Hochschule für Politik

an der Technischen Universität München

Politik der internationalen Wirtschaftsbeziehungen
POL 11000 - IPE

Foreign Aid and Other Financial Flows

into the Developing World
Entwicklungshilfe und andere Finanzflüsse
in den Globalen Süden

Prof. Dr. Tim Büthe

28 January 2021
 Remittances
• US$689b; US$528b to developing countries (2018 est.)
– exceeds private commercial flows (FDI)
– exceed government "development assistance" foreign aid
• Counter-cyclical
• Other Distinctive Features:
– senders
(private individuals: transitory migrant workers, emigrants, 2nd generation)
– recipients
(private individuals/families, local communities, widely distributed)
– channels
(money transfer; often informal channels)
– uncertainty about data
(no consistent reporting/estimating)
(c) Tim Büthe/HfP-TUM, 2018
 Foreign Aid
• Military Aid
• Humanitarian/Emergency Aid (ostensibly short-term)
• Development Aid (consciously long-term)
• Technical Assistance/Administrative Training
• including for:
– police forces
– bureaucratic organization/administrative capacity
(statistical offices)
– economic policymaking
(forecasting, central banking, antitrust enforcement)
– engineering
(esp. infrastructure: roads, bridges, etc.)
– political institution-building, esp. in transition to democracy
(building political parties, organizing elections)
Importance of Aid for Developing Countries

Source: OECD DAC

Who Gives …

www.oecd.org/dac
 Political Use of Aid:
Examples

• U.S. aid to Egypt (post-Camp David Peace Accord)

• Australian aid to Nauru (2000-2008)
• Taiwan aid to Nauru pre-2002
• PR China aid to Nauru (2002ff)
• Germany to ____ (since ___)
Why Do Countries Give Foreign Aid?

Cartoon by Michael Crawford,

from New Yorker 10 Sep 2001
 Aid to Developing Countries:
Reasons to Give …
• Donor's Physical Security
• Donor's Power/Political Influence
(positive sanctions or potential for negative sanctions)
• Donor's Economic Self-Interest
- developing market abroad for exports
- hidden subsidies for donor country producers (esp. tied aid)
• Global Public Goods Provision out of Enlightened Self-Interest
(protection of global commons, peace/stability, eradicate diseases)
• Donor's Reputation, Self-Affirmation
(incl. declamatory politics)
• Sense of Obligation/Duty
(historical or positional)
• Altruism, Pity
(improving the human well-being/welfare of others w|out personal gain)
 Aid to Developing Countries:
… and Reasons to Oppose Foreign Aid
• Moral Hazard
• Aid Dependency "Trap"
• Donor's Power/Political Influence
(positive sanctions or potential for negative sanctions)
• Donor's Economic Self-Interest
- esp. hidden subsidies for donor country producers (tied aid)
• Donor's Reputation, Self-Affirmation
(incl. declamatory politics)
• Sense of Obligation/Duty
(historical or positional)
 Your Opinion:

Should Germany ...

- increase its foreign aid,
- keep it the same, or
- decrease its foreign aid?
 German Foreign Aid
to Countries in the Global South
 German Foreign Aid
to Countries in the Global South
[ Image omitted ]
 Aid to Developing Countries:
The Numbers …
• 0.7% of GDP/GNI target for "official" development aid
(agreed by OECD countries in 1970; reconfirmed repeatedly)

• OECD-DAC Countries that met target in 2019 (prelim data):

Luxembourg (1.05), Norway (1.02), Sweden (0.99), Denmark (0.71), UK (0.7)

• Average in 2019 : 0.38% (2015: 0.30%)

• Below 0.3% in 2019:
Japan (0.29), New Zealand (0.28), Austria (0.27), Canada (0.27),
Italy (0.24), Australia (0.22), Iceland (0.24), Hungary (0.22),
Spain (0.21), Portugal (0.16), Slovenia (0.16), USA (0.16),
South Korea (0.15), Greece (0.14), Czech Republic (0.13),
Poland (0.12), Slovak Republic (0.12)
 Constant 2018 USD billion

20
40
60
80

0
100
120
140
160
180
1960
1961
1962
1963
1964
1965
1966
1967
1968
1969
1970

Source: OECD, 16 April 2020.

1971
1972
1973
1974
1975
1976
1977
1978
1979
1980
1981
1982
1983
1984
1985
1986

(a) Total DAC excludes debt forgiveness of non-ODA claims in 1990, 1991 and 1992.
1987
1988
1989
1990 (a)
1991 (a)
1992 (a)
1993
1994
1995
1996
ODA/GNI

1997
(right scale)

1998
1999
Data for 2019 are preliminary

2000
2001
2002
2003
2004
Total ODA
(left scale)

2005
2006
2007
2008
2009
2010
2011
CHART 4: TRENDS IN NET ODA FLOWS BY DAC COUNTRIES

2012
2013
2014
2015
2016
2017
2018
2019
0.00
0.10
0.20
0.30
0.40
0.50
0.60

ODA as percent of GNI

What Determines Government Aid Allocation?
Findings from Statistical Analyses

• Recipient need (inconsistent, sometimes not significant)

• Military alliances
• Geopolitical strategic importance (esp. during Cold War)
• UN voting record (prospectively or retrospectively)
• UN Security Council membership
• Trade (export opportunities for donor country)
• "Good governance" (low corruption, democracy, after CW)
 Aid to Less Developed Countries:
Useful Distinctions
• Assistance given directly (bilateral aid) or indirectly
indirectly: via international organization, such as World Bank, EU
• different "channels of delivery"
• grants vs. loans
• "official" (governmental) vs. private
OECD definition of "development assistance:" "[Financial/material] flows
to developing countries and multilateral institutions provided by official
agencies, including state and local governments or by their executive
agencies, each transaction of which meets the following tests: a) it is
administered with the promotion of the economic development and welfare
of developing countries as its main objective, and b) it is concessional in
character and contains a grant element of at least 25 per cent."
(OECD, 1969)
(c) Tim Büthe/HfP-TUM, 2018
 IOs & Economic Development
• International Bank for Reconstruction and
Development (IBRD, “World Bank”)
• UNCTAD + ~30 UN Agencies
• Changing Notion of "Development"
and Changing Scope of Activities
 Beyond Economic Development
Critiques & Shifts in I(NG)O Focus

• GDP Growth vs. Improvement in Standards of Living

• Industrialization vs. Agricultural Development
• Sustainable Development

• From Economic to Human Development

- individual vs. aggregate changes/effects
- distribution of income/wealth
- political liberty and freedom
- health
- social fabric
 What Determines IGO Aid Allocation?
Findings from Statistical Analyses

• Recipient need (more strongly than bilateral aid)

• UN Agencies: human as well as economic development need
• Region/geography (mostly overt focus of the IGO)
• UN Agencies: remote regions
• World Bank, IMF: UN Security Council membership
• "Good governance" (low corruption, democracy, after CW)
(c) Tim Büthe/HfP-TUM, 2018
 Private Foreign Aid

Based on:
Tim Büthe, Solomon Major, and Andre de Mello e Souza,
"The Politics of Private Foreign Aid: Humanitarian
Principles, Development Objectives, and Organizational
Interests in the Allocation of Private Aid by NGOs."
 Private Actors in Foreign Aid

• Transnational Aid NGOs as "Channels of Delivery"

• Transnational Aid NGOs as Agenda Setters
(Aid Advocacy NGOs)
• Private Sources of Aid Funds
- Philanthropic Foundations (Gates Foundation etc.)
- Business Charitable Giving (corporate philanthropy)
- Donations by Individuals

• Transnational and Local Aid NGOs as

Fully Private Service Providers
Bilateral Government Aid Dispersed via NGOs

0.7
USA
ODA channeled via NGO/Total ODA %

Japan
0.6

Germ.
Neth.
Sweden
0.5

Korea
0.4
0.3
0.2
0.1
0.0

2004 2005 2006 2007 2008 2009 2010 2011

Year
Source: Büthe and Cheng, 2013: 325
Transnational Private Philanthropy:
Foundations as Aid Actors
• Int'l Programs of Rockefeller, Ford, and Carnegie:
- funding for applied research;
- financing U.S. study of students from LCDs
- shaping development discourse
• "New" Foundations of "Mega-Philanthropists":
- highly specific project-funding, often massive
- indirect agenda-setting (e.g., Gates' $10billion for
vaccines against diarrhea, pneumonia, malaria, )
• fast-growing but largely unexamined role of non-
U.S. philanthropic foundations' int'l programs
 Transnational Aid NGOs as
Fully Private Actors

• Funds collected from private sources, often as a large

number of small donations by private individuals
• Allocation decision made by the NGOs independently of
any "home" governments
• Aid delivery by the transnational NGOs or local non-
governmental actors in the recipient countries
 Private Development Aid:
2 Questions
• Private-source Humanitarian/Development Aid !
• No Systematic, Aggregate Data

1.What Is the Allocation of Private Development Aid

Across Countries?
 Research Project
• 4 Inter-related development issue areas:
- health
- education
- water/sanitation/sewerage
- emergency
• Expert surveys: Largest, most important NGOs/issue
- not-for-profit
- operational; carrying out on-the-ground projects
- US-based or US fundraising

➡ 48 NGOs
• Data for 40 of 47 (annual); 2001 total: $ 1.612 billion
➡ aggregate, per capita private aid by country
for 119 low/lower-middle income countries (WB definition)
 Descriptive Statistics of the Data
Private Aid Allocation per Country

mean std. dev. min max

Private Aid per capita $1.09 $1.56 0 $9.28

Total (per country) $9.9 mil $12.9 mil 0 $61.5 mil

3 countries received no U.S. private-source aid in 2001:

- Bhutan
- Maldives
- Marshall Islands
Global Distribution of Private Development Aid
 Private Development Aid:
2 Questions
• Private-source Humanitarian/Development Aid !
• No Systematic, Aggregate Data

1.What Is the Allocation of Private Development Aid

Across Countries?

2.What Explains this Allocation of Private Aid?

• Demand > Resources
 Why?
3 Perspectives on NGO Private Aid Allocation
(1) Humanitarian Hypothesis:
• IR Constructivism: Norms " identities/roles " interests
• NGOs as "principled issue-networks"; discourses as source of roles
• Strong humanitarian discourse: imperative to help poor and needy;
➡ H1: Allocation driven by objective recipient need.
 Humanitarian NGO Discourse
• Seeking "to relieve the suffering of individuals, solely
guided by their needs" without consideration of other
criteria such as "nationality, race, religious beliefs,
class or political opinions."
(Statutes of the Int'l Red Cross and Red Crescent Movement, 1986)

• Help "the impoverished and disadvantaged

... based solely on need, regardless of their race,
religion or ethnicity" (Catholic Relief Services U.S., 1990, 2007)
• We focus on "needs ... of individuals and families in
the poorest communities in the world" (Care USA, 2001)
• "Serving the World's poorest [and] most vulnerable"
(World Vision USA, 2003)
 Why?
3 Perspectives on NGO Private Aid Allocation
(1) Humanitarian Hypothesis:
• IR Constructivism: Norms " identities/roles " interests
• NGOs as "principled issue-networks"; discourses as source of roles
• Strong humanitarian discourse: imperative to help poor and needy;
➡ H1: Allocation driven by objective recipient need.
(2) Development Hypothesis:
• Contested consequentialist discourse, with emphasis on aid effectiveness
and efficiency of outcomes;
➡ H2: Allocation driven by likely development effectiveness.
 Development NGO Discourse
• "Care strives to attack the underlying causes of
poverty through health and and education
projects" (CARE USA, 2005)
• "Enabling the world's children to realize their god-
given potential by tackling the root causes of poverty"
(World Visions 2005)

But: contested!
[ Image omitted ]
 Why?
3 Perspectives on NGO Private Aid Allocation
(1) Humanitarian Hypothesis:
• IR Constructivism: Norms " identities/roles " interests
• NGOs as "principled issue-networks"; discourses as source of roles
• Strong humanitarian discourse: imperative to help poor and needy;
➡ H1: Allocation driven by objective recipient need.
(2) Development Hypothesis:
• Contested consequentialist discourse, with emphasis on aid effectiveness
and efficiency of outcomes;
➡ H2: Allocation driven by objective recipient need.
(3) Aid Allocation as Fundraising Strategy:
• Organizational structure & competition " instrumental pursuit
of material resources and managerialist growth of the organization;
• Project selection to maximize fundraising | shifting public attention;
➡ H3: Allocation driven by media-driven perceived need.
 Operationalization (1)
• Measures of Aid Recipients' Objective Need:
• Level of economic development: GDP per capita
• Share of Population Living on <$1/day ($2/day)
• Share of Population Below Local Poverty-Line
• Human Development Index
• Physical Quality of Life Index
• Human Poverty Index

• Expectation: The richer/better-off the country, the

less aid should it receive (per capita)
 Initial Tests of the Humanitarian Hypothesis
(Alternative Measures of Objective Need)
GDP Human Develop Phys.Quality of pop living pop living Human Poverty
per capita ment Index Life Index < $1/day < local pov line Index

[OBJECTIVE –0.517*** –2.81*** –0.0280*** 0.0159** 0.0270*** 0.0241**

NEED] (.142) (.841) (.00698) (.00623) (.00805) (.0106)

Population –0.312*** –0.330*** –0.240*** –0.593*** –0.520*** –0.373**

(logged) (.0700) (.0691) (.0658) (.0973) (.0868) (.0960)

Christians 0.677* 0. 716** 0.882*** 0.707 0.549 0. 595

(as portion of pop) (.367) (.360) (.335) (.438) (.441) (.427)

Latin America 1.61*** 1.43*** 1.52*** 1.26*** 1.04** 1.69***

(.321) (.337) (.350) (.416) (.397) (.425)

constant 7.08*** 5.62*** 4.12*** 8.12*** 6.38*** 4.08**

(1.60) (1.26) (1.17) (1.69) (1.62) (1.74)

N 110 107 116 74 71 86

R2 0.4004 0.4003 0.4047 0.5216 0.5624 0.4062

Note: OLS estimates with Huber-White robust standard errors in parentheses; * p < 0.1; ** p < 0.05; *** p < 0.01, two-tailed tests. The Human
Development Index (HDI) is scaled from 0 to 1. HPI is encoded such that higher values indicate greater poverty.
 Operationalization (2)
• Measures of Expected Effectiveness/Efficiency:
• [Control of] Corruption, as measured by
- the World Bank
- International Country Risk Group
- Transparency International
• [Absence of] Civil War, Inter-State War
• [Absence of] Political Instability & Political Violence

• Expectation: The more conducive the socio-

political conditions are to effective/efficient use of
aid, the more aid should it receive (per capita)
 Tests of the Development Hypothesis
(Alternative Measures of Expected Aid Effectiveness)
WGI ICRG TI
Pol. Instability
Corruption Index Corruption Index Corruption Perception

Objective Need –0.486*** –0.606*** –0.393* –0.544***

(GDP per capita) (.149) (.184) (.201) (.143)

Corruption Index –0.204 0.217 –0.157

("control" of corruption) (.231) (.180) (.144)

Political –0.161*
Violence/Instability (.0832)

Population –0.320*** –0.430*** –0.507*** –0.238***

(logged) (.0708) (.101) (.117) (.0836)

Christians 0.686* 1.31*** 1.11** 0.852**

(as portion of pop) (.364) (.459) (.491) (.344)

Latin America 1.60*** 1.15*** 1.26*** 1.71***

(.321) (.379) (.434) (.296)

constant 6.92*** 8.87*** 9.97*** 5.80***

(1.60) (2.10) (2.51) (1.83)

N 110 75 62 109
R2 0.4048 0.5276 0.5385 0.4308
Note: OLS estimates with Huber-White robust standard errors in parentheses; * p < 0.1; ** p < 0.05; *** p < 0.01, two-tailed tests.
 Operationalization (3)
• Measures of Subjective Perceptions of Need:
• Index of media coverage for each country (114/119)
• based on qualitative content analysis of New York
Times coverage (25,463 stories, LexisNexis keyword
searches)
+ TV News index based on Vanderbilt Archive
• higher values = more frequent portrayal as place
in need of humanitarian or development aid
• 5 alternative ways of constructing the index
• Expectation: The greater the media-driven donor
perception of need, the higher will aid allocation
be
 Tests of Allocation-For-Fundraising Hypothesis

Model 15 Model 16 Model 17

Objective Need –0.578*** –0.585***

(GDP per capita) (.142) (.144)

Political –0.176*
Violence/Instability (.0854)

Media Coverage 0.0619 0.0531 0.0687

(New York Times Index 1) (.0754) (.0689) (.0669)

Population –0.302*** –0.312*** –0.248***

(logged) (.0783) (.0751) (.0893)

Christians 0.585 0.805** 0.876**

(as portion of pop) (.400) (.384) (.361)

Latin America 1.27*** 1.60*** 1.73***

(.310) (.310) (.293)

constant 3.48*** 7.33*** 6.08***

(1.27) (1.60) (1.85)

N 106 106 106

R2 0.3291 0.4236 0.4477
Note: OLS estimates with Huber-White robust standard errors in parentheses; * p < 0.1; ** p < 0.05; *** p < 0.01, two-tailed tests.
 Key Findings
Change in per capita private aid that results
Hypothesized from a 1 std dev from a 1 std dev
Explanatory decrease increase
Variable Mean from the mean from the mean

Economic
US$ 748 (p.c. GDP) + 70.1¢ – 40.9¢
development [+26.6¢ +$1.24] [-65.4¢ -19.8¢]

Political + 35.1¢ – 24.0¢

< 1 event/year
Violence/Inst [-2.5¢ +87.3¢] [-53.2¢ +4.6¢]

Media 6 (net) stories – 13.3¢ + 17.0¢

coverage suggesting need [-41.3¢ +18.5¢] [-16.9¢ +61.7¢]

Based on OLS regressions that also controlled for:

- Population size
- Christian population
- Latin America Notes:
Grey indicates: not statistically significant (we cannot be sure that the estimate captures real effect)
In brackets: 95% confidence intervals (with 95% probability, actual effect is between the min and max)
Estimates based on simulations using CLARIFY (King, Tomz & Wittenberg 2000)
 Other Factors
• Immigrants from potential recipient country in U.S. (+)
• Recipient country government corruption (–)
• U.S. government development aid (–)
• U.S. military aid (–)
• U.S. military personnel, military alliance (+)
• UN Security Council membership (–)
• Trade (with U.S.: –)
• U.S. government sanctions (–)
 More Robustness Checks
• Alternative Measures of Media Attention
• need-suggesting stories, only
• weighted index of media attention
• # of stories, irrespective of content
• all stories identified through Lexis-Nexis keyword searches
(regardless of relevance)

• Lags of Media Attention/Coverage

• Dummy Variables for Outliers/Influential Points
(≈ Selective Exclusion of Observations)
• Analysis without Latin America
• Reconceptualization of Allocation Decision
as 1-Step Decision ...
Lags of Media Coverage/Attention (NY Times)

Estimated Coefficients for Media Attention

with hd_index_2001, ln_nyt_neg_2001, 95% conf. interval
.4
.3
.2
.1
0
.1

0 2 4 6 8 10 12
Lag in Months

NYT_coefficient upperbound
lowerbound
 Lags of Media Coverage (TV News)

Estimated Coefficients for TV News Attention

with hd_index_2001, ln_tv_neg_2001, 95% conf. interval
.3
.2
.1
0
.1

0 2 4 6 8 10 12
Lag in Months

TVNews_coefficient upperbound
lowerbound
 Conclusions
• Private Development Aid:
• Private aid is substantively significant
• Allocated aid per capita varies greatly across recipient countries

• Key Analytical Findings:

• Little support for "cynical" hypothesis
• Strong and robust support for "idealistic" hypothesis

• Implications:
• Norms can outweigh material/organizational incentives
• NGOs/Private Actors: different logic of global governance
• Private development aid superior to government aid?
• Statistical testing of constructivist arguments
 Thank you

Tim Büthe
tim.buthe@hfp.tum.de

(c) Tim Büthe/HfP-TUM, 2018