www.oecd.org

D
e
ve

lo
p
m

e
n
t C

o
-o

p
e
ra

tio
n
R

e
p
o
rt 2

0
0
5

O
E

C
D

Jo
u
rn

a
l o

n
D

e
ve

lo
p
m

e
n
t, V

o
lu

m
e
7

, N
o
. 1

This book is available via SourceOECD: www.SourceOECD.org/developmentreport.

SourceOECD is the OECD’s online library of books, periodicals and statistical databases.
For more information about this award-winning service and free trials ask your librarian, or write to us at
SourceOECD@oecd.org.

The OECD Development Assistance Committee (DAC) Development Co-operation Report
is the key annual reference document for statistics and analysis on the latest trends in
international aid. It provides reflections on 2005, the “Year of Development”, when questions
about the volume and effectiveness of aid were centre stage, and includes a unique set of
facts and figures which help to contextualise the development debates and decisions of 2005.

Chapter 1 contains an overview by the Chair of the DAC, who reflects on progress made in
2005, on the prospects for turning rhetoric on more and better aid into action, and considers
the importance of two major issues: building capacity in developing countries and gender.
Chapter 2 contains a synthesis of DAC work on pro-poor growth. Chapter 3 reports on the
Paris High Level Forum on Aid Effectiveness, and argues that this key event will genuinely
make a difference. Chapter 4 outlines the main features of the aid programmes of all DAC
members, including brief information on the programmes of other donors outside the OECD.
Chapter 5 contains a new analysis of technical co-operation, one of the largest forms of
development assistance which is often poorly understood. Finally, the Annex provides the
definitive source of statistics on aid and other resource flows to developing and transition
countries, analysed by donor, recipient, sector and type.

This is the first issue to be published on line as part of our efforts to improve the accessibility
of key OECD DAC work. We aim to respond to issues in the aid community promptly and
with the best available analysis and statistics, when you most need them.

OECD Journal on Development, Volume 7, No. 1

Development Co-operation Report 2005

Extract from the OECD Journal on Development,
Volume 8 No. 2 PRE-PUBLICATION COPY

Aid Effectiveness

Overview of the Results

2006 Survey
on Monitoring
the Paris Declaration

2006 SURVEY
ON MONITORING
THE PARIS DECLARATION

OVERVIEW OF THE RESULTS

THE PARIS DECLARATION ON AID EFFECTIVENESS (March 2005) defined a number of commitments on the
part of donors and partner countries, and a set of indicators to measure progress towards 2010. Within
the Working Party on Aid Effectiveness, the Joint Venture on Monitoring the Paris Declaration is
responsible for the monitoring and follow-up of the Paris Declaration. This document is the first
Baseline Survey on Monitoring the Paris Declaration covering 34 partner countries with data
from 60 donors.

Although this is a baseline survey, a number of issues stand out, requiring improvements and actions.
Examples are:

■ The cost of uncoordinated aid is very high. There are too many actors with competing
objectives, especially in the poorest and most aid-dependent countries, leading to
high transaction costs.

■ There is still slow progress in untying of aid and technical co-operation is still
too much donor-driven.

■ Good headquarters policies are not always matched by in-country practices.
■ There is a need to strengthen country ownership. Mainly a partner responsibility,

donors can assist by capacity development and alignment to country systems.
■ A lot of work needs to be done in managing for results. Mutual accountability,

a key concept in the Paris Agenda, calls for performance assessment frameworks and
improved incentive systems in both partner and donor countries.

Increasing the impact of aid on development is fundamental and requires ambitious reforms for
both donors and partners, not least to defend the scaling up of aid. The Paris Declaration is designed
to help deliver these reforms by holding donors and partners accountable for progress in fulfilling
the commitments.

Since 2005 a lot of work has been carried out by the various actors. Donors and partner countries have
taken a number of initiatives in line with the Paris Agenda to achieve concrete improvements on the
ground. This survey gives some evidence of the challenges ahead, challenges that will require political
commitment from both donors and partner countries.

Jan Cedergren
Chair, OECD-DAC Working Party on Aid Effectiveness

Richard Manning
Chair, OECD Development Assistance Committee

FOREWORD

3THE 2006 SURVEY ON MONITORING THE PARIS DECLARATION, VOLUME 1: PRE-PUBLICATION COPY - © OECD 2007

THE 2006 SURVEY ON MONITORING THE PARIS DECLARATION was prepared under the aegis of the the
Development Assistance Committee (DAC) Working Party on Aid Effectiveness, and is based on the
work of the Joint Venture on Monitoring the Paris Declaration, co-chaired by Pham Thi Thanh An
(Viet Nam) and Gregory Briffa (United Kingdom) and then George Carner (United States).
The Overseas Development Institute (ODI), a British-based independent think tank, assisted with
the drafting. David Booth and Verena Fritz contributed to Volume 1 (Overview) and Volume 2
(Country Chapters) involved David Booth, Roo Griffiths, Alan Hudson, Kate Kilpatrick, Dinah
McLeod and Sarah Mulley.

The Working Party would like to communicate its special thanks to the governments of the 34 countries
that participated in this survey and in particular to the National Co-ordinators who managed the
process at country level:

ACKNOWLEDGEMENTS

5THE 2006 SURVEY ON MONITORING THE PARIS DECLARATION, VOLUME 1: PRE-PUBLICATION COPY - © OECD 2007

AFGHANISTAN
Mustapha Aria
ALBANIA
Albana Vokshi
Mezir Haldeda
BANGLADESH
Nargis Islam
BENIN
Rigobert Laourou
BOLIVIA
Fernando Jiménez Zeballos
BURKINA FASO
Justin Hien
BURUNDI
Pierre C. Abega Rurakamvye
CAMBODIA
Chhieng Yanara
CAPE VERDE
Miryan Vieira
Marco Antonio Ortega
CONGO
Benjamin Bonge Gibende
DOMINICAN REPUBLIC
América Bastidas
EGYPT
Fayza Aboulnaga

ETHIOPIA
Hailemichael Kinfu
GHANA
Helen Allotey
Agatha Gaisie-Nketsiah
HONDURAS
Ricardo Arias
KENYA
Jackson Kinyanju
Bernard Masiga
KYRGYZ REPUBLIC
Sanjar Makanbetov
Dinara Djoldosheva
MALAWI
Naomi Ngwira
MALI
Modibo Makalou
MAURITANIA
Isselmou Ould Sidi El Moctar
MOLDOVA
Lucretia Ciurea
MONGOLIA
Ochirkhuu Erdembileg
MOZAMBIQUE
Hanifa Ibrahimo

Mark Baldock provided the statistical analysis. The report was edited by Laura Boutin and Patricia
Rogers (Executive Summary and Chapter 3), and designed by Peggy Ford-Fyffe King. The survey was
prepared under the direction of Simon Mizrahi, Senior Policy Advisor at the OECD’s Development
Co-operation Directorate (DCD) with the assistance of Kjerstin Andreasen.

NICARAGUA
Mauricio Gómez
NIGER
Yakoubou Mahaman Sani
PERU
Agustín Haya de la Torre
RWANDA
Jean-Jacques Nyirubutama
SENEGAL
Thierno Seydou Niane
SOUTH AFRICA
Elaine Venter
TANZANIA
Joyce Mapunjo
UGANDA
Deo Kamweya
Damon Kitabire
Peter M. Ssentongo
VIET NAM
Pham Thi Thanh An
YEMEN
Nabil Shaiban
ZAMBIA
Paul Lupunga

FOREWORD 3

ACKNOWLEDGEMENTS 5

EXECUTIVE SUMMARY 9

Chapter 1 KEY FINDINGS FROM THE SURVEY 15

Chapter 2 THE SURVEY PROCESS 39

Chapter 3 CONCLUSIONS AND RECOMMENDATIONS 51

STATISTICAL APPENDICES 61

A COUNTRY DATA

(One table per Indicator) 65

B DONOR DATA

(One table per Indicator) 79

C DONOR DATA

(One table per Donor) 89

D SURVEY QUESTIONNAIRES 121

E GLOSSARY OF KEY TERMS 125

TABLE OF CONTENTS

7THE 2006 SURVEY ON MONITORING THE PARIS DECLARATION, VOLUME 1: PRE-PUBLICATION COPY - © OECD 2007

9THE 2006 SURVEY ON MONITORING THE PARIS DECLARATION, VOLUME 1: PRE-PUBLICATION COPY - © OECD 2007

IN 2005, OECD COUNTRIES INVESTED MORE THAN USD 100 BILLION to advance welfare and eradicate poverty
in developing countries. If this foreign aid is to do as much good as it possibly can, it must be used as
effectively as possible. Better aid means very different things depending on which side of the develop-
ment fence you are on. For countries that receive foreign aid, it means allocating resources to their
policy priorities and plans to promote justice, stimulate economic growth and improve social welfare
for their citizens. For countries and organisations that provide foreign aid, it means allocating funding
to countries that need it most and are more likely to make best use of it (or that need special help to
weather crises). But the true test of aid effectiveness is improvement in people’s lives.

A layperson observing today’s aid industry might be understandably baffled by the sheer number of aid
actors, funds and programmes. The last time the OECD counted, there were more than 200 bilateral
and multilateral organisations channelling official development assistance. Many developing countries
may have more than 40 donors financing more than 600 active projects, and may still not be on track
to achieve the Millennium Development Goals.

More than 100 countries and donor organisations recognised the imperative of managing aid more ratio-
nally when, on March 2, 2005, they endorsed the Paris Declaration on Aid Effectiveness, an ambitious
plan to reform the system of aid delivery.

THE PARIS DECLARATION RESTS ON FIVE COMMON-SENSE TENETS, that aid is more likely to promote
development when:
■ Developing countries exercise leadership over their development policies and plans

(ownership).
■ Donors base their support on countries’ development strategies and systems (alignment).
■ Donors co-ordinate their activities and minimise the cost of delivering aid (harmonisation).
■ Developing countries and donors orient their activities to achieve the desired results

(managing for results).
■ Donors and developing countries are accountable to each other for progress in managing

aid better and in achieving development results (mutual accountability).

In the Paris Declaration, donors and partners also committed to monitoring their progress in improving
aid effectiveness against 56 specific actions, from which 12 indicators were established and targets set
for 2010. This report presents the findings and recommendations of the first round of monitoring that
was conducted in 2006 on the basis of activities undertaken in 2005. The conclusions of this report are
based on a body of evidence from a baseline survey of 34 self-selected countries, and a comprehensive
list of donor organisations covering 36% of aid programmed across the world in 2005. The conclusions
clearly show that in half of the developing countries signing on to the Paris Declaration, partners and
donors have a long road ahead to meet the commitments they have undertaken.

THIS REPORT IS DIVIDED IN TWO VOLUMES. Volume 1 (120 pages) presents an overview of key find-
ings across the 34 countries (Chapter 1), assesses the survey process (Chapter 2), and sets out some
key conclusions and recommendations (Chapter 3). A statistical appendix provides all the data that
underpin the analysis. Volume 2 (400 pages) includes a detailed analysis for each of the 34 countries
that undertook the survey. This Executive Summary covers the key conclusions and recommendations
of the full report (Volumes 1 and 2). ■

EXECUTIVE SUMMARY

10 THE 2006 SURVEY ON MONITORING THE PARIS DECLARATION, VOLUME 1: PRE-PUBLICATION COPY - © OECD 2007

MAIN POLICY IMPLICATIONS OF THE 2006 SURVEY

HIGH EXPECTATIONS FOR REFORM

The survey shows that the Paris Declaration has
stimulated an important dialogue at country
level on how to improve aid. All donor agen-
cies have made major efforts to implement the
Paris Declaration within their organisations and
communicate its importance to their staff, and
there has been at least some implementation
activity in over 60 countries.

Regional-level initiatives to disseminate and
communicate the Paris agenda have included
five regional workshops on aid effectiveness in
2005-06 in Uganda, Mali, South Africa, Bolivia
and the Philippines. From the shores of Lake
Victoria to the heights of Santa Cruz the message
came out loud and clear: one size does not fit
all. For aid to be effective, each country needs to
determine its own priorities, pace and sequencing
of reforms.

At the same time, many partner countries in
regional workshops and elsewhere are voicing
concerns about the high transaction costs of
managing foreign aid and the slow pace of change
in donor practices. They see a strong disconnect
between headquarters policies and in-country
practices, as illustrated by continued donor-
driven technical co-operation and lack of visible
progress on untying aid. Demonstrating tangible
changes in practices is fundamental in sustaining
momentum and achieving further progress by the
next High-Level Forum in Ghana (September
2008) and the 2010 targets.

DEEPER OWNERSHIP,

MORE ACCOUNTABLE INSTITUTIONS

AND SYSTEMS

The quality of partner countries’ national devel-
opment plans or poverty reduction strategies is
one concrete measure of country ownership. The
survey findings show that national development
strategies need substantial strengthening if coun-
tries are to meet the 2010 target. In 2005, only five
countries (17% of the sample) met the six criteria
for sound operational development strategies. The
most common failing was weakness in the mech-
anisms linking budget formulation and execution
to national plans, policy priorities and results.

Improving transparency and accountability
on the use of development resources is also an
important objective of the Paris Declaration.
Strengthening the credibility of the budget as a
tool for governing the allocation and use of devel-
opment resources (domestic and external) can not
only improve the alignment of donor support,
but also permit parliamentary scrutiny of govern-
ment policies on development – which is key to
deepening ownership. In 2005, the survey indi-
cates, for nearly all countries the credibility of
development budgets is undermined by sizable
inaccuracies in the budget estimates of aid flows.
Meeting the 2010 target will require donors and
partner countries to work together so that budget
estimates are both more comprehensive and more
realistic, and aid is better aligned.

STRENGTHENING AND USING

COUNTRY SYSTEMS

The Paris Declaration encourages donors to
increasingly use strengthened country systems
(for public financial management, procurement,
environment, monitoring and evaluation, and
other country systems) so that partner countries
are empowered to develop institutions that can
implement and account for their development
policies and resource use to citizens and parlia-
ments. On average, the survey shows, in 2005,
39% of aid flows for the government sector used
country public financial management (PFM)
and procurement systems. The degree to which
donors rely on country public financial manage-
ment systems varies considerably, depending at

11THE 2006 SURVEY ON MONITORING THE PARIS DECLARATION, VOLUME 1: PRE-PUBLICATION COPY - © OECD 2007

least partly on the quality of the systems and
on such other factors as the existence of reform
programmes. Progress will depend on greater
understanding of the development benefits and
risks of using these systems, and sustained and
long-term efforts to strengthen capacity. More
countries need to use performance assessment
tools to reform and improve their systems (for
example, the Public Expenditure and Financial
Accountability (PEFA) framework for public
financial management systems). Partner coun-
tries need to take the lead in defining capacity
development priorities, and donors should direct
their technical and other assistance to imple-
menting co-ordinated strategies.

INCREASING AID EFFICIENCY

AND DONOR HARMONISATION

A key aim of the aid effectiveness agenda is to
decrease the transaction costs of delivering aid,
especially those that burden developing coun-
tries by requiring them to manage multiple
programmes using different donor procedures.

The survey provides clear evidence that the cost
of managing aid is high for partner countries.
On a business-as-usual basis, transaction costs
for partner countries can be expected to increase
significantly as the volume of aid is scaled up,
new (emerging) donors become more active and
further special initiatives are created. Donors will
need to work aggressively to reduce the transaction
costs of delivering and managing aid. They should
give special attention to: increasing complemen-
tarity and rationalising division of labour; making
greater use of local harmonisation and alignment
action plans, and of sector-wide and programme-
based approaches; expanding use of delegated
co-operation and other innovative approaches; and
reducing the number of project implementation
units and better integrating them into ministries.

Donors and partners must acknowledge that
initially there are new costs associated with doing
business differently. These costs constitute an up-
front investment in doing business more effec-
tively and should be factored into operational
budgets and allocation of staff time.

MANAGING FOR RESULTS

The commitments on management for develop-
ment results call for donors and partner coun-
tries to direct resources to achieving results,
and using information on results to improve
decision making and programme performance.
The survey suggests that translating evidence
on results into processes of policy improvement
remains a major challenge in the large majority of
surveyed countries. Countries and donors should
use performance assessment frameworks and
more cost-effective results-oriented reporting.
This, too, will require donors to invest in capacity
development and rely more on country results
reporting systems.

MUTUAL ACCOUNTABILITY

The Paris Declaration introduced the concept of
mutual accountability – that aid is more effec-
tive when donors and partner governments are
not only accountable to their respective publics
for the use of resources to achieve development
results, but are also accountable to each other
for better management of aid. The survey shows
that the work to establish specific mechanisms
for joint monitoring of aid effectiveness commit-
ments at country level is just beginning, and
more efforts will be needed to achieve the target
by 2010. Aid effectiveness issues and results need
to be discussed more explicitly at country level,
and credible monitoring mechanisms need to be
developed.

REORIENTING CORPORATE-LEVEL INCENTIVES

Most development agencies have taken impor-
tant steps to advance implementation of the
Paris Declaration, and donor field missions have
increased their local efforts to align and harmo-
nise their programmes. Nevertheless, the survey
suggests that, at the corporate level, a number
of hurdles work against donors’ ability to meet
the commitments made in Paris. For example,
in many agencies the Paris Declaration is still
principally owned by policy staff at headquar-
ters, while at country level, harmonisation tasks
are sometimes seen as getting in the way of
efforts to achieve tangible development results.
At the corporate level, policy makers are encour-
aged to review rules and procedures that get in
the way of meeting the Paris commitments. For
example, pressure to commit and disburse funds,
limited flexibility for staff to devote time to co-
ordination, and high staff turnover, which create
incentives that reward short-term benefits over
longer-term, and collective, gains.

12 THE 2006 SURVEY ON MONITORING THE PARIS DECLARATION, VOLUME 1: PRE-PUBLICATION COPY - © OECD 2007

IMPROVING MONITORING

OF THE PARIS DECLARATION

It is evident from the feedback on the baseline
survey and the challenges described in Chapter 2
that the interim progress survey planned for 2008
will need to be improved in several ways. Planned
changes include: improving the guidance; clari-
fying and standardising the definitions; comple-
menting the scope of the survey with localised
and qualitative data; expanding country coverage;
including more fragile states; strengthening the
role of National Co-ordinators; reducing the
burden to partners and donors in filling out the
survey; and ensuring that the 2007 aid disburse-
ment data is collected at country level well before
the end of the first quarter of 2008. Organising a
successful second round of monitoring, in such a
short period of time, is no small undertaking. It
will require very careful planning, communica-
tion and early involvement of donors and partner
countries. Beyond the survey, a medium-term
monitoring plan is being developed for regional
and national integration of ongoing monitoring
efforts, planned evaluations, donor self-reporting
and DAC peer reviews. The medium-term moni-
toring plan will also suggest what strategic use
could be made of the survey results at the country,
regional or international levels. ■

The survey findings and the discussions that have taken place around them point to six major priority
areas that need policy makers’ attention right now if countries and donors are to accelerate progress
towards achieving the Paris Declaration commitments.

1. Partner countries need to deepen their ownership of the development process
by engaging their citizens and parliaments more fully in planning and assessing their
development policies and programmes. They should also increase efforts to link their
plans much more closely to their budget and results frameworks.

2. Donors need to support these efforts by making better use of partners’ national
budgets to align their programmes with country priorities. They also need to improve
the transparency and predictability of aid flows by sharing timely and accurate
information on intended and actual disbursements with budget authorities.

3. Partner countries need to take the lead in determining priority programmes of capacity
development, especially those needed to improve country systems. Donors can help by
better co-ordinating their technical assistance with country priorities and fully involving
partners when commissioning technical assistance.

4. To further harmonisation, donors must work aggressively to reduce the transaction
costs of delivering and managing aid. They should give special attention to enhancing
complementarity and rationalising the division of labour; increasing use of local
harmonisation and alignment action plans; increasing use of programme-based
approaches such as sector-wide approaches and budget support; expanding reliance on
delegated co-operation and other innovative approaches; reducing the number of
project implementation units and better integrating them into ministries; and increasing
efforts on untying as encouraged by the DAC recommendation.

5. To promote managing for results, countries and donors should make greater use of
performance assessment frameworks and more cost-effective results-oriented reporting.
This, too, will require donors to invest further in capacity development and increase
their use of country results reporting systems.

6. To begin addressing mutual accountability commitments, countries and donors
should clearly define a mutual action agenda and discuss aid effectiveness progress
and development results more explicitly at country level by using country dialogue
mechanisms (e.g. revamped Consultative Group and round table meetings) and
developing credible monitoring mechanisms where needed. ■

13THE 2006 SURVEY ON MONITORING THE PARIS DECLARATION, VOLUME 1: PRE-PUBLICATION COPY - © OECD 2007

KEY POLICY RECOMMENDATIONS

15THE 2006 SURVEY ON MONITORING THE PARIS DECLARATION, VOLUME 1: PRE-PUBLICATION COPY - © OECD 2007

THIS CHAPTER PROVIDES AN OVERVIEW of the survey findings.
The survey process is described and assessed in Chapter 2.
Major conclusions, lessons and recommendations arising
from the exercise are drawn together in Chapter 3.

Like the country studies on which it is based, this chapter
draws principally on the survey returns completed in
September 2006 under the auspices of the National Co-
ordinator for each country. A second important source is the
World Bank’s 2005 Comprehensive Development Framework
Progress Report (henceforth CDF Progress Report) and the
corresponding country profiles. The updated country profiles
prepared for the Bank’s forthcoming Aid Effectiveness
Review (henceforth AER) have also been consulted. The
quantitative assessments for 2005 contained in the CDF
Progress Report are the agreed basis for Indicators 1 (country
strategies) and 11 (performance assessment systems). Finally,
the World Bank’s annual Country Policy and Institutional
Assessment (CPIA) is the source for Indicator 2a (quality of
public financial management), and OECD data are used on
aid tying. These sources are listed at the end of the chapter.

THE PARIS DECLARATION is a joint undertaking on the part of
the donor community and partner countries. The commit-
ments are highly interdependent. That is, they are only likely
to be realised through a combined effort that acts on both
sides of the problem. For example, country ownership of
development efforts depends to a significant degree on donor
behaviour, while several dimensions of alignment depend on
actions by countries. Because of this interdependence, the
baseline situation in a country can be significantly influ-
enced by the size and composition of the country’s donor
community. The reverse also applies. The baselines for each
donor agency will be affected by the pattern of its engage-
ment across countries. In presenting the results, this chapter
takes into account the possible influence of this type of factor
along with other limitations on comparability across coun-
tries and across donors, such as variations in reporting.

1 KEY FINDINGS
FROM THE SURVEY

ACRONYMS

AER Aid Effectiveness Review

CDF Comprehensive Development
F ramework

CFAA Country Financial
Accountability Assessment

CPIA Country Policy and
Institutional Assessment

DBS direct budget support

HAP Harmonisation Action Plan

IDA International Development
Association

LICUS Low-Income Countries
Under Stress

MTEF medium-term expenditure
framework

NDS National Development
Strategy

ODA official development
assistance

PAF performance assessment
framework

PBA programme-based approach

PEFA Public Expenditure and
Financial Accountability

PFM public financial
management

PIU project implementation unit

PSRP Poverty Reduction
Strategy Paper

PRS Poverty Reduction Strategy

SWAp sector-wide approach

OWNERSHIP

OWNERSHIP IS THE FIRST of the fi ve thematic head-
ings of the Paris Declaration – the apex of a
conceptual pyramid whose other building blocks
are aid alignment, aid harmonisation, managing
for results and mutual accountability (Figure 1.1).
It comes fi rst because experience shows that aid
is most eff ective when it supports countries’ own
development eff orts and policies to which leaders,
offi cials and citizens of the country are truly
committed. It is less eff ective where the policies
are donor-driven.

Th e eventual state of aff airs visualised by the
Paris Declaration is one in which partner coun-
tries exercise eff ective leadership over their devel-
opment policies and strategies, and co-ordinate
the eff orts of development actors working in
the national territory. Th is obviously has several
dimensions, some easily measured or assessed
and others not.

Th e degree to which governments take the lead in
co-ordinating aid-funded activities is the subject
of a specifi c commitment in the Paris Declaration.
It is also a variable that seems to be subject to
signifi cant changes over time. Another dimen-
sion of ownership is the degree to which coun-
tries have development strategies that are clear
and well operationalised, so that development
eff orts are eff ective and there is a robust basis for
the alignment of aid with country policies. Th is
is the focus of Indicator 1 of the baseline survey.
Th e following discussion focuses mainly on these
two linked dimensions of ownership.

LEADERSHIP IN AID CO-ORDINATION

Both the survey returns and the World Bank
sources (CDF Progress Report and AER) suggest
that governments are more inclined and capable
than they were only a few years ago to assume a
leadership role in aid co-ordination. Th is seems to
be true both in aid-dependent countries such as
Mali or Tanzania and in those that rely to a more
limited extent on external grant funding, such as
Peru. Th e trend initiated during the piloting of
the CDF – that of holding Consultative Group
meetings in country, with the national govern-
ment co-chairing the event – has become more
common. Increasingly, government offi cials also
convene sector working groups or round tables
and thereby exercise a leadership role in rela-
tion to donors in sectors, although this remains
much more common in the health and educa-
tion sectors than in fi elds where policies and aid
eff orts are more diff use. Little by little, govern-
ments are prepared to set frameworks with which
they expect donors to comply.

Indicators of the willingness and ability of coun-
tries to assert themselves in this way are included
in other sections of the survey that consider the
degree to which aid data are captured in the
national budget or the government accounts
(Indicators 3 and 7) and the use of programme-
based approaches, for which a country-led policy
framework is a necessary condition (Indicator 9).
Th e discussion of these variables points to the
degree to which country leadership is in part a

FIGURE 1.1:
The Aid
Effectiveness
Pyramid

16 THE 2006 SURVEY ON MONITORING THE PARIS DECLARATION, VOLUME 1: PRE-PUBLICATION COPY - © OECD 2007

OWNERSHIP
Partner countries

ALIGNMENT
Donors-partners

HARMONISATION
Donors-donors

MANAGING FOR RESULTS

PARTNERS
SET THE AGENDA

ALIGNING
WITH PARTNERS’

AGENDA

USING
PARTNERS’
SYSTEMS

SHARING
INFORMATION

SIMPLIFYING
PROCEDURES

ESTABLISHING
COMMON

ARRANGEMENTS

1

2

3

5

4

M
U

TU
A

L A
C

C
O

U
N

TA
B

ILITY

17THE 2006 SURVEY ON MONITORING THE PARIS DECLARATION, VOLUME 1: PRE-PUBLICATION COPY - © OECD 2007

function of donors’ willingness to give up some
of their independence of action. However, donors
cannot by themselves create this (or any other)
aspect of country ownership. The interest of the
country’s senior officials and, most importantly,
its politicians in setting the agenda for devel-
opment efforts is a variable that is not directly
captured by any of the Paris Declaration indica-
tors but is crucial to the whole venture.

OPERATIONAL DEVELOPMENT STRATEGY

The 2005 CDF Progress Report upon which the
survey report draws for Indicator 1 specifies that
an operational strategy calls for: a coherent long-
term vision and a medium-term strategy derived
from it; specific targets serving a holistic, balanced
and well-sequenced development strategy; and
capacity and resources for its implementation.
While the necessary components of a robust
development policy framework can be described
in a number of different ways, this approach from
the World Bank is largely commonsensical and
ought to be widely accepted. It gives significant
weight to the provision of resources for implemen-
tation, as well as to prioritisation and sequencing
and the derivation of medium-term objectives
from a long-term vision. The scoring reflects the
weighting of these different concerns.

Chart 1.1 shows the distribution for Indicator
1 of the countries that were both included in
the baseline survey and covered by the CDF
Progress Report. The CDF exercise only covered
International Development Association (IDA)-
eligible countries and LICUS (Low-Income
Countries Under Stress) that prepared a Poverty

Reduction Strategy Paper (PRSP) or Interim
PRSP. Appendix Table A.1 provides more detail.
The country profiles on which the scores are
based have been updated for the purposes of the
forthcoming AER, and some of the placing may
change as a result. However, in 2005 the posi-
tion was that only five countries, or 17% of the
total, had national development strategies consid-
ered “largely developed towards achieving good
practice” (category B). There were no cases of
strategies considered to “substantially achieve
good practice” (category A). From the country
descriptions, it is clear that this reflects some
rather general patterns of strength and weakness
in country policy frameworks.

In very many cases – including virtually all IDA
countries that have prepared PRSPs – govern-
ments have produced documents that begin with
a vision and derive from it a medium-term policy
framework consisting of broad fields or “pillars”
of development effort. Increasingly, the strategies
are comprehensive and reasonably well balanced,
but they tend to fall down on prioritisation and
sequencing. The latter are the key features needed
for a realistic implementation plan, given human
and material resource constraints. Some countries
do have costed targets and operational priorities.
However, it is not always the case that these are
well specified in terms of government activities,
or that there are mechanisms ensuring that prior-
itised activities actually get the required resources
and implementation capacities. This depends
on the linkage of the strategy to a fiscal policy
and budget process that raise new resources,
reallocate existing resources and stimulate the

CHART 1.1:
Do countries
have operational
development
strategies?

INDICATOR 1

A

B

E
NDS reflects little action toward
achieving good practice.

NDS incorporates some elements
of good practice.

NDS reflects action taken towards
achieving good practice.

NDS is largely developed towards
achieving good practice.17%5

0

0

18

6 21%

62% C

D

SCORENO. OF COUNTRIES CATEGORIES

National Development Strategy (NDS)
substantially achieves good practice

Source: World Bank CDF Progress Report for 2005.

LOWER SCORE ▼

HIGHER SCORE ▲

Operational value of national development strategies

18 THE 2006 SURVEY ON MONITORING THE PARIS DECLARATION, VOLUME 1: PRE-PUBLICATION COPY - © OECD 2007

THE PARIS DECLARATION visualises a situation in
which donors base their support fully on partner
countries’ development strategies, institutions
and procedures. Experience suggests that aid that
is well aligned – with country-owned policies on
the one hand and with country systems on the
other – makes a bigger contribution to develop-
ment than aid that is donor-driven and frag-
mented. As with ownership, alignment has several
dimensions, and measuring it is challenging.
Today bilateral and multilateral donors with few
exceptions base their support in at least a general
way on established country policy frameworks,
be these Poverty Reduction Strategies (PRSs),
national plans or sector strategies. However,
donor choices are only significantly restricted if
the strategies are quite well prioritised and trans-
lated into definite activities, which is the excep-
tion rather than the rule. For this reason, the
monitoring plan of the Paris Declaration aims to
set some measures of progress on alignment that
demand a greater effort on the part of donors and
imply more than a formal acknowledgement of
country policies.

The commitments on alignment are mutual, and
some of them call for joint action by donors and
governments. It is recognised that policy align-
ment needs to develop alongside the operation-
alisation of the country’s development vision, so
that the two processes reinforce each other. At the
same time, alignment of aid with country systems
entails both efforts by governments to improve the
reliability of those systems and efforts by donors
to remove barriers to their utilisation that origi-
nate in agency rules or operational habits. For this
reason, the Paris Declaration demands a stepping
up of the pace of change on several aspects of aid
alignment, but the specific targets to be achieved
by 2010 take into account the baseline situation
regarding the reliability of country systems.

BUILDING RELIABLE COUNTRY SYSTEMS

Indicator 2 will eventually cover two aspects of
country systems, public financial management
(Indicator 2a) and procurement (Indicator 2b).
In each case, the focus will be on the degree to
which existing systems adhere to broadly accepted
good practices or there is a reform programme

efficient use of resources for priority purposes.
Operationalisation in these senses is often partic-
ularly weak at local tiers of government.

The agreed target for this indicator is to raise
the proportion of partner countries with largely
or substantially developed operational strategies
(category A or B) to 75%. The baseline numbers
suggest that this is a difficult challenge, but not
a completely unrealistic one. Encouragingly,
three out of the five countries rated B in 2005 are
countries that have been in the vanguard of the
PRSP initiative, having submitted three annual
progress reports and a second-generation PRSP
by the end of 2006.

The main factor that would enable more coun-
tries to move a step up from their 2005 ratings
is a commitment by governments to using their
central resource allocation instrument, the
national budget, in a more vigorous and consis-
tent way to support agreed policy priorities.
Technical improvements in budget preparation
and execution will help in this respect. In partic-
ular, the use of medium-term expenditure frame-
works (MTEFs) is commonly seen as a key to
improving plan-budget linkage. However, this is
only the case when the MTEF is used as a policy
instrument, steering annual budgets in desired
directions, and it can only work if the policy
priorities are agreed. The benefits from technical
improvements in this area are much affected by
the position that development occupies among
the country leadership’s priorities. In practice,
strategies tend to be more operational the closer
they are to the driving political concerns of the
president, cabinet or governing party.

In summary, the Paris Declaration commitment
to strengthening country ownership of develop-
ment efforts poses a substantial challenge. The
baseline situation, with only 17% of surveyed
countries having operational development strate-
gies meeting the agreed quality threshold, leaves
a great deal to be done if the target proportion
of 75% is to be met by 2010. What needs to be
done is clear enough. It is also clear that, because
the challenges are as much political as technical,
the primary impetus for change must come from
within countries.

ALIGNMENT

19THE 2006 SURVEY ON MONITORING THE PARIS DECLARATION, VOLUME 1: PRE-PUBLICATION COPY - © OECD 2007

in place to promote improved practices. The
survey returns and CDF Progress Report and
AER country profiles contain some informa-
tion on the improvement of country procure-
ment systems, and this is reflected in the country
chapters. However, no systematic and quantified
assessments of procurement-system quality are
available at this point. Therefore, this overview
focuses on public financial management (PFM).

The assessment of PFM systems is based on a
component of the World Bank’s CPIA. CPIA
Indicator 13 is a measure of the quality of a coun-
try’s budget and financial management system.
It assesses the degree to which the country’s
arrangements include:
■ A comprehensive and credible budget

linked to policy priorities.
■ Effective financial management systems for

budget expenditure and budget revenues.
■ Timely and accurate fiscal reporting.
■ Clear and balanced assignment of

expenditures and revenues to each level
of government.

Previously not in the public domain, CPIA ratings
are now published by the Bank for IDA-eligible
countries only. Appendix Table A.2 shows the
2005 scores for the surveyed countries covered
by the published assessments. Chart 1.2 shows
the distribution across the scores. The bulk of
the countries covered are within or between the

1.0 Very weak (PFM systems)

1.5

2.0 Weak

5.0 Strong

3.5

28%

3%

0%

0%

0%

38%

17% 3.0 Moderately weak

2.5

4.0 Moderately strong

5.5

6.0 Very strong

4.5

10%

3%

0%

0%

1

0

0

0

8

11

5

3

1

0

0

SCORENO. OF COUNTRIES CATEGORY

Source: World Bank Country Policy and Institutional Assessment (CPIA), 2005.

LOWER SCORE ▲

HIGHER SCORE ▼

Quality of country public financial management systems

Moderately weak and Moderately strong catego-
ries. Some 31% of the countries have systems
considered at least moderately strong. The Paris
Declaration target is that half of partner coun-
tries move up at least half a point by 2010.

The published CPIA information does not break
the scores down by component. However, from
the survey returns and CDF Progress Report and
AER profiles, it is possible to identify for each
country the areas in which PFM reform efforts
are in place and are having effects, as well as the
areas of continuing weakness. The general picture
is that joint exercises such as Country Financial
Accountability Assessments (CFAAs) and
increasingly Public Expenditure and Financial
Accountability (PEFA) assessments are now
being widely undertaken, leading to the adop-
tion by governments of action plans for PFM
improvement. The countries scoring at the higher
end of the central range from 2.0 to 4.5 are those
that have made most headway in ensuring the
national budget is a credible estimate of actual
revenues and expenditures, and in providing
a mechanism linking budget formulation and
execution to formally agreed policy priorities.
Those countries usually have technical systems
that assist in the management and reporting
of financial flows between sector and tiers of
government. However, the main requirement is
not technical. A significant political will seems
to lie behind most successful efforts to improve

CHART 1.2:
How reliable
are country
public financial
systems?

INDICATOR 2A

20 THE 2006 SURVEY ON MONITORING THE PARIS DECLARATION, VOLUME 1: PRE-PUBLICATION COPY - © OECD 2007

PFM systems. In a number of survey countries,
the key barrier to an improved score seems to lie
in financial management at sub-national levels.

To summarise, it is again relatively clear what is
needed for countries to achieve progress in this
area: decisive leadership. On experience, moving
a country half a point up the CPIA PFM scale is
an attainable objective so long as political leaders
recognise the importance of credible public
finances to their countries’ future and transmit
this message to their officials. The biggest chal-
lenge will be ensuring that these steps are taken
by at least half of the partner countries included
in the surveys.

ALIGNING AID FLOWS

WITH NATIONAL PRIORITIES

Indicator 3 assumes that donors orient their aid in
a broad way to the objectives set out in a PRS or
national or sectoral development plan, and goes
on to ask the more specific and telling question of
how well external financial flows are reflected in
the national budget. As noted above, there is some
doubt about the ability of national development
strategy to govern the actual allocation and use
of resources in a good many partner countries.
There may also be questions about the credibility
of the budget. Nevertheless, the formulation of
the budget is a central feature of the formal policy
process in all countries. So the degree to which

donor financial contributions to the government
sector are fully and accurately reflected in the
budget provides a relevant indicator of the degree
to which there is a serious effort to connect aid
with country policies and policy processes. The
indicator is the percentage of aid disbursements
to the government sector reported by donors that
is included in the budget estimates for the same
financial year. The target is to halve the propor-
tion of aid disbursements to the government
sector not included in the budget by 2010, with
at least 85% reported on budget.

Appendix Tables A.3 and B.3 show the raw
numbers for aid disbursements and aid included
in the budget, by country and by donor. These
show that there are discrepancies in both direc-
tions, as budgets both under-include aid flows
and over-include them. In other words, budgets
are unrealistic in two opposite directions. This
needs to be taken into account in assessing the
baseline position against which future progress
will be measured.

If one focuses on the size of the discrepancies
(leaving aside their direction), one gets a measure
of the overall extent to which budgets are realistic
in respect of disbursed aid. This measure (the
baseline ratio) is 88% when calculated on a
weighted average basis (Table A.3). Chart 1.3
shows the individual averages for countries and
donors respectively.

0%

20%

40%

60%

80%

100%

0%

20%

40%

60%

80%

100%

Percentage of aid recorded in countries’ national budgets

Country-by-country Donor-by donor

In 6 countries,
less than 50% of aid is
recorded in the budget.

In the case of 6 donors,
less than 50% of their aid is
recorded in national budgets.

2010 target:
85% of aid recorded on budget

2010 target:
85% of aid recorded on budget

34 COUNTRIES LARGEST 22 DONORS

CHART 1.3:
Do national
budgets record
aid realistically?

INDICATOR 3

21THE 2006 SURVEY ON MONITORING THE PARIS DECLARATION, VOLUME 1: PRE-PUBLICATION COPY - © OECD 2007

The chart shows that the discrepancy between
the budget numbers and actual disbursements is
considerable in many of the surveyed countries,
with half of them showing a measure of budget
realism of 70% or less. For a fair number of
donors, performance is substantially below this
level (Indeed the performance, unweighted by
the volume of aid, is 45%). The agreed target of
reducing the budget inclusion gap by half means
raising the aggregate baseline ratio to 94% by
2010, with 85% as a minimum acceptable level.
This will call for substantial efforts in several
countries and by a number of agencies.

Like several of the alignment indicators, Indicator
3 is influenced by both country and donor efforts.
It also captures the effects of more than one type
of factor. According to the survey returns, in-
country donors and government officials see the
shortcomings of current practice as the compound
effect of three rather general problems:
■ Donors are not always attentive to getting

information on intended disbursements to
the budget authorities in good time and in a
usable form, resulting in systematic under-
inclusion of aid in the budget.

■ When donors do provide such information,
they are not always realistic about their
ability to disburse on schedule, resulting in a
tendency to over-estimate some kinds of aid
flows and under-estimate others.

■ Budget authorities are often not strongly
motivated or equipped to capture information
on donor disbursement intentions, or to make
realistic estimates of shortfalls, resulting in
both over- and under-counting on quite a
large scale.

In Box 1.1, there are two contrasting country
illustrations.

As this suggests, the discrepancies in the budget
numbers include a significant element due to non-
disbursement of scheduled funds or to unsched-
uled disbursements – in other words the problem
of low aid predictability, as opposed to weak
information capture. Predictability is consid-
ered directly in the discussion of Indicator 7

below. Using the numbers presented there, it is
possible to factor out the predictability problem
and focus on the degree of information capture.
This analysis, with results presented in Chart 1.4,
reveals the extent to which lack of budget realism
arises from the combination of poor reporting of
disbursement intentions by donors and limited
information capture by budget authorities.

The chart shows that budgets capture nearly all
aid for only two countries (9% of the sample)
of the donors scored. Scheduled donor disburse-
ments to the government sector are poorly
reported and/or insufficiently captured to a
greater or lesser degree in the case of 78% of the
donors. Over-estimation of scheduled disburse-
ments in the budget is restricted to smaller
donors, one regional development bank and one
large bilateral.

BOX 1.1 INCLUDING AID IN THE BUDGET:
TWO COUNTRY EXAMPLES

“ Even though there is substantial variation…, Indicator 3
shows great achievements in terms of integrating aid
flows in the national budget… Improvements in the
quality and timeliness of development partners’ projec-
tion reports to the Ministry of Finance, a better system for
gathering and maintaining this data within the Ministry
of Finance, as well as improved dialogue between the
Ministry of Finance, other ministries, departments,
agencies, regions and local government authorities…
have contributed to this progress. Nevertheless, difficul-
ties remain for the government in obtaining complete
and reliable projections from development partners
in particular for the second and third year of the MTEF
period… Challenges in closing the remaining gap
between actual disbursements and aid commitments
in the annual national budget relate to problems of aid
predictability associated in particular with the project
funding modality.” (Tanzania)

“ The discrepancies that are noted are more or less
substantial, but in general very significant. It is sometimes
impossible to explain them given the low level of detail
in the government document. Certain donors explain
the discrepancy by the fact that the table provided by
the government only presents the Special Investment
Budget, which does not take account of certain finan-
cial contributions. Another reason is probably commu-
nication problems, between the different parts of the
administration as well as between the government and
donors.” (Mali)

22 THE 2006 SURVEY ON MONITORING THE PARIS DECLARATION, VOLUME 1: PRE-PUBLICATION COPY - © OECD 2007

The differences across donors may reflect not so
much failure of notification as the different degrees
to which donors are engaged in countries where
information capture by the budget authorities is
weak. The detailed figures nonetheless reveal quite
striking differences among agencies that are involved
in similar numbers and types of countries.

To summarise, meeting the Paris objective will
call for a concerted assault on several distinct
problems which together compromise the ability
of the budget to serve as a full and accurate record
of donor financial contributions. They include
failures of both information supply and infor-
mation capture in regard to donor disbursement
intentions. They also include lack of realism, on
the part of the donors, the budget authorities or
both, about the prospects of funds being disbursed
on schedule. The precise focus of the necessary
efforts will vary somewhat among countries and
among modalities of aid delivery. Almost every-
where, however, action will be needed from both
donors and country authorities.

0%

50%

100%

150%

125%

75%

25%

14 donors
Aid in budget is less than scheduled

5 donors
Aid in budget is “about” what was scheduled

3 donors
Aid in budget is more than what was scheduled

Aid recorded in budgets as a percentage of donors’ scheduled disbursements

LARGEST 22 DONORS

Definition of accurate reporting

This chart tells us how many donors have
their scheduled aid accurately reported in
budgets.

In practice, aid recorded in budgets is
usually less or much less than what donors
expect to disburse. In this case (ratio less
than 80%), scheduled aid is said to be
under-reported.
Less frequently, aid recorded in budgets is
more or much more than what donors
expect to disburse (ratio more than 120%).
In this case, scheduled aid is over-reported.
Aid is reported accurately when aid
recorded in budget matches aid scheduled
for disbursement by donors (ratio between
80% and 120%).

CHART 1.4:
Do budgets
report donor
disbursement
plans realistically?

CO-ORDINATING SUPPORT

TO STRENGTHEN CAPACITY

In the Paris Declaration, capacity improvements
are seen critical to improving development results
as well as to achieving objectives of ownership,
aid alignment and mutual accountability. It is
increasingly recognised that capacity develop-
ment involves changes in institutional rules and
organisational systems, not just training and the
transfer of expertise. Likewise, successful capacity
development is seen as necessarily endogenous
– led by country actors with clear objectives,
making effective use of existing capacities and
harmonising external support within this frame-
work. Indicator 4 focuses on the extent to which
donor technical co-operation – as one input into
capacity development – is moving towards this
country-led model. Donor survey respondents
were asked to identify technical co-operation
support meeting all of four criteria:
■ The programmes support partners’ national

development strategies.
■ Partner countries exercise effective leadership

over the capacity development programme
supported by donors, implying clearly
communicated objectives from senior officials.

■ Donors integrate their support within
country-led programmes to strengthen
capacity development.

23THE 2006 SURVEY ON MONITORING THE PARIS DECLARATION, VOLUME 1: PRE-PUBLICATION COPY - © OECD 2007

The aggregate baseline figure of 47% for this
indicator (Appendix Tables A.4 and B.4) might
suggest that little needs to be done to achieve
the target. However, that would not be a valid
conclusion. As Chart 1.5 reveals, the dispersion
of the results both by country and by donor is far
greater than can plausibly be explained by real
differences in co-ordination levels. Conceptual

■ Where more than one donor is involved,
arrangements for co-ordinating donor
contributions are in place (for example,
pooling of resources).

The agreed target is to have 50% of technical co-
operation in co-ordinated programmes, defined
in this way, by 2010.

As the country chapters explain, several National
Co-ordinators took the view that in 2005 there
were no technical co-operation programmes in
the country meeting all four criteria, making
the true baseline zero percent. The element most
often missing was effective country leadership
based on a specific strategic vision for capacity
development. Donor respondents, on the other
hand, took the view (sometimes individually and
sometimes jointly and with government assent)
that the definition in the survey guidance was too
stringent. Very different views emerged on what
should count as co-ordination and as government
leadership. The numbers generated by the survey
therefore include technical co-operation efforts
that are co-ordinated in a relatively loose sense,
with the accent on consistency with the relevant
policy framework (e.g. a sector strategy or PFM
action plan) rather than on the presence of a
specific country initiative for capacity develop-
ment. Box 1.2 illustrates.

0%

25%

50%

100%

75%

2010 target:
50% of co-ordinated TA

In 12 countries, less than
25% of TA is co-ordinated.

0%

25%

50%

100%

75%

2010 target:
50% of co-ordinated TA

In the case of 6 donors,
less than 25% of their TA
is co-ordinated.

Co-ordinated technical assistance as a percentage of total TA

Country-by-country Donor-by-donor

34 COUNTRIES LARGEST 22 DONORS

CHART 1.5:
How much
technical
assistance is
co-ordinated
through country
programmes?

INDICATOR 4

BOX 1.2 WHAT IS CO-ORDINATED
CAPACITY DEVELOPMENT?
THREE COUNTRY APPROACHES

“ The government will dedicate itself in the framework
of sectoral dialogue with donors to… define the
co-ordinated programmes of capacity strengthening
needed to improve the impact of the sector programmes.”
(Burkina Faso)

“ All the development partners complained about
the difficulties of interpretation of the “co-ordinated
technical assistance”. Most of the donors considered
technical assistance agreed with the government
as co-ordinated… Some donors considered the bulk
or all of their assistance as co-ordinated.” (Ghana)

“ In the Nicaraguan case, it has been considered that
technical co-operation… that can be included as
co-ordinated is that which is supporting priorities
explicitly defined by the government, sector approaches
or other programme-based approaches. This co-ordination
will have been formalised in a document (for example, a
bilateral agreement, a memorandum of understanding
or resolution of a sector working group).” (Nicaragua)

24 THE 2006 SURVEY ON MONITORING THE PARIS DECLARATION, VOLUME 1: PRE-PUBLICATION COPY - © OECD 2007

differences leading to divergences in reporting
are a more likely explanation.

This is unfortunate from the point of view of
establishing a robust baseline. The survey results
on Indicator 4 provide a benchmark against which
future progress can be measured only if each
agency and country applies the same standards
of assessment year by year. This will be difficult
to ensure. On the other hand, the results may be
seen as providing a welcome wake-up call to the
international community. The Paris Declaration’s
vision of the future needs of capacity develop-
ment is clearly not just unfinished business, but
business that has barely begun.

In summary, the baseline position regarding the
Paris Declaration commitments on capacity devel-
opment is not as favourable as the figure of 47%
(against a 2010 target of 50%) might appear to
suggest (Indeed the unweighted average is only
9%). The survey has revealed quite a profound lack
of consensus on valid approaches to capacity devel-
opment and the meaning of the Paris commit-
ments in this area. This should prompt further
efforts to disseminate and evaluate the evidence
that lies behind the Paris vision. Donors and
partners should be reviewing, in the light of this
evidence, whether the expectations they currently
have of each other and of themselves on this issue
are sufficiently far-sighted and ambitious.

USING COUNTRY SYSTEMS

Indicator 5 provides a relatively unambiguous
and telling measure of the degree to which
systems alignment is taking place. Together
with the CPIA rating on PFM quality and the
anticipated assessment of procurement systems, it
provides a set of realistic targets towards which
donors and country authorities may be expected
to work together.

This said, the indicator is not entirely free of
measurement problems. After thorough review
of the data submitted, some concerns remain
about the way the survey questions on Indicator
5 were interpreted in some countries and by some
donors. For example, for several countries it is not
clear that all of the programmes included as using
country budget execution systems are accurately
described as “subject to normal country budgetary
execution procedures, namely procedures for
authorisation, approval and payment” (as specified
in the survey’s Definitions and Guidance paper).
There are also some grey areas affecting the inter-
pretation of the indicator. For example, authori-
sation procedures that are formally those laid
down in the country’s laws and public-sector
financial regulations can be subject to a large
degree of de jure or de facto delegation, meaning
that the “use of country systems” is somewhat
nominal. Together, these factors may mean that
the numbers overstate somewhat the effective use
of country systems.

31%
Auditing

31%
Financial
reporting

38%
Budget
execution

� Use of PFM systems

� Use of procurement systems

0%

20%

20%

40%

40%

60%

80%

60%

80%

Use of country systems as a percentage of aid for government
Country-by-country

Use of country PFM systems
broken down by component

34 COUNTRIES

CHART 1.6:
How much aid for
the government
sectors uses
country systems?

INDICATOR 5

25THE 2006 SURVEY ON MONITORING THE PARIS DECLARATION, VOLUME 1: PRE-PUBLICATION COPY - © OECD 2007

In Chart 1.6 (based on Appendix Table A.5)
the bars on the left represent the range of results
across countries, for the average utilisation of
three components of country PFM systems
(budget execution, financial reporting and audit)
and for country procurement systems. This shows
considerable variation in the use of country
systems across countries. In aggregate, the utili-
sation of country procurement systems is compa-
rable with the average use of PFM systems (39%
in both cases). However, there is little correspon-
dence between the two rates country-by-country.
The pie chart on the right shows how the total
measured use of the three PFM components. It
suggests that slightly more use is made of coun-
tries budget execution systems.

There is no agreed target for increasing use of
country procurement systems, as the Indicator
5 targets are layered according to a measure of
systems quality, and there is as yet no scoring
arrangement for procurement.1 For coun-
tries that have a current CPIA rating for PFM
systems between 3.5 and 4.5, the target for 2010
is to reduce by one-third the non-utilisation of
country systems, using the average rates across
the three components. If in the future country

PFM quality remains largely in the range 3.5-4.5,
aggregate utilisation will need to rise to 59%. It
is clear that compared to the unweighted average
of 33%, this will be a substantial challenge for
many donors.

This becomes apparent when the use of country
PFM systems is plotted against the agreed measure
of the quality of those systems (Indicator 2a), as
in Chart 1.7. The chart tells us two important
things. First, there may be a relationship of some
kind between the quality of a country’s PFM
systems and the use that is being made of them
by donors as envisaged by the Paris Declaration.
That is, the stronger the country systems, the
more likely donors are to use them. However, if
so, the correlation is very weak. There is a very
wide spread in use of country systems within the
group of countries that have the same quality of
PFM systems. Take for example, the eight coun-
tries that have moderately strong PFM systems
(score=4.0). Use of country systems ranges from
14% (minimum value) to 56% (maximum value).
This suggests that factors other than quality of
systems are affecting donors’ willingness to use
them. The breakdown by donor in Appendix
Table B.5 is consistent with this conclusion.

Q
U

A
LI

TY
O

F
C

O
U

N
TR

Y
P

FM
S

YS
TE

M
S

Quality of country PFM systems vs. use of country PFM systems
Country-by-country

0% 10% 20% 30% 40% 50% 60%

Findings
1. There is a correlation between quality of PFM systems and use
of PFM systems by donors. But correlation is very weak (R2=21%).

USE OF COUNTRY PFM SYSTEMS

2.0

2.5

3.5

3.0

4.0

5.5

5.0

Each dot represents a country

LOWER QUALITY

HIGHER QUALITY

2. There is a very broad range of use of country systems within
the group of countries that share the same score for quality
of PFM systems.

CHART 1.7:
What drives use
of country PFM
systems?

26 THE 2006 SURVEY ON MONITORING THE PARIS DECLARATION, VOLUME 1: PRE-PUBLICATION COPY - © OECD 2007

Although the survey data do not allow us to
confirm this, it seems likely that the use of
country PFM systems has increased in recent
years as a result of the movement towards general
or sector budget support. In a number of coun-
tries covered by the survey, budget support or a
few large programmes account for almost all of
the use of country systems. There are different
ways of viewing this. It could be interpreted as
a sign of movement towards Paris Declaration
objectives. However, it could be taken to indi-
cate that there is little commitment to the use of
country systems among donors other than those
providing budget support.

To summarise, the degree to which donors make
use of country PFM systems is highly variable, and
little of the variation is explained by differences
in the measured quality of the systems. If, as
2010 approaches, some donors continue to make
little use of country systems even when these are
rated by the CPIA as comparatively strong, the
overall performance targets on this indicator will
be hard to meet, and donor seriousness about one
of the fundamental Paris Declaration commit-
ments will be open to question.

AVOIDING PARALLEL

IMPLEMENTATION STRUCTURES

The Paris Declaration invites donors to “avoid,
to the maximum extent possible, creating dedi-
cated structures for day-to-day management and
implementation of aid-financed projects and
programmes”. Indicator 6 is a count of parallel
project implementation units (PIUs), where
“parallel” refers to having been created outside
existing country institutional structures. The
survey guidance makes a distinction between
PIUs and executing agencies and gives three
typical features of parallel PIUs:
■ They are accountable to the external funding

agencies rather than to the country imple-
menting agencies (ministries, departments,
agencies, etc.).

■ Most of the professional staff are appointed
by the donor.

■ The salary of PIU personnel often exceeds
that of civil-service personnel.

Interpretation of the survey question on this
subject was controversial in a number of coun-
tries. It is not clear that within countries all
donors applied the same criteria with the same
degree of rigour, or that across countries the same
standards were used. In a number of cases, the
descriptive part of the survey return indicates
that some donors applied a legalistic criterion of
accountability to the formal executing agency,
whereas the National Co-ordinator or other
donors would have preferred a greater recognition
of the substantive reality of accountability to the

BOX 1.3 PARALLEL PIUs:
NATIONAL CO-ORDINATORS’ VIEWS

“ The donors have encountered problems with the defini-
tion provided… 7 donors out of 14 reported a total of 24
parallel PIUs… This information, however, may not reflect
the full extent of the parallel structures used... (T)here is
no system for recording them.” (Ghana)

“ The PIU modality that is most common... is that of pro-
jects and programmes under the responsibility of two co-
directors, one national and one nominated by the donor.
While it could be argued that this is a case of a semi-
integrated PIU… it is thought that it is most common
that the management… falls almost exclusively to the
team that reports to the donor.” (Peru)

“ It has been found difficult to agree on the definition of a
“parallel unit”, which has led to very diverse and varied
responses, and changes to responses… In future, it is
recommended that the survey follows either the interna-
tional definitions of the OECD-DAC or definitions that are
appropriate to our national context.” (Senegal)

“ Some partners are making an effort to move away from
the use of parallel PIUs. For example, the World Bank-
financed projects on public financial management and
rural education… integrate some of their functions
in the respective ministries. However, these units still
cannot be considered integrated because they are not
component parts of an existing institutional structure;
salary amounts exceed those of public employees and
the donor is ultimately the one who takes the final deci-
sion….” (Moldova)

“ The principal problem in reducing PIUs is the need to
reduce at the same time the use of procedures that are
not those of the country. If a PIU is eliminated but the
co-operation agency continues to require use of its
procedures, the government departments will be over-
loaded with work….” (Nicaragua)

27THE 2006 SURVEY ON MONITORING THE PARIS DECLARATION, VOLUME 1: PRE-PUBLICATION COPY - © OECD 2007

donor. It seems possible that some respondents
allowed themselves to confuse the definitional
question (is the unit “parallel”?) with the aid-
management question (is the parallelism justi-
fied in terms of the developmental benefits and
costs?). Box 1.3 provides some illustrations.

The effect is that parallel PIUs have probably
been under-counted. Nevertheless, as with other
alignment measures, the survey has helped to
spark a local debate on the subject, and it is to
be expected that in the future it will be more
common to hear conscious and explicit consid-
eration of the trade-offs between project effec-
tiveness and strengthening mainstream project
management capacities, when projects and
programmes are being designed.

The overall target is to reduce the baseline stock of
1 828 parallel PIUs by two-thirds, to only 607, by
2010. As Appendix Table A.6 and the left-hand
side of Chart 1.8 reveal, there is considerable varia-
tion in the current count across countries, no doubt
reflecting differences in criteria as well as in the scale
of aid disbursements. Each country is expected to
contribute proportionally to the targeted reduc-
tion, whatever the basis for the baseline number.
Appendix Table B.6 and the right-hand side of
the chart confirm that reported parallel PIUs per
country vary a good deal by donor.

0

4

2

8

6

10

12

14

0

25

50

75

100

125

150

4 countries have more than 100 parallel PIUs

� Multilateral donor
� Bilateral donor

Number of parallel PIUs
Country-by-country

Average number of parallel PIUs per country
Donor-by-donor

34 COUNTRIES LARGEST 22 DONORS

CHART 1.8:
How many
PIUs are parallel
to country
structures?

INDICATOR 6

Closer analysis shows that in a few cases, these
numbers appear to be in reasonable proportion
to the aid disbursed per country. However, this is
not always the case, with a small number of large
agencies reporting a disproportionate number of
parallel PIUs. These organisations may be expected
to re-examine the reasons for their continued use
of parallel systems. Donor organisations that
have, for one reason or another, reported only a
small number of projects as currently managed
by parallel units face a different sort of chal-
lenge. They will be expected to show a two-thirds
reduction relative to their reported current posi-
tion as their contribution to implementing the
Paris Declaration commitment.

The evidence brought together by the survey
suggests that there is a will to change. A number
of large agencies, including the World Bank and
the European Commission, are frequently cited
as planning to merge existing PIUs into the
structures of ministries or agencies, with exam-
ples given. However, change will not be easy.
The principal hurdles cited in the survey returns
include the backlog of projects that were originally
set up with very little concern for alignment and
ownership goals, the interest of PIU employees
and parent ministries in the employment condi-
tions and fringe benefits that parallel units
permit, and the unwillingness of results-conscious
external funding agencies to compromise on the
quality of project implementation.

28 THE 2006 SURVEY ON MONITORING THE PARIS DECLARATION, VOLUME 1: PRE-PUBLICATION COPY - © OECD 2007

To summarise, the total of 1 828 parallel PIUs
in just 34 countries is of concern – particularly if
there are grounds for thinking this is a low esti-
mate. Even if the will to change is real, so that
new projects are increasingly managed by main-
stream government organisations, reducing this
total to 609 is going to call for a substantial effort
in which both donors and governments will need
to participate.

PROVIDING MORE PREDICTABLE AID

In the Paris Declaration, donors undertook to
provide reliable indicative pledges of aid over a
multi-year span, and disburse aid in a timely and
predictable fashion according to agreed sched-
ules. Aid predictability is generally recognised to
be an important factor in the ability of countries
to manage public finances and undertake real-
istic planning for development.

The survey data allow us to look at predict-
ability from two different angles. The first angle
is donors’ and government’s combined ability
to disburse aid on schedule. The second angle
is donors’ and government’s ability to record
comprehensively disbursements made by donors
for the government sector.

Consider first the issue of disbursement according
to schedule. Across all survey countries, donors
scheduled USD 18 984 million for disbursement
in 2005 and actually disbursed (according to their
own records) somewhat more, USD 19 406 million
(Appendix Tables A7 and B7). However, these
very aggregate figures mask the fact that there are
cases both of under-disbursement (funds not being
disbursed on schedule) and of over-disbursement
(the disbursement of unscheduled amounts). This
is true across both countries and donors.

Aid is unpredictable, in this sense, for both bilat-
erals and multilaterals. A few of the major donors
are substantial under- or over-disbursers across
the countries they support. It is unclear in these
cases whether the explanation has something to
do with the subset of surveyed countries where
those agencies are involved. A donor that gives
priority to countries where project execution is
frequently delayed may be expected to do poorly
on this measure. There may also be good grounds
for unplanned assistance at certain moments
in particular countries that account for a large
proportion of a given donor’s global spending.
If, however, these explanations do not apply, the
agencies listed on both ends of the spectrum will
want to consider how to improve their contri-
bution to the Paris Declaration objective on aid
predictability.

0%

20%

40%

60%

80%

100%

0%

20%

40%

60%

80%

100%
2010 target:
87% of aid disbursed on schedule

In 11 countries,
predictability by ratio is less than 50%.

In the case of 5 donors,
the predictability ratio
is less than 50%.

2010 target:
87% of aid disbursed on schedule

▲HIGHER PREDICTABILITY

▼LOWER PREDICTABILITY

▲HIGHER PREDICTABILITY

▼LOWER PREDICTABILITY

Aid scheduled for disbursement as a percentage of aid disbursed

Country-by-country Donor-by-donor

34 COUNTRIES LARGEST 22 DONORS

CHART 1.9:
Are disbursements
on schedule
and recorded by
government?

INDICATOR 7

29THE 2006 SURVEY ON MONITORING THE PARIS DECLARATION, VOLUME 1: PRE-PUBLICATION COPY - © OECD 2007

Consider now the government’s ability to record
disbursements. In aggregate terms, govern-
ments recorded in their accounting systems only
USD 13 851 million of the USD 19 406 million
actually disbursed according to donors. For almost
all countries and all larger donors, disbursements
were under-recorded, sometimes very substan-
tially (Appendix Tables A7 and B7). This may
reflect failure on the part of donors to notify
country authorities in the appropriate way about
their disbursements. It may equally reflect the
inability of government systems to capture and
process the information.

Indicator 7 on predictability has been designed
to encourage progress in both of these areas. In
other words, it seeks to improve not only the
predictability of actual disbursements but also the
accuracy of how they are recorded in government
accounts – an important feature of ownership,
accountability and transparency. It measures the
degree to which funds scheduled for disburse-
ment within a given year are recorded in govern-
ment accounts as disbursed that year, ignoring
the direction of any discrepancies. Calculations
of this indicator for countries and donors are
provided in the final column of Appendix Tables
A7 and B7 (the baseline ratio). Chart 1.9 shows
the results for countries and donors respectively.
It should be reminded that Indicator 7 captures

in-year rather than multi-annual predictability.
As a result it is a relatively volatile measure of
predictability that is likely to vary considerably
from year to year.

The weighted average for this indicator in 2005
was 73% (Table A7). The Paris Declaration
objective is gradually to close the gap between this
ratio and 100%, reducing it by half – so that the
ratio rises to at least 87% – by 2010.

Closing the predictability gap will require donors
and government to work increasingly together on
various fronts at the same time. Actions might
include efforts in improving:
■ The realism of predictions on volume and

timing of expected disbursements.
■ The way donors notify their disbursements to

government.
■ The comprehensiveness of government’s

records of disbursements made by donors.

The particular steps needed to improve predict-
ability of aid flows vary among aid modalities as
well as across countries. The survey returns include
some suggestions that budget support, especially
when provided within a multi-year framework
and disbursed early in the year, is friendlier to aid
predictability than project finance, where imple-
mentation delays can be unavoidable. On the other
hand, budget support faces predictability problems

0%

25%

50%

75%

100%

0%

25%

50%

75%

100%

In 5 countries,
less than 50%
of aid is untied.

Share of untied aid

Country-by-country Donor-by-donor

34 COUNTRIES 15 BILATERAL DONORS

CHART 1.10:
How much aid
is untied?

INDICATOR 8

30 THE 2006 SURVEY ON MONITORING THE PARIS DECLARATION, VOLUME 1: PRE-PUBLICATION COPY - © OECD 2007

of its own arising from performance-based vari-
able tranches, uncompleted prior actions and/or
breaches of fundamental principles. If the Paris
Declaration target is to be met, actions to reduce
the impact of these factors will need to be accom-
panied by greater collaboration between donors
and the officials responsible for country accounts,
so that the latter become progressively more able to
record aid flows accurately.

In summary, Indicator 7 is like Indicator 3 in that
it captures the joint effects of donor and govern-
ment practices. It is influenced by both the ability
to disburse funds on schedule, and the extent to
which aid can be accurately captured in country
accounting systems. The relevant target is not
out of reach, but the unweighted average of 43%
shows that considerable movement is needed in
many donor/partner relationships. As such, a
deliberate collaborative effort in each country
will be required to meet it.

UNTYING AID

The Paris Declaration reaffirms the 2001
DAC Recommendation on Untying Official
Development Assistance to Least Developed
Countries and commits the signatories to
making strides on this issue. The statistics on the
tying status of aid are held by the OECD and
are reproduced in the table A.8 (Appendix A).
It shows that in the 34 countries 75% of aid is
untied as compared to 78% in all the other coun-
tries included in the OECD records. Progress on
untying will require all bilateral donors to step up
their efforts by 2010.

IF DONORS AND GOVERNMENTS were to achieve a
complete alignment of aid flows with country-
owned policies and country systems, aid harmon-
isation would not be an issue. However, as the
baseline survey confirms, in the real world align-
ment is imperfect – for reasons that have to do
with donors and countries. In these circum-
stances, aid effectiveness is likely to be enhanced
if donors harmonise their actions and adopt as
much as possible simple and transparent common
procedures.

The baseline survey focuses on just two dimensions
of harmonisation: the use of common arrangements
within programme-based approaches (PBAs), and
conducting joint missions and sharing analysis.
Two other groups of commitments on harmonisa-
tion are included in the Paris Declaration. There
are commitments on increasing donor comple-
mentarity on the basis of comparative advantages,
with greater use of delegation; and strengthening
incentives for collaborative behaviour. The base-
line indicators are therefore proxies for a much
wider field of commitments on harmonisation.

USING COMMON ARRANGEMENTS

Indicator 9 measures the proportion of aid to
the government sector that is disbursed within a
programme-based approach. In the survey guid-
ance, PBAs are defined in such a way that this
proportion is an accurate indicator of the extent to
which common arrangements are being used. In
the words of the Definitions and Guidance paper,
PBAs share all four of the following features:
■ Leadership by the host country or

organisation.
■ A single comprehensive programme and

budget framework.
■ A formalised process for donor co-ordination

and harmonisation of donor procedures for
reporting, budgeting, financial management
and procurement.

■ Efforts to increase the use of local systems
for programme design and implementation,
financial management, monitoring and
evaluation.

HARMONISATION

31THE 2006 SURVEY ON MONITORING THE PARIS DECLARATION, VOLUME 1: PRE-PUBLICATION COPY - © OECD 2007

BOX 1.4: VIEWS ON COMMON ARRANGEMENTS
AND PBAs

“ Apart from the tiny amount of budget support in Senegal,
the donors use their own procedures for the major
part of the resources delivered through programme
approaches. The total for resources provided through
programmes ignores the fact that the donors are able
to continue using their procedures. In contrast, it is the
resources making use of common arrangements that tell
us about the level of effort being made to reduce trans-
action costs.” (Senegal)

“ Special mention should be made of a series of quite large
programmes financed with concessional loans… These
programmes have been included because they fulfill
the first three criteria for consideration as a PBA, namely:
government leadership, a single programme framework
and participation of several donors.” (Peru)

“ Apart from the Multi-Donor Budget Support, there was no
general agreement on what constituted the programme-
based approach, despite the criteria provided.… Some
donors have provided lists of the PBAs, which could not
be ascertained in line with the definition..., buttressing the
need for additional work to validate the operations of the
listed examples….” (Ghana)

These complications will need to be borne in
mind in assessing trends over the period to 2010.
As discussed in Chapter 2 on The Survey Process,
they illustrate the need for firmer guidance on the
circumstances in which it is permissible for survey
respondents to adapt the suggested definitions.

The proportion of direct budget support (DBS)
is not a Paris Declaration indicator but data were
collected on the subject, and the results are of
interest. DBS was also defined in quite restrictive

0%

30%

10%

40%

50%

60%

70%

20%

� Programme-based approaches
� Direct budget support

2010 target:
66% of aid provided as PBA

Not PBAs
57%

PBAs 43%
(of which direct
budget support
20%)

Proportion of aid provided as PBAs
Country-by-country

Proportion of aid provided
as PBAs

34 COUNTRIES

CHART 1.11:
How much aid
is programme
based?

INDICATOR 9

This set of four criteria is fairly restrictive and
include direct (general or sectoral) budget support,
sector-wide approach (SWAp) programmes and
other arrangements in which there are equivalent
efforts towards joint planning and harmonisation
of procedures.

The general Paris Declaration target for 2010
is to have 66% of aid for the government sector
using common arrangements through PBAs.
Appendix Tables A.9 and B.9 show the reported
volume and proportion of programme-based aid,
and also the breakdown between direct budget
support (defined as unearmarked funding) and
other PBAs. Chart 1.11 provides an overview of
the results by country and in aggregate.

It is clear from the survey returns that donors in
many countries had some difficulty in accepting
the suggested definition of a PBA, and usually
National Co-ordinators did not feel empow-
ered to impose a ruling. In a number of coun-
tries, a looser set of criteria was adopted on the
basis of some degree of consensus, while in some
others individual donors were permitted to follow
their own definition of a “programme”. In these
cases, at least, the true baseline numbers for use
of common procedures are lower than those
presented in the tables. Box 1.4 provides some
illustrations. This means more substantial prog-
ress is called for to achieve the target that the
baseline level of 43% might suggest.

32 THE 2006 SURVEY ON MONITORING THE PARIS DECLARATION, VOLUME 1: PRE-PUBLICATION COPY - © OECD 2007

terms in the survey guidance, limiting it to
resources managed according to the recipient’s
budgetary procedures and not subject to
earmarking (although, in the case of sector
budget support, subject to a policy dialogue that
focuses on a sector). It is not clear from the survey
returns the degree to which the guidance was
strictly followed in this respect.

Chart 1.12 provides an overview of the results
by donor. There are a few surprises in the data,
even after taking into account the possible effects
of country selectivity. A striking feature is how
many donors (no less than a quarter of the top
22) report that they are already channelling half
of their aid to the government sector through
PBAs. The descriptive survey returns suggest that
this finding reflects to an important degree the
approach that some donors took to the definition
of a PBA. Only to a more limited degree does it
reflect effective adoption of common procedures
with other donors. A sensible objective for these
agencies, as well as for those that report them-
selves as currently less involved in PBAs, would
be to concentrate on the substance of the Paris
Declaration commitment. That means moving
with deliberate speed towards the use of common
procedures, under government leadership and
where possible using government systems.

The descriptive part of the survey returns, as well
as the CDF Progress Report and AER profiles,
point to some general patterns in the adoption
of PBAs. There appears to be quite a strong trend
towards the adoption of SWAp-type arrange-
ments, not only for sectors but also for cross-
cutting institutional areas such as private-sector
development, and justice, law and order. The
sectors that move reasonably quickly towards a
full SWAp are invariably health and education.
However, joint programmes outside the social
sectors make slower and more difficult headway
towards common arrangements. Nevertheless,
where a PBA is adopted by a large group of donors
in a sector or institutional area, there is a positive
bandwagon effect. With a few exceptions, addi-
tional members are willingly and easily absorbed.

To summarise, the survey provides no grounds
for complacency with respect to the Paris
Declaration commitment on moving towards
the use of common arrangements and proce-
dures under country leadership. The use of the
programme-based approach as a proxy for joint
planning and increasingly harmonised working
may have confused the issue to some degree.
Despite the relatively encouraging numbers
generated by the survey around this indicator,
the pace of change in donor practices will need to
increase over the coming years if the Paris objec-
tive is to be achieved.

20%

40%

60%

80%

100%

0%

500

1,000

1,500

2,000

2,500

3,000

0

2010 target:
66% of aid provided as PBA

U
SD

(
M

IL
LI

O
N

S)
Volume of PBAs
Donor-by-donor

Proportion of aid provided as PBAs
Donor-by-donor

LARGEST 22 DONORS LARGEST 22 DONORS

� Budget support
� Other types of PBAs

� Budget support
� Other types of PBAs

CHART 1.12:
How much
programme-
based aid
do donors
provide?

33THE 2006 SURVEY ON MONITORING THE PARIS DECLARATION, VOLUME 1: PRE-PUBLICATION COPY - © OECD 2007

CONDUCTING JOINT MISSIONS

AND SHARING ANALYSIS

Indicator 10 measures the extent to which donors
are merging their review and analysis activities at the
country level. It counts the proportion of missions
to the country that were undertaken jointly by more
than one donor, and the share of country-analysis
exercises that were undertaken on a joint or co-
ordinated basis. The 2010 targets are 40% joint
missions and 66% joint analytical work.

Data adjustments were necessary to deal with
double-counting of both joint missions and joint
analytical work in the survey returns from many
of the countries. The results for missions are shown
in Appendix Tables A.10a and B.10a. Chart 1.13
provides an overview. The total number of
missions is strikingly high for some countries and
for some donors, notably large multilaterals. In
many cases, the number of missions appears to
be related to scale of the aid disbursed. However,
this is not always the case. Some kinds of support,
for example to governance reform, may be
mission-intensive for good reasons. At the same
time, the proportion of co-ordinated or joint
missions remains very low – an average of 16%
across countries and 24% across donors. It is clear
that the Paris Declaration targets in this area will
require intensified efforts.

This becomes more apparent when the returns are
broken down, as they are in the country chapters.
Quite a large share of the reported joint missions

is accounted for by the single area of public finan-
cial management assessment (CFAA and PEFA
in particular). A further substantial contribu-
tion comes from UN agencies, in which case it
is assumed that the bulk of the co-ordination is
internal to the UN system. Significant advances
outside these areas seem to be harder to achieve.
In a more positive vein, some partner governments
are taking a hand in the problem of mission over-
load themselves. In several African countries where
the total number of missions is large, governments
have taken the initiative by declaring mission-free
or “quiet” periods during which officials, especially
in Ministries of Finance, are able to concentrate
on budget preparation. This is an approach that
might be more widely emulated.

Appendix Tables A.10b and B.10b, together with
Chart 1.14, present the results for country analy-
tical work and joint country analytical work.
These suggest a rather greater tendency to under-
take activities jointly (in aggregate, 40% across
countries and 52% across donors). This may
reflect to some extent an understanding of the
phrase “country analysis” that is biased towards
relatively formal exercises that lend themselves
to joint working. There is, on the other hand, a
noticeable trend towards the preparation of Joint
Assistance Strategies, which may be expected to
raise the proportion of joint work in both the
mission and the analysis categories.

0

500

1500

2000

2500

3000

0

100

300

500

700

800

34 COUNTRIES LARGEST 22 DONORS

400

600

200

1000

� Co-ordinated donor missions � Individual donor missions
� Co-ordinated donor missions
� Individual donor missions

Number of missions
Country-by-country Donor-by-donor

CHART 1.13:
How many
donor missions
are co-ordinated?

INDICATOR 10A

In summary, the baseline position in joint
working may be worse for donor missions than
for country analysis, with the 2005 proportion of
17% of joint missions across countries falling well
short of the 2010 target of 40%. It is certainly
worse for missions if joint PFM assessments are
excluded. This is clearly an area in which greater
donor efforts are going to be needed if the agreed
targets are to be met.

MANAGING FOR RESULTS

One of the distinguishing features of the Paris
Declaration is an emphasis on the way improve-
ments in donor practices and country institu-
tions depend on each other and are mutually
reinforcing. This is reflected on the one hand in
the mutuality of the commitments, and on the
other in the inclusion of two major areas not
covered, for example, in the Rome Declaration on
Harmonisation and Alignment: mutual account-
ability and management for results.

The commitments on management for results
call for donors and partner countries to work
together to manage resources for the achieve-
ment of development results, using information on
results to improve decision making. Countries
are expected to develop cost-effective results-
oriented reporting and performance assessment
frameworks, while donors commit to using any
such arrangements and refraining from requiring
separate reporting. Indicator 11 measures the
extent to which the country commitment on

establishing performance frameworks has been
realised, using the scorings of the 2005 CDF
Progress Report.

Three assessment criteria are considered:
■ The quality of development information.
■ Stakeholder access to development

information.
■ Co-ordinated country-level monitoring

and evaluation.

The assessments therefore reflect both the extent
to which sound data on development outputs,
outcomes and impacts are collected, and various
aspects of the way information is used, disseminated
among stakeholders and fed back into policy.

Chart 1.15 shows the distribution of the coun-
tries participating in the 2006 baseline survey
across the CDF Progress Report five quality cate-
gories (A-E), while Appendix Table A.11 provides
further detail. Country performance assessment
frameworks are placed in category A if they
“substantially” achieve good practice, and in B if
they are “largely developed” towards good prac-
tice. The scores are for 2005 and are limited to
the 29 survey countries that were included in the
CDF exercise. The participating countries not
included are Afghanistan, Dominican Republic,
Egypt, Peru and South Africa.

On CDF Progress Report criteria, only two coun-
tries have “largely developed” performance assess-
ment frameworks compared to five with “largely
developed” operational development strategy

34 THE 2006 SURVEY ON MONITORING THE PARIS DECLARATION, VOLUME 1: PRE-PUBLICATION COPY - © OECD 2007

34 COUNTRIES LARGEST 22 DONORS

� Co-ordinated analytical work
� Individual analytical work

� Co-ordinated analytical work
� Individual analytical work

Number of country analyses

Country-by-country Donor-by-donor

0

200

600

800

1000

0

60

90

120

150

30

400

CHART 1.14:
How much
country analysis
is co-ordinated?

INDICATOR 10B

35THE 2006 SURVEY ON MONITORING THE PARIS DECLARATION, VOLUME 1: PRE-PUBLICATION COPY - © OECD 2007

A

B

E
PAF reflects little action toward achieving
good practice.

PAF incorporates some elements of
good practice.

PAF reflects action taken towards achieving

good practice

PAF is largely developed towards achieving

good practice.
7%2

0

0

17

10 34%

59% C

D

SCORENO. OF COUNTRIES CATEGORY

Performance assessment framework

(PAF) substantially achieves good practice

Source: World Bank CDF Progress Report for 2005.

LOWER SCORE ▼

HIGHER SCORE ▲

Quality of country performance assessment frameworks

CHART 1.15:
Do countries
have monitorable
performance-
assessment
frameworks?

INDICATOR 11

indicators. On the other hand, the surveyed
countries do better than the full CDF Progress
Report sample, with a substantially larger propor-
tion rated in at least the B or C categories than in
the whole sample. Countries participating in the
baseline survey in 2006 should therefore find it
easier than others to move up at least one cate-
gory. The overall target is to reduce by one-third
the proportion of countries not attaining at least
a B by 2010. This would mean raising the propor-
tion of countries having frameworks with at least
a B rating from the baseline of 7% to 38%. Nine
baseline countries need to join the two currently
scored B or A over the period.

What it takes to raise a country score by one cate-
gory is reasonably clear. In the country chapters, the
CDF Progress Report and AER country profiles
have been used to interpret the country scores. The
general picture that emerges is as follows. The supply
of survey-based data on poverty incidence and
human development variables has been improving
in most countries. This leaves the quality of admin-
istrative reporting and other sources of information
about intermediate performance variables as the
principal area of weakness on the data-generation
side. Data sharing and dissemination are somewhat
better than they were, but providing feedback loops
into policy improvement presents a major challenge
in most countries. This is at least partly because
– as discussed in connection with Operational
Development Strategies – country plans are often
weakest at the point where monitoring information
might be expected to have some purchase on policy;
that is, the specification of an implementation plan
in terms of activities to be undertaken. Actions on all
of these points are in principle quite feasible, given

sufficient high-level encouragement, and are being
in part addressed through specific programmes to
build capacity to produce and use statistics.

In summary, channelling evidence on results into
processes of policy improvement remains a major
obstacle in the large majority of surveyed coun-
tries. It is both important and very hard to achieve,
given the lack of simple technical fixes for the
problems in current arrangements for generating
and using data. Bringing the target percentage of
country performance assessment frameworks up
to the expected 2010 level will call for efforts that
are both well-coordinated and realistic about the
clear and evident barriers to performance-based
management in partner countries.

BOX 1.5: TWO AID EFFECTIVENESS
MUTUAL REVIEW MECHANISMS

“ Tanzania has recently established a Joint Assistance
Strategy for Tanzania (JAST), which replaces the Tanzania
Assistance Strategy as the framework for managing
development co-operation. It is more comprehensive…
in promoting aid effectiveness in line with the principles
of ownership, alignment, harmonisation, managing
for results, domestic and mutual accountability.
Government and development partner performance in
implementing the JAST will be regularly monitored and
evaluated on the basis of jointly agreed indicators and
targets… JAST indicators will also be drawn from the
Paris Declaration indicators, as adapted to the Tanzanian
context.” (Tanzania)

“ The Development Partnership Framework, recently
signed by the government of Moldova and donor
organisations, lays out principles, procedures and
processes aimed at greater aid alignment, co-ordination
and harmonisation. The Framework specifies concrete
activities… It will represent the basis for further mutual
assessment of progress.” (Moldova)

36 THE 2006 SURVEY ON MONITORING THE PARIS DECLARATION, VOLUME 1: PRE-PUBLICATION COPY - © OECD 2007

MUTUAL ACCOUNTABILITY

some in response to the 2006 survey, which may
to lead to the formulation of a local action plan. If
this happens, the proportion of positive responses
may increase when the survey is repeated in 2008.
This would bring closer the time when 100% of
countries have a mutual assessment mechanism,
the target that has been set for 2010.

A caution suggested by some of the country
reports is that lengthy periods sometimes elapse
between agreement on an HAP and the defini-
tion of a monitoring framework for it, with the
actual presentation and discussion of monitoring
data in an appropriate joint forum taking even
longer. A less elaborate form of agreement, with
a simpler monitoring mechanism, may be worth
considering in future initiatives.

In summary, the movement towards the estab-
lishment of specific mechanisms for joint moni-
toring of Partnership Commitments at country
level is just beginning, building in several coun-
tries on earlier experience with Harmonisation
Action Plans. It is to be expected that further
progress will already be apparent in 2008, so that
such arrangements are universal by 2010. ■

1 In accordance with paragraph 9 of the Declaration, the partnership of donors and partner countries hosted by the DAC (Working Party
on Aid Effectiveness) comprising OECD-DAC members, partner countries and multilateral institutions, met twice, on 30-31 May 2005
and on 7-8 July 2005 to adopt, and review where appropriate, the targets for the 12 Indicators of Progress. At these meetings an agree-
ment was reached on the targets presented under Section III of the present Declaration. This agreement is subject to reservations by
one donor on (a) the methodology for assessing the quality of locally managed procurement systems (relating to targets 2b and 5b)
and (b) the acceptable quality of public financial management reform programmes (relating to target 5a.ii). Further discussions are
underway to address these issues. The targets, including the reservation, have been notified to the Chairs of the High-Level Plenary
Meeting of the 59th General Assembly of the United Nations in a letter of 9 September 2005 by Richard Manning, Chair of the OECD
Development Assistance Committee (DAC).

Afghanistan

Bolivia

Cambodia

Cape Verde

Egypt

Ethiopia

Ghana

Malawi

Albania

Bangladesh

Benin

Burkina Faso

Burundi

Congo Democratic Republic

Dominican Republic

Honduras

Kenya

Kyrgyz Republic

“YES” “NO”

Countries that DO have mechanisms
for mutual review of progress (2005)

Countries that DO NOT have mechanisms
for mutual review of progress (2005)

Moldova

Mozambique

Nicaragua

South Africa

Tanzania

Viet Nam

Zambia

15 countries (44%)

Mali

Mauritania

Mongolia

Niger

Peru

Rwanda

Senegal

Uganda

Yemen

19 countries (56%)

INDICATOR 12

TABLE 1.1:
Do countries
have mechanisms
for mutual
assessment
of progress?

THE CONCEPT OF MUTUAL ACCOUNTABILITY is an
important innovation of the Paris Declaration.
It develops the idea that aid is more effective
when both donors and partner governments are
accountable to their respective publics for the use
of resources to achieve development results, and
when they are also accountable to each other. The
specific focus of the agreed indicator (Indicator
12) is on mutual accountability for the implemen-
tation of the Partnership Commitments included
in the Declaration and any local agreements on
enhancing aid effectiveness. Specifically, the
country survey returns tell us whether there exists
a mechanism for mutual review of progress on aid
effectiveness commitments.

The table indicates that a minority of surveyed
countries (41%) have in place a mechanism of this
type. Box 1.5 describes two examples. In several
countries, Harmonisation Action Plans (HAPs)
were agreed following the Rome Declaration, or
where governments have initiated the formulation
of country Aid Policies with harmonisation and
alignment dimensions. Other country responses
indicate that discussions are taking place now,

37THE 2006 SURVEY ON MONITORING THE PARIS DECLARATION, VOLUME 1: PRE-PUBLICATION COPY - © OECD 2007

LIST OF SOURCES

World Bank (2005), Enabling Country Capacity to Achieve Results: 2005 CDF Progress Report, Volume I:
Overview; Volume II: Country Profiles, World Bank, Operations Policy and Country Services,
Washington, D.C. http://www.worldbank.org/cdf

World Bank, “Draft Aid Effectiveness Profiles”, Aid Effectiveness Review, World Bank, Operations
Policy and Country Services, Washington, D.C., forthcoming,
http://go.worldbank.org/MR5I91FSP0

World Bank, Country Policy and Institutional Assessment, World Bank, Washington, D.C., http://
www.worldbank.org

OECD-DAC, International Development Statistics online, http://www.oecd.org/dac/stats/idsonline)

THE BASELINE SURVEY WAS DESIGNED TO: stimulate a broad-based dialogue (at both
country and international levels) on how to make aid more effective, promote agree-
ment on specific actions that contribute to successful implementation of the Paris
agenda at country level, and generate an accurate picture of existing aid practices as
a baseline for assessing progress.

This chapter examines how far the survey has proven to be an efficient and
effective instrument for pursuing these objectives. It considers whether they
were well served by the way the process was managed, and whether in prac-
tice there were significant trade-offs among the objectives. It concludes by
identifying the main areas in which the approach needs to be adjusted in future
rounds of the survey.

THE CHAPTER DRAWS ON TWO SOURCES: the survey returns themselves, and the insights
and opinions provided by some of the main stakeholders in the process. Important
stakeholders include the National Co-ordinators who organised the process at
country level; the donor agency respondents at country and headquarters levels; and
the DAC Secretariat and help-desk personnel at the OECD, the UNDP and the
World Bank who guided the survey process. Face-to-face or telephone interviews
were carried out with as many representatives of these stakeholder groups as possible,
to obtain a well-rounded and realistic picture of the survey process, and its principal
strengths and weaknesses.

2 THE SURVEY PROCESS

39THE 2006 SURVEY ON MONITORING THE PARIS DECLARATION, VOLUME 1: PRE-PUBLICATION COPY - © OECD 2007

SCOPE OF THE EXERCISE

The countries that participated in the survey were
a self-selected sample. From the 50-60 devel-
oping and transition countries that endorsed the
Paris Declaration, 43 expressed initial interest in
the exercise and 34 actually completed the data
collection. The sample thus includes over half
of the countries eligible to participate. It covers
36% of global country programmed ODA in
2005 (see Appendix Table A.0). The proportion
of highly aid-dependent countries included is no
doubt larger than that. Participation was volun-
tary and called for a significant effort on the part
of the country authorities. Therefore, the sample
cannot be considered representative of the rele-
vant universe. Rather, it may be assumed that it
includes a disproportionate number of countries
where there is already active interest in the Paris
Declaration process on the part of the govern-
ment, the country’s donors, or both.

This fact has been borne in mind in interpreting
the results in Chapter 1. The survey findings indi-
cate the baseline situation for a large set of coun-
tries forming the advance guard of awareness
on aid effectiveness. The results that would have
been obtained from a larger sample would very
likely have indicated a lower baseline and greater
progress to be made to meet the 2010 targets.
Anecdotal evidence from a number of bilateral
donors that have undertaken their own reviews
of performance in all their programme countries
is consistent with this assumption.

The baseline survey included only a handful of
countries classified by the OECD-DAC as fragile
states. In fact, the number was so small that it was
not possible to investigate any distinctive quali-
ties of Paris Declaration implementation in those
countries. In future rounds of the survey, it should
be possible to improve the coverage of signatory
countries in general and of fragile states in partic-
ular. This does not need to affect the validity of
comparisons within the baseline sample. From
that point of view, the most important thing will
be to ensure that countries that participated in
the 2006 survey are included again in the next
round in 2008.

BOX 2.1: A TYPICAL COUNTRY PROCESS

Country-level processes generally included four steps:

STEP 1: A launch phase, often including a workshop or
individual meetings to present the survey method-
ology. Launch workshops provided an initial platform
for discussing the Paris Declaration, the commitments
entered into, and the indicators agreed upon for moni-
toring practices and progress in aid delivery.

STEP 2: A data-collection phase (involving govern-
ment and donors). Local donor offices were frequently
involved in discussions with their headquarters and with
government. In this phase, most countries were asked
to comment on the draft findings of the World Bank
Aid Effectiveness Review, on the issues covered by Paris
Declaration Indicators 1 and 11.

STEP 3: A discussion phase around the indicators and data
(including efforts to elicit data from laggard donors). This
phase was critical. The data gathered by government
and donors diverged in many cases, leading to intense
discussion around the interpretation of some indicators.

STEP 4: A finalisation phase. The Country Worksheet was
compiled and sent off to the DAC Secretariat. In the
subsequent period of follow-up, National Co-ordinators
responded to questions concerning the data submitted
and organised country responses to the draft country
chapters.

Several regional and sub-regional meetings on aid effec-
tiveness took place in the wake of the survey, which
provided opportunities for exchanging experiences on
the survey process.

HOW THE SURVEY WAS MANAGED

At the country level, the survey was organised
to spark dialogue among government, donors
and other in-country stakeholders, while also
collecting useful information. This was reflected
in the co-ordination of the survey, the method for
putting together the information required, the
guidance provided and the process of drafting
the chapters on the countries. Box 2.1 describes a
typical country process.

In keeping with the spirit of the Paris Declaration,
the country surveys were convened by govern-
ment. Participating countries each appointed a
National Co-ordinator as leader of the survey
process. Normally, these were the heads of aid co-
ordination units in ministries of finance or foreign
affairs, or the directors’ offices in agencies for
external co-operation. In many places, individual

40 THE 2006 SURVEY ON MONITORING THE PARIS DECLARATION, VOLUME 1: PRE-PUBLICATION COPY - © OECD 2007

MANAGING THE SURVEY TO STIMULATE DIALOGUE

local donors (e.g. UNDP, World Bank, European
Commission, DFID or USAID) provided real
support, including practical help to National Co-
ordinators. However, the reports from a variety
of stakeholders suggest that governments main-
tained leadership throughout the process.

A degree of agreement among the in-country
participants was required. The country returns
took the form of a single Country Worksheet
containing both quantitative information and
qualitative commentary relating to the relevant
Paris Declaration indicators. The worksheets were
compiled by the National Co-ordinator, using the
completed two-page questionnaires submitted
by in-country donors and government. In many
cases, this involved extensive consultations to
iron out inconsistencies and reach agreement
on issues. Donors testify that communication
between themselves and National Co-ordinators
was usually open and transparent, even if there
were differences of opinion at times.

On launching the survey in May 2006, the DAC
Secretariat provided a package of materials which
steered a careful course between over- and under-
direction of the process. The materials included
an Explanatory Note, two-page questionnaires
for completion by government and donors, the
Country Worksheet (incorporating, an Excel
file for the consolidated numerical data) and
a comprehensive “Definitions and Guidance”
document. These were distributed to National
Co-ordinators and also made available through
the OECD’s website on monitoring the Paris
Declaration. The website included information
about the help desks operated by the UNDP,
the OECD Secretariat and the World Bank, and
other useful tools.

Because of the novelty of the exercise and the
communication challenges involved, the survey
started with some delay in a number of countries.
The initial deadline for submissions was 15 August
2006. A final cut-off date of 30 September was
set. Some 26 submissions were received by this
deadline, with an additional 8 arriving within the
following month, completing the sample of 34.

Chapters on the country findings were then
drafted. This proved a time-consuming task, but
was undertaken as speedily as possible in order
for initial drafts to be submitted to the coun-
tries for comment with a minimum of delay.
Checking of initial drafts led in many cases to
a further round of discussion on the status of aid
effectiveness initiatives in the country. Second
drafts, adjusted in light of corrections and addi-
tional information, were also circulated to the
countries. The final drafts of the country chapters
take full account of the comments received.

TYPES OF COUNTRY EXPERIENCE

It is clear that this process as a whole was onerous,
particularly for the country representatives.
A minimum of three person-weeks, and often
several times this amount, had to be devoted to
the task in those countries where engagement was
most intense and included a relatively wide range
of stakeholders. These estimates are based on
the conversations with National Co-ordinators.
They account only for the time that was devoted
by government staff dedicated to the survey, and
do not include time of donor staff or government
staff from other ministries or agencies who partic-
ipated in related meetings and workshops. The
task was made more difficult as it took place in
the Northern summer period, when staff of many
donor organisations take their annual holidays.

The process has been burdensome despite the
relatively narrow focus and scope of the exercise.
The questionnaires and consultations focused
only on a few proxy indicators of aid alignment,
harmonisation and mutual accountability, and
not on the full range of Paris Declaration issues.
Moreover, in most countries, the process included
government officials and multilateral and DAC
bilateral donors only. The involvement of govern-
ment departments other than central ministries
and aid co-ordination units was variable. The
goal of involving a wider range of stakeholders,
including non-DAC donors and civil society, was
achieved only in a few cases. Ghana was one of a
handful of countries where the process involved
civil society. The international community did

41THE 2006 SURVEY ON MONITORING THE PARIS DECLARATION, VOLUME 1: PRE-PUBLICATION COPY - © OECD 2007

not encourage participation by civil society at
the outset, and additional stakeholders that did
become involved did so at their own initiative.
Some foundations submitted data on the donor
questionnaire in individual countries (e.g. the
Aga Khan Foundation in Afghanistan). Box
2.2 shows some of the ways in which country
processes differed.

A question to be considered is how much of the
effort was worthwhile, either as a contribution to
the international monitoring effort or because of
the stimulus it provided to the in-country debate
on aid effectiveness. In other words, which of the
efforts brought with them benefits that are suffi-
cient to justify the costs, and which did not?

The feedback from stakeholders suggests the need
to distinguish among two or three groups of
countries participating in the survey. The first
group are those countries where discussions have
been under way for some years, at least since
the Rome Declaration, on the development of
a Harmonisation Action Plan and/or a compre-
hensive national aid policy. In the second group,
harmonisation and alignment had not been
discussed until very recently at country level. In
these circumstances, the survey brought together
the relevant stakeholders for the first time. A
possible third category consists of countries where
action on alignment and harmonisation had been
previously discussed but the process was dormant
for various reasons.

In the first group, the survey process was able to
make use of existing relationships and procedures.
This simplified the task. However, the previous
history also meant that there was little immediate
gain from the meetings. National Co-ordinators
from countries in this first category were notice-
ably more inclined to pose “so what?” questions,
and to be impatient with a survey instrument that
appeared to be limited to recording the current
state of donor practices. The potential benefits,
for these countries, are associated more obviously
with the second purpose of the survey – that of
promoting agreement on specific actions leading
to appropriate forms of implementation of the

42 THE 2006 SURVEY ON MONITORING THE PARIS DECLARATION, VOLUME 1: PRE-PUBLICATION COPY - © OECD 2007

BOX 2.2: COUNTRY VARIATIONS

The survey process varied across countries in the
following respects:

■ The scope of the questionnaires: Some National Co-
ordinators added further questions, or provided addi-
tional guidance (for example, on the requirements of a
programme-based approach or co-ordinated capacity
development). Others limited themselves to collating
the contributions.

■ The number of donors that submitted data and the
level of effort necessary to elicit responses from other
major donors in the country: Peer pressure within a
donor group was a favourable factor in some coun-
tries, but was largely absent in others. Nevertheless
the overall coverage of relevant aid flows was high at
90% ranging from 57% to 118% (Appendix Table A.0).

■ The intensity of data collection within government: In all
cases, this involved the quantitative data required to
answer the government questionnaire (on Indicators 3
and 7), but some National Co-ordinators went further
and, for example, sought line ministries’ views on the
number of parallel PIUs in sectors.

■ The length and intensity of the discussion phase: This
was affected by capacity, commitment and time
constraints, especially in countries where the initial
launch of the survey was delayed.

■ The way Country Worksheets were finalised: Some
country groups sought consensus between govern-
ment and donors, while others in accordance with the
guidance on the worksheet chose to reflect diverging
views where there was disagreement.

■ The degree of engagement with the Aid Effectiveness
Review undertaken by the World Bank: The draft
profiles prepared for the AER were helpful in
completing Country Worksheets, both in countries
where discussion on harmonisation and alignment
was recent and in those where it was well established.
However, the degree to which the AER was effectively
connected with the survey process proper seems to
have been quite variable.

Paris agenda at country level. This represents a
much tougher test, and the feedback on whether
the process contributed on balance to this objec-
tive is rather mixed, as explained below.

It is, on the other hand, clear that there were
important benefits in the second group of coun-
tries. Feedback from the National Co-ordinators
suggests that the level of awareness of the Paris
Declaration commitments is often surprisingly
low in both government and donor circles at
country level. Communication between head-
quarters policy advisers and operational staff on
Paris Declaration issues appears not to be very
effective in many agencies. However, these weak-
nesses are even more prominent where there has
been little previous discussion in the country.

Where this is the case, the interview evidence
indicates that the survey process has made a
substantial contribution. The process has raised
awareness and improved understanding at the
country level about the Paris objectives and
commitments. It has triggered debate on various
features of current aid practice, notably the role
and rationale of parallel project implementation

units (PIUs), and the scope for making greater
use of country systems. Participants have shared
their experience, among the local offices of donor
agencies and among the National Co-ordinators
of recipient countries. This has contributed to
the development of “learning networks” around
aid effectiveness.

The survey has spread the Paris message at the
operational level in donor agencies in a way that
directives from headquarters probably never
could. In many countries, it has highlighted the
variety of donor approaches and practices. Some
country offices have been prompted to consult
their headquarters on the range of aid practices
permissible under agency rules, sometimes with
surprising results. The experience of answering
the survey questions has also helped raise aware-
ness within recipient governments, including
about some of the concrete steps which might
be taken by them (e.g. Egypt, Mongolia). On
the whole, however, the effect on governments
has been limited by the fact that line ministries
and specialist agencies were in most cases only
involved at the margins.

BOX 2.3: FOCUSING ATTENTION ON KEY ISSUES

The survey helped to focus attention on important issues, including:

■ Country planning and performance management systems: Although the assessment of country systems was not fully
integrated into the survey process (see next section), strengthened dialogue on how to improve aid effectiveness
has drawn attention to weaknesses in country systems and the need for joint actions to address them.

■ Project implementation units: The survey has enriched country-level discussions on this topic, with a focus not only
on the number of parallel units, but on the rationale for using them, their effects and different views on the potential
for achieving greater integration with government structures without loss of development effectiveness.

■ Procurement systems: In recent years, many recipient countries have adopted new procurement legislation, modelled
on international best practices. Country representatives commonly feel that buy-in from donors has been inade-
quate, while donors report that legal rules may have improved, but systems are still weak. Discussion of Indicator 2b
has moved this issue up the agenda. In the next survey round, international ratings of country procurement systems
will be available to answer the question applied in 2006 to public financial management: is donor use of country
systems correlated with the strength of the systems?

■ Data systems for monitoring aid: National Co-ordinators indicate increased government interest in national aid
reporting and recording systems, including those that might allow more continuous monitoring of the aspects of
aid delivery highlighted by the Paris Declaration.

■ Co-ordinated capacity development: Many recipient country officials believe that countries have seen relatively little
benefit from technical co-operation, even though it constitutes a significant share of aid. The survey has opened
a debate about the changes needed on both donor and government sides to enable capacity development to work
better.

■ Use of common arrangements or procedures: The survey has helped to draw attention to the lack of a widely shared
understanding of the Paris Declaration commitment on common procedures.

43THE 2006 SURVEY ON MONITORING THE PARIS DECLARATION, VOLUME 1: PRE-PUBLICATION COPY - © OECD 2007

There is evidence that “dormant” country
harmonisation processes have been reinvigorated.
The survey has given actors within governments
who believe that aid effectiveness is an important
policy issue a valuable tool for reviving harmoni-
sation processes. The data generated are often
striking (e.g. the low degree to which national
procedures are currently being used). The data
reinforce the visibility of problems, helping
to bring them onto the country agenda. The
regional activities around the survey process were
instrumental in disseminating ideas and practices
regarding the possible design of national aid poli-
cies. Box 2.3 illustrates the way the survey has
focused attention on key issues in a number of
country contexts.

BENEFITS AND COSTS

In all cases, there are remaining questions about
costs associated with undertaking the survey that
did not have clear benefits and might possibly have
been avoided. Insisting upon a degree of collec-
tive agreement on the returns incurred signifi-
cant costs but did provoke dialogue, which raised
awareness and posed relevant new issues for the
participants. The same is probably true about the
iterative process of refining the country chapters
for the report. But it is less clear that the approach
taken by the guidance materials contributed to a
positive cost-benefit ratio.

The “Definitions and Guidance” document was
quite directive. For each of the agreed Paris
Declaration indicators covered by the survey, it
defined the terms used and offered additional
guidance on their interpretation. The Explanatory
Note, on the other hand, contained a more
permissive set of instructions. It recognised the
possibility that the definitions offered might not
cover all possible cases, exceptions and contingen-
cies, and invited partner authorities and donors in
each country to reach agreement on how to adapt
the definitions so that they might be consistently
applied throughout the survey. The Country

Worksheets included an instruction to record any
differences of opinion. As the discussion of find-
ings in Chapter 1 has made clear, this flexibility
in the guidance was quite widely interpreted as a
mandate for the use of different definitions.

The final section of this chapter considers the diffi-
culties this has created for consistent application
of the survey, within countries, across countries
and through time. Another question is worth
asking first. Did this flexibility help in stimu-
lating country-level dialogue? A similar question
will be considered in the next section, in relation
to the objective of getting agreement on specific
Paris Declaration implementation measures.

The testimony of the National Co-ordinators
seems quite clear. The ability of individual donor
respondents to adopt, in effect, their own defi-
nition of key terms, such as “programme-based
approach” made the process more burden-
some without any real justification. National
Co-ordinators did what they could to ensure
consistent reporting on the Country Worksheet
– for example, on what types of support to
capacity development should be regarded as
“co-ordinated”. However, there were limits to
the National Co-ordinators’ authority, particu-
larly where donor representatives were receiving
instructions from their headquarters. The guid-
ance provided by donor headquarters sometimes
overruled agreements reached at country level
between National Co-ordinators and donors.
The Co-ordinator’s role in this regard was diffi-
cult, stressful and time-consuming.

It is important to consider possible trade-offs.
However, it does not seem that the flexibility on
definitions produced useful dialogue that would
promote the Paris Declaration’s contribution to aid
management in the country. Instead, it diverted
dialogue that might have been productively
focused on the facts of the situation in the country
into a rather unproductive discussion of how to
report those facts for the purposes of the survey.

44 THE 2006 SURVEY ON MONITORING THE PARIS DECLARATION, VOLUME 1: PRE-PUBLICATION COPY - © OECD 2007

45THE 2006 SURVEY ON MONITORING THE PARIS DECLARATION, VOLUME 1: PRE-PUBLICATION COPY - © OECD 2007

IN COUNTRIES where the discussion on harmoni-
sation and alignment was already advanced, the
expectation was that the survey process would
be decidedly action-oriented. This expectation
was fulfilled in some respects. But the ability of
the survey to play an action-oriented role was
constrained by a number of factors.

IMMEDIATE AND PROSPECTIVE GAINS

The immediate gains include a widening or
deepening of existing country efforts. In some
countries, the survey was taken as an input
into established national monitoring efforts on
aid effectiveness, such as those surrounding a
Harmonisation Action Plan. In these cases, the
process drew in a wider range of stakeholders,
renewing their commitment to tracking the
actions previously agreed.

In a few such countries, additional questions were
integrated into the questionnaires with a view
to generating more “actionable” results. They
included questions about obstacles to progress.
The results of these efforts remain to be assessed,
but the practice suggests that future rounds of the
survey could be made more action-oriented for
all countries, without increasing the scale of the
exercise significantly.

In a wider range of countries, the survey has trig-
gered debates about methods of record keeping
and reporting on aid, on both the government
and donor sides. As noted in Chapter 1, data
recording systems at country level are often not
well-geared to capturing aid flows in a compre-
hensive way. There are substantial weaknesses on
both the donor and government sides affecting
the capture of information. In addition, the aid
statistics that exist tend to be disaggregated in
traditional terms (e.g. loans vs. grants) and do
not distinguish between aid modalities (projects,
programme-based, etc.) along the lines required
for Paris Declaration monitoring. In a number of
countries, the survey has drawn attention to these
issues, and led to stepped-up efforts to develop
national aid management systems.

Many National Co-ordinators indicate that
the survey results are still being “digested”. In a
number of countries, the Co-ordinators expect
that the results will be translated into an action-
able agenda in the near future. For example, there
may be plans for phasing out parallel project
implementation units, with definite commit-
ments by government to strengthen line minis-
tries, and by donors to respond in kind. The
survey question about mechanisms for moni-
toring mutual accountability has encouraged
some country aid-management teams to consider
introducing such a mechanism. The prospect of
future rounds of the survey will likely provide
an incentive for the partners to show progress over
the next 18 to 24 months.

CONSTRAINTS ON THE FORMULATION

OF ACTION AGENDAS

Two features of the 2006 survey process placed
constraints on its ability to stimulate local action
agendas. One was the separate treatment of two
broad areas of the Paris commitments – aid
delivery on the one hand and country systems
on the other. The second was the scope that the
guidance allowed for discussion to be diverted
from factual into definitional questions.

As explained in Chapter 1, the proxies for align-
ment, harmonisation and mutual accountability
(Indicators 3, 4, 5, 6, 9, 10 and 12) were dealt
with in the Country Worksheets. In this sense,
they were fully part of the country process. On
the other hand, the measures of country owner-
ship and results orientation (Indicators 1 and
11) were based on two World Bank outputs:
the Comprehensive Development Framework
Progress Report for 2005 and the forthcoming
Aid Effectiveness Review. Work on these indica-
tors was undertaken in parallel to the country
survey processes. Draft country profiles for
the AER were sent to countries for comment
during the survey period. In a number of coun-
tries, including Ethiopia, Mali, Moldova and
the Kyrgyz Republic, the draft profiles served
as an input into the Country Worksheet and
thus served to advance discussion on progress.

CONTRIBUTING TO AGREEMENT ON SPECIFIC ACTIONS

However, in other countries commenting effec-
tively on the AER profiles while also completing
the survey was found to be difficult. In these
cases, the two main parts of the survey were effec-
tively disconnected.

These constraints probably hampered the ability
of donor-partner groups at country level to draw
up action plans based on the survey experience.
In very many areas, including the development
of operational development strategies and perfor-
mance assessment frameworks, the ability to move
forward on the Paris commitments is influenced
by both donors and government. However, formu-
lating a reciprocal action agenda is more difficult
when the evidence on the two sides of the equation
has been assembled and validated in a different
way. It is clear that in future there will be gains in
terms of actionable results if the survey process can
become more integrated in this sense.

The second issue has already been touched upon.
The DAC survey team provided substantial guid-
ance to respondents, but without tightening
the definitions of the indicators fully. There was
concern that the indicators should be sufficiently
flexible to accommodate a variety of country
circumstances. Country groups had room to
reach a consensus on locally adjusted definitions
that could then be applied consistently over time.
Unfortunately, few groups seem to have been able
to reach a consensus and National Co-ordinators
were not empowered to impose a ruling. The issue
took up a good deal of debating time that might
otherwise have been devoted to discussing how to
move forward. In several countries, this was seen
as an unwelcome distraction by many stakeholders
on both the government and the donor sides.

ACCURACY OF THE PICTURE

46 THE 2006 SURVEY ON MONITORING THE PARIS DECLARATION, VOLUME 1: PRE-PUBLICATION COPY - © OECD 2007

AS A RESULT of the baseline survey, more is now
known about aid management systems and prac-
tices at the country level. Quantitative data have
been generated on a set of reasonable proxies of
the Paris Declaration commitments. These data
provide a set of benchmarks covering most forms
of official aid in 34 countries from all world
regions (formerly, such information was avail-
able only for specific modalities, such as general
budget support and for particular world regions).
The data will provide an invaluable input to forth-
coming debates on how to improve the effective-
ness of aid. The availability of such data is likely
to be a factor favouring forward movement on
aid effectiveness in countries that participated in
the first survey round and others that may join in
future rounds.

The picture provided by the results is as accurate
as possible, given the constraints that affected the
survey process. Considerable efforts have been
made to deliver high-quality data, not least by
the National Co-ordinators who compiled and
checked the returns and responded to at least
one round of follow-up queries from the OECD-
DAC Secretariat. The use of the data in the
country chapters and in Chapter 1 has taken into
consideration a range of factors influencing the
reporting. In fact, the information contained in
the survey returns on how the data were compiled
is itself an important source of insight into the
state of the Paris Declaration in particular coun-
tries. Thus, in this report, a naïve utilisation of
the data has certainly been avoided. It is to be
hoped that other users of the raw numbers will be
similarly circumspect.

That said, an important lesson from the manage-
ment of the 2006 survey is about the need for firmer
steps to ensure the standardisation of measures.
The difficulties created for National Co-ordinators
by the lack of standardisation have already been
noted. However, concern about this issue was also
common among donors, particularly in respect of
indicators on which some country-level discretion
was permitted by the guidance.

47THE 2006 SURVEY ON MONITORING THE PARIS DECLARATION, VOLUME 1: PRE-PUBLICATION COPY - © OECD 2007

It is clear in retrospect that the survey guidance
gave too much ground to the case for “localising”
concepts and measures. It is, of course, difficult
to produce watertight definitions, and there are
some genuine grey areas and classification prob-
lems needing discussion. Firmer definitions
should not be used to gloss over disagreements
about the substance of aid policy. Nevertheless,
definitions that are looser than they need to be
prevent clear discussion of policy differences and
do little to assist the technical discussion than
may be needed on the genuine borderline cases
and grey areas.

A particular concern about the 2005 baseline
established by the 2006 survey is that it may
systematically overstate the progress already
achieved in relation to the 2010 targets. If, as
a number of National Co-ordinators perceive,
donors have used the permitted leeway on defi-
nitions to place their performance closer to the
targets than would otherwise be the case, this
may create difficulties for them in demonstrating
that they have made progress over the first moni-
toring period (2005-07). If, as is recommended,
the guidance provided in the 2008 survey is
tighter and National Co-ordinators are given
greater authority, some of the results may show a
shocking – although actually illusory – reduction
in performance against the commitments. The
donor community should be prepared to meet this
eventuality in a mature and far-sighted way.

Most stakeholders in the 2006 survey seem to
take the view that firmer and clearer guidance
on the definition and local application of indica-
tors is desirable. This would help to focus discus-
sion on the things that matter – why the Paris
commitments were made, what actions can be
taken to carry them out and what the real obsta-
cles are. This means that in preparation for the
next survey round, a further effort will be needed
to clarify and disseminate the current definitions
and hammer out in advance the degree to which
they can be consistently applied.

ADJUSTING THE APPROACH IN 2008

THE BASELINE SURVEY has been a major multi-
purpose effort that has yielded important results
in the form of both substantive findings and
lessons learned. The process lessons suggest a
number of recommendations for the next round
of monitoring in 2008, discussed here and
summarised in Box 2.4.

The 2008 survey can and should involve more
countries, including a larger number of fragile
states, so that the exercise provides a more
broad-based stimulus to implementation of the
Paris commitments as well as a richer data set
with which to monitor progress. A change in
the sample need not compromise the ability to
measure changes against the baseline established
by the 2006 survey, as the 2006 baseline coun-
tries can continue to be treated as a distinct subset
for the purposes of statistical analysis.

A serious effort needs to be made to reduce those
costs incurred in mounting the survey from which
no clear benefits are expected. Such costs include
the duplication of effort arising from undertaking
the work at a time of year when donor representa-
tives are hard to contact.

BOX 2.4: SUMMARY OF RECOMMENDATIONS

In brief, the 2006 survey experience calls for:

■ Including more countries in the survey.

■ Reducing those survey costs for which no definite
benefit can be identified.

■ Assembling an indicative body of 2007 disbursement
data early in 2008.

■ Tightening the guidance on all 12 indicators.

■ Increasing the authority of National Co-ordinators.

■ Capturing additional information on practical
obstacles and ways of moving forward.

There are, however, several constraints on the
timing of the next survey that will need to
be taken into account in settling on dates for
starting and completing the survey. One of them
is the limited availability early in the year of data
on aid disbursements in the previous financial
year. Another is the deadline for reporting to
the High-Level Forum in Ghana, scheduled for
September 2008. For the 2008 survey round, the
concern to reduce transaction costs for National
Co-ordinators combines with the reporting
schedule to suggest that the survey should start
earlier than in 2006. This implies that special
efforts will be needed to assemble an indicative
body of disbursement data for 2007 no later than
the first quarter of 2008.

Not all of the unnecessary costs cited by National
Co-ordinators relate to timing. Another impor-
tant factor was the layer of complexity introduced
by the guidance encouraging local adaptations
in dealing with some of the indicators. The
2008 round of the monitoring survey should
put in place mechanisms to ensure standardised
reporting based strictly on centrally agreed defi-
nitions for the 12 Paris Declaration indicators. A
tightening of the survey guidance in this respect
would not just save time and help to generate more
robust data. As argued above, it would probably
also be more favourable than the current flex-
ibility to well-focused, action-oriented dialogue
at country level.

The survey instrument should provide scope for
feedback on areas of difficulty in applying the
definitions, but not for variations that weaken
the consistency of the survey data. In view of
the need for comparability between the 2006
and 2008 exercises, it is recommended that
feedback be collected on the ways in which any
changes in the guidance have affected the scores
for particular indicators.

48 THE 2006 SURVEY ON MONITORING THE PARIS DECLARATION, VOLUME 1: PRE-PUBLICATION COPY - © OECD 2007

The mechanisms required for implementing these
changes will need to include not just a careful
revision of the written guidance materials, but
also preparatory discussions involving donor
agency staffs at different levels as well as country
representatives. This will be important to ensure
that the principles behind tightening of the guid-
ance are understood and appreciated. In any
case, employment rotation will mean that those
responsible for gathering data will very likely be
different in future survey rounds, and this needs to
be considered. As part of this package of changes,
consideration should be given to increasing the
authority of National Co-ordinators, to enable
them to play more effectively their role as guar-
antors of the consistency of reporting.

So as not to increase the scale of the task, the
2008 survey round should focus closely on the
agreed 12 indicators. However, in two particular
respects adjustments are suggested.

First, additional attention should be given to
capturing insights into obstacles to progress on the
indicators and practical ways of moving forward.
This recommendation draws on the experience of
a few countries of using supplementary questions
to help identify a country-specific action agenda
on aid effectiveness. The experience suggests that
further information could be gathered in a cost-
effective way, with significant additional bene-
fits. It should be clear that this is not a question
of increasing the number of indicators, but of
extracting more benefits from the existing opera-
tion by encouraging country teams to make fuller
use of the spaces for qualitative comment on the
Country Worksheet.

49THE 2006 SURVEY ON MONITORING THE PARIS DECLARATION, VOLUME 1: PRE-PUBLICATION COPY - © OECD 2007

Second, there would be clear benefits from
bringing data collection for Indicators 1 and 11
(operational development strategy and results-
oriented performance assessment) within the
compass of the survey process under National
Co-ordinators. The evidence reviewed here
suggests that there would be gains for the quality
of country-level processes from ending the sepa-
ration of these two components of the Paris
Declaration monitoring effort. In particular, a
merged process for data collection would be more
conducive to a frank discussion that recognises
fully the reciprocal obligations and interdepen-
dencies involved in any forward action agenda.
For example, a revision of the AER profiles could
be conducted in-country by integrating the text
of the profiles into the survey for the 2008 moni-
toring round.

There would remain a need for standardised
measures comparable with those generated by
the CDF Progress Report and Aid Effectiveness
Review in the past. For this reason, there is going
to be a continuing need for a credible external
body such as the World Bank to take respon-
sibility for attaching scores to the descriptive
profiles, even if the profiles themselves are gener-
ated largely by an in-country process. If a joint
approach of this kind can be agreed, an early
start should be made in defining its modalities
in detail, so that they can be fully reflected in the
preparatory discussions. There is relevant experi-
ence to draw on, in which a simple scoring system
is applied to a descriptive analysis that includes a
large element of self-assessment (e.g. PEFA).

These two adjustments could help to ensure that
the 2008 round of Paris Declaration monitoring
is a rewarding process that adds substantially to
the baseline survey. Together with the proposed
efforts to involve more countries, adjust the sched-
uling and strengthen the survey guidance, they
would be expected to enhance the benefits of the
exercise at both country and international levels
while keeping its costs to a reasonable level. ■

THE PARIS DECLARATION ON AID EFFECTIVENESS (2 March 2005) is an ambitious
attempt to improve development impact through more effective use of aid. To this
end, it promotes changes to aid delivery and management practices based on agreed
principles of country ownership, harmonisation, alignment, mutual accountability
and managing for results. The Declaration goes beyond committing parties to a
clearly specified set of actions and behavioural changes. It also provides for periodic
monitoring at the country level, so that the governments of developing countries and
their development partners are increasingly accountable to each other for the progress
being made.

THIS CHAPTER DRAWS CONCLUSIONS from the first round of monitoring in 20061 and
suggests recommendations that should help achieve progress against the targets
agreed in the Paris Declaration. It is structured in four parts. The first part sets out
the key policy recommendations that deserve priority attention, the second provides
the context and the state of play in 2005, the third identifies the main challenges in
accelerating implementation of the Paris Declaration, and the last draws out opera-
tional recommendations.

The baseline survey was designed to document the situation in 2005 with a view to
benchmarking progress in 2008 and in 2010. It does not provide any direct evidence
on trends of change; that will require a second survey. It describes the position in
a self-selected sample of 34 developing countries, and does not claim to provide a
fully comprehensive assessment of aid practices across the world – though it does
cover 37% of relevant aid flows. Rather, it helps focus attention on the actions that
could be taken in the years remaining before 2010, the target date for fulfilling the
Paris commitments and a critical milestone on the road to meeting the Millennium
Development Goals in 2015. This chapter provides a summary of conclusions from
the 2006 monitoring round; a more in-depth analysis of the conclusions may be
commissioned later.