the numbers behind the stories

8/14/2019 The Numbers Behind The Stories

1/27

The Numbers

Behind the

Stories

PEPFAR Funding

for Fiscal Years

2004 to 2006

Nandini Oomman

Michael Bernstein

Steven Rosenzweig


2/27

The Numbers

Behind the

Stories

PEPFAR Funding

for Fiscal Years

2004 to 2006

Nandini Oomman

Michael Bernstein

Steven Rosenzweig


3/27

iii the numbers behind the stories

Contents

Summary 1

PEPFAR and its data 3

Issue 1. How do congressional unding restrictions (earmarks) infuencePEPFAR allocations or prevention, treatment, and care? 6

Issue 2.What is the role o aith-based organizations in PEPFAR? 9

Issue 3. What is the role o host country governments in PEPFAR? 12

Issue 4. Has PEPFAR been building local capacity by unding localrecipients? 14

Final refections 17

Annex A. Data and methods 18

Annex B. Using the data 21

Annex C. Funding data collected by PEPFAR 23

Reerences 24

Copyright 2008 by the Center for Global DevelopmentISBN 1-933286-30-X

Center for Global Development

1776 Massachusetts Avenue, NWWashington, D.C. 20036

el: 202 416 0700Web: www.cgdev.org

Support for this project was generously provided by the Bill & Melinda Gates Foundation.

Te views expressed are those of the authors and should not be attributed to the directors

or funders of the Center for Global Development.

Cover photo by Michelle Bashin/Courtesy of Photoshare.

Tis photograph is used for illustrative purposes only; it does not imply any particular

health status, attitude, behavior, or action on the part of the person appearing in the

photograph.

Editing and typesetting by Communications Development Incorporated, Washington, D.C.


4/27

pepfar funding for f iscal years 2004 to 20 06

Te Presidents Emergency Plan or AIDS Relie (PEPFAR), the largest under o

global HIV/AIDS programs, does not regularly release detailed inormation about

how the money is spent.We recommend that PEPFAR publish ocial data on

obligations to prime recipients and subrecipients to improve transparency and

accountability, permitting program cost-efectiveness analyses and comparisons

o investments with results. In addition, making data available would strengthen

coordination with other stakeholders; help host country governments plan or theuture; and make those in the policy and advocacy communities who are careully

watching PEPFAR less dependent on anecdotes and impressionistic inormation.

s

o demonstrate the value o making PEPFARspending data public, this report presents previ-

ously unreleased data acquired by the Center orPublic Integrity or 2004 to 2006 and shows how

analyzing them can clariy several much-debatedissues pertaining to PEPFARs eectiveness and

sustainability. Based primarily on 2005 data, thereport examines how unds were obligated to

PEPFARs 15 ocus countries. It shows that:Congressionally imposed unding restric-tions (earmarks) inuence how PEPFARunding in each ocus country is divided

among programs or prevention, treatment,and care, with treatment receiving the great-est share. Te authors recommend that Con-

gress make PEPFAR unds more exible byremoving the earmarksmaking alloca-

tions more responsive to circumstances ineach country, more consistent with national

AIDS plans, better oriented to local epide-miological conditions, and better coordi-

nated with other stakeholders activities.Faith-based organizations (FBOs) have re-ceived on average 10% o PEPFAR unds,nearly hal o it or treatment. Although

PEPFAR oen deends its ocus on FBOsas a means o support or local groups with

deep roots in their communities, only a smalshare o PEPFAR unding or FBOs goes to

local groups. Te authors recommend thatPEPFAR reevaluate its policy o targeting

FBOs as a way to reach local groups andinstead seek more eective strategies or

achieving that goalstrategies that shouldreect individual country contexts and the

variety o local actors providing services.Host country governments have received onaverage 13% o PEPFAR unds, more thanhal o it or prevention. While the authors

recognize that PEPFAR need not channea large share o its unds through govern-ments, they recommend that it work closely

with host country governments to ensurthat its programming helps them ull

their role as stewards o the national AIDSresponseenabling governments to help

plan and oversee PEPFAR-unded activitiesproviding them and other key stakeholders

with up-to-date inormation on PEPFAR programs, and, through timely inorma

tion sharing and joint planning, helpinggovernments to coordinate their PEPFAR

programs with the myriad other AIDS activities in their countries.


5/27

2 the numbers behind the stories

On average, 70% o PEPFAR unding hasgone to international organizations, while30% has gone to locally based groups. Al-though PEPFAR already invests signicant

sums in capacity building activities, its cur-rent strategy o promoting capacity building

by engaging local subrecipients is not su-cient, since only a small share o its total

unds is being subgranted to them. Te au-thors recommend that PEPFAR reexamine

its capacity building eorts and seek waysto build the capacity o local organizations

to receive greater shares o its undinga

critical sustainability aim, as important asachieving PEPFARs global targets.Te authors hope that these ndings will in-

orm policy debates, including those that bear onPEPFARs impending reauthorization by the U.S.

Congress. Te report also raises questions or ur-ther study, encouraging other researchers and in-

terested parties to build on the authors ndings.Te dataset accompanying the report can

be downloaded at http://www.cgdev.org/hivmonitor/PEPFARdata.


6/27


With PEPFAR due to expire in 2008, the U.S.Congress is now debating its reauthorization.President Bush has requested $30 billion over

ve years. Bills passed in the House and underconsideration in the Senate call or $50 billion.

Because the bills enjoy bipartisan support, theprogram will almost certainly be reauthorized

at $50 billion. Less certain, however, is howthe unds will be spent. Many strong and con-

icting views exist about how much PEPFARshould provide to dierent recipients and pro-

gram areas.Despite this vigorous debate, ew systematic

data are available on how PEPFAR unds arenow being divided.1 PEPFAR collects nancial

and related data about its programs, but manyo these data are not public. Even some PEP-

FAR sta do not have access to certain inor-mation, such as expenditure data or undingrecipients.

A call to PEPFAR:

Release more unding data

PEPFAR should release, at least once per year,the detailed unding and programmatic data

1. Several institutions and initiativesthe Institute o Med-

icine, the Center or Global Developments HIV/AIDS Moni-

tor, the Global HIV/AIDS Initiatives Network, and others

have undertaken case studies in recipient countries. But there

has been no systematic analysis o allocations by grant level,

program area, and recipient type at the country level.

that it already collects. Specically, or eachgrant or contract PEPFAR should disclose:

Te annual amount obligated to prime recipients and subrecipients.Te organization type or each prime recipient and subrecipient (or example: nongovernmental organization, host country

government).Te amount o unds or each strategic area(prevention, care, treatment, other).Each program area unded (or example

blood saety, condoms).2

Each activity area unded (or exampletraining, logistics).Te geographic origin o each prime recipienand subrecipient (local or international).Te programs geographic coverage area (regions or provinces reached within the program country).Te programs target populations (or example: pregnant women, orphans).Te programs stated targets (or example: somany people to be tested or HIV, so manyorganizations to be provided with technica

assistance).A comparison o targets with resultachieved (or example: so many people were

2. When PEPFAR discloses data or this and the precedin

item, it should note the amount obligated to each prime recipi

ent and subrecipient or each strategic and program area. Ta

is the ormat used in this reports accompanying dataset.

Te U.S. Presidents Emergency Program or AIDS Relie (PEPFAR) is the largest

under o global HIV/AIDS programs. A ve-year, $15 billion program, PEPFAR

has pursued three broad goals since its creation in 2003: treating 2 million AIDS

patients, preventing 7 million new HIV inections, and providing care to 10 million

people aected by the disease.

pepfar


7/27


tested or HIV, so many organizations were

provided with technical assistance).3

Releasing these data will improve transpar-ency and accountability, permitting program

cost-eectiveness analyses and comparisons oinvestments with results. It will strengthen co-

ordination among all AIDS stakeholders. Andit will help host country governments prepare

annual budgets.

Analyzing the newly

available unding data

Tis report analyzes previously unavailable

PEPFAR unding data or scal 2004 to 2006(October 1 scal year).4 By clariying our

much-debated issues pertaining to PEPFARs

eectiveness and sustainability, the analysisshows how regularly releasing such data couldimprove PEPFAR programming through evi-

dence-based research and recommendations.Te reports recommendations, based on

its ndings, can and should be used to improveuture PEPFAR programs. In addition, the au-

thors encourage other researchers to build on thendings with their own analyses o the newly

published data. (Te dataset accompanying the

3. In addition to these data, the HIV/AIDS Monitorbelieves that PEPFAR should release data on outlays to prime

recipients and on prime recipients expenditures (or details

see Bernstein and Hise 2007).

4. Te dataset accompanying this report includes details

about the rst, second, third, ourth, and sixth items on

the preceding bullet list. Note that the Center or Global

Developmentnot PEPFARcollected the data or the

sixth item (geographic origin).

report can be downloaded at http://www.cgdev.

org/hivmonitor/PEPFARdata.)Te report addresses our issues:

How do congressional unding restric-1.

tions (earmarks) infuence PEPFAR

allocations or preention, treatment,

and care?o investigate the exibil-ity o PEPFAR unding, the report

looks at the variation across countriesin the amount o money obligated to

each strategic area (prevention, care,treatment, or other programs) and

to each programmatic subarea (orexample: abstinence and be aithul,

blood saety).What is the role o aith-based organi-2.

zations (FBOs) in PEPFAR?o in-

vestigate their role in PEPFAR, thereport examines the amount obligatedto FBOs as a share o total PEPFAR

unds in each country. It looks at the program areas that FBOs are being

unded to address. Finally, it makesobservations about some o the largest

FBO recipients.What is the role o host country gov-3.ernments in PEPFAR?o investigatetheir role in PEPFAR, the report con-

siders the proportion o total PEPFARunds obligated to host country gov-

ernments, as opposed to nongovern-mental recipients, and analyzes the

breakdown o those unds by programarea.

Has PEPFAR been building local ca-4. pacity by unding local recipients?oinvestigate PEPFARs support or lo-

cally based organizations, the reportestimates the amounts obligated to

local groups, the proportion o moneysubgranted to subrecipients, and the

share o subgranted unds going tolocal groups.

Except in its investigation o unding ex-ibility across strategic areas, this reports analy-

sis is based on unds appropriated by the U.S.Congress in 2005 (2005 unds; box 1), or

which the most complete picture o total obli-gations was available. Some PEPFAR recipients

A PEPFAR obligation is a legal commitment to pay a prime recipient, either now or

in the future. For this reason obligations are technically considered commitments.But because the funds are made available for use by prime recipients as soon as

they are needed, PEPFAR obligations are in practice less like commitments than like

disbursements from other funding agencies.

A PEPFAR prime recipient receives funds through a direct transfer from the U.S.

Government. Prime recipients often transfer funds to other organizations through

subgranting. Organizations receiving such subgrants are called subrecipients. This

report refers to prime recipients and subrecipients collectively as recipients.

An appropriation is money that the U.S. Congress legally authorizes to be

transferred from the U.S. Treasury to an executive branch agency.

Box 1 Key terms


8/27


may not receive obligations o unds appropri-

ated in 2006 until 2007 or later. Tereore, anyanalysis o the unding breakdown to individualrecipients would be limited i based on data or

2006 unds.

The data in this report

Because PEPFAR operates through dierentU.S. government agencies and departments,

there are several ways to capture data on pro-gram unding. Tis reports analysis uses previ-

ously unavailable unding data rom PEPFARsmain data-capture mechanism, the Coun-

try Operational Plan and Reporting System(COPRS). Te COPRS provides a wide range o

inormation about each PEPFAR grant, includ-

ing the recipient o unds, the amount obligated,

the program areas unded, any amounts passed

on to subrecipients, and the geographic areasserved.

Te COPRS data used here constitute part

o the COPRS data or scal 2004 through2006, but not all. Tey ocus on unding obli-

gations, a term specic to the U.S. government(see box 1). Te Center or Public Integrity ac-

quired these data through legal action againstthe U.S. State Department (see annex A).5 It

then shared the data with the Center or GlobalDevelopmenta generous act or which the au-

thors o this report are grateul. (On the dataelds used in the dataset, see annex B. For in-

ormation about all data collected by PEPFARand where to nd it, see annex C.)

5. See also Bengts son and Morera (2006).


9/27


ISSUE1

An early evaluation o PEPFAR by the Insti-tute o Medicine in 2007 ound that the ear-marks were unnecessarily constraining the ex-

ibility o unds at the country level, and calledor removing them when PEPFAR is reautho-

rized in 2008.7 In contrast, PEPFAR ofcialshave argued that the earmarks direct spend-

ing towards important program areas, such asorphans and vulnerable children, that other-

wise would receive too little attention.8

Country-level allocationsto PEPFAR program areas

have generally ollowed the

earmarkswith some deviations

Te data reveal a similar pattern o PEPFAR

allocations across countries.9 hat pattern

6. Te earmarks or prevention and care are so earmarks,

meaning they are suggested. Te earmarks or treatment and

orphans and vulnerable children became mandatory in s-

cal 2006. For details on the unding earmarks see Oomman,

Bernstein and Rosenzweig (2007).

7. Institute o Medicine (2007).

8. Personal communication with PEPFAR ofcial, Febru-

ary 11, 2008.

9. Because PEPFAR determines whether earmarks are being

met by examining planned unding levels, obligations data

cannot determine conormity with congressional directives.

Instead, the important task undertaken here is to examine

how the earmarks aect obligationsthe money actually

made available to countries in a given year.

suggests that earmarks are inuencing allocations, despite variations in disease epidemiologyand other important contextual actors.

Figure 1 shows that each ocus countrygenerally appears to be ollowing the allocation

patterns set out in the congressional global earmarks or PEPFAR program areas.10

reatment programs have received the great-est share o unds in all ocus countries, many o

which approach or exceed 55% in their allocations to treatment. Similarly, prevention und-

ing has hovered around the 20% earmark, witha ew countries allocating less but most slightly

more.O the three program areas, it is treatment

whose PEPFAR allocations have been arthest rom the congressional earmark o 55%

Te average share o 2006 unds obligated ortreatment was 44% (gure 2). A likely reasontreatment programs take a long time to set up

especially in countries with inadequate inrastructure or other capacity decits. raining

providers, building health acilities, and procur-ing large drug shipments can take time and delay

putting treatment programs into practice.As these initial investments are made, how

ever, and as treatment programs are scaled up

10. Te Ofce o the U.S. Global AIDS Coordinator report

to Congress on the programmatic breakdown o unding

or its ocus countries only, not or all PEPFAR unding

combined.

Te exibility o PEPFAR unding is much debated, with particular attention to

how congressionally mandated restrictions inuence PEPFAR allocations to dier-

ent program areas. Such legislative restrictions, or earmarks, now direct PEPFAR

to spend 55% o its global unding on treatment, 20% on prevention, 15% on care

and 10% or orphans and vulnerable children.6

hw (k) pepfar v, , ?


10/27


treatment is likely to receive a greater share o

PEPFAR unds. Indeed, the average share ounds obligated to prime recipients or treat-ment grew signicantly rom 2005 to 2006,

rising rom 37.5% to 44% across the 15 ocuscountries, while the average share o obligations

or prevention ell modestly (by 3%).Although the overall unding patterns in

each country appear to be inuenced by thecongressional global earmarks, some PEPFAR

ocus countries have deviated rom the ear-marks. Four countries Cte dIvoire, Guyana,

Mozambique, and anzaniaallocated around30% o their 2006 unds or prevention. And

most countries have been obligating more than15% toward care.

Few countries have met theabstinence and be aithul

earmarkbut allocations to

such activities increased

rom 2005 to 2006

Several earmarks within program areas appear

to inuence how unds are obligated. One, re-quently debated, is the requirement that two-

thirds o all money or preventing sexual trans-mission o HIV must be al located to abstinence

and be aithul (AB) activities. Tis AB ear-mark became mandatory in 2006.11

AB unds or 2006 remained below the two-thirds earmark in most countries. Indeed, dur-

ing 2006, eight ocus countries requested andreceived an exemption rom the AB earmark on

the grounds that it did not make sense in their ep-idemiological conditions.12 Despite such exemp-

11. Te abstinence requirement, at rst only suggested,

became a mandatory requirement in scal 2006. Te congres-

sional earmark states that 33% o all prevention unds should

be devoted to abstinence-only activities. But the Ofce o theU.S. Global AIDS Coordinator released guidance or scal

2006 which asked country sta to meet this requirement by

ensuring that at least 50% o all prevention unds be used

or sexual transmission preventionand that 66% o sexual

transmission prevention unds be used or abstinence and be

aithul programs.

12. Country teams can apply or exemptions rom the AB

earmark by asserting a compelling justication. PEPFAR has

accepted all such applications to date. In 2007 and 2008, 11

countries applied and were exempted.

0

10

20

30

40

Zamb

ia

Vietna

m

Ugan

da

Tanz

ania

South

Afric

a

Rwan

da

Nigeri

a

Nami

bia

Moza

mbiqu

eKe

nya

Haiti

Guya

na

Ethiop

ia

Cte

dIvo

ire

Botsw

ana

0

10

20

30

Zamb

ia

Vietna

m

Ugan

da

Tanz

ania

South

Afric

a

Rwan

da

Nigeri

a

Nami

bia

Moza

mbiqu

eKe

nya

Haiti

Guya

na

Ethiop

ia

Cte

dIvo

ire

Botsw

ana

Funds obligated to prime recipients by strategic area for fiscal 2006 (percent)

Note: Average denotes the average percentage of funds obligated by country (not the average percentageof total PEPFAR obligations for the year).

Prevention

Care

0

10

20

30

40

50

60

70

Zamb

ia

Vietna

m

Ugan

da

Tanz

ania

South

Afric

a

Rwan

da

Nigeri

a

Nami

bia

Moza

mbiqu

eKe

nya

Haiti

Guya

na

Ethiop

ia

Cte

dIvo

ire

Botsw

ana

Treatment

Average

Average

Average

Earmark

Earmark

Earmark

Figure 1 PEPFAR unds have generally ollowedcongressional earmarks in 15 ocus countries


11/27


tions, AB unding increased between 2005 and

2006 in all but our ocus countries (gure 3).13

Te increased unding or AB activities inmost ocus countries suggests the possible inu-

ence o the AB earmark, even in countries wherean exemption rom the earmark was granted.

Although the analysis in this report cannot de-termine whether the AB earmark caused the in-

creases in AB unding rom 2005 to 2006, theassociation between them is worth exploring in

urther research.

Recommendation: Remove

unding earmarks

Tis analysis supports the Institute o Medi-

cines conclusion that the congressional ear-

marks inuence allocations, reducing the ex-ibility o PEPFAR unds.

Greater unding lexibility could make

PEPFAR allocations more responsive to cir-cumstances in each country, more consistent

with national AIDS plans, better oriented tolocal epidemiological conditions, and better co-

ordinated with the activities o other stakehold-ers. Tereore, Congress should make PEPFAR

unds more exible by removing the earmarks.

13. Te Ofce o the U.S. Global AIDS Coordinator exam-ines countries planned obligations or each year to determine

whether the planned unding levels are consistent with the

AB earmark. Since the data in this report are on actual obliga-

tions, it is possible that some ocus countries planned to meet

the AB earmark. For example, PEPFARs planned unding in

Mozambique or 2006 was consistent with the AB earmark

but actual obligations ell short o it.

Average funds obligated to prime recipients by program area for fiscal 2006

Prevention23%

Care20%

Treatment44%

Other12%

Figure 2 Average unds or treatment have been arthestrom the congressional earmark o 55%

0

10

20

30

40

50

60

70

80

Funds obligated to prime recipients for abstinence and be faithful activities for fiscal 2005 andfiscal 2006, as a share of funds obligated for sexual transmission prevention

Earmark

200

5

200

6

Nami

bia

Rwan

da

Ethiop

ia

Zamb

ia

Botsw

ana

Tanz

ania

South

Afric

aKe

nya

Moza

mbiqu

e

Cte

dIvo

ire

Guya

naHaiti

Ugan

da

Vietna

mNig

eria

Figure 3 Funds or abstinence and be aithul activities have increasedin most countriesthough most are still less than the earmark


12/27


Despite the controversy, little has been known

about how much money FBOs receive romPEPFAR or how it is spent.

FBOs have received a modest

share o PEPFAR unds

FBOs received on average 10% o 2005 obliga-

tions in the ocus countries (gure 4). Tat wasscarcely more than one-ourth the share obli-

gated to nongovernmental organizations (38%).And it was slightly smaller than the share obli-

gated to host country governments (13%), uni-versities (13%), or private contractors (12%).

Although the share o unds obligated to

FBOs has been modest overall, it has variedconsiderably across countries, ranging rom0.2% in Botswana to 31% in Kenya (see gure

4). Te reason may be that PEPFAR provides alarger share o its unds or FBOs to countries

where FBOs already have the expertise and experience to handle PEPFAR unds. Supporting

this hypothesis, the two countries with by arthe greatest share o PEPFAR unds obligated

to FBOsKenya and Namibiaalso show a very large share o such unds going to loca

groups.

Almost hal o FBO unds have

been allocated to treatment

FBOs in the ocus countries have received, on

average, more unds or treatment than or careand prevention. reatment programs averagedalmost hal (46%) o obligations to aith-based

prime recipients in ocus countries in 2005compared with just one-h or prevention and

one-third or care (gure 5). But because FBOreceived just 12% o all PEPFAR unds, they

received only 16% o total treatment programobligations in 2005.

Obligations to FBOs by program area havealso varied widely among the ocus countries. In

ve countries the share o obligations to primerecipient FBOs going to treatment programs

exceeded 80%, while in our countries it waszero.

Te Bush administrations commitment to engage aith-based organizations (FBOs)

in its work has generated both ervent praise and erce criticism. Te role o FBOs in

PEPFAR program implementation has been particularly controversial. Some advo-

cates argue that PEPFAR should do more to engage FBOs, others that FBOs wield

too much inuence and limit the AIDS response.

W -z pepfar?IS

SUE2

0

10

20

30

40

Funds obligated to FBOs by country for fiscal 2005


Average

Botsw

ana

Vietna

m

Cte

dIvo

ire

Guya

na

Ethiop

ia

Tanz

ania

Rwan

da

Nigeri

a

Moza

mbiqu

e

South

Afric

a

Ugan

daHaiti

Zamb

ia

Nami

biaKe

nya

Figure 4 PEPFAR unds to FBOs have ranged widely


13/27


Most prevention unds obligated

to FBOs have gone to abstinence

and be aithul activities

Te vast majority o PEPFAR unds obligatedto aith-based groups or prevention88% on

average in 2005has gone toward abstinenceand be aithul (AB) activities, a ocus remark-

ably consistent across countries. Indeed, ineight o the 13 countries where FBOs received

2005 unds or prevention, all obligations wenttoward AB activities.

Because obligations to FBOs made up just4% o total 2005 obligations or prevention, it is

possible that other entities programs balancedor outweighed the FBOs ocus on AB activities.

Still, FBOs have staked out an AB niche within

prevention programming. I the reauthoriza-tion o PEPFAR removes earmarks, it will beinteresting to see whether unding to FBOs or

abstinence and be aithul activities declines.

Most FBO unding has gone to

international groups, especially

Catholic Relie Services

A purported benet o unding FBOs is thatthey are locally based, with deep roots in the

communities they serve. PEPFAR documentsoen assume such a benet.14 But in only ve

countries did more than hal o the 2005 undsobligated to FBOs go to local groups. And o

all 2005 unds obligated to FBOs, just 38% on

average went to local groups (gure 6).15A single international FBO, Catholic Relie

Services (CRS), accounted or about $49 mil-

lion o 2005 PEPFAR unds, or 46% o totalobligations to FBOs.16 Te next largest FBO

recipients were the Kenyan-based Mission orEssential Drugs and Supplies ($20 million) and

the international organization World Vision($8 million).

Recommendation: Reevaluate

PEPFARs ocus on FBOs

PEPFAR oen deends its ocus on FBOs asa means o support or local groups with deep

roots in their communities. However, the

14. See or example PEPFAR (2005).

15. Since the authors lack inormation about subgranting

rom rack 1 grants, they lack complete in ormation about the

designation o rack 1 subgrantees. o estimate the amount o

unds going to local FBO subrecipients o rack 1 unds, they

have assumed that the same percentage o rack 1 unds as

country unds are subgranted to FBOs, and they have assumedthat the same percentage o rack 1 subgranted unds to FBOs

as country subgranted unds to FBOs go to local recipients.

16. Because CRS was the recipient o substantial rack 1

unding in 2005, and the authors lack inormation about the

proportion o rack 1 unding transerred to subrecipients,

they have estimated this proportion by taking the percentage

o CRSs country-managed unds that were subgranted and

applying that percentage to its rack 1 grants . Te net amount

o unding obligated to CRS is thereore an approximation

based on the best available inormation.

Average funds obligated to faith-based prime recipients by program area for fiscal 2005

Prevention

20.3%Care33.8%

Treatment45.7%

Other 0.1%

Figure 5 FBO prime recipients have received almosthal their unds or treatment programs

0

25

50

75

100

Funds obligated to local and international FBOs by country for fiscal 2005

Average

Unknown

L

ocalInternational

Tanz

ania

Nami

bia

South

Afric

aKe

nya

Nigeri

a

Rwan

da

Cte

dIvo

ire

Ugan

da

Zamb

ia

Ethiop

iaHa

iti

Moza

mbiqu

e

Vietna

m

Botsw

ana

Guya

na

Figure 6 Most countries have been providing less thanhal o their FBO unds to local groups


14/27


authors nd that only a small share o PEPFAR

unding or FBOs goes to local groups. PEPFARshould reevaluate its policy o targeting FBOs asa way to reach local groups, and should instead

seek more eective strategies or achieving that

goal. Such strategies should reect individualcountry contexts and the variety o local actorsthat provide services.


15/27


ISSUE3

As the worlds largest AIDS donor, PEPFAR

should be especially sensitive to the possible

advantages and disadvantages o sending moneythrough government. I it provides unds to thegovernment, PEPFAR must seek ways to pre-

vent the delays that have aected other donorsunding. I its money is not channeled through

the government, PEPFAR must nonethelesshelp the government to steward the national

AIDS response.1718

17. Te importance o government stewardship in coor-

dinating the AIDS response is widely accepted. See WHO

(2000).

18. Oomman, Bernstein and Rosenzweig (2007).

Many donors channel their AIDS unding through the government, believing that

such disbursement will build public capacity and support government stewardship.17

Despite these potential benets, some host country governments have been slow to

use the AIDS unds they receive.18

W v pepfar?

A modest share o PEPFAR unds

has been obligated to host country

governments

O all the 2005 money obligated to prime recip

ients in ocus countries, just 13% went to hostcountry governments (gure 7). Only Botswa

nas government (with 41%) received signi-cantly more than one-quarter o the unds obli

gated to its country.19

Although the share o PEPFAR unds going

to governments has been small, PEPFARs sheersize means that the absolute amount a govern

ment receives rom PEPFAR can rival the amounit receives rom other donors. For example, PEP

FAR gave 11% o its 2005 unds or Uganda tothe Ugandan governmentan absolute amount

o $14.5 million. In 2005 the Global Fund orAIDS, uberculosis and Malaria disbursed abou

$20 million to the Ugandan government.

More than hal o host country

government unds have been

allocated to prevention

On average, more than hal (57%) o the

amount obligated to host country governments as prime recipients has been provided

or preventionmuch o it or blood saety

19. Governments receive PEPFAR unds as both prime recip

ients and subrecipients. Some, such as Botswana, receive mos

or all o their unds as prime recipients. Others, such as Nige

ria, receive the vast majority o their unds as subrecipients.

0

10

20

30

40

50

Funds obligated to host country governments by country for fiscal 2005

Average

Zamb

ia

Ethiop

ia

South

Afric

aHa

iti

Keny

a

Rwan

da

Ugan

da

Guya

na

Nami

bia

Nigeri

a

Cte

dIvo

ire

Moza

mbiqu

e

Tanz

ania

Vietna

m

Botsw

ana

Figure 7 Host country governments have received on average13% o PEPFAR unds obligated to their countries


16/27


Governments have also received modest but

signicant unding or treatment and care.

Botswana is an exception

Botswana is the one major exception to the pat-

terns just described. O all 2005 PEPFAR undsobligated to Botswana, 41% went to the govern-

ment. A large share o that amount was allocatedto two program areas that PEPFAR rarely unds

or governments: antiretroviral drugs (38%) andcondoms and other prevention (43%).20

Botswana has an innovative public-privatepartnership to ght HIV/AIDS. Its government

is oen credited with an unusually high capac-ity, compared with its neighbors, or stewarding

a national AIDS response. Do these acts help to

explain why the Botswana government receivedsuch a high share o its countrys 2005 PEPFARunds? Further investigation is needed.

Recommendation: Make the

government a true partner

in PEPFAR programs

Although PEPFAR need not channel a large share

o its unds through governments, it should ensurethat its programming helps the government ulll

its role as steward o the national AIDS response.Because most PEPFAR unds go to nongovern-

mental entities, coordination between PEPFARand the government is critical.

PEPFAR should work closely with hostcountry governments, enabling them to help

20. Only two other host country governmentsSouth

Arica and anzaniareceived PEPFAR unding or either

antiretroviral drugs or condoms and other prevention. Com-

pared with Botswana, they received a much smaller share o

their overall PEPFAR unds or these program areas.

plan and oversee PEPFAR-unded activities.

And it should provide governmentsas well

as other key stakeholderswith up-to-dateinormation on its programs. imely inorma-tion sharing and joint planning will help gov-

ernments coordinate their PEPFAR programswith the myriad other AIDS activities in their

countries.21

21. PEPFAR country teams work with host country gov-

ernments. Te Ofce o the U.S. Global AIDS Coordinator

encourages PEPFAR country teams to include host country

government sta in program planning. Te Country Opera-

tional Plan also must be approved by the host country gov-ernment beore it can be submitted to the Ofce o the U.S.

Global AIDS Coordinator. Nevertheless, research by the

authors o this report in three countriesMozambique,

Uganda and Zambiaindicates that, despite the best inten-

tions o PEPFAR sta, host country governments are rarely

included in PEPFAR planning and that their approval o the

annual Country Operational Plans is mostly symbolic. Te

authors believe that PEPFAR could go much urther in mak-

ing the government a true partner. PEPFAR has told them

that one way PEPFAR plans to strengthen partnerships with

governments over the next ve years is by developing com-

pacts meant to ormalize the relationship between PEPFAR

and other country-level stakeholders.

Average funds obligated to host country governments by program area for fiscal 2005

Prevention57%

Care14%

Treatment18%

Other11%

Figure 8 Host country governments have received morethan hal their unds or prevention programs


17/27


Discussions o PEPFARs commitment to capac-

ity building have oen suered rom the absence

o inormation about how PEPFAR interactswith local organizations. Although the Ofce othe U.S. Global AIDS Coordinator publishes its

obligations or each country by prime recipient,it does not publicly disclose the amounts obli-

gated to subrecipients. Without such inorma-tion it has been impossible to know how much

money was reaching local organizations.22

22. []he continuing challenge or the U.S. Global AIDS

Initiative is to simultaneously maintain the urgency and

intensity that have allowed it to support a substantial expan-

sion o HIV/AIDS services in a relatively short time while

also placing greater emphasis on long-term strategic planning

and increasing the attention and resources directed to capac-

ity building or sustainability (Institute o Medicine 2007).

PEPFAR has made considerable progress toward meeting its ve-year targets by rap-

idly scaling up services. However, as the Institute o Medicine argued in its 2007

evaluation, PEPFARs next phase must work to build local capacitymoving rom

an emergency response to a sustainable response.22

h pepfar ?IS

SUE4

In the dataset accompanying this report

amounts obligated to PEPFAR subrecipients

are being published or the rst time. Tis per-mits an analysis o all unds obligated to localas opposed to international, recipients.

O total obligated PEPFAR unds,

30% on average has gone to local

recipientsbut with considerable

variation across countries

O total 2005 PEPFAR unds, 30% on average was obligated to local recipients (gure 9).2

Tat percentage is strikingly low.24

But the shareo unds going to local recipients has varied by

country. In 11 countries no more than onethird o 2005 PEPFAR unds were obligated

to local groups, with Zambia (13%) and Ethio-pia (6%) at the low end o the range. Yet in our

countriesBotswana, South Arica, Namibiaand Ugandabetween 45% and 55% o PEPFARunds were obligated to local organizations.

23. Since the authors lack inormation about subgrantin

rom rack 1 grants, they lack complete inormation abouthe designation o rack 1 subgrantees. o estimate the

amount o unding going to local subrecipients o rack 1

unds, they have assumed that the same percentage o rack

1 unds as country unds are subgranted, and that the same

percentage o subgranted rack 1 unds as subgranted coun

try unds go to local recipients.

24. PEPFAR oen cites the act that 87% o its partner

are local groups. But the authors calculations here make the

point that 87% o partners receive roughly 29% o total unds

obligated to the ocus countries.

0

25

50

75

100

Funds obligated to local and international recipients by country for fiscal 2005

Average

Unknown

Domestic

Internationa

l

Nami

bia

South

Afric

a

Ugan

da

Botsw

ana

Keny

a

Vietna

mNig

eria

Haiti

Tanz

ania

Rwan

da

Guya

na

Moza

mbiqu

e

Cte

dIvo

ire

Zamb

ia

Ethiop

ia


Figure 9 PEPFAR unds to local recipients range rom 6% to 55%


18/27


PEPFARs initial mandate suggests a pos-

sible explanation or the large share o undinggoing to international organizations. PEPFARwas designed in 2003 to respond to the emer-

gency created by rapid HIV/AIDS transmis-sion. In the rst ew years o unding its sta

channeled money to organizations that couldquickly begin implementing programs. Te

bulk o early unds thereore went to interna-tional organizations that had already made

unding agreements with the U.S. governmentand knew how to comply with its procedures. 25

I new recipients were engaged, PEPFAR sta preerred those that had demonstrated their

ability to manage and use large sums o money.In many countries most o the organizations

that met these criteria were international

groups.In some countries PEPFAR channels a sig-

nicant share o its unds to local organizations,

likely because o those countries higher localcapacity. In Uganda unding was almost evenly

split between local and international organi-zations. As detailed in a previous HIV/AIDS

Monitor analysis,26 Uganda has many local or-ganizations with high capacity or using AIDS

unding and with U.S. government connectionsthat preceded PEPFAR. Such groups were well

positioned to receive signicant amounts romthe new program.

Circumstances in Uganda may help to ex-plain the overall pattern. Te three other coun-

tries where roughly hal o PEPFAR undinggoes to local groupsBotswana, South A-

rica, and Namibiaare all thought to havelocal organizations with strong capacity orimplementing AIDS programs. PEPFAR doc-

uments explain the correlation between thecapacity o local organizations and their abil-

ity to obtain PEPFAR unds. One PEPFARdocument notes that many local indigenous

organizations have limited technical expertisein accounting, managerial and administrative

skills, auditing practices and other activitiesrequired to receive unding directly rom the

25. Oomman, Bernstein, and Rosenzwei g (2007).

26. Oomman, Bernstein, and Rosenzwei g (2007).

U.S. Government.27 A closer examination o

the issue would provide more conclusive evi-dence about PEPFARs approach to support-ing local capacity where it exists, or building it

where it does not.

O unds obligated to subrecipients,

more than hal on average has gone

to local organizationsbut local

and international subrecipients

combined have received just

19% o total obligated unds28

o increase the capacity o local organizations

as part o an overall eort to make programunding more sustainable, PEPFAR encourages

its current prime recipients to work with localsubrecipients and to build their capacity so that

they can become recipients.29 One part o thisstrategy seems to be working: a signicant share

27. PEPFAR (2008), p. 6.

28. Since the authors lack inormation on subgranting rom

rack 1 grants, all analysis o subgranting is based solely oncountry-managed unds (see annex A, note 35).

29. PEPFARs 2008 guidance to its country teams or devel-

oping Country Operational Plans states that: Eorts should

be made to support and provide technical assistance to assist

these indigenous organi zations [local subrecipients] to gradu-

ate to ull partner status and enable them to be direct recipi-

ents o PEPFAR unds. Te duciary accountability o local

organizations is essential to building sustainable capacity;

technical assistance in this area should be made available to

partner organizations. PEPFAR (2008), p. 6.

Average country-managed funds obligated to local and international subrecipients in the focus countriesfor fiscal 2005

Local55%

International42%

Unknown 3%

Figure 10 O subgranted unds, 55% on average hasbeen obligated to local subrecipients


19/27


o unding to subrecipients, well over hal (55%)

has gone to local organizations (gure 10). 30But unding data alone cannot tell whether

PEPFAR is building local capacity. Only 19%

o PEPFAR unds have been subgranted atall, meaning that local subrecipients have re-

ceived on average just 11% o PEPFAR undsin the ocus countries (gure 11). So, even i

30. Althoug h on average 55% o subgranted unds are going

to local subrecipients, there is strong variation between coun-

tries. In Namibia 95% o subrecipient unds went to local

organizations, in Guyana only 10%.

subgranting has been building the capacity o

subrecipients, it seems unlikely, given the mod-est amounts they are receiving, that they willsoon be ready to receive a signicant share o

the PEPFAR unds now provided to interna-tional recipients.

Recommendation: Expand eorts to

build the capacity o local groups

Although PEPFAR already invests signicantsums in capacity building activities, it should

reexamine those eorts. It should seek waysto build the capacity o local organizations so

that they can receive greater shares o PEPFARunding.31

PEPFARs current approach to capacity

buildingengaging local subrecipientsisnot sufcient, since only a small share o totalunds is being subgranted. Building the capacity

o local organizations is a critical sustainabilityaim, as important as achieving PEPFARs global

targets.

31. PEPFAR gures indicate that $638 million o the money

it obligated in 2007 was devoted in ull or in part to capac-

ity building activities. Personal communication with U.S.

Ofce o the Global AIDS Coordinator ofcial, February 15,

2008.

Average country-managed funds obligated to prime recipients and subrecipients in the focus countriesfor fiscal 2005

Prime recipients81%

Subrecipients19%

Figure 11 O total country-managed unds, just 19%on average has been subgranted


20/27


PEPFAR should begin to regularly disclosemore data about its unding awards, including

the type o data used or this report. PEPFARalready collects a wide range o valuable data

(see annex C). Such data should be publiclydisclosed.

Not only PEPFAR, but also other AIDS do-nors should disclose more o their unding data.3

Such disclosure could clariy important policydebates, strengthen coordination among various

AIDS stakeholders, and improve assessments oprogram impact to make aid more eective.

32. For more detailed recommendations about what und

ing data PEPFAR, the Global Fund, and the World Bank

Arica Multi-Country HIV/AIDS Program (World Bank

MAP) should release, see Oomman, Bernstein and Rosenz

weig (2007).

Tis report has analyzed unding data related to PEPFAR grants and contracts. It

has used these unding data to discuss our issues pertaining to PEPFARs eec-

tiveness and sustainability. And it has recommended actions to improve PEPFAR

programming. Te authors o the report hope that other researchers and interested

parties will use the data presented here and in the accompanying dataset to urther

clariy key issues.

f


21/27


Te data here are drawn rom the Coun-

try Operational Plan and Reporting System(COPRS), an accounting and program moni-

toring system maintained by the Ofce o theU.S. Global AIDS Coordinator. Te dataset

includes amounts obligated to prime recipi-ents and subgranted to subrecipients, with

inormation on recipient type (government,

nongovernmental, aith-based, private or-prot), unding year (year o appropriations), year o obligation, and program area (treat-

ment services, blood saety, or the like) oreach recipient, except that obligations to sub-

recipients do not include inormation aboutprogram area.

Te dataset was originally obtained romthe State Department by the Center or Public

Integrity through several Freedom o Inorma-tion Act requests and a lawsuit against the U.S.

Government, settled out o court.33

Noticingsome errors in the dataset, the Center or Public

Integrity was hesitant to release it publicly, butgenerously agreed to share it with the authors

o this report.Aer investigating the possible causes o

errors (see data limitations below), the Cen-ter or Global Development was satised thatthe overall quality o the dataset had not been

compromised. Individual errors in the data werenot signicant enough to inuence the overall

unding patterns described and analyzed in thisreport.

Te dataset being released in conjunctionwith this report builds on the Center or Public

Integrity data in three ways. First, the majorityo prime recipients and subrecipients have been

classied as either local or international,based on whether their headquarters are in the

33. See Bengtsson and Morera (2006).

host country or abroad. 34 Tese classication

were made through research by Center oGlobal Development sta and consultations

with in-country experts.Second, data on centrally awarded, or

rack 1, unding have been added.35 Te centrally awarded unding gures or 2004 and

2005 come rom data previously published by the

Center or Public Integrity, which had obtainedthem rom the State Department.36 Using thosedata, the authors o this report estimated cen

trally awarded unding gures or 2006 based onpast unding trends and the amounts allocated

to each ocus country or 2006.37

Finally, the authors used the program areas

listed in COPRS to deduce classications byPEPFAR strategic area. Users can now search

34. Almost all organizations that received unds in sca2004 through 2006 were classied by geographic origin. Fo

a ew countries, however, classications were made only o

recipients and subrecipients receiving unds appropriated in

2005 (the year o most o the data examined or this report).

35. Centrally managed (rack 1) undi ng is granted directly

rom the Ofce o the U.S. Global AIDS Coordinator in

Washington to organizations to implement programs in

country. It diers rom country-managed unding, which i

granted to recipient organizations through the eld ofce

o PEPFAR-implementing agencies in the host country. Te

great majority o PEPFAR unds are country-managed.

36. Tis dataset can be downloaded rom Bengtsson and

Morera (2006).

37. Te authors estimated the total amount obligated to each

ocus country or 2006 by taking the average ratio between

obligations and allocations or 2004 and 2005, then multiply

ing this ratio by the 2006 allocation gures or each country

Using this estimated obligation amount or 2006 they cal

culated the percentage increase in total obligations to each

country between 2005 and 2006. Finally, they applied the

percentage increase or each country to calculate the tota

estimated amount obligated to each individual rack 1 und

ing recipient or 2006.

d AnnEx

a


22/27


or unds going to prevention, treatment, care,

and other costs.

Data limitations

Te dataset has several limitations. First, it

contains data on obligations (commitments),not outlays (expenditures). So, it shows how

much was available to recipients to implementprogramsbut not how much money was actu-

ally spent.38

Second, because o time lags in the appropria-

tions process, the data do not necessarily reectthe total amount obligated to each recipient using

2006 unds. A signicant part o PEPFAR moneyappropriated in a given year is not obligated until

the ollowing yearsometimes even later. Be-

cause the dataset continues only through 2006,the authors lack data on unds appropriated byCongress in 2006 but obligated by PEPFAR in

2007. Te data thereore underestimate total obli-gations made using money appropriated in 2006.

Because some unds are not obligated until two ormore years aer the appropriation year, the data-

set also misses some obligations o unds appro-priated in 2004 and 2005. Te amount o such

delayed obligations is very small, however, com-pared with the amount that is obligated in the ap-

propriation year and the year ollowing it.39

Tird, since inormation on unding by pro-

gram area is available only or prime recipients,it is not always possible to know exactly how

much subrecipients plan to spend in particularprogram areas (though educated guesses can be

made rom the data listed or prime recipients).40

38. For details on the dierences between obligations and

outlays, see Bernstein and Sessions (2007).

39. Te percentage o money appropriated in 2004 and 2005

and obligated in 2007 or later would be very small because

almost al l unding appropriated in one year is obligated by theconclusion o the subsequent year. For example, data obtained

by the HIV/AIDS Monitor rom PEPFAR shows that 99% o

2004 unds were obligated by the end o 2005.

40. Tis limitation applies to scenarios in which a prime

recipient has been unded or multiple program areas. For

example, Catholic Relie Services received unding in 2005

or programs that would provide palliative care and support

or orphans and vulnerable children in Ethiopia. We cannot

disaggregate the unding provided by Catholic Relie Services

to its subrecipients by these two program areas.

It is principally or this reason that parts o the

analysis are ocused exclusively on unding toprime recipient organizations, without includ-ing subrecipient organizations.

Fourth, although PEPFAR requires primerecipients to report on all o their subrecipients,

several ormer PEPFAR ofcials shared their be-lie that some prime recipients do not always do

so. Since it is unclear whether prime recipientshave reported comprehensively on all subrecipi-

ents, some subrecipients may not be listed in thedataset.

Fih, three data elds are unavailable orcentrally awarded (rack 1) unds: Obligation

Year, MechID, and OrgID (see annex B ordetails on each data eld). o remedy this, the

Center or Global Development is working to

include OrgIDs or rack 1 unds in uture ver-sions o the database. For this report the authorsassume that rack 1 unds were obligated in the

year they were appropriated.Sixth, the dataset is limited to obligations

to the 15 ocus countries. Tese countries col-lectively receive the vast majority o PEPFAR

unding. But according to the U.S. Global AIDSCoordinator, more than 100 other countries

benet rom the PEPFAR program, togetherreceiving roughly 12% o PEPFAR unds.41

Data errors

Te data contain some easily detected errors.

For example, some recipients have been clas-sied under the wrong organization type. In

one instance the United Nations DevelopmentProgramme in Vietnam was classied as a non-governmental organization rather than a multi-

lateral agency. Such misclassications, probablyreecting simple human error, seem to have

occurred only in a small number o cases.Another problem, originally pointed out by

the Center or Public Integrity, is that subrecipi-ents occasionally are listed as having been sub-

granted more money under an individual grantthan was obligated to their respective prime re-

cipients. For example, in 2005 in Nigeria, thedataset shows that John Snow International, a

41. Dybul (2007).


23/27


prime recipient, was obligated $490,000 under

a particular grant, while its subrecipients arelisted as receiving $1,515,682 under the samegrant. Such errors seem to have occurred in only

a small number o cases, representing roughly3% o the grants in the dataset.42 Tey may arise

partly rom simple human errorand other ex-planations are possible. Prime recipients may be

covering the immediate costs o subgranting byborrowing unds rom other grants. Or prime

recipients may be subgranting beore they areobligated the ull value o a particular grant,

42. In a dataset containing inormation about more than

2,600 grants over three years, the authors ound just 76

errors o this type.

in the expectation that they will receive the re-

mainder in uture years.Whatever the cause o each error identied,

the authors o this report deem the datasets

overall quality and reliability high enough toallow or useul, in-depth analysis o unds ob-

ligated to the ocus countries. Te authors havenot attempted to correct any possible dataset

errors. Instead they have denoted the placeswhere errors might exist by adding a column to

the dataset in Microso Excel. Te authors en-courage people amiliar with grants identied

as having possibly erroneous gures to contactthem with inormation that might clariy the

gures. Te dataset will be updated to reectany new inormation.


24/27


Te database o PEPFAR unding or scal

20042006accessible at http://www.cgdev.org/hivmonitor/PEPFARdataprovides pre-

viously unknown details about how PEPFARmoney is allocated in the 15 ocus countries.

Tese data reveal many interesting acts aboutPEPFAR programming.

o understand the data one must have a

basic knowledge o key steps in the PEPFARbudget process (gure B1).43 o begin that pro-cess PEPFAR submits an annual budget request

to Congress or the next scal year. Aer Con-gress approves a unding amount, the Ofce o

the U.S. Global AIDS Coordinator nalizesits annual budget or country-managed unds

as well as centrally managed unds. (Country-managed unding is money distributed rom

U.S. Government in-country eld ofces to prime recipient organizations to operate pro-

grams in PEPFAR countries. Centrally man-aged, or rack 1, unding is money granted to

prime recipient organizations directly rom theOfce o the U.S. Global AIDS Coordinator in

Washington.)Aer a PEPFAR scal year budget is nal-

ized, the money must be obligated to PEPFAR

43. For more inormation, see Bernstein and Sessions (2007).

u AnnEx

b

recipients. Tis may or may not occur in the year

a grant is approved. An obligation is denedas money available to a recipient organization

either or withdrawal in advance o projecexpenditures (in a grant or cooperative agree

ment) or as reimbursement or expenditureswithin 30 days o submitting an invoice (in a

contract). PEPFAR obligations are technically

legal commitments to pay now or in the utureBut because the unds are made available or useby prime recipients as soon as they are needed

the obligations are in practice less like commitments than like disbursements rom other

unding agencies.Tis database provides inormation on PEP

FAR obligations or unding approved in scayears 2004, 2005, and 2006 and obligated in

scal years 2004, 2005 or 2006.Te database has several elds that allow

users to analyze the data along several dimensions. able B1 denes and describes each data

eld.Te issues discussed in this report are just

a ew o those that can be explored using thedata, which can be cut in many other ways to

clariy how PEPFAR money is allocated in itsocus countries.

Given the multiplicity o elds, investiga

tors can sort the data by any combination ocountry, unding year, obligation year, recipi

ent organization, organization type, programarea, geographic origin (domestic or interna

tional), individual grant, or individual recipient. For example, one could lter the data to

learn how much money was obligated to aith-based organizations in Uganda in 2005 or to

locally based nongovernmental organizationin Haiti in 2006. One could learn how much

money has been obligated to John Snow Inter-national or antiretroviral treatment. Or one

1

Congress appropriatesmoney to particular agencies or programs as part of the annual federal budget

2

PEPFAR implementing agencies (such as USAID) obligatemoney to prime recipientorganizations; prime recipients then obligate funds to subrecipient organizations

3

The U.S. Treasury outlaysmoney to recipient organizations

4

Prime recipients expendmoney or subgrantto subrecipients

This report

analyzes data

on obligations

Figure B1 Legal processes or transerring PEPFAR unds


25/27


could compare the unds obligated to Zambia

in 2006 or treatment with the unds obligatedor prevention.

We hope that examining the data in these

and other ways will prove useul to various au-diences including researchers, donors, undingrecipients, and advocacy groups.

Data feld Description

Funding year The U.S . Government fscal year ( rom October 1 t o September 30) when unding was approved by the U .S .

Congress.

Obligation yeara The year when unding was obligatedusually the same as the unding year, but sometimes a later year.

Funding source The und ing mechani sm ob li ga ting the unds. PEPFAR unds are ob li ga ted by e it her feld t eams ( country funds)or the Washington-based PEPFAR team (central funds, synonymous with Track 1 funds). PEPFAR dierentiates

between these two sources in most or all o its documentation.

Country name Country receiving a PEPFAR unding obligation.

MechIDb Unique ID given to a particular grant mechanism in the PEPFAR COPRS database. The MechID denotes a single

grant to a prime recipient and any subgrants associated with it.

OrgIDc Unique ID given to a particular prime recipient or subrecipient.

Prime partner name Name o t he pr ime rec ip ient associated wi th the grant mechan ism.

Recipient name Name o the obligation recipient. I the prime recipient is the recipient, the prime partner name and recipient

name will be the same or a particular obligation. I the obligation is to a subrecipient, the recipient name will be

the name o the subrecipient.

Recipient origin Geographic origin o the obligation recipient (domest ic or internat ional). The Center or Global Development has

researched the origins o recipients that were obligated unds in 2005 and o some that were obligated unds in

2004 and 2006.

Recipient type Category o the recipient receiving the obl igat ion (such as nongovernmental organizat ion, universi ty, ai th-based

organization, or private contractor).

P rimeOrSub Method by which sa id recipi en t recei ves the obl igat ion (pr ime = di rect ly rom a U.S. Government agency;

sub = subgrant rom a prime recipient that received the unds rom a U.S. Government agency).

Obligated amount Amount obligated to the recipient organization.

Net obligation Amount obligated to the recipient organization net o subgrants.

Strategic aread The PEPFAR strategic area (prevention, care, treatment or other) under which an obligation alls.

P rog ram a rea Prog ram a rea tha t the obliga tion t arge ts (one o 15 PEPFAR p rogram a reas ).

MechID with errors? This feld denotes possible errors that CGD has identifed in the data, such as grant mechanisms where the

amount obligated to subrecipients exceeds the amount obligated to the prime recipients under this mechanism.

See annex A or details and a discussion o the possible errors.

a. Not available or Track 1 grants. This report assumes that the obligations to Track 1 partners are made in the year when unds are appropriated.

b. Not available or Track 1 grants.

c. Track 1 unds in the database are not tagged with an OrgID. The Center or Global Development is working to include OrgIDs or Track 1 grants in the uture.

d. Program area and strategic area are available only or grants to prime recipients. The database does not have this inormation on subrecipients.

Source:Authors analysis o data as described in text.

Table B1 Data felds used


26/27


able C1 presents all o the unding data,and related inormation about grants and

contracts, that PEPFAR or other parts o theU.S. government have collected. It presents

each type o data in a new row. Te threetable columns describe whether data were

publicly available prior to the release o thisreports dataset, whether they are contained

in the dataset, and where to nd them.

Boxes shaded in yellow show data that were not previously

public but that were released in the dataset accompanying

this report.

Boxes shaded in green represent data available in this

reports dataset that were publicly available beore the

release o this report, but were difcult to access and use,

being presented as part o the thousands o pages o PDF

documents that make up annual country operational plans.

Boxes shaded in grey represent data that are publicly

available but difcult to access and use, and that are not

presented as part o the dataset accompanying this report.

Boxes shaded in blue represent data that were available

beore this report and are easily accessed on the PEPFAR

website.

Boxes shaded in red represent data that continue to be

publicly unavailable, despite being collected by the U.S.

Government.

f pepfarAnnEx

c

Publicly available rom PEPFAR beore the

release o this report and its accompanying

dataset or fscal 2004 to 2006?

Released as part o this report

and its accompanying dataset

or fscal 2004 to 2006? Where to fnd the data

General information about grants or contracts

Type o recipient (aith-based

organization, nongovernmental

organization, government, and so on)

Yes Yes Dataset released with this report, available at

[http://www.cgdev.org/hivmonitor/PEPFARdata]

Funding source (country unded

or centrally unded)


http://www.cgdev.org/hivmonitor/PEPFARdata

Geographic origin o organization

(local or international)



Program area (treatment, abstinence,

home-based care, and so on)

Yes Yes PEPFAR website at http://www.PEPFAR.gov

Activity area ( training, logistics,

human resources, and so on)

Yes No Country operational plans or each country

available at http://oia.state.gov/COP.asp

Regions and provinces covered

by each grant and contract



Planned indicators and targets or

each grant and contract



Regular updates on progress made toward

indicators and targets or each grant and contract

No No Semiannual and annual reports submitted by

PEPFAR country teams to the Ofce o the U.S.

Global AIDS Coordinator (not publicly available)

Target population (pregnant women,

health care workers, and so on)



Obligations

To each ocus country Yes Yes PEPFAR website at http://www.PEPFAR.gov /partners/

To each prime recipient Yes Yes PEPFAR website at http://www.PEPFAR.gov /partners/

To each prime recipient,

disaggregated by program area

No Yes Dataset released with this report, available at


To each subrecipient No Yes Dataset released with this report, available at


Outlays

To each prime recipient No No PSC 272 orms submitted to the U.S. Treasury

by each recipient [not publicly available]

Expenditures

To each prime recipient No No PSC 272 orms submitted to the U.S. Treasury

by each recipient [not publicly available]

Source:Authors analysis o data as described in text.

Table C1 Funding data collected by PEPFAR and where to fnd it


27/27

Bengtsson, Helena and Alejandra Fernan-

dez Morera. 2006, November 30. Diine

Interention: About this Data. Washing-

ton D.C.: Te Center or Public Integ-rity. http://www.publicintegrity.org/aids/

report.aspx?aid=816.Bernstein, Michael and Myra Sessions. 2007.A

rickle or a Flood. Washington, D.C.: Cen-

ter or Global Development.Bernstein, Michael and Sarah Jane Hise.

2007. PEPFAR Reauthorization: Improv-ing ransparency in U.S. Funding or HIV/

AIDS. Washington, D.C.: Center or

Global Development.Dybul, Mark R. 2007, November 30. Ask the

White House. Washington, D.C.: TeWhite House. http://www.whitehouse.gov/

ask/20071130.htmlInstitute o Medicine. 2007. PEPFAR Imple-

mentation: Progress and Promise: Report onthe Ealuation o PEPFAR Implementation.

Washington, D.C.: Te National Acad-emies Press.

Oomman, Nandini, Michael Bernstein and

Steven Rosenzweig. 2007. Following theFunding: A Comparative Analysis o the

Funding Practices o PEPFAR, the GlobaFund and World Bank MAP in Mozam-

bique, Uganda and Zambia. WashingtonD.C.: Center or Global Development.

Presidents Emergency Plan or AIDS Relie

(PEPFAR). 2005, September. PEPFARCommunity and Faith-Based Organizations (September 2005). Washington

D.C.: PEPFAR. http://www.pepar.gov/progress/76864.htm

Presidents Emergency Plan or AIDS Relie(PEPFAR). 2008. Country Operationa

Plan Guidance or Fiscal 2008. Mimeograph. Washington, D.C.: PEPFAR.

WHO (World Health Organization). 2000Te World Health Report 2000: Health

Systems: Improving Perormance. GenevaWorld Health Organization.

r

the numbers behind the stories

Documents