Investing in our future

The Global FundTo Fight AIDS, Tuberculosis and Malaria

UBERCULOSIS MALARIA HIV/AIDS TUBERCULOSISIS MALARIA HIV/AIDS TUBERCULOSIS MALARIA HARIA HIV/AIDS TUBERCULOSIS MALARIA HIV/AIDS/AIDS TUBERCULOSIS MALARIA HIV/AIDS TUBERC

RIA HIV/AIDS TUBERCULOSIS MALARIA HIV/AIDS TIV/AIDS TUBERCULOSIS MALARIA HIV/AIDS TUBELOSIS MALARIA HIV/AIDS TUBERCULOSIS MALARBERCULOSIS MALARIA HIV/AIDS TUBERCULOSIS M

HIV/AIDS TUBERCULOSIS MALARIA HIV/AIDS TUBELARIA HIV/AIDS TUBERCULOSIS MALARIA HIV/AI

GLOBAL FUND INVESTMENTS INFRAGILE STATES: Early Results

Acknowledgements:

This report was produced by Vinand M. Nantulya, with the assistanceof Blerta Maliqi, Valerie Manda and Mabingue Ngom.

The editorial team included Julie Archer, Beatrice Bernescut, JohnBusch, Jon Lidén, Bernhard Schwartländer and Rosie Vanek.

Design and layout by Art Gecko, artgecko@vtxnet.ch

© 2005 The Global Fund to Fight AIDS, Tuberculosis and Malaria

ISBN 92-9224-027-7

Photo credits:

Cover photos: John Rae, Gideon MendelP. 4 Robert Bourgoing (www.bourgoing.com)P. 6 Dieter Telemans (www.panos.co.uk)P. 8 Keith Lepor (www.keithlepor.comP. 10 Keith Lepor P. 12 Robert BourgoingP. 14 John Rae (www.raephoto.com)

P. 16 John RaeP. 18 Robert BourgoingP. 20 Robert BourgoingP. 22 Keith Lepor P. 23 Giacomo Pirozzi (www.panos.co.uk)P. 24 Robert BourgoingP. 30 courtesy WHO

LIST OF TERMS AND ABBREVIATIONS USED

EXECUTIVE SUMMARY

INTRODUCTION

THE GLOBAL FUND AND FRAGILE STATESStudy ObjectivesStudy DesignStudy LimitationsStudy Methodology

KEY FINDINGSGlobal Fund Investment in Fragile States Grant Performance in Fragile States Principal Recipients in Fragile States Performance by Disease in Fragile States

CONTEXTUAL ISSUES: NAVIGATING STATE FRAGILITY

CONCLUSION

APPENDIX 1: List of Fragile States

APPENDIX 2: Grants Assessed for Phase 2 Funding Through July 2005

REFERENCES

5

7

9

1111121213

1515161719

21

25

26

27

29

3

TABLE OF CONTENTSTABLE OF CONTENTS

To date, the Fund hascommitted US$ 28 millionto grants to combatHIV/AIDS, tuberculosisand malaria in Mali.

4

ACT Artemisinin-based combination therapy

ARV Antiretroviral therapy

CCM Country Coordinating Mechanism

DFID Department for International Development (UK)

FPM Fund Portfolio Manager

GSC Grant Score Cards(internal Global Fund documents outlining grant progress and results)

HBC High-burden country (used in reference to tuberculosis disease burdens)

HIPC Heavily indebted poor countries

LFA Local Fund Agent (independent consultants contracted by The Global Fund to assess and verifyprogram results as they are reported by the Principal Recipients of grants)

LICUS Low-income countries under stress

M&E Monitoring and Evaluation

MDGs Millennium Development Goals

PR Principal Recipient

SIE Strategic Information & Evaluation (unit within the Global Fund Secretariat)

TB Tuberculosis

TERG Technical Evaluation Reference Group

TRP Technical Review Panel

UNAIDS Joint United Nations Programme on HIV/AIDS

WHO World Health Organization

5

LIST OF TERMS ANDABBREVIATIONS USEDLIST OF TERMS AND

ABBREVIATIONS USED

6

Correct diagnosis of tuberculosis is an essential partof combating this epidemic. Over two years, a GlobalFund grant to Uzbekistan will support training ofmore than 3400 health care providers in TB detection.

1. In its first four rounds of funding, the Global Fundhas invested one-third of committed funds in 45 fragilestates, financing a total of 123 programs. This reportpresents initial results from an ongoing study on theperformance of grants in fragile states. The report isbased on a comparative analysis of the performanceof the 19 grants from fragile states and the 55 grantsfrom stable states that had been assessed for Phase 2funding. (Global Fund grant funding is contingent onsatisfactory performance. A full assessment of allgrants is made 15 to 18 months from the start dateof implementation to determine whether fundingshould continue for the remainder of the grant’slifespan). All 19 grants performed well: 14 grantswere graded A or B1 (measures of excellent orsatisfactory performance, respectively), and fivegrants achieved grade B2 (targets not met butdemonstrated potential). None of the 19 grants wasdiscontinued.

2. The performance by grants in fragile states wascomparable to that of the 55 grants implemented in stable states. Most of the grants in fragile states(14 out of 19) were managed by Principal Recipients(PRs) from the government sector, and these grantsperformed equally well as those managed by non-government PRs. A well-performing grant in afragile state was more likely to have a proactiveCountry Coordinating Mechanism (CCM) that metregularly with the PR and sub-recipients to reviewwork plans and progress, and which had regularcommunications with the PR and sub-recipients onthe one hand and the Fund Portfolio Manager at theGlobal Fund on the other hand. Site visits by the CCMwere important for encouraging program implementersand for verifying implementation activities and results.Coordinating the input by development partnersinto one coherent plan of action was critical, as wasthe inclusion of the non-governmental sector inimplementation. Other important factors of successincluded the creativity and initiative of the PR, timelydisbursements and verification of progress reports.

7

SUMMARYEXECUTIVEEXECUTIVE SUMMARY

8

An important component of a grant to the DemocraticRepublic of Congo to combatand prevent tuberculosis issocial mobilization throughoutall layers of society withinformation, education andcommunication activities.

3. State fragility has come to the fore in high-leveldiscussions within donor circles, as evidenced by thedevelopment of several working papers on thesubject in the last two years (1-5, 7-12, 14). The keyconcern is how development assistance can best bedelivered in these constrained environments in orderto achieve the Millennium Development Goals(MDGs). Despite this increased attention, there is noagreed definition of state fragility - most states canbe considered to be fragile in one aspect or another.Moreover, state fragility is not constant: a countrycan move in and out of fragility, making it difficult todraw up a definitive list of fragile states (6, 7). Not tomention that there is substantial stigma attached tobeing described as a fragile state.

4. The UK’s Department for International Development(DFID) has put forward a working definition of a fragilestate as one that “cannot or will not deliver corefunctions to the majority of its people, including thepoor.” The core functions listed include territorialcontrol, safety and security, capacity to manage publicresources, delivery of basic services, and the abilityto protect and support the ways in which the poorestpeople can sustain themselves (1).

5. DFID’s working definition is based on the WorldBank’s Country Policy and Institutional Assessment(CPIA) scores, which divide low-income countriesinto quintiles, or five categories of performance (8).The lowest 40 percent (lowest two quintiles)comprises a list of countries that the World Bankdescribes as “low-income countries under stress”(LICUS), and which serves as a proxy for fragility.There is, in addition, another group of countrieswhich by the DFID definition can be consideredfragile. Together, this comprises a list of 46 countrieswith a total population of 870 million people or 14 percent of the world’s population that can bedescribed as fragile states. (The list of the 46 statesregarded as fragile is given in Appendix 1).

6. State fragility presents a challenging environmentfor delivering development assistance. From theearly 1990s onwards, donors have tended to rewardlow-income countries with relatively effective

governments and stable macroeconomic policies.The poorly-performing states, i.e. the fragile states,have been confined to receiving mostly limited,short-term humanitarian aid (1, 9, 10). This trend isalso discernible in a recent decision by the G8 toprovide conditional cancellation of debt to heavilyindebted poor countries (HIPCs). Only four of the 18countries thus granted debt relief are fragile states,namely Ethiopia, Niger, Guyana and Mali.

7. There are compelling reasons for investing inpeople in fragile states. First, a large proportion ofthe world’s poorest citizens lives in fragile states.Failure to engage with these states will maintain andmay even increase levels of poverty and insecuritywhich could in turn again exacerbate the degree ofstate fragility (1, 2, 11, 12).

8. Second, fragile states have a higher diseaseburden than other low-income countries. Forinstance, the malaria death rate is nearly 13 timeshigher in fragile states than elsewhere in thedeveloping world. The proportion of people livingwith HIV is four times higher, and maternal mortalityis more than two and a half times higher. Nearly halfof the children who die before the age of five areborn in fragile states, and child mortality is two anda half times higher than in other poor countries. Toignore the states with high disease burden will onlyaggravate the burden, undermine poverty reductionefforts, and imperil achievement of the MillenniumDevelopment Goals.

9. Third, the people in these states are less likely togo to school or access essential healthcare. Denyingthe populations in fragile states access to two of themost fundamental tools for poverty reduction is tocast a long shadow over the lives and livelihoods ofthe millions living in these states and their futuregenerations.

10. Fourth, given the risk of export of disease acrossborders, of political and military instability engenderingsubsequent refugee flows, failure to implementeffective action in fragile states undermines theinvestments made in neighboring states.

9

INTRODUCTIONINTRODUCTION

10

The Global Fund has providedUS$ 35.6 million dollars to theDemocratic Republic of Congofor grants to combat HIV/ AIDS,tuberculosis and malaria.

11. The mandate of the Global Fund is to mobilizeresources on a large scale and to disburse theseresources to countries in need so that they canaccelerate and scale up the fight against these threediseases of poverty. The countries in need of theGlobal Fund’s resources are primarily the low- andmiddle-income countries with a high disease burden.This focus on low-income status and high diseaseburden, free from any other policy constraints andconditionalities, has led the Global Fund to investlarge sums of resources in fragile states. Thiscontrasts with the general current of developmentassistance from most other sources.

12. The performance and results of the grants infragile states will, to a large extent, influence theoverall success of the Global Fund and is thus ofmajor interest. The Global Fund’s large portfolio ofgrants in both fragile states and stable, low-incomecountries provides an opportunity to compare theperformance of these grants within radically differentenvironments. This is all the more important in thatthe disbursement of funds to the grantees, as well ascontinued funding decisions after the first two years ofprograms (Phase 2), are conditional on performanceregardless of state fragility (although the Phase 2evaluation allows for contextual information to beconsidered in making continued funding decisions).All its grantees are required to demonstrate goodperformance in managing and implementing theirgrants to assure continued funding.

13. Can this system work in fragile states? Thesestates have, by definition, a low capacity to managepublic resources, low absorptive capacity, inadequateinfrastructure, and a low human resource base. Insome cases, states may have few or no other donorsproviding significant funds for the fight against thethree diseases, as was the scenario in Togo when theGlobal Fund first awarded that country a grant. A keyquestion therefore is: given that the Global Fund doesnot run a differential system for performance-baseddisbursements according to political stability, canfragile states be expected to perform at the samelevel as stable states?

STUDY OBJECTIVES

14. This report presents the first information drawn froman ongoing Global Fund study on the performanceof fragile states. It is based on a study of 19 grants tofragile states which have reached 18 months or moreof program activities. The study had two objectives.First, it aimed to evaluate the suitability of the GlobalFund model for fragile states, looking into areas forpotential improvement. Then the study also aimed todraw lessons that can be used in other areas of healthdevelopment assistance, including the non-healthsectors, to advance the effectiveness of developmentaid to fragile states. It should be noted that evenwithin the category of “fragile states” there aresubstantial differences between individual countriesin ability and level of commitment concerning thedelivery of public services. Some of the results of thestudy may be attributed partly to these differences.

15. The key questions asked included: what are thekey factors that underlie well-performing grants infragile states? What factors are associated withpoorly-performing grants? The study examined thesize of grants; the characteristics of each state’sCountry Coordinating Mechanism; the number andtype of sub-recipients per grant; the managementcharacteristics of managers at implementation level;the reporting and functional relationships betweenPrincipal Recipient and sub-recipients; the relativedemand for support visits from Fund PortfolioManagers; and the roles played by technical partnersor other donor agencies.

11

THE GLOBAL FUND ANDFRAGILE STATESTHE GLOBAL FUND AND

FRAGILE STATES

STUDY DESIGN

16. The study looks at grant performance in fragilestates, based on grants assessed for Phase 2funding. These were five-year grants that hadreached the end of Phase 1 (the first two years) andhad passed through Phase 2 assessment to determinewhether funding should be continued for years 3 to 5(Phase 2). The Global Fund policy is that when agrant is approved by the Board, approval is, inprinciple, to fund the proposal for up to five years,since most of the proposals cover this time range.However, to ensure that financing is applied on aperformance basis, the funds committed by theBoard are for the first two years only. Beforecommitting to the extension of funding beyond thefirst two years, the Global Fund Secretariat criticallyreviews performance of the grant as at 18 months toassess whether funding should be approved forPhase 2.

17. The decision is based on critical assessment offinancial and programmatic performance as well as themanagement and governance of the grant. There areseveral points during the lifecycle of a grant at whichevaluation is carried out; however, the assessment atthe Phase 2 stage is a critical point involving agreater level of depth in the analysis of results. Thestakeholders responsible for grant performance arethe Principal Recipient (who carries accountability forthe grant), the Country Coordinating Mechanism (a stakeholder body responsible for submitting grantproposals and overseeing implementation), and thesub-recipients (as implementers of all or part of theprogram activities). Information on grant performancesubmitted by the CCM is, in addition, verified by aLocal Fund Agent (an independent entity contractedby the Global Fund for that purpose).

18. All the grants that had gone through their Phase2 assessments by the end of July 2005 were includedin the study. They included grants implemented inboth stable and fragile states during the same timeperiod. As of this date, 74 grants had completed thisprocess and so were included. Of the 74 grantsassessed, 19 were from fragile states. This comparesto 316 grants signed overall, of which 123 are infragile states. The relatively small number of grantsfrom fragile states which could be included in thisinitial study places substantial limitations on theweight of the conclusions of this analysis. However,the analysis is continuing as more grants continue tocome up for Phase 2 review.

STUDY LIMITATIONS

19. Based on the definition of Phase 2 (i.e., having aminimum of 15 months of implementation) this studyreviews only the grants with a lifespan longer than twoyears. Any grant awarded for a maximum of two yearswas not part of this analysis, such as the well-performinggrants to two NGOs (SANAA and KENWA) in Kenya(one of the fragile states). Likewise, five-year grantsthat were still in Phase 1 (i.e., less than 15 months’functioning) were not included. Other aspects notaddressed in this study are: the conditions precedentthat were required prior to grant agreement signature;time to completion of negotiations; and time tocommencement of implementation after receipt of firstdisbursement. This is the subject of a separate study.

12

STUDY METHODOLOGY

20. All information relevant to the Phase 2evaluation is contained in Grant Score Cards (GSC),which are compiled by the Global Fund’s StrategicInformation & Evaluation (SIE) unit at the Secretariat.This study reviewed the Grant Score Cards for allgrants that had completed the Phase 2 process. Theresearchers and authors of this study have not beenpart of the Secretariat’s Phase 2 decision-makingprocess.

21. The Grant Score Cards rate the overallperformance of grants according to one of fourcategories: A, B1, B2 or C, based on the proportionof program targets achieved and the rate of resourceutilization (defined as the proportion of fundsdisbursed against the expenditure budgets agreedat the start of implementation). Relevant contextualinformation is also taken into consideration. Category Arepresents grants reaching or exceeding expectationsand which have no or minor contextual issues; B1covers grants that have adequate performanceand/or have substantial contextual issues, but whichissues are likely to be solved over time; B2 coversgrants that show inadequate performance but havedemonstrated potential and/or major recentimprovements in the program-supporting environment;and C is for grants whose performance is unacceptablylow or have critical contextual risks (13). Grantsfalling into categories A and B1 are generallyrecommended for continued funding. Category B2grants are generally recommended for continuation,subject to certain conditions identified by theSecretariat. However, some B2-category grants maybe recommended for discontinuation, as hasoccurred with three grants thus far (to be discussedlater). Category C grants are recommended fordiscontinuation of funding.

22. In addition to performance measurement, otherinformation recorded on the Grant Score Cardsconcerns contextual issues related to program andfinancial performance. Grant management andgovernance are also reviewed. Data analysis for thestudy compared performance across grants in bothfragile and stable states. Structured interviews, guidedby a questionnaire, were conducted separately withthe Fund Portfolio Managers responsible for therelevant grants in order to gain further insights intothe contextual issues surrounding each grant. Thestudy also reviewed relevant mission reports by FPMs.

23. The limited time available did not permit interviewswith Principal Recipients and sub-recipients, LocalFund Agents, and representatives of the CountryCoordinating Mechanisms with in-depth review of allrelevant documentation prior to preparation of thisreport. This will be done later to consolidate theinitial findings.

13

14

A clinican reads an HIV-positive testresult at a public hospital. Global Fundgrants for HIV/AIDS support testing,treatment and prevention activities.

GLOBAL FUND INVESTMENT IN FRAGILE STATES

24. Over the first four rounds of funding, 123 (41 percent) of the total of 316 grants allocatedglobally were to countries defined as fragile states.Forty-five of the 46 fragile states listed in Appendix 1received at least one grant. The only exception wasthe Republic of Congo, which has yet to receive agrant from the Global Fund. The 123 grantsrepresent a commitment by the Global Fund Boardto finance interventions in the fragile states of US$ 1.1 billion over two years, and approval for aceiling of US$ 3.0 billion over the five-year lifespanof the grants. These allocations to fragile statesaccount for 35 percent of the total resourcesawarded by the Global Fund in Rounds 1 to 4 (seeFigure 1a). This is remarkable, especially whencompared to the recent trend in bilateral assistanceas shown in Figure 1b. In 2001, fragile states, whichconstitute the bottom 40 percent of countries on theCPIA score (quintiles 4 and 5), received only 14 percentof bilateral aid, whereas the top three quintiles (quintiles1, 2 and 3) received 83 percent of all bilateral aid, ornearly six times as much as the fragile states (1).

25. The proposals submitted to the Global Fund arecountry-driven, based on gaps in a country’s nationalstrategic frameworks to fight one or more of thethree diseases. The Global Fund does not setspecific allocations to individual countries inadvance.

26. The study examined whether grants to fragilestates were any different in size to those of stablestates. As shown in Figure 2, below, the studyconcludes that there is no difference in size of grantswith respect to fragile states as compared to stablestates. The distribution is more or less identical.Furthermore, grant size did not seem to influenceoverall grant performance. Grants of less than US$ 10 million performed just as well as those aboveUS$ 10 million over the two-year funding period.

15

KEY FINDINGSKEY FINDINGS

Figure 1: Allocation according to fragility status (a) Global Fund resources andas compared to (b) overall bilateral aid for 2001

% A

lloca

tion

(a) (b)100%

90%

80%

70%

60%

50%

40%

30%

20%

10%

0%GlobalFund

(2 years)

GlobalFund

(5 years)

BilateralAid

(2001)

StableFragile Figure 2: Size of grants to fragile states as compared to grants to stable states

0-5.9 6.0-10.9 11.0-15.9 16.0+

Grant Size in US$ Million

StableFragile

% A

lloca

tion

50%

40%

30%

20%

10%

0%

GRANT PERFORMANCE IN FRAGILE STATES

27. Owing to the general lack of capacity, combinedwith political instability, low absorptive capacity, andother aspects of state fragility, fragile states wouldnot be expected to do as well as stable states.Performance of grants in fragile states was examinedthrough an evaluation of Phase 2 assessment scoresfor 19 grants compared with the scores for the 55grants implemented in stable states. The results aregiven in Figures 3a and 3b, below.

28. Figure 3a shows the distribution of grants byperformance category for all the four ratings (A, B1,B2, and C). Whereas grants in the stable states morefrequently scored in the A category, fragile statestended to score a higher proportion of their grants incategory B1 as compared to grants in stable states.

29. Two grants among those implemented in thestable states scored a C. Funding for one of thesegrants was discontinued by the Board due to thisunsatisfactory performance. The other is still underevaluation by the Board. In addition, two othergrants in the stable states scored a borderline B2; a Board decision on the continuation of funding isstill pending in these two cases.

16

Figure 3a: Performance of grants in fragile states vs. stable states for allperformance ratings

A B1 B2 C

Grant Rating

StableFragile

% o

f Cou

ntrie

s

70%

60%

50%

40%

30%

20%

10%

0%

30. Since categories A and B1 are both measures ofsatisfactory performance, the two categories arecombined in the analysis given in Figure 3b, below,while categories B2 and C are given separately, asthey both reflect less-than-satisfactory performance.The results show that 75 percent of the grants infragile states scored either A or B1, as compared to81 percent of the stable states grants which scoredin the same categories. When we look at the grantsgraded together as “satisfactory”, we find that thereis little difference in performance between those offragile states and those of other countries. This isprobably the most unexpected finding of thispreliminary study.

PRINCIPAL RECIPIENTS IN FRAGILE STATES

31. One of the determinants of state fragility is thelow capacity of state organs to manage publicresources. The expectation would therefore be thatcivil society, private sector, or the UNDP, which has astanding arrangement with the Global Fund to act as“Principal Recipient of last resort”, would morefrequently take the role of PR in fragile states ascompared to stable states.

32. The results shown in Figures 4a and 4b showotherwise. For the grants in fragile states, publicsector agencies, mostly Ministries of Health, werethe predominant PR, accounting for 74 percent ofthe grants in fragile states (whereas they onlycomprise 53 percent of PRs in stable states). TheUNDP as PR accounted for about 20 percent ofgrants in both fragile and stable states. In stablestates there tends to be higher involvement ofprivate sector and civil society organizations servingas PR (29 percent). However, in the fragile states,there was only one private sector PR. Again, it isunwise to draw strong conclusions at this stage, dueto the small sample size involved, but the figures dosuggest the ability of the public sector to carry outprograms of substantial scale in some fragile states.

17

Figure 3b: Performance of grants in fragile states vs. stable states (combining categories A and B1)

A+B1 B2 C

Grant Rating

StableFragile

% o

f Gra

nts

90%

80%

70%

60%

50%

40%

30%

20%

10%

0%

81%

75%

16%

25%

2%

33. A further question was whether governmentagencies in fragile states would perform well asPrincipal Recipients, since capacity to manage publicresources is one of the determinants of state fragilityand a major concern of donors. Overall, 14 of the 19grants in fragile states were managed by governmentPRs. Based on Phase 2 performance scores, all 14 ofthese grants scored a B2 and above; there was not asingle grant scored C. Of the 14 grants managed bygovernment PRs, nine (64 percent) scored A or B1.However, the sample size is still small.

34. In comparison, grants in the stable statesshowed little difference in performance whether theywere managed by a governmental PR or by anothertype of organization. Of the 55 grants analyzed, 26 aremanaged by government PRs, and of these 85 percent(22 grants) scored in the A and B1 categories. Of the29 non-governmental PRs, 79 percent, or 23 grants,scored equally as high.

18

Figure 4a: Type of Principal Recipient in fragile states

Government74%

UNDP21%

5%Civil Society/Private Sector

Figure 4b: Type of Principal Recipient in stable states

Government53%

UNDP18%

29%Civil Society/Private Sector

PERFORMANCE BY DISEASE IN FRAGILE STATES

35. Interventions relating to the three diseases(malaria, tuberculosis and HIV/AIDS) are organizeddifferently and are supported by different partnerships,both at the global and at the country level. Of thethree, TB is probably best organized, with clear controlstrategies and a global, strong and well-organizedStop TB Partnership that supports national TB controlprograms. In the case of HIV/AIDS, the interventionsare less defined and some lack universal agreement.The interest is also much wider, attracting a widerspectrum of implementers which is both a blessingand a challenge to coordinate. Malaria lies somewherein between.

36. The study examined whether grants for thedifferent disease components performed differently.All other factors being equal, the TB grants performedexceptionally well. TB grants in the fragile states allscored A or B1; no grant scored B2 or C (see Figure 5a).The malaria and HIV/AIDS grants scored well, butdid not achieve the same level of quality, in thatthere were some which fell into the B2 category. Theresults given in Figure 5b reveal the tendency for TBgrants to also perform well in stable states. Of the 10TB grants in stable states, nine were either categoryA or B1, with only one grant in category B2.

19

Figure 5a: Grant performance by disease component in fragile states

HIV/AIDS TB Malaria

% o

f Gra

nts

100%

80%

60%

40%

20%

0%

Figure 5b: Grant performance by disease component in stable states

HIV/AIDS TB Malaria HIV/TB

% o

f Gra

nts

100%

80%

60%

40%

20%

0%

CB2A+B1

20

With Global fund support, Nigerplans to intensify prevention andtreatment efforts for HIV/AIDS andsexually transmitted infections inorder to reach at least 90 percentof the population by 2006.

37. To implement grants in fragile states is to navigatethe issues that form the essence of state fragility.Understanding the environment in which the grantswere implemented is essential for re-positioningdonor aid to these states. As a tool for betterunderstanding these issues, the Grant Score Cardsfor all 19 grants implemented in fragile states werereviewed for further insights into factors that mayhave led to the better-than-expected performanceof these grants. The areas examined included therole of the CCM in overseeing grant implementation;the role of the PR in managing activities andreporting requirements for sub-recipients; the rate oftimely disbursements to sub-recipients; oversight bythe PR over financial and programmatic activities ofsub-recipients; input by technical partners; and therole of the LFA in facilitating the process. This part ofthe study is not yet completed, but some observationsare given below.

38. A well-performing grant was more likely to havea proactive CCM which met regularly with the PRand sub-recipients to review progress. These grantsalso evidenced frequent communication with the PRand sub-recipients and with the Fund PortfolioManager at the Global Fund Secretariat. Site visits bythe PR and CCM to encourage implementation andto verify data, such as observed in the multi-countryWestern Pacific malaria grant, appears to be a criticalfactor. The PR, in the example of the Western Pacificmalaria grant just noted, was proactive in collectingdata from sub-recipients and undertook numeroussite visits, taking a hands-on approach to managingand monitoring the implementation of this grant.

39. Creativity and initiative on the part of both theCCM and PR were also important. For example, inBurundi, the CCM and the PR introduced a policywhereby Global Fund resources would be used toprocure artemisinin-based combination therapy(ACT) for first-line treatment, while funds from otherdonors such as the World Bank and UNICEF wouldbe used to buy second-line drugs such as quinine. ThePR then took the initiative to reallocate the savingsrealized from the procurement of second-line drugsto the procurement of ACT and bed nets. The CCMalso decided, on its own, to expand its composition,bringing in more civil society representatives includingorganizations of people living with the diseases. At the time of the writing of this report, plans wereunderway to expand the representation of bilateral andmultilateral agencies as well.

40. Coordination of partnerships at the country levelhas proven to be critical. Since the signing of itsHIV/AIDS grant, several other donors in Burundihave started to contribute to the government’snational strategy. The CCM has encompassed theseother development partners and has coordinatedthe efforts of these partners into a coherent nationaldisease strategy and plan of action. Recognizing thebenefits accrued from a coordinated partnership forHIV/AIDS, the Minister of Public Health, who is thePR, has embarked on fostering better coordinationamong donors and technical partners implementingthe malaria program as well.

21

CONTEXTUAL ISSUES:NAVIGATING STATE FRAGILITYCONTEXTUAL ISSUES:

NAVIGATING STATE FRAGILITY

41. Recognizing and facilitating participation by civilsociety, the private sector and faith-based organizationsis important for increasing absorptive capacity and forscaling up interventions. Public/private partnershipsare seen as critical for changing the paradigm ofdevelopment aid and for mobilizing local resourcesto respond to the three pandemics, and the GlobalFund encourages such partnerships through itsfunding model. Most of the grants in fragile stateshave embraced this concept and have activelyexpanded the number of sub-recipients considerablybeyond the traditional public sector execution. The HIV/AIDS grant in Cambodia, for example, has13 sub-recipients while the Ethiopia tuberculosisgrant has nine. The Burundi HIV/AIDS grant had thehighest number of sub-recipients, totaling 18.

42. Performance-based funding seems to prod theactors at the country level into action, perhapsstemming from fear of loss of grants and the politicaland public repercussions arising from such a loss. InEthiopia, for instance, the PR took considerable timeto understand the realities of performance-basedfunding, and this led to a failure to appreciate theimportance of timely and accurate reporting.However, between June and September 2004, thePR realized that the program was slipping. A greatersense of urgency evolved, pushing them to achieveprogram goals. Reporting drastically improved and ahigh-level (ministerial) commitment emerged forworking together to make the program succeed.Political commitment was also evident in other statesand at both political and program levels.

43. Delayed disbursement - both in terms of delayeddisbursement from the Global Fund Secretariat to thePR, and by the PR to the implementers/sub-recipients- was a major factor in poorly-performing grants. Themost frequent causes for delayed disbursement tothe PR was related to cumbersome procurementassessments of the PR by the Global Fund anddelayed procurement of drugs and other healthproducts due to bureaucratic processes at countrylevel. At the PR level, frequent causes cited fordelayed disbursements revolved around reportingissues with sub-recipients.

22

Sudan’s estimated 7.5 million cases ofmalaria per year result in approximately35,000 deaths annually. The Global Fundhas approved US$ 27 million for twogrants to combat malaria in Sudan.

23

24

In Mali, 80-90 percent of children underfive carry the Plasmodium parasite in therainy season. With Global Fund resources,the Ministry of Health and its partners aimto ensure that nearly half a million Malianwomen and children will sleep underinsecticide-treated bed nets by the end of the grant’s first two years.

44. The Global Fund is investing substantially infragile states. Approximately one-third of all theresources allocated globally in Rounds 1-4 have goneto fragile states. This applies to both the two-yearcommitments and the five-year grant period. This isprincipally because of the Global Fund’s mandatethat prioritizes flow of its resources to low-incomecountries with high disease burden, regardless ofgovernment structure, donor history or any otherfactors. This level of investment is a significantdeparture from current practice of the bilateral donoragencies, which have tended to avoid investments infragile states or to limit those investments to merelyshort-term humanitarian responses.

45. Second, the evidence from the 19 grants so faranalyzed suggests that the grants in fragile states areperforming well. This is entirely unexpected, inparticular because the majority of the PrincipalRecipients in these countries are governmentministries. It is also interesting to note that grantsmanaged by government PRs performed similarly tothose managed by non-government PRs, althoughthe grant number here is too small to draw anyreliable conclusions.

46. The findings suggest that the Global Fund’sperformance-based funding model is flexibleenough to work in fragile states, and that it isbuilding capacity as grants are being implementedwith input by technical partners. Thus, weakgovernment structures can actually be strengtheneddirectly by this model, enabling government PRs tomanage the grants well and to share resources withnon-government sub-recipients as long as there is astrong political will for change and a commitment toachieving results. This underscores the importanceof country leadership and country ownership increating change and in providing a strong basis forsustainability, as compared to direct implementationby external agencies (13, 14). If the results hold trueas more grants are assessed and are included in theongoing analysis, the Global Fund may offer aunique, performance-based model within whichother donors can engage with fragile states, inhealth and other sectors.

25

CONCLUSIONCONCLUSION

26

APPENDIX 1: LIST OF FRAGILE STATES1

1. Afghanistan

2. Angola

3. Azerbaijan

4. Burma

5. Burundi

6. Cambodia

7. Cameroon

8. Central African Republic

9. Chad

10. Comoros

11. Congo, Democratic Republic

12. Congo, Republic of

13. Cote d’Ivoire

14. Djibouti

15. Dominica

16. Eritrea

17. Ethiopia

18. Gambia, The

19. Georgia

20. Guinea

21. Guinea Bissau

22. Guyana

23. Haiti

24. Indonesia

25. Kenya

26. Kiribati

27. Lao, PDR

28. Liberia

29. Mali

30. Nepal

31. Niger

32. Nigeria

33. Papua New Guinea

34. Sao Tome & Principe

35. Sierra Leone

36. Solomon Islands

37. Somalia

38. Sudan

39. Tajikistan

40. Timor Leste

41. Togo

42. Tonga

43. Uzbekistan

44. Vanuatu

45. Yemen, Republic of

46. Zimbabwe

1 As defined by the World Bank and the UK’s Department of International Development (see Introduction on p.9)

27

APPENDIX 2: GRANTS ASSESSED FOR PHASE 2 FUNDING THROUGH JULY 2005A. Grants implemented in stable states

Grant Number

ARG-102-G01-H-00ARM-202-G01-H-00BEN-102-G01-M-00BEN-202-G02-T-00BEN-202-G03-H-00CHL-102-G01-H-00CHN-102-G01-T-00CHN-102-G02-M-00COR-202-G01-H-00CUB-202-G01-H-00EST-202-G01-H-00GHN-102-G01-H-00GHN-102-G02-T-00GHN-202-G03-M-00HND-102-G01-H-00HND-102-G02-T-00HND-102-G03-M-00IDA-102-G01-T-00JOR-202-G01-H-00MAF-202-G01-M-00MDG-102-G01-M-00MDG-202-G02-H-00MDG-202-G03-H-00MOL-102-G01-C-00MON-102-G01-T-00MON-202-G02-H-00MOR-102-G01-H-00PAN-102-G01-T-00PHL-202-G01-M-00PHL-202-G02-T-00RWN-102-G01-C-00SAF-102-G02-C-00SER-102-G01-H-00SLV-202-G01-H-00SNG-102-G01-H-00SNG-102-G02-M-00SWZ-202-G01-H-00SWZ-202-G02-M-00TNZ-102-G01-M-00UGD-102-G01-H-00UKR-102-G04-H-00ZAM-102-G01-H-00ZAM-102-G02-M-00ZAM-102-G03-T-00ZAM-102-G04-H-00ZAM-102-G05-M-00ZAM-102-G06-T-00ZAM-102-G08-H-00ZAN-102-G01-M-00ZAN-202-G02-H-00HRV-202-G01-H-00WRL-102-G01-H-00KAZ-202-G01-H-00SRL-102-G03-T-00THA-202-G03-H-00

DiseaseComponent

HIV/AIDSHIV/AIDSMalariaTuberculosisHIV/AIDSHIV/AIDSTuberculosisMalariaHIV/AIDSHIV/AIDSHIV/AIDSHIV/AIDSTuberculosisMalariaHIV/AIDSTuberculosisMalariaTuberculosisHIV/AIDSMalariaMalariaHIV/AIDSHIV/AIDSHIV/TBTuberculosisHIV/AIDSHIV/AIDSTuberculosisMalariaTuberculosisHIV/TBHIV/TBHIV/AIDSHIV/AIDSHIV/AIDSMalariaHIV/AIDSMalariaMalariaHIV/AIDSHIV/AIDSHIV/AIDSMalariaTuberculosisHIV/AIDSMalariaTuberculosisHIV/AIDSMalariaHIV/AIDSHIV./AIDSHIV./AIDSHIV./AIDSTuberculosisHIV./AIDS

PrincipalRecipient

UNDPPrivate sector UNDPUNDPUNDPPrivate Sector:Government: otherGovernment: OtherGovernment: OtherUNDPMinistry of HealthMinistry of HealthMinistry of HealthMinistry of HealthUNDPUNDPUNDPGovernment: OtherMinistry of HealthCivil SocietyCivil SocietyCivil SocietyCivil SocietyMinistry of HealthMinistry of HealthMinistry of HealthMinistry of HealthUNDPPrivate SectorPrivate SectorMinistry of HealthMinistry of FinancePrivate SectorUNDPGovernment: OtherMinistry of HealthGovernment: OtherGovernment: OtherMinistry of HealthMinistry of FinanceCivil SocietyMinistry of HealthMinistry of HealthMinistry of HealthFaith Based OrganizationFaith Based OrganizationFaith Based OrganizationCivil SocietyMinistry of HealthGovernment: OtherMinistry of HealthWorld Lutheran FederationOther government ministryMinistry of HealthCivil society

Country

ArgentinaArmeniaBeninBeninBeninChileChinaChinaCosta RicaCubaEstoniaGhanaGhanaGhanaHondurasHondurasHondurasIndiaJordanMulti-country AfricaMadagascarMadagascarMadagascarMoldovaMongoliaMongoliaMoroccoPanamaPhilippinesPhilippinesRwandaSouth AfricaSerbia & MontenegroEl SalvadorSenegalSenegalSwazilandSwazilandTanzaniaUgandaUkraineZambiaZambiaZambiaZambiaZambiaZambiaZambiaTanzaniaTanzaniaCroatiaMulti-countryKazakhstanSri LankaThailand

Grant sizein US$

12177200316664129731502173404113480003805941648070000640665922795011146512939089521417022256870554596111125834666597014720414086550331778600709031820000637471993032048117190471730000299710392387545700007244762343448784092681200000027187141285672960000004285714296333009780008790612363148922335411621214271170392001244729466149588526002307962807301378122011160003,363,974485,0006,502,0002,384,9805,993,913

PerformanceRating

AAB1B1B1B1AB1B2AB1B1B1AB2B2B2AAAB1AB1AAAAAB1AAB2B1B1CCB1B2B2B2B1B1B1B1AB1AB1B1B1B1B1B1B1B1

28

B. Grants implemented in fragile states

Grant Number

BRN-102-G01-H-00BRN-202-G02-M-00CAF-202-G01-H-00CAM-102-G01-H-00ETH-102-G01-T-00ETH-202-G02-M-00HTI-102-G01-H-00HTI-102-G02-H-00IND-102-G01-T-00IND-102-G03-H-00LAO-102-G01-H-00LAO-102-G02-M-00LAO-202-G03-T-00MWP-202-G01-H-00MWP-202-G02-M-00MWP-202-G03-T-00TAJ-102-G01-H-00TMP-202-G01-M-00ZAR-202-G01-T-00

DiseaseComponent

HIV/AIDSMalariaHIV/AIDSHIV/AIDSTuberculosisMalariaHIV/AIDSHIV/AIDSTuberculosisHIV/AIDSHIV/AIDSMalariaTuberculosisHIV/AIDSMalariaTuberculosisHIV/AIDSMalariaTuberculosis

PrincipalRecipient

Other Government ministryMinistry of HealthUNDPMinistry of HealthMinistry of HealthMinistry of HealthPrivate SectorUNDPMinistry of HealthMinistry of HealthMinistry of HealthMinistry of HealthMinistry of HealthOther government ministryOther Government ministryOther government ministry UNDPMinistry of HealthUNDP

Country

BurundiBurundiCentral Africa RepublicCambodiaEthiopiaEthiopiaHaitiHaitiIndonesiaIndonesiaLao PDRLao PDRLao PDRMulti-country Western PacificMulti-country Western PacificMulti-country Western PacificTajikistanTimor LesteCongo, DemocraticRepublic of

Grant sizein US$

48770001379212681989211124253810962600379150112460368067546972161226569249713407664127090871524338516392545303002738806242524523007446408741

PerformanceRating

B1B1B1B1B1B2AB1B1B2B2B2AB2B1B1AB1B1

29

1. Why We Need to Work More Effectively in Fragile States. Department for International Development,January 2005. (http://www.dfid.org/)

2. Fragile States Strategy. US Agency for International Development, January 2005.(http://www.usaid.org)

3. Moreno-Torres M and Anderson M. Fragile States: Defining Difficult Environments for PovertyReduction. Department for International Development PRDE Working Paper I, August 2004.

4. Berry C, Forder A, Sultan S, and Moreno-Torres M. Approach to Improving the Delivery of SocialServices in Difficult Environments. Department for International Development PRDE Working Paper III,October 2004.

5. Towards a European Consensus for Development: the European Commission approves a proposal foran ambitious development policy. European Commission, Brussels, July 2005.

6. World Bank Group Work in Low Income Countries Under Stress: A Task Force Report. The World Bank,September 2002.

7. Moreno Torres M and Anderson M. (2004) Fragile States: Defining Difficult Environments for PovertyReduction. PRDE Working Paper I. Unpublished manuscript, DFID.

8. Country Policy and Institutional Assessment (CPIA) Ratings.(http://siteresources.worldbank.org/IDA/Resources/2004CPIAweb1.pdf)

9. Levin V and Dollar D The Forgotten States: Aid Volumes and Volatility in Difficult Partnership Countries(1992-2002). The World Bank, January 2005.

10. McGillivray M. Aid Allocation and Fragile States. World Institute for Development Economics ResearchUnited Nations University, Helsinki, January 2005.

11. Chauvet L and Collier P. Development Effectiveness in Fragile States: Spillovers and Turnarounds.Centre for the Study of African Economies, Department of Economics, Oxford University; January 2004.

12. The Chairs Summary: Senior Level Forum on Development Effectiveness in Fragile States.Organisation for Economic Cooperation and Development, Development Assistance Committee,London 13-14 January 2005.

13. Investing in the Future: The Global Fund at Three Years (2005), The Global Fund to Fight AIDS,Tuberculosis and Malaria, Geneva, 2005.

14. Principles for Good International Engagement in Fragile States. Organisation for EconomicCooperation and Development, Development Assistance Committee, April 2005.

REFERENCESREFERENCES

30

ISBN 92-9224-027-7

The Global Fund to Fight AIDS, Tuberculosis and Malaria

Chemin de Blandonnet 8 1214 Vernier Geneva, Switzerland

+41 22 791 1700 (phone) +41 22 791 1701 (fax)

www.theglobalfund.org info@theglobalfund.org