Opening the black box. American Economic Review,
105(5), 215-219. https://doi.org/10.1257/aer.p20151021
Bouton, L., & Carter, D. B. (2014). Competing for global influence in the context of change. European Journal of
Political Research, 53 (2), 360-379. https://doi.org/10.1111/1475-6765.12038
Boutton, A., & Carter, D. B. (2014). Fair-Weather Allies? Terrorism and the Allocation of US Foreign Aid. Journal
of Conflict Resolution, 58(7), 1144–1173. https://doi.org/10.1177/0022002713492649
Brioni, S., & Gulema, D. (2018). Postcolonial dependencies and development cooperation in the Horn of Africa:
Ethiopia, Eritrea, Somalia, and Djibouti. Third World Quarterly, 39(8), 1482-1501.
https://doi.org/10.1080/01436597.2017.1401414
Bueno de Mesquita, B., & Smith, A. (2009). A political economy of aid. International Organization, 63(2), 309-
340. https://doi.org/10.1017/S0020818309090103
Bueno de Mesquita, B., & Smith, A. (2016). The political economy of humanitarian aid. International Organization,
70(2), 331-362. https://doi.org/10.1017/S002081831500046X
Burnside, C., & Dollar, D. (2000). Aid, policies, and growth. American Economic Review, 90(4), 847-868.
https://doi.org/10.1257/aer.90.4.847
Cabestan, J. P. (2020). Introduction: China and the geopolitics of the Horn of Africa. China and Africa Review, 2
(1), 1-17. https://doi.org/10.1177/2057891120922818
Davies, G., & Klassen, S. (2011). Aid and the political economy of policy change. World Development, 39(11),
1982-1993. https://doi.org/10.1016/j.worlddev.2011.05.017

42
Davies, R. B., & Klasen, S. (2019). Conflict, access to finance, and donor aid. World Bank Economic Review, 33
(3), 650-677. https://doi.org/10.1093/wber/lhz003
Davies, R. B., & Klasen, S. (2019). Darlings and orphans: Interactions across donors in international aid. The
Scandinavian Journal of Economics, 121(1), 243-277.
Dietrich, S. (2013). Bypass or engage? Explaining donor delivery tactics in foreign aid allocation. International
Organization, 67(3), 551-582. https://doi.org/10.1017/S0020818313000121
Dietrich, S., & Murdie, A. (2017). Human rights shaming through INGOs and foreign aid delivery. International
Studies Quarterly, 61(1), 84-99. https://doi.org/10.1093/isq/sqw052
Dietrich, S., & Wright, J. (2015). Foreign aid and democratic development in Africa. American Journal of Political
Science, 59 (1), 168-183. https://doi.org/10.1111/ajps.12142
Dietrich, S., Mahmud, M., & Winters, M. (2017). Foreign aid and voting in international organizations: Evidence
from the IWC. International Organization, 71(2), 421-455. https://doi.org/10.1017/S0020818316000340
Dollar, D., & Levin, V. (2006). The increasing selectivity of foreign aid, 1984-2003. World Development, 34(12),
2034-2046. https://doi.org/10.1016/j.worlddev.2006.04.004
Dreher, A., Fuchs, A., & Nunnenkamp, P. (2013). Aid delivery through non-governmental organizations: Does the
aid channel matter for the impact on child health? Journal of Development Economics, 105, 30-47.
https://doi.org/10.1016/j.jdeveco.2013.06.003
Dunning, T. (2004). Conditioning the effects of aid: Cold War politics, donor credibility, and democracy in Africa.
International Organization, 58 (2), 409-423. https://doi.org/10.1017/S0020818304582065
Easterly, W. (2007). Was development assistance a mistake? American Economic Review, 97 (2), 328-332.
https://doi.org/10.1257/aer.97.2.328
Easterly, W., & Pfutze, T. (2008). Where does the money go? Best and worst practices in foreign aid. Journal of
Economic Perspectives, 22(2), 29-52. https://doi.org/10.1257/jep.22.2.29
Easterly, W., & Williamson, C. R. (2011). Rhetoric versus reality: The best and worst of aid agency practices. World
Development, 39(11), 1930-1949. https://doi.org/10.1016/j.worlddev.2011.05.015
Eichenauer, V., & Knack, S. (2015). Bilateralizing multilateral aid? Aid allocation by World Bank trust funds.
Fink, C., & Radaelli, C. M. (2011). Making and breaking reputations: The politics of international humanitarian
action. International Studies Quarterly, 55 (3), 575-598. https://doi.org/10.1111/j.1468-2478.2011.00652.x
Freedom House. (2023). Djibouti. In Freedom in the World 2023 . Retrieved from
https://freedomhouse.org/country/djibouti/freedom-world/2023
Frot, E., & Santiso, J. (2009). The ambivalence of foreign aid: Confronting the challenges of the 21st century. World
Development, 37(1), 169-178. https://doi.org/10.1016/j.worlddev.2008.01.017
Frot, E., & Santiso, J. (2011). A better globalization for a better world? The governance of global value chains.
Journal of Business Ethics, 100(1), 1-7. https://doi.org/10.1007/s10551-010-0638-6
Fuchs, A., Dreher, A., & Nunnenkamp, P. (2015). Determinants of donor generosity: A survey of the aid budget
literature. World Development, 69, 169-183. https://doi.org/10.1016/j.worlddev.2014.11.014
Gartzke, E., & Naoi, M. (2011). Multilateralism and Democracy: A Dissent Regarding Keohane, Macedo, and
Moravcsik. International Organization, 65(03), 589-598.
Gartzke, E., & Naoi, M. (2011). The causes of peace: A network perspective. International Studies Quarterly, 55
(2), 377-393. https://doi.org/10.1111/j.1468-2478.2011.00633.x
Gehring, K., Langbein, J., & Sekkat, K. (2017). Aid for trade: Assessing the effects on recipient exports of
manufactures and primary commodities to donors and non-donors. Journal of Development Economics, 127, 36-56.
https://doi.org/10.1016/j.jdeveco.2017.03.002
Geneva Conventions. (1948). Geneva Convention relative to the Treatment of Prisoners of War, August 12, 1949, 6
U.S.T. 3316, 75 U.N.T.S. 135. International Committee of the Red Cross. Retrieved from
https://www.icrc.org/en/document/geneva-conventions-1949
Joyce, T., & Blankenship, B. (2023). The strategic implications of China's foreign investments: Evidence from
Djibouti. International Studies Quarterly, 67 (1), 124-142. https://doi.org/10.1093/isq/sqab004
Kalyvitis, S., & Vlachaki, E. (2010). Democratization and foreign aid. European Journal of Political Economy, 26
(4), 516-529. https://doi.org/10.1016/j.ejpoleco.2010.05.006
Kapfer, S., R. Nielsen, and D. Nielson. 2007. If you build it, will they come? Foreign aid’s effects on foreign direct
investment. In MPSA National Conference (Vol. 14).
Kennedy, M. T. (2019). International politics of the Horn of Africa, 2018: Conflict resolution, regional cooperation,
and foreign intervention. The International Spectator, 54(1), 22-36. https://doi.org/10.1080/03932729.2019.1553942
Kilama, E. L. (2018). China's impact on African development. The Journal of Modern African Studies, 56 (1), 1-26.
https://doi.org/10.1017/S0022278X1700016X
Kilby, C., & Dreher, A. (2010). The impact of aid on growth revisited: Do donor motives matter? Economics
Letters, 107(3), 338–340.

43
Kim, S. (2019). Foreign investment and aid allocation in developing countries. Journal of Development Economics,
140, 375-391. https://doi.org/10.1016/j.jdeveco.2019.03.004
Kim, S., & Menninga, E. J. (2020). When do donors cooperate and when do they compete? American Political
Science Review, 114 (1), 208-223. https://doi.org/10.1017/S000305541900062X
Kisangani, E. F., & Pickering, J. (2015). The United States, Djibouti, and the Gulf of Aden. Asian Journal of
Comparative Politics, 3 (3), 267-284. https://doi.org/10.1177/2057891115604331
Koch, S., Leiderer, S., Faust, J., & Molenaers, N. (2017). The rise and demise of European budget support: political
economy of collective European Union donor action. Development Policy Review, 35(4), 455–473.
https://doi.org/10.1111/dpr.12190
Krzymowski, A. (2022). Foreign policy and the identity of Poland in Central Europe. Polish Review of International
Affairs, 31(1), 143-163. https://doi.org/10.2478/priar-2022-0008
Lake, D. A. (2018). Hierarchy in international relations: Authority, sovereignty, and the new structure of world
politics. International Organization, 62 (1), 1-36. https://doi.org/10.1017/S0020818317000285
Lancaster, C. (2000). Aid to Africa: So much to do, so little done. The University of Chicago Press.
https://www.jstor.org/stable/10.7208/chicago/9780226468503.001.0001
Lebovic, J. H., & Voeten, E. (2009). The politics of implicit bias: How Chilean deputies prioritize public policies for
their constituents. American Journal of Political Science, 53 (1), 123-141. https://doi.org/10.1111/j.1540-
5907.2008.00369.x
Lee, C. (2022). Political versus humanitarian aid: The case of Djibouti. International Studies Quarterly, 66 (1), 89-
106. https://doi.org/10.1093/isq/sqac005
Levin, V., & Dollar, D. (2004). Does globalization increase child labor? World Bank Economic Review, 18 (1), 1-
28. https://doi.org/10.1093/wber/lhh004
Lopez, H., Papageorgiou, C., & Vourvachaki, E. (2020). Heterogeneous firms, global competition, and wage
inequality. American Economic Journal: Macroeconomics, 12 (3), 1-36.
https://doi.org/10.1257/mac.20170353
Martens, B., Mummert, U., & Murrell, P. (2002). Does aid translate into votes? World Bank Economic Review, 16
(2), 125-149. https://doi.org/10.1093/wber/16.2.125
Momani, B., & Ennis, A. M. (2012). Foreign aid, institutions, and governance in MENA countries. Middle East
Development Journal, 4(2), 1-28. https://doi.org/10.1080/17938120.2012.696189
Morrissey, O. (1993). Trade, aid, and development: A comparative analysis. Economic Development and Cultural
Change, 41 (4), 825-847. https://doi.org/10.1086/451184
Mosse, D. (2005). Global governance and the ethnography of international aid. The aid effect: Giving and governing
in international development, 12(54), 1-36.
Narang, N. (2016). Foreign aid and the future of liberal world order. Journal of Global Security Studies, 1 (1), 96-
110. https://doi.org/10.1093/jogss/ogv002
Neilson, D. (2013). The sociology of human rights: Are protest and civil rights movements effective? American
Sociological Review, 78 (4), 568-590. https://doi.org/10.1177/0003122413490420
Nielsen, R. A. (2013). Elections and human rights: A review essay. Human Rights Quarterly, 35(2), 487-501.
https://doi.org/10.1353/hrq.2013.0021
Nobuhiro Kubo-Reuters. (2016). A regional development strategy for the Horn of Africa. Journal of International
Affairs, 36(1), 205-218. https://www.jstor.org/stable/41437821
Nye, J. S. (2004). Soft power: The means to success in world politics. Public Affairs.
https://www.jstor.org/stable/10.7312/nye-12635
Papanek G. F. (1972). ‘The effect of aid and other resource transfers on savings and growth in less developed
countries’, The Economic Journal, Vol. 82, pp. 935–950
Payaslian, S. (1996). The making of the modern Armenian state: From catastrophe to democratization. Journal of
International Affairs, 49 (1), 205-227. https://www.jstor.org/stable/24357634
Piknerová, L. (2014). The role of non-state actors in international relations. SAGE Open, 4(3), 215824401454622.
https://doi.org/10.1177/2158244014546220
Powell, M. G., & Bobba, G. (2006). Conditionality without enforcement: Are IMF agreements self-enforcing?
International Organization, 60 (3), 683-718. https://doi.org/10.1017/S0020818306060222
Reisen, H., & Stijns, J.-P. (2011). Aid and public finance: A normative perspective. International Finance, 14(3),
421-438. https://doi.org/10.1111/j.1468-2362.2011.01289.x
Rowlands, D., & Calleja, R. (2015). Donor Competition for Influence in Recipient Country.
Sachs, J. (2009). A new globalization for a new world economy. Financial Analysts Journal, 65 (2), 30-41.
https://doi.org/10.2469/faj.v65.n2.6
Sen, G. (2018). Emerging economies and the shifting landscape of development assistance. Development Policy
Review, 36(3), 287-305. https://doi.org/10.1111/dpr.12282

44
SIPRI. (2019). Developments in international relations, 2018. The International Relations of the Horn of Africa,
26(2), 104-118. https://www.jstor.org/stable/26452204
Steinwand, M. C. (2015). The political economy of the aid industry: What’s gone wrong? Review of International
Political Economy, 22(2), 357-389. https://doi.org/10.1080/09692290.2014.953720
Stiglitz, J. E. (2002). Globalization and its discontents: Addressing the problems of the United States in a troubled
world. The American Economic Review, 92 (2), 1-6. https://www.jstor.org/stable/3083261
Sun, Y., & Zoubir, Y. (2016). Africa and China's 21st century maritime Silk Road. Journal of Contemporary
African Studies, 34 (2), 257-277. https://doi.org/10.1080/02589001.2016.1166520
The New York Times. (2014). How global inequality affects international relations and security. The Journal of
Strategic Studies, 24(1), 57-70. https://doi.org/10.1080/0140239032000159756
The Trumpet. (2018). The political economy of China's development strategy in Africa. Foreign Affairs, 97(5), 123-
137. https://www.jstor.org/stable/26568888
Trunkos, A. (2020). Power politics and diplomacy in the Horn of Africa: The role of soft power in regional politics.
African Journal of Political Science and International Relations, 14(4), 129-145.
https://doi.org/10.5897/AJPSIR2020.1212
UIS. (2023). Global flow of funds in 2021 and academic data collected by the Djibouti WFP
Van der Veen, A. M. (2011). Ideas, interests and foreign aid (Vol. 120). Cambridge University Press.

45
ARTICLE 2: APPENDIX A
REGRESSION ANALYSES
Table A.1: Regression Analysis on Politically-Driven Variables and ODA
Table A.2: Regression Analysis on Need Variables and ODA

46

47
CAMEROON IN CONFLICT:
ASSESSING SUBNATIONAL ALLOCATION OF HUMANITARIAN AID
ABSTRACT
In Cameroon, a terrorist insurgency in the Far North and an armed separatist conflict in
the Southwest and Northwest regions have generated one of the world’s most neglected
displacement crises. In theory, humanitarian aid could mitigate the gaps in conflict response and
recovery left by the recipient government. In practice, aid is allocated unevenly across conflictaffected areas in Cameroon, exacerbating existing inequities. Scholars suggest that this is in line
with politically-driven determinants of foreign aid but donors claim that the humanitarian aid
allocated differs from development aid and is indeed need-driven, flowing to those which need it
most. However, there are few comprehensive studies which test the determinants of humanitarian
aid specifically and its susceptibility to political interference. Comparing the two conflicts within
Cameroon addresses these issues by demonstrating the decision-making process of donors.
While the recipient government welcomes international intervention in the Far North, the
president deters involvement in the politically disfavored Anglophone areas. Therefore, donors
have to choose. In line with need-driven theory, they can allocate aid according to need and
potentially sacrifice policy gains from the recipient government. Or, in line with politicallydriven theory, donors can choose to concentrate aid to the Far North, intentionally ignoring the
intensity of need in the Anglophone areas to sustain political favor with the president. Donors
consistently choose the latter, suggesting that despite humanitarian mandates and the claims of
donors, humanitarian aid is also plagued by politics.
INTRODUCTION
Despite having a government that is fighting a civil war on two fronts with multiple
international interventions, Cameroon has been listed as one of the world’s most neglected
displacement crises by prominent humanitarian organizations for the past five years (NRC,
2024). Each year, conditions in the areas affected by the terrorist insurgency in the Far North and
the Anglophone separatist crisis in the Northwest and Southwest continue to deteriorate while
appeals for humanitarian assistance remain unanswered. In 2023, only 27% of the UN
Humanitarian Response Plan for Cameroon was funded, while Sudan’s appeal was nearly 50%
funded and Ukraine’s appeal was 64% funded (UNOCHA, 2023). If humanitarian assistance is
meant to reach the places that need it most, then why do protracted conflicts in Cameroon lack
the same access to aid as other countries in crisis?
Substantial evidence suggests that the allocation of aid is dependent on both the donorrecipient relationship and the recipient country context (Fuchs, Dreher, & Nunnenkamp, 2014).
While some aid is need-driven and intended to reach the most destitute populations, other aid is
politically-driven and serves primarily to facilitate political objectives of donor and recipient
governments. Often, these differing determinants of aid can be distinguished by the delivery and
the composition of donor contributions. For example, bilateral development aid, allocated as
money from donor governments directed to recipient governments, often reflects politicallydriven agendas (Bearce & Tirone, 2010). Alternatively, humanitarian aid, or assistance given to
provide immediate and impartial life-saving resources to populations experiencing crisis (UN
GA Resolution 60/124) is more often allocated in accordance with humanitarian principles
(OECD, 2010) and can consist of monetary, programmatic, or in-kind assistance from any type
of donor. Therefore, development aid more often reflects political motivations of donor and
recipient governments while humanitarian aid is more often driven by the needs of populations

48
(Maizels and Nissanke 1984; Burnside and Dollar 2000; Neumayer 2003). Some donors, like the
United States, even emphasize their different motivations for different types of aid allocation,
stating that humanitarian assistance is need-based and given without influence of foreign policy
priorities (Kevlihan, DeRouen, & Biglaiser 2014).
In part because of this strong assumption that humanitarian aid allocation is indeed needbased, there is a troubling dearth of studies that actually test the determinants of humanitarian aid
in ways that can reveal how politics might shape that allocation (Narang, 2016). There are even
fewer studies looking at the determinants of humanitarian aid in the subnational context – i.e. not
only examining how aid is allocated between countries, but how it is allocated within them23
.
This lack of subnational aid research has been attributed to a lack of available data – subnational
studies usually require scholars to collect and create new data at the region or project level
(Masaki, 2018; Findley, Powell, Strandoll, et al., 2011). Thus, this study provides a much-needed
empirical contribution, examining the extent to which humanitarian aid in Cameroon from 2003-
2023 is need-driven or politically driven. Using detailed regional and project-level data, the
analyses of conflict-affected areas in Cameroon contributes to both the limited research on the
determinants of humanitarian assistance and the growing literature studying the subnational
distribution aid.
If humanitarian aid is indeed driven by the needs of recipient populations rather than
political priorities of donors and recipient governments, then the amount of aid, number of
donors, and number of interventions in a region should increase with intensity of need. Needs of
recipient populations are traditionally measured through indicators of food insecurity, displaced
persons, poverty, conflict intensity, and marginalized groups. Response to these needs should
follow the existing literature and logic that highlights the altruistic mandate of humanitarian aid
as well as the need-driven claims of bilateral donors, multilateral institutions, and NGOs.
Therefore, it would be expected that aid will flow to poorer, conflict-affected areas experiencing
acute crises. Yet, data on aid flows demonstrate that responses to humanitarian appeals are not
evenly distributed across or within countries, calling into question whether politics also dominate
aid allocation in the humanitarian realm.
In Cameroon, I find that consistent with expectations of need-driven humanitarian aid,
assistance flows primarily to the conflict-affected regions. However, qualitative disaggregated
analyses demonstrate that this aid does not appear to follow the humanitarian mandate of
impartiality, neutrality, and independence. Instead, consistent with expectations of politically
motivated aid distribution, I also find that aid is not correlated with measures of inequality.
Humanitarian assistance is concentrated amidst the Far North terrorist insurgency rather than the
separatist Anglophone conflict. This implies that while donors are allocating aid to conflict
affected areas, they are choosing to avoid the politically-disfavored Anglophone populations and
are diverting aid to the Far North where they can protect their political interests by maintaining
favor with the recipient government.
In order to illustrate this decision-making process, I first review the literature and law
regarding humanitarian mandates. Then, in line with scholarship suggesting politically-driven aid
allocations, I show the uneven distribution of aid across countries. Next, I provide context for the
conflicts within Cameroon and their associated political dynamics in connection to both the
recipient government’s priorities and donor’s foreign policies. I follow with an empirical
analysis of subnational aid allocation in Cameroon and qualitative analyses of potential political
donor strategies. I finish with a sectoral study of education, explaining the role which education
has played in each of the two conflicts in Cameroon and further exemplifying how political
23 (notable exceptions include Zhang, 2004; Masaki, 2018; Nunnenkamp, Öhler, & Sosa Andrés, 2016;
Nunnenkamp, Sotirova, & Thiele, 2016, and Briggs, 2014)

49
incentives influence allocation of aid to the sector. In revealing the uneven and politically-driven
distribution of humanitarian assistance to conflict-affected areas in Cameroon, this study
concludes with a call for further data collection and analysis of subnational aid allocation and its
susceptibility to the political priorities of donors and/or the recipient government rather than the
needs of recipients.
THEORY
Within the aid community, bilateral donors, multilateral institutions, and NGOs alike
assert that humanitarian aid is need-driven and allocated to those populations which need it most.
This altruistic view of humanitarian aid is further supported by international legal frameworks
(UNGA, 46/182; Geneva Conventions, 1948; UNGA, 60/124). They provide clear conditions for
its allocation as immediate impartial life-saving assistance to populations experiencing complex
emergencies (IASC,1994). These resolutions require humanitarian aid to follow a three-principal
mandate of impartiality, neutrality, and independence (UN, 1991). Similar principles have been
adopted by multilateral institutions and select donor governments. The European Union
enshrined humanitarian aid as preventing and alleviating suffering, preserving life, and
maintaining human dignity in the face of disaster (EU Consensus, 2007). Saudi Arabia defines
their “pioneering” role in humanitarian action as “alleviating human suffering and ensuring that
all people are given the opportunity to live healthy, dignified lives” (King Salman Unified
National Platform, 2023) and France aims to “provide assistance and protection for vulnerable
populations” (French Ministry for Europe and Foreign Affairs, 2024).
In the context of conflict-affected countries, the mandate remains clear, requiring UN
organizations to act according to the humanitarian principles set forth by the aforementioned UN
resolutions and uphold the Geneva Conventions. Bilateral donors similarly suggest that they
uphold the humanitarian mandate in allocating humanitarian aid in circumstances of conflict.
Norway directs aid to countries where there are “great, unmet humanitarian needs” to “protect
civilians affected by war” (Norwegian Government, 2023) and the United States claims to use
humanitarian assistance to “provide life-saving assistance to tens of millions of displaced and
crisis-affected people” (U.S. Department of State, 2023). According to these assertions and
associated criteria for its allocation, humanitarian aid should, of course, adhere to need-driven
theory.
H₁: humanitarian aid is need-driven and allocated to areas with the higher levels conflict
intensity, food insecurity, development deprivations, and Internally Displaced Persons
While some scholars have found that donors do indeed target these crisis contexts (Sachs,
2006, Lancaster, 2005; Keck and Sikkink, 1998, Büthe et al., 2012), most of the literature
demonstrates how aid is influenced by political factors across actors (McKinlay and Little,
1978). First, bilateral donors allocate aid to further their own political, economic, and security
priorities (Berthelemy & Tichit, 2004). Second, multilateral institutions and INGOs are resource
dependent on donor governments and are thus also affected by foreign policy agendas (Cooley &
Ron, 2002; Knack, 2015). Third, recipient governments also intervene and influence aid
allocation according to their own domestic interests (Bauer, 2009; Easterly, 2007; Ziff, 2023).
In conflict-affected countries, aid may be particularly politically driven. In times of
conflict, aid can be used as a means to acquire or solidify relationships with allies (Alesina &
Dollar, 2000; Bermeo, 2017). Donors may also use aid to protect their investments or economic
interests from the crisis (Klosek, 2020). Aid can additionally be used in conflict to prevent

50
spillovers like regional instability or migration into the donor country (Duffield, 2008) as well as
to securitize against potentially expanding threats (Vaux, 2006; Boutton, 2014). Donors may also
avoid allocating aid to particular conflicts altogether if that aid cannot be used to achieve
political agendas (Lis, 2018; Narang, 2016; Mamoon, 2016). The political determinants of aid
thus suggest that countries who are either allied with donors or where conflict threatens donors
receive more aid. Therefore, scholarly consensus from cross-national studies supports
politically-driven theories and demonstrates how different countries in conflict receive different
responses from donors that aren’t necessarily associated with intensity of need (Whittall, 2015).
Most research which supports politically-driven theory, however, reflects the allocation
of development or unspecified aid–not humanitarian aid. In fact, the lack of studies which
disaggregate aid by intention and investigate only humanitarian aid has been noted by scholars
(Drury, Olson, & VanBelle, 2005; Lis, 2018). Therefore, while altruistic assumptions
surrounding humanitarian aid are justified by international law and the claims of donors, the
overwhelming evidence for politically-driven aid allocation in other contexts calls for additional
inquiry into the provision of humanitarian aid. If there is evidence that each type of donor
(bilateral, multilateral, NGO, private) across country contexts is influenced by politics regardless
of altruistic claims, then why would humanitarian aid be apolitical?
H₂: humanitarian aid is also politically-driven and allocated to areas with higher geostrategic value for donors or politically favored by the recipient government
Figure 1 demonstrates how different countries in conflict receive different responses
from humanitarian actors that aren’t necessarily associated with intensity of need (Whittall,
2015). Data and many cross-country analyses suggest that there is disproportionate attention and
aid at the onset of conflict in geo-strategic states, often diverting resources away from
‘forgotten’, prolonged, or less geo-strategic conflict contexts (Narang, 2016; Fink & Redaelli,
2011). Uneven media exposure exacerbates this inequitable allocation of aid by shaping public
perception and support according to press coverage (Van Belle, 2003; Van Belle, Rioux, &
Potter 2004;) which in turn is responded to by donor governments and resource dependent aid
organizations. As a result, distribution of aid across countries is disproportionate and inequitable.
Figure 1: Amount of Humanitarian Aid per capita per Country in 2023
Each of the countries in Figure 1 are currently in conflict but receive vastly different
amounts of aid, even when controlling for the number of people affected by the crisis. This graph
alone illustrates how humanitarian aid can be politically driven, as Yemen, Somalia, and Syria

51
have significant geo-strategic value and receive disproportionately high amounts of aid.
Similarly, countries like Ukraine and Afghanistan which have political significance for
prominent donors and global powers receive far more aid than protracted crises like those in
DRC and Cameroon. Therefore, it is imperative to consider politically-driven hypotheses of
humanitarian assistance despite the existing altruistic humanitarian principles requiring needdriven aid allocation.
If even humanitarian aid is unevenly and inequitably distributed across countries, what
about the allocation of assistance within them? Is all aid plagued by politics or can we have hope
that it will flow to the places and populations that need it most? There is an abundance of
evidence demonstrating regional disparities within states’ government spending (Hodler &
Raschky, 2014; Kramon & Posner, 2016), development (Permanyer & Smits, 2020), and
political favor (Bates, 1974). Such within-country favoritism is shown by domestic governments
not only spatially, but also according to ethnicity, religion, and political affiliation (Ejdemyr,
Kramon, & Robinson, 2018). But what about the allocation of aid from external actors? Does
humanitarian assistance reflect need, the political favor of domestic recipient governments, the
policy agendas of donors, or all of the above? Emergent within-country studies suggest that
subnational dispersion of development aid does not target the poorest places in Kenya (Briggs,
2014; Barrett, 2015), India (Nunnenkamp et al., 2016; Song, Brazys, & Vadlamannati, 2021),
China (Zhang, 2004), or Malawi (Nunnenkamp, Sotirova, & Thiele, 2016). Is humanitarian aid
in conflict contexts also politically plagued?
IS ALL AID PLAGUED BY POLITICS? THE CASE OF HUMANITARIAN AID TO CAMEROON
Cameroon is an especially well-suited case to study the subnational distribution of aid
amidst conflict with the presence of three spatially and ideologically distinct humanitarian crises.
Cameroon also has significant regional variation throughout the country across measures of need
like food insecurity, poverty, and political disfavor, making it easier to disentangle the
determinants of subnational allocation of aid. This study compares aggregated as well as detailed
project-level data on aid with incident-level conflict data from ACLED and regional need data
from the UN, IASC, and OPHI.24
While quantitative analyses are inconclusive as to the
subnational determinants of aid, qualitative analyses of disaggregated project-level data provide
important insight into the ways in which political priorities of donors may dominate the
allocation of aid to conflict-affected areas. In the context of Cameroon, these findings have
important implications given international interests, the aging autocracy, and extent of
humanitarian crises in the country.
Cameroon also provides an interesting case for studying aid allocation because unlike
most conflict-affected countries whose economies significantly suffer during crises, as a whole,
Cameroon’s economy continues to grow (African Economic Outlook, 2023). Cameroon is
considered a medium tier developing country, with a human development index ranking similar
to regional hegemons like Nigeria, Kenya, Senegal, Tanzania, and Sudan (HDI, 2024). Such
economic resilience despite costly domestic conflict, can largely be attributed to government
involvement in, and prioritization of, export industries including production in areas affected by
conflict. Cameroon ships valuable resources like crude oil, gas, cocoa, and gold mostly to China
but also to India, Italy, the Netherlands, and United Arab Emirates (Observatory of Economic
Complexity, 2022). Pursuit of such neoliberal trade policies have bolstered the economy after a
near-collapse in the 1990s, but have also contributed to systematic economic inequality within
24 The IASC is the Emergency Response and Preparedness Group of regional Inter-Agency Standing Committee for
the EU and UNOCHA. OPHI is the Oxford Poverty and Human Development Initiative. Data are also taken from
the following UN organizations: UNOCHA, UNHCR, World Bank, and World Food Programme

52
the country. As illustrated by the map of multidimensional poverty in Figure 2, wealth is
polarized between the north and south of Cameroon with over half of all impoverished citizens
living in the three northernmost regions (World Bank, 2018).
Figure 2: Map of Regional Disparities Cameroon, 2023
Regional disparities are rooted in historical and social divisions that have resulted in quite
different levels of development, access to public services, and political representation. For
example, prior to the late nineteenth century, the Far North and North regions were part of the
agriculturally focused Sao or Fulani civilizations and today remain rural. The southern coasts,
which participated in the slave trade, currently have comparatively more infrastructure and less
poverty. During the imperial era, the implementation and administration of colonial territories
left legacies with contemporary social, political, and economic implications. Through
asymmetric treaties with Duala rulers and a series of wars in the 1880s, imperial Germany
established the colony of Kamerun. At its peak, German Kamerun included dozens of distinct
populations with ethnic, religious, and linguistic differences. After Germany lost the First World
War, the territory was divided between the British protectorate of Nigeria and French Equatorial
Africa. Over the next half century, this administrative segmentation further divided the
population by cultivating separate Anglophone and Francophone identities in addition to existing
ethnic, religious, and cultural divides. It was only during decolonization that a referendum was
offered by colonial officials for British Cameroons to be absorbed into either Nigeria or French
Cameroun. The subsequent incorporation of the two Anglophone regions into the French
territory marked the contemporary borders of Cameroon. Thus, the country today is an
amalgamation of imposed colonial borders with religious, ethnic, and economic divisions across
regions.
The most significant of these divides has manifested itself in the attempted secession of
the Anglophone Northwest and Southwest regions. The armed conflict which continues today
reflects the escalation of peaceful protests in 2016 by Anglophone populations who were fighting
the expansion of central government authority over the region’s semiautonomous legal and
education systems. Failed elections in 2017 exacerbated demonstrations and violent government
crackdowns, escalating the crisis into a war fighting for an independent Anglophone state of
Ambazonia. Though the separatist rebellion includes several competing factions, most of these
armed groups are Anglophone youths drawn into militias as widespread school closures, regional
economic collapse, and historical marginalization have limited other opportunities (Nna-Emeka
Okereke, 2018). Since 2017, government forces and rebel groups alike have committed mass

53
human rights abuses, attacks on civilian populations, and prevented the resumption of public
service provision. The war has thus generated a mass exodus from Anglophone regions and a
critical humanitarian crisis amongst the populations which have remained. In total, recent reports
suggest that 715,000 people in the Northwest and Southwest require immediate humanitarian
assistance (OCHA, 2024).
The violent government response in the Anglophone regions to both initial protests and
the ensuing war has had important implications for international humanitarian organizations,
foreign policy, and donor governments looking for the favor of the current administration.
President Paul Biya, in power for nearly fifty years, has repeatedly rejected international efforts
to mediate the conflict and suggested in his 2024 New Year’s address that the “situation is
improving” (Government of Cameroon, 2024) despite evidence suggesting otherwise. While
many global powers have urged Biya to participate in peace talks, others have remained notably
silent. In 2022, President Macron of France visited Cameroon amidst a surge of violence in the
Northwest but, somewhat surprisingly, did not address the issue of the Anglophone crisis
(Mutah, 2022). The United States, who collaborates closely with Biya on containing terrorism,
officially condemned the violence and called for negotiations but then also indicted three
individuals on Biya’s behalf for funding the Ambazonian conflict from abroad (US Justice
Department, 2021). In 2018, the Chairperson of the African Union outwardly supported Biya’s
actions in Anglophone regions and advocated for the territorial integrity of Cameroon (Mutah,
2022). When taken together, these examples show how, despite impartiality being a central
principle of humanitarian assistance, international actors are hesitant to offend the administration
by responding to the Anglophone conflict. This trend lies in stark contrast to the scale of
international interventions in the Far North conflict which has and continues to receive the
largest amount of aid and attention.
Figure 3: Amount of Aid Allocated to Conflict-Affected Areas
While both the Anglophone regions and the Far North received increases in aid following
conflict onset, as Figure 3 illustrates, the Far North is allocated consistently more aid than any
other region. This includes the refugee crisis in the East where 335,000 people are taking refuge
from the protracted civil war in the Central African Republic (UNHCR, 2024). Though there is
significant need, the amount of humanitarian aid allocated to the East and Anglophone areas is
small compared to that in the Far North. This concentration of aid to the Far North began in 2014
with international recognition and response to Boko Haram as a significant terrorist threat. Boko
Haram, or the Peoples of Sunnah for Preaching and Jihad, is prominent throughout the Lake

54
Chad Basin, with costly operations in Chad, Niger, Mali, and Cameroon that preceded
international press coverage. Although their presence in Cameroon has been documented since
2009, the 2014 kidnapping of 275 girls in Nigeria generated global distress. By 2015, the
Cameroonian counter-offense was receiving international monetary and military support. A
decade later, the insurgency maintains a stronghold in the region and Boko Haram claims
responsibility for more casualties in Cameroon than in Nigeria, Chad, and Niger combined
(Oxford Analytica, 2021).
As the conflict continues, the large youth population, failures of Western-supported
government military interventions, and persistent poverty in the region have led to grassroots
responses. Vigilante armed groups fight Boko Haram sometimes in collaboration with
government forces and other times in spite of them (IRC, 2016). This, in addition to the spread of
radical Salafi Islam in the Lake Chad Basin has enabled the Islamic State West Africa Province
(ISWAP) to provide an alluring alternative to both the state and Boko Haram. ISWAP has
marketed itself as a potential governing body that is community-centric and less abusive than
government forces and Boko Haram (Stoddard, 2023). In Nigeria, these claims have been
somewhat validated with fewer cases of violence against civilians when compared to other
regional rebel groups and terrorist organizations (ibid, 2023). Despite their different islamic
affiliation, mandates, and approaches, ISWAP is often still conflated with Boko Haram in
Cameroon. Domestic operations and international military interventions are less interested in
distinguishing between the two, instead focused on combating the combined “terrorist threat”
and containing extremism in the Lake Chad Basin. H₂ would suggest that the political
prioritization of stabilization against terrorism by bilateral donors has translated to the Far North
receiving significantly more attention and aid from international actors than the needy separatist
regions and the refugee crisis in the East.
When the numbers of refugees in the East are combined with the Anglophone separatist
conflict and Far North terrorist insurgency, humanitarian reports suggest that Cameroon has
approximately 2.2 million displaced persons with an additional 2.6 million people in need of
acute humanitarian assistance. Nearly all of those requiring aid are in the aforementioned crisis
regions while the more central regions remain safe and stable. This large regional variation
across Cameroon enables investigation into the relationship between need and the spatial
distribution of aid, which in turn facilitates further analyses of the different responses by
international actors to humanitarian crises according to region. If humanitarian assistance in
Cameroon is need-driven, then aid should be allocated across the East, Far North, and
Anglophone regions according to conflict intensity–consistent with H₁ and the importance of
need as a driver of subnational aid allocation. However, if preferences of donors and domestic
politics matter more, then aid would be allocated according to political priorities as suggested by
H₂ and flow disproportionately to areas like the Far North that are either geo-strategic for donors
or politically favored by the recipient government.
MEASUREMENT & METHODS
To examine whether humanitarian aid in Cameroon is driven by need or by political priorities, I
use World Bank and UNOCHA data to predict the aid volume flowing to each region in each
year. If H₁ is correct and humanitarian aid in Cameroon is need based, I expect to see allocations
flowing to the subnational regions with the highest intensity of need and conflict. If H₂ is correct
and aid is driven by the political priorities of donors or the Cameroonian government donor's
political priorities, I expect to see aid allocated to areas favored by the Biya administration and
areas where donors can securitize their interests.
Measuring Aid Volume

55
Aligning with previous subnational studies (Briggs, 2017; Dipendra, 2020) the empirical
analyses first measure aid flows by using the logarithmic amount of aid allocated per region and
year. These data were collected and aggregated from project-level humanitarian aid contributions
reported by the UNOCHA’s Financial Tracking Service. The sample includes 923 regionally
targeted contributions from 2013 to 2023. One quarter of those projects were split between
multiple regions without specification of the amount per region and were thus divided equally
amongst the regions as was done in similar studies (Yildiz, 2021; Dreher et al.,2019; Briggs,
2017).
Measuring Need & Political Factors
Measures of need are separated across indicators of conflict intensity and subnational
socioeconomic inequality. Inequality variables include INFORM data on development
deprivation, socioeconomic deprivation, and vulnerable group presence. These variables are
complemented by additional measures of need, including poverty from the Multidimensional
Poverty Index’s annual country profiles, displaced persons from UNHCR reports, and an index
of food insecurity created from UNOCHA situation reports. Conflict data are used to measure
both the needs of the population and political factors such as counter-terrorist agendas of donors
or political disfavor by the Cameroonian government. These are collected and aggregated from
ACLED’s event-level dataset as well as indices provided by the European Commission’s
INFORM subnational data collection. Conflict variables include the number of conflict events,
ACLED’s conflict index, INFORM’s conflict intensity index, and a measure of political
violence. Selected indicators and their regional variation as of 2023 are illustrated below in
Figure 4.
Figure 4: Regional Variation of Need Variables
Empirical Results
Table 1 presents the results of two-way fixed effects regressions in which various
measures of need are used to predict aid flows at the region-year level. Two-way fixed effects
models remove confounding variation across time and regions from estimates. To avoid
overfitting, variables are grouped by category with separate models run on conflict, inequality,
and general need indicators. These additional models, which are included in the appendices,
suggest a slight relationship between need and aid allocation. As a robustness check, single
variate two-way fixed effects models are also run. The results suggest that humanitarian aid
allocation is somewhat driven by conflict, but not by other measures of need such as poverty,
food insecurity, or inequality.

56
Table 1: Single-Variate Regression Models
The fact that conflict, but not other aspects of need, predict humanitarian aid is
unsurprising given the distinction between humanitarian and development aid. Altruistic
interpretations suggest that development aid should flow to populations with high levels of
poverty and low levels of education, infrastructure, and economic opportunity (OECD, 2017).
However, the data in the sample are focused on humanitarian contributions, not development aid.
Humanitarian aid is intended to provide immediate life-saving assistance and should thus flow
mostly to places experiencing crisis (Fink & Redaelli, 2011). Therefore, the positive
relationships between conflict and aid without other indicators may support rather than negate
need-driven hypotheses because humanitarian aid appears to be channeled to the areas with the
highest intensity of conflict.
However, when development aid amounts are integrated into the data, need-driven
distribution is still absent. This lack of relationship between non-conflict need indicators and aid
aligns with prominent literature on politically-driven humanitarian assistance (Drury, Olsen, &
Van Belle, 2005) as well as recent subnational studies on the determinants of aid (Briggs, 2017,
2018; Nunnenkamp et al., 2009; Peter et al., 2016). Further support for H₂ is found in other
cross-national studies of aid allocation (Berthelemy, 2006; Claessens et al., 2009; Dollar and
Levin, 2006; Thiele et al., 2007). If the other need indicators measuring regional inequality are
removed and the study focused only on conflict, then, Table 1 would suggest that humanitarian
aid is being allocated as intended and flowing to areas affected by conflict.
The intent of this study though, is to compare aid allocation across Cameroon’s conflictaffected areas, not across all subnational regions. Following Narang (2016), additional analyses
are done to a sample which only include the conflict-affected areas of the Far North, Anglophone
Northwest and Southwest, as well as the East with its spillovers from the civil war in the Central
African Republic. Within these conflict-affected areas, all positive need-driven results disappear
and aid can no longer be explained by the number or intensity of conflicts in a region. This is

57
further illustrated by Figure 5 which shows the amount of aid per conflict-affected area per year.
Though the number of conflicts and intensity of fighting varied greatly over the past decade, the
Far North received consistently more aid than the other three conflict-affected regions.
Figure 5: Amount of Aid Allocated and the Number of Conflict Events per Year
In 2015 and 2016, the East region experienced exponentially more conflict events than
any other region as a result of surging violence in the Central African Republic. In addition to
clashes between suspected Seleka rebels from CAR and Cameroonian government forces, the
East also absorbed tens of thousands of new refugees (UNHCR, 2017) and experienced acute
food insecurity, generating the highest intensity of need in the country thus being where needdriven aid would be expected to be allocated. However, attention and aid from the international
community was concentrated in the Far North. Boko Haram had just gained global recognition as
a serious terrorist threat in Nigeria and international humanitarian actors responded through
immense increases in aid allocated to areas where Boko Haram was present. Therefore, when the
12 conflict events of the Far North are compared to the East with its 1,736 conflict events, it
appears that politically-driven responses to terrorist activity rather than need drove aid allocation.
Several cross-national studies explained similar phenomena across countries, showing how since
the onset of the ‘War on Terror’ bilateral donors and international institutions have allocated
more aid to places with terrorist insurgencies (Cosgrave, 2004) than crisis contexts without
terrorism.
A similar pattern emerges when comparing the onset of the Anglophone separatist
conflict to the terrorist insurgency in the Far North. In 2017, conflict escalation in the
Anglophone regions received an inflow of humanitarian assistance but the amount of aid
allocated was significantly less than the Far North. This pattern holds even when the Northwest
and Southwest regions are combined, as was done in Figure 5. Increased intensity of the conflict
in the Anglophone areas in 2019 and again in 2021 still received less aid than the Far North,
further supporting the idea that terrorism, rather than just the existence of conflict, is driving the
subnational distribution of humanitarian assistance within Cameroon.
Two areas of international law run counter to this idea of terrorism-driven allocation of
aid. First, there are the guiding humanitarian principles, affirmed across international

58
agreements, 25 of neutrality, impartiality, independence, and humanitarianism. These require that
assistance be without the political intention inherently attached to aid amidst terrorist
insurgencies (Wynn-Pope, Zegenhagen, & Kurnadi, 2015). Second, there are explicit terrorist
laws, such as UNSC 1373 which prevent provision of
“ any funds, financial assets or economic resources or financial or other related services
available, directly or indirectly, for the benefit of persons who commit or attempt to
commit or facilitate or participate in the commission of terrorist acts”.
Resolution 1373 amongst other similar legislation has called into question the neutrality,
impartiality, and independence of humanitarian action (Fraterman, 2014). Both international
anti-terrorism laws and those present in some donor countries outlaw even the unintentional
finance of terrorist organizations. This significantly increases the cost and risk for organizations
to extend humanitarian assistance to terrorist held territories (Mackintosh & Duplat, 2003;
Adelsberg, 2013). In the Cameroonian context, these frameworks should have disincentivized
humanitarian actors from assisting populations in the Far North.
However, there are also both need-driven and politically motivated defenses for the
allocation of aid to terrorist-affected areas. On the one hand, need-driven scholars argue that an
influx of attention and aid to contain terrorist activity can avoid its negative effects on human
development (Blomberg, Hess, & Orphanides, 2004; Gaibulloev & Sandler, 2009). On the other
hand, politically-driven theories suggest donors prioritize combatting and containing extremism
as part of strategic securitization and political agendas (Fleck & Kilby, 2010; Bouton & Carter,
2014). If need was the primary concern, then more consistent attention would have been paid to
the historically underserved Far North prior to 2014 (Boko Haram’s initial recruitment phase).
Across non-conflict measures of need, the Far North has had consistently high levels of poverty,
limited infrastructure, low voter turnout, and periods of acute food insecurity. Yet, neither
development nor humanitarian aid was targeted there over other regions until international
attention was paid to Boko Haram. Therefore, it is likely that aid was initially allocated to the Far
North in accordance with geo-strategic goals of donors and the recipient government rather than
the needs of recipients.
Donors and the domestic government both have political incentives to allocate
humanitarian assistance as an attempt to contain the terrorist insurgency. Counterterrorism has
been prioritized by domestic recipient governments, especially in autocracies, as terrorist
insurgencies have been shown to weaken state capacity, central authority, and public approval
(Wilson & Piazza, 2014; Conrad, Conrad, & Young; 2014). Since 2014, the Cameroonian
government has consistently cited the eradication of terrorism as their primary security priority,
enacting the expansive anti-terrorist law 2014/028: The Suppression of Acts of Terrorism. Since
2001, counterterrorism has also been prominent in foreign policy agendas of bilateral donors like
the United Kingdom, select OECD countries, and of course, the United States (Azam & Thelen,
2010; Kilby & Fleck, 2010). While some states respond by diverting aid away from terrorist held
territories (Ziff, 2023; Whitthall, 2014, 13), others distribute more aid to these areas as a form of
counterterrorism (Azam & Thelen, 2010). Such overlaps between donor and recipient
governments’ political priorities create a probable explanation for the concentration of aid to the
Far North.
The amount of aid allocated to the Far North may also be amplified by the comparative
lack of aid allocated to the East and Anglophone regions. Perhaps the East region should be
classified as a subnational ‘forgotten conflict’ (European Civil Protection & Humanitarian Aid
Operations, 2022; NRC, 2022; Médecins Sans Frontières, 2022). In ‘forgotten’ contexts, the
25 See for example UNGA Res. 46/182; UNGA Res. 58/114; UNSC 1265, 1894; and all the Geneva Conventions.

59
only incentive to allocate aid is humanitarian need, so international actors choose to give media
attention and financial resources instead to areas with both need and potential political gains
(Smillie & Minear 2003, 2004; Narang, 2016). As a result, ‘forgotten’ crises like that in
Cameroon’s East region remain underfunded as new conflicts capture media attention, garnering
humanitarian assistance (Scott, Bunce, & Wright 2022; Westland, 2023). With significant press
coverage from onset to escalation (Enyame, 2021) and continued attention from the international
media and NGOs (NRC, 2022; Al Jazeera, 2023; Amnesty, 2023), the Anglophone crisis cannot
quite be considered ‘forgotten’. So why does it still receive proportionately less aid than the Far
North?
A likely reason for why the Northwest and Southwest regions receive less humanitarian
aid than would be predicted given the high levels of need is that the area is politically disfavored
by the Cameroonian government. Since its onset, through censorship, arrests, and ultimate
military action, President Biya has unequivocally opposed the Anglophone movement-turnedconflict. While Biya welcomed international intervention in the Far North, he has repeatedly
rejected international involvement in mediation of the separatist conflict. If aid is politicallydriven, then donors would thus have a disincentive to give aid to the Anglophone regions for fear
of offending or undermining the recipient government. This could be one explanation for the
presence of both lower amounts of aid and fewer donors in the separatist Northwest and
Southwest regions.
In the data, 54 different donors have targeted aid to at least one of the conflict-affected
areas of Cameroon. Only a third of these donors contributed to projects that allocated assistance
to the Anglophone regions, whereas almost all actors worked in the Far North. Half of the
sample is composed of bilateral donors, with the remaining half split between UN organizations
and prominent INGOs. Figure 6 illustrates the allocation of aid by bilateral donors to the
conflict-affected areas. With a few notable exceptions, donor governments generally contributed
to projects in each of the conflict zones, though some gave more to the Far North.
All the bilateral donors, except for Sweden and Switzerland have been classified in
prominent aid literatures as politically-driven (i.e. McKinlay & Little; McGillivray). Therefore,
if aid for counterterrorism aligns with their foreign policy agendas then, the uneven distribution
of their contributions makes sense. Donor-specific scholarship on the United States (Dreher &
Fuchs, 2011; Boutton, 2014; Sexton, 2016), Australia (Wynn-Pope, Zegenhagen, & Kurnadi,
2015), as well as other cross-national studies have shown increases in aid as part of
counterterrorism initiatives (Azam and Thelan 2008, 2010; Bandyopadhyay, Sandler & Younas
2010). Yet, counter to Figure 6, Dreher & Fuchs (2011) also found that France, Italy, and
Sweden decreased the amount of aid allocated to areas with terrorist insurgencies. The United
Kingdom uniquely concentrated aid to Anglophone areas over all others. This can be attributed
to British tendencies toward maintaining close connections to former colonial territories.

60
Figure 6: Amount of Aid Allocated by Bilateral Donors to Conflict-Affected Areas from 2017-2024a
Significantly, Australia, China, and Italy only allocated aid to the Far North. In 2021, the
Biya administration announced “deepening and diversifying” ties with Italy which may explain
the absence of aid to the government-contested Anglophone regions. Similarly, China just
reached a new economic deal with Cameroon in 2023, extending existing economic and military
cooperation (The Diplomat, 2023). In this case, there is clear evidence suggesting that the
absence of aid to Anglophone regions is intentional as China denounced international
involvement in the conflict at a UNGC briefing (UN, 2019). These instances support politicallydriven theories of aid allocation as a way to strengthen diplomatic relations with a recipient
government (Bearce & Tirone, 2012).
Existing evidence predicts such politically-driven decisions by bilateral donors. However,
beyond this initial inquiry, adequate attention has yet to be paid to the impact that these donor
decisions and recipient government agency have on humanitarian assistance. For example,
humanitarian aid organizations are resource dependent and are funded primarily by these
bilateral donors (Seybolt, 2009; Bermeo, 2017; Clark 2020). So how can humanitarian actors
remain neutral, impartial, independent, and humanitarian if they are dependent on politicallydriven donors? Additionally, how do these politically-driven decisions affect the implementation
and operation of aid projects? While addressing these questions is beyond the scope of this
paper, a sectoral study of education provides some insight into how implementation by
humanitarian actors results in the distribution of aid toward politically-preferred areas and away
from disfavored populations.
THE POLITICAL PROVISION OF EDUCATION AID
In both conflicts the provision of and access to education has been compromised by nonstate armed groups. According to a UNICEF report, 1.4 million children in Cameroon need
humanitarian assistance to access education (UNICEF, 2023). Thousands of schools have
remained closed in the conflict-affected regions due to direct attacks on education facilities,
conscription of child soldiers, and internal displacement of students. In each of the conflictaffected regions, closed education facilities have been used by non-state armed groups for
training, military housing, and operations (GCPEA, 2023). With between 33-41% of schools
closed in the Anglophone regions and recruitment of school-aged children by ISWAP increasing
in the Far North, international actors consider education to be an essential element of
humanitarian assistance (HRW, 2021).

61
While education had historically been recognized as part of the development realm, it
was integrated into humanitarian assistance in the 1990s and institutionalized by the early
2000s.26 Secure education facilities and uninterrupted instruction maintains safe spaces for
children to be provided essential services and protection during crises. Therefore, bilateral
donors and multilateral institutions alike earmark education funding in emergency response
(Burde et, al., 2017). The aid allocation data used above can be further specified to include only
sectorally-targeted humanitarian contributions. When the amount of aid per region is
disaggregated by disbursement, 32% of projects focused on education, while the rest targeted
health, food security, WASH, and protection. If aid were need-driven, then there would be a
similar number of education projects in the Far North and Anglophone areas considering the
similar number of displaced students (366,650 and 367,000 respectively in 2023). However, if
aid were politically-driven as argued above, then the Far North would still receive more aid and
attention reflecting the recipient government discouraging allocation to the disfavored
Anglophone regions. The data support politically-driven aid allocation showing only 2% of
education-targeting projects in the Anglophone regions compared to 20% in the Far North.
Such a staggering difference is significant, considering that aid to education has
particularly important implications in Cameroon. First, both conflicts have targeted schools,
interrupting instruction for long periods and endangering children. Second, both conflicts
separately have ideological concerns regarding the Cameroonian government’s provision of
education. In the Far North region, both Boko Haram and ISWAP have made education a
cornerstone of their fight against the imposition of neo-colonial Western values onto African
populations. Boko Haram, often translated as “ western education is forbidden”, has exploited
marginalized populations with already unequal access to education in Cameroon and Nigeria. By
capitalizing on existing social inequities, Boko Haram has suggested that state school systems
are inherently imperialist and also ideologically incompatible with Muslim tradition. ISWAP
similarly bans Western education but approaches recruitment differently, by providing public
services and Islamic schools as alternatives to the state system.
Education-targeting humanitarian assistance to the Far North is widespread and
comparatively well-funded. Such aid allocation may be in response to need as well as the
preferences of the recipient government. However, it is also possible that targeting education in
the terrorist-affected area serves donors in a third way. These projects protect Western-style
education institutions and may be seen by donors as a means to counter conservative religious
alternatives, thus containing the spread of popular support for extremist groups. Regardless of
which political motivation dominated the decision to allocate aid to the Far North, it is clear that
there were fewer political gains from giving the same attention to the Anglophone areas.
In the Anglophone Northwest and Southwest regions, education was a primary catalyst
for the separatist movement. In 2016, the Anglophone teacher’s union went on strike, boycotting
state expansion of Francophone curricula as a violation of the scholastic autonomy which the
regions had enjoyed since their incorporation into the country in 1961. Violent government
crackdowns escalated tensions into a separatist rebellion which has since splintered into the
costly regional conflict which persists today. As of 2024, there were more than 12 separatist
factions, several of which continue to enforce a boycott on education through violence. Fearing
attacks on campuses as well as conscription and gendered violence on the way to school, the
majority of children in the Anglophone areas remain out of school. Humanitarian organizations
have called for support to protect education in these regions, but appeals remain severely
underfunded with UNICEF only having met 1.3% of its education funding goals in 2023 despite
26 For a comprehensive history of the evolution of education into humanitarian spheres see Burde, Kapit-Spitalny,
Wahl, & Guven, 2011 and Burde et, al. 2017

62
having partnerships with 29 bilateral donors in the country. This marked absence of aid in
Anglophone areas as compared to the Far North provides further support for politically-driven
explanations of donor decision-making.
Politically-driven donors prioritize relations with recipient governments and their own
policy agendas over need. While both conflict contexts have similar needs (as measured by
numbers of displaced students), there are logistical and political obstacles to providing assistance
to education in the Anglophone areas. First, the Anglophone areas are politically disfavored by
the government and the imposition of the Francophone state education system is a central issue
in the conflict. If donors provided assistance to the existing Anglophone education system, they
would be working against the government and would likely sacrifice whatever political
concessions they aim to acquire from Biya’s administration. Second, education projects and the
international actors implementing them require close government cooperation to maintain a
functioning education system and impose sustainable change (GPE, 2023). The government is
likely unwilling to provide such assistance to support Anglophone institutions, especially
considering that the conflict began, in part, because of the government’s attempt to expand
Francophone influence in education. Third, the government has expressed disapproval of any
international involvement in these disfavored regions. Therefore, if donors were to allocate aid to
education in Anglophone areas they would similarly be unable to collect policy concessions from
the government.
The diversion of aid away from the Anglophone areas and toward the Far North can thus
be seen as a strategic decision by politically-driven donors. The comparison of these two
conflicts within one country emphasizes the often understated role which domestic politics can
play in aid allocation. The Biya administration’s response to the separatist crisis is a powerful
signal to politically-driven donors. If humanitarian assistance is indeed driven by the political
priorities of donors rather than need, then there is disincentive to allocate aid to politically
disfavored areas and aid will flow according to political preferences rather than to populations in
need. This study provides insight into this decision by donors within Cameroon, but begs
additional subnational data be collected in other country contexts and more in-depth studies be
conducted.
CONCLUSION & CONTRIBUTION
While altruists and optimists view foreign aid as a means to protect and provide for
vulnerable populations, research shows that the cross-national distribution of foreign aid is often
driven instead by the political agendas of governments than by the needs of recipients. However,
donors maintain that humanitarian aid differs from development aid and is indeed driven by the
needs of populations experiencing crisis or conflict. With few comprehensive or within-country
studies of the allocation of humanitarian aid specifically and its susceptibility to political
interference, this assumption of altruistic humanitarian aid has remained largely untested. By
focusing on humanitarian aid and its subnational allocation in a conflict context, this study
contributes empirical evidence suggesting that despite humanitarian mandates, aid can still be
influenced by politics. In looking subnationally at how assistance is allocated across two
conflicts within the same country, analysis of Cameroon shows how the political agendas not
only of donors but also of recipient governments matter.
In Cameroon, a terrorist insurgency and armed separatist conflict continue to deprive
populations in conflict-affected areas of basic public services from education to food security
and safety. While the international community has responded to the crises, donors have done so
inequitably. Humanitarian assistance is concentrated in the terrorist-affected Far North, with less

63
aid flowing to the separatist Anglophone regions despite intensity of need. As the recipient
government welcomes international intervention in the Far North, President Biya’s restrictive
regime deters involvement in the separatist areas. Therefore, donors have to choose. They can
maintain the humanitarian mandate of impartiality, neutrality, and independence and allocate aid
according to need, potentially sacrificing policy concessions from the recipient government. Or
donors can choose to concentrate aid to the Far North, intentionally ignoring the intensity of need
in the Anglophone areas and sustain political favor with the president. Most donors choose to
allocate aid to the Far North and protect their own political interests. While some still provided
assistance to the Anglophone regions, almost all donors allocated more to the Farth North.
Donors who had close alliances or were negotiating economic deals with the Cameroonian
government did not allocate any aid to the Anglophone regions.
This response of donors to recipient government demands and political disfavor of
populations calls into question the principles of humanitarian aid and altruistic assumptions of its
allocation. The limited evidence that exists testing humanitarian assistance across countries in
conflict, demonstrates politically-driven distribution of humanitarian aid. However, these studies
focus only on the incentives of donors, omitting the important role which recipient governments
play in the allocation of aid. In this way, a subnational study is able to add to existing evidence
showing how even within a country, humanitarian assistance can be diverted away from
politically disfavored populations experiencing conflict. Altruistic actors interested in ensuring
that assistance reaches those who need it most, thus need to consider how political agendas of
both donors and recipient governments dominate decision-making as to where aid is allocated.

64
BIBLIOGRAPHY
Bearce, D. H., & Tirone, D. C. (2012). Foreign aid effectiveness and the strategic goals of donor governments.
Journal of Politics, 74 (1), 276-289. https://doi.org/10.1017/S0022381611001458
Bermeo, N. (2017). Development and Crisis in Post-Colonial States: Foreign Aid in Haiti and Morocco .
Cambridge University Press.
Bermeo, S. B. (2017). Aid is not oil: Donor preferences, governance, and the efficacy of foreign aid. World
Development, 99, 371-383. https://doi.org/10.1016/j.worlddev.2017.05.012
Berthelemy, J.-C., & Tichit, A. (2004). Bilateral donors’ aid allocation decisions: A three-dimensional panel
analysis. International Review of Economics & Finance, 13 (3), 253-274. https://doi.org/10.1016/j.iref.2004.03.005
Boutton, A. (2014). Securitizing development: USAID, donor fragmentation, and the securitization of US foreign
aid. Security Studies, 23 (1), 167-200. https://doi.org/10.1080/09636412.2014.874717
Briggs, R. (2014). Aid and Influence: Do Donors Help or Hinder? (2nd ed.). Routledge.
Burde, D., Guven, O., & Vaki, A. (2017). The impact of aid on education during conflict and post-conflict
transitions. International Journal of Educational Development, 52, 45-57.
https://doi.org/10.1016/j.ijedudev.2016.08.001
Burnside, C., & Dollar, D. (2000). Aid, policies, and growth. American Economic Review, 90 (4), 847-868.
https://doi.org/10.1257/aer.90.4.847
Claessens, S., Cassimon, D., & Van Campenhout, B. (2009). Aid Effectiveness and the Choice of Aid Modalities.
World Bank Publications.
Clark, A. M. (2020). Foreign aid and the pursuit of development. Oxford Research Encyclopedia of Politics.
https://doi.org/10.1093/acrefore/9780190228637.013.1380
Cooley, A., & Ron, J. (2002). The NGO scramble: Organizational insecurity and the political economy of
transnational action. International Security, 27 (1), 5-39. https://doi.org/10.1162/016228802320231155
Dollar, D., & Levin, V. (2006). The Increasing Selectivity of Foreign Aid, 1984–2003. World Development, 34(12),
2034-2046. https://doi.org/10.1016/j.worlddev.2005.11.001
Drury, A. C., Olsen, T. D., & Van Belle, D. A. (2005). Testing Theories of U.S. Foreign Policy: Comparing Foreign
Aid Decisions Across Presidents. Political Research Quarterly, 58(2), 187-202.
https://doi.org/10.1177/106591290505800201
Duffield, M. (2008). Development, security and unending war: Governing the world of peoples. Polity .
Easterly, W. (2007). The White Man’s Burden: Why the West's Efforts to Aid the Rest Have Done So Much Ill and
So Little Good . Penguin Press.
EU Consensus. (2007). European Consensus on Humanitarian Aid. Retrieved from
https://ec.europa.eu/echo/what/humanitarian-aid/european-consensus_en
Findley, M. G., Powell, J., & Strandow, D. (Eds.). (2011). Social Capital and Development Trends in Rural Areas:
Vol 23 . Springer Science & Business Media.
French Ministry for Europe and Foreign Affairs. (2024). Humanitarian aid and crisis response. Retrieved from
https://www.diplomatie.gouv.fr/en/
Fuchs, A., Dreher, A., & Nunnenkamp, P. (2014). Determinants of donor generosity: A survey of the aid budget
literature. World Development, 56 , 172-199. https://doi.org/10.1016/j.worlddev.2013.10.014

65
GCPEA. (2023). Education under attack 2023: Cameroon. Global Coalition to Protect Education from Attack.
https://protectingeducation.org/sites/default/files/documents/eua2023_cameroun_en.pdf
GPE. (2023). Cameroon Country Program. Global Partnership for Education.
https://www.globalpartnership.org/country/cameroon
HRW. (2021). Cameroon: Schools under attack in Anglophone regions. Human Rights Watch.
https://www.hrw.org/news/2021/03/29/cameroon-schools-under-attack-anglophone-regions
IASC. (1994). Review of principles of humanitarian action: Challenges, response, strategy . Retrieved from
https://interagencystandingcommittee.org/
Kevlihan, R., DeRouen, K., & Biglaiser, G. (2014). When politics prevails: Security Council members and sanctions
against rogue states. International Studies Quarterly, 58 (2), 317-328. https://doi.org/10.1111/isqu.12100
King Salman Unified National Platform. (2023). Saudi Arabia humanitarian assistance efforts. Retrieved from
https://unifiednationalplatform.gov.sa/en/
Knack, S. (2015). Aid, institutions, and governance: What have we learned? European Journal of Development
Research, 27 (4), 605-624. https://doi.org/10.1057/ejdr.2015.22
Lancaster, C. (2005). Foreign Aid: Diplomacy, Development, Domestic Politics . University of Chicago Press.
Maizels, A., & Nissanke, M. (1984). Motivations for aid to developing countries. World Development, 12 (9), 879-
900. https://doi.org/10.1016/0305-750X(84)90030-5
Masaki, T. (2018). The value of data in development. Development Finance and Institutions: Theory and Policy ,
93-108.
McKinlay, R. D., & Little, R. (1978). A foreign policy model of U.S. bilateral aid allocation. World Politics, 30 (1),
58-86. https://doi.org/10.2307/2009975
Narang, N. (2016). How power politics and international economic structures affect humanitarian aid allocation.
International Studies Quarterly, 60 (3), 536-547. https://doi.org/10.1093/isq/sqw015
Neumayer, E. (2003). The pattern of aid giving: The impact of good governance on development assistance.
Routledge .
Norwegian Government. (2023). Norwegian development aid strategies. Retrieved from
https://www.regjeringen.no/en/dep/ud/id200/
NRC. (2024). Neglected displacement crises report. Retrieved from https://www.nrc.no/
Nunnenkamp, P., Öhler, H., & Thiele, R. (2009). Donor Coordination and Specialization: Did the Paris Declaration
Make a Difference? The Journal of Development Studies, 45(8), 1264-1280.
https://doi.org/10.1080/00220380902862062
OECD. (2010). Development Assistance Committee of the Organisation for Economic Co-operation and
Development (OECD) . Retrieved from https://www.oecd.org/dac/
Peter, M., Nunnenkamp, P., & Öhler, H. (2016). Need, Merit, or Self-Interest—What Determines the Allocation of
Aid? Review of World Economics, 152(1), 177-198. https://doi.org/10.1007/s10290-015-0235-6
Sachs, J. D. (2006). The End of Poverty: Economic Possibilities for Our Time . Penguin Press.
Seybolt, T. B. (2009). Humanitarian military intervention: The conditions for success and failure. Oxford
University Press.

66
The Diplomat. (2023). China extends economic and military cooperation with Cameroon.
https://thediplomat.com/2023/04/china-extends-economic-and-military-cooperation-with-cameroon/
Thiele, R., Nunnenkamp, P., & Dreher, A. (2007). Do Donor Countries Target Aid in Line with Motivations of the
Recipient Country’s Government? An Empirical Analysis of ‘Good’ and ‘Bad’ Aid. Development Policy Review,
25(5), 531-551. https://doi.org/10.1111/j.1467-7679.2007.00383.
UN. (2019). Report of the UN General Committee on Cameroon. United Nations.
https://undocs.org/en/A/RES/74/245
UNOCHA. (2023). Humanitarian response plan funding report. Retrieved from https://www.unocha.org/
US Department of State. (2023). United States foreign assistance goals. Retrieved from https://www.state.gov/
Vaux, T. (2006). Risking humanitarian principles. International Review of the Red Cross, 88 (862), 747-758.
https://doi.org/10.1017/S1816383106001305
Whittall, J. (2015). A global justice perspective on human rights. Development in Practice, 15 (3-4), 388-396.
https://doi.org/10.1080/0961452032000094453
Ziff, A. (2023). Aid accessibility: Equity & education in Kenya. International Journal of Educational Development,
99, 102772.

67
ARTICLE 3: APPENDIX A
ADDITIONAL REGRESSION ANALYSES
Need Indicators
Table A.1: Regressions on Need Indicators and Logged Amount of Aid

Abstract (if available)

Abstract This three-article dissertation project analyzes the subnational allocation of foreign aid. I empirically test whether aid is need-driven and allocated to the places and populations with the highest intensity of need, or whether aid is politically-driven and allocated according to the political agendas of donors and recipient governments. In finding that all aid appears to be plagued by politics, I demonstrate how recipient government agency can influence donor decisions, resulting in either the mitigation or exacerbation of existing inequities. These studies collectively contribute to the literature by incorporating recipient government agency, an increasingly important and notably neglected determinant of aid into an area of research dominated by the perspectives of Western donors. If altruistic actors are able to appropriately identify, acknowledge, and assess recipient government power and preference, they can overcome obstacles to extending equitable access to aid amongst politically disfavored populations.

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

The role of local actors in international development: how incentive types affect foreign aid effectiveness

PDF

Three essays on the political economy of federal spending

PDF

By word, not deed: advancing a theory of rhetorical compliance in international law

PDF

Assessing modern conflict to monitor human rights with remote sensing: Russia's war in Ukraine

PDF

Personal agency and educational decision-making: insights from middle- and emerging-middle-income school parents in Costa Rica

PDF

Political determinants and economic effects of HIV/AIDS: a push for the multisectoral approach

PDF

The political psychology of scarcity in American society: a study of culture, discourse, and public opinion

PDF

The strategy of firms towards political elites and consumers in polarized western democracies

PDF

Forced migration as a cause and consequence of conflict

PDF

Humanitarian aid and the production of spatial knowledge and practices in the U.S.-Mexico borderlands

PDF

AID scanning & catalysis and the generation of high-affinity antibodies

PDF

Taking orders from Brussels? external actors and regional organizations in the developing world

PDF

Are you as old as you feel? A quantitative and qualitative analysis of leader age and foreign policy decision-making

PDF

A theory of status-quo terrorism: democracies in conflict and their proclivity to outsource repression

PDF

Transformation of poly(vinyl chloride) into tunable polyethylene products with branches or arene substituents

PDF

Selling virtue: how human rights NGOs and their donors work together to create a better world ... for themselves

PDF

Toward a more perfect liberalism: perfectionism in Kantian political philosophy

PDF

Redlining, neighborhood change, and individual outcomes: an exploration of how space shapes the landscape of inequality from past to present

PDF

Identity salience and political violence: centering social identity in the study of intrastate conflict

PDF

Units of agency in ethics

Asset Metadata

Creator Ziff, Alix Rebecca (author)

Core Title The allocation and accessibility of aid: the role of recipient government agency in dictating donor decisions

School College of Letters, Arts and Sciences

Degree Doctor of Philosophy

Degree Program Political Science and International Relations

Degree Conferral Date 2024-08

Publication Date 08/07/2024

Defense Date 07/30/2024

Publisher Los Angeles, California (original), University of Southern California (original), University of Southern California. Libraries (digital)

Tag AID,autocratic politics,Development,foreign aid,Foreign policy,humanitarian aid,international organizations,international political economy,terrorism

Format theses (aat)

Language English

Contributor Electronically uploaded by the author (provenance)

Advisor Graham, Benjamin (committee chair), James, Patrick (committee member), Lo, James (committee member), Schroeder, Mark (committee member)

Creator Email alixziff@gmail.com,aziff@usc.edu

Permanent Link (DOI) https://doi.org/10.25549/usctheses-oUC113998TAO

Unique identifier UC113998TAO

Identifier etd-ZiffAlixRe-13353.pdf (filename)

Legacy Identifier etd-ZiffAlixRe-13353

Document Type Dissertation

Format theses (aat)

Rights Ziff, Alix Rebecca

Internet Media Type application/pdf

Type texts

Source 20240807-usctheses-batch-1194 (batch), University of Southern California (contributing entity), University of Southern California Dissertations and Theses (collection)

Access Conditions The author retains rights to his/her dissertation, thesis or other graduate work according to U.S. copyright law. Electronic access is being provided by the USC Libraries in agreement with the author, as the original true and official version of the work, but does not grant the reader permission to use the work if the desired use is covered by copyright. It is the author, as rights holder, who must provide use permission if such use is covered by copyright.

Repository Name University of Southern California Digital Library

Repository Location USC Digital Library, University of Southern California, University Park Campus MC 2810, 3434 South Grand Avenue, 2nd Floor, Los Angeles, California 90089-2810, USA

Repository Email cisadmin@lib.usc.edu