Meta-Review of Evaluations of Development Assistance
to Afghanistan, 2008 – 2018
March 2020 | Christoph Zürcher
Meta-Review of Evaluations of Development Assistance
to Afghanistan, 2008 – 2018
March 2020 | Christoph Zürcher
On behalf of
Federal Ministry for Economic Cooperation and Development (BMZ), Germany
Division for Afghanistan and Pakistan
Introductory remarks. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5
Preface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6
1 Introduction: Will We Ever Learn? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7
2 Methodology of the Meta-Review. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
3 The Evidence Base . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12
4 Findings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14
4.1 Contextual Factors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14
4.1.1 Security . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14
4.1.2 Local Political Economy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15
4.1.3 Capacity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15
4.2 OECD DAC Criteria. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16
4.2.1 Relevance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16
4.2.2 E ciency . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16
4.2.3 E ectiveness and Impacts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16
4.2.4 Sustainability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17
4.3 Findings by Sector . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17
4.3.1 Governance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17
4.3.2 ARTF . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18
4.3.3 Sub-national Governance (including NSP). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19
4.3.4 Stabilization (including CERP) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19
4.3.5 Education . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20
4.3.6 Health . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20
4.3.7 Gender . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20
4.3.8 Sustainable Economic Development . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21
4.3.9 Infrastructure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21
4.3.10 Capacity Building . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21
4.3.11 Monitoring and Evaluation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22
4.4 Some Lessons to Learn . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23
4.4.1 What is relevance, or the fallacy of a “needs-based” approach . . . . . . . . . . . . . . . . 23
4.4.2 Taking the local context seriously. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23
4.4.3 Modest and slow is better. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24
4.4.4 Aid in insecure regions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24
4.4.5 We cannot work in the dark . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25
5 Critical Reﬂ ections . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26
6 Bibliography. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28
Imprint . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39
Abbreviations and Acronyms
ADB Asian Development Bank
ARAP Afghanistan Rural Access Program
ARTF Afghanistan Reconstruction Trust Fund
ASGP Afghanistan Sub-National Governance Project
BMZ Bundesministerium für wirtschaftliche Zusammenarbeit und Entwicklung/
The Federal Ministry for Economic Cooperation and Development, Germany
CERP Commanders Emergency Response Program
DDA District Development Assemblies
EQUIP Education Quality Improvement Program
EVAW Eliminating violence against women
IDLG Independent Directorate of Local Governance
IEC Independent Electoral Commission of Afghanistan
IRDP Irrigation Restoration and Development Project
LOT FA Law and Order Trust Fund
MBAW Making Budgets and Aid Work
NDPG National State Governance Project
NGO Non-Governmental Organization
NIBP National Institution Building Project
NSP National Solidarity Program
SIGAR Special Inspector General for Afghanistan Reconstruction
UNDP United Nations Development Program
TA Technical Assistance
Abbreviations and Acronyms
5 Introductory Remarks
The international community has been engaged in Afghanistan since 2001. The complexity of the
challenges has been exceptional in many aspects, as the amount of funding provided for reconstruction
and development was unprecedented. Afghanistan was a country in ruins with a population deeply
traumatised by more than 20 years of war and civil war. In addition, the international engagement
was a civilian and a military intervention. Development actors had to learn from scratch how to design
civil-military cooperation. On top of that they had to cope with the political realm calling for quick results
in order to win hearts and minds. Despite this disadvantageous implementation environment, it must
be said that since 2001 the Afghan partners – helped by the international community – have made huge
For the BMZ, Afghanistan is still the country receiving the biggest single share of grants. The BMZ unit
for Afghanistan has a long track record of monitoring and evaluation going back to 2005. The idea of
initiating this Meta-Review of Evaluations of Development Assistance to Afghanistan 2008 - 2018 was
motivated by a desire to increase the development impact of our portfolio and to get a solid assessment
of how international cooperation in focal areas of German cooperation has fared in Afghanistan over the
In a fragile country, and this holds true in particular for a country like Afghanistan, it is worth remem-
bering that international aid is a walk on the knife’s edge. If aid comes in adequate doses, conceptually
well designed, culturally adjusted and truly owned by the partner, a great deal can be achieved. But
it is also possible to trigger huge damage in the sense of creating distortions, destroying incentive
structures and even fuelling corruption. With regard to Afghanistan, one should take into account in
particular that international aid is not a time-invariant process. There is a unique window of opportunity
for kick starting development in a country, unleashing the dynamics of socio-economic processes
before diseconomies of scale and time cause the system to start backﬁ ring. Once people and cultural
norms are ‘spoilt’, there is no chance to start again from scratch.
In these and in many other aspects, the Meta-Review condenses a vast amount of evidence showing
what worked and what did not. It contains both challenging and even unpleasant as well as highly
valuable lessons for future programming. Reﬂ ecting on the aspirations with which the international
community started out in Afghanistan, it is also good to remind ourselves from time to time that Do no
harm is the most important and recognized principle of international cooperation. This principle is our
ethical benchmark and the daily guidance for all our e orts to improve living conditions in Afghanistan
There are small glimpses of hope for the beginning of a peace process. But poverty rates are on the rise
and the path to achieving lasting peace will be long and arduous. I am convinced that we have the ability
to learn and there is a pressing need to do the best we possibly can so as to bring hope to the people
of Afghanistan. For the time ahead we hope the Meta-Review will reveal opportunities that the inter-
national community may be able to realise.
BMZ Division for Afghanistan and Pakistan
The Federal Ministry for Economic Cooperation and Development (BMZ) has a long-standing interest
in evaluating development assistance to Afghanistan. It has conducted various strategic reviews and
evaluations of its own engagement,1 and has built up a unique aid data management system, the
German Development Tracker.2
In September 2018, the BMZ Division for Afghanistan and Pakistan commissioned a meta-review of
evaluation reports of international development assistance to Afghanistan published between 2008 and
2018. The main objective of this meta-review is to collect and summarize the experience of donors in
Afghanistan. The meta-review will provide important background information for an upcoming evalu-
ation of German development cooperation with Afghanistan. Beyond that, the results will also provide
valuable information for development cooperation in fragile and conﬂ ict-a ected states in general.
Meta-reviews are exercises in learning. This meta-review informs about what has worked and what has
not worked: Which approaches and instruments were e ective in Afghanistan? What were the impacts
and what were unintended consequences? Which results are likely to be sustainable? And what lessons
can be learned for future evaluations in similar contexts?
This meta-review summarizes the experiences from a wide range of international actors and organiza-
tions, from bilateral donors to multilateral donors to NGOs and to organizations such as the o ce of the
Special Inspector General for the Reconstruction of Afghanistan (SIGAR). These actors have provided
a large number of di erent types of evaluation reports, ranging from impact evaluations to formative
evaluations and to performance audits.
1 See Zürcher, Christoph. 2017. “BMZ Strategische Portfolio Review Afghanistan 2016”. Bundesministerium für
wirtschaftliche Zusammenarbeit und Entwicklung (BMZ), unpublished report; Böhnke, J.; Köhler, J.; Zürcher, C. (2015).
“Assessing the Impact of Development Cooperation in North East Afghanistan 2007-2013. Final Report”. Bonn/Berlin:
Federal Ministry for Economic Cooperation and Development. Available at https://www.ez-afghanistan.de/sites/default/ﬁ les/
Afghanistan_Impact_Assessment_II_en.pdf (accessed January 2, 2020);
Zürcher Christoph, Catherine Gloukhovtseva, Nora Röhner, Gregg Fy e. 2013. “ Strategische Portfolio Review Afghanistan.
Schlussbericht” Bundesministerium für wirtschaftliche Zusammenarbeit und Entwicklung (BMZ). Available at
(Accessed January 2, 2020); “Assessing the Impact of Development Cooperation in North East Afghanistan, 2005 – 2009. Final
report”. BMZ, Evaluation Reports 049. Available at https://www.ez-afghanistan.de/sites/default/ﬁ les/EvalBericht049.pdf.
(Accessed January 2, 2020).
2 See https://www.ez-afghanistan.de/sites/default/ﬁ les/DevTracker_German.pdf, and https://devtracker-afg.de/,
accessed January 2, 2020.
7 Introduction: Will We Ever Learn?
The meta-review summarizes the major ﬁ ndings from these evaluation reports. We have
arranged all evaluation reports in ﬁ ve di erent groups, and for every group we have produced
a stand-alone summary report. These are:
1. Systematic Review of Impact Evaluations of Development Aid in Afghanistan
(short title: “Impact Evaluation Report”)
2. Summary Report of Eleven Bilateral Country-Level Evaluations (short title: “Bilateral Report”)
3. Summary Report of Selected SIGAR Reports (short title: “SIGAR Report”)
4. Summary Report of Evaluation Reports by the Asian Development Bank (short title: “ADB Report”)
5. Summary Report of Selected Evaluations by Multilateral Organizations and NGOs
(short title: “Multilateral and NGO Report”)
This chapeau paper is a very condensed summary of the ﬁ ve stand-alone reports. For more details,
readers are invited to consult the ﬁ ve stand-alone reports, which also contain all the bibliographic
references to the source documents.
1 Introduction: Will We Ever Learn?
In 2006, the Center for Global Development published a widely discussed study entitled “When Will We
Ever Learn?” The study aimed to investigate why rigorous evolutions of development programs were
relatively rare, despite the ethical obligation to learn how to make aid better, and despite the commit-
ment of development organizations to provide more accountability.
More than a decade later, the same question could be asked regarding Afghanistan – one of the longest
and costliest engagements by the international community. While there have been achievements –
access to basic health and primary education has massively improved; clean drinking water is much
more widely available; roads and bridges have been rebuilt; electricity has reached many villages;
rudimentary government services are available; small, basic infrastructure and training has improved
livelihoods in rural communities – most of the more ambitious goals were missed.
Afghanistan is still engulfed in war, poverty levels have not changed, and the government has hardly
gained legitimacy or capacity. Even when we acknowledge that Afghanistan is one of the most challeng-
ing places for development assistance, the overall results of more than a decade of international aid are
It is time to learn more about what has worked, what has not worked, and why. As an attempt to distill
lessons and patterns from the hundreds of available reports, a meta-review of evaluations of develop-
ment aid to Afghanistan is certainly an exercise in learning. For this meta-review, 148 evaluation reports
published between 2008 and 2018 were analyzed. These reports provided a wealth of information and
knowledge, and readers are invited to consult our ﬁ ve stand-alone reports for more details.
Overall, one ﬁ nding stands out: the international community has repeatedly overestimated its own
capacity and the capacity of its Afghan partners to bring about rapid social change. What has worked
best are modest, locally embedded projects with immediate, tangible beneﬁ ts. What has rarely worked
are complex projects aimed at building capacity and changing behaviour. More speciﬁ cally, interventions
in basic health and education, and in improving basic livelihoods, led to results. Interventions in building
capacity for the administration, or in sectors such as the rule of law or gender, rarely worked.
In reading these 148 reports, one also realizes that the international aid community is often not good at
learning. Monitoring and evaluation systems are weak, and have hardly improved since 2002. Back in
the early 2000s, many donors pointed out that, in order to achieve meaningful and sustainable devel-
opment, more time was necessary. Fifteen years later, few sustainable results have been achieved, but
many donors continue to suggest that better results will still require more time. Few donors appear to
have changed their fundamental strategic approach, despite the fact that their own evaluations strongly
suggest that many aid programs are neither e cient nor e ective in the Afghan context.
In all fairness, the Afghan context is an incredibly challenging one, as these 148 reports vividly remind
us on almost every page. The situation on the ground was and still is characterized by a lack of basic
security; Afghan partners in government and in civil society lack basic capacities; many entrenched
political actors have little interest in real reforms. Despite these challenging conditions, there was since
the early days of the international engagement in Afghanistan tremendous political pressure on devel-
opment actors to rush in and to provide quick results. An additional layer of complexity was added by
the fact that the international engagement was from the beginning both a civilian and a military inter-
vention, and planners in headquarters as well as practitioners on the ground had to learn how to cope
with the task of civil-military cooperation. Under such circumstances, designing e ective aid programs
is a herculean task.
At the time of writing – in November 2019 – the prospects for Afghanistan’s immediate future are less
clear than ever. For the moment, the so-called peace-process has stalled; it is unclear if and when a
complete withdrawal of US troops will take place and what the implications for the intra-Afghan civil
war will be. It is not even possible to predict whether the Kabul government will hold on to power. What
seems to be clear, however, is that the international donor community is not abandoning Afghanistan.
Development cooperation will continue. Now is a timely moment for donors to reﬂ ect on what their
engagement in the future may look like, and how they may avoid some of the mistakes which were
made and which this meta-review documents.
It is our hope that this meta-review will contribute to learning, and eventually to more e ective and
e cient ways of using development aid to help Afghanistan.
Introduction: Will We Ever Learn?
9 Methodology of the Meta-Review
2 Methodology of the Meta-Review
We intended to cover various types of evaluation reports (such as impact evaluations, performance
audits, formative evaluations, bilateral country-level evaluations, etc.) produced by a wide range of
relevant bi-and multilateral actors.
In order to identify relevant studies, we conducted a systematic literature search designed to ﬁ nd any
English language publications (articles, books, conference papers, reports), published between 2008
and 2018, evaluating development interventions in Afghanistan. Six comprehensive databases were
searched: PAIS, WPSA, EconLit, IPSA, Web of Science and Academic Search Complete.
We also searched manually for studies on the websites of the development agencies of all OECD DAC
countries, on the websites of multilateral donors such as the World Bank, the Asian Development Bank,
the UN and UN agencies, and on the websites of selected NGOs with a large portfolio in Afghanistan,
among them the Aga Khan Development Network (AKDN), CARE, Mé decins sans frontiè res (MSF),
Mercy Corps, Oxfam, Welthunger Hilfe and World Vision. We also searched for relevant publications on
the website of the Special Inspector General for the Reconstruction of Afghanistan (SIGAR).
Obviously, the identiﬁ ed studies vary considerably in terms of scope, objectives and methodo-
logical rigour. This is why we arranged the selected evaluations in ﬁ ve di erent groups:
A ﬁ rst group consists of rigorous impact evaluations. Reports in this group are primarily designed
for measuring causal impacts of programs or projects, using sophisticated methods and ﬁ ne-grained
quantitative data. Typically, they do not provide information about e ciency or sustainability.
A second group consists of country-level evaluations by bilateral donors. These reports usually eval-
uate the complete aid portfolio of a national donor over a longer period of time (usually at least ﬁ ve
years). These reports are typically based on desk studies and interviews. They are designed to assess
all ﬁ ve OECD DAC evaluation criteria (relevance, e ectiveness, e ciency, sustainably and impact), but
usually cannot assess impacts because of a lack of data and methodological challenges.
A third group consists of performance audits and lessons-learned reports published by the Special
Inspector General for Afghanistan Reconstruction (SIGAR). SIGAR’s main mandate is to audit US
reconstruction programs in Afghanistan, and multilateral programs in which the US participates. Being
concerned mainly with performance audits, SIGAR’s focus is on e ciency and e ectiveness.
A fourth group consists of evaluation reports by the Asian Development Bank. The ADB supports mainly
large infrastructure projects in Afghanistan through loans and grants. ADB evaluation reports look at
four out of ﬁ ve OECD DAC criteria (relevance, e ectiveness, e ciency, sustainably) and occasionally also
discuss (but not measure) impacts.
A ﬁ fth group consists of evaluation reports by various multi-national organizations and NGOs. While
reports in this group are not in terms of their methodology set up to measure impacts, they never-
theless provide important contextual information. They also usually look at the ﬁ ve OECD DAC evaluation
criteria, but usually discuss mainly relevance and e ectiveness.
The studies in these ﬁ ve groups are based on various levels of methodological rigour. Clearly, the most
rigorous studies are impact evaluations in group 1. However, there are good reasons to also include
other studies. Not all evaluations look at impacts. There are also formative evaluations, participative
evaluations, performance audits, or country-level evaluations that operate at high levels of gener-
alization where evaluating impacts becomes impossible. All of these evaluations contain valuable
information, even if they are not suited for attributing causation. Furthermore, some interventions lend
themselves more easily to evaluating impacts than others. For example, it is relatively easy to measure
the impacts of health interventions, but very di cult to attribute causation to capacity-building mea-
sures. Thus, including only rigorous evaluations would mean losing many evaluations on interventions
for which measuring impacts is di cult.
It is important to note that the inclusion criteria are di erent for each group. The strictest and most
objective inclusion criteria were applied to the ﬁ rst group (impact evaluations). We only included studies
in this group which made a credible, transparent and methodologically solid attempt to measure the
counterfactual. In order to make sure that our selection was as unbiased as possible, two scholars
independently read the studies. Only when both scholars agreed would the study be included. In the
case of a conﬂ ict, the lead researcher, Christoph Zü rcher, made the ﬁ nal decision. Eventually, 32 impact
evaluations were deemed of good quality and included.
For other groups di erent inclusion criteria applied. We included all bilateral country-level evaluations
published after 2008, irrespective of their methodological quality. Likewise, we also included all ADB
evaluation reports in order to get good coverage of infrastructure projects. We included SIGAR reports
when they referred to one of the priority sectors of German development cooperation. Finally, we also
included in a separate group evaluation reports by NGOs and multilateral donors which did not meet
the strict criteria for impact evaluations but still contained, in our view, important lessons. Clearly, the
inclusion criteria for this group are much less objective and strict than those for the ﬁ rst group (impact
evaluations). It is important to keep these di erences in mind when reading the ﬁ ve reports.
Methodology of the Meta-Review
11 Methodology of the Meta-Review
The next table shows the inclusion criteria for each of the ﬁ ve groups.
Table 1: Inclusion Criteria
Group Inclusion Criteria
Systematic Review of Impact
Evaluations of Development Aid in
Afghanistan, 2008 – 2018
All studies with causal attribution of impacts, published in
English, 2008 – 2018 with a credible, transparent and
methodologically solid attempt to measure the counterfactual.
Note that this is a lenient deﬁ nition for causal attribution.
While most impact evaluations rely on sophisticated statis-
tical methods for assessment, our deﬁ nition also includes
studies based on careful process tracing or comparison.
Decision made by two scholars who independently
sread the full study.
Summary Report of Eleven Bi-lateral
Country-Level Evaluations, 2008 – 20183
All country-level evaluations of aid to Afghanistan by
bilateral donors published in English, 2008 – 2018.
Summary Report of Selected SIGAR
Reports, Afghanistan, 2008 – 2018
All SIGAR lessons-learned reports plus selected SIGAR
reports when they referred to priority sector of German
Summary Report of Evaluation
Reports by the Asian Development Bank,
2008 – 2018
All ADB evaluation reports on Afghanistan, 2008 – 2018.
Summary Report of Selected Evaluation
Reports by Multilateral Organizations
and NGOs, 2008 – 2018
Selected evaluation reports by NGOs and multilateral donors.
The evaluation reports may not explicitly address casual
attribution, but still contain important lessons. The decision
to include was made by the lead researcher.
Once selection and grouping of the evaluation reports was completed, we extracted the relevant infor-
mation. As we have mentioned, the reports in the ﬁ ve groups are quite di erent, hence we made no
attempt to strictly apply the same information extraction framework to all groups. However, in general
we did focus on the priority sectors of German cooperation (governance, education and vocational
training, health, water, energy, and sustainable economic development) plus two additional cross- cutting
sectors (stabilization and gender). Secondly, we collected all information about the ﬁ ve DAC evaluation
criteria (relevance, e ectiveness, e ciency, sustainability and impact). Finally, we also looked for
relevant information about the contextual factors which made it more or less likely that performance
objectives were met.
No meta-review is ever complete in its coverage and this one is no exception. Ours is limited to evalua-
tion reports published in English; we limited our systematic search to six important databases; and it
is possible, even likely, that we have not found all studies published on the websites of bilateral and
3 Germany conducted two forward-looking reviews of the BMZ aid portfolio for the period 2013 – 2017 and 2017 – 2021.
These reviews were not included in the meta-analysis since they did not meet all of the inclusion criteria for bilateral evalu-
ations, but we report the main results. Please see the box on page 7 in Part 2: Summary Report of Eleven Bilateral Country-
multilateral donors and NGOs. Many of these websites are true labyrinths, o ering little in terms of
search options. Finally, there are probably hundreds of studies that have never been made accessible
to a wider public. It is safe to assume, however, that these studies, hidden away in the electronic or
paper archives of donors, neither contain success stories nor are based on robust methods. Since all
donors are eager to share good news from Afghanistan and showcase examples of solid evaluation,
it is likely that any such studies would have been widely shared.
Despite these limitations, we are conﬁ dent that our study o ers the most comprehensive, least biased
view so far of development assistance to Afghanistan.
3 The Evidence Base
The total number of studies included in this meta-review is 148. Of these, 32 were impact evaluations
designed to attribute causation in a methodologically rigorous way. Some sectors have been subject to
more rigorous evaluation than others. For example, for the health sector, we found 21 studies overall, of
which nine were rigorous impact evaluations. For water, however, there were only six studies, with none
that could be classiﬁ ed as a rigorous evaluation. Table 2 lists the number of studies per sector, and the
number of that total that we classiﬁ ed as “rigorous.”
Table 2: Studies per Sector
Sector Total number of studies Rigorous impact
Health 21 9
Education 14 3
Gender 14 2
Stabilization (including CERP) 12 11
Sustainable economic development
(including rural development) 12 0
Governance (including rule of law,
democracy promotion, election support and
public sector reform)
ARTF 11 0
Infrastructure 10 0
Sub-national governance (including NSP) 8 4
Energy 6 2
Water 6 0
Multi-sector, country level 15 0
Other 7 1
Total 148 32
The Evidence Base
13 The Evidence Base
Table 3 reports who commissioned the included studies, and what percentage of the studies were
rigorous impact evaluations (it should be noted that these numbers are based on our sample, which
was constructed based on exogenous selection criteria, and are thus not representative for the universe
of all evaluation studies).
Looking at the evidence base, a few observations stand out. Firstly, given the length, breadth and cost of
the international engagement in Afghanistan, the overall number of evaluations – especially of rigorous
impact evaluations – seems quite small. Secondly, rigorous impact evaluations took place mainly in
the ﬁ elds of health and stabilization. The explanation for this may be that health outcomes are easily
measurable. The frequency of stabilization evaluations can be explained by the fact that the military
collected the data which made impact evaluations possible, and by the massive interest in counterin-
surgency (COIN). Thirdly, most rigorous impact evaluations were produced by independent scholars and
academics. Donors rarely commissioned complex impact evaluations. Fourthly, methodologically robust
evaluations by NGOs appear to be rather rare. And ﬁ nally, there is a rather small number of evaluations
in the ﬁ eld of sustainable economic development and rural development, which is surprising given how
important that sector is in Afghanistan.
Table 3: Commissioners of the Studies
Commissioner Total number of studies Rigorous impact evaluations
SIGAR 56 0
Multilateral 37 6
Academic/independent 23 19
ADB 13 0
Bilateral donor 13 2
NGOs 6 4
In this section, we provide important ﬁ ndings in a very condensed form. Readers are invited to consult
the ﬁ ve stand-alone reports for more details, including all references to the original sources.
When reading these condensed ﬁ ndings, it is important to remember that when we make, for example,
a statement such as “capacity building never worked”, we mean that “we have not found evidence in
the reviewed 148 reports that capacity building worked”. Obviously, we cannot exclude that there may
have been successful capacity building programs in Afghanistan – but these are not documented in the
reviewed reports, and we made considerable e orts to identify a very broad and representative sample
of evaluation reports.
The implication for practitioners is that it is now up to them to credibly demonstrate that there are more
than anecdotal successes in the ﬁ eld of capacity building. If this cannot be demonstrated, then they
should for now accept the ﬁ nding that capacity building in this context failed. Looking ahead, this then
also means that we should engage in a discussion about what is and what is not possible in a context
such as Afghanistan.
4.1 Contextual Factors
A reading of the 148 evaluation reports makes it clear that the lack of basic security was a pervasive
problem, constantly a ecting every aspect of development cooperation. The reports highlight how
di cult it was to implement and monitor development projects when sites are not accessible, or when
development workers are at risk of being targeted by insurgents. Many reports describe how a lack
of security caused delays and cost overruns. Lack of security also necessitated that aid organizations
employed security measures, which increased implementation costs. A recurring recommendation is to
better acknowledge that Afghanistan is a country embroiled in war rather than a post-conﬂ ict country,
and that donors therefore should develop a more realistic assessment of the security environment.
Despite the very di cult security situation, there was considerable political pressure on aid organi-
zations to allocate aid to the most insecure regions in the hope that aid could help to stabilize these
regions. As we know now, this was not the case, but allocating aid to insecure regions naturally made
projects less e ective and e cient, and more di cult to monitor.
Findings – Contextual Factors
15 Findings – Contextual Factors
4.1.2 Local Political Economy
These evaluation reports also vividly remind us that the political economy in Afghanistan was and still is
a major impediment to e ective development cooperation. The international engagement in Afghanistan
faced since day one a volatile and unstable political situation, a highly fragmented political elite, and a
lack of political will for reform among many members of the Afghan elites.
Many evaluation reports, especially those by SIGAR, also point to the pervasive corruption and wide-
spread rent-seeking behaviour, which undermined governance, security and service delivery, and led to
distrust and lack of legitimacy for the government.
Cultural norms also a ected programs, especially those concerning gender equality and human rights.
Many reports note the discrepancy in attitudes between donors and many segments of Afghan society
in these ﬁ elds, which often led to a lack of buy-in and little political will for reforms.
Many donors noted that Afghan ownership of development programs was a major factor for e ective
aid, but also noted that ownership was rarely high. Unfortunately, the reports do not explicitly inves-
tigate the reasons for this, but they do provide some indirect clues: ownership may have been low
because Afghan structures lacked the capacity to meaningfully “own” programs; political inﬁ ghting
politicized aid; and there was a general lack of political will to support programs in ﬁ elds such as good
governance, gender, human rights, decentralization, anti-corruption and similar ﬁ elds.
Taken together, this local political economy made it extremely di cult to design and implement
e ective aid programs.
A third recurring theme is the lack of capacity among the Afghan partners. Almost all reports stress
that Afghan governmental structures lacked the capacity to deal with aid ﬂ ows in a productive way.
Unfortunately, many donors consistently overestimated Afghan capacity, designing programs based on
largely imagined absorptive and administrative capacity.4
4 The analyzed evaluation reports contain little information about the capacity of Afghan civil society organizations,
Afghan NGOs and the Afghan private sector. One reason for this is may be that there are only very few quality evaluations
by NGOs, which typically partner with other NGOs and with civil society organizations. From what little evidence we have, we
could assume that Afghan development NGOs (which are often branches of international NGOs) have a bit more capacity than
their counterparts from the government, especially in rural development (see Summary Report of Selected Evaluation Reports
by Multilateral Organizations and NGOs, 2008–2018, and there especially Altai Consulting 2017) and FAO UN 2016). NGOs in
the sector of democracy promotion appeared to have low capacity (see Summary Report of Selected Evaluation Reports by
Multilateral Organizations and NGOs, 2008–2018, and there especially Transtec 2013 and 2015).
4.2 OECD DAC Criteria
None of the 148 evaluation reports judged that a project was not relevant. Given that Afghanistan has
needs across all development sectors, it is perhaps not surprising that all projects were seen as rele-
vant. This, however, leads us to question whether the criterion of “relevance” is useful at all in such a
context. Instead of rating a project as “relevant” when it is seen as addressing a need, it may be more
useful to assess whether a given project has any actual impact potential. In other words, relevance is
not only a function of need, but also of the probability of success given the conditions under which it
is implemented. This would require donors to develop a better understating of the conditions on the
ground, and then prioritize projects that are adequate for such a context.
4.2.2 E ciency
A common thread across all reports is the observation that the di cult context in Afghanistan made
development cooperation unusually costly. This is true for all types of projects, but especially for large
infrastructure projects. Most projects experienced implementation issues, delays and costing problems
(overruns or underutilization of funds). Most often, e ciency su ered from a lack of security, lack of
partner capacity and a local political economy which made development cooperation challenging.
4.2.3 E ectiveness and Impacts
Few studies measured actual impacts, but many assessed the e ectiveness of projects. It is clear
that e ectiveness in general was low, but there are di erences between sectors and between types
Among the more e ective interventions were those in the health and education sectors. A middle
ground is occupied by infrastructure, including small-scale infrastructure in rural areas, as most
projects were somewhat e ective in delivering intended objectives despite capacity constraints,
institutional weakness, and a volatile security environment. Interventions in sectors such as good
governance, rule of law and gender equality were rarely e ective. Mostly ine ective were stabilization
In general, smaller, modest participatory projects with an instrumental focus (for example, directly
inﬂ uencing change by providing new resources) were more e ective than large, complex projects
aimed at building capacity and changing behaviour and discourse.
Findings – OECD DAC Criteria
17 Findings – Findings Per Sector
Worrisome is that almost all reports stress that the sustainability of achieved results is very much in
question. This appears to be the case for all types of aid programs across all sectors. A lack of capacity
makes it unlikely that the Afghan government will be able to take on investments in larger projects, and
the weakness of district and community level organizations makes it equally unlikely that even small
and less complex infrastructure projects will continue to function once support ends. Also, much of
the progress which has been made in public administration and public management may not be sus-
tainable because it relied to a large extent on the so-called “second civil service”, that is on the work of
4.3 Findings by Sector
A very broad sector; governance includes capacity building, public sector and regulatory policy reform,
democracy promotion, election support, anti-corruption programs and rule of law. In general, the e ec-
tiveness of the evaluated programs in the governance sector was low.
Programs aimed at improving capacities for the Afghan central administration rarely succeeded (see
Multilateral and NGO Report, esp. pp. 11-12), and most donors, especially bilateral donors, routinely
mentioned weak state capacities as a major impediment for their work (see Bilateral Report, p. 27). In
the few instances where progress was made, it remained conﬁ ned to small silos that did not translate
to overall state capacity, and/or it was “borrowed” from the so-called “second civil service” consisting
of well-paid Afghan returnees or international consultants (Bilateral Report, p. 17). There is anecdotal
evidence that some ministries have more capacities than others (for example the Ministry of Finance is
often said to have more capacity), but the reviewed reports do not o er enough details for assessing the
relative capacities of various government branches in a systematic way.
Also not very e ective were programs aimed at increasing capacity for sub-national administration and
those meant to build up capacities for managing relations between the centre and provinces in order
to provide meaningful decentralization (Multilateral and NGO Report, pp. 12). Among the reasons for
this poor record are power struggles between and within ministries, and between centre and periphery,
which made it di cult to build capacity. Decentralization involves the redistribution of power and access
to resources. This is highly politicized, and reforms in politicized ﬁ elds are rarely e ective.Decentraliza-
tion programs were indicative of a top-down approach by donors, as there was no real demand for such
programs from within Afghanistan.
Interventions in regulatory policies were also not often successful. For example, ADB projects aimed at
reforming public administration, creating better regulatory frameworks for private sector development
and for the agriculture sector, had little e ect. ADB concluded that technical skill transfer was unlikely
because the institutional absorptive capacity of the partner was limited (ADB Report, p. 21).
The two reviewed projects in rule of law were both not successful because they were overly ambitious,
were not based on the political-economic realities on the ground, and were ideologically framed by an
unrealistic theory of change (Multilateral and NGO Report, p.13).
Finally, with regard to democracy promotion, the reports suggest that development assistance could
provide the technical capacities needed for conducting elections; however, projects aimed at democratic
awareness or democratic participation had little e ect (Multilateral and NGO Report, p.14).
In sum, programs aimed at better governance and more capacity were rarely e ective in the Afghan
context. Such programs faced too many political challenges. Factors which hampered such programs
were entrenched patronage-based practices within the government, a lack of buy-in from the govern-
ment, donor-driven top-down project design with little regard for the core institutional requirements
and demands of the partner institutions, and lack of political will of the government especially for
decentralization. Given the permanent competition for power among various networks (often ethnically
based), there was little cooperation within and among institutions, little interest in building up institu-
tional learning and institutional memory, and few incentives for cooperation with the sub-national level.
Furthermore, the frequent change of key personnel as a result of permanent power struggles exacer-
bated the problems.
Most donors saw the Afghan Reconstruction Trust Fund (ARTF) as the main vehicle for governance pro-
gramming and as an e cient channel for allocating and distributing aid money. The ARTF was also seen
as a good instrument for aid coordination, resource mobilization, and policy dialogue among donors, and
between donors and the Afghan government. However, bilateral donors found little evidence that the
ARTF was e ective at building capacity within the Afghan government. Donors also expressed concerns
that the ARTF was not sustainable and would require a complementary, gradual phase-in of Afghan
ﬁ scal responsibility (see Bilateral Report, p.16).
Many bilateral donors as well as SIGAR expressed concern about the lack of e ort by the World Bank
to monitor and evaluate outcomes and impacts of the ARTF, especially its investment window. SIGAR
criticized the World Bank for its weak monitoring of the ARTF, noting limitations in its transparency
and accounting for ARTF funding, and in the Afghan government’s ﬁ duciary controls (see SIGAR report,
Over the years, the ARTF has ﬁ nanced an impressive number of infrastructure projects. The evaluation
reports suggest that less complex infrastructure usually fared better than more complex projects.
Given the ARTF’s still relatively weak monitoring and evaluation system (despite recent improvements),
we do not know much about the outcomes and impacts. There is some evidence that implementing
and operating complex infrastructure through the ARTF is still beyond the capacity of the Afghan
Findings – Findings Per Sector
19 Findings – Findings Per Sector
Donors placed high hopes in the ARTF as an instrument for not only ﬁ nancing the operation of the
government, but also for increasing governmental capacity, improving governance, and boosting
economic growth. The reviewed evaluation reports cannot demonstrate that the ARTF has met these
4.3.3 Sub-national Governance (including NSP)
The National Solidarity Program (NSP) was the ﬂ agship project of the ARTF. Evidence suggests that
NSP contributed to an increase in services and infrastructure in rural areas but had little impact on
economic growth or local governance (Impact Evaluation Report, esp. pp.10).
NSP mobilized communities, created ownership for the projects, and mandated the representation of
women in the newly formed community development councils. There is no evidence, however, that the
formal participation of women in community-level decisions has had a tangible impact on overall gen-
der equality (Impact Evaluation Report, esp. pp.10).
NSP led to an increase in positive attitudes towards sub-national and national governments, towards
NGOs, and towards international troops. However, these positive e ects only hold in villages with a
relatively good security environment (Impact Evaluation Report, esp. pp.10).
Besides NSP, there were other programs aimed at increasing sub-national governance. There is scant
evidence about the impact of these programs, but it appears that projects rarely led to increased
capacity for sub-national administrations (Multilateral and NGO Report, esp. pp.11).
4.3.4 Stabilization (including CERP)
We have reports summarizing the experience of the stabilization programs of the UK, US, Denmark and
Canada. Assessments of those programs are usually negative; there is no evidence that stabilization
projects led to more stability in insecure regions (Bilateral Report, pp. 18). The available evaluation
reports, above all the SIGAR lessons-learned report, suggest that aid often exacerbated inter-group
tensions and attracted violence (SIGAR Report, pp. 26; Impact Evaluation Report, pp.13-18). Often, aid
was spent too fast, with a lack of oversight, and in insecure regions with little or no local governance
structures in place.
A recent systematic review (which is not part of this meta-review) found that aid only has a stabilizing
e ect when implemented in reasonably secure regions under government control. Additionally, chances
for stabilization through aid are better when projects are implemented in participatory ways – prefer-
ably through accepted local authorities – and when aid is transparent and did not beneﬁ t local power
brokers through corruption or nepotism.5 These lessons were not often applied in the stabilization
programs reviewed here.
5 Cf. Zürcher, Christoph. 2017. “What Do We (Not) Know About Development Aid and Violence? A Systematic Review.”
World Development 98 (October): 506–22. https://doi.org/10.1016/j.worlddev.2017.05.013.
Most reports agree that substantial progress has been made regarding better access for boys and
girls to primary education (Bilateral Report, pp. 24; Impact Evaluation Report, pp. 35). Well-targeted
projects did improve outcomes in primary education (Multilateral Report, pp. 24). However, the quality
of education remains problematic, a large demand for infrastructure remains, and many gains may not
be sustainable given the enduring insecurity and the lacking ﬁ nancial and bureaucratic capacities of the
Afghan government (Multilateral and NGO Report, pp. 25; SIGAR Report, pp. 18).
The available studies point to a tangible increase in access to basic health care and to a massive
improvement in such health indicators as child and maternal mortality (Impact Evaluation Report,
pp. 26; also Multilateral and NGO Report, pp. 27; and Bilateral Report pp. 25). Interventions in the
health sector were usually e ective. The reviewed reports suggest that successful interventions took
place in midwifery training, antenatal care visits, deliveries attended by health workers, conditional
cash transfers for women and community health workers, and improved family planning.
Evaluations suggest that improvements in access to services for women and girls – mainly in health
and education – have been made (Bilateral Report, pp. 19). Progress, however, is attributable to the
rehabilitation of infrastructure and the end of Taliban rule rather than the success of the gender proj-
Regarding programming for gender, donors typically reported outputs only, but remained skeptical
about outcomes or impacts. Donors noted that both the capacity and the political will of the Afghan
government and political elites for gender equality programming remained very limited as prevailing
cultural norms made progress di cult (Bilateral Report, pp. 19, SIGAR Report, pp. 25). Despite sustained
support, the capacities of the Ministry of Women’s A airs (MoWA) remained weak (Multilateral and NGO
Report, pp. 19, Bilateral report pp. 21). SIGAR noted that gender programs were not adequately moni-
tored and evaluated, which made it impossible to identify any possible impact. Insecurity, limited gov-
ernment capacity, and cultural norms also impeded any US e orts to advance women’s rights (SIGAR
Report, pp. 25).
Overall, the e ectiveness of gender programming appears to be low (Bilateral Report, pp. 19). There
are, however, pockets of modest success. Examples include rural literacy, increased access to health
and education, and better livelihoods in women-speciﬁ c activities within agriculture, such as mushroom
farming and kitchen gardening. In sum, small, modest projects embedded in traditional structures
helped to increase access to health, education, and modestly improved livelihoods for women. By
contrast, larger, more ambitious projects aimed directly at changing gender norms and relations
had no discernible impact (Multilateral and NGO Report, pp. 18-22).
Findings – Findings Per Sector
21 Findings – Findings Per Sector
4.3.8 Sustainable Economic Development
Programs supporting economic development, macroeconomic policies, and ﬁ nancial management
capacities achieved some progress in the early stages of reconstruction. For example, there was initial
growth in telecommunications, transport, and construction, but results were not sustainable nor was it
realistic to expect sustainable economic growth, given the insecure environment and the shrinking aid
ﬂ ows after 2013 (SIGAR Report, pp. 14).
Interventions aimed at promoting the private sector were rarely e ective. Evaluations cited weak insti-
tutional infrastructures and procedures, widespread corruption within the Afghan government, political
instability, and insecurity as the main reasons (SIGAR Report, pp. 14). Related, interventions aimed
at regulatory policies for ﬁ scal management and for public administration reform were also rarely
e ective. One reason for this is that technical skills transfer is not likely when institutional absorptive
capacity is limited (see ADB report, p. 21).
Support for rural development, often implemented through newly created community-level organiza-
tions, has helped to build a large amount of small infrastructure. Better access to services and basic
infrastructure such as roads, irrigation, and access to energy contributed to improved livelihoods and
has strengthened coping mechanisms, but has not led to sustainable economic growth that translated
into jobs or income opportunities (Multilateral and NGO Report, pp. 23). The capacity of the government
partner institutions in rural areas was weak, and it proved to be di cult to build more capacity.
In sum, interventions in sustainable economic development, despite some progress, have not been able
to reduce poverty rates or to promote sustainable economic growth (Bilateral Report, pp. 23).
Most of our evidence on large infrastructure projects (roads, energy, rail, airports) stems from ADB,
which noted that projects were often not e cient, but were somewhat e ective in delivering intended
objectives despite capacity constraints, institutional weakness, and a volatile security environment.
Despite frequent cost overruns and delays, caused by lack of partner capacity and an adverse security
situation, projects could be implemented. Examples include energy, road, and airport projects. How ever,
ADB considers the sustainability of all projects as less than likely, due to capacity constraints, weak
institutional capabilities, and borrower dependency on funding and technical assistance
(see ADB Report).
4.3.10 Capacity Building
Capacity building is a cross-cutting issue. A lack of capacity is at the core of a fragile state, and devel-
opment actors seek to build capacity across all levels of government and in civil society, by making
capacity building an important component of a program, or even by making it the sole focus.
Taken together, the 148 evaluation reports suggest that capacity-building measures were mostly not
successful. In the few instances where progress was made, it remained conﬁ ned to small silos, not
translating into more overall state capacity, and/or it was mainly borrowed from the so-called “second
civil service” consisting of well-paid Afghan returnees or international consultants (ADB Report, pp. 21).
There is no clear case of a highly successful capacity-building program in our sample, but quite a few
examples of rather ine ective capacity building. For example, donors agree that the ARTF had no impact
on better governance nor did it contribute to better capacity in the Afghan government (Bilateral Report,
p. 28). Likewise, three programs for decentralization and capacity building for the sub-national admin-
istration were not successful (Multilateral and NGO Report, pp. 12). Also, two reviewed capacity-building
programs for civil servants proved mostly ine ective (Multilateral and NGO Report, pp. 11).
The evaluation reports clearly suggest that the weak capacity of the Afghan administration severely
impacted project implementation, maintenance, and monitoring, and many reports called for more
capacity building e orts. However, the evaluation reports also suggest that capacity-building measures,
when part of a project, were usually not successful (ADB Report, p. 22).
The reasons for these disappointing results vary, but are mostly linked to the di cult context, as
described above in the section on governance: there was little demand for such programs, the perma-
nent competition for power among various networks hampered cooperation within and among institu-
tions, and the frequent changes of key personnel made institutional learning di cult. Most importantly,
capacity building is not e ective when there is no political will to build capacity, which is often the case
in politicized ﬁ elds (such as decentralization, which the central government opposes).
4.3.11 Monitoring and Evaluation
All bilateral evaluation reports mention that monitoring and evaluation systems of donors were weak
(Bilateral Report, pp. 30). Donors especially criticized the World Bank for its weak monitoring of the
ARTF, but country-level bilateral evaluations also admitted that their own monitoring and evaluation
systems were rarely able to measure outcomes. It is therefore not surprising that so few evaluation
reports are based on solid data. Many reports correctly point out that one explanation for weak monitor-
ing and evaluation is the lack of basic security, which made access to many project sites di cult. Yet, it
is still surprising that many donors (bilateral, multilateral, and NGOs alike) have apparently made little
progress in establishing adequate monitoring and evaluation systems since 2002, when international
engagement in Afghanistan began. As a result, we only too rarely understand what impact a program
really had, and why.
Findings – Findings Per Sector
23 Findings – Some Lessons to Learn
4.4 Some Lessons to Learn
In this last section, we brieﬂ y talk about ﬁ ve general lessons we think emerge from this meta-review.
Our selection of these lessons is to some extent subjective, but we made a considerable e ort to provide
all the underlying data in a transparent and accessible way in the ﬁ ve stand-alone reports. Interested
readers should ﬁ nd it easy to consult these data in order to ﬁ nd out more about speciﬁ c aspects of
international development cooperation in Afghanistan, and to derive their own lessons.
4.4.1 What is relevance, or the fallacy of a “needs-based” approach
Afghanistan is one of the poorest countries on the planet and it is engulfed in violence that has been
going on for more than four decades. The needs of its population are sheer endless. And, given these
needs, almost every development program seems relevant. Our analysis of the 148 evaluation reports
showed that not one of the reviewed programs was seen as “not relevant”, precisely because all of the
programs answered in one way or another to some needs. But we have also seen that by far not all pro-
grams were e ective. Actually, in the Afghan context, only a few programs were. Supporting programs
which are not e ective incurs opportunity costs – there is less aid money for programs which may have
more impact. It is therefore both rational and ethical to prioritize programs with a higher probability for
a positive impact. This then leads us to question whether the criterion of “relevance” should be based
solely on needs. We think “relevance” should be assessed, considering both needs and the probability
of having a positive impact. In other words, relevance should not only be a function of need, but also of
the probability of success given the conditions under which a program is implemented. Adopting such
a view of relevance would require donors to develop a better understanding of the conditions on the
ground, and then prioritize projects that promise to be reasonably e ective in these conditions. This
meta-review provides ample information about which programs have a high probability for impact
(and should therefore be continued) and which have a much lower probability of impact (and should
therefore be either run as pilots from which we can learn how to improve them, or cancelled).
4.4.2 Taking the local context seriously
Another recurring theme in the evaluation reports is that donors designed and ran programs which
were not appropriate for the context. Collectively, the evaluation reports make a number of suggestions
about how to adapt programs to the context. The list is long: Interventions should be less complex,
implemented sequentially, with much more ﬂ exibility to adapt. They should be based on a fair assess-
ment of security conditions on the ground. They should also be based on a realistic assessment of
partner capacities. They should not be over-ambitious, and they should be demand driven and not
donor driven. Finally, they should consider cultural norms, which may a ect many programs, among
them those concerning gender equality and human rights. However, donors rarely followed this advice.
Donors were clearly aware of the di cult local context, and rightly pointed out that it was extremely
di cult to meet performance objectives under these conditions, but we ﬁ nd little evidence that they
made strategic adaptations to their aid portfolios, to the way they delivered aid, or indeed to their expec-
tations of what could reasonably be achieved. Instead, many donors pointed out that, in order to achieve
meaningful, development results, aid ﬂ ows over much more time were necessary.
However, we are not convinced that “more of the same, but for longer” would lead to better results.
The international aid community has been engaged in Afghanistan for seventeen years now with little
signs of increased e ectiveness. We think that donors should take the di cult local context seriously,
internalize the above-mentioned suggestions, and accept that much of what we would like to do
simply may not work in Afghanistan, as long as there is no political settlement to the conﬂ ict and no
fundamental change in the political economy of the country. Instead, donors should focus on what
might work, and learn and adapt on the way. Doing this would require donors to carefully select pro-
grams and approaches they think can work, and programs and approaches that should be paused for
now. This would also require that aid organizations resist the political pressure “to do more” and to get
“quick results” – something they have not done in the past.
4.4.3 Modest and slow is better
One consequence of such an approach would be that development aid becomes more modest and
slower. The reviewed evaluation reports strongly suggest that there would be beneﬁ ts: in general,
smaller projects performed better than larger, more complex projects. Also, projects aimed at direct
results, such as building small infrastructure and providing such services as access to water and
electricity, were often e ective. Results have been less strong where programs aimed to be transfor-
mative in nature, either for capacity building or to change cultural and social norms. Furthermore, as
the evaluation reports show, there can be harm done by spending too much aid too quickly, as was
clearly the case for US aid. Government partners did not have the absorption capacities, yet programs
spent aid money quickly because the speed of spending was seen as a metric for success. However,
the windfall from aid created opportunities for power brokers to increase their inﬂ uence in villages,
cities, and within the government itself. This fuelled corruption, cultivated an environment of impunity,
and weakened the rule of law. In our view, a big lesson that emerges from a close reading of the 148
evaluation reports is that aid only has a fair chance of being e ective in Afghanistan when programs
are modest, rather small than large, do not assume unrealistic partner capacities, are aware of the
cultural context, do not spend aid money too fast, do not spend aid money in insecure regions, and are
equipped with solid performance measurements and the means to track these measurements with
baselines and follow-up data.
4.4.4 Aid in insecure regions
Much of the aid for Afghanistan has been directed to its most insecure regions. Unfortunately, aid for
stabilization does not work in highly insecure regions. Aid can only have a positive impact when it is
injected in reasonably secure regions under government control. Additionally, chances for stabilization
through aid are better when projects are implemented in participatory ways – preferably through
accepted local authorities – and when aid is transparent and does not beneﬁ t local power brokers.
In the absence of these conditions, the evidence strongly suggests that aid in insecure regions will
fuel conﬂ ict. This may happen because aid will exacerbate intercommunal or interethnic tensions. Or,
in regions where insurgent groups are present and able to act, aid will trigger a strategic reaction:
insurgents will either shut down aid projects because they fear that aid may increase the cooperation
Findings – Some Lessons to Learn
25 Findings – Some Lessons to Learn
between local communities and the government, or insurgents will try to regulate aid ﬂ ows in order to
“tax” aid and increase their prestige and legitimacy among local communities as enablers of aid. Some-
times insurgents also target roads or bridges when they think that they give the military an advantage.
Such strategic responses will lead to more immediate violence (since shutting down aid projects or
“taxing” requires violence or the threat of violence) and it will increase the capabilities of insurgents for
future violence. In the light of this observation, we think that development actors should not implement
aid in localities where insurgents retain meaningful capabilities to coerce and tax local communities
and aid workers.
4.4.5 We cannot work in the dark
Finally, almost all evaluation reports mentioned the lack of adequate performance measures and solid
data. Weak monitoring and evaluation systems have seriously hampered our ability to really understand
what impacts the billions of aid dollars had in Afghanistan. It is true that to some extent the di cult
security context is to blame for this. But it is still surprising that many donors (bilateral, multilateral,
and NGOs alike) have made little progress in establishing better monitoring and evaluation systems
since 2002, when international engagement in Afghanistan began. We know that we are not the ﬁ rst to
say this – but we think that investing in monitoring and (impact) evaluation will greatly help to make aid
more e ective, either by identifying approaches which have worked or, at the very least, by identifying
approaches which are not e ective under the given conditions.
5 Critical Reﬂ ections
The ﬁ ndings of this analysis are ﬁ rst and foremost an invitation to reﬂ ect. They are a starting point for
a discussion, not an end result. Some of the questions that arise are: Why do we see more success in
some sectors, and less in other sectors? Does that mean that success in the more “di cult” sectors
(such as governance, gender, or capacity building) is simply not achievable under the given circum-
stances? Or does it mean that with better planning and better implementation, results could have been
better? If the latter is true – which I believe it is – then what are realistic objectives and realistic time
frames, under the given circumstance? Finally - where should we invest our aid going forward? Where
do we think we can have a meaningful impact in a reasonable timeframe? And, no less important, where
do we think we need to invest so that future developments will become possible, even if we cannot
expect to see impacts in the immediate future? The meta-review does not provide easy answers to
these questions. But it prompts us to critically reﬂ ect on them.
The results, sobering as they may seem at ﬁ rst glance, can and do o er some guidelines by pointing
to ﬁ elds where impacts are more or less probable. We should take these insights seriously, but at the
same time we should keep things in perspective.
Evaluations – especially rigorous ones - measure success as the extent to which a clearly deﬁ ned objec-
tive has been met. Success is therefore always dependent on how it is deﬁ ned. In Afghanistan, ambi-
tions and expectations were too high, and in a way, many programs where therefore set up for failure,
since it was never likely that they could reach all their ambitious objectives. But this does not mean that
a program has had no impact at all. Examples abound. ARTF may have had little impact at increasing
good governance and overall institutional capacity. But it did enable the government to pay salaries
to teachers and administrators, to build schools or to repair roads. Rural development programs
may not have created employment on a massive scale, but they did improve livelihoods and provided
some poverty relief. Gender programs may not have empowered women, but they did help to improve
women’s literacy and provided some additional incomes for women. All of these modest achievements
add up and pave the way for future development – a way that would be blocked without these “small”
Furthermore, evaluations have a relatively short time frame. They usually take place immediately after
a program has ended. Often that is not enough time for impacts to unfold. Longer time frames would
quite possibly unearth more successes.
Also, do we always look at the right place for successes? Often evaluations assess programs which
are dear to the hearts of donors. Hence many evaluations may look at things such as progress in good
governance, democracy, or gender. Considerably fewer evaluations look at less fancy and less visible
things such as: has a ministry acquired more technical expertise to run its daily a airs? Are salaries
paid? Are water meters installed? But these more tangible things can make a big di erence in the lives
Critical Reﬂ ections
27 Critical Reﬂ ections
Another bias stems from the simple fact that some things are easier to measure than others: For
example, child mortality, or water consumption, or the number of enrolled students in primary schools
are easy to measure, hence successes are easily documented. By contrast, increased capacities, better
bureaucratic processes, more knowledge, better training, changed attitudes or better education are
di cult to measure, yet such changes add up and will manifest themselves in increased human capital
which in turn will change the social fabric over time and enable development in the future.
Taken together, these “small things” explain the seemingly paradox observation that while many aid
programs appear to be only marginally successful in the light of the meta-review, the overall situation in
Afghanistan today is still remarkably di erent – and better – than it was in the early days of 2002.
The review suggests that the simpler, easier things have worked better than the more ambitious, com-
plex things. For example, donors were able to reduce child mortality, but the justice sector has hardly
improved. Wouldn’t that mean that all donors should invest all their resources in the easy things, and
avoid the complex things? Shouldn’t we all go after the low-hanging fruits? I believe that this would be a
Afghanistan, just like any other developing country in the midst of conﬂ ict, needs both the small things
as well as a gradual systemic change. The meta-review does not imply that aid cannot and does not
support this systematic change. It merely shows that our e orts were not as e ective as we had hoped.
This, I believe, should prompt us to be even more alert so that we don’t miss the next window of oppor-
tunity; it should prompt us to carefully balance interventions that have a high probability of immediate
impacts with long-term interventions that are necessary steps for future development, even if we won’t
see immediate impacts. And it should prompt us to identify ﬁ elds which should be avoided as long as
the context makes success unlikely.
The best possible development cooperation I think, is the one that constantly reﬂ ects on these issues
and is ready to adapt and improve. It is my hope that this meta-review contributes to such an endeavor.
Note: Reports in bold denote rigorous impact evaluations
Bilateral Country-Level Evaluations
Australia. (2017, September). Aid program performance report 2016 – 2017: Afghanistan. Australian
Government, Department of Foreign A airs and Trade.
Canada. (2015, March). Synthesis report: Summative evaluation of Canada’s Afghanistan development
program ﬁ scal year 2004 – 2005 to 2012 – 2013. Department of Foreign A airs, Trade and Development,
Denmark. (2012). Evaluation of Danish development support in Afghanistan. Ministry of Foreign A airs,
Finland. (2007, November). Evaluation: Finnish aid to Afghanistan. Ministry for Foreign A airs of Finland,
Department for Development Policy.
https://www.alnap.org/system/ﬁ les/content/resource/ﬁ les/main/erd-3613-full_0.pdf
New Zealand. (2013, March). New Zealand’s achievements from 10 years of development assistance in
Bamyan, Afghanistan. New Zealand Ministry of Foreign A airs and Trade. https://fyi.org.nz/request/3507/
Norway. (2012, March). Evaluation of Norwegian development cooperation with Afghanistan 2001 – 2011.
Norwegian Agency for Development Cooperation.
Norway. (2016). A good ally: Norway in Afghanistan 2001 – 2014. Norwegian Ministry of Foreign A airs
and Ministry of Defence. O cial Norwegian Reports NPOU 2016:8 (English translation from Norwegian).
Sweden. (2017). Summary of the report of the inquiry on Sweden’s engagement in
Afghanistan 2002 – 2014. Swedish Government, SOU. https://www.government.se/492dc9/
Sweden. (2018, April). Review of Sida’s support to Afghanistan: Lessons and conclusions from
7 evaluations. Swedish International Development Cooperation Agency.
United Kingdom. (2009, May). Evaluation of DFID’s country programmes: Afghanistan 2002 – 2007.
United Kingdom, Department for International Development.
United States. (2011, June 8). Evaluating U.S. foreign assistance to Afghanistan. A majority sta report,
prepared for the use of the committee on foreign relations. United States Senate.
Country Programs of Multilateral Donors
Independent Evaluation Group. (2013).Evaluation of World Bank programs in Afghanistan,
2002–11. Washington, DC: World Bank Group.
UNDP. (2014, July). Assessment of development results: Evaluation of UNDP contribution in Afghanistan.
Independent Evaluation O ce, United Nations Development Programme.
Afghanistan Reconstruction Trust Fund (ARTF)
Management Systems International (MSI). (2017). Year I annual report (November 2015 – October 2016).
Afghanistan Reconstruction Trust Fund.
Management Systems International (MSI). (2017). Year II annual report (November 2016 – October 2017).
Afghanistan Reconstruction Trust Fund.
Scanteam. (2008). Afghanistan Reconstruction Trust Fund: External evaluation. Final report. World Bank.
Scanteam. (2012). ARTF at a crossroads: History and the future. Final report. World Bank.
Scanteam. (2017). Taking charge: Government ownership in a complex context. External review.
Afghanistan Reconstruction Trust Fund. World Bank
Sida Decentralized Evaluation. (2015). Review of Afghanistan Reconstruction Trust Fund, ARTF, internal
and external studies and evaluations in Afghanistan – ﬁ nal report.
Beath, A., Fotini, C., & Enikolopov, R. (2013). Do elected councils improve governance? Experimental
evidence on local institutions in Afghanistan. World Bank.
Beath, A., Fotini, C., & Enikolopov, R. (2015). The national solidarity program: Assessing the e ects of
community-driven development in Afghanistan. World Bank.
Beath, A., Fotini, C., & Enikolopov, R. (2017). Can development programs counter insurgencies? Evidence
from a ﬁ eld experiment in Afghanistan. World Bank.
Chiwara, R. M., Afridi, H. R., & Jawhary, A. M. (2014). Afghanistan subnational governance programme
phase II, January 2010 – December 2014: Final evaluation report. United Nations Development Program
Collin, C., Jawhary, A. M., & Robertson, L. (2014). UNDP–Afghanistan CPD outcome 3 evaluation:
Final evaluation report. United Nations Development Program (UNDP).
Komorowska, K. (2016). Citizen voice in Afghanistan: Evaluation of National Solidarity Programme III.
Rao, M. P., & Alam, T. (2014). Final external evaluation report: National institution building project.
United Nations Development Program (UNDP) Sub-National Governance.
Saed, I. (2017). Local governance programme (LoGo), October 2015 – November 2017: Midterm evaluation
report. United Nations Development Program (UNDP).
Transtec. (2013). Evaluation report: Enhancing the capacity for inclusive local governance through
synergies and sustainable linkages between communities and government in North Afghanistan.
Rule of Law
Abbott, K., & Naderi, A. (2017). Independent evaluation of the UNDP JHRA phase II project.
Puric, O., Brooks, J., & Massoudi, K. (2018). Mid-term review of the Afghanistan access to justice
Gomez, C., & Baker, T. (2015). Final evaluation of the enhancing legal and electoral capacity for tomorrow
(ELECT) phase II: January 2012 – July 2015. United Nations Development Program (UNDP).
Transtec. (2013). Evaluation report: Citizens’ platform for democratic debates and dialogues in
Afghanistan. UDF-AFG-09-316. UNDEF.
Transtec. (2015). Evaluation report: Involving women and youth CSOs in strengthening democratic debate
and public news media around elections in Afghanistan. UDF-AFG-12-508. UNDEF.
Beath, A., Christia, F., & Nikolopov, R. (2013). Empowering women through development aid:
Evidence from a ﬁ eld experiment in Afghanistan. https://dspace.mit.edu/handle/1721.1/83869
Bernard, A. (2014). Forward-looking strategic evaluation of the UNICEF-supported female literacy
programme (2010–2013) in 34 provinces of Afghanistan. UNICEF.
FAO UN (Food and Agriculture Organization of the United Nations, O ce of Evaluation). (2016). Cluster
evaluation of: Strengthening the role of women in agriculture development for improved household food;
Strengthening policy development and coordination for food and nutrition security in Afghanistan; and
Support of extension systems, household food and livelihood security. Project Evaluation Series.
Haarr, R. N. (2015). External evaluation of the elimination of violence against women (EVAW) special fund
2008 – 2014. UN Women Afghanistan Country O ce.
Mushinga, M., & Fattahi, Z. (2016). UNDP gender equality project II: Final evaluation report. United Nations
Development Program (UNDP).
Ojha, G. P., & Fattahi, A. Z. (2015). Women’s empowerment and gender equality project: Mid-term
evaluation project. United Nations Development Program (UNDP).
Sayara Research. (2015). Impact evaluation report: EU empowerment project. Oxfam. Intervention.
Sloot, H., & Becker, S. (2013, December). Rights in crisis campaign Afghanistan project e ectiveness
review. Associate Consultants, The Coalition Factory.
Transtec. (2014). Evaluation report: Raising awareness about women’s social, political, and economic
rights in Afghanistan. UDF-AFG-10-379. UNDEF.
Watkins, F., & Azarbaijani-Moghaddam, S. (2019). Evaluation of UN women country programme in
Afghanistan: Final report. Sida Decentralized Evaluation.
Altai Consulting. (2017). AFD agriculture programs in Afghanistan (2005 – 2014): Final report.
AFD Evaluation 65.
Bhattacharjee, A., Postgate, D., & Andersen, H. (2013). Evaluation of CPD outcome 6: Diversiﬁ ed
livelihoods, private sector development and public–private partnership. UNDP Afghanistan.
Emmott, S. (2018). Support Afghanistan livelihoods and mobility (SALAM): Midterm evaluation report.
United Nations Development Program (UNDP).
Emmott, S., & Jawhary, A. M. (2014). Evaluation of the national area-based development programme
(NABDP) in Afghanistan. United Nations Development Program (UNDP).
Hussain, T., & Wasim, M. (2017). Mid-term review of UNDP GEF-LDCF2 project Afghanistan (27 April
2014 – 30 June 2017): Strengthening the resilience of rural livelihood options for Afghan communities in
Panjshi Balkh, Uruzgan and Herat Provinces to manage climate change-induced disaster risks. UNDP.
Burde, D., & Linden, L. L. (2012). The e ect of village-based schools: Evidence from a randomized
controlled trial in Afghanistan. National Bureau of Economic Research (No. w18039).
Integrity Watch Afghanistan. (2018). Education compromised? A survey of schools in
10 provinces of Afghanistan.
Samuel Hall Consulting. (2015). School-in-a-box 2015 evaluation. The Womanity Foundation.
UNESCO. (2016). Evaluation of UNESCO’s role in education in emergencies and protracted crises:
The e ects of police literacy training in Afghanistan.
UNICEF. (2015). Let us learn (LUL): Formative evaluation. UNICEF Afghanistan Country O ce.
UNICEF. (2017). Evaluation of improving street-working children’s access to education and livelihood
support for their families.
Adams, G. (2015). Honing the proper edge: CERP and the two-sided potential of military-led
development in Afghanistan. The Economics of Peace and Security Journal, 10(2), 53 – 60.
Beath, A., Fotini, C., & Enikolopov, R. (2017). Can development programs counter insurgencies? Evidence
from a ﬁ eld experiment in Afghanistan. MIT Political Science Department.
Böhnke, J., & Zürcher, C. (2013). Aid, minds and hearts: The impact of aid in conﬂ ict zones. Conﬂ ict
Management and Peace Science, 30(5), 411 – 432.
Child, T. (2014). Hearts and minds cannot be bought: Ine ective reconstruction in Afghanistan. The
Economics of Peace and Security Journal, 9(2), 43 – 49. doi:10.15355/epsj.9.2.43
Chou, T. (2012). Does development assistance reduce violence? Evidence from Afghanistan.
The Economics of Peace and Security Journal, 7(2), 5 – 13.
Fishstein, P., & Wilder, A. (2012). Winning hearts and minds? Examining the relationship between aid and
security in Afghanistan. Feinstein International Center. http://ﬁ c.tufts.edu/publication-item/winning-
Gordon, S. (2011). Winning hearts and minds? Examining the relationship between aid and security in
Afghanistan’s Helmand Province. Feinstein International Center.
Haysom, S., & Jackson, A. (2013). “You don’t need to love us”: Civil–military relations in Afghanistan,
2002–13. Stability: International Journal of Security and Development, 2(2).
Karell, D. (2015). Aid, power, and grievances: Lessons for war and peace from rural Afghanistan. The
Economics of Peace and Security Journal, 10(2), 43 – 52.
Mercy Corps. (2015). Does youth employment build stability? Evidence from an impact evaluation
of vocational training in Afghanistan. https://www.mercycorps.org/sites/default/ﬁ les/MercyCorps_
Plumb, R. I., Shapiro, J. N., Crisman, B., Singh, M., & Mao, J. (2017). Stabilization in Afghanistan: Trends in
violence, attitudes, well-being and program activity. RAND Corporation.
Sexton, R. (2016). Aid as a tool against insurgency: Evidence from contested and controlled territory in
Afghanistan. American Political Science Review 110(4), 731–749.
Anwari, Z., Shukla, M., Maseed, B. A., Wardak, G. F. M., Sardar, S., Matin, J., Trasi, R. (2015). Implementing
people-centered health systems governance in 3 provinces and 11 districts of Afghanistan: A case
study. Conﬂ ict and Health, 9(1). doi:10.1186/1752-1505-9-2
Carvalho, N., Salehi, A. S., & Goldie, S. J. (2013). National and sub-national analysis of the health beneﬁ ts
and cost-e ectiveness of strategies to reduce maternal mortality in Afghanistan. Health Policy and
Planning, 28(1), 62 – 74. doi:10.1093/heapol/czs026
Central Statistics Organization (CSO), Ministry of Public Health (MoPH), and ICF. (2017). Afghanistan
demographic and health survey 2015. Kabul, Afghanistan: Central Statistics Organization.
Edmond, K. M., Yousuﬁ , K., Anwari, Z., Sadat, S. M., Staniczai, S. M., Higgins-Steele, A., Smith, E. R. (2018).
Can community health worker home visiting improve care-seeking and maternal and newborn care
practices in fragile states such as Afghanistan? A population-based intervention study. BMC Medicine,
Engineer, C. Y., Dale, E., Agarwal, A., Agarwal, A., Alonge, O., Edward, A., Peters, D. H. (2016).
E ectiveness of a pay-for-performance intervention to improve maternal and child health services
in Afghanistan: A cluster-randomized trial. International Journal of Epidemiology, 45(2), 451 – 459.
Lin, A. (2016). Stimulating demand: An assessment of the conditional cash transfer project in Afghanistan.
(Unpublished doctoral dissertation). Johns Hopkins University, Baltimore, Maryland.
Lin, A., & Salehi, A. S. (2013). Stimulating demand: E ects of a conditional cash transfer programme on
increasing maternal and child health-service utilisation in Afghanistan,
a quasi-experimental study. The Lancet, 381, S84. doi:10.1016/S0140-6736(13)61338-0
Mayhew, M., Ickx, P., Stanekzai, H., Mashal, T., & Newbrander, W. (2014). Improving nutrition in
Afghanistan through a community-based growth monitoring and promotion programme:
A pre–post evaluation in ﬁ ve districts. Global Public Health, 9(sup1), S58 – S75. doi:10.1080/17441692.2
Rao, K. D., Waters, H., Steinhardt, L., Alam, S., Hansen, P., & Naeem, A. J. (2009). An experiment with
community health funds in Afghanistan. Health Policy and Planning, 24(4), 301 – 311. doi:10.1093/
Society for Sustainable Development of Afghanistan (SSDA). (2017). Evaluation of the WASH in schools
(WinS) programme (2008 – 2014): Evaluation report. Kabul, Afghanistan: UNICEF Afghanistan.
Speakman, E. M., Shaﬁ , A., Sondorp, E., Atta, N., & Howard, N. (2014). Development of the community
midwifery education initiative and its inﬂ uence on women’s health and empowerment in Afghanistan: A
case study. BioMed Central 14, 111.
Turkmani, S., Currie, S., Mungia, J., Asseﬁ , N., Javed Rahmanzai, A., Azfar, P., & Bartlett, L. (2013).
Midwives are the backbone of our health system: Lessons from Afghanistan to guide expansion of
midwifery in challenging settings. Midwifery 29, 1166 – 1172.
Witvorapong, N., & Foshanji, A. I. (2016). The impact of a conditional cash transfer program on the
utilization of non-targeted services: Evidence from Afghanistan. Social Science & Medicine, 152, 87 – 95.
World Health Organization. (2015). Trends in maternal mortality, 1990 to 2015: Estimates by WHO, UNICEF,
UNFPA, World Bank Group and the United Nations Population Division. Geneva, Switzerland: WHO.
Zainullah, P., Ansari, N., Yari, K., Azimi, M., Turkmani, S., Azfar, P., Bartlett, L. (2014). Establishing
midwifery in low-resource settings: Guidance from a mixed-methods evaluation of the Afghanistan
midwifery education program. Midwifery 30, 1056 – 1062
Cole, P. (2018). Assessing the impact of a renewable energy programme in Bamyan, Afghanistan:
The value of a capability approach. Energy for Sustainable Development, 45, 198 – 205.
Shoaib, A., & Ariaratnam, S. (2016). A study of socioeconomic impacts of renewable energy projects
in Afghanistan. Procedia Engineering, 145, 995 – 1003.
Loschmann C., Parsons, C. R., & Siegel, M. (2015). Does shelter assistance reduce poverty in
Afghanistan? World Development, 74, 305 – 322.
Asian Development Bank (ADB)
ADB. (2011). Afghanistan: Multisector program. Validation report. Independent Evaluation Department.
ADB. (2012). Afghanistan: Agriculture sector program. Validation report. Independent Evaluation
ADB. (2012). Afghanistan: Andkhoy – Qaisar road project. Validation report. Independent Evaluation
ADB. (2012). Afghanistan: Emergency infrastructure rehabilitation and reconstruction project. Validation
report. Independent Evaluation Department.
ADB. (2012). Equity investment Afghanistan: Afghanistan International Bank. Validation report. Independent
ADB. (2012). Afghanistan: Regional airports rehabilitation project (Phase 1). Validation report. Independent
ADB. (2012). Country assistance program evaluation: Islamic Republic of Afghanistan. Independent
ADB. (2013). Afghanistan: Fiscal management and public administration reform program. Validation report.
Independent Evaluation Department.
ADB. (2014). Afghanistan: Private sector and ﬁ nancial market development. Validation report. Independent
ADB. (2014). Afghanistan and Tajikistan: Regional power transmission interconnection project. Validation
report. Independent Evaluation Department.
ADB. (2015). Afghanistan: Hairatan to Mazar-e-Sharif railway project. Validation report. Independent
ADB. (2016). ADB guidelines for public sector operations. Independent Evaluation Department.
ADB. (2017). Afghanistan: Power transmission and distribution project. Validation report. Independent
ADB. (2017). Afghanistan: Validation report of the country partnership strategy ﬁ nal review, 2009
to mid-2015. Independent Evaluation Department.
Special Inspector General for Afghanistan Reconstruction (SIGAR)
SIGAR. (2009, October). Women’s participation in elections. SIGAR Audit 10-1.
SIGAR. (2010, January). Energy sector. SIGAR Audit 10-4.
SIGAR. (2011, March). Economic and social development/NSP. SIGAR Audit 11-8.
SIGAR. (2011, October). Governance and development/agriculture. SIGAR Audit 12-1.
SIGAR. (2012, July). Infrastructure. SIGAR Audit 12-12.
SIGAR. (2013, April). Health services in Afghanistan. SIGAR Audit 13-9.
SIGAR. (2013, July 30). Contract oversight. SIGAR Quarterly Report.
SIGAR. (2013, September). Health services in Afghanistan. SIGAR Audit 13-17.
SIGAR. (2014, January). State Department support of Afghan justice sector. SIGAR 14-26-AR.
SIGAR. (2014, April). Afghanistan’s water sector. SIGAR 14-52-AR.
SIGAR. (2014, April 30). Corruption. SIGAR Quarterly Report.
SIGAR. (2014, July 30). Sustainability. SIGAR Quarterly Report.
SIGAR. (2014, December). U.S. e orts to support Afghan women. SIGAR 15-24-AR.
SIGAR. (2015, April). Afghanistan’s extractive industries. SIGAR 15-55-AR.
SIGAR. (2015, July). Rule of law. SIGAR 15-68-AR.
SIGAR. (2015, July 30). Conditionality. SIGAR Quarterly Report.
SIGAR. (2016, January). Afghanistan’s extractive industries. SIGAR 16-11-AR.
SIGAR. (2016, January). Review letter: USAID-supported health facilities in Kabul. SIGAR 16-09-SP.
SIGAR. (2016, January 30). Economic challenges. SIGAR Quarterly Report.
SIGAR. (2016, April). U.S. e orts to support Afghan education. SIGAR 16-32-AR.
SIGAR. (2016, April). Lessons from the coalition: National experiences from the Afghanistan reconstruction.
SIGAR. (2016, June). Review letter: USAID-supported health facilities in Badakhshan. SIGAR 16-40-SP.
SIGAR. (2016, July 30). Electriﬁ cation. SIGAR Quarterly Report.
SIGAR. (2016, September). Corruption in conﬂ ict: Lessons learned from the U.S. experience in Afghanistan.
SIGAR. (2016, October). U.S. e orts to support Afghanistan’s road infrastructure. SIGAR 17-11-AR.
SIGAR. (2016, October 30). Afghan women. SIGAR Quarterly Report.
SIGAR. (2016, November). Schools in Herat Province: Observations from site visits at 25 schools.
SIGAR. (2017, January). USAID support for Afghanistan’s health care. SIGAR 17-22-AR.
SIGAR. (2017, February). Land reform. SIGAR 17-27-AR.
SIGAR. (2017, March). Schools in Balkh Province: Observations from site visits at 26 schools.
SIGAR. (2017, March). USAID-supported health facilities in Ghazni Province: Observations from site visits to
30 locations. SIGAR 17-34-SP.
SIGAR. (2017, July). USAID-supported health facilities in Takhar Province: Observations from sites visits to
35 locations. SIGAR 17-51-SP.
SIGAR. (2017, September). Health facilities in Nangarhar Province, Afghanistan: Observations from visits at
four facilities. SIGAR 17-67-SP.
SIGAR. (2017, October). 2011 AIF projects. SIGAR 18-10-AR.
SIGAR. (2017, November). USAID-supported health facilities in Khost Province, Afghanistan:
Observations from 20 site visits. SIGAR 18-13-SP.
SIGAR. (2017, December). Schools in Faryab Province, Afghanistan: Observations from site visits at
17 schools. SIGAR 18-17-SP.
SIGAR. (2018, January 30). Mineral development. SIGAR Quarterly Report.
SIGAR. (2018, February). Schools in Kabul Province, Afghanistan: Observations from site visits at
24 schools. SIGAR 18-31-SP.
SIGAR. (2018, April). Afghanistan reconstruction trust fund. SIGAR 18-42-AR.
SIGAR. (2018, April). Schools in Kunduz Province, Afghanistan: Observations from site visits at 6 schools.
SIGAR. (2018, April). Private sector development and economic growth: Lessons from the U.S. experience
in Afghanistan. SIGAR 18-38-LL.
SIGAR. (2018, May). Afghanistan’s anti-corruption e orts. SIGAR 18-51-AR.
SIGAR. (2018, May). Stabilization: Lessons from the U.S. experience in Afghanistan. SIGAR 18-48-LL.
SIGAR. (2018, June). Review: USAID-supported health facilities in Kandahar Province, Afghanistan:
Observations from site visits at 9 schools. SIGAR 18-55-SP.
SIGAR. (2018, July). Regional agriculture development program. SIGAR 18-65-AR.
SIGAR. (2018, August). Schools in Parwan Province, Afghanistan: Observations from site visits at
14 schools. SIGAR 18-67-SP.
SIGAR. (2018, September). USAID’s promoting gender equity in national priority programs.
SIGAR. (2018, September). Bridges in Baghlan Province, Afghanistan: Six of eight bridges constructed or
rehabilitated by DOD remain in generally good, usable condition; two appeared to have structural issues
needing attention. SIGAR 18-70-SP.
SIGAR. (2018, October). State Department’s good performer’s initiative: Status of six completed projects in
Takhar Province. SIGAR 19-02-SP.
SIGAR. (2018, December). Bridges in Kabul Province, Afghanistan: Six Bridges are generally in good
condition, but Afghan government lacks budget for sustained maintenance. SIGAR 19-08-SP.
Christoph M. Zürcher
University of Ottawa
Graduate School of Public and International A airs
120 University, Ottawa, ON K1N 6N5, Canada
T +1 613-562-5800-8997
F +1 613-562-5241
Christoph M. Zürcher, Ottawa, Canada
Barbara Reuter, Oberursel, Germany
E barbarareuter-graﬁ email@example.com
This publication contains links to external websites. Responsibility for the content
of the listed external sites always lies with their respective publishers.
The present report was commissioned by the BMZ. The responsibility for the content
lies solely with the publisher and does not reﬂ ect the views of the BMZ.
On behalf of
Federal Ministry for Economic Cooperation and Development (BMZ), Germany
Division for Afghanistan and Pakistan