G l o b a l E n v i r o n m e n t F a c i l i t y

GEF/ME/C.30/6
November 2, 2006
GEF Council
December 5-8, 2006

Agenda Item 9






EVALUATION OF THE GEF ACTIVITY
CYCLE AND MODALITIES

JOINT EVALUATION OF THE GEF EVALUATION OFFICE AND THE EVALUATION
OFFICES OF THE IMPLEMENTING AND EXECUTING AGENCIES OF THE GEF

(Prepared by the GEF Evaluation Office)

---

Recommended Council Decision


The Council, having reviewed document GEF/ME/C.30/6, Evaluation of the GEF
Activity Cycle and Modalities and the management response (GEF/ME/C.30/7) takes
note of the conclusions of the evaluation that:
(a) the GEF activity cycle is not effective, nor efficient, and that the situation has
grown worse over time; nor is it cost-effective;
(b) GEF modalities have not made full use of trends towards new forms of
collaboration that serve to foster ownership and promote flexibility, efficiency
and results.
The Council agrees with the management response that no gains would be achieving by
streamlining the current project cycle at the margins. Therefore, the Council requests the
Secretariat, in consultation with all the GEF entities, to present for Council
consideration in June 2007 one or more options for a new project cycle and associated
business processes, with the objective of processing a proposal from identification to
start of implementation in less than 22 months. This wo uld include the following
decisions on the recommendations of the evaluation:

(a) ensure quality through greater emphasis on results-based management during
implementation with a corresponding reduction of detailed information that is
currently required in the formulation and appraisal stages of the cycle.
(b) ensure that the project identification phase is focused on establishing project
eligibility, resource availability and country endorsement;
(c) the Work Program as presented to Council should move towards the strategic
level;
(d) ensure endorsement by the CEO of fully documented project proposals on a
rolling basis.
The Council requests the GEF Evaluation Office to report through the Management
Action Record on the follow-up to this decision.



ii

---

Table of Contents

EXECUTIVE SUMMARY...................................................................................................................1

1. BACKGROUND
...................................................................................................................4


2. SCOPE AND METHODOLOGY......................................................................................................5

3. MAIN FINDINGS AND CONCLUSIONS..........................................................................................6

4. RECOMMENDATIONS ...............................................................................................................14

5. ISSUES FOR THE FUTURE..........................................................................................................19

ANNEX

.................................................................................................................21





iii

---

EXECUTIVE SUMMARY
1.
The objective of this evaluation has been to help improve the effectiveness, efficiency
and cost-effectiveness of GEF operations, especially the activity cycle which is widely regarded
as complex, long and costly. Such problems have arisen largely because GEF policies,
requirements and decision points have been loaded on top of the standard project cycles of each
of the Implementing and Executing Agencies. Although the need to streamline and simplify the
cycle has been highlighted periodically since soon after the GEF began, repeated streamlining
efforts have so far had limited impact.
2.
The evaluation methodology included reviews of key documents, partner and
stakeholder interviews, a stakeholder survey, and exploratory studies within selected partner
agencies of (i) harmonization and simplification opportunities and (ii) alternative delivery
modalities. Field work was undertaken in 17 countries. Existing information systems were
unable to provide reliable data on the time projects spent moving through different phases of the
cycle, and the evaluation therefore designed and assembled a database for the full universe of
GEF projects and proposals (1,926) with the milestone dates of GEF decision points in the cycle.
The data have been grouped according to the Replenishment period of approved projects.
3.
This evaluation has been conducted jointly by evaluation offices of the Implementing
and Executing Agencies, supported by the GEF coordinating units of these agencies as well as
the GEF Secretariat. The Council allocation of US$ 150,000 for this evaluation was matched by
generous contributions in kind, in contracts and in funding of the partners in the evaluation,
which enabled a considerable deeper scope than anticipated. Special mention should be made of
the active involvement of the evaluation units of ADB, UNEP, UNIDO, UNDP and the World
Bank through field visits and specific technical work. The evaluation has relied heavily on full
agreement between partners through a discussion of methodology and findings in a series of
workshops. A workshop with all key stakeholders in New York City in September 2006
provided input to possible solutions and comments on the draft report.
4.
The findings are strongly interlinked. It is often unclear what is cause and what is effect.
Is the cycle not effective because it is not efficient? Or is it not efficient because it is not
effective? In other words: are delays caused by the inability of the system to take timely
decisions, or is the system unable to take timely decisions because of the delays? Is the GEF "out
of date" because the cycle is ineffective and inefficient, or is the cycle ineffective and inefficient
because the GEF is out of date? If one cause should be singled out, the evaluation finds that
disclosure of information and transparency in the GEF has been uneven both to
management and to stakeholders. This has been noted in many other evaluations and in OPS3 ­
for the activity cycle it has meant that no reliable overview existed of what was actually
happening and what this meant for the efficiency and effectiveness of the cycle. However, rather
than going into difficult debates about ultimate causes, the evaluation presents the picture of the
cycle as it was on the closing date for the evaluation data on the first of January 2006.
5.
The evaluation reached the following conclusions :
(a)
The GEF activity cycle is not effective and the situation has grown worse

1

---

(b)
The GEF activity cycle is not efficient and the situation has grown worse
(c)
The GEF activity cycle is not cost-effective
(d)
The GEF modalities have not made full use of the trends in its Agencies and
partner countries towards new forms of collaboration, fostering ownership and
promoting flexibility, efficiency and results.
6.
The GEF is not effective in leading projects through the full activity cycle in a timely
manner. As a result, the proportion of new ideas in each GEF Replenishment period has
decreased, with a backlog of older project proposals in the pipeline.
7.
The evaluation finds that the average length of the activity cycle ­ the time taken for a
project to be identified, prepared, approved and launched ­ increased for projects approved
during each of the last three GEF Replenishment periods. The main growth in elapsed time is
found before project approval, for concept review, formulation and appraisal. Delays in
processing GEF projects are primarily due to structural and institutional constraints related to
GEF complexity, including poor connections between the time-bound GEF decision points and
the Agency cycles, and additional steps and procedural requirements. Elapsed time is also
affected by project-specific circumstances and the Agency project cycles, the nature of the GEF
projects and local circumstances, but these cannot consistently account for the overall increase in
elapsed time.
8.
Given the relatively long GEF cycle, the evaluation considered the possibility that longer
preparation times were resulting in higher-quality projects. The analysis shows no relationship
between the time spent by project proposals in the activity cycle and subsequent performance
ratings either during project implementation or after project completion. This means that the
additional time proposals spend going through the GEF activity cycle does not lead to more
successful projects. Furthermore, the GEF internal cost-effectiveness is decreasing, since the
cycle now takes more time, money and effort than it used to, with lower effectiveness.
9.
The most significant GEF modalities are full-size and medium-size projects, with
associated Project Development Facilities (PDF). The GEF has seen a proliferation of new types
of modalities, which is linked to the inability of the regular activity cycle to respond efficiently
to different needs. There is a need for greater clarity by the GEF in terms, definitions, application
and policies regarding different types of projects and modalities. There is considerable demand
from countries for a long term vision and programming for the GEF, including for
(a) programmatic frameworks, (b) umbrella projects and (c) tranched and phased projects. The
evaluation finds that these are useful tools that should be pursued in a more systematic and
coherent manner.
10.
The cycle management of the GEF is lagging behind international good practice. The
success of the GEF-4 replenishment period and the RAF will depend, in part, on the mechanisms
that are developed to enable the GEF to provide timely support. Fortunately, recent
developments lay a foundation for a new way of doing business, including increased awareness
of environmental issues, capacity within the GEF partnership in addressing GEF concerns and

2

---

managing projects, emphasis on oversight and national ownership, and increased co-funding that
changes the leading role of the GEF.
11.
The evaluation provides the following overarching recommendation:
(a)
no easy fix will improve the activity cycle ­ what is needed is a radical re-
drawing of the cycle, maintaining the quality and attributes for GEF funding;
This redrawing of the cycle should include the following specific
recommendations:
(b)
a shift towards Results Based Management will ensure quality during
implementation and enable a dramatic reduction of detailed "blue print"
information that is currently required in the formulation and appraisal stages of
the activity cycle ;
(c)
the identification phase should be kept to a minimum of establishing project
eligibility, whether resources are in principle available and whether the concept is
endorsed by recipient countries;
(d)
the Work Program as presented to Council should move towards the strategic
level; and
(e)
fully documented project proposals should be endorsed by the CEO on a rolling
basis.
12.
This reflects a re-think of the cycle with the overarching goal of keeping it short and
increasing transparency, predictability as well as decreasing transaction costs, as envisaged in the
GEF Instrument. In addition, the evaluation also recommends closer observation of trends in
international harmonization and simplification by the GEF in future, especially in the context of
changing roles with the introduction of the Resource Allocation Framework. Of specific
relevance to the cycle and modalities and the GEF innovative and catalytic role, the evaluation
finds that that cost-effectiveness of project preparation funding requires further analysis, and that
the GEF should examine the effectiveness of a country based pilot program that uses sector
based approach to enable GEF to assess whether there are advantages to pursuing this approach
in the future.
13.
The full evaluation report containing the detailed analysis is published on the website of
the GEF Evaluation Office and can be found at www.thegef.org (choose Evaluation Office, On-
going Evaluations, and then Joint Evaluation of the GEF Activity Cycle). Technical papers for
the evaluation will also be published on the website.

3

---

1.
BACKGROUND
14.
In support of its mission to achieve global environmental benefits, the Global
Environment Facility (GEF) has committed over $6 billion in grants to more than 1,800
approved projects in 140 countries since 1992. This includes $5,537 million for 716 full-size
projects (FSPs); $267million for 326 medium-size projects (MSPs); and $330 million to 821
approved Enabling Activities. Additionally, preparatory resources worth almost $90 million have
been allocated to develop full- and medium size proposals that have not received approval yet.
15.
From the outset, the GEF has operated with three Implementing Agencies (IAs): the
World Bank, the UN Development Programme (UNDP) and the UN Environment Programme
(UNEP). In 1999 the GEF Council gave seven other agencies direct access to GEF resources
with the status of Executing Agencies (ExAs)1. In 2003 the GEF Council approved the current
arrangement whereby those ExAs with direct access can submit proposals to the GEF Secretariat,
while ExAs with indirect access can submit proposals only through one of the three IAs.
The GEF Activity Cycle and Modalities
16.
The bulk of GEF support has so far been provided through projects, based on
submissions of proposals from countries through the IAs and ExAs (hereafter called Agencies).
The identification, preparation and implementation of GEF projects takes account of GEF
criteria and policies, GEF and Agency policies and procedures, the GEF Scientific and Technical
Advisory Panel (STAP), global environmental conventions, and national needs and priorities.
17.
In practice, GEF requirements are superimposed on the standard project cycles of each of
the Agencies. These cycles have five common phases: concept development, preparation;
appraisal; approval and supervision; and completion and evaluation. In addition, projects will go
through an approval process in recipient countries.
18.
The most significant GEF modalities (i.e., vehicles for the disbursement of funds) are
full-size and medium-size projects, with associated Project Development Facilities (PDF). Other
modalities include enabling activities and variations such as the national capacity self-
assessments, programmatic approaches, targeted research, umbrella projects, phased and
tranched projects, as well as project variations funded by special funds.
19.
The GEF activity cycle is widely regarded as complex, long and costly. Almost since the
GEF began, the need to streamline and simplify the cycle has been highlighted by numerous
evaluations, by the overall performance studies, by the GEF Council and by many of the GEF's
partners and stakeholders. Recent GEF replenishment negotiations emphasized that the GEF
should be "...making its processes more expeditious, streamlined and efficient...". The proofs
of underperformance presented to Council in many evaluations and other documents, were
mainly partial, and partners therefore expressed a need to better understand the underlying
causes. As a result, this evaluation was welcomed by all partners in the GEF to present the full
overview. The idea for the evaluation came from an Executing Agency, and was subsequently
funded as a special initiative by the GEF Council as well as supported by the GEF partner
agencies as a joint effort.


4

---

2.
SCOPE AND METHODOLOGY
20.
The objective of this evaluation is to help improve the effectiveness, efficiency and cost-
effectiveness of GEF operations. The evaluation aimed to (i) identify and analyze the strengths
and weaknesses in the GEF activity cycle and modalities; (ii) identify the constraints that need to
be addressed in order to improve the efficiency of GEF procedures, operations and systems; and
(iii) make recommendations to contribute to simplifying GEF operations.
21.
The evaluation has given particular attention to two areas of concern: first, the early
phases of the activity cycle, from concept identification through preparation, appraisal and
approval as far as project start-up; second, the full-size (FSP) and medium-size projects (MSP),
which absorb most of the financial resources. The parties also agreed to focus in-depth analysis
on projects approved in the GEF3 Replenishment period, closed projects from GEF2 and all
jointly implemented projects, as these are most recent and relevant, with reliable data. The
evaluation has also analyzed the impact on the cycle of the GEF's increase in scope and
complexity over time, which generally corresponds to GEF Replenishment periods.
22.
The evaluation methodology included revie ws of key documents (including the policies
and regulations of the GEF, the Agencies, as well as previous evaluations), partner and
stakeholder interviews, a stakeholder survey, and exploratory studies within selected partner
agencies of (i) harmonization and simplification opportunities and (ii) alternative delivery
modalities. Field work was undertaken in 17 countries.
23.
Existing GEF and Agency information systems were unable to provide reliable data on
the time projects spent moving through different pha ses of the activity cycle, a basic information
requirement for this evaluation. To develop usable data as a basis for analysis, the evaluation
therefore designed and assembled a database reflecting the situation in the GEF on 1 January
2006 for the full universe of GEF projects and proposals (1,926) with basic project parameters.
GEF projects do not all follow the same trajectory in the cycle. While all projects are approved at
Work Program Entry, the point of origin of project ideas is not available. The database therefore
captures the milestone dates of GEF decision points as projects progress from, where available,
pre-pipeline identification, PDFA Approval, Pipeline Entry or PDFB approval to project
effectiveness (i.e., their launch).
24.
The data concerning elapsed times, effectiveness of projects moving through the cycle,
value added at the various stages, have been grouped according to the Replenishment period in
which they were approved. This perspective is justified on two grounds: first of all, the cycle
differs per replenishment period in steps, requirements and criteria. Secondly, each
replenishment period has its own specific policy goals, within the framework of the overarching
goals of the GEF as set by the Instrument. In other words: the Council and the CEO need to be
able to see how old project proposals are that they are asked to approve, how these proposals
went through their respective formulation phases, how they added value on the way and how
they fit into the goals of the current cycle. This perspective is more relevant to Council and CEO
than a perspective which would start from the "birth" of project ideas.
25.
The evaluation considered the possibility that preparation times may result in higher-
quality projects. Quality is difficult to measure as GEF proposals are being developed, however,

5

---

especially as there are no systematic or quantitative mechanisms for quality assurance during the
project development process. The evaluation team therefore used the application of GEF
Operational Principle s as a proxy for project quality.2
26.
Led by the GEF Evaluation Office, this evaluation has been conducted jointly by
evaluation offices of the Implementing and Executing Agencies, supported by the GEF
coordinating units of these agencies as well as the GEF Secretariat. The GEF Evaluation Office
has also conducted a parallel evaluation to assess the experience of the seven ExAs in GEF
cooperation and project development and implementation (see Council paper).
3.
MAIN FINDINGS AND CONCLUSIONS
27.
Before presenting the substance of the findings, it is noted that this evaluation will not
identify one primary cause or party responsible for the underperformance of the activity cycle.
The underperformance turned out to be caused by a multifaceted set of issues, linked to a
complex series of events and involving many if not all actors in the GEF. There is no scapegoat
and no quick fix.
28.
Considering the seriousness of the findings below on elapsed time and its negative
effects, it may be surprising that action has not been taken previously to remedy the situation.
The evaluation finds that disclosure of information and transparency in the GEF has been
uneven both to management and to stakeholders. The GEF information management systems
have not been reliable to generate information on project status and elapsed time, and reporting
on this subject has not been systematic or fully transparent. Hence, it has been difficult for
stakeholders to do anything but complain about the complexity in an uninformed way, and
impossible to ascertain accountability for delays and negative effects. There are clearly
significant opportunities to expedite the activity cycle by sharing information on projects under
preparation in a more consistent and timely way, yet there is no certainty that this is in the
process of being adequately addressed.
29.
Based on the 2004 APR and the Country Portfolio Evaluation, the Council reiterated, in
June 2006, its decision of June 2005 that "the transparency of the GEF project approval process
should be increased" and requested the GEF Secretariat to reinforce its efforts to improve this
transparency. The GEF Secretariat was also requested to take steps to improve the information
mechanisms in the GEF to make essential operational information available at the national level.
These recommendations remain urgent. The main areas where transparency is lacking include:
(a) key GEF policies, strategies and criteria of programming, and (b) Project management
tracking of project progress and status. Also, transparency on operational policies is missing,
especially on GEF eligibility and procedures. For example, access to the GEF Operational
Manual remains limited to the GEF Secretariat.
30.
The findings are strongly interlinked. If the cycle is not effective in producing new
projects, it is by definition inefficient for the projects that were dropped or cancelled along the
route to approval or project completion. However, cause and effect are by no means certain. Is
the cycle ineffective because it is inefficient? Or is it inefficient because it is ineffective? Is there
a lack of value added because of the inefficiencies in the cycle, or is the lack of value added a
root cause for the inefficiencies? Is the GEF "out of date" because the cycle is ineffective and

6

---

inefficient, or is the cycle ineffective and inefficient because the GEF is out of date? It is
important to state that the evaluation did not find any significant causal relationships between the
four areas of cycle effectiveness, efficiency, cost-effectiveness and modalitie s. There is no single
key factor that emerges as the reason why the cycle is ineffective, inefficient, not cost-effective
and producing projects that are not in line with the most modern modalities that the
Implementing and Executing Agencies of the GEF employ. Rather, there are many mutually
reinforcing factors that together add up to the cycle as it currently is. It is the sum of the parts
that leads to serious concerns.
1.
The GEF activity cycle is not effective and the situation has grown worse
31.
For the purposes of this Evaluation, the objective of the activity cycle is to produce
projects, preferably good projects in a timely manner. A cycle can be considered effective if it
achieves this objective; and its various phases produce their respective outputs such as concepts
for the identification phase and project documents by the development phase.
32.
The GEF cycle is not effective in producing projects in a timely manner. At each
cycle phase, outputs are either not produced, or the GEF takes a long time in reaching a decision
to clear the project to move to the next phase (Chart 1 in the Annex).3 This has implications for
the age of the GEF portfolio. For example, 46% of the full-size project proposals that have
entered the pipeline since 1992 have yet to begin, reflecting that projects can take up to three
years from concept to project start-up. Twenty-five percent of the projects that have recorded
pipeline entry dates in GEF-1, before 1999, are still active.
33.
The proposals
that are presented for
Table 1. Cumulative proposals and approvals per GEF Replenishment Period4
approval during a
replenishment period


GEF Pipeline



are increasingly from
GEF Replenishment Period
GEF-1
GEF-2
GEF-3
Totals
%
an earlier period.
GEF-1
proposals
62


62
100%
Table 1 shows that 27
(1995-98)
approvals
35


35
56%
project proposals
GEF-2
proposals
27
271

298
100%
dating from GEF-1
(1999-02)
approvals
16
105

121
41%
came up for approval
GEF-3
proposals
11
166
320
497
100%
in GEF-2. Sixteen of
(2003-06)
approvals
9
132
97
238
48%
them were approved in
that period and 11 were carried over into GEF- 3, in which period 166 projects dating from GEF-
2 were still under consideration. Although a high number of new ideas were entered into the
pipeline during GEF-3 (320 project proposals), the majority of approvals in this period (132)
were dating from GEF-2. The approval rate of GEF-3 shows an improvement compared to GEF-
2, from 41 % for GEF-2 to 48 % for GEF-3, but this is due to the high level of approvals in GEF-
3 for GEF-2 proposals. A large number of projects (259) still await approval and could pre-
determine the early approvals in GEF4, which would contain 2 left-over ideas from GEF-1, 34
from GEF-2 and no less than 223 from GEF-3. This backlog must have an effect on the
innovative and catalytic nature of the GEF.

7

---

34.
The activity cycle is becoming less
and less effective in timely producing new
Table 2. Proportion of Proposals Approved within the
ideas for implementation. Table 2 shows that
same Replenishment Period
the proportion of new ideas in each


Totals
%
replenishment period has decreased. Whereas
GEF-1
proposals
62
100%
in the GEF-1 period 56 % of approvals

approved in GEF-1
35
56%
concerned new ideas (35 out of a total of 62
GEF-2
proposals
298
100%
proposals), in the GEF-2 35 % concerned new

approved in GEF-2
105
35%
ideas (105 out of a total of 298 proposals, and
GEF-3
proposals
497
100%
in GEF-3 the proportion of new versus all

approved in GEF-3
97
19%
proposals has gone down to less than 20 % (97 out of 497). The proportion of new ideas
approved versus new ideas proposed also decreases over time, from 56 % in GEF-1 (35 of 62) to
39 % in GEF-2 (105 of 271) to 30 % in GEF-3 (97 of 320). The GEF-3 percentages will improve
slightly over time, but will not reach the level of GEF-2.
35.
The GEF is not effective in leading projects through the full activity cycle. After 16 years
of GEF, 210 full-size-projects are recorded as complete, i.e. the ratio of completed projects is
16 % of all 1292 full-size project proposals. The completion rate for earlier GEF replenishment
periods is, of course, higher (for example, 46% for GEF1 full-size-projects). While there are no
established standards for completion rates of a portfolio, it implies that there remains a relatively
limited pool of completed projects from which lessons learned can be generated and impact can
be expected.
36.
The evaluation finds that the average elapsed time during implementation is not a major
cause of concern. For 191 closed GEF full-size-projects, the expected duration for
implementation was 47 months (4 years), with an average overrun of 9.2 months. The
implementation periods are, however, not commensurate with the preparatory phases. When
considering the entire lifespan of the closed projects from Pipeline entry to actual closing, 43%
of the projects' life span was spent in pre-implementation (that is, being prepared).
37.
Not all projects in the cycle will finish. A total of 238 projects and proposals have been
dropped, aborted or cancelled, i.e. the ratio of rejected proposals is 18% of the full-size-project
proposals. Fifty full-size projects have been cancelled during implementation (3% of all
proposals) for various project-specific and justifiable reasons In one sense, a certain proportion
of dropped and cancelled projects is to be expected if the project is a risky undertaking, and
could be a sign of cycle effectiveness in weeding out undesirable projects. However, the
evaluation finds that the length of the cycle stages until project start leaves GEF proposals more
vulnerable to changing circumstances and priorities. For example, 288 of the project rejections,
(79%) occurred before Pipeline entry, which seem higher than common practice. PDF resources
worth almost $16 million were allocated to proposals that were subsequently dropped or aborted
before approval.
38.
The effectiveness of the cycle must also be considered in light of its underlying
objectives that the cycle should be producing good projects. The GEF invests considerable effort
and funds into the development of proposals, through PDF-A, B and C funding, numerous
checkpoints for appraisal and approval, with the expectation that projects entering the pipeline

8

---

should have a reasonable chance of approval. It does not operate under a foundation model with
open and transparent competition for funding proposals, and a consequently low rate of approval.
39.
What would a reasonable chance of implementation be? The fourth Replenishment policy
recommendations accepted the possibility that 25 % of its projects would not perform
satisfactory. However, it would seem reasonable that the norm for satisfactory outcomes could
be translated into a similar norm for the activity cycle: 75 % of project proposals should lead to
implementation. Currently, less than 40% of proposed projects have started implementation by
January 2006. Future changes are uncertain given the new pipeline management mechanisms
under GEF-4.
2.
The GEF activity cycle is not efficient and the situation has grown worse

40.
The evaluation finds that the average length of the activity cycle ­ the time taken for a
project to be identified, prepared, approved and launched ­ increased for projects approved
during each of the last three GEF Replenishment periods.
41.
Given the long queue of projects being processed at every stage of the Cycle, a
significant number of GEF projects are moving slowly
Table 3. PDFA to Effectiveness
through the cycle. This trend becomes more pronounced when
projects use GEF preparatory resources. Full-size projects
GEF Phase
Months
Projects
approved during GEF-1 took an average of 36 months to
GEF1
36
17
move through the full cycle from approval for PDF-A funding
GEF2
50
15
for Concept development until project launch ("project
GEF3
66
12
effectiveness"). This already lengthy preparation time
increased to 50 months for GEF-2 projects and to 66 months for GEF-3 projects (see Table 3
and Chart 2 in the Annex).
42.
This trend is well-known, although not in quantified form, as per the Survey conducted
for this evaluation, 74% of 289 stakeholders feel that the GEF cycle duration compares
unfavorably with other donors. Agencies try to reduce
elapsed time by skipping the PDF-A phase and entering
Table 4. Pipeline to Effectiveness
proposals directly into the pipeline. The relevant period is
GEF Phase
Months
Projects
then `Pipeline Entry to Effectiveness', which rose from 37
GEF1
37
36
months for GEF1 to 42 months for GEF3 (3.5 years) (see
GEF2
39
90
Table 4). The main growth in elapsed time is found before
GEF3
42
110
project approval, for concept review, formulation and appraisal. The overall elapsed time for
GEF-3 is deflated, because Agencies use, of course, other sources or their own time and energy
to substitute for the PDF-A phase. The GEF does not record the dates for concept development
by the Agencies or project proponents without PDF funding. If, however, one assumes an
optimistic figure of no more than five months for concept development without PDF-A (i.e. less
than PDF-A time in GEF-1), the average time to Effectiveness in GEF-3 would rise from 42 to
47 months.

9

---

43.
However, many projects approved in GEF-3 have not
Table 5. Updated Average (estimate):
yet completed the cyc le to project effectiveness. The figures
Pipeline to Effectiveness
in Tables 3 and 4 above present the situation of elapsed time
GEF Phase
Months
Projects
as of January 2006. When taking account of elapsed time
GEF1
37
36
until 1 October 2006 for 90 projects that are still waiting final
approval for effectiveness, the updated estimates of average
GEF2
39
95
time from Pipeline Entry to Effectiveness rises to 44 months
GEF3
44
200
for projects approved during GEF-3 time (see Table 5). This is a low estimate; many approved
GEF-3 projects have not yet reached effectiveness, which means that time for them continues to
pass. The times for GEF-1 and 2 however, remain constant. Again, including a concept
development phase without PDF-A would bring the average time to Effectiveness in GEF-3 from
44 up to 49 months.
44.
In the case of MSPs, the total project cost of approved MSPs (up to $1 million) is
roughly 8% of FSPs, while the preparation time averages 60% of that of FSPs. This implies that
each dollar committed to an MSP requires four times the preparation effort of an FSP. Not
surprisingly, this has discouraged Agency staff as well as many country stakeholders from
pursuing this type of project, despite indications by the 2001 MSP evaluation that they generate
positive impacts.
45.
The time for approved projects is `path dependent', which means that it depends on the
process by which each project gets to various decision points. The evaluation notes that elapsed
time is exacerbated due to the shifting, and often increased, GEF-requirements over successive
periods which cause `delayed' proposals having to be further re-processed so that they may fit
into new requirements. Delays in processing GEF projects are primarily due to structural and
institutional constraints related to GEF complexity:
· The progress of projects through the GEF activity cycle has been impeded in a variety of
ways as the GEF has become more complex. The GEF has seen a broadening of the
network of diverse stakeholders; an increase of additional cycle phases, steps and
requirements for projects; and growth in the number of modalities used by the GEF from
one main modality to more than fourteen; the introduction of new substantive dimensions
such as focal areas and strategic priorities; an a constant evolution of interpretation of
definitions of key concepts.

· Duplication and lack of synchronization in the cycle stand out as the most important
factor in elapsed time. Poor connections between the time-bound GEF decision points
and the Agency cycles are a major cause of delays and inefficiencies. The GEF steps of
concept phase, PDF formulation and approval, and Council and CEO approval steps are
additional and disruptive to the flow of the Agency regular cycles. The frontloading of
GEF design requirements in the cycle compounds the disconnect, and is accompanied by
repetition and efforts to fit these elements into design later.

· GEF procedures represent significant add-ons to the Agencies' existing project cycles,
such as cofinancing letters, analysis of incremental costs, specific formats and
summaries, the GEF operational principles, additional reviews, etc. Consequently, GEF
projects take longer than IA and ExA standards in all phases before the project starts.

10

---

· Project proponents have to navigate past a chain of gatekeepers in order to have their
projects approved, with proposals frequently returned for reformulation. Not all of these
gatekeepers apply the frequently changing GEF principles and policies in a consistent
manner, resulting in interruptions to the activity cycle with little gain. The project
preparation process is therefore iterative and inconsistent, with the proponents of
proposals facing considerable uncertainty. Project appraisal is subject to delays due to
multiple, and often duplicative, reviews necessitating extensive discussions and
correspondence on eligibility, design and content of project document56. In spite of the
rigorous requirements and increasing appraisals, projects still arrive at Council to be
rejected or subject to additional reformulation.

· Elapsed time is also affected by project-specific circumstances and the Agency project
cycles, the nature of the GEF projects and local circumstances, but these cannot
consistently account for the overall increase in elapsed time. In short, some outlier
projects take a long time while others take a short time. There is potential to gain
efficiency; for example, 17 projects pipelined in 2005 were approved in 2006, with an
average of 8 months. Any inefficiency in an Agency cycle step tends to balance out over
time by compensating strengths in other steps of the cycle, so that no internal Agency
cycle step is a consistent bottleneck. Similarly, there is no dominant pattern in linkages
between elapsed time and complexity in focal area, regions, country circumstance or
budget size.

· Compounding these factors, the evaluation team notes a significant lack of trust between
the GEF partners, notably the GEF Secretariat, the Agencies and the Council. While
some degree of mutual skepticism might have been expected initially in such a novel and
complex undertaking as the GEF, it seems regrettable that such mistrust persists after a
decade and a half. Despite all of the detailed policies and procedures put in place,
together with the evident growth in capacity of the Agencies and GEFSEC over this
period, duplicate reviewing and micromanagement still appear to take place throughout
the system.

3.
The GEF activity cycle is not cost-effective

46.
Given the relatively long GEF activity cycle, the evaluation considered the possibility
that longer preparation times were resulting in higher-quality projects, i.e. well-designed and
highly successful. The evaluation scope did not entail a full cost-effectiveness analysis that
compares relative expenditure versus outcomes associated with two or more courses of action.
47.
Longer preparation time has not resulted in better projects. Are the delays and
efforts for GEF projects justified and acceptable because projects are of corresponding high
quality? The analysis shows no relationship between the time spent by project proposals in the
activity cycle and subsequent performance ratings either during project implementation or after
project completion. This has two implications: (a) there is no validation that weak proposals take
longer to formulate and cause delays; and (b) the additional time proposals spend going through
the GEF activity cycle does not lead to more successful projects within the GEF portfolio. The

11

---

long periods of preparation, appraisal and approval cannot be considered cost-effective if they
make no notable difference to performance.
48.
In comparison to quality
Box: WB portfolio performance indicators, FY 2006:
dimensions for other similar projects,
· Projects at risk: 12% for GEF and 14% Bank-wide.
the analysis suggests that the additional
· IEG ratings of satisfactory outcomes: 86% for closed
GEF documentation, review and
GEF projects assessed between 2003 and 2006, and
approval requirements do not add to the
82% Bank-wide.
quality of the portfolio. For example,
· The likelihood of sustainability: 71% for closed GEF
projects assessed between 2003 and 2006, and 85%
information from the GEF World Bank
for Bank-wide.
unit on portfolio performance indicators
Source: GEF/World Bank
for the World Bank's GEF program are
generally at the same levels as for IBRD/IDA (see Box).

49.
Furthermore, the GEF internal cost-effectiveness is decreasing, since the cycle now takes
more time and effort than it used to, with similar budgets, results and scope. There is room for
gains in achieving better or the same results with less resources in terms of money and time.
50.
GEF projects continue to experience the same design and implementation challenges as
other aid projects. For example, past PPRs and GEF Annual Performance Reports indicated
several issues regarding project formulation, including over-ambitious and complex design;
failure to sufficiently assess the underlying problem or risks; and weak planning for
sustainability and replication. The 2004 Program Study for International Waters found that
"Inadequate project design has been a problem cited in a number of project midterm and final
evaluations". The 2005 APR established that 58% of projects comply with the Council
expectations on M&E arrangements at the point of CEO Endorsement. There are also examples
of projects proposed for Work Program Entry which appear outside the expected technical area
or comparative advantage of a particular Agency.
51.
Qualitative assessments show that a considerable energy is spent on obtaining quality on
paper but with limited added value in substantive terms. Such "paper evidence" includes the
annex on Incremental Cost Analysis (ICA), reviewed by the ICA evaluation and the quest for
cofinancing letters. The evaluation visited several project proponents who had obtained 22 or
more letters ­ and still the projects were not approved. The field visits heard that "Some
comments received for the proposal and project appraisal from the GEF Secretariat were more
related to the writing style and language and not to the content or substance of the proposal
(Laos, Senegal)", or of "ideas received are often completely out of place" (Slovakia, Egypt). A
universal complaint is that the focus on correct GEF language calls for the use of external
experts, by some called "GEF gurus", which represents a barrier to quality elements of national
ownership and drivenness. A Mexican stakeholder expressed that "It seems that GEF projects
...have to go back and forth to get the right words".
52.
The cycle cost-effectiveness is further reduced by the fact that the cycle delays tend to
cause a number of negative effects. One observation made by a survey respondent captures a
widely-held view: "as the rules became stricter, the stages from concept development, project
preparation and project appraisal tend to drag, resulting in the withdrawal of good proposals by
proponents who could not afford to wait, and lost opportunities for government ownership". The

12

---

long process in formulation until approval often reduces the quality of the project by making it
outdated by project start. The GEF procedures on re-submission in case of changes discourage
redesign to secure project quality. One result of delays in appraisal and approva l is the gap,
often up to 18 months, between the completion of PDF-financed project preparation and the
beginning of implementation, with a critical disruption for project staff in recipient countries, as
the GEF does not permit the use of resources after agency approval until project effectiveness.
53.
The GEF Council has responded to growing concerns about the length and complexity of
the activity cycle by encouraging simplification, better coordination and the imposition of strict
time limits. However, the Council has not so far indicated that the technical standards of project
preparation, appraisal and approval should be relaxed or the barriers to entry lowered to offset
the increasing complexity of and demands on the cycle.
4.
The GEF modalities have not made full use of the trends in its Agencies and
partner countries towards new forms of collaboration, fostering ownership and promoting
flexibility, efficiency and results.
54.
GEF has seen a recent proliferation of new types of modalities, including special funds
as well as new and overlapping terms and practices for existing modalities. This growth is
linked to the inability of the regular activity cycle to respond efficiently and flexibly to different
needs. This has led to confusion among stakeholders, misunderstandings between partners, and
concern of the Council. These concerns apply in particular to programmatic approaches,
umbrella projects, the targeted portfolio approach, corporate programs, phased and tranched
projects, country programs and programming frameworks, subprojects, partnership approaches,
financial mechanisms and disbursement mechanisms. There is a need for greater clarity by the
GEF in terms, definitions, application and policies regarding different types of projects and
modalities.
55.
There is considerable untapped demand from countries for a long term vision and
programming for the GEF, in line with the nature of global environmental benefits, the RAF,
donor harmonization practices and country priorities. Country visits and stakeholder
consultations reflected two main implications for modalities for such support: (a) programmatic
frameworks, (b) umbrella projects and (c) tranched and phased projects. The evaluation finds
that these are useful tools that should be pursued in a more systematic and coherent manner.
Lessons learned from Previous Efforts to Streamline and Simplify
56.
Streamlining efforts have had limited impact. The GEF's growing complexity has not
been mitigated by the discontinuation of any significant steps or requirements in the activity
cycle. The analysis shows that virtually none of the several attempts made since 1998 to reform
and simplify GEF procedures have made a notable difference in expediting the activity cycle.
The CEO endorsement for MSPs has shortened the MSP cycle, which, however, still remains
long compared to Agency cycles for similar and larger projects. Evaluations, most recently the
2005 OPS3 and the 2005 APR, continue to highlight concerns on the time for a GEF project to
begin implementation. This suggests that the potential time savings to be gained by refining
current procedures have limited prospects for delivering significant improvements without more
fundamental changes in the way that GEF does business.

13

---

57.
It is now clear that many of the expectations and claims made by earlier GEF cycle
reform efforts within both GEFSEC and the Agencies were not based on a full appreciation of
the underlying problems. As a result, their expectations tended to be over-optimistic and
underlying institutional incentives given relatively little attention. Most efforts towards
streamlining in the GEF have resulted in additional requirements designed to mitigate the
negative impacts of existing requirements.
4.
RECOMMENDATIONS

Recommendation 1: No easy fix will improve the activ ity cycle ­ what is needed is a radical re -
drawing of the cycle, maintaining the quality and attributes for GEF funding

58.
The excessive length of the GEF activity cycle has left all stakeholders frustrated while
eroding the credibility of the GEF as an attractive partner to work with to support the global
environment. Moreover, there is a perception among partners that the situation is deteriorating,
and disappointment that attempts at remedies have failed in the past. The cycle management of
the GEF is lagging behind international good practice, and ultimately impedes the achievement
of impact since it is taking longer and longer to make projects operational on the ground. In an
increasingly competitive environment for resources, public sector agencies need to clearly
demonstrate important development results with positive effects to decision-makers and
beneficiaries. The success of the GEF-4 replenishment period and the RAF will depend, in part,
on the mechanisms that are developed to enable the GEF to provide timely support.
59.
Were the GEF activity cycle to be developed from scratch today, it seems inconceivable
that anything resembling the current system would be proposed. The time for adjustments or fine
tuning has passed. Now there is a need for an overhaul, to "wipe the slate clean" and to re-think
the cycle with the overarching goal of keeping it short and increasing transparency, predictability
as well as decreasing transaction costs. The GEF needs sufficient flexibility to address the
changed context of international cooperation and to support global environmental benefits in a
dynamic manner.
60.
Fortunately, the GEF operational context that guided the design of the activity cycle has
changed since 1991. Several recent developments lay a foundation for a new way of doing
business, and point to solutions that go beyond tinkering with the cycle as it functions today. The
institutional framework now contains mechanisms for oversight and validation, broadened
capacity among partner Agencies, extensive experienc e with GEF project development, a
renewed focus on national ownership and endorsement, and an increased emphasis on targets and
indicators. Furthermore, the increase in co-funding leads to a change of the GEF as the partner
drawing attention to global environmental benefits, to the GEF as a relatively minor financer of
the support, that no longer can insist, as one stakeholder expressed it, on "calling all the shots"
for other partners. All these elements require a fundamentally different approach in the cycle.
61.
The evaluation itself cannot redefine the activity cycle. However, it can formulate the
principles which should guide this effort and it can propose a few key moments in a new cycle.
The following principles can be applied under a banner of overall simplification:

14

---

a. Consistency with the GEF Instrument, section VII on operational modalities. There is
a need to go back to the fundamental intentions behind GEF management, which has
been diluted over time. On GEF projects, the Instrument prescribes endorsement by the
CEO before final project approval provided it is consistent with the Instrument and GEF
policies and procedures.

b. Employing the comparative advantages of the different parts of the GEF system,
including the Council, the Secretariat, the recipient countries, the Agencies, STAP, and
the Evaluation office, as appropriate at the various points in the cycle. The number of
partners and changed roles, increased project and context complexity and increase in
procedural requirements have not only led to duplication of effort, but also left gaps that
need to be filled. The increasing complexity and growing portfolio call for an increased
focus of the Council, as well as the GEF Secretariat, on strategy and policy, portfolio
monitoring and program results verification. Since 1991, the awareness of environmental
issues, capacity in addressing GEF concerns and managing projects have been on the rise
within the GEF partnership. The evaluation has also identified strides in simplification of
Agency systems for operations and identified that the Agencies have policies or
requirements that are compatible with the main GEF Operational Principles. There is
significant scope to utilize "certified" agency systems for operations and design that
would enhance efficiency and effectiveness. The national partners are assuming new
responsibilities for greater ownership and participation. The STAP has proposed reforms
to increase its relevance to project quality. All these partners must be empowered to fulfil
their roles within an environment of trust, transparency and accountability.

c. Working within the emerging Resource Allocation Framework, with a corresponding
deployment of resources in the activity cycle towards the project implementation phase;
as recommended in OPS2, a shift in emphasis from an "approval culture" to a culture of
"quality and results". The overall portfolio could benefit from more programmatic
approaches as requested by countries, which at the same time would reduce the
administrative workload. Results based management is also dependent on strong
partnering around results, and on harmonization efforts to maximize the impact of
assistance. Any changes to the cycle must also fit the needs of all focal areas and regions,
both subject to RAF and on project-by-project approval. The focus on results and country
leadership is also a good opportunity for scale-up and replication of current programmatic
approaches, based on lessons learned.

d. Establishing performance benchmarks for measuring efficiency and effectiveness of
GEF operational policies and procedures, as well as agency score-cards and enforceable
time standards. Such systems of checks and balances, and clear definition of validation
roles should accompany the devolution of responsibilities in the formulation, appraisal
and approval phases. Full transparency is a precondition for performance measurement
systems and accountability for compliance with deadlines; to provide consistent and
comprehensive information that is available to all parties.

e. Ensuring a regular monitoring and clean-up of proposals in the activity cycle which will
make timely decisions to discontinue proposals that are in danger of obstructing the

15

---

pipeline: forever in re-development because additional formulation or information is
sought, or sidelined because the circumstances are not right to move forward, or good
ideas which turn out to be very difficult to operationalize. A regular cleaning up of the
proposals in the cycle will lead to increased ease of flow and quicker decisions on
projects that are well designed. Conversely, the GEF Secretariat or others would not ask
for modifications on proposals more than once so as to avoid causing elapsed time.

f. Allowing scope for proposals that are well embedded in programmatic approaches,
whether national, regional or in focal areas, that ensure that individual projects benefit
from interacting with other projects.

62.
When applying these principles, the following recommendations identify elements that
would allow for a complete restructuring of the activity cycle and modalities.
Recommendation 2: a shift towards Results Based Management will ensure quality during
implementation and enable a dramatic reduction of detailed "blue print" information that
is currently required in the formulation and appraisal stages of the activity cycle

63.
The GEF should deepen its move towards Results Based Management (RBM) that started
by its introduction of the Resource Allocation Framework (RAF), harmonization of the
evaluation function, and the ongoing development of a framework for portfolio monitoring. In
particular, development effectiveness should be pursued (a) directly, for tasks directly relevant to
the GEF Secretariat, by simplifying the framework and the steps of the GEF activity cycle; and
(b) indirectly, by ensuring that the GEF partner agencies are supported in their own
simplification efforts. The aim should be to improve predictability; focus on program-level
outcomes and results; and decrease transaction costs. After introduction of the RAF,
harmonization of the evaluation function and introducing results indicators and portfolio
monitoring, the GEF is well positioned to move to the next level of RBM and thereby
streamlining the cycle, through three main pillars:
· A comprehensive results-based management framework for the GEF to be
implemented in GEF-4 which will incorporate monitoring and reporting at three levels:
corporate, programmatic (focal area) and projects. Delays in project start and
implementation, and compliance with M&E, would also be tracked by the annual
Portfolio Performance Report. Issues best covered by monitoring include cost-
effectiveness, flexibility, participation and ownership, resource mobilization and progress
towards outcomes.
· The application of the Monitoring and Evaluation Policy and the system of
performance measurement provided by the independent GEF Evaluation Office, with
support of the Agency evaluation units. In particular, this system now provides for
systematic conduct and assessment of project evaluations, as well as impact evaluations,
country portfolio evaluations and the review of the focal area GEF-4 strategies that
incorporate all projects. Aspects of M&E quality, project-at-risk systems, and quality-at-
entry are also covered. Issues best covered by evaluation include sustainability,
replication, actual co-financing mobilized and impact.

16

---

· The development of the new Management Information System approved at the
November 2005 Council. From the cycle perspective, a new data system can be simple;
but it needs discipline and consistency with established business practice. In particular, it
is indispensable to track the exact progress of a project through the cycle. It is now time
that the GEF takes full advantage of modern communication opportunities to become
more service-oriented to stakeholders by providing accessible information regarding its
policies and procedures on the GEF website. Although still a work in progress, the Joint
Evaluation data base will be available to facilitate further analysis, with the potential to
provide a relatively low-cost tool to support management oversight of the activity cycle.
64.
In turn, the GEF requirements for project design and content can be revisited and
drastically simplified. Rather than mechanically following detailed design and reporting
requirements, partners will be expected to be more closely involved in strategic choice of M&E
mechanisms; adaptive management; reporting on GEF concerns, and follow-up and learning. In
particular, the evaluation has identified design elements that already form part of the Agencies'
and government's regular project design process: incorporation of lessons learnt, project
consistency with national or other plans and priorities; identification of major stakeholders, and
planning for their involvement, including safeguards for marginal groups; principles and policies
for national ownership, stakeholder participation and disclosure; as well as analysis of likely
sustainability and risks. This would allow the GEF projects to benefit from the capacities of the
Agencies and national project proponents to develop regular non-GEF projects with similar
quality at much faster process.
Recommendation 3: The identification phase should be kept to a minimum of establishing
project eligibility, whether resources are in principle available and whether the concept is
endorsed by recipient countries

65.
The step of identification of GEF assistance is crucial to a smooth project development
process. The difficulty in determining GEF eligibility (or "GEF-ability" as some stakeholders
express it) is a key factor both in delays, in dropped project proposals and in wasted efforts that
could be put to better use for the global environment. The GEF has, after more than a decade,
developed considerable experience in determining whether a project proposal has incremental
components. The knowledge is, however, not easily accessible, and open to individual
interpretation. In effect, proposed projects are subject to (a) incremental justification;
(b) strategic priorities; (c) other preferences applied by the GEF Secretariat to shape the
portfolio; and (d) financial resource limitations. These crucial elements would need to be judged
on relatively limited information, otherwise the identification phase would turn into an appraisal
phase.
66.
The GEF has already started a shift towards emphasizing identification. The underlying
premise of the RAF is country-drivenness in the identification process. The CEO introduced a
Project Identification Form (PIF) in August 2006, to be submitted by the Agencies to the GEF
Secretariat. The proposal must obviously fit with the endorsed pipeline for climate change and
biodiversity projects. The PIF effectively subsumes the need for a Concept Brief for pipeline
entry. Evolution has already overtaken the Pipeline Entry phase, by which the RAF endorsement
process de facto constitutes the pipeline. The same principles could be applied to other focal

17

---

areas. The proposals must, however, be made publicly available for transparency for the Council,
local stakeholders and other partners.
67.
The purpose of an early identification should revert to the original vision of the concept
phase, as discussed by the GEF in 1995, that is, to ascertain whether the proposal is eligible for
GEF financing, no more, no less. The identification should provide sufficient assurance that a
likely good project, in line with global and national priorities, can be developed. It should not
provide assurance that the project is well designed, as it has not been formulated yet. Unless the
GEF exercises restraint in asking for planning details at this early stage of the process, past
experience points to the risk of further delays of later re-design based on reality.
Recommendation 4: The Work Program as presented to Council should move towards the
strategic level
68.
A work program should be presented to Council for consideration on a more strategic
level. As envisaged in the Instrument, the work program should be prepared among the
Secretariat, Implementing Agencies, executing agency, and in cooperation with eligible
recipients. Originally, the focus of the work program for approval by the Council included an
indication of the financial resources, and the principles of cost-effectiveness, country-drivenness
and flexibility (Instrument paragraph 4). The management of financial resources has increased in
importance, and the lack of systematic overview of financial status has caused further delays in
the cycle, as exemplified by the recall of 2006 approvals for lack of liquidity.
69.
Several scenarios are possib le, which could fill voids in portfolio management that are
currently not addressed, and in turn support quality project development. The work program
content could also include an overview of country RAF strategies, eligibility criteria and
priorities for individual projects, lessons learned from the portfolio and update on RAF-4
strategies, co-financing plans, program outcome indicators, brief summary of focal area potential
projects or list of PIFs, or other. As far as proposals are concerned, the work program would
provide information of the composition of the pipeline as approved by the CEO on the basis of
the Project Identification Forms. This would also present a greater opportunity to address how
the cycle generates a portfolio of projects that are mutually supportive, as the `work programme'
approval process could look across PIFs at portfolio level synergies. In other words, the whole
portfolio should be greater than the sum of the individual parts.
Recommendation 5: Fully documented project proposals are endorsed by the CEO on a
rolling basis
70.
It is recommended that GEF projects are subject to a rolling endorsement by the CEO
before final project approval, as envisaged in the Instrument. If the endorsement is smooth, the
Agency would be able to schedule its own approval shortly thereafter. If the principle of full
disclosure of project documentation, reviews and project status is observed, appraisal can be
used applying modern techniques of on-line review on a rolling basis, available to all. Inclusion
of comments in the project document should not be a matter of editorial amendments, but should
be addressed by the Agency and project proponents for the Agency approval, be posted on the
GEF website, to be verified by spot checks and M&E mechanisms later. The CEO should be at

18

---

liberty to submit a project to Council if it is felt that the project raises a sensitive or policy issue,
while trying to ensure that policies are not driven by project proposals.
71.
In line with the principles of results-based management, appraisal should focus on a
sound strategy to achieve results and the management mechanisms to monitor and adapt.
Distortional incentives of promoting increasingly detailed comments could be avoided by non-
objection approvals regarding consistency with GEF policies and procedures. Full and
transparent use of online electronic tools would allow the CEO to inform the Council and other
stakeholders of all proposals submitted for endorsement, thus allowing Council members to
inform themselves on the project documentation and raise any objections if they feel that these
are warranted.
5.
ISSUES FOR THE FUTURE

72.
The roles played by the GEF partners are not only complex but are changing. The GEF
Secretariat has portfolio monitoring responsibilities; STAP is in the process of reform; and the
GEF Evaluation Office became an independent entity in 2003. The roles of countries in GEF
decision making on individual projects are becoming more significant. National Focal Points
must now take on pipeline planning. Donor countries have increasingly required project design
changes before approving projects, while recipient country procedures are becoming
significantly more influential with the introduction of the Resource Allocation Framework. As
countries take a more direct role in setting priorities for and monitoring their overall GEF
resource use, more decisions are being made at a national level, requiring a synchronization of
the GEF activity cycle with national needs and practice. Without dramatic changes in other steps
as recommended above, the cycle can be anticipated to lengthen further as a result.
73.
The GEF has not taken full advantage of learning opportunities on project management
within the international development community. For example, the arrangement of the UN
system for a common country presence and the internal harmonization effort among all UN
agencies at the country level could provide an entry point for the GEF to strengthen country-
level support to the implementation of the RAF. This suggests closer observation of trends in
international harmonization and simplification by the GEF, for example through participation
in the relevant working groups and task forces or through agreements for mutual information
exchange on organizational matters, with active follow-up on procedural changes.
74.
The cost-effectiveness of project preparation funding requires further analysis. The
use of PDF funds to prepare both full-size and medium-size GEF projects has grown in both
relative and absolute terms. A significant proportion of PDFs spend a relatively long period of
time being implemented and GEF information systems are currently unable to either track the
progress (or otherwise) of PDFs through the activity cycle or to measure their results. Thus, the
value of the PDF contribution to developing GEF projects is hard to assess, although the analysis
shows no relationship between the presence of PDF funding and subsequent performance ratings.
The procedures for obtaining PDF funds also cause delays.
75.
While comparisons of projects with and without PDFs do not indicate obvious
advantages for the use of PDFs, if the GEF and Agencies continue to require relatively complex
project documents with frequently changing requirements, there may not be an alternative to

19

---

some form of project development facility in financing project preparation. Not many countries
would find it worthwhile to build up their own capacity and Agencies may naturally be reluctant
to advance funds for this purpose. If on the other hand, the recommended revamp of the cycle
does lighten requirements and shortens the gap between formulation and project start, the needs
for formulation support should change.
76.
While the GEF is the largest environmental fund and the only fund to target incremental
costs for global environmental benefits, it is also mandated to be innovative and catalytic. In this
context, GEF should examine the effectiveness of a country based pilot program that uses sector
based approach to enable GEF to assess whether there are advantages to pursuing this approach
in the future.

20

---

ANNEX


Chart 1: Current Distribution of Full-Size-Proposals in the Activity Cycle by Pipeline-Entry date (1991-2006) - by Proportion

FY for
Total No.
Pipeline PDFB Dropped Approved
CEO
IA
Active Cancelled Complete

Pipeli
Entering
Endorsed
Approved
Phase of Pipeline
ne
Pipeline
Entry
Entry
per FY
Entry date not
available
NA
325
2%
1%
4%
3%
2%
3%
29%
8%
48%



6%
6%
78%






1991
Pilot
18


11%






to
Pipeline
1994






3%
1%
4%
25%
4%
46%






1995
GEF1
72
1%
1%
14%
to






Pipeline
1998






11%
5%
18%
34%
5%
2%






1999
GEF2
319
6%
5%
14%






to
Pipeline
2002






2003
79
11%
27%
6%
14%
8%
19%
11%
4%

2004
76
8%
43%
3%
22%
5%
14%
3%

1%
GEF3
2005
119
14%
71%
1%
10%
2%
2%
1%


Pipeline
2006
51
57%
39%

2%
2%




Status of
Pipelined
100%
8%
17%
7%
8%
3%
9%
22%
5%
20%
Proposals (%)

Status of
Pipelined
Proposals (No.)

1059
86
178
79
87
36
98
235
50
210

21

---

Chart 1. Low Effectiveness:

What does Chart 1 show?
The current position in the Activity Cycle of all 1059 full-size-projects that have entered the Pipeline is reflected in this Chart. The total number of
projects pipelined within each (fiscal) year, is defined at the beginning of each row, for earlier periods these are aggregated these into four year
cohorts. The Chart then provides a measure of the speed of projects while moving horizontally through the Cycle by disaggregating the group of
pipelined projects and categorizing them under their most current status ­ showing how far down the row they have moved since pipeline entry.

The Chart reflects the entire range of project status a given cohort is currently positioned in. It reflects in particular on the management of the
Activity Cycle, on whether decisions to approve a project for the next stage or to weed it out were taken, or whether a project is `languishing'.
Implications of this distribution are discussed in detail in Chapter 4.

Does lack of pipeline entry date change the findings?
The proposals without pipeline entry dates (the first row in the table) capture a substantial proportion of the older GEF Portfolio, these are omitted in
the rest of table due to lack of recorded pipeline entry data. This row represents up to 85% of the Pilot phase, 74% of GEF1, 42% of GEF2 and 6% of
GEF3. The majority (53%, 170 FSPs) are UNDP projects, 85% of which are either complete (92 FSPs) or active (53 FSPs). Another 91 are Bank
projects, 42 of which are complete and 14 active; and 32 are joint IA projects of which 12 are complete and 15 are active. This row illustrates that
the while a larger proportion than described in the Chart are closed, older projects are also still active.





22

---

Chart 2: Time Spent at each Stage as Full-Size-Projects move through the Activity Cycle,

by GEF Replenishment Phase in which Projects entered the Work Program


6.4





5.5 months
3 months
7 months
17 months
2mo
months












4 projects
18 projects
51 projects
123 projects
118 proj
112
proj




ects
Average Time spent by full-size-projects approved in GEF1, in the Activity Cycle


8
5.5
months
1.4 mo
18 months
17 months
1



mo
months












16 projects
88 projects
141 projects
185 projects
179
144 proj


proj
Average Time spent by full-size-projects approved in GEF2, in the Activity Cycle

14
3
31 months
10 months 1.4 2.4
months
months
mo
mo













28 projects
202
204 projects
169 projects
135
110
projects


proj
proj




Average Time spent by full-size-projects approved in GEF3, in the Activity Cycle

LEGEND:




Pipeline to PDFB
PDFB Approval to
Project Approval
CEO
Agency
PDFA Approval
Project Approval
to CEO
Endorsement
Approval to
to Pipeline Entry
Approval
Work Program Entry
Endorsement
to IA Approval
Effectiveness
-
Pipeline Ideas

Effectiveness
Pre


23

---

Chart 2. Activity Cycle Processing Time: An Exposition
· Which projects are reflected in Chart 2, and how is each cohort defined?

The entire universe of 704 full-size-projects approved between FY 1995 and mid-FY2006 has been divided into three cohorts based on the four-year GEF replenishment
phase they were approved in (i.e. their year of Work Program Entry). This follows the traditional classification of GEF projects and most commonly defines the set of
prerequisites they have fulfilled while being processed. These requirements have varied and increased over successive GEF phases. The GEF1 cohort contains 136 projects
(98% are active, cancelled or complete), GEF2 contains 210 (90%) and GEF3 254 projects (24%). The early Pilot Phase (1991-94) has been excluded from the analysis.
· What does the Chart show?
The Chart represents the average calendar time (in months) it has taken projects within each cohort to move from one stage in the Cycle to the next. Based on this
representation, it shows that projects approved in recent years have spent increasing amounts of processing time in the GEF Cycle. The Chart further identifies processing
stages within the Cycle that take more or less time; this is compared with other processing stages and over time. There is indication that, on average, projects are taking longer
in the pre-Approval period (before entering the Work Program), and then perhaps moving faster post-Approval. In particular, the Chart marks out an `orange' zone, the project
preparation-period between PDFB Approval and Work Program Entry which has shown the most significant increase over time per approved project.
· How robust are these trends?
The entire universe of full-size-projects is included here, making for robust representation. Corroborating the vigor of the Chart, these trends have been identical across
Agencies. The increase in time from PDFB Approval to Project Approval in GEF3 is statistically significant. It is noted that some newer projects have not yet moved right
through the entire Cycle and that due to data restrictions dates for a few ­ of the older- projects are not recorded; yet the available processing time for most of the universe of
projects is measured and the number of projects within each phase with complete information is recorded to signal robustness. Over time, as the entire cohort of projects
within each GEF phase becomes effective, it is possible that the declining time -trend in the post-Approval periods may ease-off as slower projects reach effectiveness. The
trend for the pre-Approval period is however, relatively fixed for the GEF cohorts represented in the Chart given that, by definition, all projects within these cohorts have
received approval (a ll projects are reflected here, except for those included in the last June 2005 Work Program), future approved projects will fall under GEF4 replenishment.
· Can the total processing time taken by a project be calculated by adding up each Stage?
The average total processing time across the GEF replenishment cohorts is affected by the number and nature of projects that have completed all the various phases in the
Cycle. Elapsed time cannot therefore accurately be calculated simply by adding the various time -slices. Below are representations of this total time. Projects using Block-A
preparatory funding (12% of projects between GEF1 to GEF3) start the Cycle from PDFA Approval, and the total processing time for these is indicated in the first row (from
PDFA to Effectiveness). The second two rows represent the total time for projects that do not have PDFA funding, but may have accessed Block-B and Block-C funding
(38%, 67% and 80% of GEF1, GEF2 and GEF3 respectively have PDFB components; 2% of all projects have PDFC resources). Row 3 removes projects with unusually high
processing times from the cohorts in Row 2, but projects with unusually low processing times are included, providing estimation of averages that have higher probability.

Total Processing Time Across the GEF Activity Cycle
Total Average (Months)
No. of Projects
Standard deviation (months)
Median(months)

GEF1
GEF2
GEF3 GEF1 GEF2 GEF3
GEF1
GEF2
GEF3
GEF1
GEF2
GEF3
1. PDFA Approval to Effectiveness
36
50
66
17
15
12
23
19
23
40
44
52
2. Pipeline to Effectiveness
37
39
42
36
90
110
22
17
20
37
37
38
3. Pipeline to Effectiveness (without
Outliers)
33
38
41
34
88
108
16
15
19
29
37
40




24

---

1 Asian Development Bank (ADB), African Development Bank (AfDB), European Bank for Reconstruction and Development (EBRD), Inter American Development Bank (IADB),
UN Food and Agriculture Organization (FAO), UN International Fund for Agriculture and Development (IFAD), and UN Industrial Development Organization (UNIDO).
2 These principles cover Incremental Costs for Global Environmental Benefits, Country Ownership, Cost-Effectiveness, Flexibility, Full Disclosure, Public Participation, Catalytic Role
and Monitoring and Evaluation.
3 Based on the 1,059 projects that entered the GEF pipeline from 1991 to 1 January 2006 (including informal GEF-1 pipeline).
4 The Pilot Concepts (17) have been excluded. Only concepts with recorded Pipeline dates are included (325 post-pipeline concepts without dates excluded). Includes concepts that are
currently PDFB (175), Pipeline (82) and Pending (2). Concepts that are pre-Pipeline, or were dropped before WKPE or rejected before Pipeline entry are not included.
5 Several Agencies have highlighted the particular challenges in the timing of appraisal phase, which is especially heavy for GEF projects with comments from GEFSec, other
Agencies, convention secretariats, STAP and the Council, in addition to internal Agency feedback and country comments.


25

---