ARTICLE IN PRESS

0149-7189/$ - se

doi:10.1016/j.ev

�CorrespondE-mail addr

kari.hartwig@y

(M. Merson).1For orphans

both parents, an

to include ch

threatened by

Evaluation and Program Planning 31 (2008) 51–60

www.elsevier.com/locate/evalprogplan

Government–NGO collaboration and sustainability of orphans andvulnerable children projects in southern Africa

Alana Rosenberg�, Kari Hartwig, Michael Merson

Center for Interdisciplinary Research on AIDS, Yale University School of Medicine, 135 College Street, Suite 200, New Haven, CT 06510, USA

Abstract

Given current donor attention to orphans and children made vulnerable by HIV/AIDS, and the need for a new framework that

recognizes the complementary roles of nations and non-governmental organizations (NGOs), this analysis reviews NGO-operated

community-based orphans and vulnerable children (OVC) projects in Botswana, Lesotho, Namibia, South Africa, and Swaziland. There

has been a lack of attention within the field of evaluation to inter-organizational relationships, specifically those with government

agencies, as a factor in sustainability. We analyzed evaluations of nine OVC projects funded by the Bristol-Myers Squibb Foundation for

the influence of government–NGO collaboration on project sustainability. For eight of the nine projects, evaluations provided evidence

of the importance of the government partnership for sustainability. Government collaboration was important in projects designed to

help families access government grants, initiate community-based solutions, and advocate for OVC rights through legislation.

Government partnerships were also critical to the sustainability of two projects involved in placing children in foster care, but these

showed signs of tension with government partners. In addition to the more common factors associated with sustainability, such as

organizational characteristics, donors and NGOs should concentrate on developing strong partnerships with local and national

government agencies for the sustainability of their projects.

r 2007 Elsevier Ltd. All rights reserved.

Keywords: Orphans and vulnerable children; Sustainability; Government; AIDS; Africa; Non-governmental organizations

1. Introduction

In sub-Saharan Africa, more than 12 million childrenunder 18 years of age have been orphaned by AIDS(UNAIDS/WHO, 2004). These children face increasedeconomic, medical, nutritional, and psychosocial depriva-tion. They often drop out of school or attend irregularly;lose their inheritance rights; and suffer from stigma withinthe community (Global AIDS Alliance, 2005). Newfunding initiatives have recently scaled up the globalresponse to the problems facing orphans and vulner-able children (OVC),1 largely through community-based

e front matter r 2007 Elsevier Ltd. All rights reserved.

alprogplan.2007.08.005

ing author. Tel.: +1203 764 4353; fax: +1 203 764 8469.

esses: alana.rosenberg@yale.edu (A. Rosenberg),

ale.edu (K. Hartwig), michael.merson@duke.edu

, we include children who have lost their mother, father or

d children under 18 years of age. We use the term ‘‘OVC’’

ildren whose survival, well-being or development is

HIV/AIDS (UNAIDS/UNICEF/USAID, 2004). While

responses. This scale-up occurs within a current interna-tional aid culture that often polarizes non-governmentalorganizations (NGOs) and local governments, and assumescivil society has a comparative advantage to affect changeand reach vulnerable populations (Hartwig, 2001; Tvedt,2006). While this has contributed to a proliferation ofNGOs worldwide, it has also pitted governments againstNGOs as they compete for donor funds. As Tvedt (2006)points out, this theoretical positioning of the state versusthe NGO often neglects to take into account that manyNGOs are funded in part through the state and/or mustwork collaboratively with the state to perform theirmissions successfully.Given the current donor attention to orphans and the

need for a new framework that recognizes the complementary

(footnote continued)

the term ‘‘OVC’’ may not be appropriate at the local level because of its

potential to stigmatize, we use it here to correspond with national action

plans, which use the term.

ARTICLE IN PRESSA. Rosenberg et al. / Evaluation and Program Planning 31 (2008) 51–6052

roles of state and non-state actors, this analysis reviews theevaluations of nine NGO-operated community-based OVCprojects in Botswana, Lesotho, Namibia, South Africa,and Swaziland and specifically focuses on the theme ofgovernment–NGO collaboration and its influence onproject sustainability.

A range of definitions for sustainability has been devel-oped and analyzed (Cassidy & Leviton, 2006; Johnson,Hays, Center, & Daley, 2004; Olsen, 1998; Scheirer, 2005).We make use of a simple definition developed by Bossert(1990): ‘‘the continuation of activities and benefits achievedduring the project after the donor’s funding has ceased.’’Cassidy and Leviton note evaluators’ lack of attention to theinter-organizational relationships important to capacitybuilding and sustainability. Specifically, there has been alack of attention in the literature to partnerships that includegovernment agencies in relation to NGO project sustain-ability. An exception is Bossert’s review of US fundedprojects in Central America and Africa, which citesgovernment financing, integration into the national imple-menting agency, and a mutually respectful negotiatingprocess with governments as important for sustainability.We examine the issue of government partnerships in relationto NGO project sustainability using the example of OVCprograms in southern Africa.

(footnote continued)

2. Government responses for OVC in southern Africa

The countries of Botswana, Lesotho, Namibia, SouthAfrica, and Swaziland face some of the highest rates oforphanhood in the world. The percentage of children whowere orphans ranged from 12% (Namibia) to 20%(Botswana) in 2003 and this is expected to increase to18% (Namibia) to 24% (Botswana) in 2010. Theseincreases represent the largest UNICEF estimates oforphan growth in all of sub-Saharan Africa. In SouthAfrica in 2003, there were 2.2 million orphans, a figure thatis estimated to increase to 3.1 million by 2010. AIDS is amajor contributing factor to orphanhood in these coun-tries; the percentage of orphans who lost a parent to AIDSranged from 50% in South Africa to 75% in Botswana in2003 (UNICEF, 2005).

These five nations legislated OVC policies in 1999(Botswana), 2005 (Lesotho, Namibia, South Africa) and2006 (Swaziland).2 All policies call for the protection ofinheritance rights; medical care, nutritional support,counseling and psychosocial support; and the endorsementof community-based care. All except the Lesotho policydirectly address social grants, and all except the SouthAfrica policy explicitly mentioned the provision of schoolfees for OVC.3

2Republic of Botswana (1999). Government of South Africa (July 2005;

July 15, 2005); Government of Namibia (2005); The Kingdom of

Swaziland (2006); Lesotho Ministry of Health and Social Welfare (2005)..3The South African policy refers to the South African Schools Act No.

84 of 1996, which makes school attendance compulsory for learners

The extent of policy implementation, however, is anothermatter. For example, in Botswana, only 5000 of thecountry’s 10 000–20 000 orphans living with HIV/AIDS areon anti-retroviral therapy, and the country is lacking inHIV-testing facilities for children (IRIN News-Botswana,2006). The foster care system in South Africa is over-burdened and cumbersome. Although OVCs are entitled toa social grant until they reach the age of 14 years (USDepartment of State, 2006), only a small proportion ofeligible children receive these.4 Only 25 000 of Namibia’sestimated 150 000 OVC received a monthly grant from thegovernment.5 In Lesotho and Swaziland, rates of children’sprimary school attendance are dismally low, with 65% and70% of children attending, respectively (U.S. Departmentof State, 2006). Clearly, though governments are com-mitted to helping vulnerable children, they need theadditional support of NGOs, the private sector, andinternational donors to help all children access theresources and services to which they are entitled.

3. Description of the program

For the past 4 years, the Center for InterdisciplinaryResearch on AIDS (CIRA) at Yale University has servedas external evaluator to Bristol-Myers Squibb Foundation(BMSF)’s Secure the Future (STF) initiative. STF repre-sents a commitment of over $115 million to find sustainablesolutions for women, children and communities affected byHIV/AIDS in southern and western Africa. CIRA’sMonitoring and Evaluation Unit (MEU) was establishedthrough a grant from BMSF to provide evaluation servicesand training in monitoring and evaluation for STFgrantees. Local and international consultants with 5–10years of evaluation experience and knowledge of the NGOand health sectors, including HIV/AIDS, are contracted bythe MEU to conduct evaluations. Initial planning sessionswith project staff are followed by a site visit of 3–5 days.Site visits include document review and interviews withorganization leaders, staff, partners from other organiza-tions, and recipients of services. Focus groups withconstituents are also conducted. Evaluation reports areusually 30–60 pages and review the project’s strengths andweaknesses with respect to organization and managementas well as with respect to the achievement of objectives.Reports also provide recommendations and lessonslearned.Since 1999, STF has funded 69 projects in southern

Africa through the Community Outreach and Education(COE) fund. A request for proposals (RFP) mechanismallowed international, national, and community-based

between the ages of 7 and 15 years and provides for learners to be

exempted from the payment of school fees under certain conditions.4The government is working to address the problems faced by families

trying to access foster care grants. (IRIN News-South Africa, 2006).5The Namibian government has set up a trust fund to supplement the

monthly state grants. (IRIN News-Namibia, 2005).

ARTICLE IN PRESSA. Rosenberg et al. / Evaluation and Program Planning 31 (2008) 51–60 53

non-governmental organizations to compete for grants,and a Technical Advisory Committee (TAC) was convenedto review the proposals. The TAC included local andinternational HIV/AIDS experts as well as relevantgovernment, university, and NGO representatives fromall five southern African countries.

STF categorizes 29 of the 69 COE projects as includingservices for children affected by AIDS. Total funding forthese 29 projects equals almost $5 million. These projectsincluded psychosocial support; material support; trainingfor OVC, their caregivers, or organizations servingchildren; community mobilization for the care of OVC;and prevention of HIV/AIDS among youth. Some projectswere in response to a special RFP on OVC that wasdeveloped in 2001, which focused on three major areas ofunmet need: capacity building, civic education, andpsychosocial support.

The BMSF criteria for funding included, among others,the degree to which collaboration among different dis-ciplines and institutions would be established, and thedegree to which the project was consistent with governmentprograms. Ministries of Health received proposals toreview for alignment with national priorities and endorse-ment before final funding decisions were made (Secure theFuture, 2004). This practice is not uncommon amongdonors; however, it may account for the prominence ofNGO–government collaboration in the current review.

4. Methods

For the current analysis, we focused on a subset of STF-funded COE projects—those that were entirely focused oncaring for orphans and vulnerable children. Governmentcollaboration is particularly important for OVC work, asrecognition of the need to take action for OVC well-beingwas late in coming, and governments need the comple-mentary work of NGOs to put policies into effect.

We eliminated 12 of the 29 funded OVC projects becausethey were mainly prevention programs targeted at youth(6), general community HIV/AIDS programs only partiallyfocused on OVC (5), or a school-based health screeningprogram (1). We eliminated an additional 8 because noevaluations of them had been conducted (4) or they wereevaluated before January 2003 (4), as these evaluations hadlimited documentation about projects’ relationships withgovernment. The evaluations of the remaining nine projectsare the basis for this analysis.

The 9 projects reviewed represent a diversity of inter-ventions and organizations, but all have in common astrong emphasis on community-based solutions. Two faith-based providers of residential care for children, in Lesothoand South Africa, were funded to pilot community careapproaches (Sites 2 and 6). Two other faith-basedorganizations were funded to provide psychosocial supportto OVC, through day-care programs with communityoutreach in Botswana (Site 1), and camps and youth clubsin Namibia (Site 4). Four projects, in South Africa,

Swaziland, and Lesotho, aimed to build the capacityof communities and the organizations serving them tocare for OVC; 3 of these were implemented by localchapters of international organizations (Site 3 in Lesotho,Sites 8 and 9 in Swaziland) and the other was implementedby a partnership within one South African provincebetween CBOs, government and private sectors (Site 5).Lastly, an academic institution in South Africa was fundedto disseminate information about government grants andother social services (Site 7). Projects varied in length ofimplementation from 18 months to 3 years, and fundingamount from $55,000 to $273,000. Interestingly, length ofimplementation was inversely related to amount offunding. It is also important to note that sustainabilitywas not related to project duration or funding amount. SeeTable 1 for institutions, projects, and amounts funded.Seven of the 9 evaluated OVC projects had both an

interim evaluation and a final evaluation. Two had onlyinterim evaluations. The evaluations of these 9 projectswere carried out between March 2003 and July 2006 andserve as the core data for this analysis. We coded commentsin evaluation reports that were related to governmentcollaboration, major challenges, sustainability, and specificcontributions the project made to the care of OVC.Subcategories were also created. For example, withingovernment collaboration, we examined different types ofpartners and the nature of collaboration, the division ofroles between government and NGO, and the governmentrelationship as it pertains to project sustainability.A limitation of our methodology is that evaluators use the

term ‘‘sustainability’’ in reports without a previously agreed-upon definition. Shediac-Rizkallah and Bone (1998) categor-ize indicators of sustainability as 3 types: those that measurecontinued health benefits, those that measure institutionali-zation within an organization, and those that measurecontinued capacity of a community. The nature of theevaluations used for this analysis limits us to the secondcategory: institutionalization. While institutionalization maybe related to benefits for program participants, this link isnot guaranteed. For this analysis, we modified the version ofthe definition developed by Bossert (1990)—the continuationof activities and benefits achieved during the project after thedonor’s funding has ceased—to include only the continua-tion of activities, as concretely measuring outcomes forparticipants was beyond the directive of evaluators.We believe this definition is broad enough to include the

meaning of sustainability for our team of evaluators whilecapturing sustainability from the perspective of institutiona-lization. It is difficult for evaluators to ascertain how manyproject activities may be continued and for how long.Because evaluations took place during or at the completionof project funding, an evaluator’s comments about sustain-ability during or at the final stage of a project are based oninformed projections, but are nonetheless only projections.We contacted the leadership of the 9 projects in May

2006 to complete a short questionnaire about theircollaboration with government entities and about the

ARTICLE IN PRESS

Table 1

Description of the 9 orphans and vulnerable children projects evaluated

Site Organization Country Project funded by STF Amount No. of

years

1 Faith-based provider of

nationwide community counseling

services

Botswana Funded orphan day-care projects in two towns to integrate the

provision of psychosocial support to children through trained

volunteers and staff.

$178 904 2

2 Faith-based provider of residential

care for OVC

Lesotho Funded to establish an efficient foster care/adoption system

for babies in organization’s care; advocate for changes to

related national policies; and begin a community outreach

component to capacitate a community outside Maseru to care

for its OVCs.

$54 836 3

3 Local chapter of an international

children’s organization

Lesotho Funded to advocate for children’s rights in Lesotho through a

situation analysis, development of a directory of OVC service

providers, initiation of a coalition of NGOs focused on child

rights, training of NGOs in issues related to children’s

programs, and empowering children through child focus

groups in 6 villages.

$69 430 3

4 National Christian counseling

organization with expertise in

counseling children

Namibia Funded to provide psychosocial support to OVC through

establishing experiential learning camps and kids clubs and

training group leaders in communities. Activities based in

Windhoek.

$144 572 3

5 Partnership between CBOs,

government and private sectors

South

Africa

Funded an umbrella organization of CBOs in the Kwazulu-

Natal province to strengthen their care for orphans and

vulnerable children through identifying and supporting OVCs

with food, clothes, uniforms, school fees, foster care grant

access, income-generation activities for families caring for

OVC.

$272 644 1

6 Faith-based provider of residential

and community care for OVC

South

Africa

Funded to assist in the re-integration of children back into

their communities and provide long-term care to AIDS-

affected children, particularly during their teenage years, via a

home- and- community centered bridging program in Durban.

$166 177 3

7 Academic institution South

Africa

Funded to improve caregiver’s ability to access grants and

health care information in a peri-urban and rural area through

training of NGO and CBO employees and implementing

health camps to disseminate information from government

departments.

$227 777 2

8 Local chapter of international

training and research organization

Swaziland Funded to aid OVCs in ten communities by training mentors,

providing psychosocial support, income generation

opportunities, and educational grants and to develop a

directory of OVC service providers and carry out national

advocacy activities.

$215 616 2

9 Local chapter of an international

children’s organization

Swaziland Funded to establish Orphans and Vulnerable Children

Committees (OVCCs) in 4 regions to address issues pertaining

to children’s rights and OVC, and to establish peer education

and support networks for in-school and out-of-school OVC.

$107 457 3

A. Rosenberg et al. / Evaluation and Program Planning 31 (2008) 51–6054

sustainability of their STF-funded projects. For example,the survey asked which project activities the organizationcollaborated with government on for the STF-fundedproject. It also asked about the current status of theSTF-funded project. Four surveys were returned fromprojects in Lesotho, Namibia, South Africa, and Swazi-land. The low response rate may have been due to email asthe main method of communication and time lapsed sincefunding termination. The information shared was used tosupplement information available in evaluations.

5. Findings

In the following section, we present the findings of ouranalysis regarding the government collaboration and how

it relates to project sustainability. Please note thateven when evaluators found projects to be sustainableor found hope for sustainability, the project may have beenonly partially successful in degree or scope of achievement.Table 2 gives an overview of some key project achieve-ments and challenges noted by evaluators. NGOs trainedcaregivers and OVC, provided psychosocial and materialsupport, facilitated access to government grants and caredfor children’s daily needs, while facing the challengesof operating in limited resource settings and facilitatingtrue community engagement. Site 2 succeeded in itsadvocacy efforts for improved adoption legislation,but faced challenges in implementing its communityoutreach plans, as this was a new scope of work forthe organization. For Site 5, the community-based

ARTICLE IN PRESS

Table 2

Key quantitative achievements and challenges from OVC project evaluation reports

Site Key quantitative achievements Key Challenges

1 37 center staff trained on OVC care and support Sheer volume of demand required operating at maximum

capacity98 community education and mobilization sessions organized for awareness of OVC

needs, children’s rights, and center activities A withdrawal of promised funding by another

organization led to a halt in payments for electricity, staff

salaries, etc.80 OVC attended workshops and 340 OVC attended camps

Shortages of staff, space, and equipment386 children enrolled at centers

2 62 children placed with adoptive, foster or biological families Replenishment of HBC kits by DOH slow

Only 2 of the 24 children at the center at the time of the evaluation had been there more

than a year

Difficulty in implementing community outreach

component

Timely support of DSW workers- with heavy workloads-

critical

3 5 capacity building workshops organized, 441 attendees (some counted more than once)

of which 124 were children

Not all workshops translated into practical results

78 children provided with psychosocial supportChild groups organized in 1 out of 3 target areas

803 children provided with material support via 24 NGOs/CBOsMuch stronger and longer support needed for a strong

child-led movement

8 child focus groups created Limited implementation of planned agricultural activities

to benefit OVC

4 30 group leaders trained by Philippi in psychosocial support Inadequate management support and training

1696 OVC reached through 11 Experiential Learning Camps Lack of supervision and monitoring for Kids Clubs

14 Kids Clubs established; 1764 OVC registered with these clubs Quality of the psychosocial support cannot be determined

without visits372 young people trained in listening and responding skills

122 young people received basic training in psychosocial support

5 80 children received school fees Registration of more than 12 000 OVC raised expectations,

but capacity to reach only a small portion of these40 children received school uniformsDifficulty ensuring the success of IGAs478 families received food parcelsInsufficient number of staff for coverage of large district61 children’s grant applications processed

53 children placed in foster care

6 Identified, secured, and refurbished 5 community care houses Ongoing difficulty of identifying, recruiting, and securing

host and community care families for long-term

placementsServed 150 children during the 3-year program period

High turnover rates of trained, licensed social workers and

caregiver staff

Facilitated reintegration of 40 children into their biological families

Coordinated 36 foster placements and 4 adoptions

Established long-term community care placements for 33 children

7 Baseline survey of 300 households and NGOs/CBOs conducted Community groups’ participation in the organization of

the health fairs difficult to effect10 courses organized, 180 individuals attendedDifficulties in implementing an inherited program,

designed by another,now defunct,agencyOne two-day Health Camp organized for 816 participants, 6 government departments,

13 NGOs/CBOs

Two medium size one-day Health Camps organized for about 100 participants each

8 Trained 140 children in life skills Not all identified communities have begun their projects/

slow start-upTrained 27 mentors on life skillsMentor frustration with hungry/homeless children’s

situations and limited resources to helpReached 140 households of OVC with weekly support from mentors

9 Child-to-Child Clubs established in 8 primary schools OVCCs discouraged and doubt their purpose when unable

to mobilise resources49 OVC and 7 primary school teachers trained in life-coping skillsPublic service delivery in rural Swaziland insufficient2 workshops conducted for 35 traditional leaders on care and supportLinkages between the OVCCs and local government not

sufficiently explored48 parents and guardians trained to impart life skills education

OVCCs and OVC reached 2613 students, 70 teachers and 158 adults with awareness

raising sessions on HIV/AIDS and children’s rights

A. Rosenberg et al. / Evaluation and Program Planning 31 (2008) 51–60 55

organizations that make up the network funded by STFwere able to assist very modest numbers of childrencompared with the 12 000 orphans registered in theorganization’s database.

5.1. Government collaboration

Fig. 1 highlights the different government partners withwhom NGOs collaborated, and the number of NGOs

ARTICLE IN PRESSA. Rosenberg et al. / Evaluation and Program Planning 31 (2008) 51–6056

collaborating with each type of partner. One or morerepresentatives from the Ministry or Department of SocialWelfare or Social Development were partners for all 9projects. One or more representatives from the Ministry orDepartment of Health (MOH or DOH) were involved in 7projects. Other national and local partners were lesscommon. Fig. 2 lists the types of NGO–governmentcollaboration evident in the 9 projects analyzed. Theseincluded: participation in training (7); funding for activitiesrelated to the BMS project (4); collaborative advocacy andlegislation development (3); assistance in accessing govern-

Advocacy

dev

Participation in conducting

training

Funding

0

1

2

3

4

5

6

7

8

9

Type of Collaboration rel

Nu

mb

er

of

Pro

jects

Fig. 2. Types of Governme

0

1

2

3

4

5

6

7

8

9

Min

or

Dept

of

Socia

l

Welfare

/Devm

t

Min

or

Dept

of

Health

Oth

er

National

Part

ners

Local/P

rovin

cia

l

Govt

Part

ners

Types o

Nu

mb

er

of

Pro

jects

Fig. 1. Types of Gov

ment grants/entitlements (3) and placement of children inadoptive or foster care families (2).The separation of NGO and state identities and roles,

and sometimes resulting tensions, were evident in evalua-tions. The leadership of a faith-based organization thatprovides community-based counseling services in Botswa-na stressed to the evaluator their desire not to rely ongovernment funding. They felt doing so would meanhaving to provide services unrelated to religious messages,and they did not want to compromise their Christian values(Site 1).

and legislation

elopment

Assistance in accessing

government

grants/entitlements

Placement of children in

adoptive or foster care

families

ated to BMS-funded project

nt–NGO collaboration.

Nat'l

AID

S C

oord

Body

Dept

of

Agriculture

Min

istr

y o

f

Education

Dept

of

Hom

e

Affairs

f Partners

ernment partners.

ARTICLE IN PRESS

6Martel, P. Site 2 Final Evaluation Report. STF Monitoring and

Evaluation Unit. December 2005.

A. Rosenberg et al. / Evaluation and Program Planning 31 (2008) 51–60 57

For the two projects involving foster care and adoption,tensions arose between the NGO and the state. In Lesotho,the Department of Social Welfare (DSW) perceived anNGO’s desire to process adoption cases and follow-up withadopted children (Site 2) as challenging its autonomy. Theevaluator wrote of the necessity to increase the resourcesand capacity of the DSW rather than allowing the NGO totake over this state function. The same dilemma wasdescribed in another evaluation of Site 6, where NGO staffwere witnessing ill-treatment of children placed in fostercare. As children were officially within the custody ofthe state, the NGO staff could not intervene, but, as theevaluator notes, they were required ethically to act. Sherecommended staff notify the local Department of SocialDevelopment immediately, and assist in finding a solutionif possible.

In the evaluation of Site 3, the evaluator reports thatboth the government and the NGO set out to create twodifferent directories of OVC services, resulting in anoverlap of activity.

For projects involved in helping families to accessgovernment grants, evaluators described complementaryroles of NGOs and government partners. In an evaluationof an umbrella of community-based organizations(CBOs) for OVC care in the Kwazulu–Natal regionof South Africa, the CBOs provided food baskets tofamilies that were waiting for their application forgovernment grants to be processed (Site 5). Anotherexample of complementary roles of government andNGO partners occurred with Site 7, also in South Africa,where the organization’s request for the Department ofLabor (DOL) to provide training led to the utilization ofthe DOL’s previously underspent budget line for training.As a result of the project’s contact with governmentofficials, the Provincial Head of Home Affairs organizedcampaigns in the area and, between August and October2004, processed over 8000 applications for birth certificatesand national identification cards, a necessary step inaccessing entitlements.

5.2. Government collaboration and project sustainability

This section examines the effect of a project’s relation-ship with the host government on the project’s sustain-ability. We previously described types of governmentpartners and types of collaboration; we saw no clearpattern between sustainability and type of partner or typeof collaboration, possibly due to the small sample.

For one project, Site 3, the involvement of thegovernment was not emphasized as important to thesustainability of the project. The project aimed to mobilizesupport and build skills for the protection of children’srights in Lesotho. Some aspects of the project, such as thecreation of an NGO coalition for children’s rights, and thereinforcement of activities by existing children’s rightsorganizations, were seen as potentially sustainable, but nolink to government partnership was emphasized.

For the 8 other sites, the evaluators noted and explainedthe importance of partnership with government forthe sustainability of the project. For some of these projects,evaluators considered the project to have achievedsustainability and for others, they saw reason to beoptimistic about sustainability. For still others, evaluatorssaw sustainability as a challenge or were not able to predictthe project’s ability to survive after STF funding wasconcluded, but did see the role of the government as criticalfor sustainability.Evaluators considered Sites 1, 2, and 4 to have achieved

sustainability. Site 1, supported by STF for 2 OVC day-care centers with a psychosocial component, had obtainedfurther funding at the time of the final evaluation. Sincethe funder was a foundation set up by the government ofBotswana, government endorsement of this project wasclearly essential to its continuing work. Similarly, Site 4obtained funding to continue its psychosocial supportproject for OVC through funding from PEPFAR and theGlobal Fund for HIV, TB and malaria. Since theNamibian Ministry of Health and Social Services is theprimary recipient and administrator of Global Fundmonies, government collaboration can be seen as essentialto the project’s continuation. On a scale of 1 (unimportant)to 5 (important) the Director of this NGO gave a rating of3 to the importance of collaboration with the governmentto the success of the STF-funded project.For Site 2 in Lesotho, the STF funding enabled the

hiring of a social worker to play an advocacy role andpartner closely with the DSW. Through this relationship,they helped to develop the Child Protection and Welfarebill, which defines children’s rights and child abuse,stipulates more severe punishments for abusers, and detailsadoption rules. The evaluator notes, ‘‘the bill willguarantee more rights to the children; and strengthen thelegal basis on which such institutions as (Site 2) operate inthe country’’.6 Therefore, the project can be seen to havesucceeded in the long-term care of OVC through thislegislation, largely due to their partnership and advocacyrole with the DSW. The Director of the organization gave arating of 4 on a scale of 1 (unimportant) to 5 (important),when asked how important the collaboration with thegovernment partner is to the success of the STF-fundedproject.For 2 other sites, evaluators were hopeful about

sustainability. For Site 5, an umbrella organization ofCBOs helping OVCs in Kwazulu–Natal, South Africa, theevaluator’s optimism was due to the multisectoral andcommunity-based approach employed; the acquisition ofan additional year’s funding from another foundation; andthe interest expressed by other donors around the time ofthe evaluation. As a main component of the project wasassisting families in accessing grants, she directly links thesuccess of the project with government collaboration: ‘‘All

ARTICLE IN PRESSA. Rosenberg et al. / Evaluation and Program Planning 31 (2008) 51–6058

[sub] projects to be assisted in developing strong andeffective links with the DSW as this is the core of success ingrants’’.7 She also lists as a challenge the project’s need tostrengthen its relationship with the Department of Agri-culture for the success of income generating activities(IGAs). Therefore, government collaboration can be seenas key to this project’s sustainability and ability to have animpact beyond the current funding. Site 7 was also viewedoptimistically for potential sustainability by the evaluator.A major component of this project was administeringhealth camps to disseminate information from governmentdepartments about grants and other benefits that OVCsand OVC caretakers are entitled to. Given the objectives ofthis project, the support and involvement of governmentpartners is necessary for the success and sustainability ofthis project.

For the remaining three projects, the evaluators ex-pressed concern about or difficulty in ascertaining thestatus of project sustainability but noted the importance ofthe governments’ assistance for the success of each. At Site6, designed to re-integrate children from institutional careinto communities in South Africa, additional funding hadnot yet been secured. The organization has recentlyreceived a letter from the Durban regional office instruct-ing them to withdraw from placing children into theCommunity Houses. Organization leadership does notunderstand this recent change in regional policy. Itdistinguishes between the supportive attitude of theNational Department of Social Development and thedifficulties they have with the Provincial Department ofSocial Development. On a scale of 1 (unimportant) to 5(important) the Director of this NGO gave a rating of 2 tothe importance of collaboration with the government to thesuccess of the STF-funded project. However, the localgovernment and NGO’s non-alignment of strategy clearlyhas implications for the sustainability of this project, sincethe children being cared for by this organization are wardsof the state. For Site 8, a project to provide support forOVC through mentors within the community, the evalua-tor found it difficult to ascertain prospects for sustain-ability due to the slow roll-out of activities. However,she noted that ‘‘If the government was to support mentorswith the stipend when funding comes to an end therewas hope from [project] staff that the project would besustained’’,8 clearly highlighting the importance of thegovernment partnership for this project.

Finally, for Site 9, the evaluator writes, ‘‘The sustain-ability of the model will rely heavily on the establishmentof working relationships between OVC Committees(OVCCs), government structures and traditional autho-rities’’ and notes that OVCCs will need to accessgovernment resources to survive. She also talks of the

7Mabude, Z. Site 5 Interim Evaluation Report. STF Monitoring and

Evaluation Unit. July 2004.8Mabude, Z. Site 8 Final Evaluation Report. STF Monitoring and

Evaluation Unit. August 2005.

government’s management role: ‘‘OVCCs also have to beheld to account by both government and traditionalauthoritiesy.More needs to be done, through and byOVCCs, to establish linkages between communities on theone hand and between communities and governmentauthorities at the administrative district and at the nationallevel on the other hand’’.9 The evaluator, therefore, foundproject sustainability to depend on government collabora-tion. The program manager of the project rated govern-ment collaboration as 3 on a scale of 1 (unimportant) to 5(important) for the success of the project. With regard tolong-term impact of the project, the program managerwrote that the BMSF-funded project contributed immen-sely to the development of the OVC National StrategicPlan, National Monitoring and Evaluation Plan andNational Action Plan for OVC. These contributions havethe potential to benefit OVC in Swaziland beyond the STFgrant timeframe, and are due to close collaboration withgovernment.

6. Conclusions

In our review of 9 projects to support orphans andvulnerable children in southern Africa, evaluations of 8projects emphasized the importance of a relationship withthe host government in project sustainability and making along-term contribution to the care of OVC, while only oneproject showed the possibility of project sustainabilitywithout close government collaboration. The integral roleof the government partner was seen in Botswana, Lesotho,Namibia, South Africa, and Swaziland.For South Africa, 2 projects worked with government

representatives to help families access government grants,thereby directly enabling government to fulfill its missionof providing such grants. Another South African projectinvolved children who were wards of the state. In Botswanaand Namibia, it was a government-funding mechanismthat allowed the continuation of the project, and inLesotho and Swaziland, projects contributed to positivechanges in legislation to protect children. In Swaziland, 2separate projects’ evaluations pointed to the critical role ofthe government’s support at the community level forproject sustainability, through supporting OVC mentors inone project and OVC committees in another.While our sample is small, the results of our analysis

point to varying patterns of government collaboration fordifferent types of project objectives. The projects analyzedhere highlight the importance, but also potential tension, ofworking with government on projects that handle fostercare or adoption placements. NGOs can be helpful inassisting governments to fulfill their mission of providinggrants to affected children, and in this content analysis,governments and NGOs worked well together on suchprojects. NGOs have acted as a liaison between families

9Fourie, A. Site 9 Final Evaluation Report. STF Monitoring and

Evaluation Unit. August 2005.

ARTICLE IN PRESSA. Rosenberg et al. / Evaluation and Program Planning 31 (2008) 51–60 59

and government for grant access; provided materialsupport to families waiting for their grant applications tobe processed, and acted as a catalyst for government actionin processing applications. When NGOs initiate commu-nity-based solutions, such as OVC mentors or committees,the government’s buy-in, and later financial support, iscritical to sustainability. The government’s role in fundadministration, such as the Global Fund or national trustfunds, links governments and NGOs as NGOs seeksustainability. Finally, NGOs can be seen as playing animportant advocacy role for OVC rights, as many civilsociety actors have done for the rights of people living withHIV/AIDS.

While some researchers have looked at the importance ofpartnerships for strengthening national public healthsystems (Padgett, Bekemeier, & Berkowitz, 2005; Reich,2002; Walt, Pavignani, & Gilson, 1999), organizationalpartnerships have received less attention with respect toNGO project sustainability. Our analysis has suggestedthat it is critical for donors, NGOs and CBOs to workclosely with local and national government partners forsustainable projects and long-term improvements in thecare and welfare of OVC. Those engaged in the evaluationof programs should analyze the types, processes, andstrengths of NGO–government collaboration, or lackthereof, as well as other types of organizational partner-ships. Investigations also need to consider alternatives forsustainability: when collaboration with governments is notpossible or desirable, what enhances likelihood of projectsustainability? Government policies, structures and leader-ship vary tremendously for the countries in question, andcomplementary analyses to capture this diversity and itsimplications for collaboration are needed. To reach thegoal of ensuring the rights of young citizens, close linkagesbetween civil society and government agencies are oftenneeded, and all parties should work towards a cleardelineation of appropriate activities to enhance thesustainability of meritorious and worthwhile projects.

7. Lessons learned

Through this analysis we can suggest, in retrospect, anumber of operational changes that would have improvedthe evaluation efforts of our unit. To assess sustainability,funders and evaluators should conduct evaluation after theproject’s funding has terminated, rather than the morenormal immediate final evaluation. For this analysis,evaluations done 6 months or 1 year since completion offunding would have allowed a more concrete measurementof ‘‘the continuation of activities after funding has ceased.’’In addition, to the extent possible, evaluators shouldidentify concepts needing predefined indicators. Evaluatorsshould be required to use well-defined indicators whenmeasuring project success. Our lack of a predefinedindicator for sustainability hindered this analysis. Finally,while process evaluation is considered less desirable thanimpact evaluation, careful qualitative analysis of process

evaluations can lead to valuable lessons learned for thosefunding and implementing public health projects.

Acknowledgments

The authors would like to thank the Bristol-MyersSquibb Foundation for supporting the use of evaluationdata for this analysis. We would also like to thank theauthors for their evaluations on which this analysis isbased: Auriol Ashby, Annalize Fourie, Theophelus Han-dura, Zonke Mabude, Tamra Madenwald, Pierre Martel,and Zethu Mathebeni. This analysis would not have beenpossible without their thoughtful insight and knowledge ofNGO sustainability. We would also like to thank theprogram directors who provided additional informationabout their collaborations with government partners orallowed the names of their organizations to be listed in thismanuscript: Patrick Vorster and Leslie Kemp of St.Philomena’s Children’s Home, Mandla Mazibuko of Savethe Children Swaziland, Rajas Naidoo of Division ofCommunity Paediatrics, University of Witwatersrand,Fiona Napier of Save the Children, UK for their Lesothooperations, and Marianne Olivier, of Philippi TrustNamibia, in addition to others wishing to remainanonymous. We gratefully acknowledge Laura Crawford’suseful suggestions and editing assistance and SimphiweMngadi’s determined search for the status of OVC policies.Finally, we thank the communities and staff involved in theprojects discussed here, and recognize their contributionsto the lives of children in need.

References

Bossert, T. J. (1990). Can they get along without us? Sustainability of

donor-supported health projects in Central America and Africa. Social

Science and Medicine, 30(9), 1015–1023.

Cassidy, E. F., & Leviton, L. C. (Eds.). (2006). Guest editorial: The

relationships of program and organizational capacity to program

sustainability: What helps programs survive? Evaluation and Program

Planning, 29, 149–152.

Fleming, K., Zeitz, P., Global AIDS Alliance. (October 17, 2005)

Advocacy brief: Remember the children: Global fund round 6 in

2006. A critical opportunity to scale up programs for orphans and

vulnerable children. .

Government of Namibia. (2005). National policy on orphans and other

vulnerable children.

Government of South Africa, Ministry of Social Development. (July

2005). Policy framework for orphans and other children made

vulnerable by HIV and AIDS South Africa.

Government of South Africa, Ministry of Social Development. (July 15,

2005). National action plan for orphans and other children made

vulnerable by HIV and AIDS South Africa for 2006–2008.

Hartwig, K.A. (2001). Tanzania’s organizational response to the AIDS

pandemic: An embedded case study of Arusha Region, 1989–1999.

Unpublished dissertation, University of North Carolina at Chapel

Hill, School of Public Health, Chapel Hill, NC.

IRIN News (2005). ‘Namibia: Policy aims to assist OVC’, February 10

/www.irinnews.org/report.asp?ReportID=50210S.

IRIN News (2006). ‘Botswana: More funds needed to curb HIV/AIDS-

related child deaths’, April 20 /www.irinnews.org/report.asp?ReportID=

5880S.

ARTICLE IN PRESSA. Rosenberg et al. / Evaluation and Program Planning 31 (2008) 51–6060

IRIN News (2006). ‘South Africa: Government adopts more focused

approach to help orphans’, February 21 /www.irinnews.org/report.

asp?ReportID=51835S.

Johnson, K., Hays, C., Center, H., & Daley, C. (2004). Building capacity

and sustainable prevention innovations: A sustainability planning

model. Evaluation and Program Planning, 27, 135–149.

The Kingdom of Swaziland. (2006). National plan of action for orphans

and vulnerable children (2006–2010).

LesothoMinistry of Health and Social Welfare, Department of Social Welfare.

(April 29, 2005) National policy on orphans and vulnerable children.

Olsen, I. (1998). Sustainability of health care: A framework for analysis.

Health Policy and Planning, 13(3), 287–295.

Padgett, S. M., Bekemeier, B., & Berkowitz, B. (2005). Building

sustainable public health systems change at the state level. Journal of

Public Health Management Practice, 11(2), 109–115.

Reich, M. R. (2002). Reshaping the state from above, fromwithin, from below:

Implications for public health. Social Science and Medicine, 54, 1669–1675.

Republic of Botswana, Ministry of Local Government Lands and

Housing, Social Welfare Division (1999). Short term plan of action

on care of orphans in Botswana 1999–2001.

Scheirer, M. A. (2005). Is sustainability possible? A review and

commentary on empirical studies of program sustainability. American

Journal of Evaluation, 26(3), 320–347.

Secure the Future (2004). Our story—care and support for women and

children with HIV/AIDS. Bristol-Myers Squibb Company and Bristol-

Myers Squibb Foundation.

Shediac-Rizkallah, M. C., & Bone, L. R. (1998). Planning for the

sustainability of community-based health programs: Conceptual

frameworks and future directions for research, practice and policy.

Health Education Research, 13, 87–108.

Tvedt, T. (2006). The International aid system and the Non-Governmental

Organizations: A new research agenda. Journal of International

Development, 18, 677–690.

UNAIDS/UNICEF/USAID (July 2004). Children on the Brink 2004:

A joint report of new orphan estimates and a framework for

action.

UNAIDS/WHO (2004). Report on the global AIDS epidemic—4th global

report.

UNICEF (2005). Excluded and invisible, State of the world’s children,

2006.

US Department of State (March 8, 2006). Country reports on human

rights practices 2005. accessed on March 14, 2006 at /http://

www.state.govS.

Walt, G., Pavignani, E., Gilson, L., & Buse, K. (1999). Managing external

resources in the health sector: Are there lessons for SWAps? Health

Policy and Planning, 14(3), 273–284.