76223 v1 H N P D i s c u s s i o n P a p e R THE ROLE OF FAITH-INSPIRED HEALTH CARE PROVIDERS IN SUB-SAHARAN AFRICA AND PUBLIC-PRIVATE PARTNERSHIPS Strengthening the Evidence for Faith-inspired Health Engagement in Africa, Volume 1 Edited by Jill Olivier and Quentin Wodon November 2012 THE ROLE OF FAITH-INSPIRED HEALTH CARE PROVIDERS IN SUB-SAHARAN AFRICA AND PUBLIC- PRIVATE PARTNERSHIPS Strengthening the Evidence for Faith-inspired Health Engagement in Africa, Volume 1 Edited by Jill Olivier and Quentin Wodon November 2012 Health, Nutrition and Population (HNP) Discussion Paper This series is produced by the Health, Nutrition, and Population Family (HNP) of the World Bank's Human Development Network (HDN). The papers in this series aim to provide a vehicle for publishing preliminary and unpolished results on HNP topics to encourage discussion and debate. The findings, interpretations, and conclusions expressed in this paper are entirely those of the author(s) and should not be attributed in any manner to the World Bank, to its affiliated organizations or to members of its Board of Executive Directors or the countries they represent. Citation and the use of material presented in this series should take into account this provisional character. Enquiries about the series and submissions should be made directly to the Editor, Martin Lutalo (mlutalo@worldbank.org). Submissions undergo informal peer review by selected internal reviewers and have to be cleared by the Task Team Leader’s (TTL's) Sector Manager. The sponsoring department and author(s) bear full responsibility for the quality of the technical contents and presentation of material in the series. For information regarding the HNP Discussion Paper Series, please contact Martin Lutalo at mlutalo@worldbank.org or 202-522-3234 (fax). © 2012 The International Bank for Reconstruction and Development / The World Bank 1818 H Street, NW Washington, DC 20433 All rights reserved. ii Health, Nutrition and Population (HNP) Discussion Paper THE ROLE OF FAITH-INSPIRED HEALTH CARE PROVIDERS IN SUB- SAHARAN AFRICA AND PUBLIC-PRIVATE PARTNERSHIPS Strengthening the Evidence for Faith-inspired Health Engagement in Africa, Volume 1 Edited by Jill Oliviera and Quentin Wodonb a School of Public Health and Family Medicine, University of Cape Town, Cape Town, South Africa b Education Department, World Bank, Washington, DC, United States Collection prepared in support of the World Bank’s work on private service delivery in human development. Abstract: As African governments, donors, and a wide range of organizations increase their efforts to reach the Millennium Development Goals (MDGs) and set the agenda for the post-MDGs era, the role of non-state providers of health care is gaining new attention. In Africa, the largest non-state networks of providers are often faith-inspired. But how important is the role of faith-inspired institutions (FIIs) in health care provision in Africa? How substantial are their market share and reach to the poor? How affordable are the services provided by FIIs to households? How satisfied are households with these services? What are some of the interesting and innovative experiences that have been documented in terms of FIIs providing quality services to underserved populations? Beyond facilities-based care, which types of non-institutionalized initiatives emerge out of communities of faith that are generative of health? How can these initiatives be mapped, understood and leveraged for better health and development? The objective of this edited series of three World Bank HNP Discussion Papers is to gather tentative answers to such questions. This first volume in the series focuses on assessing the role and market share of faith-inspired providers and on assessing the extent to which they are involved in and benefit from public-private partnerships. Keywords: Health, service delivery, faith, public-private partnerships, Africa Disclaimer: The findings, interpretations and conclusions expressed in the paper are entirely those of the authors, and do not represent the views of the World Bank, its Executive Directors, or the countries they represent. Correspondence Details: Dr. Quentin Wodon, World Bank, 1818 H Street NW, Washington, DC, 20433. Email: qwodon@worldbank.org. iii Table of Contents ACKNOWLEDGMENTS…..……………………………………………………………vi FAITH-INSPIRED HEALTH CARE PROVIDERS IN SUB-SAHARAN AFRICA AND PUBLIC-PRIVATE PARTNERSHIPS: A BRIEF OVERVIEW …...…………………...1 Jill Olivier and Quentin Wodon. CHAPTER 1 - MARKET SHARE OF FAITH-INSPIRED HEALTH CARE PROVIDERS IN AFRICA: COMPARING FACILITIES AND MULTI-PURPOSE INTEGRATED HOUSEHOLD SURVEY DATA …….……………………….……….11 Jill Olivier and Quentin Wodon CHAPTER 2 - MARKET SHARE OF PRIVATE HEALTH CARE PROVIDERS IN AFRICA: COMPARING DEMOGRAPHIC AND HEALTH SURVEYS WITH MULTI- PURPOSE INTEGRATED SURVEYS …………………………….…………….....….27 Quentin Wodon, Minh Cong Nguyen and Clarence Tsimpo CHAPTER 3 - HÔPITAL MATLABOUL FAWZAINI: AT THE INTERSECTION OF DIASPORA, FAITH, AND SCIENCE IN TOUBA, SENEGAL ………………………41 Ellen E. Foley and Cheikh Anta Babou CHAPTER 4 - CONTRACTING BETWEEN FAITH-BASED HEALTH CARE ORGANIZATIONS AND THE PUBLIC SECTOR IN SUB-SAHARAN AFRICA ….60 Delphine Boulenger, Francoise Barten and Bart Criel CHAPTER 5 - HALF A CENTURY YOUNG: THE CHRISTIAN HEALTH ASSOCIATIONS IN AFRICA ……………………………………………………….…71 Frank Dimmock, Jill Olivier and Quentin Wodon CHAPTER 6 - PHARMACEUTICAL SERVICE DELIVERY IN CHURCH HEALTH SYSTEMS IN AFRICA: A CROSS-COUNTRY ANALYSIS ……………………….104 Heather Budge-Reid, Donna Asiimwe Kusemererwa and Anke Meiburg CHAPTER 7 - FAITH-INSPIRED HEALTH CARE PROVISION IN GHANA: MARKET SHARE, REACH TO THE POOR, AND PERFORMANCE …………….118 Jill Olivier, Mari Shojo and Quentin Wodon CHAPTER 8 - APPROACHES FOR INTEGRATING PRIMARY HEALTH CARE IN REPRODUCTIVE HEALTH PROGRAMS IN UGANDA …………………………..134 Lorna B. Muhirwe iv CHAPTER 9 - INCREASED FUNDING FOR AIDS-ENGAGED (FAITH-BASED) CIVIL SOCIETY ORGANIZATIONS IN AFRICA? …………………………...…..143 Jill Olivier and Quentin Wodon CHAPTER 10 - CHRISTIAN ORGANIZATIONS’ PLACE IN MULTISECTORAL HIV/AIDS RESPONSE: KENYA, MALAWI AND THE DEMOCRATIC REPUBLIC OF CONGO ……………………………………………………………………..…..167 Jill Olivier and Beverley Haddad v Acknowledgments This work benefitted from funding from the World Bank as part of a work program on the role of private providers of education and health services. Special thanks are due to Robin Horn, Nicole Klingen, and Rakesh Nangia for supporting this research, as well as to Daniel Cotlear, Frank Dimmock, Mattias Lundberg, Rosalia Rodriguez Garcia, Halsey Rogers, and Karima Saleh for serving as peer reviewers. The editors are grateful to the authors for their contributions towards this collaborative collection, and to Martin Mpungu Lutalo and Erika Yanick for their help in producing this volume. vi FAITH-INSPIRED HEALTH CARE PROVIDERS IN SUB-SAHARAN AFRICA AND PUBLIC-PRIVATE PARTNERSHIPS: A BRIEF OVERVIEW Jill Olivier and Quentin Wodon University of Cape Town and World Bank INTRODUCTION This study is comprised of a three volume series on strengthening the evidence for faith- inspired engagement in health in sub-Saharan Africa. An increasing level of interest in the role of faith in development has generated much debate and dialogue at the international and national levels over the last decade. Despite difficulties in communication and differences in cultures within such debates, there has been a continued reaffirmation of the potential benefits that faith-inspired communities can bring towards efforts to achieve the Millennium Development Goals (MDGs), especially in the areas of health. Yet the evidential field in this area is riddled with uneven data, informational gaps, mismatching frames of reference and frequently conflicting opinions and agendas. There is still a lack of shared understanding on how the direct and indirect impact of religion and faith-inspired institutions (FIIs) is to be measured and understood. The partial and fragmented nature of the body of evidence is one of the greatest obstacles to further cooperation at the policy level – given that policy recommendations often require systematic and sustained supporting evidence which is rarely available here. The field of enquiry is diverse, with few foundational texts. As a result, newcomers tend to struggle to find the points of access; and generally, substantial time and resources are spent in costly meetings re-treading introductory ground, rather than pushing the debate and evidence further. This logically results in repeated calls for more systematic research and more theoretical structuring of the field of enquiry. Recently, and as a result of such calls, more research has been undertaken, mainly around two questions: 1) how does faith or religion impact on individual health-related behaviors, and 2) what tangible infrastructure is held by faith-inspired groups that can be better leveraged for health and development. The objective of this volume, as well as the other two volumes in the series, is to make available some of the recent research on the second of those two questions, with a focus on faith-inspired health engagement in sub-Saharan Africa. PREVIOUS LITERATURE The diverse interests reflected in this series range from papers with an institutional focus (such as the originating ‘mission hospital’ model) to community level responses to health and illness. There is a substantially greater body of knowledge available on the former than the latter. For example, in his seminal book, The Quest for Health and Wholeness, James McGilvray (1981) described ground-shaking events for faith-inspired institutions (FIIs) from the 1960s to the 1980s. These include the changes brought by political 1 independence and the role of the Christian Medical Commission (CMC) and the Tübingen meetings in shaping new thinking on faith-inspired health services, the vision of primary health care (PHC), and the formation of the Christian Health Associations (CHAs). McGilvray also described the state of the evidence on FIIs – mainly based on studies by the CMC in 1963-1964. He provided what would become pioneering national estimates of medical facilities contributed by church groups: “43 percent of the national total in Tanzania, 40 percent in Malawi, 34 percent in Cameroon, 27 percent in Ghana, 26 percent in Taiwan, 20 percent in India, 13 percent in Pakistan and 12 percent in Indonesia�. Although he then added, “However, one should not read too much into the above ratios because, at the time of the surveys, this church-related sector was a very disparate group which, with few exceptions, had no collective existence�. While these estimates could be debated, the core questions McGilvray reported from the 1960s and 1980s are still relevant and mostly unanswered today. These are questions about the nature of FIIs’ engagement in health, their role in facilities-based versus primary or preventative care, what it means to be a faith-inspired or ‘Christian’ provider, whether it is possible to bear the costs of a ‘pro-poor’ mission, whether FIIs can continue to be sustainable given new financial contexts and constraints, and queries about the ‘value- added’ of FIIs. All these questions have long been identified, but how far have we come to answering them with solid evidence over the last half century remains an open question. To a large extent, the literature from the 1960s to the 1990s still holds relevance for those seeking to understand the engagement of faith-inspired communities and institutions in health intervention in Africa today. Many faith-inspired providers remain deeply connected to the aims and traditions forged in the early health systems that they often created. For example, important lessons can be found in the literature addressing mission- based health services and the changes they faced as African countries became independent and national health systems were restructured. This literature emphasizes the effect of different colonial administrations – for example contrasting the English and French approaches – on the development of faith-inspired institutions (FIIs). In the 1990s the literature showed an increased interest in ‘faith-based organizations’ (FBOs) and the newly divided public and private health sectors, with FIIs being recognized as significant, and usually clustered as private-not-for-profit (PNFP) providers. Review suggests four main characteristics of this literature from this period: (1) recognition of the role FIIs historically played in health provision, although not aligned with national systems; (2) ‘market share’ estimates indicating that FIIs provide a high share of health care; (3) reports of weaknesses of ‘FBOs’ such as dogmatic resistance to particular health strategies, lack of management capacity, or resistance to evaluation of finances; and (4) statements of possible ‘added value’ such as unique reach, trust and access into communities, resources such as volunteers and community leadership, networks, and means to motivate staff and sustain quality services. 2 All those questions are revisited in this series on strengthening the evidence for faith- inspired health engagement in Africa. But apart from discussing the role of FIIs in health service delivery through facilities-based care, the series also includes a focus on non- institutionalized initiatives that have emerged out of local faith communities. The last decade has seen a substantial expansion in the literature on this area, with some gaps remaining, but also with the emergence of isolated informational ‘nodes’. This literature aims to assess how these mostly informal initiatives can be mapped, understood and leveraged for better health and development, and how they interact with facilities-based care. The role of non-institutionalized or informal community initiatives on health and development is most evident in the area of HIV/AIDS. The impact of the HIV/AIDs pandemic on research and data relating to religion and public health indeed cannot be over-estimated. The civil society response to HIV/AIDS, and a skewing of vertical funds and FIIs’ focus towards HIV/AIDS have contributed to the emergence of a large diverse literature addressing the ‘religious response to HIV/AIDS in Africa’. This literature reflects multiple approaches and lenses on the poorly defined ‘faith sector’ – often incorporating all HIV/AIDS-engaged faith-inspired institutions and initiatives – and at other times focusing only on faith-inspired nongovernmental organisations (NGOs), civil society organizations (CSOs), community-based organizations (CBOs) or congregational activities. Yet while the role of informal groups, and the potential for large networks of faith communities to make a difference has often been advocated, it has rarely been demonstrated through empirical evidence. It is possible that the intense focus on ‘the religious response to HIV/AIDS in Africa’ has been at the cost of a better understanding of the daily and ‘holistic’ health and development activities of faith-inspired institutions and initiatives, and especially of local faith communities. Whether one considers facilities-based care or informal community health-related activities, there has also been a growing emphasis in recent years on context- and country-specific evidence, recognizing that broad generalizations are rarely useful for policy. African states have starkly different histories resulting in different patterns of civil society and public health engagement in health, as shown by Schmid et al (2008) in comparing Mali (where ‘FBOs’ are few and Islam does not often manifest through formal health services) to countries such as Zambia which are inundated with a complex range of FIIs. There is better recognition of the complexity of the effect of religion itself – for example the pluralistic health-seeking behaviours of patients simultaneously utilizing traditional, religious and medical systems. Work has also focused on the provision and utilization of pharmaceuticals by FIIs - with the Ecumenical Pharmaceutical Network (EPN) and the World Health Organization (WHO) conducting several important early baseline studies. While encouraging, these new pockets of literature do not yet generate a sudden wealth of new evidence. Even today, most work on religion, development and public health in Africa ends with caveats that the ‘missing evidence-base demands more research’. During literature landscaping reviews by the African Religious Health Assets Programme (see ARHAP 2006) efforts were made to collate existing data from ministries, large 3 studies and expert advisors with standard bibliographic searches. This demonstrated the obstacles FIIs and stakeholders face: in-country datasets were often missing, not electronically available, or only accessible through personal relationships. Even for data that was accessible, challenges remained. More is known about FIIs in countries that have Christian Health Associations; less is known about health-engaged faith-inspired NGOs, CSOs and congregations; more is known about Anglophone countries and large FIIs than about district- and community level initiatives whether connected to FIIs (such as mobile health services) or not (such as informal community care groups). Significantly more is said about Christian health providers than the health activities and services that emerge through communities of other faiths. It is important to note that the interest in FIIs – both formal and informal – is not located in one specific discipline, and available evidence is often difficult to use comparatively – with most studies resorting to a qualitative integration of data. There are few standard measures applied to the ‘faith sector’, and no shared typology or classification of ‘FBOs’. In addition, many FIIs have been propelled from not having to monitor their funding streams (trusting in historical partnerships), to being suddenly faced with calls for harmonization and standardized monitoring and evaluation (M&E) requirements. While generalization is not fair, it has been noted that many FIIs lack M&E capacity, leaving little time for additional research to ‘make the case’ for the comparative value of FIIs. To sum up, the landscape of research in this field is progressively changing. Faith- inspired health providers have forged stronger collaboratives and thereby a stronger voice and presence – as can be seen through the emerging role of the CHAs in Africa, and emerging research on this role. More researchers have begun to focus on faith-inspired engagement in health – albeit from a number of different perspectives: some looking at local initiatives and others focussing on specific responses, such as those of faith-inspired communities and institutions to HIV/AIDS. More work has also now been completed through qualitative lenses, foe example anthropological perspectives on health-seeker behaviours, user preferences and health worker motivations. And there has also been an increase in quantitative data relating to FIIs from Ministries of Health and the CHAs. Questions relevant to FIIs have recently been included by the WHO in their Services Availability Mapping survey – although it will take time before this data becomes available. Data from nationally representative household surveys have now been used in a systematic way to measure the market share, reach to the poor, cost, and satisfaction with FIIs as compared to public and private secular providers. These data have the advantage of linking the uptake of services with individual and household variables, including poverty status. While many household surveys distinguish between private and public providers, some now also permit the identification of faith-inspired providers specifically, a feature that has been relied upon in several contributions for this series (see Wodon 2013). OBJECTIVE OF THE THREE VOLUMES IN THIS SERIES Our purpose in this series of three HNP discussion papers is to ‘round up’ various analytical perspectives and emerging research on faith engagement in health in Africa 4 from a range of researchers and practitioners from the ‘North’ as well as the ‘South’. What is shared is a common interest in uncovering what might be distinctive about faith- inspired health initiatives and institutions. While we mainly focus on Africa, the questions that are raised are likely to be of interest for other regions of the world as well. The authors of the various chapters rely on different kinds of research strategies and perspectives. Some of the offerings are full-length analytical articles, reporting on new evidence, and others are shorter notes summarizing recent work and perspectives. The work of World Bank authors has been deliberately interspersed with those of other academic and practitioner partners in order to provide a more layered perspective. It is also necessary to make note of the varied uses of terminology throughout this series. There are continued and unresolved debates relating to the appropriate terminology best applied to ‘faith-inspired institutions and initiatives’ (also called faith-based organizations, religious entities and the like). As editors we have not standardized the use of terminology in each contribution, as the varied usage reflects the complexity of the issue and the variety of interests which intersect at this point. The series is structured into three volumes: a first volume on the role and market share of faith-inspired providers and public-private partnerships, a second on satisfaction and the comparative nature of faith-inspired health provision, and the third on mapping of faith- inspired provision and the extent to which faith-inspired providers reach to the poor. While the various perspectives represented in this series reflect the complexity of the field of inquiry at the intersection of religion, health, and development, the authors do share the belief that the future of genuine dialogue on religion and development – at local and international levels – depends on the availability of robust evidence. Continued efforts are needed to document and to describe, applying a variety of lenses, and building dialogue based on knowledge of what works – and what works best. It is no longer quite true that nothing or little is known about religion, health, and development – or how faith-inspired individuals, communities and initiatives engage in the provision of health in development contexts. Building on previous work, we hope that useful new evidence has been provided in this series and that a strong case can be made in support of continued efforts for developing this systematic evidence base. OVERVIEW OF VOLUME ONE This collection, the first in the three volume series, consists of nine papers which focus on the presence of faith-inspired providers as they are faced with rapidly changing and increasingly complex national health systems. The first two papers in the collection are devoted to various ways of assessing the market share of faith-inspired health care providers in Africa. First, Olivier and Wodon use recent facilities data and multi-purpose integrated household survey data to compare the resulting estimates of the ‘market share’ of FIIs in the provision of health care services in Africa under both approaches. While estimates based on facilities data, especially for hospitals, suggest that the market share of FIIs is at 30 percent to 40 percent, estimates 5 from multi-purpose integrated household surveys are typically at less than ten percent. A number of potential explanations for these large differences are provided. Both types of estimates suffer from reliability concerns and limits. More rigorous data collection is necessary, as are standardized and systematic approaches. However, observing the two types of estimates alongside one other provides a more balanced view of the market share of FIIs in health care systems as a whole. The authors conclude that based on this analysis, the presentation of any generalized market share estimates without strong caveats are problematic – whether this comes from facilities or multi-purpose integrated household survey data. They suggest that such broad statements currently do more damage than good, acting as a barrier to the uncovering of other evidence of significance such as performance, quality of services, or impact on those most vulnerable. Chapter two by Wodon, Nguyen, and Tsimpo provides a closer look at the data on the market shares of various health care providers obtained from different types of household surveys. Building on the evidence provided in chapter one, the question asked is whether estimates of market share for faith-inspired health care providers obtained from multi- purpose integrated household surveys are themselves reliable. The idea is to compare market share estimates obtained from those surveys with estimates obtained for the private sector as a whole in Demographic and Health Surveys. While Demographic and Health Surveys do not identify separately faith-inspired providers, they do identify private and public providers separately, so that comparisons can be made. Overall, the findings suggest that on average, the market share estimates for public and private providers of health care are similar in both types of surveys, which in turn suggests that the market shares obtained in multi-purpose integrated surveys may not be too far off the mark. At the same time, beyond cross-country averages, there is substantial variation in estimates between surveys at the country level, suggesting again the importance of collecting better data on those issues and triangulating findings from various sources. The third paper by Foley and Batou in this collection, entitled “Hôpital Matlaboul Fawzaini: at the intersection of diaspora, faith, and science in Touba, Senegal�, provides the reader with a necessary reminder that while the bulk of the available evidence on religion and health in Africa has an unavoidably Christian focus, there are significant other faith traditions engaged in health provision. The authors present a case study of a development initiative spearheaded by the members of a transnational diaspora, the creation of a medical hospital in the holy city of Touba in central Senegal. Although the construction of the hospital is decidedly a philanthropic project, Hôpital Matlaboul Fawzaini is better understood as part of the larger place-making project of the Muridiyya and the pursuit of symbolic capital by a Mouride dahira. The hospital illuminates important processes of forging global connections and transnational localities and underscores the importance of understanding the complex motivations behind diaspora development. The authors argue that the hospital initiative represents a distinct form of development that is not easily incorporated into state schemas for urban development or public health. The hospital’s history reveals the delicate negotiations between state and non-state actors and diaspora organizations and the complexities of public-private partnerships for development. Evidence of the Mouride dahira’s commitment to building the holy city of Touba, their ability to persevere in the face of significant technical and 6 financial obstacles, and their recent internal crisis of leadership are crucial to understanding this Muslim Sufi brotherhood’s resilience, dynamism, adaptability, and future trajectory. In the next article, “Contracting between faith-based health care organizations and the public health sector in sub-Saharan Africa�, Boulenger, Barten and Criel report on a study conducted on behalf of Medicus Mundi International from 2007 and 2009 on contractual arrangements between faith-based hospitals and public health authorities in four sub-Saharan African countries: Cameroon, Tanzania, Chad and Uganda. The authors note that faith-based health providers have historically not been contractually recognised or established within government or public health systems. They regard contracting as a critical step towards the development of effective and equitable health care delivery systems, and the ultimate integration of faith-based facilities in public health systems. The chapter describes several different contracting experiences between faith-based facilities and the public health sector at the national level. Although experiences are varied, the authors note that comparative analysis can identify common features, providing an interpretative lens for the assessment of contracting policies between faith- based and public sectors in sub-Saharan Africa. They conclude that the success of the relationship between faith-based facilities and public health authorities appears to lie more in the quality of the partnership processes at the central level, than in the operational contracts themselves. They note the need to raise awareness among stakeholders on the crisis in the current contracting landscape, and the need to dramatically improve knowledge and expertise in designing, implementing and monitoring contractual arrangements – especially in the face of increasingly complex health systems. Following this, Dimmock, Olivier and Wodon provide an overview of the increasingly important role of umbrella networks of faith-inspired health providers: the Christian Health Associations (CHAs) in “Half a century young: The Christian health associations in Africa�. The authors note how the CHAs have become a solid presence in the collaborative environment of African health systems. Established through sometimes trial-and-error attempts to draw together disparate faith-based health providers who were disconnected from each other, and also unaligned with national health systems, CHAs have evolved into a particular kind of collaborative effort with a very specific role. CHAs now network faith-inspired health providers and facilities; advocate for a proper recognition of their work; negotiate with governments; build capacity among members; and in some cases channel and report on substantive funds. In this paper, the authors provide a brief recounting of the history of the CHAs and how they were established, as well as a basic typology of CHAs according to three (highly stylized) conceptual stages of their development. This is followed by a discussion of some of the challenges facing CHAs today, based on self-reports from the CHAs. These challenges include increased strain on often weak systems, a human resource crisis, reduced funding from traditional sources, challenges in raising core funding for long-term operations, challenges in gaining government support and cost recovery, and a perceived erosion of Christian values in the work environment. In the face of such challenges, the CHAs play an 7 increasingly important role in negotiating, channeling and raising support – and in directing the future direction of faith-inspired services. Budge-Reid, Asiimwe Kusemererwa and Meiburg then provide insight from a cross- country analysis conducted by the Ecumenical Pharmaceutical Network: “Pharmaceutical service delivery in church health systems in Africa�. The authors note that faith groups play a significant role in providing health care around the world although their exact contribution has not been well defined and few systematic studies have been done about pharmaceutical service delivery at the facility level in the church sector. The baseline studies on access undertaken by EPN between 2005 and 2008 and summarized in this article were a foundational attempt to address the information gap in this area. The studies were conducted in order to investigate compliance by church health services with guidelines on efficient and effective pharmaceutical services. Church health facilities (numbering 363) representing over 20,000 beds and 4 million outpatients and with budgets totaling more than US$ 40 million were surveyed. Results of the study show some areas that are working well and others requiring urgent attention to improve effective and efficient provision of pharmaceutical services. In the next article Olivier, Sojo and Wodon look more closely at the provision of health services by faith-inspired providers in one West African country, Ghana. Relying on administrative, household surveys and qualitative data, the authors ask three questions about the services provided by faith-inspired or ‘mission-based’ health care providers in Ghana, asking: what is the market share of faith-inspired providers as compared to other types of providers; are there differences in market shares among the poor between faith- inspired providers and other types of providers; and how satisfied are patients with the services received and why are patients choosing faith-inspired providers for care? Estimates of the market share of faith-inspired are much lower with household survey data than with facilities data on hospital beds, and various factors explaining such differences are explained. Faith-inspired and public providers appear to be serving the poor roughly equally to public providers, while private providers tend to serve more the higher socio-economic groups than either faith-inspired or public providers. Qualitative data collected in six facilities suggests that the satisfaction with the services received in faith-inspired facilities is high, especially in areas such as respect paid to patients. Finally, the authors note that the reasons that lead patients to choose faith-inspired providers appear not to be related directly to religion per se, but to the quality of the services provided, including through the values of dignity and respect for patients that these facilities exhibit. The Executive Director of the Uganda Protestant Medical Bureau (UPMB), Muhirwe then describes the context in which UPMB operates. The paper, “Approaches for integrating primary health care in reproductive health programs in Uganda�, describes UPMB’s approach towards aligning religious health services with national systems through improved primary health care (PHC). The author notes that Uganda entered into the primary health care arena rather late due to the political upheavals of the late 1970s and early 1980s. By the time Uganda embraced primary health care the public health system was in a fragile state and health care was provided in large part by the religious 8 non-governmental sector now called the private-not-for-profit sub-sector. In the early 1980s, UPMB revised its constitution in response to the Alma Ata declaration of 1979 to pave the way for Primary Health Care Programs and Community Based Health Care activities. Muhirwe describes the case of a particular PHC program undertaken by UPMB and concludes that this project demonstrates PHC is still relevant to faith-inspired providers, with the principles of PHC leading to greater benefits to communities (in this case, leading to increased utilization of reproductive health services and increased knowledge and awareness among target communities). Monetization of community based health programs and inadequate community health financing initiatives remain the biggest barrier to scaling up such models. In chapter nine, entitled “Increased funding for AIDS-engaged (faith-based) civil society organizations in Africa?,� Olivier and Wodon review the evidence on the comparative extent to which faith-based civil society organizations (FB-CSOs) have benefited from increased funding related to the HIV/AIDS response in Africa. First, the available literature is reviewed on whether FB-CSOs have benefited from such funding. The authors find the arguments vigorous, but the evidence inconclusive. Next, the paper provides fresh analysis of a survey carried out in six Southern African countries to compare the profile and sources of funding of FB-CSOs against CSOs that are non- religious or ‘secular’. It is important to be aware of the at times artificial distinctions made between faith-based and ‘secular’ structures, given the often integrated presence of religion in the lives of civil society actors and their institutions – especially in Africa. However, it is still useful to consider this particular distinction, impacting as it does on current policy discussions and strategies for civil society engagement. While the data of this particular study is mostly representative of a cluster of well-established ‘CSOs’, the evidence suggests that FB-CSOs have been able to benefit as much as other CSOs from enhanced funding opportunities. The paper concludes with a discussion of the challenges that remain for supporting smaller and less formal FB-CSOs and initiatives operating at a local community level. Finally, Olivier and Haddad report on perceptions of the position and role of Christian organizations in multisectoral HIV/AIDS response in Africa in their paper entitled “Christian organizations’ place in multi-sectoral HIV/AIDS response: Kenya, Malawi, and the Democratic Republic of Congo�. The authors note that while partnership with FBOs for improved HIV/AIDS response in Africa sounds logical and useful, there are persistent questions about the ‘faith sector’ and a history of isolationism from other civil society collaboratives. The article examines the possibilities and challenges of collaborative relationships in the struggle against HIV/AIDS in several countries in Africa, specifically between Christian religious entities and national governments and donors. These collaborative partnerships are considered within the framework of the Three Ones policy promoted by UNAIDS. The paper summarizes some of the broader comparative findings that emerge from a research study conducted in Kenya, Malawi, and the Democratic Republic of the Congo (DRC). The authors highlight the very real challenges facing such collaboration as well as the potential such partnerships bring to the HIV/AIDS response. They note that Kenya, Malawi and the DRC represent three different ‘stages’ of multisectoral collaboration. They conclude that this particular study 9 highlights the importance of ‘trust’ and genuine partnership – over and above the policies, mechanisms, and collaborative processes that have been implemented. REFERENCES African Health Assets Programme (ARHAP). 2006. Appreciating Assets: The Contribution of Religion to Universal Access in Africa. Cape Town: African Religious Health Assets Programme. Haddad, B., editor. 2011. Religion and HIV and AIDS: Charting the Terrain. Pietermaritzburg, South Africa: Cluster Publications. McGilvray, J. 1981. The Quest for Health and Wholeness. Tübingen: German Institute for Medical Mission. Schmid B., E. Thomas, J. Olivier, and J. R. Cochrane. 2008. The Contribution of Religious Entities to Health in Sub-Saharan Africa. Cape Town: African Religious Health Assets Programme. Wodon, Q. 2013. Faith and Human Development: Education and Health Services in Africa, New York: Palgrave MacMillan. 10 CHAPTER 1 MARKET SHARE OF FAITH-INSPIRED HEALTH CARE PROVIDERS IN AFRICA: COMPARING FACILITIES AND MULTI- PURPOSE INTEGRATED HOUSEHOLD SURVEY DATA Jill Olivier and Quentin Wodon University of Cape Town and World Bank This paper relies on facilities and household survey data to estimate the ‘market share’ of faith-inspired institutions (FIIs) in the provision of health care services in Africa. While estimates based on facilities data, especially for hospitals, often suggest that the market share of FIIs is at 30 percent to 40 percent, estimates from household surveys are typically at less than ten percent. A number of potential explanations for these large differences are provided. Both types of estimates suffer from limits, but observing the two types of estimates alongside one other provides a more balanced view of the market share of FIIs in health care systems as a whole than is the case for any single type of measure. INTRODUCTION Market share estimates - typically in the 30 percent to 40 percent range - are frequently wielded as the most concrete evidence of faith-inspired activity and impact in the area of health care provision in sub-Saharan Africa. In fact, there are few speeches, reports or articles which do not make some mention of market share, attaching some percentage to the significance of faith-inspired healthcare provision in the continent as a whole, or in specific countries. Such statements usually appear in the early stages of a text, and are the basis on which a further argument is made, for example that this sector therefore requires further attention or resources from governments and donors. A systematic review of this literature (Olivier and Wodon 2012a) reveals that these statements have currently reached the point of becoming almost indisputable ‘truth’, as they are now frequently made without any referencing. There is of course an empirical basis for these statements, but it is not as strong as commonly believed. Several attempts have been made at synthesizing the data on market share of faith-inspired health provision in Africa – usually resulting in a table listing country estimates (see for example De Jong 1991, Dimmock 2005 and 2007, Chand and Patterson 2007, Gilson et al 1994, Green et al 2002, Grills 2009, Hanson and Berman 1994, Hecht and Tanzi 1993, Kawasaki and Patten 2002, Robinson and White 1997, Rookes 2009, Schmid et al 2008, Turshen 1999). It should be noted however that most of these authors provide caveats to this data, naming the unevenness of the data and the inherent challenges to synthesizing mismatching measures. Indeed, any generalized statement of faith-inspired market share is immediately undermined by the fact that barely any of the evidence is measured utilizing the same indicator or measure (if such is at all apparent), and becomes more a case of ‘comparing oranges, apples, and elephants’. 11 There are several reasons why it is difficult to obtain reliable, robust, and comparable market share estimates. Firstly, the data on faith-inspired health service provision is embedded within different evidence clusters. In different countries (and often in different studies of the same country), FIIs get differently counted as part of: private, public, non- state, voluntary, private-not-for-profit (PNFP), traditional, government-assisted health facilities, non-governmental (NGO), or civil society sectors (see Batley 2010, Schmid et al 2008). None of these classifications are without their own problems – for example, the WHO (2006) notes that the definition of public and private providers is increasingly blurred “…as medical goods and services flow between public, commercial, philanthropic traditional and informal providers.� However, FIIs have a particular tendency to fall into the grey areas. For example in some countries a hospital owned by a faith-inspired group can also be classified as a public (or district) health facility; or a government-owned parastatal hospital can be operated by a faith-inspired governing body (see CSSC 2007, DeRoeck 1998, Gilson et al 1994, Munishi et al 1995). In many countries, health facilities are co- owned between faith-inspired and government providers, and cross-subsidization and the co-sponsoring of staff wages and costs have created a complex environment for assessment (see Schmid et al 2008, Batley 2009). In fact, many FIIs consider their character and internal philosophy to be more ‘public’ than ‘private’. 1 FIIs, who are usually non-profit and often profess a central intention to work with the rural poor, find little comfort in being counted as part of a private for-profit sector serving an urban elite. Further murkiness comes from the fact that some FIIs do pursue profits (see Gilson et al 1994). There is also some grey area as to which FIIs are private-non-profits and which are part of the ‘traditional’ health sectors – with multi-modal religious practices making a clear distinction difficult in some places, especially when looking at informal community levels. In Zambia, for example, traditional healers often run local ‘healing hospital facilities’ and are becoming more integrated with biomedical health systems through referral practices (see ARHAP 2006). In addition, evidence on FIIs is also embedded in different types of studies, among others on private-public partnerships, HIV/AIDS responses, community development, and other disease-specific surveys, making cross-analysis difficult. A similar complexity has developed around the nomenclature of FIIs – with no clarity or consensus on how FIIs should most appropriately be named or classified. In this way, some FIIs are mapped and remapped again, and others remain invisible, lost between the cracks of research agendas and frameworks. It is also often difficult to know what exactly most market share estimates refer to. Many estimates are stated as percentage of ‘health care’ or percentage of the ‘national health system/service’ (NHS) – with only a few estimates being precise and showing specific 1 There are implications for these classifications. It has been noted, for example, that the classification of the religious non-profit services as ‘private’ has encouraged an atmosphere of competition, rather than collaboration, with the public sector (see Schmid et al 2008). 12 indicators such as number of hospital beds, number of doctors or nurses, number of training schools, number of health centers, or percentage population served, in-patient days and out-patient attendances, number of congregations, orphans and vulnerable children reached, or health and education services provided. Different measures also represent a number of different interests and ‘levels’ – a common characteristic of literature addressing FII’s engagement in health and development, which takes a transverse slice across a number of different interests and fields. For example, studies of the faith-inspired response to HIV/AIDS in Africa commonly address response of formal health facilities, primary health care programs, networks and non-governmental and community institutions, congregations, as well as informal community initiatives within the same piece of research. The most common measure is based on the number of hospital beds or number of health facilities which, as Hanson and Berman (1998) argue, has been so far the most readily available data. However, such data are typically used to make broader statements about faith-inspired market share – despite the fact that the evidence is primarily based on a cluster of inventories of health facilities and hospital beds owned or operated by members of the national faith-based health networks (NFBHNs) 2 in a few key countries. There is significantly less known about the non-facility-based, informal and community level of healthcare provision. All this speaks to a complex evidential landscape, in which the little data there is on market share – fits poorly together, making cross-country or regional comparison difficult – and results unavoidably in specific pieces of evidence being utilized for broader claims than for which they are usually designated. In this context, the objective of this paper is to take one (partial) step towards a more comprehensive assessment of the role of FIIs in health care provision in Africa by comparing and interpreting the market share estimates obtained with facilities data (and especially hospital beds) with alternative estimates obtained from nationally representative households surveys that tend to take into account health systems as a whole. The results are striking: while estimates based on facilities data suggest that the market share of FIIs in Africa is at 30 percent to 40 percent, estimates from household surveys are typically at less than ten percent. Beyond providing both sets of estimates, we discuss several of the factors that help in understanding why the estimates vary so much between the two data sources (for work comparing market share estimates for private providers of health care, both faith-inspired and private secular, in sub-Saharan Africa using different types of household surveys, see Wodon et al. 2012, and Wodon, 2013). 2 The term ‘national faith-based health networks’ (NFBHNs) has gained some traction so we will leave it as is. NFBHNs are country-level providers of health services, or networks of health service providers. Apart from a few outliers, most are Christian Health Associations (CHAs). The core functions of NFBHNs are to support health services provided by their members through their activities in advocacy, technical assistance and training, capacity building, resource mobilisation, research, M&E, joint procurement and equipment maintenance, and communication. Although the NFBHNs and their members face various challenges such as financing and workforce concerns, they are generally considered to be exemplars of the positive impact of collaboration (Schmid et al 2008, CHAK 2006, CSSC 2007, Dimmock 2007 – see also www.africachap.org). 13 We realize that the issue of market share is a potentially inflammatory topic, with different parties having vested interests in such estimates which are perceived to have a significant impact on collaborative engagement. We must therefore be clear that this article does not set out to lambast any current estimates – the majority of which are self- admittedly based on inadequate evidence. We also do not conclude with a newly synthesized estimate of faith-inspired market share for health provision in Africa as a whole, or in any of the specific countries we are looking at. Rather, we simply suggest that there is a lot to gain in looking at different sources of data to assess the role of FIIs in national health systems, and also actually in moving beyond arguments or advocacy based on market share, which are of limited value, to start looking at some of the more important policy questions about the role of FIIs. The rest of the paper is structured as follows. In section two, we review some of the existing estimates of the market share of FIIs that have been provided in the literature on the basis of facilities data. In section three, we provide new evidence on alternative market share estimates obtained from nationally representative household surveys. In both sections, we also discuss some of the assumptions that lie beneath the estimates, and some of the factors that may help explain the large difference in market share estimates between facilities data and household surveys. In concluding, we suggest how moving beyond the debate or advocacy around market shares may be beneficial for research aiming to assess the respective roles and complementariness of the public, private non- religious, and faith-inspired sectors. ESTIMATES BASED ON FACILITIES DATA Africa-wide estimates of the market share of faith-inspired health care providers are plagued by comparability issues, and estimates based on particular indicators are typically used for much broader claims (Olivier and Wodon 2012a). For example, an estimate for hospital beds provided by one faith-inspired network may be used by others as an estimate of all faith-inspired health-relevant activity in that country or the entire region. By contrast, estimates obtained at the country level are at least potentially more useful for informing policy. Ideally, for any given country, one would like to have a comprehensive assessment of the scope and scale of all health-related services provided not only by government facilities and FIIs, but also by private-for-profit providers as well as other (non-religious) not-for- profits (NGOs), community-based organizations and initiatives - including in areas such as the response to HIV/AIDs. This is however not feasible in practice. The evidence to- date is for the most part based on estimates of the role of FIIs, and especially Christian Health Associations, in national health services, and based primarily on the share of hospital beds located in facilities owned or operated by Christian Health Associations, as compared to beds located in facilities owned by then public sector, typically through Ministries of Health. Table 1 provides examples of such estimates. The implied market share of FIIs ranges from 10-20 percent in Chad to 50-70 percent in the DRC, with most estimates falling in 14 the 30-40 percent range (this is the case for Ghana, Kenya, Lesotho, Liberia, Malawi, Nigeria, Rwanda, Sierra Leone, Tanzania, Uganda, and Zambia). There thus seems to be substantial convergence in existing estimates on the market share of FIIs, with again most estimates reflecting the share of hospital beds or out-patient care provided by FIIs as compared to public facilities. The question is whether these estimates capture a large enough share of the total delivery of health care in the respective countries. We would argue that interpreting the data in table 1 as valid estimates of the share of services provided by faith-inspired facilities within national health systems is problematic for at least three reasons. First, within formal care delivery mechanisms, a large share of health services are provided by other types of facilities than hospitals, such as clinics and health centers, as well as maternity homes and facilities from government-run community-based health planning and services which are primary health care focused services (sometimes with mobile units). Estimates of the market share of FIIs based solely or principally on hospital beds may lead to overestimating the role of FIIs to the extent that the share of hospital-based care among all services provided is often higher among FFIs such as the Christian Health Associations than is the case for public facilities. Indeed, when one looks at the share of total facilities operated by FIIs, one often gets a lower estimate than when referring to hospital beds only. Second, in part because the private non-religious sector often operates independently of Ministries of Health in many countries, data on the role of private non-religious facilities are often missing, even if one restricts the analysis to measures such as hospital beds. That is, most of the estimates of the share of FIIs are based on comparing FIIs with public sector facilities, without factoring in the existence of similar facilities operated by private non-religious providers (and for that matter also without factoring in some religious providers that are not part of large federations such as the Christian Health Associations – this is often the case for Islamic clinics and hospitals.) A third issue with the reliance on statistics on hospital beds, pharmaceuticals, outpatient care and for that matter also the number of formal facilities operated by different networks is that a large share of health care is provided by other types of providers that are not included in such statistics. At least two different groups must be mentioned here. First, many countries have a significant traditional sector that often operates alongside orthodox biomedical care, for example with patients mixing plural health-seeking modalities. While studies on religion and health-related behaviors recognize the role of traditional practices, this is rarely addressed in the literature on the market share of FIIs. Second, self-medication has also been noted to be a significant practice in many countries – given the limited availability of doctors. Some studies have shown that self-medication with privately purchased drugs may in some countries represent the most common treatment after home remedies. As noted by Bennet et al (1997), “household surveys indicate that drugs purchased from local drug sellers or pharmacies are used to treat approximately 53 percent of illness episodes in Burkina Faso … studies on general and low-income populations in Kenya, Nepal, Rwanda, Thailand and elsewhere show similar high rates of medication with drugs acquired in the private sector.� 15 Table 1: Examples of national health service (NHS) market share estimates based on facilities data Country Share (%) of NHS Reference Chad Faith-based care ~20% national health coverage Boulenger et al 2009 NFBHN [UNAD] 10% national health coverage DRC Church related institutions 70% health services ECC 2007 FBOs ~50 % health services provided and facilities owned Kintaudi 2006 Ghana All FBOs (Christian & Muslim) 40% national health services EPN 2005 CHAG 35-40% national health care CHAG 2006 Kenya NFBHN [CHAK and KEC] 40% national health services Mwenda 2007 Lesotho NFBHN [CHAL] 40% national health service Green et al 2002, MOH- Lesotho 2007 Liberia Christian Health Networks ~46% national health sector Chand and Patterson 2007 Malawi Church 40% services Robinson and White 1998 NFBHN [CHAM] 37% health services Mhango 2006 NFBHN [CHAM] 20% national health infrastructure MOH-Malawi 2001 Nigeria CHAN 40% national health services CHAN 2007 Rwanda Church-affiliated facilities 45% hospitals and 35% primary care facilities CCIH 2005 Sierra Leone NFBHN [CHASL] 30% national health services Dimmock 2007 Tanzania NFBHN [CSSC] 48% national health service Green et al 2002 NFBHN [CSSC] ~26% all health facilities, 40% hospitals, and 50% Todd et al 2009 health services in rural areas Uganda NFBHNs [Christian and Muslim: UPMB, 50% beds, 60% hospital services, HERA 2005 UCMB and UMMB] 42.3% hospitals, 22% lower level health facilities, 70.7% nursing/midwifery schools NFBHN [UPMB & UCMB] 50% national health service Green et al 2002 Diocese and parishes 70% all PNFP (lower-level units & hospitals) MOH-Uganda 2001 Zambia NFBHN [CHAZ] 50% rural health care provision Nussbaum et al 2005 30% total health care provision NFBHN [CHAZ] 30% all health services MOH-Zambia 2002 Zimbabwe NFBHN [ZACH] 45% national health service Green et al 2002 Christian Hospitals 68% total bed capacity Benn 2003 Source: Compiled by the authors. 16 These three factors are likely to lead to substantial overestimation of the market share of FIIs in national health systems when the main data source being used refers to hospital beds and formal facilities. At the same time, one should also note that other factors may go in the other direction and leading to higher market shares for FIIs. As noted early on by Hanson and Berman (1994), estimates of health service provision, measured by the number of providers (beds or facilities) may be misleading if utilization levels differ significantly between different sectors and contexts. To the extent that faith-inspired facilities have higher (or lower) utilization rate (say because of a higher or lower quality of services, at least as perceived by patients), they may provide a higher (or lower) share of hospital-based care than suggested by statistics on hospital beds. There are however only a few localized studies – mostly dated, which address utilization of these services, and it is difficult to generalize from the limited evidence available. For example Mwabu (1986) reported findings from a household survey looking at provider choice in Kenya which suggest that after the initial visit, mission clinics dominated other facilities as a treatment source. In Burundi, there is dated evidence that mission facilities are twice as heavily-utilized by outpatients as government facilities (World Bank 1983). But Banda and Simukonda (1994) suggests a lower utilization of religious facilities in Malawi (based on hospital bed data from the MoH). 3 In addition, it has also been noted that FIIs tend to be engaged in a range of activities that stretch beyond formal health services because of their ‘holistic’ focus on health. But data on those community-based efforts are rarely available. And overall, the factors that would lead to higher market shares for FIIs in national health systems are likely to be overridden by the above three main reasons why the market share of FIIs is likely to be overestimated when relying solely on facilities data. In order to assess to what extent this may be the case, we turn in the next section to alternative estimates of market share based on household surveys. ESTIMATES BASED ON HOUSEHOLD SURVEYS To date, nationally representative household survey data in which households are asked about the type of health care facility that they use when seeking care have not been drawn much into the discussion about the market share of FIIs in Africa. This may be in part because the surveys most frequently used for work on health and development, the Demographic and Health Surveys (DHS) implemented in similar ways in most African countries at regular intervals. DHS simply do not distinguish between faith-inspired and non-religious providers of care; they only distinguish between public and private providers, often suggesting that private providers provide a large share of all care in Africa, but this does not help us very much here. For this paper, instead of using DHS data, we rely on the main multi-purpose and nationally representative household surveys implemented in approximately 30 African countries. These are the surveys that are used for poverty measurement, or for analyzing 3 See Olivier and Wodon (2012b) in volume 2 of this collection for some further studies on utilization in relation to patient satisfaction. 17 the links between education and employment. But these surveys also have detailed health modules that provide information among others on whether household members have been sick, ill, or injured in the recent past (typically over the last two or four weeks), whether they did seek care, and if so, where they went for care. As shown in table 2, in about half of the surveys that we looked at, there is enough information on the type of provider consulted by households to identify public, private non-religious, and private faith-inspired providers 4. Table 2: Identification of FIIs in the health modules of selected household surveys Country Identification Country Identification Benin (QUIBB 2003) No Liberia (QUIBB 2007) No Burkina Faso (QUIBB 2003) No Malawi (HIS-2 2004) Yes Burundi (QUIBB 2006) Yes Mali (QUIBB 2006) Yes Cameroon (ECAM 2007) Yes Niger (ENBC 2007) Yes Cape Verde (QUIBB 2007) No Nigeria (LMS 2003/2004) Yes Chad (ECOSIT2 2003/04) Yes ROC (QUIBB 2005) Yes Cote d'Ivoire (ENV 2002) No Rwanda (EICV 2001) No DRC (123 survey 2004/05) No Senegal (ESPS 2005) Yes Gabon (QUIBB 2005) No Sierra Leone (SLIHS 2003) Yes Ghana (GLSS5 2005/2006) Yes Swaziland (SHIES 2009) Yes Guinea (QUIBB 2007) No Togo (QUIBB 2006) No Kenya (KIHBS 2005) Yes Zambia (LCMS IV 2004) Yes Source: Compiled by the authors Table 3 provides our estimates of the market share of public, private faith-inspired, and private non-religious providers in the fourteen countries where the questionnaire provides the necessary information to do so. This table displays significantly lower estimations for faith-inspired market than those base on hospital beds or health facilities mentioned earlier. The estimates range from a market share of 1.5 percent in Niger to 15.1 percent in Cameroon. In addition, it appears that the market share of the private non-religious sectors, which includes here not only private non-religious formal facilities, private chemical stores, and pharmacies, but also traditional healers and private doctors, may be much larger than that of FIIs. On average, across the 14 countries, the market share of faith-inspired providers is below 6.0 percent. This is likely to be too low, because some of the countries that are known to have a very large faith-inspired health care sector, such as the Democratic Republic of Congo, are not included in the sample (simply adding the Democratic Republic of Congo would raise the average market share of faith-inspired providers by several percentage points). Still, the market shares based on multi-purpose integrated household surveys do suggest substantially smaller market shares than those provided based on facilities data such as the share of beds owned by the CHAs. 4 More details on how this is done in each of the survey where the information is available given the way questions are asked in survey questionnaires are provided in Wodon et al (2012) in this volume, where the market share estimates obtained from multi-purpose integrated household surveys are compared to private market share estimates for private providers as a whole in Demographic and Health Surveys. 18 What could explain such different results? Obviously, much of the difference is likely to result from differences in the universe of health care being considered. In Ghana for example, hospitals account for less than a third (31.6 percent) of all consultations in the survey. Assuming that hospital beds or outpatient care are good proxies for the overall supply of care of hospitals (which itself would be a strong assumption, given that a large share of hospital care does not necessarily require hospital beds), a hospital bed market share for FIIs of a third might be diluted into a market share of about 10 percent for health care as a whole when a broader universe of care is taken into account, as is done in the surveys. The market share could be lower when accounting for the role of private sector facilities which are often not accounted for when estimating the market share of FIIs based on facilities data. This suggests that the household survey-based estimates of market share may not be completely out of sync with the reality on the ground. Still, even then the survey-based estimates look small against the current wisdom of those working on the ground – for example those in the Ministries of Health or Christian Health Associations. Another explanation might then be that the identification of faith-inspired providers by households in the surveys is partial, with some households considering faith-inspired facilities as either private non-religious facilities, or with more likelihood as public facilities, especially when mission hospitals are considered as district or regional hospitals by Ministries of Health and accordingly funded by the government. That could very well be the case, and in that case, efforts to better identify facilities would be required in household surveys in order to obtain more reliable data. At the same time, it is unclear how large this problem might be, and the extent of the problem is likely to depend on the specific country and survey questionnaire. Although this does not fully address this question, results in Wodon et al (2012) obtained from a comparison of market share estimates for private providers of healthcare in Demographic and Health Surveys and in Multi-purpose integrated household surveys suggest that broadly, there is coherence between the various surveys even though the questionnaires tend to differ. Another factor explaining differences between facilities and household survey-based estimates of market shares may be the fact that in table 1, most of the countries listed belong to Anglophone Africa, and the two countries that are Francophone (Chad and the Democratic Republic of Congo) are both conflict-affected countries where FIIs have helped fill the void in service delivery left by weak governments. By contrast, in table 3, we have a more balanced sample with six Francophone countries and eight Anglophone countries (if Cameroon is included in that second group). Many (although not all) Francophone countries have lower FII market share – and it is a well-established fact that colonial administrative policy had an important role in facilitating the growth of FIIs in Anglophone countries. Most of the strongly established Christian Health Associations are thus also located in Anglophone Africa. The fact that much of the literature has so far focused on Anglophone Africa may thereby have led to a bias upward in the assessment of the role or market share of FIIs in Africa as a whole. 19 Table 3: Estimates of Market Shares from Household Surveys Faith- Private Public Survey Period in Number of Consultations inspired Non-religious Market Population questionnaire Consultations per person Market Market Share (%) Share (%) Share (%) Burundi, 2006 8,237,232 Last 4 weeks 1,778,654 0.216 69.3 11.5 19.2 Cameroon, 2007 18,083,282 12 months 7,318,156 0.405 44.9 15.1 40.0 Chad, 2003/04 7,393,259 Last 4 weeks 795,874 0.108 53.1 10.7 36.2 Ghana, 2005/06 22,216,866 Last 2 weeks 1,615,726 0.073 44.4 6.6 49.0 Kenya, 2005 35,494,317 Last 4 weeks 6,666,834 0.188 49.0 4.2 46.8 Malawi, 2004 12,329,494 Last 2 weeks 2,745,456 0.223 36.9 3.9 59.2 Mali, 2006 12,317,562 Last 4 weeks 1,267,931 0.103 68.5 1.0 30.5 Niger, 2007 13,427,990 Last 4 weeks 1,583,052 0.118 77.6 1.5 20.9 Nigeria 2003/04 126,482,035 Last 2 weeks 8,380,632 0.066 50.2 1.9 47.9 Republic of Congo, 2005 3,551,500 Last 4 weeks 946,993 0.267 44.0 4.0 52.0 Senegal, 2005 12,012,657 Last 4 weeks 1,417,784 0.118 - 2.3 - Sierra Leone, 2003/04 4,825,118 Last 2 weeks 546,559 0.113 60.1 2.0 37.9 Swaziland, 2009/10 1,018,358 Last 4 weeks 138,562 0.136 - 13.2 - Zambia, 2004 10,987,778 Last 2 weeks 922,788 0.084 55.0 6.1 38.9 Average - - - - 55.2 5.8 39.0 Source: Authors’ estimations Note: For Swaziland, the population is obtained from the World Bank Development Database Platform. Comment: the fact that the consultation rates per person are of a similar order of magnitude for the different countries is reassuring on the validity of the data. Data on the choice of provider of health services are based on the last consultation in a given period of time; one thus expects a higher ‘last’ consultation rate over a one year period as in Cameroon (because it is more likely that a person was sick at least once over a one year period) than over a period of two weeks, as in Nigeria or Sierra Leone. 20 CONCLUSION This paper was meant to contrast estimates of the market share of FIIs in Africa based on facilities data and household surveys. A number of explanations for why the estimates are so different have also been provided. We are certainly not claiming here that estimates from nationally representative household surveys are somehow better than those based on facilities data. Both types of estimates have strengths and weaknesses. The ‘real’ estimates of market share probably lie somewhere in between these estimates. It is because so far the facilities-based estimates were virtually the only ones used in the broader advocacy and policy circles that we felt it was important to also provide estimates based on household surveys. What we find problematic is the over-reliance on one kind of estimates, especially when facilities-based market share estimates for a very specific cluster of health providers are utilized to make broad claims related to the role of all FIIs in national health systems and development more broadly. Utilization-oriented market share estimates based on household surveys were not shown here in order to undermine the significance of FIIs, or the estimates of their role based on the number of facilities or beds operated. These estimates remain important. But consideration of both types of estimates should challenge how existing market share estimates have been used for advocacy purposes, and we believe that more caution is needed in interpreting these data. Any generalized or basic estimation of market share can be easily challenged. What is necessary is to derive more complex estimates of market share that integrate different measures and data sources. Faith-inspired market share is unfortunately a volatile topic - especially when wielded in fragile and resource-constrained collaborative contexts. The historical lack of alignment between government and FII services has led to such estimates being used as blunt instruments for advocating for or against a stronger role for FIIs. For some FII advocates, a higher estimate of the FII market share has been thought to be beneficial in order to increase funding and policy influence, or to foster greater independence for FIIs. Conversely, it could be that some government agencies might have seen benefits in lower estimates of FII market share, for example to limit requests for financing and staffing costs. Such tendencies on both sides are however not helpful, as they get in the way of establishing systematic and rigorous service delivery assessment systems where the contribution of both public, FII facilities is fully recognized, as is that of other non-public and non-religious service providers. More rigorous data collection is necessary, as are standardized and systematic approaches. But beyond debates on market share, what is even more important is to measure the quality of services provided, the cost of services for users, and the extent to which different types of service providers reach the poor who still often lack access to services and have difficulties in affording such services. What is also needed is a detailed assessment of the extent to which policy reforms affect public and faith-inspired providers differently. For example, changes in cost recovery policies may have very different impacts on different types of providers depending on how they are funded. A number of efforts have recently been launched to better document the role of FIIs – some 21 are led by multilateral agencies such as WHO and the World Bank; others are led by FIIs; still others are country-led. One can hope that a few years from now, the evidence base on which policy in this area can draw will be stronger. The main conclusion from our analysis is that that the presentation of generalized market share estimates without caveats are problematic – whether this comes from facilities or household survey data. In our opinion, broad statements on market share currently do more damage than good. The distortion and malleability of these estimates is less a result of faulty research than an indicator of powerful agendas. Market share discourse and confrontation might be acting as a barrier to the uncovering of other evidence of significance – such as performance, quality of services, or impact on those most vulnerable. Beyond the issue of market share, there is still a worrying lack of basic evidence on facility-based faith-inspired healthcare, never mind the more complex range of informal and community level activities. The estimation of faith-inspired market share remains a valid research endeavor, but there are other important research agendas to pursue. REFERENCES Annan, I., and S. Essuman. 2006. Results of Ghana Baseline Survey: Maximising Access to Essential Medicines for Church Health Services and their Clients. Nairobi, Kenya: Ecumenical Pharmaceutical Network. African Health Assets Programme (ARHAP). 2006. Appreciating Assets: The Contribution of Religion to Universal Access in Africa. Cape Town: African Religious Health Assets Programme. Banda, E. E. N., and H. P. M. Simukonda. 1994. “The Public/Private Mix in the Health Care System in Malawi.� Health Policy and Planning 9(1): 63-71. Banda, M., E. Ombaka, S. Logez, and M. Everard. 2006. Multi-country Study of Medicine Supply and Distribution Activities of Faith-based Organizations in sub- Saharan African Countries. Geneva: World Health Organization and Ecumenical Pharmaceutical Network. Batley, R., and C. Mcloughlin. 2009. State Capacity and Bon-state Service Provision in Fragile and Conflict-affected States. Birmingham, Report prepared for DFID by the Governance and Social Development Resource Center. _____. 2010. “Engagement with Non-state Service Providers in Fragile States: Reconciling State-building and Service Delivery.� Development Policy Review 28(2):131-54. Benn, C. 2003. “Why Religious Health Assets Matter.� Paper presented at ARHAP: Assets and Agency Colloquium, Pietermaritzburg, South Africa, August 2003. 22 Bennett, S., J. D. Quick, and G. Velásquez. 1997. Public-private Roles in the Pharmaceutical Sector: Implications for Equitable Access and Rational Drug Use. Geneva: The World Health Organization. Boulenger, D., B. Keugoung, and B. Criel. 2009. “Contracting Between Faith-based and Public Health Sector in Sub-Saharan Africa: An Ongoing Crisis? The Cases of Cameroon, Tanzania, Chad and Uganda.� Antwerp, Institute of Tropical Medicine, Study commissioned and published by the Medicus Mundi International Network. CCIH. 2005. GRHAM Mapping: Rwanda. Bureau des Formations Médicales Agréees de Rwanda. Christian Connections for International Health. [Accessed 01/10/2007] Christian Health Association of Ghana (CHAG). 2006. Annual Report: June 2005 - May 2006, Accra, Ghana. Christian Health Association of Ghana. [Accessed 10/09/2007]. Christian Health Association of Nigeria (CHAN). 2007. Christian Health Association of Nigeria website. [online]. [Accessed 15/08/2011]. Chand, S., and J. Patterson. 2007. Faith-based Models for Improving Maternal and Newborn Health. Baltimore, USA: USAID-ACCESS. Christian Social Services Commission (CSSC). 2007. “Christian Health Association’s Conference Planning Document.� Tanzania: (CSSC). _____. 2007. “Christian Health Associations at Crossroad Toward Achieving Health Millenium Development Goals.� Christian Health Association’s Conference, 16-18 January 2007, Dar es Salaam. Tanzania : (CSSC) Christian Social Services Commission, pp.1-28. De Jong, J. 1991. “Nongovernmental Organizations and Health Delivery in Sub-Saharan Africa.� World Bank Working Paper, Washington, DC: Population and Human Resources Department, World Bank. DeRoeck, D. 1998. “Making Health-sector Non-governmental Organizations More Sustainable: A Review of NGO and Donor Efforts. Bethesda, MD: Partnerships for Health Reform Project, Abt Associates Inc. Dimmock, F. 2005. “Christian Health Associations in Africa.� Paper Presented at Christian Connections for International Health Annual Conference, Washington, DC. _____. 2007. “Faith-Based Health Networks in Africa.� Paper Presented at Christian Connections for International Health Annual Conference, Washington, DC. 23 ECC. 2007. Eglise du Christ au Congo website. < http://ecc.faithweb.com/> [Accessed 03/08/2007] Gilson, L., P. D. Sen, S. Mohammed, and P. Mujinja. 1994. “The Potential of Health Sector Nongovernmental Organizations - Policy Pptions.� Health Policy and Planning 9(1): 14-24. Government of Malawi. 2001. National Health Accounts (NHA): A Broader Perspective of the Malawian Health Sector, 1998/9 Financial Year. Lilongwe, Malawi: Ministry of Health and Population, Republic of Malawi. Green, A., J. Shaw, F. Dimmock, and C. Conn. 2002. “A Shared Mission? Changing Relationships Between Government and Church Health Services in Africa.� International Journal of Health Planning and Management 17:333-53. Grills, N. 2009. “The Paradox of Multilateral Organizations Engaging with Faith-based Organizations.� Global Governance 15: 505-520. Hanson, K., and P. Berman. 1994. “Non-government Financing and Provision of Health Services in Africa: A Background Paper. Washington, DC: United States Agency for International Development. _____. 1998. “Private Health Care Provision in Developing Countries: A Preliminary Analysis of Levels and Composition.� Health Policy and Planning 13(3):195-211. Hecht, R. M., and V. L. Tanzi. 1993. “The Role of Non-governmental Organizations in the Delivery of Health Services in Developing Countries.� Unpublished report, Washington, DC: World Bank. HERA. 2005. “Public Private Partnership for Equitable Provision of Quality Health Services.�, Technical review. Reet, Belgium: Health Research for Action for the Ministry of Health, Government of Tanzania. Kawasaki, E., and J. P. Patten. 2002. “Drug Supply Systems of Missionary Organizations Identifying Factors Affecting Expansion and Efficiency: Case Studies from Uganda and Kenya.� Boston MA: Boston University. Kintaudi, L., Derstine P and Baer F. 2004. “Faith-based Co-management of Health Zones and Umbrella Projects in DR Congo.� APHA presentation, (APHA) American Public Health Association Meeting, 2004. Kintaudi L. 2006. “Churches and Health Care in DR Congo.� SANRU presentation available online [Accessed 15/08/2011] 24 Mhango, D. 2006. “Innovative Recruitment and Retention Strategies.� Presented at Global Health Council. Washington DC, 31st May. Ministry of Health (MOH)-Zambia. 2002. “Zambia National Health Accounts 1995- 1998.� Lusaka: MOH/ Central Board of Health. Government of Lesotho. MOH-Lesotho. 2007. “Christian Health Association of Lesotho.� Ministry of Health, Government of Lesotho website. [Accessed 24/10/2007]. Munishi, G. K. 1995. “Social Services Provision in Tanzania: The Relationship between Political Development Strategies and NGO Participation.� InService Provision Under Stress in East Africa, edited by J. Semboja and O. Therkildsen. London: James Currey. Mwabu, G. 1986. “Health Care Decisions at the Household Level: Results of a Rural Health Survey in Kenya.� Social Science and Medicine 22(3): 315-319. Mwenda, S. 2007. “HR Crisis in Kenya, the Dilema of FBOs.� Church Health Association Conference, Dar es Salaam/Bagamoyo Tanzania. Christian Social Services Commission. Nussbaum, S., editor. 2005. The Contribution of Christian Congregations to the Battle with HIV and AIDS at the Community Level. Oxford, Global Mapping International for Oxford Centre for Mission Studies. Olivier, J., and Q. Wodon. 2012a. “Playing Broken Telephone: Assessing Faith-Inspired Health Care Provision in Africa.� Development in Practice, 22 (5-6): 819-834. _____. 2012b. “Satisfaction with Faith-inspired Health Care Services in Africa: Review and Evidence from Household Surveys.� In J. Olivier and Q. Wodon, editors, The Comparative Nature of Faith-inspired Health Care Providers in sub-Saharan Africa, HNP Discussion Papers, World Bank, Washington, DC. Republic of Uganda. 2001. “Facility-based Private not-for-profit Health Providers: A Quantitative Survey. Kampala: Ministry of Health in Collaboration with Uganda Catholic, Muslim, and Protestant Medical Bureaux, National Health Consumers Organization and Tropical Business Research. _____. 2001. “Policy for Partnership with Facility-based Private not-for-profit Health Providers.� Kampala: Ministry of Health, Republic of Uganda. Robinson, M., and G. White. 1998. “The Role of Civic Organisations in the Provision of Social Services: Towards Synergy.� In New Patterns of Social Provision in Low Income Countries, edited by G. Mwabu, C. Ugaz and G. White. Oxford: Oxford University Press. 25 Rookes, P. 2009. “Commitment, Conscience or Compromise: The Changing Financial Basis and Evolving Role of Christian Health Services in Developing Countries.� PhD thesis, Birmingham: University of Birmingham. Schmid, B., E. Thomas, J. Olivier, and J. R. Cochrane. 2008. “The Contribution of Religious Entities to Health in Sub-Saharan Africa.� Cape Town: African Religious Health Assets Programme. Todd, S., G. Brubaker, S. Chand, C. Franzén, C. Hafner, A. Kimambo, and P. Pamba. 2009. “Human Resources-geographical Information Systems Data Development and Systems Implementation for the Christian Social Services Commission of Tanzania.� Final report, Washington, DC: USAID and The Capacity Project. Turshen, M. 1999. Privatizing Health Services in Africa. New Brunswick: Rutgers University Press. Wodon, Q. 2013. Faith and Human Development: Education and Health Services in Africa, New York: Palgrave MacMillan. Wodon, Q., M. C. Nguyen, and C. Tsimpo. 2012. “Market Share of Faith-inspired and Private Secular Health Care Providers in Africa: Comparing DHS and Multi-purpose Integrated Surveys.� In J. Olivier and Q. Wodon, eds, The Role of Faith-inspired Health Care Providers in sub-Saharan Africa and Public-Private Partnerships, HNP Discussion Papers, World Bank, Washington, DC. World Health Organization (WHO). 2006. The African Regional Health Report. Geneva, Switzerland: World Health Organization. World Bank. 1983. Country Population and Health Sector Reviews (1983–1992). Unpublished country memoranda, Washington, DC: The World Bank. 26 CHAPTER 2 MARKET SHARE OF FAITH-INSPIRED AND PRIVATE SECULAR HEALTH CARE PROVIDERS IN AFRICA: COMPARING DHS AND MULTI-PURPOSE INTEGRATED SURVEYS Quentin Wodon, Minh Cong Nguyen and Clarence Tsimpo World Bank Much of the evidence used to-date to back up statements about the market share of faith-inspired providers of health care in sub-Saharan Africa comes from data on health care facilities, and especially on the share of hospital beds held by Christian Health Associations in the countries where these associations operate. In those countries, estimates of the market share of faith-inspired health care providers based on hospital beds or similar measures are the 30 percent to 50 percent range. On the other hand, the evidence available from multi-purpose integrated household surveys that ask households where they go for health care and that identify specifically faith-inspired providers in survey questionnaires tells a different story, with lower market shares for faith-inspired facilities. One could ask whether the evidence from these multi-purpose integrated household surveys is itself robust. The objective in this chapter is to assess whether this is the case. Specifically, the idea is to compare market share estimates obtained from different types household surveys, by considering not only multi-purpose integrated surveys, but also Demographic and Health Surveys for which country coverage is larger. The findings suggest that market share estimates for faith- inspired healthcare providers are of a similar order of magnitude in both Demographic and Health Surveys and multi-purpose integrated surveys. INTRODUCTION Market share estimates have often been used as a rather blunt instrument in order to advocate on behalf of faith-inspired providers of healthcare in sub-Saharan Africa to help convince governments and donors to provide more support (financial and otherwise) to them. The problem is that if existing market share estimates are not based on strong evidence, such advocacy efforts may be more detrimental than useful (Olivier and Wodon 2012a), and the available evidence seems at first to be contradictory. Many statements have been made over the years to the effect that about half of all health services in sub-Saharan Africa may be provided by faith-inspired facilities and other faith-affiliated institutions. Examples of statements include that of past World Bank President James Wolfensohn who suggested that “Half the work in education and health in sub-Saharan Africa is done by the church� (quoted by Kitchen 2002), and a recent UNFPA (2009) report stating that that “there is clearly an important parallel faith-based universe of development, one which provides anywhere between 30-60 percent of 27 healthcare and educational services in many developing countries.� There is of course some empirical basis for these statements, but it is often misunderstood. As discussed by Olivier and Wodon (2012a, 2012b), much of the evidence to-date backing up such statements comes from data on the share of hospital beds or related measures in countries where faith-inspired providers such as the Christian Health Associations (CHAs) have a strong presence. Indeed, most estimates of the hospital bed market share of faith-inspired providers in those countries are in the 30 percent to 50 percent range (see De Jong 1991, Dimmock 2007, Chand and Patterson 2007, Gilson et al 1994, Green et al 2002, Grills 2009, Hanson and Berman 1994, Hecht and Tanzi 1993, Kawasaki and Patten 2002, Rookes 2009, Schmid et al 2008, Turshen 1999). However, the evidence provided by Olivier and Wodon (2012b) from multi-purpose integrated household surveys that ask households where they go for healthcare tells a somewhat different story, with substantially lower market shares obtained for faith- inspired facilities. What could explain such different market share estimates between data based on facilities and hospital beds and data from household surveys? As discussed by Olivier and Wodon (2012b), four main factors play a role in the differences between facilities and household survey based estimates of market share. First, within formal healthcare delivery mechanisms, a large share of health services are provided by other types of facilities than hospitals, such as clinics and health centers, as well as maternity homes and facilities from government-run community-based health planning and services which are primary healthcare focused services (sometimes with mobile units). Estimates of the market share of FIIs based principally on hospital beds may lead to overestimating the role of FIIs to the extent that the share of hospital-based care among services provided is often higher among FIIs such as CHAs than in public facilities. Second, in part because private secular providers often operate independently of Ministries of Health, data on their role are often missing, even if one restricts the analysis to measures such as hospital beds. That is, most of the estimates of the share of FIIs are based on comparing FIIs with public sector facilities, without factoring in the existence of facilities operated by private non-religious providers (and for that matter also without factoring in some religious providers that are not part of large federations such as the Christian Health Associations – this is often the case for Islamic clinics and hospitals.) A third issue with the reliance on statistics on hospital beds, or on measures of the use of pharmaceuticals and outpatient care, is that a large share of healthcare is provided by other types of providers that are not included in such statistics. At least two different groups must be mentioned. First, many countries have a significant traditional sector that operates alongside orthodox biomedical care, for example with patients mixing plural health-seeking modalities. While studies on religion and health-related behaviors recognize the role of traditional practices, this is rarely addressed in the literature on the market share of facilities-based faith-inspired providers. Second, self-medication has also been noted to be a significant practice in many countries – given the limited availability of doctors. Some studies suggest that self-medication with privately purchased drugs in some countries represents the most common treatment after home remedies. 28 Finally, many of the countries for which data are available on the share of hospital beds owned by faith-inspired facilities belong to Anglophone Africa. In many (although by no means all) Francophone countries, faith-inspired providers have lower market shares, in part due to colonial policy. The fact that much of the literature has so far focused on CHAs in Anglophone Africa may have led to a bias upward in the assessment of the market share of faith-inspired providers in the region as a whole. These four factors are likely to lead to overestimation in the market share of faith-inspired facilities in health systems as a whole when the main data source being used refers to hospital beds. One could ask however whether the evidence from multi-purpose integrated household surveys provided in Olivier and Wodon (2012b) is itself robust, and useful for the purpose of interrogating these issues. The objective in this chapter is to assess whether this is the case. Specifically, the idea is to compare market share estimates obtained from different types household surveys, by considering not only multi-purpose integrated surveys as done by Olivier and Wodon (2012b; see also Wodon, 2013), but also Demographic and Health Surveys (DHS) for which country coverage is larger. The next two sections provide the findings from the analysis, and a brief conclusion follows. EVIDENCE FROM DEMOGRAPHIC AND HEALTH SURVEYS While DHS surveys do not identify separately faith-inspired healthcare facilities, they do distinguish between three broad types of facilities: public facilities, private facilities, and others. Questions on the type of provider used are asked to individuals for various types of health needs, including fever/cough and diarrhea, as well as contraceptives, antenatal care, delivery, and other. The list of providers in the three broad categories changes slightly depending on the type of care being sought, but in the case of diarrhea for example, the public sector includes government hospitals and clinics, government health centers, government health posts, mobile clinics, fieldworkers, and other public providers. In the case of the private medical sector, the list consists of private hospitals and clinics, pharmacies, private doctors, mobile clinics, fieldworkers, other clinics, maternity homes, and other private medical care. Finally, the ‘other’ category includes shops and markets (and thereby self-medication), traditional practitioners, and drug peddlers. Because most visits to health facilities are related to fever/cough and diarrhea (these are more frequent occurrences in a household than a birth delivery), these are the data reported here first, with additional data provided next. Table 1 provides the estimates of public, private, and other service provision for fever/cough and diarrhea obtained for 36 different countries – while the data for some of the countries are a bit dated, most of the surveys were implemented in the last ten years. In the case of diarrhea treatment, the largest market share for the private medical sector is obtained for Nigeria (45.67 percent market share for visits related to diarrhea treatment, and 54.34 percent market share for visits related to a fever or a cold), while the smallest market share is obtained for Mozambique (2.67 percent for diarrhea and 2.94 percent for fevers/colds). On average, the private medical sector market share is at 17.4 percent for diarrhea treatments and 24.28 percent for fevers/colds (these statistics are simple 29 averages without taking into account different population weights for different countries). Recall that these market shares include both private secular and faith-inspired facilities. By contrast, the average public market share is above 50 percent for both types of illnesses. In other words, the average market share of faith-inspired health facilities that would be inferred from the DHS data should be well below 17.4 percent. For example, if it were assumed that private secular provision in the various countries represents about 40 percent of total private facilities-based provision, then the average market share for faith- inspired providers would be at 10 percent (we will come back to this below). 30 Table 1: DHS Market Shares by Provider for Diarrhea and Fever/Cough (%) Diarrhea treatment Fever and/or cough Public Private Other Public Private Other Burkina Faso 2003 59.94 4.81 35.25 64.26 8.69 27.06 Benin 2006 39.64 15.89 44.46 37.54 20.51 41.95 Burundi 1987 20.15 19.28 60.58 15.93 19.97 64.10 Congo Democratic Republic 2007 74.16 20.15 5.69 37.80 53.83 8.36 CAR 1994-95 72.93 13.98 13.09 79.00 11.70 9.30 Congo (Brazzaville) 2005 46.30 10.94 42.76 57.45 15.39 27.16 Cote d'Ivoire 1998-99 48.24 6.17 45.59 61.83 9.33 28.84 Cameroon 2004 37.08 19.32 43.60 40.62 29.20 30.18 Ethiopia 2005 73.37 18.79 7.84 72.46 24.83 2.71 Gabon 2000 60.00 25.06 14.94 58.85 29.05 12.09 Ghana 2008 56.06 34.28 9.66 54.69 36.79 8.52 Guinea 2005 54.66 5.18 40.16 49.83 10.31 39.86 Kenya 2008-09 64.24 24.71 11.05 61.14 29.41 9.45 Comoros 1996 46.15 17.69 36.15 58.78 20.61 20.61 Liberia 2007 37.71 39.80 22.49 45.47 38.60 15.93 Lesotho 2009 60.11 30.40 9.49 50.58 42.43 6.99 Madagascar 2008-09 64.21 21.32 14.48 59.37 29.17 11.46 Mali 2006 53.14 6.67 40.19 49.36 12.57 38.07 Malawi 2010 71.90 15.03 13.07 69.18 18.21 12.61 Mozambique 2003 86.15 2.67 11.18 83.54 2.94 13.52 Nigeria 2008 36.52 45.67 17.81 35.94 54.34 9.71 Niger 2006 47.22 4.69 48.08 40.15 33.14 26.71 Namibia 2006-07 86.61 8.62 4.77 76.06 19.73 4.22 Rwanda 2005 64.71 16.11 19.18 68.41 14.89 16.70 Sudan 1989-90 50.79 11.59 37.62 - - - Sierra Leone 2008 71.05 19.40 9.55 61.97 27.46 10.58 Senegal 2005 52.43 8.55 39.02 62.07 16.73 21.20 Sao Tome and Principe 2008-09 74.30 22.98 2.72 76.43 19.69 3.88 Swaziland 2006-07 68.33 27.87 3.80 69.16 28.11 2.73 Chad 2004 20.51 2.92 76.57 18.36 4.36 77.29 Togo 1998 48.36 5.46 46.18 45.15 9.81 45.04 Tanzania 2010 22.83 30.50 46.68 22.35 34.18 43.47 Uganda 2006 - - - 47.10 51.53 1.38 South Africa 1998 72.75 25.61 1.64 65.30 33.66 1.04 Zambia 2007 80.71 10.44 8.84 74.28 16.34 9.38 Zimbabwe 2005-06 64.76 16.66 18.59 49.69 22.18 28.13 Average 56.80 17.41 25.79 54.86 24.28 20.86 Source: Authors using DHS surveys. What about some of the other indicators of market share that can be obtained from other types of health-related consultations observed in the DHS data? As shown in table 2, the private medical sector, both private secular and faith-inspired, on average accounts for 28.0 percent of the sources of modern contraceptive methods (the market shares are 54.8 percent for the public sector and 17.2 percent for the '‘others’ category), but only for 9.2 percent of family planning for non-users of modern contraception methods (the estimates are 86.3 percent for the public sector, and 4.5 percent for others). 31 Table 2: DHS Market Shares by Provider for Contraception/Family Planning (%) Source of family planning Source of modern for non-users of modern contraceptive methods contraceptive methods Public Private Other Public Private Other Burkina Faso 2003 53.93 13.77 32.30 91.19 4.83 3.97 Benin 2006 43.11 35.48 21.42 82.88 11.45 5.67 Burundi 1987 86.72 10.46 2.82 - - - Congo Democratic Republic 2007 21.52 60.64 17.84 58.68 33.45 7.87 CAR 1994-95 50.96 32.78 16.26 - - - Congo (Brazzaville) 2005 24.57 29.52 45.91 85.86 10.05 4.08 Cote d'Ivoire 1998-99 26.53 55.09 18.38 - - - Cameroon 2004 21.42 25.27 53.31 73.60 20.75 5.65 Ethiopia 2005 79.50 16.73 3.77 93.72 5.50 0.78 Gabon 2000 26.86 51.42 21.71 - - - Ghana 2008 41.68 54.11 4.21 90.76 8.98 0.26 Guinea 2005 36.19 19.55 44.26 71.04 14.08 14.88 Kenya 2008-09 58.19 36.09 5.73 92.92 5.57 1.52 Comoros 1996 77.83 7.83 14.35 Liberia 2007 54.24 32.53 13.23 78.26 19.06 2.68 Lesotho 2009 56.32 27.88 15.80 88.52 10.95 0.53 Madagascar 2008-09 73.19 19.75 7.07 92.58 6.07 1.35 Mali 2006 53.00 37.66 9.34 75.03 18.93 6.04 Malawi 2010 74.10 21.89 4.00 94.76 4.80 0.44 Mozambique 2003 48.12 7.53 44.35 97.22 0.63 2.15 Nigeria 2008 23.65 61.16 15.19 90.55 7.34 2.10 Niger 2006 68.78 24.60 6.62 89.16 5.93 4.92 Namibia 2006-07 75.05 10.19 14.77 98.40 0.85 0.75 Rwanda 2005 - - - - - - Sudan 1989-90 58.45 35.81 5.74 - - - Sierra Leone 2008 51.93 40.09 7.99 89.99 7.77 2.24 Senegal 2005 69.84 21.96 8.20 96.40 3.17 0.42 Sao Tome and Principe 2008-09 86.86 1.78 11.36 99.07 0.63 0.30 Swaziland 2006-07 48.00 32.85 19.15 82.01 10.88 7.11 Chad 2004 59.99 12.80 27.22 58.48 5.50 36.02 Togo 1998 48.57 14.98 36.45 Tanzania 2010 63.56 21.74 14.70 87.50 7.75 4.74 Uganda 2006 35.06 52.08 12.85 89.02 9.93 1.05 South Africa 1998 84.51 14.56 0.93 Zambia 2007 68.83 16.71 14.46 97.05 2.37 0.57 Zimbabwe 2005-06 67.85 22.04 10.11 84.85 12.37 2.78 Average 54.82 27.98 17.19 86.28 9.24 4.48 Source: Authors using DHS surveys. 32 Finally, in table 3 the market share of the private sector is 6.8 percent for the place of birth delivery for the last birth (45.8 percent for the public sector, and 47.4 percent for others, in part because of a large number of deliveries at home), and 10.6 percent for antenatal care visits (83.2 percent for the public sector, and 6.2 percent for others). Again, faith-inspired providers account for only part of the role played by the private medical sector, so that overall the DHS surveys indicate lower market shares for faith-inspired than estimates based on facilities data such as the share of hospital beds owned by CHAs. Table 3: DHS Market Shares by Provider for Delivery and Antenatal Care (%) Place of delivery Antenatal care Public Private Other Public Private Other Burkina Faso 2003 39.72 0.76 59.52 - - - Benin 2006 65.93 14.59 19.48 - - - Burundi 1987 - - - - - - Congo Democratic Republic 2007 50.90 21.81 27.29 - - - CAR 1994-95 45.65 3.20 51.15 - - - Congo (Brazzaville) 2005 75.89 7.96 16.15 - - - Cote d'Ivoire 1998-99 44.13 0.88 54.99 - - - Cameroon 2004 43.00 18.96 38.05 - - - Ethiopia 2005 5.78 0.61 93.61 87.99 8.20 3.81 Gabon 2000 67.86 17.79 14.35 - - - Ghana 2008 50.74 9.45 39.81 87.84 10.98 1.17 Guinea 2005 30.07 1.80 68.13 - - - Kenya 2008-09 35.31 12.01 52.68 83.13 15.43 1.43 Comoros 1996 42.40 0.11 57.49 - - - Liberia 2007 30.20 11.56 58.24 64.40 16.77 18.83 Lesotho 2009 52.61 4.20 43.19 70.30 7.47 22.23 Madagascar 2008-09 34.44 3.20 62.36 86.26 5.98 7.76 Mali 2006 45.16 2.38 52.46 - - - Malawi 2010 59.37 16.82 23.81 78.92 19.16 1.92 Mozambique 2003 51.55 0.26 48.19 99.03 0.84 0.13 Nigeria 2008 21.31 15.25 63.43 64.89 26.75 8.36 Niger 2006 17.70 0.46 81.84 Namibia 2006-07 77.31 5.11 17.58 90.26 8.34 1.39 Rwanda 2005 49.37 0.91 49.72 97.05 1.91 1.04 Sudan 1989-90 - - - - - - Sierra Leone 2008 23.33 2.86 73.81 84.78 6.39 8.82 Senegal 2005 59.97 4.27 35.76 - - - Sao Tome and Principe 2008-09 80.43 0.25 19.32 98.59 0.25 1.16 Swaziland 2006-07 60.36 5.59 34.05 70.20 25.31 4.49 Chad 2004 11.58 0.52 87.90 - - - Togo 1998 45.99 4.44 49.57 - - - Tanzania 2010 25.40 2.90 71.71 69.89 4.85 25.26 Uganda 2006 33.50 13.13 53.37 87.66 11.09 1.26 South Africa 1998 77.87 8.48 13.65 - - - Zambia 2007 45.63 4.94 49.43 92.53 6.53 0.93 Zimbabwe 2005-06 56.67 12.93 30.40 84.41 14.73 0.86 Average 45.80 6.78 47.43 83.23 10.61 6.16 Source: Authors using DHS surveys. 33 COMPARISON WITH MULTI-PURPOSE INTEGRATED SURVEYS How do estimates obtained from DHS data compare with those obtained from multi- purpose integrated household surveys that identify separately faith-inspired providers from private secular providers? In order to answer this question, estimates of the market share of faith-inspired and private secular providers were obtained by Olivier and Wodon (2012b) using multi-purpose household surveys for 14 countries. Details on the surveys and the categories of health care providers identified in each survey are provided in the annex. Before presenting the results, three brief comments are in order. Firstly, in many cases, NGOs have been included together with faith-inspired providers. This is because in a few countries, NGOs are lumped together with faith- inspired providers in the way questions are asked in the surveys. To keep the data consistent, where NGOs are identified separately, they have then also been considered as faith-inspired providers. However, because the market share of NGOs is typically much smaller than that of faith-inspired providers, this should not lead to any major bias in the results, and in addition, at least some NGOs are faith-inspired. Secondly, traditional and faith healers have been included in the ‘private secular’ category in order to focus the assessment of faith-inspired providers on facilities-based care. Considering traditional and faith healers as private secular providers is a bit of a misnomer given that ‘faith’ often inspires traditional healers, but the term ‘private secular’ has been kept in order to simplify the terminology. Because the role of traditional and faith healers is often small in comparison to that of pharmacists and chemical stores (due to self-medication), this again should not lead to a large bias in most countries. Thirdly, because of the way questions are asked in the surveys, the identification of faith-inspired providers in some countries may be better than in others, but it is difficult to make a precise country-level assessment of the quality of this identification, so this will not be attempted here. Table 4 provides the results already presented by Olivier and Wodon (2012b). The order of magnitude for the market share estimates tend to be comparable to the estimates obtained from DHS surveys, at least on average. Indeed, the average market share of public facilities-based care across the 36 countries for which DHS estimates are available is 55.9 percent for consultations related to fevers/coughs and 56.8 percent for consultations related to diarrhea. This compares to an average market share for public facilities in the multi-purpose integrated surveys of 55.2 percent. In the subset of countries where multi-purpose household surveys identify separately private secular and faith-inspired facilities, the average market share of faith-inspired providers is 6.0 percent. As already mentioned this is probably on the low side because some countries where faith-inspired providers are known to have very large market shares such as the Democratic Republic of Congo are missing. Overall, it would seem to be fair to say that in the multi-purpose integrated surveys, the market shares for faith-inspired in healthcare systems broadly defined may well at 10 to 15 percent, with a market share of other private facilities-care being of a similar order of magnitude if one considers in turn the DHS data. Note finally, in table 4, the market share of the private secular sector broadly conceived, which includes here not only private secular medical care, but also chemical stores, pharmacies, and traditional healers, tends to be larger than that of faith-inspired 34 providers. This is not surprising given that essentially the definition of the private sector in table 4 includes the ‘others’ category in the tables based on the DHS data. Table 4: Market Shares from Multi-purpose Integrated Household Surveys (%) Public healthcare Faith-inspired Private secular providers providers providers Burundi, 2006 69.3 11.5 19.2 Cameroon, 2007 44.9 15.1 40.0 Chad, 2003/04 53.1 10.7 36.2 Ghana, 2003 43.1 3.7 53.2 Ghana, 2005-06 44.4 6.6 49.0 Kenya, 2005 49.0 4.2 46.8 Malawi, 2004 36.9 3.9 59.2 Mali, 2006 68.5 1.0 30.5 Niger, 2007 77.6 1.5 20.9 Nigeria 2003/04 50.2 1.9 47.9 Republic of Congo, 2005 44.0 4.0 52.0 Senegal, 2005 65.0 2.3 32.7 Sierra Leone, 2003-04 60.1 2.0 37.9 Swaziland, 2009-10 66.4 13.2 20.3 Zambia, 2004 55.0 6.1 38.9 Average 55.2 5.8 39.0 Source: Authors using multi-purpose integrated surveys. See also Olivier and Wodon (2012b). For easier reference, table 5 provides a synthesis of the average results obtained across countries with the two types of surveys, focusing on diarrhea and fevers/colds in the case of DHS data since consultations related to these two illnesses tend to be more frequent than for the other types of illnesses and health needs identified in the DHS. Table 5: Comparison of Average Market Share Estimates Between Surveys (%) Public All private Other Faith-inspired All other Total facilities facilities private facilities Private DHS, Fevers/coughs – 36 countries 54.9 24.3 20.9 - - 100.0 DHS, Diarrhea – 36 countries 56.8 17.4 25.8 - - 100.0 Multi-purpose – 14 countries 55.2 - - 5.8 39.0 100.0 Source: Authors. Importantly however, while on average the results obtained with the DHS and the multi- purpose integrated surveys tend to be of the same order of magnitude, as illustrated by similar average market shares for public providers of healthcare in the two sets of surveys, there is a lot of variation in estimates at the level of individual countries depending on which data source is used. This is best illustrated with country level public market shares which are displayed in Figure 1 for the set of countries for which both DHS and multi-purpose household surveys are available. The scatter plot in Figure 1 suggests that for many countries in the sample, there are large differences in public market share estimates depending on the type of survey used, and the relationship between the two sets of public market shares is essentially flat (it is actually negative). The likely reason for this lack of fit between the two sets of estimates is that what is measured is often different in the DHS and multi-purpose integrated surveys, both in terms of the types of illnesses considered and in terms of the definitions of facilities. The 35 questions are asked in the multi-purpose integrated surveys in a different way from what is asked in the DHS surveys. In addition, the comparison of estimates also suggests that the data is somewhat noisy, possibly because of recall problems on the part of households. Thus, while it is legitimate to draw general conclusions about the order of magnitude of faith-inspired providers from those data, in that their market share is likely to be below the levels suggested by facilities data, caution is needed for work at the country level, and data from household surveys must be triangulated whenever feasible. Also, the substantial variations in estimates between surveys suggest that more efforts are needed to collect good data on the respective market shares of various types of providers. Figure 1: Comparison of DHS and Multi-purpose Surveys 90 Public market share in DHS syrveys, diarrhea (%) 80 70 60 50 40 30 20 10 0 0 10 20 30 40 50 60 70 80 90 Public market share in multi-purpose surveys (%) Source: Authors. CONCLUSION The purpose of this short paper was to compare market share estimates for private facilities in DHS and multi-purpose integrated surveys in a large sample of sub-Saharan countries. The DHS surveys have the advantage of being comparable and standardized, and they are also available for a large set of countries. The multi-purpose integrated surveys selected for this analysis have the advantage of distinguishing between faith- inspired and private secular facilities, which is not the case in the DHSs. On average, the results obtained for the countries in the various sample suggest that the market share of faith-inspired providers in health systems broadly defined are below traditional estimates based on facilities data, and especially on the number of hospital beds owned by CHAs. Several reasons already discussed by Olivier and Wodon (2012b) account for this result. Firstly, faith-inspired providers often have larger shares of hospital beds than their share of all health care facilities. Secondly, the market shares of CHAs in hospital beds often 36 do not account for the role of private secular health facilities. Thirdly, estimates based on the CHAs tend represent the situation of countries where faith-inspired providers have traditionally been strong, with lower market shares in other countries. Finally, estimates of market shares based on facilities-based care do not account for the role of a range of other providers, including traditional healers, pharmacists, and chemical stores. These various factors tend to generate lower market shares in household surveys than in databases that rely on hospital beds or closely associated measures. As in Olivier and Wodon (2012b), the take-away from this analysis should not be that facilities-based estimates are proven ‘wrong’ by household surveys. Household surveys might underestimate the market share of some faith-inspired facilities when households do not know that their facility is faith-inspired. In addition, from the point of view of a Ministry that supports faith-inspired facilities in their service delivery activities, the market share of these facilities in the universe of faith-inspired and public facilities (as measured through hospital beds among others) is important for allocating funding, and probably more so than estimates of market share that take the broader health system into account. Thus, various types of estimates – whether based on facilities or household surveys – can and should be used for various purposes at the country or local level. Furthermore, even if estimates from DHS and multi-purpose integrated surveys tend to converge on average, there is a lot of variation in the estimates obtained from different sources at the level of individual countries, in part because different surveys measure different things. Clearly, when designing household surveys, more efforts should be undertaken to collect better data on the respective roles of various types of health care providers, and for this more detailed survey questionnaires would be highly welcome. REFERENCES Chand, S., and J. Patterson. 2007. Faith-based Models for Improving Maternal and Newborn Health. Baltimore, USA: USAID-ACCESS. De Jong, J. 1991. “Nongovernmental Organizations and Health Delivery in Sub-Saharan Africa. World Bank Working Paper, Washington, DC: Population and Human Resources Department, The World Bank. Dimmock, F. 2007. “Faith-Based Health Networks in Africa.� Paper presented at Christian Connections for International Health Annual Conference, Washington, DC. Gilson, L., P. D. Sen, S. Mohammed, and P. Mujinja. 1994. “The Potential of Health Sector Nongovernmental Organizations - Policy Options.� Health Policy and Planning 9(1): 14-24. Green, A., J. Shaw, F. Dimmock, and C. Conn. 2002. “A Shared Mission? Changing Relationships between Government and Church Health Services in Africa.� International Journal of Health Planning and Management 17:333-53. 37 Grills, N. 2009. “The Paradox of Multilateral Organizations Engaging with Faith-based Organizations.� Global Governance 15: 505-520. Hanson, K., and P. Berman. 1994. “Non-government Financing and Provision of Health Services in Africa: A background paper.� Washington, DC: United States Agency for International Development. _____. 1998. “Private Health care Provision in Developing Countries: A Preliminary Analysis of Levels and Composition.� Health Policy and Planning 13(3):195-211. Hecht, R. M., and V. L. Tanzi. 1993. “The Role of Non-governmental Organizations in the Delivery of Health Services in Developing Countries.� Unpublished Report, Washington, DC: World Bank. Kawasaki, E., and J. P. Patten. 2002. “Drug Supply Systems of Missionary Organizations Identifying Factors Affecting Expansion and Efficiency: Case Studies from Uganda and Kenya.� Boston MA: Boston University. Kitchen, M. 2002. “World must Coordinate Efforts, End Waste, Says Wolfensohn.� UN Wire http://www.unwire.org/unwire/20021025/29832_story.asp. Olivier, J., and Q. Wodon. 2012a. “Playing Broken Telephone: Assessing Faith-Inspired Health Care Provision in Africa.� Development in Practice, 22 (5-6): 819-834. _____. 2012b. “Market Share of Faith-inspired Health Care Providers in Africa: Comparing Facilities and Multi-purpose Integrated Household Survey Data.� In J. Olivier and Q. Wodon, eds, The Role of Faith-inspired Health Care Providers in sub-Saharan Africa and Public-private Partnerships, HNP Discussion Papers, World Bank, Washington, DC. Rookes, P. 2009. “Commitment, Conscience or Compromise: The Changing Financial basis and Evolving role of Christian Health Services in Developing Countries.� PhD thesis, Birmingham: University of Birmingham. Schmid, B., E. Thomas, J. Olivier, and J. R. Cochrane. 2008. “The Contribution of Religious Entities to Health in Sub-Saharan Africa.� Cape Town: African Religious Health Assets Programme. Turshen, M. 1999. Privatizing Health Services in Africa. New Brunswick: Rutgers University Press. United Nations Population Fund (UNFPA). 2009. “Guidelines for Engaging Faith-based Organizations (FBOs) as Agents of Change.� New York. Wodon, Q. 2013. Faith and Human Development: Education and Health Services in Africa, New York: Palgrave MacMillan. 38 Annex Table 1: Multi-purpose Integrated Household Surveys Used for Cross-country Comparisons Year Survey Name Number of Country households Burkina Faso 2007 Enquête Annuelle sur les conditions de vie des ménages (EACVM-QUIBB) 8496 Burundi 2006 Enquête Questionnaire des Indicateurs de Base du Bien être (QUIBB) 7046 Cameroon 2007 Enquête sur les Conditions des Ménages Camerounais II (ECAM) 11391 Chad 2003/04 Deuxième Enquête sur la Consommation et le Secteur Informel au Tchad (ECOSIT) 6697 Dem. Rep. of Congo 2004/05 Enquête 1-2-3 (123 survey) 12098 Ghana 2005/06 Ghana Living Standards Survey, Fifth round (GLSS) 8687 Kenya 2005 Kenya Integrated Household Budget Survey (KIHBS) 13158 Malawi 2004 Malawi Integrated Household Survey (HIS) 11280 Mali 2006 Enquête Légère. Intégrée auprès des Ménages (ELIM) 4494 Niger 2007 Enquête nationale sur le budget et la consommation des ménages (ENBC) 4000 Nigeria 2003/04 Nigeria Living Standards Survey (LMS) 19158 Republic of Congo 2005 Enquête Congolaise auprès des Ménages pour l'évaluation de la pauvreté (ECOM) 5002 Senegal 2005 Enquête de Suivi de la Pauvreté au Sénégal (ESPS) 13568 Sierra Leone 2003 Sierra Leone Integrated Household Survey (SLIHS) 3720 Swaziland 2009 Swaziland Household Income and Expenditure Survey (SHIES) 3167 Uganda 2010 Uganda National Household Survey (UNHS) 6775 Zambia 2004 Zambia Living Conditions Monitoring Survey (LCMS) 19315 Source: Authors. 39 Annex Table 2: Identification of the Various Types of Healthcare Facilities in Multi-purpose Integrated Household Surveys Public Faith-inspired and NGOs Private secular Burundi Public hospital, Public clinic Missionary hospital, Private hospital, Private clinic, Pharmacy, Private doctor, Missionary clinic Private Sage-femme, Traditional healer Cameroon Public, Para-public Private religious Private non-religious Chad Public health center, Public district Religious health center and Private health center, Private clinic, Private doctor or hospital, HGRN/Liberty NGO, Religious/NGO hospital dentist, Home-based care Ghana Public Private religious Private non-religious Kenya Referral hospital, District/provincial Missionary hospital/dispensary Private dispensary/hospital, Private clinic, Traditional hospital, Public dispensary, Public health healer, Pharmacy/chemist, Kiosk, Faith healer, Herbalist, center Other Malawi Government health facility Church/mission facility Private health facility, Local pharmacy, Local grocery for medicine, Treatment with traditional healer, Treatment with faith healer, Help from relatives, Other Mali Public hospital, CSCOM, CSRef, Other Religious health center Private doctor or dentist, Private care facility, Private clinic, public facilities Traditional healer/marabou, Pharmacy, Other private facilities or NGOs Niger Public hospital, Integral health center, Private Christian/NGO facility Private hospital/clinic, Private health facility, Pharmacy, Maternity, Health post Traditional healer/marabou, Others Nigeria Federal Government, State Government, Religious Body Industrial, Private, Other Local Government Republic of Congo Public hospital/clinic, Integral Health Church Private hospital/facility, Private doctor or dentists, Post Traditional healer, Pharmacy, Other Senegal Hospital/Clinic/Dispensary, Christian/NGO Private Doctor/Dentist, Healer/Marabou, Sage Hospital/Health Center, Femme/Neighborhood nurse, Pharmacy/Pharmacist, Other Dispensary/Health Post, Health Hut Sierra Leone Government NGO, Missionary, Catholic Private, Other Govt Hospital, Govt Health Centre, Govt Mission Hospital, Mission Private Hospital, Private Doctor, Traditional Healer, Swaziland Clinic Clinic Pharmacy, Other Zambia Government Institution Mission Institution Industrial Institution, Private Institution, Pharmacy, Relatives, Neighbors, Friends, Traditional Healers, Other Source: Authors. 40 CHAPTER 3 HÔPITAL MATLABOUL FAWZAINI: AT THE INTERSECTION OF DIASPORA, FAITH, AND SCIENCE IN TOUBA, SENEGAL Ellen E. Foley and Cheikh Anta Babou 5 Clark University and University of Pennsylvania In this paper we present a case study of a development initiative spearheaded by the members of a transnational diaspora, the creation of a medical hospital in the holy city of Touba in central Senegal. Although the construction of the hospital is decidedly a philanthropic project, Hôpital Matlaboul Fawzaini is better understood as part of the larger place-making project of the Muridiyya 6 and the pursuit of symbolic capital by a Mouride dahira 7. The hospital illuminates important processes of forging global connections and transnational localities and underscores the importance of understanding the complex motivations behind diaspora development. We argue that the hospital initiative represents a distinct form of development that is not easily incorporated into state schemas for urban development or public health. The hospital’s history reveals the delicate negotiations between state and non-state actors and diaspora organizations and the complexities of public-private partnerships for development. Evidence of the Mouride dahira’s profound commitment to building the holy city of Touba, their ability to persevere in the face of significant technical and financial obstacles, and their recent internal crisis of leadership are crucial to understanding this Muslim Sufi brotherhood’s resilience, dynamism, adaptability, and future trajectory. INTRODUCTION While many parts of the world continue to experience a set of transformations commonly glossed as globalization, much of the African continent occupies a tenuous position in this new economic landscape. Many urban and rural regions suffer from the neglect of diminished states that no longer have the means or the mandate to address escalating poverty, inadequate public services, or the dearth of livelihoods. In response to worsening conditions, many Africans are emigrating to larger cities, different countries, and new continents in search of social and material opportunities that they lack at home. Once thought of as a straightforward path from departure to assimilation, migration is now understood as a process in which migrants establish social formations within a 5 A version of this paper was previously published in African Affairs 110/438 - the authors and publishers are thanked for their permission to republish the paper in this collection. 6 The Muridiyya is a Sufi order that originated in Senegal in the early 20th century. It is the most widespread and fastest growing Muslim brotherhood in Senegal. 7 Dahiras are associations of Mourides with a shared allegiance to a particular marabout or location. 41 transnational space as they continue to participate in cultural, economic and political life in their home and host countries (Levitt and Jaworksy 2007). The establishment of African diaspora communities throughout Europe and North America has fuelled new kinds of circulation of people, goods, and values (Babou 2002, Bava 2003, Diouf 2000). In spite of the sociological richness and complexity of these diasporas, one dimension in particular has caught the imagination of the international development community - the large and increasing volume of remittances that migrants send home. Globally, the value of migrant remittances is greater than official development assistance and foreign direct investment; recorded remittances doubled between 2002 and 2007 (Ratha 2007). 8 African migrants have participated in this growing trend. Although North Africa receives the greatest volume of remittances, in 2007 sub-Saharan Africa received US $10.8 billion. The top five receivers were Nigeria, Kenya, Sudan, Senegal and Uganda (Ratha and Xu 2008). In 2007 alone Senegal received nearly US $1 billion dollars from its diaspora. As remittances and diaspora projects now overshadow traditional forms of development aid, scholars and development professionals have begun to assess the impact of remittances and the social forces behind them. Some question whether transfers promote equity and sustainable development or aggravate inequity (Merz et al 2007). Others suggest that diasporas function as part of the unfolding of global capitalism that has usurped state-led ‘development’ (Mohan 2006). Grillo and Riccio (2004) highlight the potential of translocal development in which non-state entities in the North and South, including migrant associations, collaborate on projects. Still others argue that diaspora initiatives might transform the very notion of development itself, and urge caution when attempting to incorporate migrant efforts into normative understandings of development (Mercer et al 2008). In spite of these questions, enthusiasm for ‘diaspora development’ remains high. Remittances have not yet attained the status of magic bullet, but they are seen as a promising source of resources for attaining the Millenium Development Goals’ and as a means to ‘leverage funds for development’ (Maimbo and Ratha 2005, Ratha 2007). In this paper we seek to move beyond the current enthusiasm for diaspora development to examine its numerous paradoxes. We do so through an analysis of Hôpital Matlaboul Fawzaini, an initiative spearheaded by a Mouride dahira whose members live primarily in North America and Western Europe. 9 Although the hospital project is decidedly philanthropic, we contend that Hôpital Matlaboul Fawzaini is better understood as part of the larger ‘place-making project’ of the Muridiyya and the pursuit of symbolic capital by an increasingly powerful dahira (Babou 2007, Copans 1988, O’Brien 1971, Sy 1960).The notion of place-making draws attention to the processes by which Touba’s material transformation (physical expansion, demographic growth, creation of ritual spaces) becomes infused with religious significance. The dahira’s insistence on 8 Most scholars estimate that officially recorded remittances make up 50% or less of total remittances. 9 Dahira are mutual assistance groups of Mouride disciples that range from dozens to thousands of members. 42 achieving Shaikh Amadou Bamba’s holy vision of Touba first and foremost, and meeting the needs of its residents second, underscores the importance of understanding the complex and multiple motivations behind diaspora development. By building Hôpital Matlaboul Fawzaini, the Mouride dahira entered the realm of large capital projects the likes of which typically fall under the mandate of the developmentalist state. Nonetheless, we argue that this project represents a distinct form of development, and one that is not easily incorporated into state schemas for urban development or public health (Mercer et al 2008). Equally important, the hospital’s history reveals the delicate and often contentious negotiations between state and non-state actors and diaspora organizations. The hospital project’s complicated and costly road to success highlights the pitfalls of the neoliberal age in which ‘development’ has been relegated to actors in the private sphere who attempt to close the gap between needy populations and an ineffective state (Mohan 2006). Our analysis of how dahira are positioning themselves as development players remaking the Mouride homeland also offers a new dimension to scholarship on the Muridiyya.10 The Mourides were once known as a Sufi order comprised largely of rural peasants who laboured on their own fields and in the fields of their religious leaders (Copans 1988). Over the past 30 years they have transformed into an increasingly urban, transnational tariqa (Sufi order) whose followers have come to rely on local, national, and international commerce as their primary economic strategy. Touba, the religious capital of the Mourides, founded by Sheikh Amadou Bamba Mbakke in the 19th century, is now Senegal’s second largest city, with close to one million inhabitants. Recent research on the Mourides examines their migration across Africa, Europe, and North America, how they reproduce their institutional structures in new social settings, and how they have invented new circuits of accumulation (Diouf 2000). As a project of the Mouride disapora, the hospital illuminates the processes through which Mourides emigrants forge global connections and capitalize on the opportunities afforded by their transnational social networks. The dahira’s profound commitment to building the holy city of Touba, their ability to persevere in the face of significant technical and financial obstacles, and their internal crisis of leadership around the time of the hospital’s completion are crucial to understanding the Muridiyya’s resilience, dynamism and adaptability. ‘MONEY AND KNOW-HOW ARE NOT THE SAME’: HÔPITAL MATLABOUL FAWZAINI Hôpital Matlaboul Fawzaini is the first privately-funded hospital ever built in Senegal by a Muslim organization, the dahira of the same name. Mouride dahira have been a familiar feature in Senegal for the past several generations. Dahira initially helped ease 10 Three major trends can be discerned on the scholarship on the Mourides. Early scholars were concerned with the order’s political significance as an instrument for the adaption of the Wolof ethnic group to colonial rule. A second trend focused on the economic dimension of the Muridiyya and emphasized its contribution to the expansion of the colonial cash crop peanuts. A later trend explores the urbanization of the order and the emergence of a global Mouride diaspora (Babou 2007). 43 the sense of dislocation of newly arrived Mouride dwellers in Senegalese cities (Babou 2002). Dahira in Africa, Europe, and North America form networks that facilitate travel, insertion into host communities, and employment. While some dahira focus on mutual assistance among members, others contribute to public works projects and serve as civil society partners for the state and non-governmental organizations. Members of Matlaboul Fawzaini collected annual dues and special contributions for the hospital project in Africa, Europe, and North America over a period of ten years. Between the time of groundbreaking and the completion of the hospital in 2003, the various chapters of the dahira contributed an astounding 5 billion FCFA (US $10 million). 11 Upon its completion, Matlaboul Fawzaini handed the hospital keys to the Mouride Khalife-General at the time, Serigne Saliou Mbakke. 12 Serigne Saliou designated the Ministry of Health as the structure that would become responsible for ensuring the hospital’s ongoing operations. Five individuals, including the hospital director and his chief of staff, arrived at the hospital structure early in 2005 with the charge of making it operational; they found empty buildings with no furniture, no medical equipment, and no supplies. Less than 18 months later the hospital employed nearly 200 clinical and auxiliary personnel, the majority of whom were being paid with the hospital’s own revenue. It has since become a Level Three National Hospital under the jurisdiction of the Senegalese Ministry of Health, and it is the only hospital to have this technical designation outside of the capital city Dakar. The earliest formulation of the organization that was to become Matlaboul Fawzaini emerged under the leadership of Daam Ndiaye, a dynamic Mouride emigrant who settled in Spain in the mid 1980s (Thioune 2007). He became president of the Mouride dahira of Madrid, and was the first Senegalese to own a boutique in Spain. In 1990, upon the occasion of Serigne Modou Bousso Dieng Mbakke’s visit, Ndiaye launched the idea of a project that would be funded by all of the Mouride dahira in Spain: a project to build a hospital in Touba. 13 In 1991 Ndiaye received the Khalife-General’s permission to begin the project and he created the Association des émigrés sénégalais au service de Cheikh Amadou Bamba (Gueye 2002). Members of Ndiaye’s association began collecting funds for the hospital, but financial support faltered after a year. After convening a meeting in Touba in 1994 with at least a 11 The hospital project is exceptional for its duration, scale, and the amount of money raised. In his study of Mexican hometown associations, Orozco (2003) found that most groups attempted projects that cost US $10,000 or less and that few associations last longer than ten years. 12 The Khalife-General is the religious head of the Mourides. After Bamba’s death in 1927 his sons and then grandsons succeeded one another in this position at the top of the organization. 13 The late Serigne Modou Bousso Dieng Mbakke was the elder son of Serigne Falilou Mbakke, second caliph of the Muridiyya (1945-1968). Serigne Modou Bousso first grabbed the limelight when he challenged the leadership of his uncle, Caliph Abdou Lahad (1968-1988). They reconciled and he gained notoriety for the supposed efficacy of his prayers to help people get visas, especially for Europe. He travelled abroad frequently and had a large following among the Mouride diaspora. 44 thousand Mourides in attendance, Ndiaye finally launched a dahira with sufficient institutional heft to manage the project. The dahira was named Matlaboul Fawzaini, which references a poem written by Amadou Bamba in the 1880s after he discovered the site that would become Touba. Matlaboul fawzaini means ‘the search for two happinesses’ (in this world and the next) and it conveys Bamba’s utopian vision for the city (Gueye 2002, Ross 2006). The poem is essentially a supplication: Bamba prayed that God bless Touba and its inhabitants and that He make of Touba a sanctified city of light, faith, peace and prosperity. The new dahira set as its central purpose the realization of Bamba’s prayer. Once Matlaboul Fawzaini was established, Mourides became members by purchasing membership cards in sums determined by geographic location and based on an estimation of the relative affluence of the Mouride community in question. Membership cards in African countries cost considerably less than cards sold in Europe and North America. Membership dues were renewed annually in the form of sas, a longstanding Mouride fundraising practice. 14 Matlaboul Fawzaini soon became the first dahira to take the form of a non-governmental organization (NGO) with bylaws, board members and a salaried staff. It is now recognized by the government of Senegal as a tax-exempt non-profit organization. The dahira has over 60,000 members with national chapters and a global headquarters in Touba. Daam Ndiaye served as president throughout the 1990s and beyond the completion of the hospital project (Gueye 2002). The hospital’s construction began in 1994 on a site donated by the Khalife-General. A key turning point came two years later when Serigne Saliou selected a group of medical professionals to tour the construction site and to provide feedback on the project. Taking advantage of the Minister of Health’s visit to Touba, a strategic meeting was organized that assembled dahira members, representatives of the Ministry of Health, and other technical advisors. Although this meeting was an occasion for representatives of the state to influence the hospital project, these experts were hand-selected by the Khalife-General and many of them were Mourides. The man who became the first hospital director, Falilou Diop, was one of the more prominent Mouride civil servants who attended the meeting (Saar 2006). This technical consultation, which occurred after the dahira had spent approximately US $1.2 million dollars, was clearly long overdue. The advisors concluded that the construction to date was completely inappropriate, and they had the unpleasant task of telling the dahira members that the new structures needed to be torn down. As Falilou Diop explained, the technical team had two options. They could ignore the problems and let construction continue, or they could provide honest feedback about the numerous problems in evidence (Diop 2006). (The building materials were not up to code, doorways could not accommodate stretchers and other medical equipment, and so on). 14 The Wolof word sas can be translated as load, assignment, or duty. It refers to a cash sum or amount of work that Mouride leaders request from their disciples individually or collectively to complete a task. During the course of building the hospital the highest annual sas requested by the dahira amounted to 100,000 CFA ($200) (Thioune 2007). 45 The technical team told the taalibe, “what you have built, it is too small. It is no good. If you want this to be a first-rate hospital, you have to knock down what has been built and start over� (Saar 2006). Heated discussions between the technicians and the dahira members followed, with the taalibe accusing the experts of being ‘soxor’ (malicious or mean). There was talk of converting the hospital into a medical centre, but the dahira members decided to persevere and build a Level III hospital. The Khalife-General contributed to the compromise by donating more land so the dahira could expand the structure’s footprint. Nearly 80 percent of the structure was demolished and construction began again with guidance from the technical team. Underlying the willingness of the dahira to overcome this setback was the acknowledgement that this medical structure was being built to honour the memory, vision, and spiritual guidance of the Mouride’s founding saint. The divine inspiration behind the project contributed in large part to the dahira’s determination to see its completion. Although the dahira finished the hospital, the poor formulation of the original plans has had lasting effects. Doctors and administrators concur that aside from financial problems, the current size of the hospital is their main operational constraint (Dieng et al 2006, Saar 2006). There is not enough space for the current medical personnel and no room for expansion to add other specialties. In 2006 the physical therapy unit was operating in the laundry facility and other specialties were doubled-up and practising in the same wings. This lack of space continues to pose problems for the hospital’s day to day operations (Sheff 2009). As the chief of staff reflected, “money and know-how are not the same.� In spite of the dahira’s extraordinary capacity to mobilize millions of dollars for a large project, they did not have the technical expertise to ensure its success. Without the intervention of the Khalife-General, who was the impetus for the technical consultation, many more millions would have been spent on a potential failure. When the construction of the hospital was nearly complete in 2002 the dahira members presented it to the Khalife-General, who in turn gave the hospital to President, Abdoulaye Wade. Serin Saliou told Wade that the Mourides had built the structure to serve the populace of Touba, but that it was his desire that the Ministry of Health oversee its operation. The Khalife-General’s gift of the hospital to the state, in the person of Senegal’s first Mouride president, is charged with numerous socio-political ramifications. 15 It is an admission of the brotherhood’s inability to run a complex medical structure, but it also served as an opportunity to leverage state resources for Touba. The Mourides have always been suspicious of state encroachment in Touba and Sheikh Saliou’s attitude was consistent with the tacit division of labour that governs Mouride relationships with the Senegalese government. Mourides have invested millions of dollars to refurbish the Great Mosque, to establish a modern Islamic library, and to build 15 The history of the Mourides is one of confrontation and accommodation to the French colonial state, and compromise with the post-colonial state for influence over Senegal’s citizens and disciples (O’Brien 2003, Robinson 2001, Villalon 1995). 46 glamorous houses for distinguished guests, but they have also shied away from shouldering investments they see as the state’s responsibility. With the Matlaboul Fawzaini hospital, dahira members agreed to raise money for the hospital’s construction, but after this initial investment the Khalife-General delegated the cost of running the hospital to the state. In late 2004 the Minister of Health, Dr. Issa Mbaay Samb, contacted Falilou Diop, then the director of one of the most renowned hospitals in the country. In Diop’s words, “Samb approached me, he told me that the Mourides have built the hospital, and they want me to come run it� (Diop 2006). Diop agreed to leave his prestigious post in Dakar and he arrived in Touba in late 2004 with a team of four other people. They spent over US $200,000 fixing the remaining construction mistakes and bringing the hospital up to code; the team successfully transformed a set of empty buildings into a functioning hospital in less than four months. President Wade and the Minister of Health inaugurated the hospital on March 25, 2005. By 2006 the hospital employed 190 people, only 60 of whom were civil servants paid by the Ministry of Health. 16 The hospital paid the remaining staff with funds generated by its own activities.17 Hôpital Matlaboul Fawzaini offers a significant example of the capacity of overseas emigrants to conceptualize, fund, and implement projects that would normally fall within the portfolio of the state. The financial and human resources mobilized by the dahira bolster the arguments of scholars who see in diaspora organizations tremendous potential to become development actors. Yet the pursuit of ‘development’ is not the raison d’être of Matlaboul Fawzaini. Instead it has dedicated itself to a central Mouride objective: building the holy city of Touba as the realization of the founding saint’s holy vision. Matlaboul Fawzaini’s hospital is less an achievement of ‘development’ than of bringing Touba a step closer to Bamba’s vision for the city, which increases the symbolic capital of the dahira at the same time. Throughout the Muridiyya’s history realizing Bamba’s dream for Touba has served as its central project. Bamba established the city after being visited in a vision by the angel Gabriel, who told him that the site would become a spiritual metropolis of universal significance (Gueye 2002, Ross 2006). After Bamba’s death in 1927 his sons and successors began the actual construction of the city. The Great Mosque in Touba, the site of Bamba’s mausoleum, was constructed over several decades and is the largest mosque in sub-Saharan Africa. Touba is a spiritual and an economic beacon; for the past 50 years it has attracted rural migrants seeking relief from drought and ongoing economic crisis. Its population growth over the past 50 years has astounded scholars and politicians alike - Touba counted 2,127 residents in 1958, and this number grew to 125,000 by the late eighties. The population reached 300,000 by 1993, and current estimates are between 800,000 one million (Gueye 2002). 16 By 2008 the hospital staff had grown to 260 employees (Sheff 2009). 17 Like all government health structures in Senegal, the hospital is run on a model of partial cost recovery. It charges user fees for medical appointments and for hospitalization, and also sells pharmaceuticals and other medical supplies to patients. 47 Not only has the building city of Touba been the central project of the Mourides, but it has been accomplished with little assistance from non-Mourides, including the government of Senegal. Touba is a socio-political enclave within the Senegalese state; the Khalife-General and his advisors oversee Touba’s expansion and development. The Muridiyya reserves the right to alienate and distribute land in Touba. Urban settlement has proceeded in successive waves of lotissement organized by the khalifes (Ross 2006). The bureaucratic and rationalizing impulse of the Senegalese state has never successfully reached Touba and it has remained largely illegible and impenetrable to outsiders, from state authority to development actors (Scott 1998). In spite of its unique administrative status, Touba faces similar challenges to other rapidly expanding locales. The demands of urban life pose new challenges for the Mouride leadership, particularly in the context of extending infrastructure such as sanitation, health care, and housing to meet the needs of Touba’s growing population. What distinguishes Touba from other urban centres is that religious leaders can deploy the religious devotion of Mouride disciples to mobilize vast sums of money and labour for major urban projects. In the absence of state authority and state resources to manage Touba’s development, a social field has emerged in which Mouride associations jockey for prestige and baraka (blessings) from the Khalife-General through their efforts to construct the holy city. As the good works of Mouride dahira are playing an increasingly important role in managing Touba’s development, the ability to marshal the earning power and labour of dahira members has catapulted several large dahira, like Matlaboul Fawzaini and Hizbut Tarqiyaa, into positions of prominence. Touba is also a site where personal ambitions and dreams are realized. Regardless of their country of residence, most Mourides attempt to make the annual pilgrimage to Touba on the occasion of Magal, the celebration of Bamba’s return from a seven-year exile in Gabon. Magal provides an occasion for receiving Bamba’s baraka through the mediation of his successors and for renewing allegiance to one’s marabout (Coulon 1999). In addition to the annual Magal, many Mouride migrants build impressive villas in Touba as expressions of their financial success, which also serves as evidence of the baraka that Bamba bestows upon his hard-working and pious followers. The final aspiration of Mouride taalibe is to be buried in the holy city after death, thereby securing Bamba’s blessings as they move to the afterlife. Touba is a complex signifier for Mourides, and a space that has become meaningful through the execution of both collective and individual projects. Several significant changes in the internal dynamics of the Muridiyya underpin the emergence of the hospital project as the central aspiration of a large collective of overseas Mourides. With the growth of the Mouride diaspora, Touba itself is becoming an increasingly global locale, that is, a locale produced at the intersection of global movements of people, material resources, and ideas (Gille and Riain 2002, Sassen 2007). As migrants spend longer periods of time away from Senegal, Touba has come to occupy a prominent place in the imagination of transnational Mourides. In response, taalibe who share a deep commitment to achieving Bamba’s vision are forging new social 48 connections and transcending traditional scales of human interaction to reproduce dahira abroad and to create the holy city at home. The increasing reliance on migration as a key economic strategy for Mouride disciples has reinforced attachment to Touba as the centre of the global Mouride diaspora, “The logic governing Mouride mobility requires the constant presence of the Touba sanctuary…There is an undeniable concomitance between the construction of the point of reference [Touba] and mobility� (Diouf 2002). The dahira that launched the hospital project illuminates the convergence of these recent trends. As a transnational dahira with over 60,000 members in Europe and North America, Matlaboul Fawzaini offers its adherents an extensive network for insertion into host communities and adaption to life abroad. At the same time, the dahira’s commitment to being an engine of development in Touba reveals its members’ spiritual attachment and nostalgia for the holy city, and their desire to participate in the ongoing construction of Touba even from afar. The elaborate organizational structure of the dahira, and its success in attaining the status of a legal non-profit organization, reflect attempts by Mouride taalibe (and Mouride intellectuals in particular) to modernize the brotherhood’s activities and reputation at home and abroad (Diouf 2000). Although the hospital is a development project of sorts, it is also the culmination of migrant longing for the spiritual homeland, a reflection of their commitment to achieving Bamba’s vision of the holy city, and a display of the dahira’s financial and symbolic capital. BARAKA OR BIOMEDICINE? THE HOSPITAL AS A UNIQUE THERAPEUTIC STRUCTURE Just as the motivations behind the hospital’s construction are multiple and complex, the hospital has taken shape as a distinct place where science, faith, technology, and baraka converge. This convergence occurs in social relations at the hospital as well as in its physical structure - the hospital’s departments are organized around a central courtyard that is equipped both with a mosque and a helicopter landing pad. The hospital’s origin as a taalibe-initiated project shapes how Mouride patients interact with the medical personnel, how they assess the quality and cost of the care they receive and their sense of ownership of the hospital itself. Hospital personnel feel that they have a mandate to heal the sick and to relieve suffering, and to emulate the pious work ethic and solidarity that produced the hospital in the first place. A sense of a divine mission underlies therapeutic encounters and other social interactions at the hospital. When asked what might be different about the hospital, given its origins as a religious project, most employees were quick to stress that it is a modern biomedical facility that functions like any other hospital (Bogol 2006, Dieng et al 2006, Seck 2006). The prestige of being the only Level III hospital outside of Dakar is central to the pride that medical personnel and dahira members take in Hôpital Matlaboul Fawzaini. Yet in the wake of assertions about the hospital’s modernity and technical sophistication, evidence of its particularity emerge. “There is the tendency for hospital staff to transform themselves into taalibe when important marabouts arrive at the hospital�, one doctor acknowledged (Dieng et al 2006). In the presence of high-ranking marabouts, some hospital staff forgo professional conduct and enact taalibe-sheikh relations by prostrating themselves on their hands and knees in front of their religious superiors (who are also patients). 49 Staff narratives offer a clear indication that there are numerous personal and professional motivations behind their work at the hospital. All of the hospital’s top administrators had distinguished careers before arriving at Hôpital Matlaboul Fawzaini. Given Touba’s distance from Dakar, which for Senegal’s intellectual class represents ‘all of Senegal’, these doctors and administrators relinquished far more prestigious, comfortable, and in most cases more lucrative positions to work at the hospital. Some of the attraction came from wanting to be part of something entirely new. “Building a functioning hospital from scratch is not something you often get to do in the course of a normal medical career� explained one doctor (Dieng et al 2006). For the Mouride intellectuals working at the hospital, mystical and religious motivations weighed heavily in their decision to relocate to Touba. In recounting their stories, they convey that their professional work at the hospital is infused with religious significance and animated by the Mouride ethic of xidma. 18 One young doctor described his choice in the following manner: “We came to Matlaboul Fawzaini leaving behind us greater opportunities for material advancement, but as young men who love our country and have a desire to serve it, and as disciples of Sheikh Amadou Bamba, who suffered so much for his people, we thought that our place was here…to bring our contribution to the well-being of our compatriots and fellow disciples� (Dieng et al 2000). Other doctors and medical personnel described their willingness to forgo the comforts of Dakar for the chance to contribute in such a fundamental way to the residents of Touba and to realizing Bamba’s vision. Touba exerts a strong pull on Mourides in the diaspora and throughout Senegal. Just as a sense of being called to fulfil Bamba’s mission features prominently in dahira members’ stories about the hospital project, the Mouride staff reflect on their own attachment to Touba. This mystical connection figured prominently in the story of the Human Resources Director, who had spent most of his career far from both Dakar and Touba. In one interview he stressed that he had known for some time that he was going to end up in Touba, largely because of successive dreams in which he was being called upon by various deceased Mouride sheikhs (Ndiaye 2006). He spoke of the conversations he had with them in these dreams, and of his growing sense that Touba was calling him. “When I got my assignment to come here� he said, "I wasn’t in the least surprised.� Even non-Mouride staff report being influenced by the sense of devotion that Mouride taalibe bring to their work at the hospital and in Touba more broadly. Touba looms large in the Senegalese landscape because of it occupies a space outside the bounds of the state and because of the renowned religious zeal of Mouride taalibe. Touba’s distance from Dakar, its extreme heat, and the local ban on alcohol, tobacco, movies, and music create an image of a puritan and inhospitable place. This reputation extends to the world of donors and NGOs - there is almost no presence of any of the major development agencies 18 Xidma means service to others, and it was used by Amadou Bamba alongside the concept of amal (labor) and kasb (gains) to refer to work. Xidma differs from the other concepts in that in Bamba's view, it signifies work in the name of God for the sake of rewards in the hereafter. 50 that are active in Senegal in the holy city. These agencies (and to some extent state structures) are reluctant to accept the necessary compromises and sharing of power that operating in Touba would require. For all of these reasons, one young doctor spoke of his disappointment after he had been assigned to the Touba Health Centre for his four-month medical residency. He assumed that daily life in Touba would be tedious and difficult, and he hoped to be relocated as soon as possible. The doctor continued, “I told my supervisor I would try to make it through the first month, but then I would have to be transferred elsewhere." Instead, after only a few weeks, he was ‘seduced’ by the atmosphere in Touba, by the Mourides’ organizational capacity, and by their willingness to work hard for the benefit of the population. He concluded, "Like many of my colleagues, coming to this hospital was a challenge that I took with some misgivings. But we are rewarded every day by what we have been able to achieve in only one year on behalf of the population of Touba� (Dieng et al 2006). He added that he can imagine spending a good portion of his career in Touba. In spite of their varied itineraries and different motivations for accepting their posts at the hospital, there was a widespread consensus among the staff that they are all working harder in Touba than they would in most other medical facilities. They are driven to meet the acute needs of the population of Touba, and also by the sense that they are working to fulfil a divine mission that Matlaboul Fawzaini entrusted to them. The origins of the project as an homage to Serigne Touba continues to shape the daily practices of the personnel who are now charged with carrying out its medical and administrative functions. For the dahira members and the hospital staff, the technologically sophisticated hospital is a symbol of a modernizing Touba, but it is also the concrete realization of Bamba’s vision and evidence of his continued blessing of the holy city. DAHIRA AND PUBLIC-PRIVATE PARTNERSHIPS: AN ALTERNATIVE TO STATE-LED DEVELOPMENT? In spite of its rocky beginning and operational constraints, the hospital addresses an acute need for biomedical infrastructure in Touba itself and the surrounding region. There has been little to no expansion of government medical services in Senegal for the past two decades. Hôpital Matlaboul Fawzaini is one of two hospitals in the region of Juurbel with a total population of almost 2 million residents.19 In the first year of operation the out- patients department received 5440 patient visits. While the hospital provides an example of transnational civil society filling the gap between citizen needs and state capacity, the short and long-term viability of this kind of development remains uncertain. The dahira’s responsibility for the hospital ceased in 2002, and the current administrators and staff face ongoing challenges. The most evident set of problems at the hospital are those related to its current capacity in light of Touba’s population growth. Administrators estimate that the structure needs to 19 The Henrich Lubker hospital built by German cooperation in Diourbel in the 1960s was barely functional throughout the 1980s and 1990s. European donors approved funding to rehabilitate the hospital in 2004. 51 double in size (from three to six hectares) to achieve its potential as a hospital with level three ranking. The director of human resources can’t hire additional staff because of the shortage of space (Ndiaye 2006). Doctors express frustration with the space constraints; except in rare cases they need to discharge patients after three or four days because of the bed shortage. Both clinical personnel and administrators blame these problems on the lack of foresight with which the dahira undertook the project. The other major impediment to the hospital’s functioning is related to finances, and the fragile balance of responsibility between the Ministry of Health, the hospital administrators, and the Muridiyya for the hospital’s operating costs. During interviews in 2006, the hospital staff asserted that the hospital falls within the administrative hierarchy of the state. As salaried civil servants and they want to make claims on the Ministry of Health to obtain additional personnel and financial support for the hospital. Yet at that time, the hospital received an annual budget of US $660,000 from the Ministry of Health which covered approximately 50 percent of its costs. The remaining expenses were being covered by the hospital’s own revenues. The hospital is now accumulating debt and must occasionally divert patients to other medical structures to obtain pharmaceuticals (Sheff 2009). In spite of the hospital’s status as a state structure, the hospital administrators are keenly aware of the social and political stakes involved in receiving financial assistance from both the state and the Mourides. They currently have many needs and they are continually sending requests to the Health Ministry for additional support (Saar 2006). The unofficial response is that the hospital should go to the Khalife-General and ask the Muridiyya to make up the difference between the hospital’s overhead and state financing. There are several ways to read this response. It could be understood in the broader context of the muted political tension spurred by Mouride aspiration for autonomy versus the state’s desire to extend its control over the Mouride heartland. In the recent past, this tension was expressed in the form of Mouride resistance to the introduction of state- sponsored French education in Touba and opposition to ending the gratuity of water in the holy city. It also could be an allusion to the weak capacity of the state, or an attempt to test the Sufi order’s seemingly inexhaustible ability to mobilize labour and capital. For their part the hospital administrators understand that returning again and again to the Muridiyya for financial assistance could be perceived as a subtle allegation that the Health Ministry is failing to support the hospital. In the potentially contentious realm of state-tariqa politics, such a critique could easily result in backlash against the hospital itself and a loss of existing state funding. The political ramifications of these funding sources are perceived differently by actors inside and outside of the hospital. During a tour of the hospital the chief of staff spoke openly about the hospital’s financial challenges, the shortcomings of the hospital’s current structure, and the seemingly endless need for more money (Saar 2006). On the tour was a high-ranking marabout who was part of Serigne Saliou’s circle of advisors. This marabout, who had not previously visited the hospital, suggested that Serigne Saliou could easily remedy any of the hospital’s needs. In his words, “If Matlaboul Fawzaini 52 can’t continue financing the hospital, someone else will take over. We can get other taalibe to provide the funding.� Although the chief of staff sees his financial challenges as symptomatic of Senegal’s underfinanced health system, his religious counterpart alluded to the ability of the Mouride leadership to tap into the vast capacity of the brotherhood. His enthusiasm for the hospital contrasts with the apparent government reservations about increasing its funding. As a dahira initiated the project its success can be counted as part of the broader legacy of the Muridiyya. The structure is definitively a Mouride hospital. The social and political capital accrued by the dahira and the Khalife-General from successfully bringing such an ambitious project to fruition will never be extended to the state. In light of the permanent association between the hospital and the taalibe who financed its construction, the state has little to gain by increasing its support of the hospital. Given the historical conflation of the state with the ruling party in Senegal, state legitimacy stems from the ruling party’s monopoly over state resources and their distribution through party networks. In exchange citizens reward party leaders with electoral fidelity. This clientelist system leaves little room for state enthusiasm about civil-society initiated projects, which offer as an implicit critique of the state’s shortcomings. While it would have been difficult for President Wade to deflect the Mouride’s gift of the hospital to the citizens of Senegal, the insufficient funding the hospital has received to date reflects limited means and state ambivalence about supporting a Mouride project. The current financing reflects a compromise in which the state provides just enough financial support to deflect accusations that it has abandoned Touba. The state’s ambiguous relationship with the hospital is even more intriguing since the government continues to show a commitment to enhancing the city’s infrastructure. In the past state investment in Touba was a means of ensuring continued loyalty of the Khalife-General, and therefore of securing his ndigal which guaranteed Mouride support in presidential elections. 20 President Wade has devised his own project for Touba for which he has pledged an investment of 200 billion CFA over ten years to build new roads and to extend the electrical grid and water supply. Wade’s attitude indicates that the state is more interested in garnering credibility through its own initiatives in Touba than in supporting Mouride projects, even when the latter respond to the immediate needs of the population. 21 20 See Beck (2001) for an analysis of how the state has historically used investment in infrastructure to leverage political support. 21 In 2009 the Khalife-General’s spokesman issued a scathing critique of the lack of government support for Touba. He highlighted the paltry financing of Matlaboul Fawzaini hospital that was built ‘entirely by talibe’ and receives the least amount of funding of any Senegalese hospital. (Walfadjri, ‘Gestion des affaires de Touba: Le porte-parole du khalife general des mourides decrie le laxisme du gouvernement’ (20 November 2009). Page available at (20 November 2009). 53 WHAT FUTURE FOR DIASPORA DEVELOPMENT IN SENEGAL AND BEYOND? Within the context of Senegal’s deep Islamic history and vibrant Sufi landscape, the Mourides have long attracted the attention of scholars and researchers for their dynamism and their ability to transcend major shifts in Senegal’s political economy over the past 100 years. Recent scholarship has examined how this Sufi order, and particularly Mouride dahira, have been remarkably adept at harnessing new social and economic conditions. The Mouride diaspora offers an example of ‘globalization from below’ whereby “small players, as opposed to mega-corporations, make use of the opportunities offered by globalization� (Mohan and Zack-Williams 2002). In contrast to early theorizing about globalization which anticipated cultural homogenization, the Mourides have continued to “appropriate the possibilities offered by globalization…to take advantage of it and to be borne by it in every sense of the word� (Diouf 2000). Mouride dahira reflect changing social circumstances internal to the Muridiyya and the ways that the Sufi order has capitalized, literally and figuratively, on transnational migration as an economic strategy and source of symbolic capital (Babou 2002). Matlaboul Fawzaini is not unique in its transnational membership, but it provides a remarkable example of the potential scale and capacity of diaspora associations. After a slow beginning, the dahira established a successful transnational bureaucracy that effectively coordinates chapters in a half-dozen countries with tens of thousands of members. By becoming a formal NGO registered in Senegal, it has numerous advantages over more ad-hoc diaspora associations, including its minimal tax burden and its ability to import materials and equipment duty-free. The vast sums of money raised for the hospital, the successful completion of the construction project, and the ability to sustain the project for more than a decade are unprecedented for a Muslim organization in Senegal. In contrast, secular NGOs in Senegal run by migrants that focus on translocal development have had difficulty mobilizing resources for their projects (Grillo and Riccio 2004). Without the motivation of religious piety or a sense of allegiance to a collective spiritual project, most migrants use their remittances for family welfare and expect the state to provide for basic needs and infrastructure. 22 ‘Development’ as an end in itself does not appear to be a compelling objective for savings or investment. In spite of its successes, over the past several years Matlaboul Fawzaini has suffered from a crisis of leadership. Accusations of corruption levied against several of its members and ensuing struggles to wrest control of the dahira from the current leadership have stalled attempts to pursue equally ambitious projects. This crisis reflects new tensions between the increasingly numerous descendants of Amadou Bamba and a dynamic generation of young Mouride disciples. While the former claim a right to power and leadership based on saintly genealogy, the latter stake their credibility on their ability to help fulfil their sheikh’s ambitions for Touba. 22 See Sheff 2009 for a comparison of Matlaboul Fawzaini and l’Association pour le codéveloppement (ASCODE), a secular NGO with many Mouride members. 54 The leadership conflict also illustrates the growing fault line between ‘arabisants’ (Mouride intellectuals with a Koranic education and little or no literacy in French) and ‘francisants’ (Mourides with degrees from francophone institutions). Daam Ndiaye, the dahira’s first president, stems from the arabisant group, and his authority has been challenged by francisants. While francisants have little interest in the more traditional dahira that help rural migrants adapt to urban settings, Matlaboul Fawzaini’s tax-exempt NGO status offers a platform for lucrative business opportunities and as such has become an arena of contention. 23 Rather than being co-opted by the state or harnessed for state- coordinated development projects, the dahira faces internal threats to its continued operation. In addition, as with all things in Touba, future development efforts must receive the blessings of the Khalife-General. The passing of Serigne Saliou Mbakke in 2008 and the installation of Mouhamadou Lamin Bara Mbakke as Khalife-General ushered in a new era for the Muridiyya. It is too early to assess the extent to which the new Khalife- General will embrace and facilitate diaspora or dahira initiated projects in the holy city. The longevity and future efficacy of Matlaboul Fawzaini are far from certain. As a product of African transnational organizing, Hôpital Matlaboul Fawzaini illustrates the varied motivations behind diaspora projects, the distinctive processes by which these projects come to fruition, the tremendous potential of diaspora philanthropy and the numerous problems that can plague its execution. As scholars have found in studies of hometown associations in Mexico, Cameroun and Tanzania, these associations often have limited skills in project management, and their identification of need does not always coincide with the most immediate development needs of their home community (Mercer et al 2008, Orozco 2003). The difficulties that plagued the Mouride hospital project echo these concerns- the project was poorly conceptualized, the dahira did not have the technical expertise to design a hospital, and over US $1 million was wasted on inadequate construction before outside technical consultants helped redesign the building plans. Although the region lacked a top-tier hospital, the health of Touba’s inhabitants could easily be improved with access to clean drinking water, adequate nutrition and shelter, a robust sanitation system, and comprehensive primary health care. 24 The rise of diaspora-sponsored projects raises a nexus of questions about the development role for states in the neoliberal moment, competition and collaboration between state and non-state actors, and the very future of the development paradigm itself. Some scholars express concern that “such projects disproportionately burden migrants and make them responsible for functions that rightfully belong to states� (Levitt and Nyberg-Sorensen 2004). Okome (2007) argues that regardless of what remittances 23 A similar leadership dispute between arabisants and francisants emerged in the dahira Hizbut Tarqiyya and the Khalife-General ultimately stepped into resolve the conflict (Beck 2009). 24 Touba’s residents rely on potable well-water brought in from outlying villages by donkey cart, and its lack of a stable water system has facilitated several outbreaks of cholera in the past few years. There are not enough primary health clinics to serve the needs of the city’s growing population. 55 can accomplish in African settings, they should be considered only an adjunct to state efforts to combat poverty and marginalization. Those most enthusiastic about diaspora development see the potential for states to regain some of their diminished resources, to leverage the value of remittances to obtain credit in international finance markets, and to partner with dynamic diaspora groups that can conceptualize and execute small-scale projects (Ratha 2007, Ratha and Maimbo 2005). In spite of the global trend for ‘development’ funded by private capital investment, the Senegalese state displays tremendous ambivalence about diaspora-initiated projects. In an interesting twist on state-brotherhood relations, the transfer of the hospital to the Ministry of Health afforded the Mourides an opportunity to secure additional state resources on terms of their choosing. Bestowing the hospital structure on the Senegalese government forced it to accept fiscal and technical responsibility for its functioning. Although in many situations non-state actors are replacing the state’s historical functions, this case suggests that diaspora associations can wrest new commitments from the state by successfully completing infrastructure projects. Yet the Senegalese government has only partially embraced the hospital, primarily because it stands to gain little legitimacy from the success of a Mouride project. The case of Matlaboul Fawzaini highlights the distinct processes and outcomes of diaspora projects. The hospital stemmed from a unique constellation of religious motivations, economic opportunities, and transnational collaboration. The possibility of replicating this development strategy, even by Matlaboul Fawzaini itself, is far from obvious. Even the most ardent supporters of diaspora development acknowledge that remittances are not public money but personal flows, and states may have little ability to direct these resources towards sustainable or equitable development projects (Levitt and Jaworksy 2007, Ratha and Maimbo 2005). States and non-state actors may also have competing motivations, as seen in the reluctance of the Senegalese state to expend resources that won’t reinforce its own legitimacy. Given the limited reach and competence of many African states, the enthusiasm for remittances as a lifeline for the poor and a panacea for a development comes as no surprise. Our analysis reinforces the conclusion that there is no replicable model for international development or public-private partnerships to be derived from the complex and contingent projects pursued by transnational migrants (Grillo and Riccio 2004, Mercer et al 2008). In spite of the Mouride penchant for innovation and mobilizing vast material resources, it is unclear whether a Sufi organization (or any non-state actor) can manage a city that will soon have over a million residents. Touba’s status as a religious enclave creates particular challenges for its development and for state-Mouride collaboration. Although Matlaboul Fawzaini undertook a project designed to benefit Touba and the surrounding region, we should remain mindful that remittances and diaspora philanthropy privilege some families, communities, and regions while others are doubly-marginalized by state neglect and the lack of resources transferred by transnational migrants. As a product of the uneven expansion of global capitalism (which in many cases has overtaken 56 state-led development) the economic investments of diaspora associations are bound to be equally sporadic and unsystematic (Mohan 2006). The extent to which the individual and collective savings, projects, and dreams of Africans living in global diasporas can remedy the ills of a marginalized continent remains to be seen. REFERENCES Babou, C. A. 2002. “Brotherhood Solidarity, Education and Migration: The Role of the Dahiras Among the Mouride Muslim Community of New York.� African Affairs 101: 151-170. _____. 2007. Fighting the Greater Jihad: Amadu Bamba and the Founding of the Muridiyya of Senegal. 1853-1913, Athens, OH: Ohio University Press. Bava, S. 2003. “De la Baraka aux Affaires: Ethos économico-Religieux et Transnationalité chez les Migrants Sénégalais Mourides.� Revue Européenne de Migrations Internationales 19 : 69-84. Beck, L. 2001. “Reining in the Marabouts? Democratization and Local Governance in Senegal.� African Affairs 100: 601-621. Copans, J. 1988. Les Marabouts de L’arachide. Paris: Harmattan. Coulon, C. 1999. “The Grand Magal in Touba.� African Affairs 391:195-210. Dilip, R., and Z. Xu. 2008. Migration and Remittances Factbook 2008. World Bank, Washington, DC. Dilip, R. 2007. “Leveraging Remittances for Development.� Paper presented at the Second Plenary Meeting of the Leading Group on Solidarity Levies to Fund Development, Oslo, Norway. Diouf, M. 2000. “The Senegalese Mouride Trade Diaspora and the Making of a Vernacular Cosmopolitanism.� Public Culture, 12 (3): 679-702. Grillo, R., and B. Riccio. 2004. “Translocal Development: Italy-Senegal.� Population Space and Place 10: 9-111. Gueye, C. 2002. Touba : La Capitale des Mourides, Paris: Karthala. Interview by the authors with Issa Ndiaye, 31 May 2006, Touba, Senegal. Interview by the authors with Mustapha Thioune, Founding Member of Matlaboul Fawzaini, 12 July 2007, New York, United States.. 57 Interview by the authors with S. Seck, Hospital Social Work Director, 30 May 2006,Touba, Senegal.,. Interview by the authors with Elhadj Dieng Boly Diop, Amadou Diouf, Abdoulaye Wone, and Aynina Gueye, 13 May 2006, Hospital physicians, Touba, Senegal.,. Interview by the authors with Falilou Diop, 20 May 2006 Hospital Director, Touba, Senegal.,. Interview by the authors with J. Bogol, Matlaboul Fawzaini physician, 17 May 2006 Touba, Senegal.,. Interview by the authors with M. Saar, Hospital Chief of Staff, 12 May 2006 Touba, Senegal.,. Levitt, P., and B. Nadya Jaworksy. 2007. “Transnational Migration Studies: Past Developments and Future Trends� Annual Review of Sociology 33: 129-156. Levitt, P., and N. Nyberg-Sorensen. 2004. “The Transitional Turn in Migration Studies.� Global Migration Perspectives, Working Paper, Global Commission on International Migration, no 6. Maimbo, S., and D. Ratha, eds. 2002. Remittances: Development Impact and Future Prospects, World Bank, Washington, DC. Mercer, C., B. Page, and M. Evans. 2008. Development and the African Diaspora: Place and the Politics of Home. New York: Zed Books. Merz, B., L. Chen, and P. Geithner, editors. 2007. Diasporas and Development, Cambridge. Harvard University Press. Mohan, G., and A. B. Zack-Williams. 2002. “Globalization from Below: Conceptualizing the Role of the African Diasporas in Africa’s Development.� Review of African Political Economy 92: 211-236. Mohan, G. 2006. “Embedded Cosmopolitanism and the Politics of Obligation: The Ghanaian Diaspora and Development.� Environment and Planning A 38: 867-883. O’Brien, D. C. 1971. The Mourides of Senegal: The Political and Economic Organization of an Islamic Brotherhood in Senegal. Oxford: Clarendon Press. _____. 2003. Symbolic Confrontations: Muslims Imaging the State in Africa. New York: Palgrave Macmillan, 2003. Okome, M. O. 2007. “African Diasporas. In B. Merz, L. Chen and P. Geithner, eds, Diasporas and Development, Cambridge: Harvard University Press. 58 Orozco, M. 2003. “Hometown Associations and their Present and Future partnerships: New Development Opportunities? Report commissioned by USAID, Washington, DC. Sheff, R. 2009. The Problem with Eating Money: Remittances and Development within Senegal’s Mouridiyaa. Unpublished Honor’s thesis, Macalester College, 2009. Robinson, D. 2001. Paths of Accommodation: Muslim Societies and French Colonial Authorities in Senegal and Mauritania 1880-1920. Athens, OH: Ohio University Press. Ross, E. 2002. Sufi City: Urban Design and Archetypes in Touba. Rochester: University of Rochester Press. Sassen, S. 2007. A Sociology of Globalization. New York: W.W. Norton & Company. Scott, J. C. 1998. Seeing Like a State, New Haven. CT: Yale University Press. Sy, C. T. 1969. La confrérie Sénégalaise des Mourides. Paris: Presence Africaine. Villalon, L. 1995. Islamic Society and State Power in Senegal: Disciples and Citizens in Fatick. Cambridge: Cambridge University Press, Cambridge. Zsuzsa, G. and S. O’Riain. 2002. “Global Ethnography.� Annual Review of Sociology 28: 271-295. 59 CHAPTER 4 CONTRACTING BETWEEN FAITH-BASED HEALTH CARE ORGANIZATIONS AND THE PUBLIC SECTOR IN SUB-SAHARAN AFRICA Delphine Boulenger, Francoise Barten and Bart Criel Institute of Tropical Medicine and Radboud University Nijmegen Faith-based health providers have historically not been contractually recognised or established within government or public health systems. Medicus Mundi International (MMI) became engaged in an international debate focused on the repositioning of faith-based health facilities in national health systems in the 1990s. Through this, it was recognised that contracting with faith-based health facilities was a critical step towards the development of effective and equitable health care delivery systems, and the ultimate integration of faith-based facilities in public health systems. This paper presents the main findings of a study commissioned by MMI and conducted by the Institute of Tropical Medicine between 2007 and 2009 on contractual arrangements between faith-based hospitals and public health authorities in four sub-Saharan African countries: Cameroon, Tanzania, Chad and Uganda. INTRODUCTION In 2007, the network organisation Medicus Mundi International (MMI) 25 commissioned the Institute of Tropical Medicine (ITM, based in Antwerp) to conduct a study on the developing trend of contracting between faith-based hospitals and public health authorities in four sub-Saharan African countries. 26 Contracting can be described as "a voluntary alliance of independent or autonomous partners who enter a commitment with reciprocal obligations and duties, in which each partner expects to obtain benefits from the relationship" (WHO 1997). Faith-based health providers have historically not been contractually recognised or established within government or public health systems. MMI became engaged in an international debate focused on the repositioning of faith-based health facilities in national health systems in the 1990s. Through this, it was recognised that contracting with faith-based health facilities was a critical step towards the 25 MMI is an international network composed of a private not-for-profit organizations working in the field of international health. The network’s key strategic approach is to strengthen the health system as a whole. Strengthening the Private Not-For-Profit (PNFP) health sector is an essential aspect in this endeavor. For MMI, the contracting of faith-based health facilities is a means to an end – the end being the development of effective and equitable health care delivery systems via the integration of faith-based facilities in public health systems. 26 This article summarizes findings which are reported more extensively in: Contracting between the public and the faith-based sector in sub-Saharan Africa: an ongoing crisis? The case of Cameroon, Tanzania, Chad and Uganda (Boulenger et al 2009) 60 development of effective and equitable health care delivery systems, and the ultimate integration of faith-based facilities in public health systems. Since the late 1980s, contracting has become central to health sector reforms (Palmer 2000). Contracting is now a tool that is widely used to enhance the performance of health systems in developed and developing countries: not limited to the purchase of services, but used to formalize all kinds of relations. In the health sector, contracting is now seen as a strategy in and of itself, as a core element of a systemic reform, under which governments expand their attention beyond service delivery to additional roles. Carrin et al (1998), note that the contractual approach is a powerful policy tool. For instance, through contracting, private not for profit (PNFP) providers, guided by a public purpose, can be integrated into national health care delivery systems (Giusti et al 1997). Although, more recently, there have also been some controversial contractual experiences with output-based incentive schemes (Meessen et al 2007, Eldridge and Palmer 2009) that serve as a warning. There are thus several types of contractual relations: variously based on the scope and nature of the contract (public or private), and the parties involved. Perrot (2006) proposed a generic classification of contracts in three ways: firstly, delegation of responsibility (set up so that rather than directly managing the health services it owns or undertaking to develop health coverage itself, the state delegates an entity to take over this task); secondly, purchasing of services (a health actor entrusts a partner with providing services in exchange for payment, rather than providing the service itself), and thirdly, cooperation (partners share resources needed to work together towards a common goal while respecting one another’s identity). The contractual arrangements we are studying here, contracts between faith-based district hospitals and governments, fall under the first category of ‘delegation of responsibility’. The types of contracts and the modalities for establishing contractual arrangements may differ considerably. A central element, however, is the degree of enforceability of the contract. Generally speaking, a contract is a binding commitment – ‘enforceable’ in the legal sense. That means that non-fulfilment of the clauses by one of the parties can lead to penalties, and ultimately the parties can invoke the commitments before the courts. The contract may contain provisions for these penalties and for the means of enforcing them. Some contractual arrangements do not follow this rule; in which case they are referred to as a ‘relational contract’ (MacNeil 1978). Relational contracts attach great importance to the relationship between the contracting parties, thus waiving a certain degree of detail in favour of the spirit of the agreement reached. They rely primarily on trust and flexibility. We describe below some of the differing contracting relationships that have developed in SSA. However, there is great need for further research into the impact of contractual relationships on the performance of health systems in low- and middle-income countries. When it comes to contracting with the PNFP sector in Africa, public health authorities see it as a powerful opportunity to improve and sustain national health coverage, especially in rural areas. The expectations from the PNFP sector, on the other hand, are to 61 gain explicit recognition of their contribution to the health sector and to be subsidised accordingly by the public sector. Today, however, little evidence is available on the impact of conventional, input-based contractual arrangements between the two sectors. The paucity of the available evidence left questions as to whether the current contracting experiences between faith-based facilities and public health authorities actually work? And if they work, what makes them work? And if they do not work, what are the reasons for this lack of success? MAIN FINDINGS: A DIFFICULT RELATIONSHIP The study described here sought to fill this gap and begin to answer some of these questions. It was framed as a set of case study evaluations carried out in 2007-2009. The study included both Anglophone (Tanzania and Uganda) and Francophone (Cameroon and Chad) countries because of their distinct historical, medical and cultural contexts. Semi-structured interviews (numbering 100) were carried out at different levels (central, regional and district) with stakeholders from both public and faith-based sector. Additional desk review and interviews was also conducted. Cameroon, Chad and Tanzania provided examples of more ‘conventional’ contracting agreements: where faith- based hospitals have taken on the role of a district hospital or situations where a faith- based organization is entrusted with the management of a health district. However, the Uganda case study was conducted in a slight different fashion, and focused instead on contracts being negotiated between faith-based hospitals and PEPFAR (President’s Emergency Plan for Aids Relief) recipients. The intense negotiation around this relationship provided a particular learning potential which warranted this different focus. Cameroon In Cameroon, the PNFP sector provides 40 percent of the overall national health care. It is mainly constituted by faith-based providers linked to three different organizations: Organisation Catholique de la Santé au Cameroun (OCASC), Conseil des Eglises Protestantes au Cameroun (CEPCA), and Fondation Ad Lucem (FALC). Contracting developed in the early 2000s with isolated pilot cases: such as faith-based hospitals getting a district referral status, recognition of the churches’ role in health care delivery, and a public focus on underserved areas. From 2001, gradual steps were undertaken towards formalization of de facto contracting policies. A major event was the Contrat de Développement et de Désendettement (C2D) project launched in 2003, which brought in financial resources to give real content to the contractual arrangements. A partnership strategy was developed later (2003-2006) and model documents were established from 2007. This study investigated the Tokombéré hospital setting – a Catholic 160 bed hospital situated in a rural area in the extreme-northern province of the country. Part of the OCASC network, the hospital is owned by the Maroua-Mokolo Diocese. Tokombéré hospital has been characterised by strong leadership from expatriate hospital directors bringing in external resources. Tokombéré is reported to have a good reputation, extending beyond its district boundaries. Since the early 1990s, the hospital has played the de facto role of district hospital, formalized by a contract between the diocese and the 62 Ministry of Health (MoH) in 2002. However, the contract’s objectives have remained vague with a poor definition of the respective obligations and responsibilities. There was, for instance, no specification of the mechanisms of allocation of funds to the hospital, no reference to any authority of the hospital on the public health centres, and no reference to the specific faith-based nature of the hospital. The monitoring and evaluation mechanisms were poorly developed, communication between the stakeholders was not well organised, and there was no structure operating as a functional, problem-solving organ. Moreover, there was a low level of collaboration between the health centre network and the hospital, seriously hampering the functioning of the local district system in a more integrated and systemic way. For example, there was a reported failure of the MoH to respect its commitments in terms of subsidies to be paid, allocation of staff, official recognition of the hospital as district hospital despite regular requests from the hospital’s medical director. This case points to a role of the faith-based hospital of partial substitution rather than one of complementarity. The hospital has functioned mainly on its own resources and the formal contract has basically reiterated the pre-existing situation, without major changes in terms of mutual relationships. It is clear that the level of knowledge on the contracting technicalities and on the institutional mechanisms needed to streamline these arrangements was insufficient. The contracts require revision and up-dating, taking into account other similar experiences in the country. Finally, there is also a major concern about sustainability – in particular about what will occur after the C2D project has ended. Tanzania After independence in 1961, Tanzania adopted an approach of free health care provision by public health services, and the Arusha Declaration (1967) began a health sector reform aiming at ensuring social and health services to the marginalized populations in the rural areas. The Tanzanian government is the main provider of health services and owns approximately 64 percent of all health services. The Tanzanian faith-based (or voluntary) sector is the second largest health care provider in Tanzania after the government. The private-for-profit (PFP) sector was banned in 1977, but has since re-emerged after the health reforms in the 1990s which also re-introduced user fees. It is worth noting that approximately 87 percent of all health services in Tanzania are dispensaries, and that hospitals and health centres account for 9 percent and 4 percent respectively. Collaboration between the faith-based sector and the government developed rapidly after independence, under President Nyerere’s mandate. The government’s control has increased over the faith sector, not without tension, although religious freedom has been maintained. In the health care sector this translated in the public recognition of the crucial role played by (rural, isolated) faith-based health facilities in terms of coverage. The collaboration between government and faith-based sectors was formalised in 1972 with the adoption of a decentralized, pyramidal health system: a number of faith-based hospitals then acquired the status of District Designated Hospital (DDH), sealed by a formal contract. This enabled the government to compensate for the shortage of public facilities while avoiding duplication. Contracts also guaranteed public funding of the DDH’s recurrent expenditures. 63 After Nyerere’s death, a Memorandum of Understanding (MOU) was negotiated by the churches. This enabled collaboration to continue, while offering more protection to the faith-based sector against public absorption (or forced nationalizations, as sometimes occurred under Nyerere’s rule) and enabling access to external financing sources. Further steps gradually led to the adoption of a Public Private Partnership (PPP) as an official policy, still referred to in key documents and embodied by several governing organs. Moreover, old DDH contract models were revised in 2005 in accordance with the decentralization policy and a new type of operational contract was created in 2007 for private (Voluntary Agencies) and public facilities, excluding hospitals. The Christian faith-based health sector today is well represented in the public health arena by the Christian Social Services Commission (CSSC) – an umbrella network which enjoys official recognition - and by its five regional coordination bureaus. This field research part of this study focused on Nyakahanga DDH (NDDH): a Lutheran hospital in the remote Kagera region in the north west. NNDH currently provides 200 beds and has been the property of Karagwe diocese since 1912. The hospital officially became a DDH in 1992, and the accompanying contract is of the earlier model and has not been revised to fit the 2005 revisions. Consequently, the diocese’s public counterpart remains the MoH, whereas DDHs created since 2005 have been decentralized to deal with Local Government authorities. The contract also lacks a number of elements that are provided for in the newer model, such as: a proper definition of the terminology and concepts referred to in the contract; the principle agreement of a monitoring and evaluation system to follow-up contracting policies; the replacement of the old Board of Governors (BoG) by the Hospital Committee as a facility’s representative body; or the backing up by a solid legal framework. The current management of NDDH’s contracting relationship with the MoH takes place under the auspices of the BoG, but this body does not function in an optimal way. Monitoring of the contractual relationship is not properly done and supervisions remain erratic. In terms of provision of drugs, the public system faces frequent stock-outs, for which NDDH compensates through basket-funding and its own resources. Any available funds are almost completely absorbed by the provision of care at the expense of capital investments. These and other problems have led to a negative perception of the contractual relationship from the perspective of both the hospital and the diocese. Their trust in the contracting relationship has been deeply undermined, and threatened withdrawal has been considered in order to enforce improvement. The Tanzanian contracting model is impressive because of its long-standing character and its extensive coverage. There is however a need to adapt to the evolving context. There are major problems of information and communication and many important policy documents are simply not available, especially at the decentralised levels of the health system. Moreover, the current contracts established with DDHs are in contradiction with the decentralization of the health system, resulting in impaired management and lack of problem resolution capacity. Decentralization itself remains partial and incomplete, with unclear distribution of responsibilities – leading to dysfunctional communication lines. Tanzanian faith-based facilities face growing difficulties resulting from the decrease in external financial and technical support in a context of increasing demands placed on 64 health services by the HIV epidemic and a shortage of human resources due to migration and lack of retention policies. The limited capacity of current contracting arrangements to adequately compensate for this situation carries the seed for a deterioration of the partnership climate at facility levels. Chad Christian churches and health providers in Chad are relatively young, but their facilities now cater for about 20 percent of the national health coverage; half of them being provided by Catholic hospitals and health centres under the umbrella of the Union Nationale des Associations Diocésaines (UNAD). Faith-based Christian facilities mainly concentrate in the South after they filled a gap left by public authorities as a consequence of civil war. The UNAD hospitals and health centres were integrated in the health map from 1993 onwards, in line with Primary Health Care policies. Catholic facilities were integrated as a result of a request from religious authorities, and legalization of church structures and the signing of first contracts gradually modified the initially informal collaboration. Steps towards partnership formalization were taken as soon as 1999, with contracting being one of Chad’s national health policy strategic orientations. A contracting policy was elaborated from 2001, which considers delegation of public service mission to hospitals as well as delegation of health districts’ management to PNFP organizations. In practice, most existing contracts were signed with faith-based organizations, mainly for full delegation of district management, inclusive of potentially existing public district hospitals. This ambitious interpretation is rarely observed elsewhere. Contracting experiments are set in the context of health sector decentralization which, however incomplete, forms the background of the contracting policy. The Catholic Church’s social sector is also decentralized, with the UNAD coordinating technical bureaus (Bureau d'Etudes, de Liaison des Actions Caritatives et de Développement - BELACDs) which are in turn responsible for coordination at a diocesan level. The BELACDs bear responsibility for management activities in case of delegation of health district administration to the Catholic Church. Framework agreements are established at central level and operational contracts at peripheral level. This field research considered the contractual delegation of Moïssala health district’s management to the BELACD of Sarh. The district is located in Southern Chad, and its capital, Moïssala, is home to the district hospital. The current situation is the result of a process that began in 1992 with the transfer of a Catholic hospital’s equipment and human resources to the moribund district hospital of Moïssala. This project (TRABEMO) was followed by different contracts which gradually delegated the management of the health district and district hospital to the BELACD of Sarh. Financial and technical support of external partners sustained this evolution. Those far-reaching agreements were made possible thanks to the preexisting dialogue between public and faith-based sectors, but also because of the limited range of the public sector in covering the South of Chad. The well-recognised experience of the Catholic Church and the willingness of external partners to support the project further facilitated the contractual delegation of management authority to the BELACD. 65 The generally positive reaction of the public sector can be explained by: the poor state of the health system after civil war; the pre-existence of a dialogue between the two sectors; and the public recognition of the role of the faith-based health providers. Operational contracts aim at ensuring the commitment of the government (towards provision of human resources, infrastructure, tax exemptions and training of PNFP counterparts), and the commitment of PNFPs (to the implementation of the national health policy). Participation of both in decision-making is not formally foreseen, but takes place in practice. In addition, sensitization activities have been conducted and a training workshop was organised in 2004 and was well-attended by all key stakeholders. In total, a comprehensive regulatory and operational framework has been established in Chad, far more so than in the other three countries investigated. A good atmosphere characterised by mutual respect prevails between the two sectors. The open-mindedness of the government actors, the quality of the contracting framework, and the support provided through the contracts are particularly valued by the PNFP sector. However, real weaknesses are still apparent. Uganda The faith-based health sector in Uganda owns about 30 percent of the country’s health facilities, the majority belonging to Catholic and Protestant churches. These networks are represented by the different denominational health platforms or Medical Bureaus. 27 Pressure resulting from the decrease in financial and human resources pushed PNFPs to seek a formalized partnership with the public sector after a long period of informal collaboration. Grand principles of public-PNFP collaboration were set in an MOU established in 1998, but partnership policy documents drafted by the Medical Bureaus in 2003 still await legal approval. Faith-based hospitals nevertheless receive public subsidies, although these are felt to be well below the level required. Medical Bureaus collaborate actively in order to facilitate legal recognition by the public authorities and to promote the development of a meaningful partnership framework. An additional source of concern for the Medical Bureaus is the emerging trend of the PEPFAR funding arrangements to directly contract with faith-based facilities, bypassing not only the Medical Bureaux but also the government authorities. Uganda became a PEPFAR focus country in 2004. With a budget exceeding 280 million $US in 2008, the US initiative is by far the biggest donor for HIV/AIDS-related funding in Uganda. Recent concerns have been raised about the channels by which PEPFAR steers such funds. Monies largely remain out of sight of the public budget, thereby impairing the planning capacity at the MoH level. The problem is further aggravated by poor leadership at MoH level, and the faith-based Medical Bureaus are even less involved. Overall, both the public and the PNFP authorities feel that they are being bypassed and lack the information required to adequately steer or engage in the process. The field research focused on two faith-based hospitals involved in contracting agreements with PEPFAR recipients. Saint-Joseph’s Hospital (SJH) is a facility owned 27 Uganda Protestant Medical Bureau (UPMB), Uganda Catholic Medical Bureau (UCMB), Uganda Muslim Medical Bureau (UMMB) 66 by the Gulu diocese and located in Kitgum, Northern Uganda. Since 2005, contracts have been signed with three different Ugandan recipient organisations of PEPFAR funds for HIV/AIDS engagement. The agreements are felt to be constraining: they are extremely detailed, characterised by precise, indicator-bound objectives and activities, rigid descriptions of respective responsibilities and highly demanding monitoring and evaluation procedures. There are reports from SJH that PEPFAR contracts have led to some distortion in the supply of care and the allocation of human resources in favour of HIV/AIDS-related activities. Overall, the involvement of local public health authorities in these contractual arrangements remains limited. However, there are also noted positives: these contracts are accompanied by regular training, intense monitoring and evaluation activities, and exchange opportunities with other beneficiary facilities; reporting duties contribute to the development of a reflexive attitude amongst the providers; and contracts are respected by the donor. Overall, contracts with PEPFAR are well appreciated by the local faith-based representatives because of their predictability and trustworthiness. The other example is Kabarole Hospital (KH), property of the Anglican church of Uganda - a relatively modest facility located in Fort Portal, Western Uganda. The first contract with PEPFAR was set up in 2005 and included prevention, treatment and care activities of HIV/AIDS. It is significant to note that this is the only source of external support which KH receives and it represents half of the hospital’s annual budget. Many of the observations made above for SJH also apply to KH. Local health authorities remain largely positive, seeing PEPFAR interventions as a welcome complement to the limited resources currently available, and providing a valuable contribution in terms of data generation. Sources of worry include the issue of sustainability of this support, the absence of fall-back strategies, the rigidity of donors, the lack of harmonization with existing procedures and policies, and limited information sharing. KH critically voices the preferential focus on HIV/AIDS activities skewing the overall offer of care and unbalancing staff allocation. What was striking was the difference in perceptions of PEPFAR contracts between central and peripheral level health authorities, as well as between MoH and faith-based sector representatives. While the contracts are relatively well appreciated at the peripheral levels of the health system, there is huge frustration at the central level, where there is a strong imperative for a systems approach to address existing inequities between provinces or districts. This frustration is also explained by the lack of involvement of the MoH and the Kampala-based Medical Bureaus in the design and monitoring of the contracts. The PEPFAR programmes tend to develop as autonomous strategies that run in parallel to existing home-grown programmes and health policies. The problems of weak leadership at MoH level, and the incomplete decentralisation process, further compound the situation. The still unsatisfactory relationship between public and faith-based providers may well lead the latter to favour policies and contracts that secure their immediate survival. For example, when faith-based facilities at a district level increasingly opt for the more predictable agreements with external funders such as PEPFAR. This may well bode ill for the future of faith-based engagement in PPP, and for the capacity of the national health system to ensure health for all. 67 CROSS-CUTTING ISSUES Although these cases evidence a variety of contracting practices, comparative analysis can identify a number of common features, providing an interpretative lens for the assessment of contracting policies between faith-based and public sectors in sub-Saharan Africa. The contribution of the faith-based sector to health care provision in all four of these countries is significant (20-40 percent of total health care coverage), and the relevance of the sector should not be under-estimated. The faith sector is especially present in marginalized rural areas where government services are rare. In all settings there is a focus on further integrating faith-based health providers in the public health system. Contractual arrangements between faith-based facilities and the Ministry of Health thus acquire a particular significance as a means to promote a systemic and integrated approach in developing national health systems. Current contracting experiments between the public and faith-based health sectors face great difficulties. Moreover, awareness of these challenges, particularly among public sector actors, is strikingly low. The sometimes problematic state of the relationship between faith-based and public sectors can be as a result of the following factors, clustered into three main areas. Firstly, there is a lack of preparation and participation. Agreements arrive at the peripheral level of the health system in a top-down manner; rarely incorporate lessons from preceding experiments; and are rarely accompanied by adequate training or coaching. Secondly, the contracting documents have many shortcomings - marked by incompleteness, poor integration in existing frameworks, and further aggravated by the absence of any mechanisms for revision. This leads to a heterogeneous contracting landscape, sometimes in contradiction with existing policies, where various non-harmonized types of agreements co-exist. Thirdly, all these country- cases revealed a strong dichotomy between the central and the peripheral level of the health system, further fragmenting the contracting landscape and pointing to the incomplete and immature character of the decentralization processes. This negatively affects contracting experiences by impairing the follow-up of agreements, the establishment of structural responses to address difficulties, and the capitalization processes of past experiences. The scarcity of financial and human resources is hardly alleviated by the signature of agreements. Governments do not always respect their commitments, or do so only to a limited extent, thereby putting faith-based facilities under great strain as they try to fill financial gaps. Contracts deliver as expected if and when they are effectively backed by sufficient resources, as shown by the examples of PEPFAR in Uganda and the one of Moïssala’s first agreement in Chad. Success of the relationship between faith-based facilities and public health authorities appear to lie more in the quality of the partnership processes at the central level, than in the operational contracts themselves. As far as the more conventional agreements are concerned, the relational character of the agreements leads to proper acknowledgment of pre-existing situations (such as a faith-based facility playing the role of a district hospital) rather than leading to innovative organisational arrangements. At best, the current format of contracting experiments seems to offer an inadequate answer to the severe crisis the PNFP and faith-based health providers are facing. These difficulties seriously affect faith- 68 based providers and remain largely underestimated by the public sector. The contracting agreements read, with some nuances, as a locus for disappointing and imbalanced relationships, benefiting the public sector to some extent while pressurising faith-based providers. This situation reveals a real risk of disintegration of the current partnership dynamic between the public sector and faith-based providers in many sub-Saharan African countries. This is demonstrated by worrying signs, such as was seen in Chad and Tanzania where faith-based providers are threatening to move away from existing agreements. The priority of immediate survival and the search for direct results stimulate the development of bilateral relations with external donors, at the potential expense of further integration in the overall health system. Some churches have begun to question the very notion of ‘partnership’ and whether their involvement in the provision of healthcare can be maintained in the future at all. CONCLUSION The particular case of PEPFAR contracts in Uganda provides an interesting contrast to the more conventional contracting forms between public health authorities and faith- based health providers. The PEPFAR contracts are of course in line with the program’s purpose and disease-focused character. Responses in this study suggest that these contracts tend to lack adaptation to the national policy context, problematically bypass national structures, do not incorporate lessons learned from the past, and lack transparency for outsiders (including the MoH). On the positive side, however, PEPFAR contracts are appreciated by the operational actors because of their strong reliability and because of the quality of their monitoring and evaluation mechanisms. These are important assets which are often lacking in the other types of contracting experiences faith-based providers have established with public health authorities. Our study points to the need to raise awareness among stakeholders on the crisis in the current contracting landscape, and the need to dramatically improve knowledge and expertise in designing, implementing and monitoring contractual arrangements. On the faith-based side, further professionalization of health management is a critical requirement. Adaptation and capacity to deal with increasingly complex health systems requires strong administrative, managerial and technical skills. It may also require a larger delegation of managerial authority at facility and diocesan levels than is currently in place. Historical agreements need to be revisited in order to adapt them to meaningful and sustainable partnership models. Updated national inventories of contracting experiences could contribute to the development of a strong knowledge-base and improved institutional memory. The latter, essential to enable appropriate capitalization of knowledge and knowledge translation, is currently largely absent. Finally, the experiences described briefly here, point to the need for tailored and continuous support to build capacity and address these various situations. Central level policies and models, however complete, do not sufficiently guarantee successful implementation and follow- up. Specific training, technical support and continuous steering at district level are needed so as to ensure that the arrangements in place are well-adapted to local needs. 69 REFERENCES Boulenger D., B. Keugoung, and B. Criel. 2009. Contracting Between the Public and the Faith-based Sector in sub-Saharan Africa: An Ongoing Crisis? The Case of Cameroon, Tanzania, Chad and Uganda. 1 & 2. Report. MMI-ITM. Carrin, G., M. Jancloes, and J. Perrot. 1998. “Towards New Partnerships for Health Development in Developing Countries: The Contractual Approach as a Policy Tool.� Tropical Medicine & International Health, 3(6): 512-14. Eldridge, C., and N. Palmer. 2009. “Performance-based Payment: Some Reflections on the Discourse, Evidence and Unanswered Questions.� Health Policy & Planning, 24: 160-66. Giusti D., B. Criel, and X. De Bethune. 1997. “Viewpoint: Public vs. Private Health Care Delivery: Beyond the Slogans.� Health Policy & Planning 12: 193-8. MacNeil, I. 1978. “Contracts: Adjustments of Long-term Economic Relations Under Classical, Neo-classical and Relational Contract Law.� Northwestern University Law Review, 72: 854-905. Meessen, B., J. P. Kashala, and L. Musango. 2007. “Output-based Payment to Boost Staff Productivity in Public Health Centres: Contracting in Kabutare District, Rwanda.� Bulletin of the World Health Organization, 85(2): 108–115. Palmer, N. 2000. “The Use of Private-sector Contracts for Primary Health Care: Theory, Evidence and Lessons for Low-Income and Middle-Income Countries.� Bulletin of the World Health Organization, 78(6)� 821-29. Perrot, J. 2006. “Different Approaches to Contracting in Health Systems.� Bulletin of the World Health Organization, 84(11): 859-66. Taylor, R. J. 2003. “Contracting for Health Services.� In A. Harding and A. S. Preker, editors, Private Participation in Health Services, Health, Nutrition and Population Series, World Bank, Washington, DC. World Health Organization (WHO). 1997. “The Contractual Approach: New Partnerships for Health in Developing Countries.� Technical document: Geneva. 70 CHAPTER 5 HALF A CENTURY YOUNG: THE CHRISTIAN HEALTH ASSOCIATIONS IN AFRICA Frank Dimmock, Jill Olivier, and Quentin Wodon Presbyterian Church USA, University of Cape Town, and World Bank Christian Health Associations (CHAs) – umbrella networks of faith-inspired health providers – have become a solid presence in the collaborative environment of African health systems. Established through sometimes trial-and-error attempts to draw together disparate faith-based health providers who were disconnected from each other, and also unaligned with national health systems, CHAs have evolved into a particular kind of collaborative effort with a very specific role. CHAs now network faith-inspired health providers and facilities; advocate for a proper recognition of their work; negotiate with governments; build capacity among members; and in some cases now channel and report on substantive funds. In this paper we provide a brief recounting of the history of the CHAs and how they were established, as well as a basic typology of CHAs according to three (highly stylized) conceptual stages of their development. This is followed by a discussion of some of the challenges facing CHAs today, based on self-reports from the CHAs. INTRODUCTION Over the last fifty years, Christian Health Associations (CHAs) 28 – umbrella networks of faith-inspired health providers – have become a solid presence in the collaborative environment of African health systems. Established through sometimes trial-and-error attempts to draw together disparate faith-based health providers (also sometimes called mission-based providers or church health services) who were not only disconnected from each other, but also unaligned with national health systems, CHAs have evolved into a particular kind of collaborative effort – and have established a very specific role for themselves: networking faith-inspired health providers and facilities; advocating for a proper recognition of their work; negotiating with governments; building capacity among members; and in some cases now channeling and reporting on substantive funds. CHAs do not operate in a vacuum, so to understand the challenges they face, one must also understand some of the broader challenges facing national health systems as a whole. 28 This area of interest is a terminological minefield. Our main focus in this article is the Christian Health Associations (CHAs) which are also sometimes called national faith-based health networks (NFBHNs). The members of these CHAs are called many different things in different contexts. Usually clustered as private-not-for-profit (PNFP) providers, they are sometimes called ‘mission-based providers’ (MBPs), ‘church health services’ (CHS), or the non-descriptive ‘faith-based organizations’ (FBOs). Importantly these are facility-based providers of ‘modern’/biomedical health care services. 71 Such challenges include under-resourcing, the difficulty of operating in hardship areas, the Africa-wide human resources for health (HRH) crisis, the impact of HIV/AIDS, challenges relating to corruption, or the lack of management capacity. Despite ambitious restructuring, and health and development goals which have been set to reduce the burden of disease and poverty-related illness, health systems in many African countries are overstretched, understaffed and under-resourced. In addition, many of the countries in which CHAs operate are ‘fragile’ and face severe economic and political challenges. Some are experiencing prolonged conflict (e.g. Liberia, Sudan and Uganda) while others have experienced economic isolation (e.g. Zimbabwe). These crises have jeopardized national public health care and essential social services, and in many cases, the CHAs and their members have evolved in response by filling public service gaps. Several recent studies have shown that CHAs have a unique collaborative place and role in national health systems – particularly in the context of the weak health systems in which they often operate. Better data has been collected on the work of CHAs, including their relationships with governments and public health service providers. Many CHAs have also begun to improve their information and reporting systems, and (as a result) have become more visible in national surveillance systems. In addition to this emerging background information, this article builds on the findings of four main sources (and as such these will not be referenced directly again): (1) A questionnaire-based survey of CHAs in Africa which was circulated to CHA representatives at their 4th Biennial CHAs Assembly in Kampala Uganda in 2009 and then followed-up at the 5th CHA Assembly in Accra Ghana in 2011, to identify challenges and opportunities for maintaining and strengthening their role within national health systems; with responses received from 18 networks in 16 countries of sub-Saharan Africa (Dimmock 2011). 29 (2) A study by ARHAP for the Bill and Melinda Gates Foundation in 2008 which conducted desk review on national faith-based health networks (NFBHNs) in 24 SSA countries, with more detailed primary data collection on CHAs in Uganda, Mali and Zambia (Schmid et al 2008). (3) A study by ARHAP for Tearfund and UNAIDS which also gathered primary data on CHAs in Kenya, Malawi and the Democratic Republic of Congo (Haddad et al 2008). And (4) results from a series of papers prepared at the World Bank on the market share, reach to the poor, cost, and satisfaction vis à vis the services provided by CHA member institutions (see Olivier and Wodon 2012 in this collection). It is important to note that there is no standard list of CHAs. Ultimately, we are focused here on those entities which have self-selected themselves as ‘CHAs’, as can be seen in their presence at CHA meetings, or as members of the newly formed African Christian Health Associations Platform (ACHAP). 30 These are all national-level networks of faith- inspired health providers, although they have critical differences. Many CHAs have formed as a ‘health desk’ of a national Christian Council or denominational body; some have broken away and become nongovernmental organizations (NGOs) in their own 29 Including Benin, Cameroun, Central African Republic, Chad, Ghana, Kenya, Zambia, Ethiopia, Lesotho, Malawi, Sudan, Tanzania, Togo, Uganda, Zimbabwe and Liberia. Countries with CHAs not reporting include DRCongo, Nigeria, Sierra Leone and Senegal. 30 see www.africachap.org 72 right, while others remain an arm of a broader network of organizations. CHAs also have very different membership structures: sometimes the health facilities (e.g. the hospitals) are the members, sometimes the members are the supporting congregation or churches, and some CHAs have included NGOs and community-based organizations (CBOs) as members as well – so long as they are engaged in health service provision. Therefore, while some CHAs are directly engaged in the management of health services, others exist mainly to build capacity among members. This also demonstrates the constant tension in the literature and data, which often blurs the work of the CHAs and the challenges facing them, with that of their members. This paper specifically addresses CHAs, not the ‘faith sector’ more broadly in each country, which typically includes a much broader range of health-engaged faith-inspired institutions (FIIs). We have also not included other non-Christian NFBHNs, the most prominent example being the Uganda Muslim Medical Bureau (UMMB) which was established in 1998 and is said to network 5 hospitals and over a hundred health centers (and works in strong collaboration with the Protestant and Catholic NFBHNs in Uganda). Neither have we included some of the denominational and regional networks who sometimes participate as CHAs, but are not national-level networks of health providers. For example, in South Africa, where mission hospitals were nationalized by the government in the 1970s, networks such as the South African Catholic Bishops Conference and the South African Council of Churches play a CHA-like role, coordinating large numbers of health-engaged faith-inspired institutions and initiatives. Similarly, in Swaziland, where there is no functional CHA, the Swaziland Church Forum on HIV and AIDS plays this role, with several denominations, hospitals and clinics affiliated. Indeed, there are a large number of substantial faith-inspired national and regional networks (denominational, pharmaceutical, interfaith, HIV/AIDS councils and the like) which in some contexts function almost as CHAs do. The paper is structured as follows. In section two we provide a brief recounting of the history of the CHAs and how they were established, together with a discussion how CHAs have had to adapt to changes in funding sources since the 1960s. In section three we present some of the main characteristics of CHAs today, introducing a basic typology of the CHAs according to three highly stylized conceptual stages of their development (emergence, professionalization and integration). We recognize that this is an area of investigation that is notoriously lacking in systematic data. Even the basic estimates of the comparative presence of the CHAs in national health systems provided here should be considered with caution - they represent an attempt to establish some baseline (sometimes based on disparate data), rather than definitive estimates. 31 We also acknowledge that no two CHAs are alike, even within a specific stage of the ‘life cycle’: with characteristics shaped by their history and country context. Some are loose networks (sometimes newly formed or fragile), while others are strongly organized collaborative with direct partnerships with governments or donors - for example, CHAZ is a primary recipient of The Global Fund to Fight AIDS, TB and Malaria (GFATM) in Zambia. Still, despite such differences, CHAs do share certain challenges, and we provide this basic typology to 31 See Olivier and Wodon (2012) in this collection which describes this evidential landscape in more detail. 73 assist in framing the discussion of key challenges faced by CHAs in sections three and four that follow. Some challenges, such as constrained funding or human resources, are logically shared with other (non-religious) health providers, yet may impact CHAs differently. Other challenges tend to be more specific to CHAs and their members, such as concerns about historical funding sources, and fundamental questions as to whether the core intentions of CHAs’ members (such as the desire to provide quality health services to the poor), might be under threat. A BRIEF HISTORY OF CHAS AND THE CHALLENGE OF FUNDING CHAs have different historical trajectories (see Annex 1): some have evolved as a health desk to a Christian Council, others have formed more recently based on the example of other CHAs in Africa. Yet an important influence in the shaping of the CHAs came in the 1960s, with the Christian Medical Commission (CMC) of the World Council of Churches (WCC). One account of that era comes from James McGilvray, the first director of the CMC, and a missionary who began this work by encouraging the churches in the Philippines to form a coordinating body of faith-based health work in 1958. In his seminal 1981 booklet, The Quest for Health and Wholeness, McGilvray describes the ground-shaking events for church health services in this time (1960s-1980s): the changes brought by African independence and changed missionary contexts, and the efforts to re- imagine the role of church health services. The proceedings of a CMC meeting in 1968 describe the crisis facing the Church’s healing ministry as follows: “Today, many of these (church) institutions suffer from multiple problems: steeply rising costs, limited staff, inadequate administrative systems, and obsolescence. There are crippling limitations of resources with which to meet those problems. These institutions often function in isolation, not co-ordinating their activities with each other or with government. Governments meanwhile develop plans for providing universal health care, but neither do they take into account nor benefit from a representative voice from the churches, because there rarely is such a representative voice…the orientation of hospital work toward the service of only those who come to the institution, rather than reaching out to serve all in a surrounding community, has meant that many in need have not been served at all...a re-orientation of Christian medical work is obviously required� (in McGilvray 1981). These concerns turn out to be somewhat ‘prophetic’ or at least still very relevant today. From 1963-1964, McGilvray, with national church bodies, conducted surveys of church- related health services in several countries (McGilvray 1981). These led to a series of key meetings, commonly called the Tübingen meetings (although not all were held in Tübingen, Germany) led by the CMC and designed to shape new thinking on church- based health provision. The surveys generated significant collaborative interest, and also what would become pioneering national estimates of medical facilities contributed by church health providers: “43 percent of the national total in Tanzania, 40 percent in Malawi, 34 percent in Cameroon, 27 percent in Ghana, 26 percent in Taiwan, 20 percent in India, 13 percent in Pakistan and 12 percent in Indonesia�. Importantly however, McGilvray then adds, “however, one should not read too much into the above ratios 74 because, at the time of the surveys, this church-related sector was a very disparate group which, with few exceptions, had no collective existence.� These efforts in the 1960s highlighted the importance of establishing collaborative networks or bodies to address a lack of collective existence, and in particular a lack of representation at the national level. This was particularly important as countries began to move towards independence – as a result of which national health systems were being reassessed and there was a greater need for church health providers to be represented and negotiate together as a group. This led to the formation of several CHAs. For example, in the case of the CMC-led ‘Tübingen’ meeting of April 1967 in Legon, Ghana, it was “…resolved to form an Association of church-related hospitals and clinics which would co-ordinate all church-related medical programs both Catholic and Protestant...This body would represent a united voice in negotiations with the government and would make a concerted effort to employ Ghanaian doctors in its hospitals and give high priority to the training of nationals…It was also decided that…churches should explore new avenues of service in community health as distinct from…the individualistic approach through curative medicine as practiced in hospitals.� From this meeting a coalition, formed as a voluntary professional association, was formed in 1967: the Christian Health Association of Ghana (CHAG). A similar process occurred in Uganda (1955-1957), Cameroon (1957) and Malawi (1966) - see Annex 1 for more detail about these formative events. More CHAs followed in the 1970s: Zambia (1970), Democratic Republic of the Congo (1971), Zimbabwe (1973), Nigeria (1973), Lesotho (1973-1974), Botswana (1974), Sierra Leone (1975), Liberia (1975), and Rwanda (1975). Other CHAs have since been established in Benin (1985), Kenya (1987), Central African Republic (1989), Mali (1992), Tanzania (1992), and Togo (1994). More recently, the Christian Health Association of Sudan (CHAS) has been developing (2004-2008), and there are new partnerships growing in Senegal and Ethiopia. It is important to note that these establishment dates are specific to CHAs (or CHA-like networks), and are not reflective of the historical presence of church health services or other kinds of denominational networks in these countries. Some CHAs were formed when other bodies were renamed or reshaped, and of course, many mission-based providers have been present in these countries since (pre)colonial times. Since the first CHAs were established, church health services have seen the context or landscape around them change dramatically (discussed below). While it is extremely difficult to generalize about all countries in Africa – or the development of all CHAs – broadly speaking it is useful to mention that colonial administration structures had an important impact on church health providers. Distinct patterns can be seen between church health services (and their CHAs) depending on the colonial administration from which they emerged. For example, it has been acknowledged that Anglophone and Francophone countries experienced distinctly different health system management, different attitudes towards missionary health and education activities, and as a result different inherited modern health systems, with Francophone systems being noticeably ‘weaker’ than Anglophone counterparts. Another significant common factor for church health services was that alongside the independence movement in Africa in the 1950s and 75 1960s was a similar movement to indigenize churches (Green et al 2002). In some countries, this resulted in a shifting of the ownership and management of the ‘mission- based’ health facilities from the international bodies to the national denominational churches. While it may be difficult to generalize about financial support for all CHAs – crudely put, church health services and their CHAs have experienced a significant ebb and flow of financial resources, with much of the last half century spent scrambling to tap into and become accustomed to new funding pools and sources. In many African countries, church health providers expanded facilities significantly in the first half of the twentieth century. Based on a 1996-1997 survey of CHAs in 11 countries (9 in Africa and 2 in Asia), Asante et al (1998) found that the peak founding years of the surveyed hospitals were 1930-1967 (with a decline towards 1967). In Ghana, for example, independence from colonial administration resulted in the rapid expansion of church health services – with many new church health facilities built after 1957, including the hybrid ‘agency hospitals’ established in the late 1950s and built by the Government in “…what were then rural areas, and handed…over to religious organizations to run…despite being funded by government, (these) were able to reflect the religious nature of their Churches� (Rasheed 2009). 32 However, at the same time, many church health services also started to see a decrease in external funding flows from originating traditional sources such as those historic relationships with international denominational bodies (see McGilvray 1981, Ewert et al 1990). Importantly, they also saw a decrease of ‘in-kind’ contributions of equipment, drugs, and externally funded technical staff such as long-term medical missionary staff (Green et al 2002, McGilvray 1981). Van Reken (1990) notes that medical mission has leveled off since 1925, and gradually decreased since then. Not only has international mission declined, but there has also been a shift from long-term postings to short-term assignments. CHAs have noted that medical missionaries not only brought skills but also created a strong north-south partnership bringing other resources and some budgetary relief for hospitals (CSSC 2007). The shift to short-term mission has had a severe impact on church health services and CHAs, not only in relation to reduced financial support and partnership (thus threatening the sustainability of church health services), but also in an increased burden on local management. Since independence and with intensification in the 1980s, governments (and international donors) started to implement different plans and strategies for strengthening health care – all of which church health services and CHAs have had to adjust to. For example, health sector reforms such as those led by the International Monetary Fund and World Bank, the new divisions of the health system into ‘sectors’ (e.g. public, private-for-profit, private- not-for-profit), and different strategies to implement universal health care (e.g. making public health services free, or implementing user fees), have all impacted on church 32 This raises an important point. In many countries, there are different kinds of ‘hybrid’ facilities – jointly owned or managed between different partners (between different denominations, or between government and church health services). 76 health services who had traditional ways and means of operating and recovering costs. Since the 1980s, new funding avenues also appeared for church health providers, but it is unclear to what extent CHAs and their member facilities were initially able to tap into them given that the expansion in multilateral and bilateral development assistance was mainly directed at governments. Atingdui (1995) does note that the 1980s and 1990s “…saw significant growth, especially at local levels, of charitable, relief, and development activities carried out by nonprofit organizations affiliated with the Catholic, Presbyterian, Anglican and Methodist Churches� - however, it is unclear whether such growth relates to the church health services, or rather the ‘faith sector’ NGO activities more broadly. As a consequence of changing sources of support, church health providers have increasingly sought government funding in order to survive. This has not been a simple shift, as Green et al (2002) note, many church health providers found it extremely difficult to shift away from a structure “…where the majority of external income comes from those motivated to promote religious activities, to one where there is a greater contribution from secular sources such as bilateral and multilateral donors, international NGOs and national government as well as user charges.� Many individual faith-inspired institutions are still reluctant to align themselves with the government, for example in terms of the priorities that they should adopt in their activities (Green et al 2002, Schmid et al 2008). We do not assess the funding patterns here further, but it is important to note that these shifting funding landscapes and the need for renewed (and more technical) relationships with governments have strongly established the role and function of the CHAs. THE CHAS TODAY: BASIC DATA AND TYPOLOGY CHAs were conceived as national umbrella networks of Christian health providers - mainly tasked to draw together the various Christian health providers so as to improve coordination of services, reduce duplication, and, perhaps most importantly, provide a more consolidated platform from which church health providers can dialogue and collaborate with the government. There are a number of similarities among all or most CHAs, namely: (1) The CHAs are the umbrella body and as such do not usually ‘own’ or manage the health facilities themselves; (2) The member facilities are usually classified as private-not-for-profit (PNFP) providers, although there are outliers; (3) The nature of the member facilities’ operations is more ‘public’ than ‘private’ in that they customarily state a mission to provide quality health services to all – especially the poor in hardship areas; (4) CHAs and their members state a mission of being engaged in health care provision as motivated by their faith and Christian values (e.g. a Christian mandate to serve the poor as a concern of justice and equity) 33; (5) Characteristically, CHAs and their members are simultaneously engaged in a complex arrangement of many different 33 Asante (1998) notes five fundamental principles are commonly cited in Christian healthcare provider’s mission statements – and highly valued by all CHAs: Should be dedicated to the promotion of human dignity and the sacredness of life; Should assist all in need, with a preferential option for the poor and marginalised; Are meant to contribute to the common good; Should exercise responsible stewardship; and should be consistent with the teachings and moral principles of the church. 77 networks, including at the national, civil society and denominational levels; and (6) CHAs and their members are usually engaged in many different kinds of health and development-related activities, not only medical service delivery. The core functions of a typical CHA include: advocacy (for example, for planning and policy making); communication and health information; technical assistance and training; capacity building or institutional strengthening (for example, strategic planning, organizational development, human resource management); resource mobilization and administration; research; monitoring and evaluation (monitoring and evaluation, establishing standards); joint procurement (for example, drug procurement) and equipment maintenance. Typically, each CHA has a secretariat that is responsible for liaising between the church health services and various Government ministries and other partners to address these core functions. Clergy, health professionals and representatives from Ministries of Health (MOH) often participate in the managing boards of the CHAs. In turn, the more established CHAs are usually represented on a number of Government and civil society committees and boards. Each CHA member is usually also part of a complex web of historical and institutional relationships: with local communities, as well as local, national and international institutions. In fact, a key role of CHAs that is not commonly highlighted is their role in managing and negotiating a complex array of relationships and initiatives – especially in relation to representing and simplifying these to outsiders and stakeholders. CHA members also typically have very different communication and decision-making processes which it takes significant (and unacknowledged) skill to coordinate. CHA secretariats (as umbrella networks) do not always have the full authority necessary to ensure that members act appropriately, especially with regards to the submission of plans, budgets, and financial statements – requiring ongoing internal negotiation. For example, in the case of Ghana, Rasheed (2009) notes that each denominational group within CHAG has its own decision-making and communication arrangements: “…the Presbyterians have the best scope to coordinate implementation among their members. Within their group, policy and funding are decided upon centrally and regionally. The Catholics operate using a fully decentralized system, with each diocese in charge of local policy and funding. All in all, coordinating the various entities for decision making is far from simple, and depends on the political and technical acumen of the representatives of each…� Rasheed (2009) concludes that CHAG’s Secretariat requires more capacity to coordinate these complex relationships. While each CHA is unique, we find it useful to characterize the various CHAs according to their level of development, which turns out to be closely related to the level of development of the country in which a CHA operates, at least on average. This is illustrated in table 1, which provides basic data on the CHAs as well as broader country characteristics, as well as in table 2, which provides perhaps what could be considered as the simplest possible conceptual typology of the CHAs according to three stages in their development: emergence, professionalization, and integration (explained below). Note that in tables 1 and 2, the data on the CHAs is based in part on a recent internal survey of CHAs conducted by Dimmock (2011). We included most of the CHAs in the table, apart 78 from those which had missing facilities data. Countries not included that do have a CHA (albeit newly formed) include Angola, Burkina Faso, Ethiopia, Niger and Senegal. Some basic information on all CHAs, including those not listed in tables 1 and 2 is provided in the annex. The CHAs that tend to be least developed – for example the less active CHAs, or the CHAs that do not have a MOU with their MOH – tend to be located in countries that have very low income levels (as measured by GDP per capita in purchasing power terms) and/or have been affected by conflict and weak governance. More generally, on the basis of the data in table 1, the basic typology presented in table 2 distinguishes between CHAs according to three stages in their life cycle: emergence, professionalization, and integration. Emergence indicates that the CHAs are still in the process of being formed, or are at a latent stage of activity if formed. Professionalization suggests a movement towards a stronger role for CHAs in a country, together with more formal relationships with the MOH, as well as an important role in capacity building for member facilities. Integration reflects a stage where faith-inspired facilities tend to be fully integrated in national health systems, so that the role of CHAs can shift from securing funding to exerting broader influence. We are not suggesting that all CHAs need to go through these three stages, indeed, some of the countries listed as being at the integration stage may not have gone through an obvious multi-year CHA professionalization stage, and some of the CHAs listed in the professionalization stage already undertake functions that are more akin to the integration stage. Still, the typology begins to illuminate how CHAs’ priorities may differ under different circumstances. Professionalization Consider first the group of countries with low levels of income characterized as being in a stage of professionalization, in some cases already very advanced. These are the core members of ACHAP, the CHAs that were among the first to be established, and which tend to have an especially high (self-described) share of health provision in their countries, with these figures typically based on a perceived comparative share of hospital beds or facilities. 34 Most CHAs in this group already have an MOU with their respective Ministries of Health, or are in the process of negotiating one. These are also countries were the number of CHA facilities per million inhabitants is the highest; with a high ratio of hospitals to health clinics; and a similarly high ratio of training facilities to the sum of CHA hospitals and health clinics. Although life expectancy in these countries is not higher than in fragile states (in part due to HIV/AIDS), the number of hospital beds per 1,000 inhabitants and spending on health care is higher than in fragile states, and this is also the case at the margin for the number of physicians per 1,000 inhabitants. Because these CHAs are well-established and professional, but at the same time still receive limited funding from the state in many countries, one of their key objectives is to secure better financial (and other) support from MOHs, which is why MOUs are indeed so important. This focus on securing support is represented in table 2 by the arrow emerging from the CHAs towards external stakeholders, but returning to internal stakeholder since 34 This measure tends to overstate the share of all health care accounted by CHAs, see Olivier and Wodon (2012). 79 the bulk of the support that is requested from the state is to help fund the care provided by member facilities. However, some of the best managed and most advanced CHAs in this group also aim to exert a broader influence on their countries’ health policies and practices, for example, as is the case of CHAG in Ghana. Emergence Consider next the group of countries characterized as ‘fragile’, due either to conflict or major problems of governance leading to a ‘failed’ state. Most of the countries in this group have very low levels of income in part due to conflict, although Sudan has been faring better, mainly due to oil (for Zimbabwe, recent data on GDP per capita adjusted for purchasing power parities are not available). Some of the CHAs in this group were established early, but often did not ‘take off’, in part because of conflicts which disrupted the ability to organize and perhaps also reduced the need to negotiate with the state (in several of these countries, the state almost gave up its role in health care provision during conflict periods, which led in some cases to a very prominent role an ‘market share’ for faith-inspired health providers, as is the case in the DRC). Other CHAs, such as Chad and Sudan, were established much more recently. Typically, with the exception of the DRC, the market share of CHAs in health care is lower in fragile states than in the low income group, and a higher share of services are provided through health centers than hospitals, at least in terms of the number of facilities and probably because many of these countries have large rural populations (Zimbabwe being an exception, but that country started from a much higher income base until recently). The countries are also characterized by a lower availability of facilities per million inhabitants (again, with the exception of Zimbabwe) as well as a lower number of beds or physicians per thousand inhabitants. The CHAs in these countries have a more limited number of training facilities available in comparison to other CHA facilities, but this does not mean that they play a smaller role in this area given that the ability of governments to train health care professionals is limited in fragile states. Because in many fragile countries CHAs have been constrained in their development by conflict circumstances, a key priority at this time is basic internal organizing, which is necessary when aiming to secure better support from the state (and donors). This is represented in table 2 by an arrow emerging from the CHAs and going towards their internal stakeholders. Integration The third group, consists of middle income countries with small populations. These CHAs often do not have formal MOUs with the state, and typically the facilities operated by faith networks in these countries are already well integrated (and funded) in national health systems. The CHAs in the three countries in this group were created later than those in the low income group, perhaps because in middle income countries with better developed health systems there was less immediate need for the creation of CHAs in order to negotiate support from the state. The CHAs networks in middle income countries also tend to have a smaller market share of health care (bed) provision, again possibly because of better provision by the state. As a result, these countries also have a smaller number of CHA hospitals and health centers per million inhabitants than is the case in the low income group. The ratio of hospitals to health clinics among CHA facilities is higher in these countries, probably because the countries tend to be more urbanized, but the 80 CHAs do seem to play a key role in the training of health personnel as suggested by the ratio of training to other CHA facilities which is as high as in the low income group. In terms of broader health systems characteristics, not surprisingly these are countries were the number of beds and physicians per thousand inhabitants is highest, with also much higher levels of spending on health per capita. Yet life expectancy is not necessarily higher, in large part due to the burden of HIV/AIDS especially in the cases of Lesotho and Swaziland. It is difficult to highlight the main priority of the CHAs in these countries as data are less available than is the case in low income countries. However, we can surmise that to the extent that the CHA member facilities are already well integrated into national health systems, possibly a priority could (or should) be to exert influence on the countries’ broader health policies and practices, for example in order to help disseminate/share the ‘comparative values’ that tend to characterize faith-inspired health care. This is represented schematically in table 2 by an arrow going from the CHAs towards external stakeholders, and especially government agencies. One should not read too much in this very basic typology – and there are important differences between CHAs within the three groups. Other countries where CHAs are being created or considered are not included in the typology, and this is especially the case for Francophone (and Islamic-majority) countries where the market share of faith- inspired facilities tends to be much lower, and the historical circumstances of health care provision were very different. What the data in table 1 and the typology in table 2 seek to illuminate is that there is not only a lot of diversity between CHAs, but also common characteristics and challenges that are worth considering. The priorities associated to the three groups of countries in table 2 tend to reflect a quasi-natural process through which after organizing internally, and after securing external support for their services, CHAs would then shift to a different agenda related to influencing health policies and practices on the basis of their core values and experiences. This also suggests that there is potential for CHAs in the stages of ‘professionalization’ to assist those in the stages of ‘emergence’, based on lessons learned through experience. Although the challenges at the three stages of the life cycle of CHAs are different, in the next section we raise challenges that have been identified as important to all CHAs (although most clearly identified by the core group in the stage of ‘professionalization’). 81 Table 1: Basic Data on CHAs and Selected Health Indicators by Country, Circa 2010 PPP Self- Number Number of Ratio of Ratio of Number GDP declared of CHA health CHA Number CHA PPP health of per CHA Number CHA care facilities hospitals of CHA training to spending Number Life physicians capita market of CHA health (1)+(2) to health training health care per capita of beds expectancy per Year in US$ share hospitals centers per million centers facilities facilities in US$ per 1,000 at birth 1,000 Population Country MoU est. (2009) (beds) (1) (2) inhabitants (1)/(2) (3) (3)/[(1)+(2)] (2005) inhabitants (years) inhabitants (millions) DRC - 1971 319 50% 89 600 10.4 15% 20 3% 31 0.80 48 0.11 66.0 Liberia - 1975 396 10% 6 67 18.5 9% 3 4% 53 0.70 59 0.01 4.0 CAR - 1989 757 20% 2 62 14.5 3% 19 30% 32 1.20 47 0.08 4.4 Sierra L. - 1975 808 30% - - - - - - 106 0.40 48 0.02 5.7 Togo - 1994 850 20% 3 39 6.3 8% 0 0% 57 0.85 63 0.05 6.6 Chad Yes 2009 1300 - 8 130 12.3 6% 2 1% 94 0.43 49 0.04 11.2 Sudan - 2008 2210 30% 4 - 0.1 - - - 161 0.70 58 0.28 42.3 Zimbabwe - 1973 - 35% 80 46 10.1 174% 15 12% 56 - 45 0.16 12.5 Fragile 1987 949 28% 10.3 8% 8% 73.8 0.72 52.2 0.09 Malawi Yes 1966 794 37% 27 142 11.1 19% 10 6% 50 1.10 54 0.02 15.3 Rwanda - 1975 1136 40% - - - - - - 102 1.60 51 0.02 10.0 Mali - 1992 1185 2% - - - - - - 66 0.57 49 0.05 13.0 Uganda Yes 1957 1217 12% 42 491 16.3 9% 19 4% 115 0.39 53 0.12 32.7 Tanzania Yes 1992 1362 42% 89 815 20.7 11% 24 3% 68 1.10 56 0.01 43.7 Zambia Yes 1970 1430 40% 36 110 11.3 33% 9 6% 68 1.90 46 0.06 12.9 Lesotho Yes 1973 1468 40% 8 72 38.7 11% 4 5% 133 1.33 45 0.05 2.1 Benin Yes 2002 1508 40% 6 20 2.9 30% 28 108% 65 0.50 62 0.06 8.9 Ghana Yes 1967 1552 42% 58 104 6.8 56% 10 6% 104 0.93 57 0.09 23.8 Kenya Yes 1963 1573 - 74 808 22.2 9% 24 3% 68 1.40 55 0.14 39.8 Nigeria - 1973 2203 40% 147 2747 18.7 5% 28 1% 136 0.53 48 0.40 154.7 Cameroun Yes 1957 2205 40% 30 150 9.2 15.4% 3 1% 122 1.50 51 0.19 19.5 Low inc. 1972 1495 37% 15.8 20.3% 14% 93.7 1.11 52.6 0.10 Swaziland - 1998 4998 - 3 27 25.3 11% 1 3% 312 2.10 46 0.16 1.2 Namibia - 1978 6410 - 6 - 2.8 - - - 384 2.67 62 0.37 2.2 Botswana - 1974 13384 18% 2 6 4.1 33% 2 25% 1341 1.81 55 0.34 1.9 Middle inc. 1983 8264 18% 10.7 22% 14% 678.9 2.19 54.3 0.29 Source: Compiled by the authors based on data from CHAs and World Bank Development Indicators database. Notes: Countries are ranked by PPP GDP in US$ in 2009. Sudan and Zimbabwe are included in group 1 despite higher GDP levels because of the conflict in Sudan and the situation in Zimbabwe. Mali is included in Group 2, but not included in average statistics for that group. Blank cells indicate that data are not available. 82 Table 2: Typology of CHAs according to their state of development Stage 1: Stage 2: Stage 3: Emergence Professionalization Integration List of Fragile: Low income: Middle income: countries DRC; Liberia; CAR; Sierra Malawi; Rwanda; Uganda; Swaziland; Namibia; (ranked by Leone; Togo; Rwanda; Chad; Tanzania; Zambia; Botswana increasing Sudan; Zimbabwe Lesotho; Benin; Ghana; level of GDP Kenya; Nigeria; Cameroon per capita) (plus Mali) Country PPP GDP pc: $949 PPP GDP pc: $1,479 PPP GDP pc: $8,264 average Life expectancy: 52.2 Life expectancy: 52.6 Life expectancy: 54.3 characteristics Bed rate: 0.7 per 1,000 Bed rate: 1.1 per 1,000 Bed rate: 2.2 per 1,000 Physician rate: 0.09 Physician rate: 0.10 Physician rate: 0.29 PPP Health sp. pc: $74 PPP Health sp. pc: $94 PPP Health sp. pc: $679 CHA average Year established: 1987 Year established: 1972 Year established: 1983 characteristics Share with MOU/Eq.: 1in8 Share with MOU/Eq.: 9 in 12 Share with MOU/Eq.: 0 in 3 National (bed) share: 28% National (bed) share: 37% National (bed) share: 18% (Ho+HCs)/million: 10.3 (Ho+HCs)/million: 15.8 (Ho+HCs)/million: 10.7 Ho/HCs: 8% (excl. Zimb.) Ho/HCs: 20.3% Ho/HCs: 22% Training/(Ho+HCs): 8% Training/(Ho+HCs): 14% Training/(Ho+HCs): 14% CHA Priority: Internal Organizing Securing Support Exerting influence stakeholders (also valid for advanced (member CHAs from low income facilities) countries) CHAs External stakeholders (government, donors, etc.) Source: Compiled by the authors based on data from CHAs and World Bank Development Indicators database. CHALLENGES FACED BY CHAS AND FAITH-INSPIRED MEMBER FACILITIES Having described in broad terms the history of the establishment of the CHAs and their current characteristics, we now shift to some of the challenges they face. Although the CHAs in Africa clearly operate in very different funding and health contexts, representatives of CHAs do point to several shared challenges. This is evident in the broader emerging literature, as well as from a small survey of CHAs carried out by Dimmock (2011) from which the quotations in this section are taken. Increased demand equals increased strain on health providers Church health providers commonly state that they have recently experienced greatly increased demand (for health services) which in turn has put added pressure on them and the health system. This is both in relation to those countries with stronger systems (such as Ghana, where the implementation of a national health insurance has had positive 83 impact, but has also placed greater strain on the providers), and those in more fragile contexts, such as Zimbabwe. As ZACH/Zimbabwe notes, as a result of the instability and isolation the country has experienced, “…partners lost confidence and moved to neighboring countries with vibrant economics…The demand for health increased due to poverty and increased disease burden. The demand for health care meant that hospitals needed to increase their capacities to provide services, however due to poor economic performance critical shortages…forced hospitals to scale down and provide basic care.� Human resources for health crisis All church health providers share the challenges of a continental-wide human resource crisis, especially in relation to human resources for health (HRH). All of the CHAs indicate difficulties in competing with governments and international NGOs for staff. Several say that a competitive salary is the best way to retain staff, but that this is also one of the greatest challenges. At the same time CHAs have successfully implemented (sometimes innovative) incentive strategies such as continuing education for staff, and motivation such as giving credit where credit is due, and good working conditions. As stated by UPMB/Uganda: “Unfortunately we are not in a good position to compete favorably with government, INGOs and NGOs. Salaries in these sub sectors are much higher than in our network…(however) professional staff get more job satisfaction working with us because facilities and drugs are available.� A particular problem for the CHAs and their members is the loss of long-term medical mission staff. However, CHAs have become heavily engaged in the human resource crisis, through negotiation with government (see below), and also through engagement in a CHA-HRH technical working Group. Reduced funding from traditional sources Even today, CHAs continue to feel the effects of a reduction in traditional sources of support and funds from affiliated religious groups in the West. For example, CHADCath notes, “Yes, the funding is getting more and more difficult…The general opinion is that there are other countries in Africa suffering a lot (more)…�, and CHADProt similarly says, “…the funding that we get from our Christians partners from Europe (decreased) during the last year. They say that people in Europe don’t give more money like the last past years and ask us to focus on the local opportunities of fundraising.� There are many other such examples of how the loss of traditional funding sources continues to hurt the church health providers and their CHAs. For example, Boateng (2006) notes that CHAG facilities find their financial sustainability seriously threatened due to increased demands for services against declining donations from traditional sources, sometimes uncertain support from government, and low cost recovery in member facilities. Targeted funding not allowing for long-term or core activities All CHAs and their members now find themselves heavily dependent on local and international donor support – and increasingly dependent on conditional grants and targeted project funding. (Ironically, this trend has further weakened their relationship with traditional church partners). In Uganda, for example, the UCMB reported that 49 percent of their funding during 2007/8 was comprised of project funds: “It is true that donor (project funds) are increasing…(but) 80 percent were for HIV/AIDS only. So I 84 cannot really say the main work of the hospitals is depending on donor funds.� CHAs note the difficulties and detrimental effects that targeted funds (especially HIV/AIDS funds) have on broader health provision. For example, CHAM/Malawi note: “Yes, funding for projects has increased, but funding for core programming has decreased.� CHAZ/Zambia reported that project-funded donors are often more interested in short- term technical inputs than in long-term investment in developing local human resource capacity. This places additional stress on under-staffed health programs and encourages competition within the health sector. UCMB/Uganda, CHAK/Kenya, CHALe/Lesotho and ASSOMESCA/CAR all expressed caution with regard to the need to balance attention given to administering specific project funding with the priority tasks of providing integrated and essential health services. The demands of reporting and accountability for donor funding have also increased – stretching the capacity of church health providers and increasing the role of CHAs in capacity building. Government support, responsibility and cost recovery A key role of CHAs has become the negotiation of appropriate and sustained support from government. Several CHAs are now heavily dependent on government subvention for covering payroll and operating expenses within their facilities (for example, CHALe/Lesotho, CHAZ/Zambia, CHAM/Malawi, CHAS/Sudan). The CHAs have played an important role in the negotiation between church health services and the governments – especially in relation to proving the significance of the church health services to the governments. Dimmock (2011) surveyed CHAs about the likelihood of their handing services over to government, and what this would mean to denominational bodies or churches. Most of the responses were strongly against this notion – also noting that most governments did not have the capacity to manage CHA facilities in addition to their own. They also cited the trust local communities had in church health services, and that transfer of the facilities to governments would mean a loss of credibility in the broader healing ministry of the church. For example, CAM/Cameroon: “In Cameroon, the churches were the first in the area of health. People trust us a lot. It will be very difficult to accept (a very big failure).� However, some CHAs noted that it was increasingly difficult to maintain independent health services in the current financial climate – and that it was mainly their poor experiences of handing over services and their fear that whoever took over would do a worse job, with poorer quality or not serve the poor as well, which kept them engaged. Says UCMB/Uganda: “This is not an option, at least for the foreseeable future. We believe we have a duty as Christians to fulfill the mission of Christ…People in situations of instability would be the greatest losers as Churches have provided resilience to health care for them when everybody else left or could not.� Erosion of Christian values Most importantly, a constant challenge, relating to all of the above is that CHAs and their members feel that it is increasingly difficult to maintain their Christian mission and values in the face of new constraints and integration with public and private services. This is felt broadly, in terms of searching for financial sustainability to continue to be oriented towards the poor (private urban hospitals are certainly more profitable). For example, UCMB/Uganda notes “…the poor rural hospitals are more dependent on the 85 conditional grants from government…Drops in (primary health care conditional grants) are increasingly forcing facilities to try to increase user fees. In turn this affects the principles of our mission, universality and preferential option for the poor.� This tension is also felt in relation to new partnerships and conditional support of governments and donors – which often have a different ‘vision’ or operational culture. CHAK/Kenya notes: “We must resist the temptation of getting donor funding from sources that would compromise our faith and values. We have to be firm with government on the minimum acceptable standards for our values.� This challenge is also felt in relation to operational decisions, such as the kind of staff that gets hired. CHAZ/Zambia expressed a perceived erosion of Christian values in the services of their members. This was related to the shortage of professional staff and relaxation of recruitment criteria reflecting religious values. UPMB/Uganda noted that “The biggest threat to values lies in the secondment of staff to our health facilities. These staff are often recruited and then deployed by government with no consideration whatsoever for the values and work ethics of the receiving faith-based institution…their social values and work ethic are sometimes in conflict with the organizational culture of the institution to which they have been deployed.� Clearly, the above list of challenges suggests that CHAs and their members are operating in complex and apparently rapidly changing circumstances. However, it is rather disconcerting to note that many of the above challenges were already raised in the 1960s, and have still not been resolved. For example, McGilvray’s (1981) account raised most of the challenges in relation to the nature of church health services, their role in facilities- based versus PHC/preventative care, what it means to be a ‘Christian’ provider, whether it is possible to bear the costs of a ‘pro-poor’ mission, whether church health providers are sustainable given new financial constraints, and queries about evidence of their ‘value-added’ in modern health systems. What has changed, however, is the strengthening presence of the CHAs in this negotiation, especially in the group of middle income countries outlined earlier. In these countries, CHAs have become active in negotiating these challenges to partners and in working to mitigate these effects – both directly and indirectly. ROLE OF CHAS IN NEGOTIATING, CHANNELING AND RAISING SUPPORT In this last section, we consider the relationships between CHAs and their external stakeholders, and especially the negotiations taking place between CHAs and MOHs (in particular, focusing on the CHAs at the stage of professionalization characteristic of the low income countries group - where securing support from the state is paramount). The issue of financial and other forms of support (such as capacity building) is important not only for the CHA member facilities, but also for the CHA secretariats themselves – and is often a distinctly different fund-raising endeavor. When CHAs were asked how they ensure their own future financial viability (CHAs specifically, not their members), they indicate the following strategies: developing business plans, reducing staff, cutting expenses, outsourcing some services, negotiating with government for additional support – and interestingly, many CHAs are now engaging in direct income generation to support 86 the CHA (secretariat) activities. For example UPMB/Uganda and CHAK/Kenya have both investment properties that provide some revenue, and others are involved in guesthouses, office rental, drug supply and distribution, and corporate health service contracts. 35 The increased engagement with government and donors (and attendant donor requirements) has cemented the role of the CHAs, in terms of their unique position in negotiation with government and partners, and building technical capacity in their members. It is important to note that this role for the CHAs has been developed gradually, with much trial and error, and continues to evolve. Internally, negotiating the role of the CHA has caused some tensions, especially with regards to whether the CHA holds funds and is involved in the management of health facilities or whether it mainly builds capacity and channels resources to members. This was experienced in Kenya, where the role of the CHA has been renegotiated several times. CHAK/Kenya: “From 1946 to the early 80s CHAK received and channeled grants from MOH/Government Budget to Protestant Church health facilities. In the mid-80s to mid-90s CHAK enjoyed huge funding from international donor partners including bilateral partners and experienced uncontrolled growth in response to donor funding. The pressures of fulfilling donor demands gradually shifted its role to implementation to the extent that it at times worked in competition with member health facilities. In 1996, CHAK underwent a major paradigm shift from an implementer to a facilitator. Priority was refocused on advocacy, capacity building, networking, communication and facilitation.� In Zambia, CHAZ has evolved from being an umbrella network, to now being a primary recipient of the Global Fund, even receiving the bulk of Global Fund support for Zambia in 2006 (58 percent committed, and 56 percent received), and in turn dispersed money to “411 local FBOs to fight AIDS, 73 local FBOs to fight TB and 75 local FBOs to fight malaria� (GFATM 2008). At the start of the Global Fund grants in 2003, CHAZ had 23 employees and by the end of 2008 had 82 employees. 36 The role of negotiator of partnership between the national government and respective church health providers is perhaps the most significant role that has evolved for CHAs. There is some evidence that several of the CHAs in the professionalization (low income group) have managed to establish strong collaborative relationships with government, evidenced by successful and ongoing negotiation around issues such as financial support and human resources (see examples below). Several studies indicate that in the case of HIV/AIDS multisectoral collaboration, some CHAs hold a stronger collaborative relationship with government than other FBOs and NGOs thanks to a dual pathway to government, through the Ministry of Health (in relation to their medical response, e.g. 35 CHAs have also become significantly involved in negotiation and operation of pharmaceutical provision. (e.g. CHAZ/Zambia, CSSC/Tanzania, BUFMAR/Rwanda, ECC/DRC, CHAM/Malawi). Other medical supply organisations include: Mission for Essential Drugs and Supply (MEDS) – jointly owned by CHAK/Kenya and KEC/Kenya; the Joint Medical Store (JMS) – a joint venture of UCMB/Uganda and UPMB/Uganda; ASSOMESCA/CAR which operates a regional drug distribution agent system in the Central African Republic, with customers (church member groups) in CAR, DRC and the Congo; and CHANPHarm operated by CHAN in Nigeria (Schmid et al 2008). 36 See www.theglobalfund.org 87 ART), and also through national AIDS councils as a civil society representative. “Although the CHAs and their members face various critical challenges such as financing and workforce concerns…these associations are exemplars of the positive impact of collaboration, networking and resource sharing…in countries that do have such national faith-based health networks, there is stronger collaboration between FBOs, as well as between FBOs and secular groups - in particular a stronger advocacy role with government. The NFBHN appears to be a valuable type of FBO that draws together different faith-health activities, and provides support in a variety of ways� (Haddad et al 2008). The evolution of relations between CHAs and governments is demonstrated by the negotiation surrounding the establishment of MOUs between CHAs and their national governments (usually the Ministries of Health). We specifically mean those MOUs which frame and lay down the terms of a specific relationship between the CHAs and that government (not standard legal policies outlining the roles of NGOs or private sector providers more generally). As outlined in Table 3 below, these MOUs characteristically formalize the relationship between the CHA and the government; acknowledge the important work of the CHA members; and often formalize some level of reciprocal partnership and support (such as waivers on import taxes for medicines and supplies, budgetary or human resource support, and access to training opportunities). These MOUs are indicative of the collaborative role that CHAs are increasingly playing – as chief representative and negotiator of groups of church health providers. There are also several examples of CHAs successfully negotiating their way through specific crises. For example, several networks (e.g. CHAM/Malawi and CHAG/Ghana) have negotiated service contracts at the district or local level, or have agreed on ‘designated District Hospitals’ (e.g. CSSC/Tanzania, UCMB/Uganda and ZACH/Zimbabwe) through which church hospitals, subsidized by government, act as public hospitals. Several CHAs continue to negotiate on the issue of human resources – for example, curtailing the inappropriate secondment or ‘luring away’ of staff, or negotiating access to government training opportunities for church health staff. These relationships however require continuous negotiation. That is, the presence of an MOU does not automatically result in stronger practice – or adherence to the terms of the MOU. To improve adherence to the MOU, some countries have also established joint committees, including government representation on managing Boards (for example, CHAM/Malawi and CHALe/Lesotho) or have an official in the Ministry of Health assigned to liaise with the private sector (CHALe/Lesotho). Most CHAs have reported that the process of establishing the MOU was important in identifying and aligning the relationship between the Ministries of Health and CHAs. In Ghana, this relationship is held up as one of the most positive examples of public-private partnership in the country. Indeed, the rest of the private sector is described as feeling left out of the mainstream government thinking and planning (Rasheed 2009, Makinen et al 2011). Rasheed (2009) describes the gradual strengthening of the CHAG-MOH relationship as follows: “CHAG worked with government, gaining trust and proving its usefulness. In particular, CHAG participated in restructuring exercises within the health 88 sector, represented FBOs within the SWAP, and helped with mapping services including describing the extent of the human resource crises. CHAG worked with government to work out strategies to ensure health service coverage for rural areas and conceptualize national health insurance policies. CHAG also substantively increased collaborations with government in training. It was during the administration of 2003-2006/7 when things finally came together.� The relationship between CHAG and the MOH was strengthened slowly over time, and the MOU is reflective of the fact that CHAG has sought and secured a seat at the Ghanaian health policy table (Rasheed 2009). Even though described as comparatively strong (especially in comparison to Christian Health Associations in other African countries), the relationship between CHAG and the MOH should not be taken for granted, or assumed to be without its own tensions and obstacles. For example, Dovlo et al (2005) describes the context in which the national health insurance scheme (NHIS) was initiated by the MOH and notes that at that time CHAG tried to initiate a dialogue on health insurance but failed to get an audience before the NHIS law was passed. However, there appear to be more positive examples than negative, and the fact that such tensions are openly expressed suggests that the collaborative relationship between CHAG and the MOH includes a healthy degree of debate that goes beyond surface-level ‘dialogue’. This is just one country example, but generally CHAs have to continue to work to maintain good partnerships, proving the ‘value’ of their members to the broader system. Also, despite this stronger role, CHAs (as a function) still face significant internal challenges. For example, reporting on a CHA meeting in 2006, Mandi lays out the main challenges as being: lack of co-ordination among CHAs when lobbying since they usually approach governments independently and not as a united front; CHAs do not have adequate lobbying or negotiating powers; there is a lack of trust between governments and CHAs; CHAs fear that if they partner with governments, they will be absorbed and lose their identities; and that governments view church health providers as direct competitors rather than partners (Mandi 2006). 89 Table 3: MOUs between CHAs and governments in Africa Country- CHA MOU Status Benin - AMCES no Have agreement with government as an “NGO of public utility� since 2008. Cameroon - yes MOU signed 2007: monitored through a committee with representatives of Protestants, Catholics, others and the MOH. Further FEMEC negotiations underway (draft proposal for state assistance). CAR - no A convention signed 1995: used to get drugs, equipment, and vehicles from abroad without paying taxes, but no longer in effect. ASSOMESCA Chad - Catholic yes MOU signed 1994: to cooperate with the NHS, including the importation of drugs without paying any import tax (formally renewed annually) Chad - Protestant yes MOU signed 2006: for the importation of medicines and supplies without taxes. 2008 further contract signed for support of hospital running costs. Ethiopia - CHAE no New CHA Ghana - CHAG yes MOU signed 2003, with administrative addendum in 2006 - – indicates shared responsibilities and some government financial assistance Kenya - CHAK no MOU in development - awaiting government review and finalization. Lesotho - CHALe yes Service agreement in 2002, MOU signed 2007: Monitored through agreed structures in different committees (e.g. HR, quality, legal) – but no predetermined frameworks with indicators. Liberia - CHALi no Not in development. Malawi - CHAM yes MOU revised in 2002: indicates government’s responsibility to provide health services to the nation, and CHAM’s role to complement the Government’s efforts. Towards this end, the government undertakes to provide financial assistance to CHAM units, monitored through CHAM-MOH quarterly review meetings. Sudan - CHAS no Government said MOU not necessary – advised CHAS to get legal status in order apply for government funding and service agreements. (Status obtained 2008) Tanzania - CSSC yes Two MOUs signed in xxx: Service agreement on the provision of health services, database sharing, joint supervision and joint review. Togo - no MOU has been designed, but not signed by government. APROMESTO Uganda - UCMB no MOU in being developed – after legal framework for PPPH policy approved by Cabinet. Uganda - UPMB no But collaboration with government on health service provision and capacity support is advanced. Zambia - CHAZ yes MOU revised in 2004: government provides 75% of the CHA facilities running costs, seconds almost all health professionals and supplies essential medicines. CHAZ in the sector advisory group and in various other committees (e.g. NAC and CCM of the Global Fund) as well as in health-related statutory bodies and overall planning and budgetary processes. Zimbabwe - no MOU drafted and waiting on Management Board Review and inputs from MOHCW. ZACH Source: Authors’ compilation, based in part on Dimmock (2011), Schmid et al (2008), and Haddad et al (2008) 90 THE NEXT HALF CENTURY FOR CHAS – WHICH WAY FORWARD? Maintaining and strengthening health systems will necessitate effective and efficient partnerships within the health sector, and beyond. It will call for increased intersectoral cooperation (rather than competition), regarding the scarce human and financial resources. It will also require the identification, resourcing and rational use of religious health assets, both tangible and intangible, to positively impact health outcomes. Now, perhaps more than ever before, it is vital to align public health policies with primary health care (PHC) principles and objectives to promote equity, access and fair health for all. Despite the numerous challenges currently facing African CHAs and their members, these associations will likely remain the primary partner to government health services for a long time to come. We have outlined several challenges facing CHAs and their members – some challenges that are contextual, and some which are specific to what church health providers are trying to be. In all this, the developing role of the CHAs as a particular collaborative function has been demonstrated as key. This function is very much in line with what was imagined in the 1960s. However, what does this mean for the future – and for what CHAs might need to become in the future? We have named several important functions, such as continued and improved capacity building, negotiation, and skills building. The literature and the CHAs themselves outline two other functions that may become even more important in the future. Wrestling with the weightier questions It is clear that CHAs are seen to be centrally involved in wrestling with the weightier questions about the role, function and sustainability of church health services. Said differently, CHAs have become the forum for consideration of what it means to be a faith-inspired health provider in modern health systems. The CHA annual meetings have replaced the function of those other mechanisms such as the CMC for engagement with these concerns. CHAs have been tasked by their members, not only with operational issues such as funding or capacity-building, but also with some of the more difficult questions about the role of the church in health, the potential of (re)orientation towards PHC, or weighting community health and health promotion versus hospital-based care. Technical assistance, information gathering and sharing It has been widely noted that many church health providers lack the information systems and the necessary capacity to document their work. We cannot judge whether this capacity is less or more than other public or private providers. Suffice it to say that this is a significant need – and one that CHAs are seen to play an important role supporting their members. This is not only relating to M&E, but also to the ability to evidence the work they do, and also utilize information systems to improve their own work. There are also clearly many lessons that need to be shared between CHAs – for example, on how they deal with these challenges named above. So far, this sharing (between CHAs) has mainly occurred at annual meetings, and supported by outside partners. 91 The CHAs tend to have a few key partners with whom they work in relation to technical assistance (for example the development of information systems). It was concerning that when the African CHAs were asked who knew of technical assistance (TA) available to CHAs and who accessed such – specifically from country partner offices (e.g. EU, World Bank, UN agencies) of the 14 CHAs that responded (Dimmock 2011), only 8 knew of TA available to CHAs. For example, CHAM/Malawi noted TA was available in Malawi, but not to CHAs; CHAS/Sudan noted TA was only available to the government; and CHAZ/Zambia noted that TA was seen to be available, but that “the Association has not yet accessed this support�. Only two CHAs had accessed TA in the past: UCMB/Uganda noted that they usually did not access TA, but had done so in the past, for example “…from a European partner in developing proposal for EU funding in the past�; and ZACH/Zimbabwe was the only CHA who noted they were currently gaining TA by “working with UN agencies, World Bank, GFATM, European Union�. During the CHAs Assembly in Tanzania in January 2007, a significant step was taken, when it was agreed that a continent-wide platform for African national associations should be established: the African Christian Health Associations Platform (ACHAP). Its objectives are to enhance advocacy, facilitate technical assistance and support, networking and communication of ideas for coordination and capacity building (the hub of which is mobile, and currently hosted in Kenya by CHAK). Fundamentally, what this platform represents is an acknowledged need to share information, experiences and best practices among CHAs. This need is supported by the available literature, which reiterates the importance of ‘evidence-based’ advocacy if CHAs are to take the role of negotiator further (see Olivier and Wodon 2012). While some CHAs would see the ACHAP platform as taking on the main information gathering role, there is still not enough capacity in the platform for this to happen. In addition, the country-specific information gathering role (as potentially played by the CHAs) cannot be supplanted by a regional structure, simply because while there are certainly shared traits between CHAs, they operate in highly context-specific environments. For example, we have focused mainly on national level collaboration here, but there are also urgent questions being asked about the role of church health providers and CHAs at a district and local level – especially as countries increasingly decentralize health services. District-level collaboration requires a whole different range of partnerships and functions, and different system of information gathering and support. This means that each CHA needs to become increasingly involved in data gathering and information system building and also negotiate the partnerships necessary to build this capacity when necessary (for example reaching out to partners for TA or to universities for research support). This also means an increased role for CHAs in ‘chronicling’ their experiences, from their negotiations with particular partners, to highlighting best practices of their members. Given the rapid development of CHAs and the lessons they have learned over the last half century, another potentially powerful task for CHAs (and the new platform) might be to take a more proactive role in sharing these lessons with a broader range of networks, in particular the emerging non-Christian NFBHNs. In Ghana, for example, the Islamic 92 health providers have a much smaller market share of facilities – and also a significantly weaker relationship with government, even though some hospitals function as district hospitals (Miralles et al 2003). They are also clustered in areas in which the CHAG providers are not. This might certainly be an opportunity for CHAG (and other CHAs) to take on a stronger role of inter-faith collaboration. Supporting innovation and flexibility Another potential function for attention of the CHAs might be increased support of innovation and flexibility. Many church health providers have been noted as being particularly innovative: in the way they respond to need with limited resources, in the ways they connect with community, or the health-providing traditions and operations that have been built over decades of in-context trial and error. Many church health providers could also be considered to be ‘rural health specialists’, advising others (including government) on what it takes and means to operate health services in hardship areas. However this role is not properly documented or recognized. Another concern which has been raised as CHA members become increasingly integrated into national health systems (surely a good thing) – is that they may be losing some of their characteristic strengths in the process. For example, Gilson et al (1997) noted that church health providers “tend to enjoy greater flexibility, adaptability and innovativeness than government providers because they are not governed by rigid bureaucratic procedures.� Gilson et al argued that while increased collaboration and contract compliance with the public sector makes it easier for governments to integrate private providers in national systems, it may also impose constraints that affect the very nature of the private sector, including its flexibility and innovativeness. “Formalizing the government/church relationship through contracts may, thus, change the nature of church providers and so undermine some of their comparative advantages over government� (Gilson et al 1997). This means that a potential function for CHAs might to take on the challenge of protecting the innate flexibility and innovativeness of church health providers (even while negotiating with government for improved integration), and partnering with others to document and demonstrate this more effectively. CONCLUSION These suggestions of potential areas for future focus emerge from the CHAs themselves (it is certainly not for us to say what role CHAs should play in the future). The CHAs have benefitted from increased attention over the last half century. They have established a particular and important role for themselves – especially in African health systems which require constant negotiation and intricate collaboration between a variety of public and private providers. The challenges and suggestions above all hint at an increasingly complex role for CHAs going forward – and a role which will require more diverse and complex capacities for CHA staff. When CHA members sometimes respond that they ‘just want to get on with the healing’ – the CHAs are then handed the task of looking forward. This role extends beyond that of an ‘umbrella network’ which organizes meetings or facilitates dialogue between partners, and beyond that of a standard intermediary organization which channels funds or builds technical capacity among members. In resource constrained environments, it now seems that it is the CHAs which 93 have mainly been tasked (by their members) with the role of looking forward beyond immediate need – to take the longitudinal view of where faith-inspired health providers should be in the next fifty years. REFERENCES Asante, R. K. O. 1998. Sustainability of Church Hospitals in Developing Countries: A Search for Criteria for Success. Geneva: World Council of Churches. Atingdui, L. 1995. “Defining the Nonprofit Sector: Ghana�, in L. M. Salamon and H. K. Anheier, editors, Working Papers of the Johns Hopkins Comparative Nonprofit Sector Project. Baltimore: The Johns Hopkins Institute for Policy Studies. Boateng, J., H. Surnye, A. Mensah, B. Boateng, P. Nyarko, N. Munguti, and J. H. Bratt. 2006. “Costs of Reproductive Health Services Provided by Four CHAG Hospitals, Frontiers in Reproductive Health Program and Family Health International Supported by USAID. Boulenger, D., B. Keugoung, and B. Criel. 2009. “Contracting Between Faith-based and Public Health Sector in Sub-Saharan Africa: An Ongoing Crisis? The Cases of Cameroon, Tanzania, Chad and Uganda.� Antwerp, Institute of Tropical Medicine, Study commissioned and published by the Medicus Mundi International Network. Christian Health Associations at Crossroads (CSSC). 2007. “Christian Health Associations at Crossroad Toward Achieving Health Millenium Development goals.� Christian Health Association’s Conference, 16-18 January 2007, Dar es Salaam. Tanzania: (CSSC) Christian Social Services Commission, pp.1-28. _____. 2007. “Christian Health Association’s Conference Planning Document.� Tanzania: (CSSC) Christian Social Services Commission. Dimmock, F. 2007. “Faith-Based Health Networks in Africa.� Paper presented at Christian Connections for International Health Annual Conference, Washington, DC. _____. 2011. “Maintaining and Strengthening Religious Health Assets: Challenges Facing Christian Health Associations in the Next Decade.� In B. Schmid, J. R. Cochrane, and T. Cutts, eds., When Religion and Health Align: Mobilizing Religious Health Assets for Transformation, Cluster Publications: Pietermaritzburg. Dovlo, D. Y. 2005. “Social Dialogue in the Health Sector: Case study Ghana.� Sectoral Activities Programme, Working Paper Series, WP 234, Geneva: International Labour Office, World Health Organization. Ewert, D. M., and A. Merrill, editors. 1995. A New Agenda for Medical Missions. MAP International. 94 GFATM. 2008. “Report on the Involvement of Faith-based Organizations in the Global Fund.� The Global Fund to Fight AIDS, Tuberculosis and Malaria, Geneva, Switzerland. Gilson, L., J. Adusei, D. Arhin, C. Hongoro, P. Mujinja, and K. Sagoe. “Should African Governments Contract Out Clinical Health Services to Church Providers?.� In S. Bennett, B. McPake and A. Mills, eds., Private Health Providers in Developing Countries: Serving the Public Interest?, London and New Jersey: Zed Books. Green, A., J. Shaw, F. Dimmock, and C. Conn. 2002. “A Shared Mission? Changing Relationships Between Government and Church Health Services in Africa.� International Journal of Health Planning and Management 17: 333-353. Haddad, B., J. Olivier, and S. De Gruchy. 2008. “The Potential and Perils of Partnership: Christian Religious Entities and Collaborative Stakeholders Responding to HIV and AIDS in Kenya, Malawi and the DRC.� Cape Town: Africa Religious Health Assets Programme. Makinen, M., A. Sealy, R. A. Bitran, A. Adjei, and R. Munoz. 2011. “Private Health Sector Assessment in Ghana.� World Bank Working Paper Series No.210.: World Bank, Washington, DC. Mandi, F. 2006. “Planning, Development and Supporting the Faith�based Fealth Workforce.� African Christian Health Associations' Human Resources for Health Mini� Forum, Nairobi, Kenya, The Capacity Project and Medicus Mundi Int. McGilvray, J. 1981. The Quest for Health and Wholeness. Tübingen: German Institute for Medical Mission. Miralles, M. A., F. Aboagye-Nyame, et al. 2003. « Access to Essential Medicines: Ghana, Strategies for Enhancing Access to Medicines Program.� Arlington, VA: Center for Pharmaceutical Management, Management Sciences for Health. Olivier, J., and Q. Wodon, 2012. “Estimating the Market Share of Faith-Inspired Health Care Providers in Africa: Comparing Facilities and Household Survey Data.� In J. Olivier and Q. Wodon, eds, The Role of Faith-Inspired Health Care Providers in Sub-Saharan Africa and Public-Private Partnerships., HNP Discussion Paper, World Bank, Washington, DC. Rasheed, F. 2009. “Collaborations Between Public Health Systems and Faith-Based Organizations: Memorandum of Understanding Between the Ministry of Health and the Christian Health Association, Ghana.� Programme on Partnerships and UN Reform - World Health Organization, Geneva. Schmid B., E. Thomas, J. Olivier, and J. R. Cochrane. 2008. The Contribution of Religious Entities to Health in Sub-Saharan Africa. Cape Town: African Religious Health Assets Programme. 95 Van Reken, D. E. 1990. “Medical Missions and Development of Health.�, in D. M. Ewert and A. Merrill, editors. A New Agenda for Medical Missions, MAP International: 23-24. 96 ANNEX: Basic information for Christian Health Associations in Africa Country Full name DOB Angola CICA: Conselho das Igrejas Cristãs em Angola [Council of Christian Churches in Angola, Christian Medical Commission] Establishment and Members: Established in 1977 as an ecumenical institution with the purpose of 1977 being the Angolan faith community's guiding institution, and prophetic voice of its member institutions towards local and international governments. CICA's mission is to contribute to the peace process and to strengthen ecumenism in Angola. ~ 20 church members CICA Work: provides guidance and technical assistance in community assistance and development, literacy, vocational training, community health (including HIV and AIDS), development of youth, ecumenical co-operation, and peace education and reconciliation. Benin AMCES-Bethesda: Association des Œuvres Médicales Privées Confessionnelles et Sociales au Bénin [‘Association of Private Church Medical and Social Works in Benin’ Establishment and Members: In 1985 the ‘Association of Private Church Hospitals of Benin’ 1985 (AHPOC) with was established – later instituted as AMCES the ‘Association of Private Church Medical and Social Works in Benin’. There were five initial facility-based members (hospitals). There are now ~20 members with 30 related health facilities. AMCES Work: AMCES’ focus is to promote and foster cooperation, understanding, exchange of information and experience, consultation among members and with the Beninese government and all governmental or non-governmental national and international; to organize collaboration; solve common problems; promote work of all non-profit providers; and defend the moral and material interests of its members. Websites:www.amces-benin.org ⋅ In Cameroon, ECMA was also created as a platform to coordinate all private, non-profit providers – religious (Catholic, Protestant and Muslim) and secular (association and NGOs) members. ⋅ Bethesda is a faith based hospital that was started in Benin in 1990 by the Council of Churches of Benin with an aim to "improve the physical and social wellbeing of the individual and the family at accessible and reduced rates". Website: http://www.bethesdabenin.org/ Botswana AMMB: Association of Medical Missions for Botswana Establishment and Members: Established in 1973-1974, AMMB began as an ecumenical arm of the 1974 Botswana Christian Council – before developing as an association in its own right. Work: AMMB is a loose association of Christian health facilities. Cameroon CEPCA: Conseil des Eglises Protestantes du Cameroun [Council of Protestant Churches of Cameroon – Health Department] Establishment and Members: CEPCA was established in 1941 (as FEMEC), with its Department of 1957 Health established in 1957 and reconstituted as CEPCA in 1968. A federation of 11 protestant churches and missions, managing 30 hospitals, 150 health centers and 3 nurse training schools. CEPCA Work: CEPCA has several departments, including those focused on education, women and communication. The Department Health assists churches in the management of health facilities, organizes seminars on hospital management of certain diseases, distributes grants when the state gives, serves as a link with the ministry. CAR ASSOMESCA: Oeuvres Médicales des Eglises pour la Santé en CentrAfrique Establishment and Members: Established in 1989 as an ecumenical collaboration with 16 member 1989 groups: 8 Protestant and 8 Catholic. Participating members now include Lutherans, the national work associated with Baptist Mid-Mission, Grace Bethren, Catholic, Swiss Pentecostal, Apostolic, Swedish Baptist and a few independent groups. Work: ASSOMESCA coordinates and builds capacity among members – and operates a drug distribution system in CAR, with customers (church member groups) in CAR, DRC and Congo. 97 Chad AEST/UNAD-sante: Union Nationale des Associations Diocésanes de secours et développement Establishment and Members: Newly forming collaborative 2009 DRC ECC-DOM: Eglise du Christ au Congo - Direction des Oeuvres Médicales & SANRU Establishment and Members: The Protestant Church of Zaire (currently known as ECC – Eglise du 1971 Christ au Congo) came into existence in 1971 with around 60 member communities. At the same time, a medical office, the Direction des Oeuvres Médicales (DOM) was created to coordinate the health work of the ECC members and to serve as the liaison with the Ministry of Health. In 1999, in a major move, the DRC Ministry of Health formally turned over responsibility for health care in 60 zones (of a total of 306) to a coalition of mostly faith-based non-governmental health organisations – with ECC-DOM as implementing partner (the project named SANRU (Projet Santé Rurale). ECC currently co-manages 65 of the 515 health zones in the DRC, with more than 50 hospitals and several hundred dispensaries. The ECC network of 64 member communities includes Anglican, Presbyterian, Evangelical, Baptist, Pentecostal, Methodist and Mennonites. Health Work: Through its member communities, ECC-DOM provides community-base health care, hospital and dispensary-based care. ECC also plays an important collaborative and networking role, uniting most protestant congregations and health efforts. ⋅ SANRU includes ECC-DOM as well as the Salvation Army, Kimbanguist Medical Department, and the Catholic church. website: www.sanru.org Ethiopia CHAE: Christian Health Association of Ethiopia Establishment and Membership: CHAE is currently in the formative stages. Other new ⋅ Also networking in Ethiopia is CRDA (Christian Relief and Development Association) – established in 1973 as an umbrella organisation of 212 NGOs and FBOs. Said to coordinate around 89 hospitals and several hundred lower units). Today it focuses on development, capacity building, advocacy and networking. ⋅ Also networking is EECMY-DASSC (The Ethiopian Evangelical Church Mekane Yesus Development and Social Service Commission) website: www.eecmydassc.org.et/health.htm Ghana CHAG: Christian Health Association of Ghana Establishment and Members: CHAG was founded in 1967 as Voluntary Professional Association. 1967 CHAG is open to any Christian church-related medical institution in Ghana, which is recognized as such by the Ministry of Health. Founding members: Ghana Catholic Bishops Conference, Christian Council of Ghana, Ghana Pentecostal Council. Institutional members: the hospitals and clinics, which belong to the founding members and share in the responsibilities and benefits. Associate members: other Church-related institutions which share in the aims and objectives of the association and share only some limited benefits and responsibilities. Facilities: 152 institutions: 56 hospitals, 83 primary health care bodies and 8 health worker training centres - most belonging to the Catholic church, then the Presbyterians, and then the SDAs. CHAG Work: CHAG is an umbrella organization that coordinates the activities of the Christian Health Institutions and Christian Churches’ Health programmes in Ghana. It is a body through which all or most of the Christian Church related health facilities/programmes liaise with the Ministry of Health to ensure proper collaboration and complementation of the government efforts at providing for the health needs of Ghanaians. Main activities: policy analysis, advocacy & lobbying, capacity building of members, networking & public relations (or public image building), translating government policies in operational terms for members to implement. website: http://www.chagghana.org 98 Kenya KEC: Kenya Episcopal Conference (CHC: Catholic Health Commission) Establishment and Members: Although Catholic health care provision in Kenya dates back to the 1957 early 1900s – KEC was founded in 1957, with the CHC established in 1967. The Catholic Health 1967 Commission provides oversight and co-ordination of 456 Catholic health facilities affiliated to 25 Catholic Diocese in Kenya – as part of the church’s social and pastoral mission. These include 53 hospitals, 82 health centres and 311 dispensaries. In addition, the Church has Community Health Programs that offer Mobile Clinics, HBC and care for OVC. KEC-CHC work: CHC provides oversight, advocacy, lobbying and representation, capacity building, networking, and management to its members. CHC member facilities provide a holistic and wide range of preventive, rehabilitative and curative health services and programs. website: www.kec.or.ke Kenya CHAK: Christian Health Association of Kenya Establishment and Members: CHAK’s history dates back to the 1930s when it was established as a 1930 Hospitals’ Committee of the National Council of Churches of Kenya (NCCK) - as an umbrella 1987 organisation of health facilities or programs owned by Christian denominations or missionary groups who were providing health services in Kenya. Re-constituted as a non-profit organization (CHAK) in 1987 – membership includes 33 affiliated Protestant church denominations, responsible for 455 facilities - 25 hospitals, 48 health centres, 324 dispensaries, 10 nursing training colleges and 58 church health programs. CHAK Work: CHAK’s core functions on behalf of its members are advocacy & representation, capacity building, health care technical services, technical support, networking, communication and HIV/AIDS programs. CHAK member facilities provide a wide range of preventive, rehabilitative and curative health services. website: www.chak.or.ke ⋅ CHAK and KEC jointly own MEDS (Mission for Essential Drugs and Supplies) - which provides essential drugs and medical supplies, as well as training of church and other not-for- profit health facilities in the management and appropriate use of drugs. “MEDS’ current clientele of more than 1,500 health facilities in Kenya and other countries (Sudan, DRC, Ethiopia, Somalia, and Tanzania) ⋅ CHAK was recently appointed to host the first secretariat for the African Christian Health Assocations Platform (ACHAP) Lesotho CHALe: Christian Health Association of Lesotho Establishment and Members: In the 1960s, physicians from various mission hospitals began 1974 meeting to discuss common problems, and in the 1970s the Christian Council of Lesotho and Oxfam advocated a more formal organization. The Minister of Health also urged a more formal association in 1973, and in 1974 CHALe was founded as a voluntary association of Christian churches providing not-for-profit health care services to the Basotho, particularly in hard to reach places around the country. Members include: The Anglican Church of Lesotho, Assemblies of God Church, Bible Covenant Church, Lesotho Evangelical Church, Roman Catholic Church, and the Seventh Day Adventist Church of Southern Africa. These manage around 72 health centres and 8 hospitals. CHALe Work: CHALe assumed operational responsibility for health services in 8 of 18 health service areas. CHALe’s service programme comprises of three components: capacity building, primary health care and medical services, infrastructure improvement – as well as several projects including: HIV/AIDS prevention and control programme, primary health care programme (MCH/FP, nutrition, environmental health), child protection, rural health development and rural clinic improvement programs. Liberia CHALi: Christian Health Association of Liberia Establishment and Members: Founded in 1975, is an ecumenical umbrella body of Liberian 1975 churches involved in the health sector. 6 hospitals, 67 health centres. CHALi Work: CHALi supports members work in several fields, namely: drug supply, PHC, capacity building, water and health, family education and HIV/AIDS control. CHALi is currently being revitalized with a new constitution and strategic plan being drafted. 99 Malawi CHAM: Christian Health Association of Malawi Establishment and Members: CHAM was established in 1966 as the ‘Private Hospitals Association 1966 of Malawi’ - following a meeting of the WCC and church leaders in Malawi. In 1991 the Association changed its name to CHAM to reflect its Christian identity and its focus on broader health ministry. Membership consist of 18 different Catholic and Protestant churches and church organizations (i.e. ecumenical and interdenominational organizations that operate health facilities). There are also 12 associate members, and facilities are currently at 27hospitals, 142 health centres and 10 training institutions. CHAM Work: CHAM works for the improvement and expansion of health facilities, facilitation of inter-denominational cooperation, collaboration with government and other organisations, the development and coordination of training programs including nursing schools, the provision and coordination of support services. website: www.cham.org.mw Mali APSM: Association Protestante de la Santé au Mali (Association of Evangelical and Protestant Groupings of Mali) Establishment and Membership: An associate member of the AGEMPEM, the APSM (Association 1992 of Evangelical and Protestant Groupings of Mali, Association Protestante de la Santé au Mali) was established in 1992 by the Health Personnel of the Protestant Churches and Missions in Mali. It is non-political, inter-denominational, non-profit, nongovernmental organization – set to serve all the people of Mali and also evangelical Christians. APSM Work: APSM’s objectives are to encourage its members to provide quality health care to the Malian population, and facilitate the coordination of the medical activities of the churches and missions. The APSM designs, amongst others, literacy and health education programmes for women and children. In the area of HIV/AIDS, APSM undertakes prevention exercises, and has produced information and awareness-raising documents in Bambara, the most widely-spoken national language. website: www.apsmmali.org Namibia CCN-ECN: Council of Churches in Namibia Establishment and Members: Formed in 1978, CNN is an ecumenical body that is focused on 1978 religious, education and social concerns. Members manage around 8 hospitals in Namibia. Nigeria CHAN: Christian Health Association of Nigeria Establishment and Members: CHAN was established in 1973 by founding members: The Catholics 1973 Bishops Conference of Nigeria (CBCN); The Christian Council of Nigeria (CCN); and The Northern Christian Advisory Council of Nigeria (NCMAC) to facilitate cooperation between Member Institutes (MIs) and to help build capacities in order to better serve the Health needs of Nigerian population. Members include ~400 registered Member Institutions (MIS) operated by 15 denominations. These manage 147 hospitals, around 3000 community facilities, and 28 training centres. CHAN Work: As an umbrella network, CHAN works to coordinate and assists the health services of its members. Primary Health Care Services (PHCS): activities include training of village health workers and traditional birth attendants, nutrition, immunization, maternal and child health care, growth monitoring, water and sanitation, management training for various health workers of different levels, HIV/AIDS and STD control and AIDS care activities, holistic health care activities and program development, a resource centre with a bookstore. CHAN also manages a drug supply service, CHANPHARM – which is responsible for essential drugs importation, production and supply to member institutions. Website: www.channigeria.org 100 Rwanda BUFMAR: Bureau des Formations Médicales Agréees de Rwanda [The Office of Church-affiliated Health Facilities in Rwanda] Establishment and Members: Established in 1975, BUFMAR is an umbrella organisation that 1975 represents both the Catholic and Protestant churches and their health facilities and programmes throughout Rwanda. It is comprised of 24 Christian churches and services with 120 health facilities. Church-affiliated health facilities represent 45% of hospitals and 35% of primary level care facilities (health centres, dispensaries, health posts). BUFMAR work: BUFMAR coordinates and supports member facilities. Senegal EPSCM: Eglise Protestant du Senegal Commission Medicale Establishment and Members: A newly established Protestant umbrella group in Senegal. Still in new formation. Sierra CHASL: Christian Health Association of Sierra Leone Leone Establishment and Members: Established in 1975, members are heads of churches and health 1975 institutions. Facilities include hospitals and health centres Sudan CHAS: Christian Health Association of Sudan Establishment and Members: CHAS has been slowly forming since 2002, with initiating members 2008 such as the Sudan Council of Churches, and reshaping the function of older Health Desks and Health Secretariats (CEAS - Church Ecumenical Action in Sudan – has recently been taken over by CHAS). The task to facilitate the establishment of CHAS became more visible in 2004 with technical and financial support from ICCO and later EED and Caritas Australia. The envisaged goal is to evolve CHAS from a network of Christian health organizations into a legally registered entity that functions fully as a Health development arm of the Sudan Council of Churches. (Note, the Sudanese context and the role of CHAS is rapidly changing – especially in South Sudan where international NGOs are currently rehabilitating health facilities destroyed during the civil war. CHAS work: Still forming, the vision is that CHAS will support Christian health care program with technical support, budget proposals, financial management, the provision of medical supplies. www.chasudan.org Tanzania CSSC: Christian Social Services Commission Establishment and Members: Established in 1992, The Christian Social Services Commission is an 1992 umbrella body that brings together the Tanzania Episcopal Conference (TEC), representing the Catholic Church, and Christian Council of Tanzania (CCT), representing about 14 Protestant Churches and 10 Church related Organizations. CSSC coordinates 89 hsopitals, 815 health centers and dispensaries, and 24 training centers. CSSC has two executive organs, the Christian Medical Board of Tanzania (CMBT) and the Christian Education Board of Tanzania (CEBT) for health and education respectively. CSSC work: CSSC is involved in fostering ecumenical cooperation in matters regarding social services provided by Tanzanian Churches, lobbying and advocacy with government towards improving the environment for provision of church related services among other things. CSSC is also involved in education through the establishment of various schools. website: http://www.cssc.or.tz Togo APROMESTO: L'Association Protestant des Oeuvres Medico-sociales du Togo [The Protestant Association Medico-Social Works of Togo] Establishment and Members: Established in 1994, bringing together 7 churches of the Christian 1994 Council of Togo. Facilities include 3 hospitals and 39 HC/lower units. APROMESTO work: Co-ordinating the action of the health centres and hospitals belonging to these churches, sensitising the faithful in the struggle against AIDS, train the nursing personnel and resolving the health problem through concerted actions. APROMESTO encourages the team of health facilities to establish a psychosocial care unit and draw up AIDS projects in order to address the epidemic more effectively. 101 Uganda UPMB: Uganda Protestant Medical Bureau Establishment and Members: Established in 1957, UPMB is an umbrella, private, not-for-profit, 1957 Faith Based Organisation (FBO) of Protestant Churches and Church-related organizations involved in Health care in Uganda. Originally established to provide co-ordination and collaboration between medical care institutions affiliated to the Protestant churches in Uganda and the Ministry of Health – it then became the official channel for disbursing government grants-in-aid to the hospitals, evolving into a national umbrella organisation. Facilities include 15 Hospitals, and 251 HC/lower units, and 7 training institutions. UPMB work: SUPMB is involved in advocacy and lobbying for policy on behalf of its members, capacity building, publicity and networking and support and supervision. website: http://www.upmb.co.ug/ Uganda UCMB: Uganda Catholic Medical Bureau Establishment and Members: Established in 1955-1956, with the main purpose of overseeing the 1955 procurement of medical drugs and equipment and distributing aid provided by the colonial 1956 government to the voluntary health sector. With the establishment of the Uganda Episcopal Conference, UCMB became the technical arm of the Conference’s Health Commission with co- ordination of health units organised on the intermediate (diocesan) and national levels. Facilities managed are: 27 hospitals, 240 HC/lower units, 12 training institutions. UCMB work: The UCMB invests in human resource management, financial management, health management information systems, assistance to dioceses to compile strategic plans, and quality improvement by adopting a gradually more sophisticated accreditation system called “faithful to the mission�. The strengthened capacity pays off in an improved performance at unit level (in terms of utilisation, cost, quality of care) and a better negotiation position with the MoH at central and local level. website: www.ucmb.co.ug ⋅ In 1979, the Catholic Medical Bureau and the Protestant Medical Bureau jointly established a drugs’ procurement agency: the Joint Medical Stores. The JMS is now an autonomous organisation in whose Board sit the representative of the founding bodies (the UPMB and the UCMB). Zambia CHAZ: Churches Health Association of Zambia Establishment and Members: Founded in 1970 (then as CMAZ: The Churches Medical Association 1970 of Zambia) through the merging of the Medical Committee of the Christian Council of Zambia and the health department of the Zambia Episcopal Conference, following a recommendation by the World Council of Churches. CHAZ acts as an umbrella organization, representing the interests of church administered health institutions in Zambia. Membership includes hospitals, health centres, faith based organizations and community based programs – in total 135 affiliates representing 16 different Catholic and Protestant churches. Facilities include: 30 hospitals and more than 36 hospitals, 81 HC/lower units (+ 29CBOs), and 9 training institutions CHAZ work: The stated mission of CHAZ is to be committed to providing technical, administrative and logistical services for affiliate members to serve communities with holistic quality health services that reflect Christian values, so that people live healthy and productive lives. CHAZ provides members with representation and advocacy, administrative and logistical support, technical support, and resource mobilization assistance. website: www.chaz.org.zm 102 Zimbabwe ZACH: Zimbabwe Association of Church Related Hospitals Establishment and Members: Founded in 1974, ZACH is the ecumenical medical arm of Christian 1973 churches in Zimbabwe – registered as an NPO or public voluntary organization. ZACH is accountable to the Heads of Christian Denominations (HOCD) in regard the running of Church Health Institutions/Hospitals. Facilities include 80 hospitals, 46 HC/lower units, and 15 training institutions ZACH work: ZACH represents the link between Head of Christian Denominations (HOCD), Ministry of Health and Child Welfare (MOH & CW) and other Health Providers and Agencies. Objectives include the promotion of Christian medical care; to facilitate and co-ordinate cooperation between member institutions and the Ministry of Health & Child Welfare and partners; to coordinate the planning, implementation and evaluation of projects and programmes; to assist member institutions in staff recruitment and development; to source funds and support; to assist member institutions to enhance their management capacity; to keep member institutions abreast and updated on management trends of Health delivery. 103 CHAPTER 6 PHARMACEUTICAL SERVICE DELIVERY IN CHURCH HEALTH SYSTEMS IN AFRICA: A CROSS-COUNTRY ANALYSIS Heather Budge-Reid, Donna Asiimwe Kusemererwa and Anke Meiburg Ecumenical Pharmaceutical Network Faith groups play a significant role in providing health care around the world although their exact contribution has not been well defined and few systematic studies have been done about pharmaceutical service delivery at facility level in the church sector. The baseline studies on access undertaken by EPN between 2005 and 2008 as summarised in this article, were a foundational attempt to address the information gap in this area. The studies were done to investigate compliance by church health services (CHS) with guidelines on efficient and effective pharmaceutical services. Church health facilities (numbering 363), representing over 20,000 beds, over 4 million outpatients and with budgets totalling more than 40 million US dollars were surveyed. Results of the study show some areas that are working well and others requiring urgent attention to improve effective and efficient provision of pharmaceutical services. INTRODUCTION Faith groups play a significant role in providing health care around the world although their exact contribution has not yet been adequately defined (Olivier and Wodon 2012). In Africa, churches are most actively involved in health care and medicines distribution in the sub-Saharan region (ACHA 2010) but even then wide variations exist and for countries like South Africa, for example, church-based health care is virtually non- existent after such hospitals were nationalized in the 1970s (Coovadia et al 2009, Patel et al 2009). Most church based health providers describe themselves as existing to serve the poor, marginalized and un-reached (CHAL 2008, CHAM 2007, CHAG 2008) and as such the activities of these churches tend to be more concentrated and more significant in rural areas. 37 Any health system is expected to deliver appropriate health products and services in an equitable, reliable and efficient manner. From a public health perspective, one of the key measures of health system effectiveness is its ability to provide medicines. Faith based organizations in three African countries (Malawi, Kenya and Uganda) were, in the late nineties, estimated to handle between 20 percent and 40 percent of their country’s flow of pharmaceuticals (Kawasaki and Patten 2002). However not enough data is available on the exact contribution of the different church-based players to medicine supply in Africa. Church-run pharmaceutical systems vary greatly in size, 37 Church-based health care outside of Africa is less often addressed – for example, in Papua New Guinea churches are said to provide up to 50% of the country’s health services and are well integrated within the national health system (AusAid 2009). 104 nature of operation and impact within the country in which they are situated (Banda et al 2006). In some countries the systems are well developed with one agency undertaking procurement, warehousing and distribution for the entire sector. In these cases the central units such as Mission for Essential Drugs and Supplies (MEDS) in Kenya, the Christian Health Association of Nigeria’s CHANMedi-Pharm in Nigeria and the Joint Medical Store (JMS) in Uganda are significant players, supplying pharmaceuticals to their national markets and beyond (MEDS 2008, Kawasaki and Patten 2002). These organizations usually also offer broader services such as quality-testing of medicines and training on rational medicine use (MEDS 2008). Others such as Catholic Drug Centre in Accra which is a pharmaceutical unit of the National Catholic Health Service of Ghana have had mixed fortunes and are currently not operating at full capacity (Ballou-Aares et al 2008). Among the success factors identified for church-run medical supply systems are transparent procurement procedures, motivated staff (even when these staff are seconded from government health sector) and competitive pricing (Banda et al 2006). In general the church sector supply system rarely operates as a distinct entity and more often than not is deeply enmeshed with other national and regional supply systems including public and private-for-profit providers (Ensor and Witter 2001 in Bennett et al 2005, Ballou- Aares et al 2008, CHAK 2008). Supplying medicines to health facilities is a crucial step in providing quality services to patients. However, pharmaceutical service delivery at a facility level through church-based providers has not been systematically studied. Between 2005 and 2009, EPN undertook a series of access baseline studies as a first step to address this information gap – the results of which are reported here, utilizing a cross- country perspective. 38 ACCESS BASELINE STUDIES: METHODOLOGY, LIMITATIONS AND ANALYSIS Between 2005 and 2009, the Ecumenical Pharmaceutical Network (EPN) undertook five full baseline studies (in Ghana, Ethiopia, Malawi, Uganda and Zambia), and three partial studies (in Nigeria, Tanzania and Togo) which provided additional comparative data. EPN is an independent non-profit Christian networking organization that works to increase access to medicines and pharmaceutical services through church health systems. 39 The studies were undertaken to analyze the ‘compliance’ of church health services (CHSs) with guidelines on efficient and effective pharmaceutical services (EPN 2005). The guidelines are a tool intended to support health facility managers, pharmaceutical personnel and others involved in pharmaceutical services delivery to provide a quality service. The survey covered 363 church health facilities, representing over 20 000 beds, over 4 million outpatients and budgets totaling more than 40 million US dollars. The country information was collected using four tools: a church- related health service self-assessment survey; a faith-based drug supply organization survey; a guided desk review for each country; and guided assessment workshops held in up to 10 hospitals in each country (the workshops always emphasizing health service staff as central to the identification of priorities, and that their ideas should be the basis of 38 More extensive reporting of these findings can be found at www.epnetwork.org 39 EPN is by nature an international network of associations, institutions and individuals who have an interest or are involved in the delivery of just, equitable and compassionate quality pharmaceutical services. 105 intervention). Through self-assessment questionnaires, 127 hospitals, 178 clinics and 58 health posts were surveyed. In addition, 50 focus groups were held with church leaders (although not reported on here); more than 50 guided assessment workshops were held with hospital staff; and 5 national feedback meetings were also held. Given the lack of preceding base-line studies, there were several challenges to this work that placed limitations on the results. Despite pressure from management committees and senior staff, getting fully-completed self-assessment surveys (questionnaires) was challenging - also indicated by the high number of ‘don’t know’ responses. Translation into French also presented some difficulties since a few of the cross-check questions resulted in identical translations. In some countries, significant time passed between project stages, and in some contexts, the practical challenges of communication made the data-gathering process difficult. Nevertheless, sufficient data was available for analysis. Utilizing Epi Info™, a database was created for each country, and other data were analysed separately. Country reports were fed back to stakeholders in feedback meetings in each country, where priority areas were identified. SUMMARY OF RESULTS The cross-country analysis presented in this article is drawn from the raw data from five core countries, combined with data from a further three countries, where applicable. This data was reanalysed using Epi Info™ to calculate overall results across all the countries. The findings on compliance with the guidelines on efficient and effective pharmaceutical services has been organised according to the six building blocks of the WHO’s Framework for Action for Strengthening Health Systems (WHO 2007). The original EPN guidelines around which these studies were framed are now categorized according to these WHO health system strengthening (HSS) building blocks – allowing for results to be considered in relation to this broader focus. An average score is calculated to give an overall level of compliance with the guidelines associated with each building block. Table 1: Analysis of issues/guidelines addressed under each building block Building Block Issues addressed by the guidelines 1. Service Delivery Quality assurance policies, disaster preparedness procedures, rational use of medicines 2. Health Workforce Presence of pharmaceutical personnel, assessment of pharmaceutical unit work, representation of pharmaceutical function in decision making 3. Information Access to key pharmaceutical information, cross institutional information sharing 4. Medical Products & Supply systems, medicines management best practices, medicines donations Technologies 5. Financing Pricing policies, government subsidies , tax exemption, pro-poor ethic 6. Leadership & Governance structures and access to information, Transparency mechanisms, Governance community involvement, medicines and therapeutics committees, church leader engagement Source: Authors. 106 In summary, a consolidation of the individual scores on compliance using the building block framework shows that as far as pharmaceutical services are concerned, building blocks 1 (service delivery) and 4 (medical products and technologies) appear to be the most problematic for CHSs. The sector also fares poorly in leadership and governance where the guidelines address transparency, community involvement and understanding of health facility owners of their roles and responsibilities (Figure 2). Building block 5 (financing) had an overall score of 63 percent for the guidelines which included pricing policies, government subsidies, pro-poor ethic and tax exemptions available to CHSs. Although the guidelines included do not reflect the availability of funding nor accounting capacities, the fairly high score does show strength in this area. Figure 1: Overall guideline compliance scores by building block 100% 80% 60% 40% 54% 48% 63% 20% 38% 36% 41% 0% Block 1: Block 2: Block 3: Block 4: Block 5: Block 6: Service delivery Health workforce Information Medical Financing Leadership and products, governance vaccines & technologies Source: Authors. Issues affecting access More than 50 guided self-assessment meetings were held with hospital staff. A problem tree tool was used to help participants develop their understanding factors affecting access to medicines, and to use that understanding as an input into intervention development that would maximize access to medicines. Key issues identified in self- assessment workshops that affect access to medicines (combined per country): ⋅ Ethiopia (9 workshops): funding ⋅ Ghana (9 workshops): personnel, funding, management, rational use of medicines (limited public awareness) ⋅ Malawi (10 workshops): poverty (funding), infrastructure (funding), donor fatigue (funding); lack of skilled staff, staff turnover, HIV and AIDS resource drain (funding), low-skilled management, lack of medicines and therapeutics committee, poor management information systems ⋅ Nigeria (8 workshops): staffing problems ⋅ Uganda (10 workshops): poor availability of medicines, staff losses, funding ⋅ Zambia (9 workshops): staff retention 107 DETAILED FINDINGS ON COMPLIANCE WITH SELECTED GUIDELINES There are many factors that make CHSs non-comparable such as differences in denominational character, funding sources, management approaches and contexts. However, it is possible to draw out common trends across the CHSs studied, such as similarities in funding challenges, increasing demand for services, and significant shortages of pharmacy staff. Although the level of compliance varied from low to a maximum of 80 percent for the highest compliance calculated, there was evidence in each country of at least one facility that scored highly for each guideline. The following section gives detailed findings on compliance with selected guidelines whose full implementation should have a significant impact on access to medicines, and for which the measures investigated most comprehensively reflect the issue the guideline addresses. Building block 1: Service delivery Quality assurance policy in place and implemented: 67 percent (223 of 334) have written standard operating procedures (SOPs) for receiving delivery of medicines, with 71 percent of these indicating that they are fully implemented (23 percent only partially implemented), putting the use of SOPs for receiving medicines at 47 percent. 62 percent have written SOPs for issuing of medicines to pharmacies or wards; with 77 percent of these facilities indicating that they are implemented, putting the use of SOPs issuing medicines at 48 percent. Disaster preparedness procedures in place: 27 percent have a plan for what to do in the event of a large-scale natural disaster, such as flooding, or conflict-based emergencies. Building block 2: Health workforce At least one pharmaceutically trained person per facility: 51 percent of all facilities recorded to have pharmaceutically trained staff employed. The total number of staff reported was 613. Facilities were asked about the percentage of their pharmaceutically trained staff that were trained by the government as opposed to private training schools. 128 facilities answered this question and the average response was 20 percent. However, 58 percent of these responses were scores of zero. Table 2: Number of pharmaceutically trained personnel Nurse with some How many pharmaceutically trained Pharmacy Pharmacy pharmacy personnel work in your facility? Pharmacist technician assistant training 40 Number of staff 48 111 183 271 Percentage of all facilities reporting this 11 16 24 30 cadre of trained staff Source: Authors. 40 For the purpose of this study a nurse with some pharmaceutical training is one who in addition to their nursing qualification had received training in any aspects of pharmaceutical service provision 108 Table 3: Training received by pharmaceutical personnel How many pharmaceutical personnel Pharmacist Pharmacy Pharmacy Nurse: some % of have the following training? 41 technician assistant pharmacy total training Quantification of medicines needs 40 72 69 85 43 Medicines store management 45 81 76 112 51 Quality assurance 40 82 54 79 42 Rational use of medicines 54 79 64 129 53 Unit costing 37 72 65 98 44 Record and data management 46 75 63 136 52 Financial management 38 33 25 68 27 Compounding of medicines 38 87 59 84 44 Dispensing 45 94 112 221 77 Source: Authors. Building block 3: Information Access to key pharmaceutical information (available in the pharmacy area, and available in the facility): Essential medicines lists and formularies were the reference books most readily available. Table 4: Available pharmaceutical information Publications available % Yes In the pharmacy? Essential Medicines List (national or WHO) 66 National (or WHO) formulary 52 Standard treatment guidelines 64 New and obsolete medicines list 29 List of medicines registered in the country 28 International medicines pricing indicators 28 Managing medicines supply manuals 37 In the facility? National drug policy 43 MTC guidelines 25 WHO essential medicines publications 30 Rational use of medicines information 45 Source: Authors. Building block 4: Medical products, vaccines and technologies Implementation of standard operating procedures for procurement: 55 percent had written SOPs for procurement, and 68 percent of this group actually implement them fully. This puts the use of SOPs for procurement at 38 percent. Compliance with best practices for medicines storage and management: 67 percent indicated that there is a checklist for good storage conditions for medicines, and 85 percent of this group said that this checklist is actually in use, putting the use of a checklist at 57 percent. 55 percent have written SOPs for procurement, and 68 percent of this group actually implement them fully. This puts the use of SOPs for procurement at 38 percent. Facilities were asked how often stock-taking reports on medicines are 41 Respondents were simply asked to state whether pharmacy staff had received training in a particular area irrespective of the duration or nature of the training 109 compiled. Of these, 31 percent reported stock-taking more than once a month, 38 percent once a month and 31 percent less frequently than once a month. Of the sample list of essential medicines provided in the survey, facilities experienced stock-outs of the following items most commonly: 42 Table 5: Stock-outs of essential medicines Do you ever run out of any of the following essential medicines? % Whitfield ointment (benzoic acid compound) 49 Povidone iodine 45 Amoxicillin 37 Tetracycline eye ointment 35 Co-trimoxazole 31 Quinine injection 31 Paracetamol tablets 31 Oral rehydration salts (ORS) 30 Procaine penicillin injection (often not standard treatment) 29 Chlorpheniramine 26 Aspirin 23 Ferrous salts 23 Mebendazole tablets 23 Folic acid tablets 22 Sulfadoxine-pyrimethamine tablets 21 Source: Authors. Table 6: Stock-outs of medical supplies for clinics Does your clinic ever run out of any of the following supplies? % Syringe disposable with needle 10cc 40 Bandage crepe 40 Gloves, examination latex non-sterile disposable 39 Syringe disposable with needle 5cc 36 Tape plastic adhesive microperforated 36 Gauze absorbent 33 Source: Authors. Table 7: Stock-outs of medical supplies for hospitals Does your hospital ever run out of any of the following supplies? % Surgical sutures 42 I.V. cannula 20G 37 Urine collection bag for adults 2000cc 37 Gloves surgical latex rubber sterile 33 I.V. giving set 21 Source: Authors. 42 Because of the need to make this survey as simple and as widely comparable as possible, specific formulations, (e.g. of strengths) were not used. The term “run out of� was used so that, in the perception of whoever was completing the form, the concept of “not being available for use� would be standard. WHO and EPN staff advised that including formulations and strengths could cause confusion, especially as dosages can be made up from a variety of strengths. What was required from these questions was a broad statement of availability of a medicine, so each one was only named. The non-availability of a medicine for any period is deemed unacceptable and is therefore a stock-out. 110 Building block 5: Financing Pricing policies in place and operationalized: 72 percent indicated that a pricing policy was in place for medicines and was able to give the percentage of margins that were applied. 64 percent indicated that consultations were charged on a fixed price basis (95 percent of the total) or calculated based on the time and services required. 59 percent answered that they would reduce the price charged for a medicine if the cost was reduced. Government subsidies extended: 68 percent indicated that part of the overall salary costs for their facility was paid by the government, ranging from 0.1 percent to 100 percent, with an average of 37 percent. Some 50 percent indicated that the government subsidized some of the medicines they provide. Pro-poor ethic in evidence: 90 percent of facilities indicated that they provide free or subsidized medicines and treatment to the poorest. 90 percent of this group were able to describe how their facilities achieved this. 298 facilities responded to the question “What percentage of people seeking health treatment at your facility in the end cannot pay?� with an average result of 18 percent of people not being able to pay for treatment. Building block 6: Governance and leadership All facility ‘owners’ have maximum understanding of roles, best practice, and management information: 78 percent have a Board or Committee, made up of individuals other than staff, who have overall responsibility for the running of the facility. They meet, on average, every 3 to 6 months, and members come from the synod, church, public sector, community services and business. 33 percent of the Boards or Committees had received training in the last five years. 48 percent of facilities with a Board or Committee were able to describe the information given to the Board, including annual and quarterly reports on activities and finances. 88 percent have an annual planning process that identifies such things as budgets, shortfalls, targets or plans for growth or improvement. 58 percent thought that their facility could survive with current levels of funding for at least two years. Transparency mechanisms in place in support of ‘Health for All’: 88 percent stated they had an annual planning process. 68 percent produce audited financial statements, but only 53 percent of these reports include pharmacy-related activities separate from other activities. This gives an overall result of 36 percent. 94 percent produce annual reports, and 46 percent of these reports include pharmacy-related activities separate from other activities. This gives an overall result of 43 percent. 66 percent have written staff recruitment policies, with 62 percent indicating that the policy is fully implemented. This gives an overall result of 41 percent. 51 percent have a monitoring and evaluation policy with 63 percent of these facilities indicating that the policy is implemented. This gives an overall result of 32 percent (26 percent of respondents to the question indicated that the policy was partially implemented). 43 43 The French translation of this question omitted the word ‘sanitaire’ from the phrase ‘formation sanitaire’, which may have led respondents to read the question as ‘is there a written M&E policy for your training’, rather than for ‘your health facility’. 111 Effective community involvement system in place: 69 percent reported ‘Yes’, with a range of descriptions of how the community was involved, coded in the table below. However, full involvement of the community would normally be indicated by community membership of the hospital management committee – in this case, 33 percent of facilities have an effective community involvement system in place. 184 facilities answered the question “What has been the main issue raised by the community?� Table 8: Most common descriptions of community involvement Form of involvement % Management committee, includes all community representation on hospital management 33 committees Outreach, includes all forms of ‘public health involvement’ between the health facility and 16 the community (such as immunization and community health training) Consultation, includes suggestion boxes, surveys, and contact with local councils and 13 community health workers Labour, includes donations of effort by the community to the health facility (such as 3 cleaning the compound and building new wards) (Described activity did not involve the community) 2 (No description given) 33 Source: Authors. Table 9: Most common reported issues raised by the community Type of issue % Affordability of medicines and services 24 Availability and quality of beds and services (additional structures or upgrading to the next 20 level of service delivery) Other issues 14 Transport, including ambulances and distance to facility 10 Availability and quality of staff 9 External to the facility, such as sanitation, water supply, income generation 7 Outreach, such as malaria prevention, immunization 7 Stock-outs of medicines and supplies 6 Waiting times at the facility 5 Attitude of staff 2 Source: Authors. Functioning Medicines and Therapeutics Committees (MTC) in hospitals: 44 Of the responding facilities, 126 were hospitals. 57 percent of hospitals said they have a MTC. 80 percent of these meet at least every three months; 84 percent of these keep minutes; 57 percent of these have written terms of reference; 70 percent of these have adapted the essential medicines list to fit local needs. Overall, only 15 percent of hospitals were able to answer all questions to satisfy having a functioning MTC. 44 Facilities with more than 49 beds. 112 CONCLUSION There are still challenges to improving adherence to good practice in pharmaceutical service delivery in the CHS sector. While much is known about what should be done, less is understood about why there are failures in implementation in some CHSs and not in others. In every country, there are examples of CHSs that are doing some things well – they are not the majority, but they exist and that suggests it is possible to succeed. More needs to be done to understand what the success factors are and therefore improve the quality and quantity of implementable known solutions. Although these surveys faced challenges, they do provide a baseline for comparison, and this methodology can be reused in order to assess changes over time. However, until more comparative studies are carried out in other health facilities, it is not known whether these problems are specific to CHSs or are similar across all health services in a country. The use of the WHO framework on the six building blocks of a health system to present the results of the study has limitations since the guidelines were developed prior. The scope of issues addressed by the guidelines is, in all cases, only a small component of what would be required to describe the full state of service as far as a particular building block is concerned. However it is does provide indications that for areas like financing, the CHSs are investing a lot more because their very survival depends on it while for softer service delivery issues like having the correct policies and procedure in place they are still lagging behind. There are country-specific concerns not addressed here. For example, the participants at the Uganda national feedback meeting agreed on the need for action in the areas of rational use of medicines and SOPs. They noted that there was already significant work being done, for example, the Joint Medical Store (JMS), Uganda Catholic Medical Bureau (UCMB) and Uganda Protestant Medical Bureau (UPMB) all had current programmes in these areas, as did WHO and others. However, it was recognized that still more could be done. For example: WHO and other leaders in the field have developed comprehensive supporting documentation for medicines storage and management and MTCs. Participants felt that more extensive distribution of these would be useful, and asked if WHO or JMS could set up a distribution system for these materials, as well as for the other information on pharmaceuticals. Distribution of information is costly, and it was hoped that a distribution system could be developed. There was also concern about the lack of coordination between the main actors. For example, no actor was working on these issues in the west of Uganda. It was also clear that there was no real coordination in terms of activities and working together, nor any real knowledge of each other’s activities. Furthermore, participants noted that while there was lots of training - there was no knowledge of what was working, or why it worked or did not. For example, it was not known how many institutions were able to set up a MTC and maintain it as a functioning committee, nor was it known what had caused some to fail following MTC training in a hospital or SOP support. A better understanding of local circumstances, and ongoing monitoring and evaluation is required to increase the effectiveness and sustainability of interventions. 113 There is evidence that introducing pharmaceutical systems (where they are absent) that serve the CHSs and strong, vibrant umbrella organizations for the church health facilities like CHAs are particularly important (Vogel and Stephens 1989). Additional support could improve the rational use of medicines, MTCs in hospitals and medicines storage and management and would be welcomed in all countries. Some aspects of CHSs are influenced by the government; some by the nature of the individual facility; and some by the fact of being church-owned. This diversity of influences needs to be reflected in interventions. There is need to develop approaches to interventions that are more financially effective and efficient, and include monitoring, evaluation and lesson dissemination. The feedback component of this project was particularly useful in the five primary countries – used to prioritise the actions required to address the gaps identified in the baseline studies. Three key findings emerged from these meetings. Firstly, each country group prioritized interventions differently, so there is no single intervention that can be selected for an overall programme for all countries. However, where there were similar prioritizations, it may be possible to develop a common approach that would work across several countries. Secondly, where results in a particular area for a country were high, issues in those areas were not identified as priorities in the feedback meetings. The chosen priorities always came from areas with the lowest guideline results, but did not correlate exactly with the very lowest scores. This suggests that the results of the baseline study are not a direct indication of what interventions would be prioritized by those who are closest to the issues. The baseline study followed by a local participant workshop gives a better country-specific prioritization. Thirdly, in each country participants quickly recognized that priorities needed to be addressed together as a group. Single-target interventions do not reflect the interrelations between problems. For example, the first priority would probably not succeed unless the other priorities were also addressed. In all five feedback meetings, it was stressed by participants that identified intervention methodologies should all be carried out together – in spite of the fact that most donors and actors prefer to support targeted, single-issue interventions. The need to integrate interventions in a coordinated plan that maximizes impacts through economies of scale, as well as cumulative benefits was stressed. How sustainable are the church health systems in Africa, and internationally? There are vital issues looming, many outside the scope of this investigation into pharmaceutical services. However, these concerns were raised during the course of the study – concerns which impact on access to medicines, but are indicative of a wider problem of survival for CHSs. There is some indication that addressing issues such as fundraising capacities, institutional management capacities and representational voice in dealings with governments will not only support the survival of CHS in general but will also strengthen national health systems, and improve pharmaceutical service delivery and access to medicines. Furthermore we would suggest that integration of church health facilities as part of the national health system and therefore involving them in national planning and strategy implementation would positively impact on the sustainability of CHSs as well as on national health goals and targets. However, in some countries such integration would 114 require changes in thinking from both sides – governments, policy makers and partners financing health, and also the church sector. REFERENCES African Christian Health Association (ACHA), 2010. ACHA Documents and Publications. http://www.africachap.org Annan, I., and S. Essuman. 2005. “Results of Ghana Baseline Survey: Maximizing Access to Essential Medicines for Church Health Services and their Clients.� 1–30. Nairobi, Kenya: Ecumenical Pharmaceutical Network (EPN). AusAid, 2009. “Australian Aid to Health Service Delivery in Papua New Guinea, Vanuatu and Solomon Islands.� Evaluation report. Canberra: Australian Agency for International Development. Ballou-Aares, D., et al. 2008. Private Sector Role in Health Supply Chains. Review of the Role and Potential for Private Sector Engagement in Developing Country Health Supply Chains. The Rockefeller Foundation. http://www.zlc.edu.es/default.aspx?info=00021A. [Accessed 26 January 2010] Banda, M., et al. 2006. Multi-country Study of Medicine Supply and Distribution Activities of Faith-Based Organizations in Sub-Saharan Africa Countries. Geneva: World Health Organization and Ecumenical Pharmaceutical Network. Bennett, S., J. D. Quick, and G. Velásquez. 1997. Public-private Roles in the Pharmaceutical Sector: Implications for Equitable Access and Rational Drug Use. Geneva, Switzerland: World Health Organization. Christian Health Association of Ghana (CHAG). 2008. Annual Report, Accra: CHAG Christian Health Association of Kenya (CHAK). 2008. Annual Report. 2008, Nairobi: CHAK Christian Health Association of Lesotho (CHAL). 2008. 34th Annual General Meeting report. Mophato-oa-Morija: 2008 Christian Health Association of Malawi (CHAM). 2007. Annual Report 2007, Lilongwe: CHAM Coovadia, H., et al. 2009. “The Health and Health System of South Africa: Historical Roots of Current Public Health Challenges.� Lancet 374(9692): 817-34. Earl, S., F. Carden, and T. Smutylo. 2001. Outcome Mapping: Building Learning and Reflection into Development Programs. Ontario, Canada: IDRC. 115 Ecumenical Pharmaceutical Network (EPN). 2004. EPN “Guidelines for Efficient and Effective Pharmaceutical Services.� http://www.epnetwork.org/access-to-essential- medicines-guidelines Kawasaki, E., and J. Patten. 2002. “Drug Supply Systems of Missionary Organizations Identifying Factors Affecting Expansion and Efficiency: Case Studies from Uganda and Kenya.� Geneva: WHO/EDM. Konde Lule, J., S. Mutyaba, and B. Sabiti. 1996. Sustainability of Church Hospitals in Uganda. Kampala: UPMB. MEDS. 2008. Annual Report and Financial Statements. 2008. Nairobi: MEDS Monroe, M. C. 2001. “Evaluation’s Friendly Voice: The Structured Open-Ended Interview.� Applied Environmental Education and Communication Preview issue, Autumn. Olivier, J., and Q. Wodon. 2012. “Playing Broken Telephone: Assessing Faith-Inspired Health Care Provision in Africa.� Development in Practice, 22 (5-6): 819-34. Patel, A., et al. 2009. “Drug Quality in South Africa: Perceptions of Key Players Involved in Medicines distribution.� International Journal of Health Care Quality Assurance 22(5): 547-60. Vogel, R. J, and B. Stephens. 1989. “Availability of Pharmaceuticals in Sub-Saharan Africa: Role of the Public, Private and Church Mission Sectors.� Social Science and Medicine 29(4): 479–86. World Health Organization (WHO). 2007. Everybody’s Business: Strengthening Health Systems to Improve Health Outcomes-WHO’s framework for action. Geneva: World Health Organization. 116 ANNEX: OVERALL GUIDELINE COMPLIANCE SCORES BY BUILDING BLOCK 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% [ Building block 1: Service delivery ] Quality assurance policy in place and implemented 8 23 Disaster preparedness procedures in place Compliance w ith rational use of medicines 'guidelines' 2 Access to medicines f or a f acility passed on to patients 7 [ Building block 2: Health w orkf orce ] At least one pharmaceutically-trained staf f member per 22 17 20 18 f acility Regular assessment of pharmaceutical unit w ork Representation at regional and national levels in relevant debates Pharmaceutical f unction represented at all levels of discussion [ Building block 3: Inf ormation ] 15 19 Access to key pharmaceutical inf ormation Indicators Cross-institutional inf ormation sharing [ Building block 4: Medical products, vaccines & technologies ] Implementation of standard operating procedures f or 6 procurement Functioning medical supply system 3 Compliance w ith best practices f or medicines storage 1 and management 16 Compliance w ith medicines donations guidelines [ Building block 5: Financing ] Pricing policies in place and operationalized 9 12 11 10 Government subsidies extended Pro-poor ethic in evidence Tax exemptions available to CHS [ Building block 6: Leadership and governance ] All 'ow ners' w ith maximum understanding 5 Transparency mechanisms in place in support of 'Health 14 13 f or All' Ef f ective community involvement system in place Functioning Medicines and Therapeutics Committees 4 (MTC) in hospitals 21 Church leaders' aw areness of key messages 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% Percentage compliance Source: Authors. 117 CHAPTER 7 FAITH-INSPIRED HEALTH CARE PROVISION IN GHANA: MARKET SHARE, REACH TO THE POOR, AND PERFORMANCE Jill Olivier, Mari Shojo and Quentin Wodon 45 The World Bank This paper relies on administrative, household surveys and qualitative data to answer three questions about the services provided by faith-inspired health care providers in Ghana, asking: (1) what is the market share of faith-inspired providers as compared to other types of providers; (2) are there differences in market shares among the poor between faith-inspired providers and other types of providers; and (3) how satisfied are patients with the services received and why are patients choosing faith-inspired providers for care? While estimates based on facilities data, especially for hospitals, suggest that the market share of faith- inspired providers is at 30 percent to 40 percent, estimates from household surveys are at less than ten percent. The market share among the poor of faith- inspired providers appears to be similar to that of public providers, but higher than that of private non-religious providers. The qualitative data suggests that the reasons that lead patients to choose faith-inspired providers are not related directly to religion per se, but rather (perhaps indirectly) to the quality of the services provided, including (but not only) through the values of dignity and respect for patients that these facilities exhibit. INTRODUCTION It is commonly accepted that faith-inspired institutions (FIIs) provide a substantial share of health services in sub-Saharan Africa. To substantiate this perception, one would ideally like to have a comprehensive assessment of the scope and scale of all health- related services provided not only by government facilities and faith-inspired providers , but also by private-for-profit providers and other non-religious not-for-profits (NGOs), community-based organizations and initiatives - including division into engagement in particular response such as HIV/AIDs. Such comprehensive overviews are unfortunately not available at this stage. It is nevertheless possible to take one (partial) step towards such comprehensive assessments in specific countries by comparing and interpreting the market share estimates for the health care services provided by various types of providers obtained with both facilities and household survey data, and to measure the facilities’ ‘reach to the poor’ (understood here as a comparative market share assessment of various types of 45 This work benefitted from the support of the Country Management Unit for Ghana, Liberia and Sierra Leone as well as the Human Development Network at the World Bank. The authors are grateful to Ishac Diwan and Rakesh Nangia for supporting this research. 118 providers among segments of the population according to their level of well-being, and especially among the poor). In addition, qualitative work can help reveal the reasons why patients tend to choose one type of provider versus another. The objective of this paper is to do precisely this in the case of Ghana. The main faith- inspired providers 46 in Ghana is the Christian Health Association of Ghana or CHAG, an umbrella organization that coordinates the activities of Christian Health Institutions and Christian Churches’ health programs in Ghana. The federation was founded in 1967 as a Voluntary Professional Association with the assistance of the World Council of Churches, the Catholic Bishops Conference and the Christian Council of Zambia. It is a body through which most of the Christian health facilities liaise with the Ministry of Health to ensure proper collaboration and complementation of government efforts at serving the health needs of the population (Schmid et al 2008, Boateng 2006, CHAG 2003, 2006, and 2008, Dieleman and Hilhorst 2009). CHAG is open to any Christian medical institution recognized as such by the Ministry of Health in Ghana. The founding members are the Ghana Catholic Bishops Conference, the Christian Council of Ghana, and the Ghana Pentecostal Council, while the institutional members are the hospitals and clinics that belong to the founding members and share in the responsibilities and benefits of CHAG. Associate members are other Church-related institutions which share in the aims and objectives of the association and in some limited benefits and responsibilities. The number of CHAG Member Institutions or facilities has grown from 25 in 1967 to 182 in 2011, now including 58 hospitals. The bulk (some 70 percent) of the institutional members are associated with the Catholic Church. A significant event in the relationship between CHAG and the Ministry of Health (MoH) was the government-commissioned Adibo Committee in 1975 which led to public subsidies for the salaries for Ghanaian staff working in mission hospitals. Another watershed moment came in 2003 at the signing of a Memorandum of Understanding (MOU) between CHAG and the MoH. The MOU awards CHAG member hospitals district hospital status and holds them accountable for fulfilling health service contracts, in exchange for which they receive support for salaries and other subventions. An addendum was added to the MOU in 2006, developing administrative instructions for implementation. Today, CHAG is recognized as an Agency with its own personnel by the MoH, parallel to the Ghana Health Service (GHS). In many ways, the collaboration between CHAG and the MoH represents a model that could inspire similar agreements in other countries. Most of the information available on FIIs – and indeed, the basis for most of existing market share assessments – refers to CHAG facilities. There are, however, other faith- inspired providers who are not part of this network – although little is known about them. For example, among Islamic providers, the Ahmadiyya Muslim Mission (AMM) operates six hospitals in Kaleo (Upper West), Techiman (Brong Ahafo), Asokore and Kokofu (Ashanti), Swedru (Central), and Daboase (Western). The Ahmadiyya Muslim community also provides other health-related services including clinics, medical aid 46 In Ghana, faith-inspired providers are also commonly called ‘mission-based providers’ – and faith-inspired facilities are frequently classified as ‘mission hospitals’ – leading to some terminological challenges. 119 programs, and short-term health delivery projects (Salisu and Prinz 2009, Makinen et al 2011, Miralles et al 2003, Samwini 2006) as well as homeopathic medicine. Ahmadiyya appears to have few contacts with or support from the government, even if some of its hospitals function as district hospitals (Miralles et al 2003). Other Islamic initiatives are mentioned occasionally in the literature, such as a clinic run by the Islamic Republic of Iran (ICRO in Salisu and Prinz 2009). Yet overall, these other faith-inspired health care providers, whether Christian or Islamic, are small in comparison with those under the CHAG federation. Given this context, we focus here on facilities associated with CHAG, and the purpose of this paper is to rely on administrative, household surveys, and qualitative data to answer three specific questions about the services provided by FIIs in Ghana: (1) what is the market share of faith-inspired providers as compared to other types of providers (section two of the paper); (2) are there differences in market shares among the poor between faith-inspired providers and other types of providers (this is also discussed in section two); and (3) how satisfied are patients with the services received, and why are patients choosing faith-inspired providers for care. The main conclusions are that estimates of the market share of FIIs from household surveys are at less than ten percent, versus 30 percent to 40 percent when using partial information on facilities data. The market share among the poor of FIIs appears to be similar to that of public providers, but higher than that of private non-religious providers. The reasons that lead patients to choose faith- inspired providers are not related directly to religion, but rather to the quality of the services provided, including values of dignity and respect for patients that the facilities exhibit. MARKET SHARE AND REACH TO THE POOR Estimates based on facilities data In Ghana, the most commonly cited estimates of the market share of faith-inspired providers are based on hospital beds. Estimates from the Ministry of Health suggest that CHAG facilities provide 28 percent of all hospital beds. The total share of hospital beds among FIIs is 29.3 percent when Islamic facilities are added. However, the data on which these estimates are based are incomplete. Makinen et al (2011) found that data for private for-profit providers was particularly absent and that the data on the CHAG facilities was the exception, based on the stronger collaborative relationship between CHAG and the MoH. The other facilities tend to operate independently of the MoH, and are thus not included in current data gathering. This implies that the market share for both CHAG and public facilities tends to be overestimated, although it is not possible at this stage to assess how substantial this overestimation might be. Table 1 provides a summary of other estimates of the share of faith-inspired providers in Ghana, as discussed in the literature. Most estimates of the market share of FIIs are in the 30 percent to 40 percent range. Some of the estimates are based on hospital beds, but others rely on outpatient care and the consumption of pharmaceuticals among others. For example, the Ecumenical Pharmaceutical Network (Annan and Essuman 2005) estimated on the basis of a survey on the consumption of pharmaceutical products by different types 120 of facilities that faith-inspired providers accounted for about 40 percent of total pharmaceutical consumption. Nimo and Wood (2005) suggested similarly that FIIs served around 40 percent of the population, supplying an estimated 30 percent of beds and 35 percent of outpatient care (quoted in Rasheed 2009). Marek et al (2005) suggested that 37 percent of inpatient admissions in hospitals (Accra excluded) were provided by CHAG, and CHAG has provided evidence that its hospital occupation rates tend to be higher than those of government hospitals (CHAG 2006). Previous statements and studies (see for example Miralles et al 2003) have yielded similar results. A few other studies, including older ones, are also included in Table 1, again with similar results. As a result of these various estimates, the MOU signed between CHAG and the government of Ghana explicitly stated that, “This collaboration recognizes the pivotal role of the private health sector, which provides about 42 percent of Ghana’s health care services and has been growing rapidly in recent times, as the engine of growth in the country’s socio- economic recovery programme� (Ghana-MoH and CHAG 2006). Table 1: Selection of stated market share estimates for faith-inspired care in Ghana Descriptive unit Estimate Source Mission hospitals 50% outpatient care; >25% beds nationally, Bradley (in De Jong 1991) 46% in six under-privileged Northern regions Mission hospitals (A third) 33% beds World Bank 1993 Church 25% beds and 40% population served Robinson and White 1998 Church hospitals 34% medical work Matomora 1995 CHAG Approx. 40% national health service Green et al 2002 Catholic 27% share health care Annan and Essuman 2005 Other Christian 11% share health care Muslim 1-2% share health care Mission facilities 40% pop., 30% beds, 35% outpatient care Nimo and Wood 2005 (in Rasheed 2009) CHAG 35-40% national health care CHAG 2006 CHAG 25% NH Services Dimmock 2007 Christian Health ~34% of NHS (national health sector) Chand and Patterson 2007 Networks NFBHN (CHAG) ~34% NHS Schmid et al 2008 Christian Health Services 40% NHS Rookes 2010 Source: Compiled by authors There is thus apparently considerable convergence in the existing estimates in the literature on Ghana, but this is not too surprising given that to a large extent, the estimates reflect the same reality – that is, data based on out-patient care as well as the consumption of pharmaceutical products tend to be very closely related to estimates of hospital beds because out-patient care and pharmaceuticals are used primarily by hospitals. The question is whether these estimates capture a large enough share of the total delivery of health care in the country. We would argue that interpreting the data in table 2 as valid estimates of the share of services provided by FIIs within the whole health system of the country is problematic for three reasons. First, within formal care delivery mechanisms, a large share of health services are provided by other types of facilities than hospitals, such as clinics and health centers, as well as maternity homes and facilities from the government-run community-based health 121 planning and services which are primary health care focused services sometimes with mobile units. CHAG members primarily provide services through hospital facilities. For the year 2008, the share of district and other hospital facilities operated by FIIs was at 16.3 percent. But the share of all types of clinics operated by FIIs was much lower at 9.1 percent, and if one adds maternity homes and community-based health planning and services, the share of FIIs in all non-hospital facilities fell further to 5.5 percent. Overall, the share of FIIs in all types of facilities listed was at 6.6 percent. An analysis by Kissah- Korsah (2008) of more than 2,163 health institutions including most if not all of the CHAG facilities (the author identified 180 Christian facilities) suggests similarly that 53.5 percent of all facilities were governmental or quasi-governmental, with 38.0 percent being private non-governmental, and only 8.5 percent Christian facilities. Estimates of the market share of faith-inspired providers based solely on the number of facilities could also lead to underestimating the role of faith-inspired providers if only because hospitals tend to serve more patients per facility, but this still helps to put statistics on the number of beds into broader perspective. A second issue with the reliance on statistics on hospital beds, pharmaceuticals, outpatient care and for that matter also the number of facilities is that a large share of health care is provided by other types of providers that are not included in such statistics. At least two different groups must be mentioned here. First, Ghana has a significant traditional sector that often operates alongside orthodox biomedical care, for example with patients mixing plural health-seeking modalities (Kissah-Korsah 2008, Van den Boom et al 2004). While studies on religion and health-related behaviors recognize the role of traditional practices, this is rarely addressed in the literature on FIIs. Some have suggested that in the roughly 25 years since the introduction of ‘on the spot’ payment for health delivery, more than half of the country’s patients have turned increasingly to traditional health care and self-medication (Van den Boom et al 2004, Salisu and Prinz 2009). Second, self-medication has also been noted to be a significant practice – given the limited availability of doctors and pharmacists. There has been a trend towards the use of services provided by chemical stores especially. Self-medication has many potential dangers, including in terms of consumption of leftover and often expired drugs as well as untrained chemical sellers taking experts roles (Van de Boom 2004, see also Ballou-Aares et al 2008). Data from the fifth round of the Ghana Living Standards survey (GLSS5) implemented in 2005-2006 help document the role of traditional healers and chemical stores. Nationally, it turns out that hospitals account for 31.6 percent of all consultations, followed by clinics that account for 28 percent of consultations. Both maternity homes and pharmacies have small markets shares. Traditional healers do not appear explicitly in this cut of the data, but their market share is somewhat limited (this group shows up primarily in the categories of providers identified as consultant’s home, patient’s home, and other.) By contrast, the role of chemical stores is very large, as they account for 29.4 percent of consultations, and an even larger share of consultations among the bottom quintiles (i.e. poorer segments) of the population, as is the case for other non-formal mechanisms of care delivery. In other words, on the basis of the types of care used in the country according to the GLSS5 data, the high market share of faith-inspired providers in terms 122 of hospital services would be somewhat diluted when considering a broader definition of the health sector. Thirdly, it has been noted in many other neighboring countries, that faith-inspired as well as other community-based organizations tend to be engaged in a range of activities that stretch beyond formal health services. In the case of FIIs, this might be as a result of a ‘holistic’ focus on health – or because the FIIs (such as CHAG members) are commonly tied into broader developmental activities as a result of their ties to local communities and denominational bodies. Certainly in Ghana, the FIIs are involved in a range of ‘development’ activities, such as micro-financing and sanitation (De Jong 1991, Schmid et al 2008), and less formal community-based organizations play a large role in specific areas such as HIV and AIDS. But data on those community-based efforts are rarely available. All this to say that it is difficult to assess the scope and scale of community- based organizations and networks’ health-related activities based on simplistic estimates of hospital bed market shares, or for that matter based on traditional household surveys. Estimates based on household surveys Market share estimates can also be obtained from household surveys, such as the GLSS5. The categories of providers in the GLSS5 are a bit different. In that survey, a first question is related to the type of facility used by households. A second question asks whether the facility is public, private-religious, or private non-religious. The market share of faith-inspired obtained from the survey is provided in table 2. It is estimated at close to 7 percent, almost reaching 8 percent when one excludes traditional healers and chemical stores. This is of course much lower than what is obtained on the basis of facilities data. What could explain such different results? Much of the difference is likely to be as a result of the universe of health care being considered. If hospitals account for less than a third of all consultations in the GLSS5 data, assuming that hospital beds or outpatient care are good proxies for the overall supply of care of hospitals (which itself would be a strong assumption, given that a large share of hospital care does not necessarily require hospital beds), a hospital bed market share for faith-inspired providers of a third might be diluted into a market share of about 10 percent for health care as a whole (or slightly more when considering faith-inspired clinics as well) when a broader universe of care is taken into account, as is done in the surveys. This suggests that the household survey based estimates of market share may not be completely out of sync with the reality on the ground. Still, even then the survey-based estimates look small against the current wisdom of those working on the ground – for example those in the MoH or CHAG. An additional explanation might be that the identification of faith-inspired providers by households is partial only, with some households considering faith-inspired facilities as either private non-religious facilities, or with more likelihood as public facilities, especially when mission hospitals are considered as district hospitals and accordingly funded by the government. 123 Table 2: Share of patients by type of provider used, 2005-2006 GLSS5 (%) Residence Area Welfare Quintile Total Accra Other urban Rural Q1 Q2 Q3 Q4 Q5 Including chemical, traditional providers, etc. Public 52.0 44.8 43.2 46.0 43.6 43.5 44.2 45.1 44.4 Private religious 3.9 7.8 6.4 7.0 5.5 7.1 6.6 6.8 6.6 Private non-religious 44.1 47.5 50.5 47.0 50.9 49.4 49.3 48.2 49.0 Total 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 Excluding chemical, traditional providers, etc. Public 55.1 64.0 63.2 69.2 65.3 62.8 64.3 57.2 62.5 Private religious 4.2 8.7 8.0 6.3 6.4 8.9 7.6 8.4 7.8 Private non-religious 40.8 27.3 28.9 24.6 28.3 28.3 28.2 34.4 29.8 Total 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 Source: Authors’ estimation using GLSS5 2005-2006 survey Apart from this stirring up the debate on market share, is the question of whether some types of providers reach the poor better than others, as this relates directly to efforts by the government to adapt the entire health system so that it better serves rural and marginalized areas. Here again, the common perception is that faith-inspired providers play a special role vis-à-vis the poor, in part because of their preferential option for the poor, and its corollary of faith-inspired services and facilities being physically located in marginalized areas. As CHAG (2006) put it, “CHAG members cater for an estimated 35- 40 percent of the national population, mainly in the hard to reach rural places in Ghana.� Or as De Jong (1991) argued, faith-inspired services tend to be “…particularly represented in poorer, more remote areas out of commitment to serve the under privileged (e.g., religious missions often state this explicitly) or because they can fill a gap in such areas not already met by government services. In Ghana, for instance, while missions provide 25 percent of total hospital beds in the country, they provide about 46 percent of beds in the six under-privileged northern regions (Bradley 1989 in De Jong 1991)…Missions tend to be disproportionately represented in the less privileged areas.� The same argument is made almost thirty years later by Ballou-Aares et al (2008) who suggest that “the public sector channel is most active in more densely populated areas, which are also relatively more affluent. The faith-inspired channel is an important source of health care to less affluent or poor people in far-flung areas…geographic access for less affluent people in rural areas is a problem with access being limited to the southern part of Ghana.� Serving marginalized areas was certainly the intention of many faith- inspired providers when facilities were set up. However, current evidence on whether faith-inspired facilities today still reach the poor proportionately more than other types of providers is thin, as this is difficult to establish relying solely on facilities data limited for the most part to hospitals and clinics. Because household surveys include data on the socio-economic characteristics of households, they can be used to assess which facilities reach various segments of the population classified according to their level of well-being. In the GLSS5, well-being is measured according to the level of consumption per equivalent adult. Importantly, even if some faith-inspired providers are misclassified in the surveys, to the extent that the likelihood of such misclassification is similar for all faith-inspired facilities, this should 124 not affect substantially estimates of the extent to which various types of facilities reach the poor. Consider the case where faith-inspired providers serve the poor more than other providers in the specific sense that the share of their services obtained by the poor is higher than is the case for other providers. If some faith-inspired facilities are misclassified by households as public facilities in a quasi-random way (the probability of misclassification is similar for all faith-inspired facilities), then the share of the beneficiaries that are poor in faith-inspired facilities would not be affected. As for public providers, the erroneous inclusion of some faith-inspired providers in their pool would lead to a higher share of beneficiaries of public facilities identified as poor than warranted, but the bias should be small because the number of faith-inspired facilities misclassified as public facilities would be small as a proportion of the total number of public facilities. This is because the market share of faith-inspired facilities is smaller than that of public facilities, and because only a subset of faith-inspired facilities would be misclassified. In addition, if it turns out that the profile of beneficiaries according to level of well-being is similar between faith-inspired and public facilities, the bias would even be smaller. The data in table 2 suggest relatively few differences between public and faith-inspired facilities in terms of whether they reach various segments of the population according to level of well-being, but substantial differences with private non-religious facilities, as would be expected given that these facilities tend to be more expensive. Specifically, according to the GLSS5 data, faith-inspired providers do have a higher market share in rural than in urban areas. As to whether faith-inspired providers have a higher market share among the poor, the proportion of patients relying on faith-inspired providers seems slightly lower in the bottom than in the top quintiles (as a proportion of the overall demand for care among those groups), but the differences are not very large. Thus, it would be fair to suggest that according to the results from the survey, faith-inspired providers do not necessarily have a proportionately higher market share among the poor than among better off households than is the case for the government (the public provider) – but of course, faith-inspired providers do reach patients from poor households. It must be noted that the estimates provided in table 2 are estimates of the market shares of various providers within population groups, such as households grouped according to their level of consumption. The estimates show that the market share of FIIs in the various quintiles is similar across quintiles, as is the case for public facilities. This can be interpreted as showing that FIIs and public facilities comparatively reach the poor in a similar way. But it does not mean that both types of facilities serve the poor and better off households equally. As discussed in Coulombe and Wodon (2012a) in their benefit analysis of public health spending in Ghana, households from the top quintiles of consumption benefit from a much larger share of the health services provided by both public and FIIs facilities than households in the bottom quintiles. That is, both public and FIIs facilities serve the poor less well than they serve the better off (see also Coulombe and Wodon 2012b on the geographic location of FIIss). 125 PERFORMANCE Existing studies Detailed comparative data between faith-inspired-, public-, and other private providers of health services are for the most part not available in Ghana. It is most difficult to compare FIIs against the other private providers such as other NGOs or for-profit providers. Makinen et al (2011) have recently completed a mapping assessment of the private health sector in Ghana, and note that there is still no comprehensive data on the size and configuration of the private sector, with the exception of CHAG – whose closer relationship with the MoH means that there is slightly more known about CHAG than the other private providers. Certainly there is not nearly enough available data to embark on a more comprehensive comparative of equitable access to health care (availability, affordability and acceptability). What remains then are stand-alone comparatives, providing individual pieces to the broader puzzle that is the Ghanaian health system, and one of the objectives of the papers in this collection is precisely to start conducting more in-depth work on comparing public, private religious, and private non-religious health care providers. On the basis of admittedly incomplete data, it can be seen that based on some measures FIIs seem to be doing slightly better than public providers, while in other areas they may be doing slightly less well. As a first example of comparative analysis, consider the statistics provided in Table 3. The table shows that CHAG district hospitals had on average higher occupancy rates in their facilities than public hospitals. This might reflect a preference on the part of the population in using FIIs (as also suggested by Shojo et al 2012), better management, or simply location advantages. While location advantages tend to be historically driven, other explanations would suggest higher performance by CHAG in this specific area. Table 3: Beds and Occupancy Rates in CHAG and Public District Hospitals, 2006 Western Central Greater Volta Eastern Ashanti Brong Northern Upper Upper Accra Ahafo East West Number of Beds CHAG 474 387 67 967 930 1,084 1,119 324 253 336 Government 756 678 683 1,137 1,180 932 252 340 575 493 Occupancy Rates (%) CHAG 74.9 61.9 43.2 56.6 51.4 57.5 57.8 53.1 42.0 Government 46.5 44.0 47.8 43.8 44.2 44.5 62.5 57.7 45.0 46.1 Source: CHAG (2006). As a second example of comparative assessment, consider the data on the pricing of medicines provide in Table 4 from META Ghana (2009). The data suggests that the prices of various drugs tend to be slightly higher in the rural faith-inspired sector than in the rural public facilities. For four out of seven drugs, the prices in FII facilities were more than ten percent higher than those in public facilities. For two of the drugs, the prices were more than ten percent higher in public facilities, while for the last drug, the differences in prices were below ten percent. This suggests on average that drug prices in FIIs are slightly higher in FIIs than in public facilities. This also appears to extend to patient prices more generally. Miralles et al (2003) suggested that patient prices in FIIs 126 were on average twelve percent higher than in the public sector. Makinen et al (2011) find that out-of-pocket spending by patients at CHAG facilities is higher than at private and public providers. However, they also note that “lower price as a provider attribute is most frequent among CHAG providers� – suggesting that while out-of-pocket costs might be higher at CHAG facilities, the public perception may well be that they are lower. This is discussed in more detail in the Shojo et al (2012), but it is important to note that this may have changed in recent years with the adoption of the National Health Insurance Scheme which has substantially reduced out-of-pocket costs overall. For the most part, comparative based only on the cost for patients of pharmaceuticals suggests slightly better performance for the public sector (in terms of lower costs for patients), although some of the broader issues that affect the sector as a whole should not be obscured by the somewhat limited differences between FIIs and public facilities. Indeed, it has been suggested that in general, the price of pharmaceuticals may be too high in Ghana. The MoH, as cited in Salisu and Prinz (2009) stated that “A comparative study of medical procurement by the MoH and the faith-inspired health sector…showed that it would take approximately four wage days for a person to purchase medicines from the private sector and almost three from the mission sector…the study also found out that beside many complications such as change of prices with change of deliverer, both government and missions often had to buy at more than double the market procurement prices� (Salisu and Prinz 2009). Table 4: Comparison of drug prices by sector (2007/2008) Inter- Rural Mission Rural Public Rural Private NHIS national Sector Sector Sector reimbursement reference (median price) (median price) (median price) prices Price (GHC) Medicine IRP Median Ratio Median Ratio Median Ratio Reimbursed Ratio name Price to Price to Price to prices 2008 to IRP (GHC) IRP (GHC) IRP (GHC) IRP (GHC) Ciprofloxacin 0.0292 0.2000 6.84 0.1750 5.98 0.1700 5.81 0.2 6.84 Clotrimazole 0.0077 0.1650 21.51 0.0584 7.61 0.1075 14.02 0.16 20.86 Diclofenac 0.0055 0.0350 6.40 0.0400 7.32 0.0300 5.49 0.1 18.30 Mebendazole 0.0156 0.4750 30.39 0.3500 22.40 0.7000 44.79 1.2 76.79 Phenytoin 0.0048 0.0800 16.69 0.1000 20.86 0.06 12.52 Quinine 0.0768 0.1175 1.53 0.2500 3.26 0.1250 1.63 0.28 3.65 Injection Ranitidine 0.0229 0.1200 5.24 0.1250 5.46 0.1000 4.36 0.2 8.73 Source: META Ghana (2009). As a third example of analysis aiming to compare the practices and performance of FIIs versus public and other private facilities, Annan and Essuman (2005) use a baseline survey carried out as part of a project to improve access to essential medicines for FIIs and their clients. The study found that the overall trend for FIIs in Ghana was one of improvement. They note Ghana has a well-functioning drug supply system, and relatively high numbers (at least in comparison with other African countries) of pharmacists, pharmaceutical technicians, and pharmaceutical assistants per hospital. Yet the study also revealed differences between FIIs and other providers. The study suggested that at a 127 national scale, FIIs tended to have lower drug prices for patients than other providers (the difference from the studies mentioned earlier may be related to the fact that the study by Annan and Essuman was national in scope), they also tended to rely on drug sales revenues more than other providers to cover their costs, and they tended to have more difficulties in managing their stocks of drugs. FIIs also did not have established rational drug use policies. The above three studies on comparative performance suggest some differences between FIIs and public facilities, but these differences tend to be small, and not always consistent between studies. It is likely that with the further integration of CHAG facilities in the national health system that has taken place over the last few years, some of the differences between FIIs and public facilities may have been reduced further. What is clear though is that both FIIs and public facilities struggle with common problems such as a shortage of qualified staff, a high cost of inputs, limited cost recovery which affects financial sustainability, and increased competition from other private providers. Such competition is observed not only for patients, but also for staff. Yet before turning to these problems, it is necessary to consider the relationship between CHAG and the government, and how it has evolved, especially during the last decade. Other existing studies Recent qualitative data collected by Shojo et al (2012, in the second volume of this collection) from fieldwork among staff and patients in six FIIs in two locations in 2010 provide more insights into the performance of FIIs, as well as the reasons why patients rely on their services. A few results from that study are worth summarizing here. The data revealed that patients using FIIs were satisfied with the quality of staffs, hygiene in the facilities, and cost (which are now lower thanks to the NHIS), but less so with the availability of proper accommodation, technical equipment, and medicines. The situation was more difficult for clinics and hospitals not yet accredited with the new NHIS (National health Insurance Scheme) being implemented since 2004, probably in part because this resulted in higher costs for patients. More precisely, quality of care was the main reason for choosing facilities according to the qualitative fieldwork. Among patients in Christian clinics/hospitals, two thirds said that quality was the main reason for choosing the clinic/hospital, and close to sixty percent mentioned that workers are skilled, knowledgeable, competent, dedicated, and patient; in short they appreciated the quality of the staff. For patients in Islamic clinics/hospitals, the most common answer was the quality of workers followed by the quality of service, with location coming third. Respect for patients came in strongly as a key reason for choosing FIIs. “Here we are treated with respect. They listen to us well and understand all of our problems. They take their time to talk to us in a polite way. You don't regret spending your money at this hospital. Even if they don't have all the equipment, the way they handle makes me feel comfortable� (Female Muslim patient, Islamic clinic); “I have heard that they are a top quality hospital and they are very serious with their work and they treat patients with care and respect� (Male Christian patient, Christian hospital). 128 To get at the reasons for choosing providers differently, patients were asked to share the advantages that they see in using FIIs. In Christian facilities a third of patients cited “quality of workers� as the main advantage of the facilities, followed by “assistance for the poor� (25 percent of respondents) and “quality of service� (19 percent). Among patients in Islamic facilities, the most common answer was “worker’s skills and quality� (44 percent) followed by “location� (31 percent). Two other reasons were mentioned: “Assistance for the poor/orphans� and “quality of service� by 12.5 percent of respondents. The availability of assistance for the poor, while not a leading criterion for the choice of provider, was also mentioned by facility staff. As a Director at an Islamic clinic explained, “What is the target population of this clinic? Elders come, youth come, children come, and pregnant women come… any kind of category. The majority of people who come to this clinic are Moslem, but we have non-Moslem too. They are Christian or believe traditional religion. Also we have both poor and somehow middle income group. Majority of the patients are actually poor. That is one of main reason of establishment of this clinic. People are facing financial problems, unemployment and deprivation. Their monthly income is low. We try as much as possible to subsidize our services.� By contrast, few patients mentioned religion as determinant or deterrent for choosing FIIs. This is apparent in answers provided to questions shown in table 5, and it emerges also from interviews: “I am Christian but came to this Islamic clinic not because of my religious beliefs but because the clinic works well� (Female Christian patient, Islamic clinic); “I will seek health care from even a Christian health facility if that is of high quality but not go to a traditional priest� (Male Muslim patient, Christian clinic); “My religious beliefs do not affect my choice of health care for me and my family. I am Moslem and I have been attending a Catholic clinic in the past, so religion doesn't matter to me. Any clinic where I can receive effective medical care, I will go� (Male Muslim patient, Islamic clinic). Most patients had no problem in using services provided by FIIs from a different faith: “If they will take good care of me to get well, I don't care what faith is behind them� (Male Christian patient, Christian hospital); “I use Islamic clinic here even though I am Christian because I believe that it is providing gravity health care and not about changing me to Moslem� (Male Christian patient, Islamic clinic). When patients mentioned values or faith, this was done typically in general terms and in a positive way: “As an Islamic community this clinic is seen as a good model of what Islam can do for Moslems. It is providing health care as well as spiritual care for the people� (Male Muslim patient, Islamic clinic); “They try to increase the faith of patients who come to this clinic, so it is good. It boosts the moral of patients and increases their faith. Even though I am Moslem, I like it so much� (Male Muslim patient, Christian clinic). 129 Table 5: Patients’ values and choice of health care provider, 2010 Patients who use a Patients who use a clinic that belongs Questions clinic that belongs to a different to the same religion religion Do your religious beliefs and values affect your choices Yes: 0 % Yes: 10.8% regarding healthcare? Are you willing to use health care services at a clinic which is Yes: 100% Yes: 89.1% grounded in a faith different from your own? Do you think that the health clinic/hospital should provide Yes: 18.1% Yes: 33.3% spiritual guidance and counseling to the patients? Source: Shojo et al. (2012). The desire to serve communities as a whole without specific reference to faith of patients also emerged from interviews with FII staff. “We serve all mankind. We accept patients who belong to different religion. The vision of the national catholic health services is to provide high quality health care in the most effective, efficient and innovative manner, specific to the needs of the communities we serve and at all times acknowledging the dignity of the patient� (Director of a Christian facility). “There was no clinic around here before. We established this clinic to assist poor community in this area. Most of the people in this area are Moslem, but our target population is entire community. We accept everyone…Personally I am Christian, but I am working at Islamic clinic as a doctor. I don’t care the patients' religion. Whatever they believe, we are fighting for our own goal to support the people's health� (Doctor at an Islamic clinic). CONCLUSION The objective of this paper was to answer three questions. Firstly, what is the market share of FIIs as compared to other types of health care providers in Ghana? Existing estimates based on administrative data on hospital beds suggest that FIIs account for 30 to 40 percent of health care provision, but estimates from household surveys are at less than 10 percent. Existing estimates are likely to be biased upward because they rely on hospital beds (or related measures such as outpatient care and the consumption of pharmaceuticals), a sub-sector where faith-inspired providers have traditionally been especially active. But this ignores large segments of health care provision, including smaller health facilities, traditional healers, as well as pharmacists and chemical stores, some of which have grown substantially in recent times. Estimates of the market share of FIIs based on household surveys may well be underestimated, in part because some faith- inspired facilities may have been considered by households as public facilities given that in many districts, faith-inspired hospitals serve as district hospitals. The second question was: do FIIs reach the poor proportionately more than other types of provider? The household survey data suggests that the market share of FIIs among the poor is similar to that of public providers, and higher than that of private non-religious providers who tend to have a higher market share among better off segments of the population. Thus faith-inspired and public providers appear to be serving the poor roughly equally to public providers, while private providers tend to serve more the higher socio-economic groups than either faith-inspired or public providers. This does not mean 130 however that FIIs and public providers serve all segments of the population equally. Most health care facilities continue to serve the better off more than they do the poor, in that the share of services received by poorer segments of the population is much smaller than their population share. The third question was about the level of satisfaction of patients with the services received from FIIs, and the motivation for choosing faith-inspired providers. Qualitative data collected in 2010 in six facilities by Shojo et al (2012) suggest that the satisfaction with the services received in faith-inspired facilities is high, including in areas such as respect paid to patients. Subjective satisfaction does not measure quality per se, but it is an important indicator and it appears indirectly from the qualitative data that faith- inspired facilities may have a comparative advantage at least in terms of the attention paid to patients. More data would be needed to confirm this, but it is encouraging for faith-inspired facilities. It also appears that faith-inspired facilities try to help the poor afford the cost of care. Finally, and this is also related to the question of quality, religion itself does not seem to be a direct factor for the choice of faith-inspired facilities. Many patients use services from clinics and hospitals that are affiliated with a different faith from their own, and the main reason for the choice of facility is precisely the perception that they provide services of quality. This of course raises questions about the indirect impact of religion or faith on quality of care. REFERENCES Annan, I., and S. Essuman. 2005. “Results of Ghana Baseline Survey: Maximizing Access to Essential Medicines for Church Health Services and their Clients.� Nairobi, Kenya, Ecumenical Pharmaceutical Network EPN. Ballou-Aares, D., A. Freitas, et al. 2008. Private Sector Role in Health Supply Chains: Review of the Role and Potential for Private Sector Engagement in Developing Country Health Supply Chains. The Rockefeller Foundation, Dalberg and MIT Zaragoza. Boateng, J., H. Surnye, et al. 2006. “Costs of Reproductive Health Services Provided by Four CHAG hospitals.� Frontiers in Reproductive Health Program and Family Health International. Christian Health Association of Ghana (CHAG). 2003. Annual Report: 2003. Accra: Christian Health Association of Ghana. _____. 2006. Annual Report: June 2005–May 2006. Accra: Christian Health Association of Ghana. _____. 2008. Annual Report: 2008. Accra: Christian Health Association of Ghana. Chand, S. and J. Patterson. 2007. Faith-based Models for Improving Maternal and Newborn Health. Baltimore: USAID-ACCESS. 131 Coulombe, H., and Q. Wodon. 2012a. “Benefit Incidence of Public Health Spending in Ghana: Comparing Public and Faith-inspired Providers.� In J. Olivier and Q. Wodon, eds, Mapping, Cost, and Reach to the Poor of Faith-inspired Health Care Providers in sub-Saharan Africa. HNP Discussion Papers, World Bank, Washington D.C. _____. 2012b. “Mapping Religious Health Assets: Are Faith-inspired Facilities Located in Poor Areas in Ghana?� In J. Olivier and Q. Wodon, editors, Mapping, Cost, and Reach to the Poor of Faith-inspired Health Care Providers in sub-Saharan Africa. HNP Discussion Papers, World Bank, Washington, DC:. De Jong, J. 1991. “Nongovernmental Organizations and Health Delivery in Sub-Saharan Africa.� Working Paper, Population and Human Resources Department, World Bank, Washington, DC. Dieleman, M., and T. Hilhorst. 2009. Quest for Quality: Interventions to Improve Human Resources for Health among Faith-based Organisations. CORDAID and Royal Tropical Institute, The Hague, The Netherlands. Dimmock, F. 2007. “Faith-Based Health Networks in Africa.� Presentation at the Christian Connections for International Health Annual Conference, Washington, DC. Ghana-Minstry of Health (MoH). 2006. “Medicine prices in Ghana: A comparative Study of Public, Private and Mission Sector Medicine Prices.� Accra, Ghana Ministry of Health, Health Association International Kenya, and the World Health Organization. _____. 2010. “Independent Review Health Sector Programme of Work.� 2009, Accra, Ghana Ministry of Health MOH. Ghana-Ministry of Health (MoH) and Christian Health Association of Ghana (CHAG). 2006. “Memorandum of Understanding and Administrative Instructions.� Accra, MOH and CHAG. Green, A., J. Shaw, F. Dimmock, and C. Conn. 2002. “A shared mission? Changing relationships between Government and Church Health Services in Africa.� International Journal of Health Planning and Management 17:333-53. Kissah-Korsak, K. 2008. “Spatial Accessibility to Health Care Facilities in the Ajumako- Enyan-Essiam and Upper Denkyira Districts in the Central Region of Ghana.� Norsk Geografisk Tidsskrift - Norwegian Journal of Geography 62: 203-09. Makinen, M., S. Sealy, R. A. Bitran, S. Adjei, and R. Munoz. 2011. Private Health Sector Assessment in Ghana. World Bank, Washington, DC. Marek, T., C. O’Farrell, et al. 2005. Trends and opportunities in public-private partnerships to improve health service delivery in Africa, Africa Region Human Development Working Paper Series, World Bank, Washington, DC. 132 Matomora, K. S. M. 1995. The District Health Systems Approach. Contact 143:7-14. META Ghana. 2009. META Ghana CSO Presentation at META Ghana Forum, 17 December 2009. Miralles, M. A., F. Aboagye-Nyame, et al. 2003. “Access to Essential Medicines: Ghana.� Prepared for the Strategies for Enhancing Access to Medicines Program, Arlington, VA, Center for Pharmaceutical Management, Management Sciences for Health. Shojo, M., C. Tsimpo, and Q. Wodon. 2012. “Satisfaction with and Reasons for Choosing Faith-Inspired Health Care Providers in Ghana.� In J. Olivier and Q. Wodon, eds, The Comparative Nature of Faith-inspired Health Care Providers in sub-Saharan Africa, HNP Discussion Paper, World Bank, Washington, DC. Rasheed, F. 2009. “Collaborations Between Public Health Systems and Faith-Based Organizations: Memorandum of Understanding Between the Ministry of Health and the Christian Health Association, Ghana.� Programme on Partnerships and UN Reform, Geneva: World Health Organization. Robinson, M., and G. White. 1998. “The Role of Civic Organisations in the Provision of Social Services: Towards Synergy.� In G. Mwabu, C. Ugaz, and G. White, eds, New Patterns of Social Provision in Low Income Countries, Oxford: Oxford University Press. Rookes, P. 2009. “Commitment, Conscience or Compromise: The Changing Financial Basis and Evolving Role of Christian Health Services in Developing Countries.� University of Birmingham. Salisu, A., and V. Prinz. 2009. Healthcare in Ghana. Wien, Austrian Centre for Country of Origin & Asylum Research and Documentation, ACCORD and Austrian Red Cross. Samwini, N. 2006. The Muslim Resurgence in Ghana since 1950: Its Effects Upon Muslims and Muslim-Christian Relations. Piscataway: Transaction Publishers. Schmid, B., E. Thomas, J. Olivier, and J. R Cochrane. 2008. The Contribution of Religious Entities to Health in sub-Saharan Africa. Cape Town: African Religious Health Assets Programme. Van den Boom, G. J. M, N. N. N. Nsowah-Nuamah, and G. B. Overbosch. 2004. “Healthcare Provision and Self-medication in Ghana.� Amsterdam: Working Paper Series, Centre for World Food Studies. World Bank. 1993. World Development Report 1993: Investing in Health. World Bank, Washington, DC. 133 CHAPTER 8 APPROACHES FOR INTEGRATING PRIMARY HEALTH CARE IN REPRODUCTIVE HEALTH PROGRAMS IN UGANDA Lorna B. Muhirwe Uganda Protestant Medical Bureau This paper, written by the Executive Director of the Uganda Protestant Medical Bureau (UPMB), describes the context in which UPMB operates - in particular approaches for aligning religious health services with national approaches to reproductive health intervention through improved primary health care (PHC). Uganda entered into the primary health care arena rather late due to the political upheavals of the late 1970s and early 1980s. By the time Uganda embraced primary health care the public health system was in a terrible state and health care was provided mainly by the religious non-governmental sector now called the private-not-for-profit sub sector. In the early 1980s, UPMB revised its constitution in response to the Alma Ata declaration of 1979 to pave the way for Primary Health Care Programs and Community Based Health Care activities. INTRODUCTION Uganda is a small land-locked country in East Africa with a total area of 236,040 square kilometers. The population is estimated at 30 million (2009), the third fastest growing population in the world at a rate of 3.2 percent per annum. Large families tend to be much poorer than smaller families, in part because of limited access to land and asset shortages. Despite achievements in economic growth, in comparison to countries with similar levels of development Uganda remains an outlier with regard to reproductive health indices. Maternal and infant mortality and morbidity, though declining, remain unacceptably high, with the maternal mortality rate (of 435 deaths per 100,000 live births) translating to about 6000 women dying every year due to pregnancy-related causes. The infant mortality rate is 67 per 1000 live births and neonatal mortality rate is 29 per 1000 live births; 89 percent of deliveries occur in rural areas and only 37.7 percent of deliveries in rural areas are assisted by skilled birth attendants (compared to 80.4 percent in urban areas.) In 1978, the World Health Organization (WHO) led an international consensus on the need to change global health policy. That year, Health Ministers from around the world agreed on the launch of the primary health care (PHC) strategy, accompanied by the ambitious slogan: “Health for all by 2000�. The basic tenets of the PHC strategy were: that health care should be made accessible to all and health systems should be designed in such a way that communities could afford health care. Community participation was fundamental to the primary health care strategy as was intersectoral collaboration and equity which was defined as “the fair possibility for everyone to access health services and achieve a healthy life.� Uganda entered into the PHC arena rather late due to the 134 political upheavals of the late 1970s and early 1980s. By the time Uganda embraced PHC, the public health system was in a terrible state and health care was provided mainly by the religious non-governmental sector, now described as the private-not-for-profit (PNFP) sector. However, PHC has since become the backbone of the health system. For example, in 1999 it was the basis of the National Health Policy; programs such as the Uganda National Expanded Program for Immunization, and the Essential Drugs Program proved successful by extending PHC; and the Uganda National Minimum Health Care Package was elaborated and is comprised of interventions including preventive interventions that address the major causes of the burden of disease. Key implementing partners of the government of Uganda are the networks of the religious medical bureaus, of which there are three: The Uganda Protestant Medical Bureau (UPMB), the Uganda Catholic Medical Bureau (UCMB) and the Uganda Muslim Medical Bureau (UMMB). These networks of private-not-for-profit (PNFP) health facilities tend to have a strong rural presence and have historically grown from self- determined needs of congregations. The PNFP facilities are acknowledged for their strong links to communities through well-established and vibrant community based health care (CBHC) programs. In addition, the medical bureaus provide access to substantial networks of parent religious bodies and congregations; providing a potential mechanism to reach thousands of people with weekly messages on reproductive health. Increased partnership and collaboration with these service delivery points is of crucial importance to stakeholders working to improve reproductive health outcomes for the people of Uganda. In the early 1980s, UPMB revised its constitution in response to the Alma Ata Declaration of 1979 to pave the way for PHC programs and community based health care (CBHC) activities. This move opened up membership into UPMB to include lower level health units and CBHC programs. UPMB has several other PHC-focus areas, for example, a project begun in 2006 for improving reproductive health through PHC activities. REACHING GIRLS AND WOMEN WITH QUALITY SRH INFORMATION AND SERVICES In partnership with Interact Worldwide (IW), in April 2006, UPMB implemented a three- year sexual and reproductive health (SRH) project – with funding from the Big Lottery Fund (BLF). The overall theme of the project was Reaching Girls and Women with Quality SRH Information and Services in 10 Districts in Uganda, and was implemented by ten UPMB member facilities through their existing facility-based and CBHC programs. The project covered ten districts in Uganda and targeted women, girls, boys and men in the communities as well as community leaders and community health workers (CHWs). The main technical aims were to: 1) Increase the quality, appropriateness and accessibility of SRH services for women and girls in target communities (through training, improving skills of facility and community based health workers and strengthening the provision of SRH health services at facility level). 2) To increase the knowledge and awareness of SRH rights among women, girls, boys and men, thereby contributing to greater gender equity and empowerment (through advocacy, 135 improving support for SRH issues among communities; and behavior change communication, improving health-seeking behaviors among women girls boys and men). SERVICE UTILIZATION As a result of this project, capacity of implementing facilities to provide sexual and reproductive services within the Uganda Minimum Health Care Package was greatly enhanced through provision of equipment and training of staff. Personnel were able to acquire skills in post abortion care (PAC). The project also facilitated maternity wings (Maternal and Child Health building), installation water tanks and other equipment such as manual vacuum aspiration machines, delivery kits, maternity beds and infant warmers. Capacity building was done early in the project and allowed facilities to increase the quality and range of SRH services both at facility level and in outreach activities before demand was generated. The sequence of activities ensured that capacity was built met the increased facility based service demand. Table 1: Utilisation of services over the project life M F Total No. Receiving routine counselling & testing for HIV services 14,100 45,121 59,221 No. treated for Opportunistic Infections 6,343 15,221 21,564 No. treated for STIs (Syphilis, abnormal discharges etc) 2,329 5,599 7,928 No. started on Septrin Prophylaxis 1,834 14,936 16,770 No. started on ARVs 409 3,705 4,114 No. counselled for prevention of STIs and HIV/AIDS 11,550 30,568 42,118 Source: Authors. As far as range of services is concerned, this was widened due to procurement of equipment that did not exist before - key among them a lithotomic bed and manual vacuum aspiration machines which led to the strengthening of post abortion care services. All project sites reported an increase in outreach services. The quality of services improved due to procurement and installation of several equipment and infrastructure development which contributed greatly to the increase in staff morale and utilisation of SRH services. Based on antenatal and postnatal care (ANC) exit interviews, there was a gross improvement in quality of ANC as seen from a number of indicators. Despite reduction in mean time spent with the health personnel (by recall) during the ANC visit, we saw an increase in proportion of ANC clients who discussed ANC issues as well family planning. A total of 670 women and girls were seen at the health facilities with complaints related to sexual and gender based violence. Of these, 415 were referred to other related services like the Local Councils or police. This was one of the services previously unavailable in the health facilities although it is articulated in the minimum health care package of Uganda. 136 Figure 1 Source: Authors. RESPONSIVENESS TO CLIENT NEEDS: YOUTH FRIENDLY SERVICES At baseline it was noted that young people were not utilizing available SRH services, both at community and facility levels. That is, despite a high need and demand for services by the youth, the utilization was low. There were also no services targeting youth in relation to sexual and reproductive health rights (SRHR) – despite serious related concerns in the community, such as: ⋅ Most girls and boys were not aware of their rights relating to sexual and reproductive health. ⋅ Parental guidance in SRH issues to adolescents was very low, mostly due to lack of knowledge. Only 43 percent of girls and 27 percent of boys had discussed sex with their parents; and only 49 percent of girls and 17 percent of boys had discussed pregnancy with their parents. ⋅ According to the baseline survey, only 16 percent of parents had enough knowledge to discuss sexual matters with their teenage children. Against this background, a specific youth program was started, to intervene to change behaviors. UPMB conducted training of 3 health workers from each implementing facility including the focal person and one youth community volunteer. In addition UPMB provided training in communication skills and peer counseling to 3 youths identified by each of the implementing facilities. Following the training implementing facilities submitted budgets to UPMB that were funded using project funds to establish ‘youth corners’. Young people were mobilized and provided information as well as SRH services as a result. For example, during the subsequent survey, 45.7 percent of the young people interviewed reported a willingness to return to the facility for reproductive health services because of the availability of a local youth corner or peer-educator. 137 Today youth-friendly services (YFS) are available in participating health facilities due to: ⋅ The Introduction of youth corners in the facilities: Midwives trained by UPMB in the provision of YFS, goes to each youth corner once a week to provide antenatal services, encourage young mothers to seek services, and provide voluntary counseling and testing for HIV/AIDs. In this way, young people can now more easily access SRH services without fear, as they are easily attended by fellow peers or referred to specific trained health workers who would respect confidentiality when necessary. This has helped to reduce stigma among the youth, and some parents now visit youth corners together with their adolescent children for advice. ⋅ Improved friendly attitude of the trained health workers in the provision of YFS: In a health facility in Arua district, youth have come up with a unique service of starting a specific antenatal clinic for expectant adolescent mothers who are in early- or forced marriages, are pregnant as a result of rape, or pregnant out of ignorance. These youth women are usually illiterate or have dropped out of school, and some are HIV positive. This service is uniquely designed to serve those who fear going to hospitals for antenatal care due to fear of stigma. At community level, evaluators found a significant number of youths arrived for SRH services as a result of introducing free youth friendly services at the health facilities and recreational activities in the form of sports activities. Generally, the community felt satisfied with the services provided. For example, in Goli HC III where the SHR project was implemented, Joseline* a 16 year-old girl from Jupa Riek village described how the project had impacted on her life. Joseline had become pregnant and in the process developed a swelling that protruded on her private part, and she did not know what caused the swelling. “...It grew so big and made me uncomfortable to the extent that I could not even work and appear in public. Our community village health volunteer came over and counselled me and she referred me to Goli HC III. In fact, she accompanied me, where I was received and admitted. I was treated and little did I know that it was an STI. Since then I have given birth to a normal baby and my husband was also treated. I continue accessing post-natal care services after birth and the services offered by Goli HC III have contributed to my family’s good health.� Facility utilization data also showed a general increase in the uptake and of care, support and treatment services by young people. Figure 2 shows the increase in utilization of selected reproductive health services by girls at four implementing sites during the first three months (April, May & June) of year 3 (2008) of project implementation in comparison to year 2 (2007). 138 Figure 2: Utilization of selected reproductive health services by young girls Source: Authors. ACHIEVING COMMUNITY PARTICIPATION Male involvement A main strategy of this project was to involve men in SRHR issues and activities. Encouraging men to come for HIV testing with their wives was found to be particularly successful; the channels used for communication and the messages put forward also exposed benefits of knowing ones status. Men were mainly reached through community outreach, the church, Fathers Union meetings, radio programs and community meetings. Ishaka hospital adopted the practice of providing incentives, rewarding every man who accompanied his wife for counseling and testing, family planning, prevention of mother to child transmission (PMCTC) or immunization procedures. In order to encourage men to keep accompanying their wives, every accompanying man was removed from the queue, given priority service and provided with a mosquito net. In Nebbi district, Goli health facility organizes Fathers Day Meetings, once every three months in each participating village, which targets men, sensitizing them to the importance of supporting and accompanying their wives as they seek SRH services. Once the local councils became involved in mobilization it became compulsory for every man to attend. Providing information The first two years of the project focused on the sensitization of communities to change their reproductive health beliefs and behaviors - especially those based on traditional norms, cultures, myths, misconceptions and negative attitudes towards SRH services. These beliefs greatly hindered the rights of the communities, especially women and girls, from gaining knowledge of SRH services and thus hindering their access to and utilization of these services. The lack of knowledge, accessibility and utilization of the services was directly linked to SRH-related problems in the communities such as: HIV/AIDs, STI infections, gender-based discrimination and violence, increased maternal mortality, unwanted pregnancies, complications related to abortions, early school leaving, and early marriages. Over 350 community health workers (42 percent women) have been trained to act as champions for SRH issues in the targeted communities. Appropriate user-friendly communication materials were used in these communities including posters 139 and T-shirts distributed to implementing facilities. Behavioural change and communication interventions included drama, video shows, community discussions, and radio spots regarding SRH. Table 2 shows the population in the project area who received empowerment through information, education and the communication strategy of the project. In all age groups and geographical locations, more females than males received appropriate information. Table 2: Population receiving appropriate information by age, sex and location Name of health unit Youth 10 – 24 years Adults > 25 years Total Male Female Male Female Kyetume 549 500 28 174 1,251 North Kigezi 36 135 44 187 402 Ruharo Mission 350 818 1052 2672 4892 Kolonyi 2470 1878 2721 2439 9508 Ishaka 282 570 799 1164 2815 Goli 1515 2445 1363 2788 8111 All Saints Kagoma 262 421 210 385 1278 Kuluva 385 340 364 430 1519 Total 5849 7107 6581 10239 29,776 Source: Authors. Community involvement in design, implementation and monitoring The project adopted as a key approach the use of community health assets or resources such as community health workers, peer educators and traditional birth attendants (TBAs) who were equipped with skills to undertake SRHR community level activities. These community-based structures supported SRHR activities and increased mobilisation for the utilisation of services. In addition they networked communities the catchment area’s health facility. As a strategy to change SRH issues and behaviors in the communities, the project focused on involving community members from project start to finish. The participating health facilities included service user and community perspectives in planning and delivery of services during the project cycle. Local participation was encouraged, for example through the following methods: ⋅ Information, education and communication (IEC) materials developed through a participatory approach ⋅ Behavioural change communication (BCC) programs developed, utilizing drama groups already in communities ⋅ Technical working group supporting the development of the BCC strategy supported ⋅ Project advisory committee meetings held (district national level) ⋅ Monthly meetings with community health workers and District Health Officers ⋅ Community discussion forums ⋅ Focus group discussions during the annual project reviews Women and girls were significantly involved in the design, development, and use of information, education and communication (IEC) materials within target communities. For example, of the 234 individuals who participated in the development of IEC materials in western Uganda: 68 percent were women and girls; 18 percent were men and boys; and 14 percent were community leaders, CHWs, and facility-based health workers. In 140 eastern Uganda a total of 115 persons participated in the development of IEC materials. Of the 115 participants, 26 percent were adolescents, 35 percent were women, 18 percent were men, and the rest were facility-based community health workers and local leaders. 41 drama groups were trained to act as a channel for dissemination of the messages - 32 percent of drama group leaders were women. By Year 3 of the project, 92.2 percent (333/361) of the young people interviewed knew where to find HIV testing services; this was slightly higher than 86.7 percent (333/384) noted during the second year. The level of awareness of rights among the adolescents was high at 82.8 percent comparing favorably with the finding of 80.2 percent that was obtained during the second year. Over half (58.2 percent, or 217/373) of the young people reported that they had had an HIV test in the last two years. This reflected an increase of about 7 percent from the findings of the second year. Knowledge of HIV status was also high among adults, 77.7 percent (283/64) of them reported having taken an HIV test within the past two years. Though this is not significantly different from that of year two APR (78.1 percent) it still reflected a high and sustained uptake of care among the adults. Sustainability and affordability The project activities described here were not implemented in isolation, but integrated with existing health care services provided by the UPMB health facilities. Therefore, when funds supporting these activities have been expended, most of the remaining activities are self-supporting and will therefore continue. Referral and linking services will be maintained, forming part of the sustainability measure that will benefit the target populations. Although all the facility-based focal persons and peer educators were trained at the UPMB centre, health facilities were responsible for cascade training of community health workers and traditional birth attendants at community level. The training was found to be relevant and appropriate to SRHR activities in terms of content, pedagogical approach, duration, medium of instruction and follow up. Local training at the health facility level was seen to be appropriate because it limited costs by using local resource persons. The project was initiated by training community resource persons. These community structures have helped to link SRH services at the respective health facilities and the community beneficiaries and mobilizing the community to utilize the services. The target group will continue accessing the services. In addition, the training of peer educators will help in voluntarily disseminating SRH information to fellow youths. We hope that services offered by the network of resource persons will outlive the project life span as could be seen at the time of the end evaluation. The orientation exercise at the beginning of the project was broad and involved a wide range of stakeholders such as religious-, community- and political leaders, and these networks should endure. 141 CONCLUSION The following conclusions emerge from the analysis: ⋅ Improved access to health services is dependent on capacity-building of community leaders and health workers. This leads to improved skills of health providers and improved attitudes towards clients (the health worker’s attitude determining the effective delivery of services). ⋅ The involvement of community-, district-, and religious leaders at every stage of project development is important to the project’s survival and sustainability. Community empowerment and participation is critical for improving reproductive health services. ⋅ Use of IEC materials and provision of memory aids during the counselling session, use of posters, flipcharts and illustrated booklets (pre-tested for comprehension and cultural acceptability, especially for groups with a low literacy rate), assist both providers and clients in the sharing of key information. Illustrated take-home materials can be used during counselling to help clients recall instructions later and also to further disseminate accurate information. Video shows for young people on SRH issues were very helpful in reinforcing the right behaviours at a community level. ⋅ Involving the target community in the development of their own IEC messages was important, as they know their needs and the language in which such messages should be conveyed. These IEC messages were found to have a significant impact on behaviour. ⋅ Male involvement has considerable potential to improve health seeking behaviour for women and children. ⋅ The availability and supply of services must be made to match the increased demand which comes as a result of community mobilisation. ⋅ Mechanisms to hold policy-makers and service providers accountable such as community score cards are vital for sustained improvements. ⋅ Consistent documentation of best practices is critical for replication and scale-up. This project demonstrates that primary health care (PHC) is still relevant and approaches to health care that employ the principles of PHC lead to greater benefits to communities. The PHC approach utilized here, focused on the provision of better quality essential reproductive health services. There was a reciprocal cycle of information put into place: client needs were gathered through surveys and monthly meetings; and information was in turn made available to communities through community health workers, drama groups and district advisory committees. The reproductive health services provided were therefore increasingly responsive to client feedback and needs; and communities were encouraged to monitor the project progress and designed IEC materials themselves. The project demonstrated increased utilization of reproductive health services and increased knowledge and awareness among target communities. Monetization of community based health programs and inadequate community health financing initiatives remain the biggest barrier to scale up of such models. 142 CHAPTER 9 INCREASED FUNDING FOR AIDS-ENGAGED (FAITH-BASED) CIVIL SOCIETY ORGANIZATIONS IN AFRICA? Jill Olivier and Quentin Wodon 47 The World Bank This paper considers the evidence on the comparative extent to which faith-based civil society organizations (FB-CSOs) have benefited from increased funding related to the HIV/AIDS response in Africa. First, we review the literature on whether FB-CSOs have benefited from such funding, and find the arguments vigorous, but the evidence inconclusive. Next, we rely on a survey carried out in six Southern African countries to compare the profile and sources of funding of FB-CSOs against the broader collection of CSOs (non-religious or ‘secular’). It is important to be aware of the at times artificial distinctions made between faith- based and ‘secular’ structures, given the often integrated presence of religion in the lives of civil society actors and their institutions – especially in Africa. However, it is still useful to consider this particular distinction – impacting as it does on current policy discussions and strategies for civil society engagement. While the data of this particular study is mostly representative of a cluster of well- established ‘CSOs’, the evidence suggests that these FB-CSOs have been able to benefit as much as other CSOs from enhanced funding opportunities. We conclude, with a discussion of the challenges that remain for supporting smaller and less formal FB-CSOs and initiatives operating at a local community level. INTRODUCTION While civil society organizations (CSOs) have been recognized as being of critical importance to the HIV/AIDS response by multilateral and national agencies as well as national governments, the extent to which they have been able to access targeted sources of funding remains a much-debated concern. There is evidence that funding for well- established CSOs has increased significantly over the last decade, but it has also been suggested that local-level and less formal CSOs still remain largely disconnected from donor funding streams and unaligned with national strategies (see Birdsall and Kelly 2007, Rodriguez-Garcia et al 2011). Concerns have been raised as to whether the substantial international funding provided for HIV/AIDS has indeed ‘trickled down’ effectively to the local level – and how this may play out in the future especially in the face of increasing resource constraints. 47 We are especially grateful to Kevin Kelly (CADRE) for sharing the dataset used in this paper, and for resulting discussion and comment. This paper also benefited from discussion and comments from Rosalia Rodriguez-Garcia and René Bonnel. 143 As part of the interest in the community response to HIV/AIDS, there are also as yet unanswered questions about the availability of funding for faith-based civil society organizations (FB-CSOs), 48 and how this relates to their assumed particular capabilities (as compared to CSOs that do not self-identify as faith-based, from here on, simply titled ‘CSOs’). The work of FB-CSOs has clearly become more visible over the last two decades, and a more substantial literature has grown landscaping the response of FB- CSOs to HIV/AIDS, particularly in Africa (see Olivier and Wodon 2012). Certain ‘characteristic’ strengths and weaknesses have been observed - especially in relation to FB-CSOs operating at a community level, such as: a particular presence and connectedness to community, or a particular lack of capacity for evaluation and documentation or poor representation in national structures (see Difaem 2005, Haddad et al 2008, Keough and Marshall 2007). One must wonder whether this increased attention paid to FB-CSOs at the policy level in relation to HIV/AIDS engagement has in turn increased their ability to access funding. Given that we are at a critical point – where both HIV/AIDS funding and community engagement are under the spotlight – it is timely to consider whether the interest in FB-CSOs has resulted in specific resourcing, or whether this has remained largely at the dialogue level. The objective of this paper is to take stock of what we know about the magnitude and characteristics of donor and other sources of funding towards FB-CSOs in relation to HIV/AIDS-response, as compared to CSOs more broadly. We first provide a brief review of the literature on this topic, which appears to be filled with opinions, but as yet inconclusive. Next, we (re)analyze data collected by Birdsall and Kelly (2007) among CSOs in six southern African countries (Lesotho, Malawi, Mozambique, Namibia, Swaziland and Zambia). Birdsall and Kelly have produced a detailed and well recognized report on these data. They found that the increased interest towards CSOs has indeed resulted in increased funding from various sources between 2000 and 2005. But they also suggest that “funding bottlenecks have often resulted in resources not reaching communities in adequate volumes, or reaching groups that are particularly vulnerable or high risk.� They go on to describe a ‘funding funnel’, where large NGOs receive the bulk of donor support at a national level, with this narrowing down considerably as increasingly limited funds trickle down to local CSOs. This OSISA dataset remains one of the most substantial of its kind. However, Birdsall and Kelly focused mainly on the broader CSO landscape, and did not substantially tease out the comparative characteristics and funding patterns of the FB-CSOs included in their dataset. We therefore work to extract this information from their dataset, to see if it can cast any further light on the specific funding patterns and resourcing of FB-CSOs. The results, as described below, suggest that among the relatively well established CSOs included in the dataset, there are in fact few dramatic differences between FB-CSOs and CSOs, with FB- CSOs benefitting as much from donor and other funding. 48 There are on-going and as yet unresolved concerns about terminology in relation to ‘faith-based organizations’ (also called religious entities, faith-inspired initiatives or institutions, or the like). In this paper we utilize the term FB-CSO to indicate a classification in this particular dataset under discussion – between those CSOs who self-identify as ‘faith-based’ and those that do not (non-religious CSOs). 144 Of course, as Birdsall and Kelly already noted, such observations do not mean that all CSOs (or FB-CSOs) involved in the community response to HIV/AIDS have benefited to the same extent from increased funding. There is a lot of heterogeneity within civil society – and particularly within the sometimes bewildering clusters of faith-inspired institutions and initiatives (commonly called the 'faith sector'). While more is known about formal FB-CSOs (for example, those that are nationally registered, or have visible infrastructure), much less is known about informal FB-CSOs and initiatives (such as congregations and community initiatives or projects run by informal faith groups at the periphery of congregations – see Schmid et al 2008). As one would expect, more structured, formal CSOs are better able to access donor funding mechanisms. Yet in any given community in Africa one may find a complex web of initiatives, also with, for example, an array of international faith-based and secular agencies both funding and running programs at a local level. This, often in parallel to programs being run through multiple faith-affiliated sources: from large congregations, through denominational offices, from denominational networks, as well as initiatives motivated by faith-inspired community members and completely unaffiliated with any organization. The common argument is that these more complex and less formal initiatives may be having just as much impact, but are less likely to have access to the funding streams which mainly recognize CSOs of the ‘NGO variety’ (and of course, sometimes also because of the perceived ‘dangers’ of funding initiatives of a faith-based character). In the final discussion we raise some of these broader concerns – and consider what it would mean for AIDS-related funding to reach mostly informal faith-based community initiatives. REVIEW OF THE LITERATURE Renewed interest in FB-CSOs For many years, before the HIV/AID epidemic became a central concern, FB-CSOs were often invisible to the international gaze. This was true for both the large national-level NGOs (such as the mission hospitals and health services), as well as for the local community-based organizations with a faith character and congregational initiatives. There has since been substantial work done to make these entities and their activities more visible: the inclusion of FB-CSOs in policy and documentation has become the norm and high level dialogue with religious leaders more prevalent. Bilateral and multilateral organizations such as the World Bank, PEPFAR, DFID and GFATM have all made significant public gestures towards the ‘faith sector’, and have hosted workshops and seminars aimed at identifying information gaps and overcome obstacles that have limited the sector’s contributions to HIV and AIDS programming (Benn 2011, Olivier et al 2006, Taylor 2005a and 2005b). 49 49 For example, the World Bank has held several international workshops to help FB-CSOs access funding from national HIV/AIDS programs including the World Bank MAP program – such as the one in Addis Ababa in 2003, then in Accra in 2005 (Keough and Marshall 2007). UNAIDS holds many consultative programs with FB-CSOs – as can be seen in the report Partnership with faith-based organizations: UNAIDS strategic framework (UNAIDS 2009). The Global Fund to Fight AIDS, Tuberculosis and Malaria has similarly held several workshops and meetings expressly to facilitate FB-CSOs access to Global Fund resources, as is expressed in Report on the involvement of faith-based organizations in the Global Fund (GFATM 2008). PEPFAR and USAID have similar statements on their websites. 145 Much of the international ‘rediscovery’ of FB-CSOs has, in fact, been related to HIV/AIDS - with the idea that faith communities are not only an important entry point for intervention, but also that FB-CSOs have a particular potential or comparative advantage for HIV/AIDS response. 50 This potential is usually described as being rooted in their connection to community, their access, reach or trust, their longevity or reliability (Olivier and Wodon 2012). In their study of AIDS-engaged CSOs in sub-Saharan Africa, Birdsall and Kelly (2007) state: “There is a widespread prevailing belief that CBOs and FBOs are an under-utilized resource for expanding the reach of services to the poorest of the poor and ‘spending money where it most helps’…CSOs have the trust of their communities and can therefore work effectively on personal and intimate issues. This view is promoted particularly strongly by PEPFAR, which sees FBOs as possessing particular ability to ‘influence the attitudes and behaviors of their community members buy building on relationships of trust and respect.’ High levels of religious affiliation and the role of churches in delivering health services make them ‘crucial delivery points for HIV/AIDS information and services.� But has the renewed interest in FB-CSOs generated additional funding for their programs and interventions? As is common in this field of inquiry, where reliable and systematic data are hard to come by, there are two directly opposing discourses prevalent as regards to the funding of HIV/AIDS-engaged FB-CSOs. The first opinion is that the increased interest in FB-CSOs has resulted in significantly increased resources and access to funding (the most commonly stated example is the PEPFAR program); and the directly opposing view is that FB-CSOs receive comparatively less support than ‘secular’ CSOs – precisely as a result of their faith-based character (Olivier 2010). Generalizations about ‘FBOs’, ‘faith sector activities’ and in particular whether or not FB-CSOs are ‘at the international funding table’ are not entirely useful - as every national (and district-level) context is different. For example, the post-conflict Francophone context of the Democratic Republic of Congo has a particular humanitarian-medical focus to its HIV/AIDS response, with heavy donor involvement and several large FB-CSOs managing large sectors of the health system (Haddad et al 2008). By contrast, in the Muslim-majority Mali context, fewer FB-CSOs operate, although there is substantial engagement in HIV/AIDS through religious leaders’ education programs and the like instead (Schmid et al 2008). Malawi illustrates still a different context with the characteristics of a mature ’AIDS industry’ with a broad range of donors and a large number of FB-CSOs (of the NGO variety) involved (Haddad et al 2008). 50 It might even be argued that the HIV/AIDs response has helped form multisectoral collaboration more broadly. Those countries with more prevalent HIV/AIDS epidemics now have more mechanisms for interfaith collaboration in place than those that do not. Certainly countries such as Zambia, Malawi, Kenya and Uganda, now have several mechanisms (such as Interfaith AIDS Councils) to better strengthen the national ‘faith sector’ response and representation. On the other hand, while funding has helped many FB- CSOs in expanding their services, it also has created tensions. Haddad et al (2008) note obstacles to effective interfaith and multi-sectoral collaboration including in some countries a lack of real representation (for example on interfaith AIDS councils); competition for funding between FB-CSOs; interfaith rivalry; and the sometimes dogmatic and conservative attitudes of FB-CSOs. 146 Perceptions of increased funding for FB-CSOs Despite the general inadequacy of data and evidence on FB-CSOs, there is a strong perception at a policy level that FB-CSOs have also benefited from the higher levels of funding targeted towards CSOs in recent years. Bonnel et al (2011) describe the main sources of donor funding for HIV/AIDS-engaged CSOs in general (such as PEPFAR, GFATM, the World Bank’s MAP) - and these mechanisms have been described with specific attention to FB-CSOs as well (see Haddad et al 2008, Keough and Marshall 2007, Taylor 2005a, 2005b, 2006, 2007). We summarize here some of the issues most pertinent to FB-CSOs. PEPFAR (the President’s Emergency Plan for AIDS Relief) has clearly, and at times controversially, paid special attention to FB-CSOs. Keough and Marshall (2007) summarize some of the controversy (such as the great ‘condom issue’) and note that PEPFAR currently takes the stance that funding is only based on standard criteria of merit and capacity, but that faith groups are also acknowledged as essential in PEPFAR’s activities. There have been some questions as to whether PEPFAR’s funds are overly weighted towards Christian evangelical groups in the past. PEPFAR does not make broad estimates of total comparative allocation to FB-CSOs, however, it is possible to look at the main recipients manually and conclude that a significant number of primary and secondary recipients are indeed nominally faith-based. 51 The Global Fund (The Global Fund to Fight AIDS, Tuberculosis and Malaria – GFATM) has entered into the FB-CSO comparative allocation debate more cautiously. In the literature on the faith-based response to HIV/AIDS in Africa, the Global Fund frequently gets mentioned as an example of inadequate funds being allocated to FB-CSOs, calculated as a percentage of total disbursements. However, there is a lack of clarity as to just what percent of funds from GFATM should be considered high or low – given the poor data on FB-CSO response or impact to compare such funding provisions against. As an indication of the level of interest in this topic – clearly responding to this debate, the Global Fund released a Report on the Involvement of Faith-based Organizations in the Global Fund (GFATM 2008) in which it argues that GFATM has always recognized the important role of FB-CSOs, and that in 2006, nine FB-CSOs were allocated funds as principle recipients, with an additional 488 FB-CSOs as sub-recipients. In addition, the Global Fund also noted that in 2006, of the 120 Country Coordinating Mechanisms (CCMs), 94 had at least one FB-CSO representative. The report also notes that the allocation to FB-CSOs is regionally different, with the highest percentage of monetary resources being channeled to FB-CSOs in West and Central Africa (11.8 percent). Even for those countries with a comparatively lower percentage of resources going to faith- based sub-grantees, GFATM argues that this was mainly because in these areas a larger number of FB-CSOs received small grants. For example, in Eastern Africa FB-CSOs received 2.4 percent of funds, representing 61 FB-CSO sub-grantees. The emphasis of the Global Fund (2008) on sub-grants provided to ‘sub-sub-recipients’ is legitimate. The most apparent example of this is the Churches Health Association of 51 See http://www.pepfar.gov/budget/partners/index.htm 147 Zambia (CHAZ). Together with another CSO (ZNAN – The Zambia National AIDS Network), CHAZ received the bulk of Global Fund support for Zambia in 2006 (58 percent committed, and 56 perceived received). According to the Global Fund, CHAZ in turn dispersed money to “411 local FBOs to fight AIDS, 73 local FBOs to fight TB and 75 local FBOs to fight malaria.� Many others have held up the Zambian country mechanism as a success story – and CHAZ in particular as an exemplar of best practice. Of course, it should be noted that CHAZ is not dependent entirely on Global Fund monies. It receives funds from a variety of different sources – and in turn sub-grants to a wide network of partners. For example, DanChurchAID provided US$6 million in support for local FB-CSOs and NGOs, including CHAZ, over the period 2002-2005, working through multi-year partnerships with local NGOs, many of which are faith- based, and all of which is targeted towards CSOs (Birdsall and Kelly 2007). This is simply noted to provide an example of the complexity of faith-inspired actors, including international NGOs, national NGOs and umbrella bodies, grantees, sub-grantees, and sub-sub-recipients. Parsing out which funds come from faith-based/secular sources, and which then are dispersed to faith-based/secular CSOs is a complex and challenging process. The DanChurchAID example also introduces another lesson from the literature, namely the rapid growth and presence of faith-based international NGOs (INGO). It is, in fact, difficult to argue that the ‘faith sector’ is not receiving sufficient funds in the face of the increased allocation to the large FB-INGOs. Consider the case of World Vision International (WVI) which has dramatically expanded its scope of work over the last decade. WVI demonstrates the adaptability of some INGOs - in each country context engaging in a different funding relationship: sometimes operating as a primary recipient, sometimes as a secondary recipient; sometimes partnering with local institutions to provide services and at other times acting as a local NGO or provider themselves. Just as one example, in 2006, WVI had a total Global Fund portfolio (as a primary or secondary recipient) of more than US$130 million. The rapid expansion of INGO recipients has created some tensions at the local level. FB-CSOs have noted how important INGOs now are for their support and existence – and also the difficulties that occur in managing these partnerships. For example, local FB-CSOs note the challenges that occur when INGOs begin a funding process in collaboration with local partners, but then often ‘turn into the competition’ once the funds become a reality (Haddad et al 2008). Less evidence is available on the World Bank’s MAP (Multi Country HIV/AIDS Program) support for FB-CSOs, because the World Bank does not disaggregate its data that way, and provides funding to governments as opposed to CSOs directly. Yet, as noted in Bonnel et al (2011), indirect support to CSOs has been substantial, and it is likely that a substantial share of those CSOs were FB-CSOs. There are also examples of interventions where funding was initially granted to governments, but then provided by the governments to large FB-CSOs with the World Bank’s blessing. One example is the ‘DREAM’ project managed by lay Catholic Sant’Egidio Community, which was launched in Mozambique in 2002 (Keough and Marshall 2011). 148 Perceptions of inadequate funding for FB-CSOs At the same time, while many FB-CSOs seem to have benefited from increased funding for HIV/AIDS-response, this does not mean that they are ‘well-funded’, or that they benefit from their ‘fair share’ of HIV/AIDS funding. A global assessment of FB-CSOs’ access to resources for HIV/AIDS reports that “despite substantial efforts and good will by all, churches and other faith-based organizations have not yet been consistently successful in accessing resources for their response to HIV and AIDS from international funding agencies� (Difaem 2005). In this study, the following statement is said to reflect the opinion of many from around the world: “Theoretically, church organisations should be able to access resources because of their good track record, their close links to communities, their emphasis on positive values, their enormous human resources and their credibility and sustainability as institutions. In practice, government agencies tend to keep most of the resources from international donors to themselves, and donors happily go along...Sometimes churches create obstacles for themselves, because they have hang-ups on policy issues such as condoms, or they are judgmental towards people living with HIV and AIDS� (study respondent in Difaem 2005). The perception that FB-CSOs are not getting their fair share is common at the international dialogue level. In a recent report from a collaborative meeting convened by the Center for Interfaith Action on Global Poverty, the authors state: “…while comprehensive data on the scale of development resources channeled through faith entities is lacking, many in the sector suggest that funding is not commensurate with the share of services they provide…during the first eight rounds of Global Fund grant- making, faith-inspired organizations received only 3.1 percent of disbursements. This level of funding would seem to be far below a fair share for the sector, given that one in five organizations involved in HIV/AIDS programming is faith-based, and that faith- based organizations provide an estimated 40 percent of HIV/AIDS treatment and care in sub-Saharan Africa. Some FBOs are concerned that funding for faith entities, as well as NGOs generally, could further decline as donors increasingly concentrate on government health systems strengthening and direct budget support for governments� (CIFA 2011). 52 A number of reasons are usually provided to account for the difficulties of FB-CSOs in accessing HIV/AIDS resources. For example, some secular governments are resistant to funding FB-CSOs; funders fear their resources might be utilized for proselytism; FB- CSOs are historically distrustful of government and international funding processes; they lack capacity for dealing with complex funding proposals and evaluations; and it is particularly difficult to hold FB-CSOs accountable (see Haddad et al 2008; Taylor 2005a, 2005b, 2006, 2007). The perception that FB-CSOs receive comparatively less support for 52 This particular quote from the CIFA report, which addresses the percentage of services provided by the ‘faith sector’ versus the percentage of funds provided, raises however another key concern. As we have argued elsewhere (Olivier and Wodon 2012), broad generalizations about the percentage of services being provided versus the percentage of funds being allocated to ‘FBOs’ or the ‘faith sector’ generally is a common advocacy device in this field of research. We have argued that most market share estimates are often biased on the high side, and not only problematic, but also counter-productive given the inadequate information available and the heterogeneity of the ‘faith sector’. The broad and complex nature of the ‘faith sector’ means that making any broad statement about services provided or funding flows can be challenged. 149 HIV/AIDs response is based on a complex history, where many felt the role and presence of FB-CSOs was generally not well acknowledged at national and international levels. For example, in a UNFPA (2004) report, speaking about Malawi, the authors note: “Most faith-based organizations and religious institutions involved in HIV/AIDS prevention and care feel that they have been marginalized to a large extent by the Government and NGOs. Many international organizations regard faith-based organizations as extremists and untrustworthy, which has discouraged religious institutions and hindered the formation of long-term partnerships…in some cases they are met with resistance in their search for partners and funding because of their stance on condoms.� And as Keough and Marshall (2007) note, when funding is provided, this can also be seen as problematic “for some secular groups, the growing partnership, and the associated funding flows, between faith-based organizations and governments/donors, remains a deeply divisive topic. This is especially relevant in the United States, as it pertains to the Bush administration’s deliberate strategy to channel resources to faith-based organizations - virtually all Christian - within the context of…PEPFAR." Several authors have also noted the difficulties of different ‘languages’ between faith- based and ‘secular’ stakeholders (Keough and Marshall 2007, Olivier 2010). While stakeholders such as donors tend to speak in technical terms, FB-CSOs tend to frame their actions more discursively, in terms of motivation, or values. Taylor (2005) also notes a “fundamental mismatch between the values and bases of operation of local faith- based initiatives, and those of donors…local faith-based initiatives have something significant to contribute…(but) perceive that the funders do not understand their basis of operation and the values that lie behind their work, proposal design and implementation excludes them, and monitoring and evaluation systems are not adequate to track whether resources are reaching the poorest people or being used effectively.� As one Bishop in Zambia stated it “Churches don’t have programmes, they have church activities� (cited in Taylor 2005). This mismatch of ‘languages’ is felt most acutely when dealing with the resourcing of local informal FB-CSOs, who are deeply rooted in their particular faith. Finally, the issue of funding gaps for FB-CSOs is related to the fact that the resourcing landscape for FB-CSOs has changed dramatically – and many organizations have trouble keeping up with these changes. For those FB-CSOs who have been in existence for a long time, many traditional funding sources have dried up. For example in Kenya, Mandi (2006) describes the funding crisis in the 1990s where “much of the support FBOs were getting from the big congregations, churches and donors, as well as the assistance received from the government from as far back as the fifties and sixties, came to an end.� The area of international ‘philanthropic’ and especially religiously-motivated funding is an area where there is a particular dearth of information, but the changes that have taken place have clearly often had a negative impact on many historically rooted FB-CSOs. 150 FORMAL FB-CSOS: ANALYSIS OF THE CADRE-OSISA DATA The literature briefly outlined in the previous section therefore cannot cast much light on how significantly FB-CSOs are benefitting from HIV/AIDS-related donor funding. This is in part because much of the evidence on which the literature is based is incomplete or anecdotal – and lacking in systematic data. What is clear is that FB-CSOs, along with CSOs generally, find themselves in a new era, with new donor and national funding mechanisms coming into place, and new expectations that they participate in such collaborative arrangements. In this section, we use data collected by Birdsall and Kelly (2007) on the main HIV/AIDS-related funding streams benefitting CSOs in six southern African countries. We start by describing the methodology used by the authors to collect their data, and then compare basic statistics of the FB-CSOs in the sample. Methodology The data used in this section was collected by Birdsall and Kelly (2007) for a study commissioned by the Open Society Initiative for Southern Africa (OSISA). Details on the methodology are available in the detailed report prepared by these authors, so that only a few pointers should be necessary here. A four-page questionnaire was sent to a quasi- nationally representative sample of established CSOs working in HIV/AIDS response in five southern African countries (Lesotho, Malawi, Mozambique, Namibia, Swaziland, and Zambia). A list of CSOs working on HIV/AIDS was established in each country using information from AIDS coordination networks, National AIDS Coordinating Authorities (NACAs), and granting and sub-granting institutions. While efforts were made to reach both NGOs and CBOs (community-based organizations), the sample appears more representative of well-established organizations, as opposed to small informal ones, which is important when interpreting the results. Some 633 questionnaires were sent, with a response rate of 69 percent (439 responding organizations). In the dataset graciously provided to us by the authors, data are available on 369 organizations. The first two pages of the questionnaire provides a basic profile of the organization in terms of its characteristics, history, staff, services provided. The next two pages are devoted to funding. Our analysis is based on a sample of 349 organizations out of the 369 in the data set, because we consider only those organizations that stated whether there were is associated with a church or faith-based in orientation (there were 20 missing values). Out of the 349 organizations, 117 were FB-CSOs, and 232 CSOs which did not identify themselves as FB-CSOs (we name this group secular- or ‘S-CSOs’ from here for convenience, although this is admittedly not ideal). We rely on the subjective classification by respondents as to whether they are faith-based or not, acknowledging that this is somewhat problematic and there is likely a high degree of heterogeneity among all CSOs. 53 In two countries 53 There are in the literature no standardized typologies for FBOs and this gets particularly messy when looking at the local-community level, where there are a complex array of international non-government organisations (INGOs), national NGOs, local CBOs, networking bodies, intermediaries, congregational initiatives and informal care groups in operation – and about which we know substantially less than those operating at a national level (Olivier 2011). The classification of whether something is faith-inspired or not is especially difficult at the level of local communities where religion is part of everyday life and action. 151 (Lesotho and Swaziland), the full sample of CSOs identified as working on HIV/AIDS were included in the sample, but this was not the case in the other four countries, where among all organizations identified, a random sample of 120 organizations was selected. Despite the fact that in the four larger countries not all identified organizations were sampled, following Birdsall and Kelly (2007) we did not weight the data to account for the fact that one organization sampled in one of the larger countries would represent a larger number of organizations (since we do not have a clear census of all organizations anyway, weighting would be imprecise). It should also be noted that not all organizations answered all of the questions in the questionnaire. This begs the question as to whether the missing values should be treated as zero values or ‘no’ answers, or as true missing, in which cases all statistics would be computed only on those organizations that answered a specific question. For simplicity, we used this latter approach, thus treating missing values as true missing data. A different treatment could of course yield different results, prompting caution in the interpretation. This is why more than emphasizing point estimates, we are instead discussing the comparison between FB-CSOs and (non-FB) S- CSOs which, under normal circumstances, would be less sensitive to the missing values issue. Basic characteristics This section provides basic statistics on some of the characteristics of the CSOs in the CADRE database. In most cases, there are relatively few differences in those characteristics between S-CSOs and FB-CSOs. For both groups, 55 percent of the CSOs are located in a town or city which is as an administrative center for surrounding areas or towns, and 45 percent in a rural village or small town. Again, for both groups, close to three fourths (74 percent for S-CSOs and 73 percent for FB-CSOs) work in more than one community. Approximately nine in ten CSOs, whether they are faith-based or not, have an office or work from premises that can be visited by the public, and more than nine in ten organizations have a bank account. The number of years of existence of the CSOs and of experience in working on HIV-AIDS is also similar for both types of CSOs, as shown in table 1, with approximately 40 percent of the organizations created since 2001, and close to 60 percent having started to work on HIV-AIDS since then, suggesting that the increase in funding in this area indeed led to the creation of (particular kinds of) CSOs as well as existing CSOs emphasizing more HIV/AIDS in their work (see below). Table 1: Years of experience of CSOs and work on HIV/SAIDS (%) Started operations Started work on HIV/AIDS Up to 1991- 1996- 2001- Up to 1991- 1996- 2001- 1990 95 00 06 All 1990 95 00 06 All S-CSOs 15.1 15.0 29.6 40.3 100 3.2 7.7 29.6 59.5 100 FB-CSOs 19.6 9.8 32.2 38.4 100 5.5 7.3 30.9 56.4 100 Source: Authors’ estimation using CADRE database. Most classification strategies (including self-identification as ‘faith-inspired’ in surveys) have weaknesses and different studies employ different schema for inclusion or exclusion of FB-CSOs, making comparison risky. 152 There are a few areas where one observes differences between the two types of CSOs. The proportion of FB-CSOs that have branches or programs in other countries, at 18 percent, is higher than for S-CSOs, at 10 percent, and the proportion of FB-CSOs that are part of an HIV/AIDS association or coordinating network/body is also slightly higher for FB-CSOs, at 90 percent, versus 83 percent for S-CSOs. Also, 72 percent of FB-CSOs also conduct activities not related to HIV/AIDS, versus 64 percent of S-CSOs. This suggested that in the sample, FB-CSOs tend to be slightly more international, connected to other organizations working on HIV/AIDS, and active in other areas than is the case for S-CSOs. Another difference between the two types of organizations is that as expected, S-CSOs tend to have a higher ratio of paid staff (full-time or part-time) to the number of volunteers working for the organization than is the case for FB-CSOs. This is true for both national and international staff (see table 2). Table 2: Average number of national and international staff and volunteers Citizens of your country International staff Full-time, Part-time Unpaid Full-time, Part-time Unpaid paid staff paid staff volunteers paid staff paid staff volunteers S-CSOs 15.7 9.8 90.7 0.5 0.1 0.6 FB-CSOs 6.6 4.4 108.8 0.3 0.2 1.9 Source: Authors’ estimation using CADRE database. The literature also suggests that FB-CSO activities still often remain unaligned with larger health systems and national HIV/AIDS coordination, and that this is especially the case for response at a community level (see ARHAP 2006, Agandjanian and Sen 2007, Birdsall 2005, Haddad et al 2008). CSOs engaged in HIV/AIDS activities typically tend to focus primarily on 'care and support' - including home based care (HBC), care of orphans and vulnerable children (OVC), as well as prevention (including behavior change, awareness or education). By all accounts, FB-CSOs generally tend to have this same focus, with a particular emphasis on HBC and care of OVC (see Haddad et al 2008, Foster 2004). The obvious exception is that of the faith-based health services, such as those of the Christian Health Associations, which are often more integrated into health systems and are also more strongly engaged in treatment (ART). A specific feature of FB-CSOs noted in the broader literature is that they tend to have HIV/AIDS activities embedded in a holistic range of health and development services. This may explain why some FB-CSOs have vigorously argued that they find vertical HIV/AIDS funding particularly problematic as it does not facilitate broad-based service provision (see Haddad et al 2008, Schmid et al 2008). Also, beyond formal FB-CSOs, many faith-inspired initiatives take place at a 'sub-congregational' level, or at the periphery of the congregation – this is the case of care and support initiatives run by women's groups, or spontaneous caring activities (ARHAP 2006). Agadjanian and Sen (2007) note that much of the congregational-level assistance visible in their study cohort in Mozambique was small in scale and episodic, neither organized nor controlled by the church leadership. They found that congregational leadership was involved only in larger-scale actions that required the pooling of resources. 153 In relation to the activities of CSOs, the CADRE-OSISA survey distinguished between: the prevention of HIV/AIDS (condoms, PMTCT, VCT, education, communication); treatment, care and support (nutrition, home based care, counseling, support for people with HIV/AIDS); impact mitigation (work with orphans and others in need of social assistance, income generation, poverty alleviation); HIV/AIDS management (training, co-ordination, capacity building, M&E, systems development); policy development, advocacy, research; and acting as a channel for funds to service delivery organizations. As shown in table 3, FB-CSOs are somewhat more active in treatment, care, and support, as well as in impact mitigation and HIV/AIDS management than is the case for S-CSO. S-CSOs are slightly more active in prevention, as well as policy, advocacy and research. But overall, differences in activity profiles tend to be small. Table 3: Areas of activity related to HIV/AIDS (%) Little or no Some Much Primary All activity activity activity activity S-CSO Prevention of HIV/AIDS 4.6 22.0 36.2 37.2 100 Treatment, care and support 13.0 30.8 32.2 24.0 100 Impact mitigation 23.8 20.4 25.7 30.1 100 HIV/AIDS management 20.3 39.6 28.4 11.7 100 Policy, advocacy, research 50.8 34.3 7.2 7.7 100 Channel for funds 83.8 8.4 4.8 3.0 100 FB-CSO Prevention of HIV/AIDS 11.0 17.4 33.9 37.6 100 Treatment, care and support 10.3 18.7 37.4 33.6 100 Impact mitigation 7.3 22.0 26.6 44.0 100 HIV/AIDS management 25.2 29.9 24.3 20.6 100 Policy, advocacy, research 57.1 30.6 8.2 4.1 100 Channel for funds 72.7 16.2 5.1 6.1 100 Source: Authors’ estimation using CADRE database. The survey also asked CSOs which target groups they served. One should be cautious about the estimates in table 4 because of the missing data issue – it is more likely here than in other parts of the questionnaire that a missing may actually be interpreted as a ‘no’ value, in which case the proportions provided in the table are overestimated. But there are at least two features of the data that appear quite robust. First the data suggest that in general, FB-CSOs serve a larger number of target groups than is the case with S- CSOs, and this may be related to the fact that FB-CSOs are also more likely to run other programs and may thus be able to provide services to more target groups than is the case for S-CSOs (this should not be interpreted as FB-CSOs necessarily serving more persons – rather that they serve more varied target groups). In addition, the ranking of the various groups in terms of the likelihood of being served by both FB-CSOs and S-CSOs is very similar. Thus, apart from the fact that FB-CSOs may be able to reach more varied target groups, there are again relatively few differences between the two types of CSOs. 154 Table 4: Target groups related to HIV/AIDS (% reaching the target group) FB-CSOs S-CSOs Difference Women and girls 95.1 91.4 3.7 HIV-positive people 89.1 84.3 4.8 Street children 86.5 58.1 28.4 Farm workers 86.2 75.0 11.2 Rural people 85.7 73.9 11.8 Elderly people 84.6 65.4 19.2 Informal economy workers 80.0 61.3 18.7 Fishermen and fishing communities 73.3 46.9 26.4 Substance abusers 70.0 46.7 23.3 People with disabilities 66.7 51.3 15.4 Prisoners or their families 65.0 27.5 37.5 Minority groups 62.5 42.9 19.6 Commercial sex workers 60.8 44.7 16.1 Migrants 56.3 28.6 27.7 Informal urban areas 56.3 38.5 17.8 Long distance transport workers 47.2 32.1 15.1 Uniformed services 44.8 33.7 11.1 Refugees or internally displaced people 34.3 15.9 18.4 Men who have sex with men 26.9 5.7 21.2 Miners 23.1 14.0 9.1 Pregnant women 13.0 20.0 -7.0 Source: Authors’ estimation using CADRE database. Funding for CSOs The analysis of the basic characteristics of CSOs based on the questions available in the CADRE-OSISA survey suggests relatively few major differences between FB-CSOs and S-CSOs. We now turn to the data on funding. Table 5 provides selected data on expenditure and funding for the period 2001-2005. First, it is clear that average levels of spending on HIV/AIDS among the CSOs in the sample have increased sharply over time, with the average level of spending among S-CSOs at US$160,141 in 2005, versus US$150,613 for FB-CSOs. By contrast, the corresponding amounts for both groups were about three times lower in 2001. The number of grants received has also increased over time (that information is not collected for 2002 and 2004). Note that the average funding per CSO is substantial – this confirms that the sample includes mostly established organizations, as opposed to local informally-run community based interventions. Available studies suggest that FB-CSOs typically access a broad array of funding streams for sustainability, including government funds, external church donations, external development agency funds, local donations, donations from other local organizations, mother bodies or faith networks, as well as individual charitable donations (see Birdsall and Kelly 2007, Haddad et al 2008). There are also reportedly important differences among differently affiliated FB-CSOs. Mainline congregations tend to have more organized international links through denominational structures, while 'revival' or 'healing' churches (such as Zionist or Pentecostal) congregations tend to have less hierarchical structures and different funding mechanisms (Schmid et al 2008, Agadjanian and Sen 2007). 155 The survey provides a few interesting findings in this area, as it asks whether the CSOs benefited from funding from specific types of donors. There may be a bit of inconsistency in the data here, in that when summing up the support (yes-no answers) declared from the various types of donors in each year, one gets a much larger number of funders than indicated by the CSOs in the direct question “From how many different sources did you receive grants in each of the following years?� for which data are reported in the second row of table 5. For example, in 2005, one gets an average number of sources of funding estimated from the information on the various types of donors of 5.1 for S-CSOs and 5.5 for FB-CSOs, as compared to the values of 2.68 and 2.89 in table 4. Part of the difference may again be explained by the issue of missing values (statistics computed on subsets of the sample will overestimate the total number of funders in case some of the missing values represent no funding). Birdsell and Kelly (2007) themselves highlight the difficulties inherent in tracking funding flows through these responses – especially with regards to funds that flow from recipient to sub-recipient to sub-sub- recipient (the CSOs sometimes naming the originating source, and sometimes the recipient ahead of them in the line as the funding source). Therefore, instead of presenting the data on the types of funders as direct percentage of organizations that appear to benefit from a specific source of funding among respondents without missing value, these percentages have been scaled into indices, considering the likelihood of a S- CSO benefitting from funding from a foreign donor or an international institution as the baseline. The data on funding source by type are thus to be interpreted as relative odds ratios with the comparison being international donor funding in 2001 for S-CSOs. Three observations can be made on the relative odds ratios of funding by type of donor. First, the odds ratios are systematically higher in 2005 than they are in 2001, indicating that likelihood of funding has increased for all types of donors. Second, the largest increase in the likelihood of funding over time has been from national, provincial or district HIV/AIDS structures (the increase in relative odds ratios is from 0.24 to 1.04 for S-CSOs, and from 0.38 to 1.03 for FB-CSOs). Third, FB-CSOs tend to report slightly more different funding sources than is the case for S-CSOs, which confirms the findings on the number of grants from different sources. 54 Still, overall, differences between FB- CSOs and S-CSOs are again small. 54 Although for the number of grants from different sources, S-CSOs were better placed in 2001; by contrast, on the odds rations, FB-CSOs were better placed throughout the period under review, with very few exceptions. Out of 30 potential comparisons of odds rations (six donor types and five years of data), FB-CSOs fare worse on .likelihood of funding on only two occasions – the funding from national, provincial or district HIV/AIDS structure in the last two years of data, with the difference being small. 156 Table 5: Expenditure levels and types of organizations funding CSOs 2001 2002 2003 2004 2005 S-CSOs Average total expenditure on HIV/AIDS (US$) 49,201 69,763 94,175 121,892 160,141 Number of grants from different sources 1.80 - 2.15 - 2.68 * Foreign donor or international institution 1.00 1.01 1.08 1.08 1.10 * Government department or ministry 0.57 0.63 0.76 0.76 0.77 * National, provincial or district HIV/AIDS structure 0.24 0.45 0.91 0.97 1.04 * Other NGO 0.62 0.73 0.87 0.99 1.04 * Services provided (fees from users) 0.71 0.75 0.86 0.89 0.90 * Local sources (businesses, churches or charities) 0.77 0.83 0.86 0.91 0.95 FB-CSOs Total expenditure on HIV/AIDS (US$) 56,642 63,932 104,296 133,818 150,613 Number of grants from different sources 1.63 - 2.18 - 2.89 * Foreign donor or international institution 1.06 1.07 1.08 1.09 1.12 * Government department or ministry 0.68 0.68 0.91 0.85 0.96 * National, provincial or district HIV/AIDS structure 0.38 0.83 0.91 0.94 1.03 * Other NGO 0.91 0.95 0.98 1.03 1.11 * Services provided (fees from users) 0.97 1.00 1.00 1.04 1.04 * Local sources (businesses, churches or charities) 0.94 0.97 1.02 1.02 1.04 Source: Authors’ estimation using CADRE database. Note: * indicates that the variable is expressed as an index value – see text for explanation. Another question relates to the type of funding received by class of expenditure (table 6). The survey distinguishes between salaries, stipends or incentives; office and administration costs (such as rent, electricity, telephone); program costs, including supplies (such as home-based care kits, gloves, rapid test kits, transport, training costs); and equipment or vehicles. Again, the differences between FB-CSOs and S-CSOs are limited, even though FB-CSOs tend to have a higher likelihood of benefitting from funding than S-CSOs. Not surprisingly the category least eligible for funding is equipment or vehicles. The other three categories tend to be equally likely to be supported by external assistance (one might have expected that program costs would be more likely to be funded than administration, but this does not appear to be the case). Table 6: External financial assistance by type of expenditure (%) No funding Some funding Full funding All S-CSOs Salaries, stipends or incentives 45.8 31.8 22.4 100 Office and administration costs 37.1 42.3 20.6 100 Program costs, including supplies 42.3 38.0 19.8 100 Equipment or vehicles 61.6 18.6 19.8 100 FB-CSOs Salaries, stipends or incentives 36.2 44.7 19.3 100 Office and administration costs 29.4 39.7 30.9 100 Program costs, including supplies 24.7 52.6 22.7 100 Equipment or vehicles 59.3 26.7 14.0 100 Source: Authors’ estimation using CADRE database. Information was also gathered on broader changes in the funding environment (table 7). There is clear recognition among both types of CSOs that the availability of funding has increased between 2001 and 2005, with close to half of the CSOs stating that this was the 157 case, versus approximately 30 percent stating that funding had decreased or greatly decreased. The increase in funding availability as well as program expansion has also meant that the time allocated for fundraising has also increased, with more than 60 percent of both types of CSO stating that this was the case, versus less than 20 percent stating that time for fund-raising had decreased or decreased greatly. About 20 percent of the CSOs of both types did not perceive changes in either the availability of funds, or the time allocated for fund-raising. Table 7: Other changes in funding environment over last five years (%) Greatly Stayed Greatly All decreased Decreased the same Increased increased Availability of funds S-CSO 13.3 16.8 19.4 43.9 6.6 100 FB-CSO 12.1 19.8 22.0 35.2 11.0 100 Time for fund-raising S-CSO 5.1 9.3 23.2 38.4 24.1 100 FB-CSO 5.9 12.8 19.6 41.2 20.6 100 Source: Authors’ estimation using CADRE database. In table 8, the number of grant proposals submitted by FB-CSO is higher than for S- CSOs (6.8 proposals versus 5.0), which also means that the number of responses received and the number of grants approved is also slightly higher for FB-CSOs. The success rates for proposals is however slightly higher for S-CSOs at 33 percent, versus 27 percent for FB-CSOs. As to whether the activities run by the CSOs are driven by donor funding, close to two thirds of both types of CSOs indicated that this was very much the case, versus 10 percent stating not at all. Table 8: Success rates in funding proposals and dependency on funding (%) Number of proposals for funding Activities driven by funding opportunities Prepared Response Approved Not at all A little Very much All S-CSO 5.0 2.9 1.6 9.9 25.6 64.6 100 FB-CSO 6.8 3.5 1.9 10.8 25.2 64.0 100 Source: Authors’ estimation using CADRE database. Four additional questions from the survey are reported on in table 9. The first is whether the CSOs feel that donor priorities for funding have changed, with almost half of the CSOs stating that this is the case. The second question is whether CSOs have started new programs mainly because funding was offered for those activities, with about a third of the organizations stating that this was the case. The third question is whether CSOs have cut back on any areas of activity because of absence of funding, with approximately 60 percent responding in the affirmative, suggesting that while funding has indeed increased, there are also clear limitations set on the available funding. Finally, CSOs are asked about the proportion of their planned program that is already funded for the next 12 months. In many cases, the proportion seems rather small, suggesting a high level of vulnerability of both types of CSOs to any decrease in HIV/AIDS funding. Overall, while we will come back to some of the questions regarding the civil society response to HIV/AIDS in the next section, it is actually striking how similar the profile of the FB-CSOs and S-CSOs included in the CADRE-OSISA database are similar, at least on average. 158 Table 9: Perspectives on budgets and funding security (%) Perspectives on budget Share of funding needed secured for next year Donors New Cuts 0-25% 26-50% 51-75% 76-100% All Priorities Funding in Programs S-CSO 45.79 29.55 63.18 45.23 24.12 20.6 10.05 100 FB-CSO 42.53 33.64 57.8 43.69 20.39 23.3 12.62 100 Source: Authors’ estimation using CADRE database. INFORMAL FB-CSOS: KNOWLEDGE GAPS THAT REMAIN The conclusion from the previous section is that well-established or formal FB-CSOs operating at a community or national level tend to have the same access to donor resources than their S-CSO counterparts, and that they may also not be that different in other respects as well. However, this dataset addresses a particular kind of FB-CSO: those of the ‘NGO’ variety. The question remains as to the recognition received by the more ‘messy’ kinds of FB-CSOs, especially those operating at local levels and in less formal ways. It has been argued that community-level FB-CSOs, especially those of the congregational-initiative variety – are providing significantly more HIV/AIDS response than they are being supported for. This question cannot be analyzed with the CADRE- OSISA dataset, but it is worth discussing. Foster (2002) made the argument for the lack of funding for informal FB-CSOs in a study of community support to OVC. Foster describes a range of informal and everyday activities undertaken to support OVC in Africa - often started by small groups of individuals in a context of non-existent or weak public services, he describes these as “non-sensational and almost invisible to outsider and insider alike.� Later, for UNICEF and World Conference of Religions for Peace, Foster (2004) conducted a six-country study of the work of FB-CSOs supporting OVC. Based on interviews with 686 FB-CSOs (mostly congregations) in Kenya, Malawi, Mozambique, Namibia, Swaziland and Uganda, the study identified close to 350 initiatives that support more than 139,400 OVC. These FB-CSOs draw upon the help of more than 7800 volunteers, mostly through community-based initiatives involving spiritual, material, educational and psychosocial support. Foster estimates that this represents only a tiny proportion of the FB-CSOs working on OVC. Eighty-two percent of the initiatives identified were operating at a community level, through small congregation-based projects supporting on average less than 100 children each. More than half of these initiatives had been established after 1999; most have been initiated by community members themselves and did not receive any external support. As Foster (2004) notes: “Many congregations indicated that their only source of support consisted of contributions made by the members of their congregations. Faith-based community groups raise finances and materials to contribute to the families of vulnerable children…(But) the actual amounts of money raised by many initiatives are small and the ability of initiatives to provide meaningful material support to destitute families is limited. A few congregations received funding from their (religious coordinating bodies) but due to inadequacy of resources…funds received were minimal and did not meet the 159 needs…� Foster concludes that local-level FB-CSOs involvement in OVC is expanding rapidly without financial and technical assistance, and that, contrary to assumptions, the work is well-organized and not under-capacitated administratively. However, Foster also found that lack of funds was the major limitation facing 52 percent of the FB-CSOs. Funds were required mainly to provide direct assistance to children for school uniforms or food, or to provide incentives or transport costs for volunteers. Other studies have similarly argued that it is at the local community-level that FB-CSOs are most active, and most under-supported. A study of AIDS-engaged FB-CSOs in Zambia and Lesotho mapped a plethora of informal community initiatives and support groups providing care and support from physical care, to transportation and food parcels for those affected by HIV/AIDS – often utilizing only their own resources (ARHAP 2006). Agadjanian and Senn (2007) also focused on local congregations in Mozambique, and found that congregations were generally under-resourced, with low capacity for further engagement and few financial or material resources to share. Interviewees for this study reported that, when resources could be offered to the broader community, they usually were put toward transportation fares to the clinic or hospital, medical fees, drug costs, and, especially, funeral expenses. In an analysis of 162 community level FB-CSOs in South Africa, Birdsall (2005) noted that in contrast to a subset of long-established congregations, many faith-based projects have been initiated more recently: "especially over the last five years, there has been a largely spontaneous and often locally funded explosion of congregation and community level activity to respond to the HIV/AIDS crisis in many countries.� Again, there is not enough data to speak too strongly, but other studies have similarly reported that separate from the general growth of CSOs emerging as a result of HIV/AIDS funding (and FB- CSOs of the NGO variety as in the dataset above), there appears to have been an expanding subset of FB-CSO response - operating at a local community level (often located in rural areas), and primarily supported by the communities in which they are based, rather than tapping into the increased HIV/AIDS funding. For example, in Lesotho, a significant number of community support groups were identified in the ARHAP (2006) mapping study: these were informal, self-initiated, usually self-funded, deeply religious though not formally linked to any religious structure, and were among the most important local health providers for HIV/AIDS. While there is not enough data to compare the funding strategies of FB-CSOs country- by-country, it is suggested in the literature that community-based FB-CSOs do generally have different funding strategies from formal CSOs. For example, examining the South African national database of AIDS-engaged NGOs, Birdsall (2005) found that FB-CSOs source their support from a variety of sources with ‘donations’ (from within the church, from community members, and from other sources) by far the most commonly cited form of support: “No congregations report receiving support from government sources, from national or international donors, or from the private sector.� In a broader study of this database, Birdsall and Kelly (2005) note that while 40 percent of CSOs involved in the HIV/AIDS response reported receiving some funding from the government – none of the smaller informal FBOs in the survey did. Another example, from Malawi shows that in 160 the 2004 and 2005 period, 35 percent of all funds disbursed by the National AIDS Council of Malawi went to NGOs – of that, 25 percent to NGOs, 10 percent to CSOs and one percent to FBOs (Birdsall and Kelly 2007). In another study, Munene (2003) noted that 79 percent of churches and Christian NGOs responding to HIV/AIDS in Namibia received no outside funding. 55 All of these examples hint at different funding strategies and opportunities for small FB-CSOs than for CSOs more broadly. The lack of funding for small informal community-based organizations is not surprising - they often do not have clear structures and accountability mechanisms that would allow funding, or they simply lack the capacity to tap into resources. Difaem’s (2005) study of FB-CSOs’ access to HIV/AIDS resources found that lack of capacity was indeed the main obstacle. Lack of capacity is frequently described as relating to the areas of proposal writing, management of large-scale projects, monitoring and evaluation, and financial management. These are “key obstacles in FBOs not being able to apply successfully for grants from big donors…there are positive examples where agencies were able to access funding due to the availability of capacity…However, the majority of respondents mentioned a lack of capacity as major obstacle.� The logical implication of this is not only that there needs to be a stronger focus on capacity building targeted at these initiatives that are informal, but also that there needs to be improved understanding of the intermediary mechanisms that could support such initiatives. Foster’s (2004) study of local support of OVC suggested to better finance ‘religious coordinating bodies’ (RCBs) that may be well-placed to play a more significant role in supporting congregation-level work: “Congregations have the capacity to implement OVC support activities and receive funds but most receive no external support. Funding should therefore be provided through small grants funds operated by RCBs to support activities initiated by congregations. Donors should ensure that a majority of RCB funding is spent at community level.� (Foster 2004). Taylor (2005) also argues that where funds have reached local congregational initiatives, this has been “where there is a facilitating intermediary that understands their situation but is also able to respond to the requirements of higher level funding processes. In Zambia, this has been achieved through setting up a dedicated stream of funding for faith-based organisations. In Kenya, it has come with decentralization to enable community groups to access funds.� Others are similarly arguing that what is urgently required is a better understanding of this ‘intermediary’ role – not only for the channeling of funds, but also the ‘incubation’ of local initiatives without destroying them (Cochrane 2011). In relation to FB-CSOs, this intermediary role can be enacted by a number of different entities: religious coordinating bodies, denominations, government ministries and platforms, primary grantees and sub-grantees (such as World Vision or Christian Health Associations), and international partners (INGOs). What is clearly required then is 55 In countries where Islam is prevalent, Zakat and other direct payments from Islamic communities play a large role in the funding of FB-CSOs. In Chipata, Zambia, the local Muslim community did not have many associations engaged in HIV and AIDS response, but was financing a wing of the local government hospital as part of their social responsibility (ARHAP 2006). Yet these types of payments are rarely accounted for or recognized. 161 improved evaluation of how such support does or does not flow down to the community level, and what impact it has, or how effective local FB-CSOs indeed are. CONCLUSION The evidence to-date on whether FB-CSOs have been able to access various sources of funding for HIV/AIDS has been limited. Our analysis of the survey implemented by Birdsall and Kelly (2007) suggests that among formal and established CSOs working on HIV/AIDS in developing countries, donor funding has increased and is now significant, and it has enabled the CSOs to expand their activities. The results suggest also that the profile of relatively well established FB-CSOs and S-CSOs (those most likely to be included in this dataset) is rather similar, both in terms of the areas on which they work, and in terms of their sources of funding as well as expenditure levels. This suggests that within the broader so-called ‘faith sector’, formal FB-CSOs that may actually look and act similar to secular counterparts may well be on par with secular CSOs in terms of the sources of funding that they are able to access. Yet while strides have been made for enhancing funding mechanisms that are inclusive of different types of formal CSOs, whether faith-based or secular, there are challenges, especially for smaller informal initiatives. Small, idiosyncratic and informal community initiatives are more difficult to know, measure and support, but they are fundamental to the support provided to those who suffer from HIV/AIDS, or are at risk. It is likely that a larger share of these initiatives are faith-inspired than is the case for formal CSOs, given that the bulk of community-level work often spontaneously emerges out of ‘congregations’ or linked to individuals motivated by personal faith. These activities are frequently driven by a local faith leader, or a collection of women who may congregate after choir practice. These activities are often locally funded, and need support if they are to be sustained. This local response may actually be where the real strength of the 'faith sector' lies - in the wealth of programs and initiatives that are rooted in community and which have a double impact of service provision and behavioral change potential through faith commitment. But how to reach these groups and find appropriate ways to support them remains a challenge for policy interventions. It may well be a mistake to try to ‘secularize’ these informal activities, or even to formalize them. In fact, it is likely to be both impossible and undesirable to attempt to disentangle these activities from the forces of ‘faith’ which motivate and congregate these individuals and communities (ARHAP 2006). However, in relation to international funding strategies and impetus, if there is any lingering resistance to fund activities which seem too ‘churchy’ – or too closely related to proselytism – then this is where that reluctance will be found (Olivier and Clifford 2011). This reluctance certainly seems to be less evident in relation to large national and international FB-NGOs, who often are (increasingly) indistinguishable as faith-based or not in terms of their operations, as suggested by the analysis provided in this paper. So the remaining question is really what to do about informal initiatives responding to HIV/AIDS. One clear suggestion that warrants further consideration is the role of 162 intermediary or bridging entities who can mediate between the national and international funding structures and the smaller often innovative community-level activities. Such bridging financial mechanisms have been noted in a variety of forms, including distinctively faith-based exemplars, such as some denominational structures, some national faith-based health networks, and some FB-CSOs who provide an 'incubation' function to local initiatives. There is clearly no one-size-fits-all funding strategy for supporting this segment of the 'faith sector' response to HIV/AIDS, especially at a community level. But there is still some significant learning to be done to understand how FB-CSOs are distinctive service providers as a result of their faith-inspired nature, and what funding allocations and strategies should be enacted as a result. Finally, one last comment should be made about the current context for funding formal CSOs, whether faith-based or not. The data provided in this paper confirms that in African countries with a substantial HIV/AIDS epidemic, there has been a rapidly scaled up response to HIV/AIDS from CSOs, particularly over the period 2000-2005. This response has been observed across the full range of CSOs, from formal national-scale CSOs and networks to the proliferation of new community-level initiatives and programs, and among both faith-based and secular organizations. While some of these initiatives argue that they emerged as a response to need, it is also clear that many emerged as a result of the greatly increased availability of HIV/AIDs funding over that period. Yet the data also show that many of the CSOs that have been created to respond to HIV/AIDS, or that have included HIV/AIDS in their programs in part because of the availability of funding, remain fragile. In a context where funding for HIV/AIDS is becoming more scarce, the questions about whether those newly created CSOs will be able to survive, or how they might be ‘redirected’ so that their capacity and experience is not lost remains worryingly unclear. REFERENCES Agadjanian, V., and S. Sen. 2007. “Promises and Challenges of Faith-Based AIDS Care and Support in Mozambique.� American Journal of Public Health 97 (2):362-366. African Religious Health Assets Programme (ARHAP). 2006. Appreciating Assets: The Contribution of Religion to Universal Access in Africa. Cape Town. Benn C. 2011. “The Continued Paradigm Shift in Global Health and the Role of the Faith Community.� In B. Schmid and et al, eds, When Religion and Health Align: Mobilizing Religious Health Assets for Transformation, Cluster Publications: Pietermaritzburg. Birdsall, K. 2005. “Faith-Based Responses to HIV/AIDS in South Africa: An analysis of the Activities of Faith�Based Organisations (FBOs) in the National HIV/AIDS database. Johannesburg: Centre for AIDS Development, Research and Evaluation. 163 Birdsall, K., and K. Kelly. 2005. “Community Responses to HIV/AIDS in South Africa: Findings from a Multi�country Survey.� Johannesburg: Centre for AIDS Development, Research and Evaluation. _____. 2007. “Pioneers, Partners, Pproviders: the Dynamics of Civil Society and AIDS Funding in Southern Africa.� Johannesburg: Centre for AIDS Development, Research and Evaluation and the Open Society Initiative for Southern Africa. Bonnel, R., et al. 2012. Funding Mechanisms for Civil Society: The Experience of the AIDS Response. World Bank Study, World Bank, Washington, DC. Center for Interfaith Action on Global Poverty (CIFA). 2010. “Many Faiths, Common Action: Increasing the Impact of the Faith Sector on Health and Development - A Strategic Framework for Action. Washington, DC. Cochrane, J. R. 2011. “Trustworthy Intermediaries: Role of Religious Agents on the Boundaries of Public Health.� In B. Schmid and et al, eds, When Religion and Health Align: Mobilizing Religious Health Assets for Transformation, Cluster Publications: Pietermaritzburg. German Institute for Medical Mission (DIFAEM). 2005. “Global Assessment of Faith- based Organisations’ Access to Resources for HIV and AIDS Response.� Tubingham: German Institute for Medical Mission. Foster, G. 2004. “Study of the Response by Faith-Based Organizations to Orphans and Vulnerable Children.� New York: World Conference of Religions for Peace and United Nations Children Fund. GFATM. 2008. “Report on the Involvement of Faith-Based Organizations in the Global Fund.� Geneva: The Global Fund to Fight AIDS, Tuberculosis and Malaria. Haddad, B., J. Olivier, and S. De Gruchy. 2008. “The Potential and Perils of Partnership: Christian Religious Entities and Collaborative Stakeholders Responding to HIV and AIDS in Kenya, Malawi and the DRC.� Cape Town: Africa Religious Health Assets Programme. Keough L,, and K. Marshall. 2007. “Faith Communities Engage the HIV/AIDS Crisis: Lessons Learned and Paths Forward.� Washington DC: Georgetown University, Berkley Centre for Religion, Peace and World Affairs. Mandi, F. 2006. “Planning, Development and Supporting the Faith-Based Health Workforce.� In African Christian Health Associations' Human Resources for Health Mini-Forum, Nairobi: The Capacity Project and Medicus Mundi International. Munene, J. A. W. 2003. “A Situational Analysis of the Church's Responses to HIV/AIDS.� Nairobi: PACANet. 164 Oliver, J. 2010. “In Search of Common Ground for Interdisciplinary Collaboration and Communication: Mapping the Cultural Politics of Religion and HIV/AIDS in sub- Saharan Africa.� Cape Town: University of Cape Town, Olivier, J. 2011. “`An FB-oh?’: Mapping the Etymology of the Religious Entity Engaged in Health.� In J. R. Cochrane, B. Schmid and T. Cutts, eds, When Religion and Health Align: Mobilizing Religious Health Assets for Transformation, Pietermaritzburg, South Africa: Cluster Publications. Olivier, J., and P. Clifford. 2011. “Religious Community Care and Support in the Context of HIV and AIDS: Outlining the Contours.� In B. Haddad, editor, Religion and HIV/AIDS: Charting the Terrain, Pietermaritzburg, South Africa: UKZN Press. Olivier, J., and Q. Wodon. 2012. “Layers of Evidence: Discourses and Typologies for Faith-Inspired Community Responses to HIV/AIDS in Africa.� In J. Olivier and Q. Wodon, eds, Mapping, Cost, and Reach to the Poor of Faith-inspired Health Care Providers in sub-Saharan Africa, HNP Discussion Papers, World Bank, Washington DC. Olivier, J., J. R. Cochrane, and B. Schmid. 2006. “ARHAP Literature Review: Working in a Bounded Field of Unknowing.� Cape Town, South Africa: African Religious Health Assets Programme, University of Cape Town. Rodriguez-Garcia, R., R. Bonnel R, N. N’Jie, J. Olivier, F. Pascual, and Q. Wodon. 2011. Analyzing Community Responses to HIV and AIDS: Operational Framework and Typology, Policy Research Working Paper No. 5532, World Bank, Washington, DC. Schmid, B., E. Thomas, J. Olivier, and J. R Cochrane. 2008. “The Contribution of Religious Entities to Health in sub-Saharan Africa.� Cape Town: African Religious Health Assets Programme. Taylor, N. 2005a. “Many Clouds, Little Rain? The Global Fund and Local Faith-Based Responses to HIV and AIDS.� Teddington: Tearfund. _____. 2005b. “The Warriors and the Faithful: The World Bank MAP and Local Faith- Based Initiatives in the Fight Against HIV and AIDS.� Teddington: Tearfund. _____. 2006. “Working together? Challenges and Opportunities for International Development Agencies and the Church in the Response to AIDS in Africa.� Teddington: Tearfund. _____. 2007. “DFID, Faith and AIDS: A Review for the Update of Taking Action.� London: UK Consortium on AIDS and International Development. 165 Tearfund. 2006. “Faith Untapped: Why Churches Can Play a Crucial Role in Tackling HIV and AIDS in Africa.� Teddington: Tearfund. United Nations Population Fund (UNFPA). 2004. “Culture Matters: Working with Communities and Faith-Based Organizations: Case Studies from Country Programmes.� New York. 166 CHAPTER 10 CHRISTIAN ORGANIZATIONS’ PLACE IN MULTISECTORAL HIV/AIDS RESPONSE: KENYA, MALAWI AND THE DEMOCRATIC REPUBLIC OF CONGO Jill Olivier and Beverley Haddad University of Cape Town and University of KwaZulu Natal 56 Partnership with faith-based organizations (FBOs) for improved HIV/AIDS response in Africa sounds logical and useful. However there are lingering questions about the ‘faith sector’ and a history of isolationism from other civil society collaborative. This article examines the possibilities and challenges of collaborative relationships in the struggle against HIV/AIDS in several countries in Africa, specifically between Christian religious entities and national governments and donors. These collaborative partnerships are considered within the framework of the Three Ones policy promoted by UNAIDS. The argument presented here distils and summarizes some of the broader comparative findings that emerge from a research study conducted in Kenya, Malawi, and the Democratic Republic of the Congo. We consider here both the real challenges facing collaboration and the potential such partnerships bring to HIV/AIDS response. INTRODUCTION There has been a strong drive for multisectoral collaboration for HIV/AIDS engagement for many years now. Within that, an emerging interest area has been the potential for improved engagement with religious leaders and religious entities (REs - often called faith-based organizations or ‘FBOs’ 57) for improved multisectoral response. There is, of course, a general drive to ensure that civil society is engaged as a ‘more equal’ partner in multisectoral collaboration. For example, a review of progress on the Three Ones by UNAIDS in 2005 found that in most countries reviewed, civil society is not an equal partner – and “even where consultative procedures have been put in place in relation to national plans, key stakeholder groups – such as those representing women, NGOs, faith- based organizations (FBOs) and people with HIV – are not fully engaged� (Birdsall and Kelly 2007). However, REs have historically often been working in isolation in health and development, unaligned and without the acknowledgement of national- or 56 This paper reports on research supported by Tearfund and UNAIDS (full findings and recomendations can be read in Haddad et al 2008). The authors acknowledge the advisors and participants who gave generously of their time and perspectives during the research process. 57 The term (Christian) religious entity (RE or CRE) seeks to capture the broad range of tangible RHAs, incorporating religious facilities, organizations as well as practitioners, both bio�medical and traditional. This encompassing term is necessary in order to be able to speak to the more traditional religious entities such as faith�based organizations, as well as those more amorphous entities such as individual traditional healers. 167 international-level policy makers. Certainly, in the early stages of the HIV/AIDS pandemic, REs were often ignored or treated with caution and even suspicion. Although such tensions remain, perspectives have mostly changed from around 2000, with REs getting significantly more interest at a national and international policy level. 58 As Benn (2003) notes: “…for decades [they] were at best tolerated but not actively supported. Now there is almost a competition among big secular donors to fund the best programs.� The increased interest is most commonly framed around an interest in their potential contribution to health and development, their ties to community, and their potential impact on behavior change. However, because of such diverse and rapidly evolving opinions towards REs, and their potential in HIV/AIDS response, it is now unclear just how far dialogue with REs has resulted in improved or integrated multisectoral practice. Or asked differently, do religious stakeholders (or the ‘faith sector’) experience multisectoral collaboration differently to other civil society partners at a national level? Little is actually known about collaborative practice with REs in the context of HIV/AIDS in Africa. There are certainly increased reports of religious leaders being involved in national HIV/AIDS campaigns, more research reports on the ‘religious response to HIV/AIDS in Africa’ (see Olivier et al 2006), and ‘FBOs’ are now increasingly named as being represented on national aids councils and bodies. For example, GFTAM (2008) notes that by 2006, most Country Coordinating Mechanisms (CCM) had FBO representatives and nine FBOs were allocated funds as principle recipients of GFTAM. However, there is limited analysis of how collaboration occurs in context: how REs are partnering at a national and local level with donors, governments, other NGOs and communities, and what impact this has on practice. In 2007 Tearfund and UNAIDS commissioned the African Religious Health Assets Programme (ARHAP) to conduct a three-country study to examine the relationship between the work of REs and the national AIDS strategy, as well as the strengths and weaknesses of the relationships with other collaborative stakeholders. The Promise and Perils of Partnership (Haddad et al 2008) reports on the possibilities and challenges of collaborative relationships in the struggle against HIV/AIDS in Africa, specifically between Christian religious entities (CREs) on the one hand, and national governments and donors on the other. These collaborative partnerships were considered within the framework of the Three Ones policy then promoted by UNAIDS, the principles of which include one agreed action framework, one national coordinating authority, and one monitoring and evaluation system. The goal of this research was to engage with nationally-representative CREs in Kenya, Malawi and the Democratic Republic of the Congo (DRC) to better understand their contribution to the promotion of Universal Access. Three main research methods were utilized: desk review; participatory workshops (two in each country); and self- administered questionnaires. The participatory phase of this research process was designed to identify findings related to the context in which CREs are working, their 58 See Olivier and Wodon (2012) for a more detailed outline of the changes in perspectives towards REs in the context of HIV/AIDS. 168 work in promoting Universal Access, the strengths and weaknesses of collaborative partnerships between CREs and other stakeholders, and the challenges and potential of those collaborations. Participants were separated into two workshop levels: identified representatives of main CREs engaged in HIV/AIDS at a national scale, and stakeholder representatives from funders, government and other NGO and religious groups). This separation importantly provided participants the freedom to respond anonymously on sensitive issues, although still tested by peers. This approach has some limitations – for example, a narrowed focus on Christian entities (although other faith groups were included as participant stakeholders); a narrowed focus to the main CREs with a national- level AIDS-response response (limiting perspective on community-based initiatives); and the expected limitations of secondary literature and data for establishing a comparative baseline. In addition, while these countries were chosen in order to provide some country-comparative perspective, findings should certainly not be generalized to all of Africa – especially given the acknowledged diversity of religious- and country contexts. In the following article we draw from this study some of the main comparative findings. We do not present the country-specific findings and recommendations, nor do we address the findings on the nature of HIV/AIDS activities undertaken by these CREs). The purpose here is to reflect on the nature of REs’ engagement in multisectoral collaboration on HIV/AIDS at a national level. FINDINGS ON MULTISECTORAL COLLABORATION WITH CHRISTIAN RELIGIOUS ENTITIES Country comparison is a challenging exercise especially considering that this research strongly confirmed how contextually-specific the nature of each ‘faith sector’ is, and how varied the work of CREs tends to be. It is, however, possible to note some broad similarities between Kenya, Malawi and the DRC - in relation to factors underpinning their collaborative context. Each of these sub-Saharan African (SSA) countries have established HIV/AIDS epidemics which are acknowledge as being tied to broader issues of poverty, socio-economic challenges, and other diseases such as malaria and TB. Each country faces particular national challenges that exacerbate their epidemic, for example, in Malawi poverty and food security are key, in the DRC conflict and security, in Kenya the challenges of development and financing of ART (as well as political turmoil 59). They all have severely weakened health systems and, among other challenges, are faced with significant human resource crises in the health sector. They are all also Christian majority countries, and have significant histories of mission-based religious involvement in health and education (although DRC with its Francophone colonial past is slightly different to Anglophone Kenya and Malawi). The result of which is that in all three countries, REs are significantly engaged in health (and education) provision. In addition, all three countries have a characteristically complex ‘faith sector’ which includes an array of international organizations, national organizations, ecumenical or denominational bodies, health desks, networks, community organizations, congregations and other initiatives. 59 It should be noted that the fieldwork for this study took place during an upsurge of political and social turmoil in Kenya (2007-2008). 169 All three countries also have (at least on paper) a national emphasis on multisectoral response to HIV/AIDS, and a national commitment to collaboration driven by the principles of the Three Ones. Indeed, in a study of SSA countries, Birdsall and Kelly (2007) note that multisectoral collaboration has been strengthened to a point where it is one of the key strategies worldwide, and “nearly all national plans are now multisectoral in design.� There are also common challenges to multisectoral response. As the National AIDS Control Council (NACC) in Kenya notes (2005): “Kenya is implementing a successful multi-sectoral response to HIV/AIDS…There is an increasing understanding and willingness to cooperate among stakeholders across Government, civil society, the private sector and development partners…But progress cannot be taken for granted; enormous challenges remain.� All three countries juggle a complex donor-partner environment, with many different donors engaged in HIV/AIDS-response, and related funding being driven through different channels (albeit with a recent demand for harmonization through basket funds.) Typically, CREs engaged in health provision in all three countries have faced significantly changed financial contexts over the last three decades – with traditional sources of funding being reduced. As Mandi (2006) notes, “much of the support FBOs were getting from the big congregations, churches and donors, as well as the assistance received from the government from as far back as the fifties and sixties, came to an end.� However, it is also worth noting that as a result of newly available funds (and need) – there has also been a noticeable ‘explosion’ of civil society organisations engaged in HIV/AIDS (including CREs) since around 2000 – which was confirmed by study participants. 60 In this context, we consider some of the broad comparative findings raised about multisectoral collaboration on HIV/AIDS inclusive of CREs in Kenya, Malawi and the DRC. LACK OF (EXPECTED) ANTAGONISM TOWARDS CRES In contrast to reports of national-level and international antagonism or indifference towards REs that would have resulted in them being sidelined apart from civil society or multisectoral response, one key finding of this research was that in all three countries there was substantial and consistent goodwill displayed towards CREs by stakeholders – even when stakeholders were given several opportunities to express reservations anonymously. This may seem a ‘light’ and common-sense finding. However, given the tensions in the past, and the reported national and international disinterest and suspicions of CREs – it was expected that stakeholders would display greater reservation – resulting in more CREs being sidelined or working in isolation. Instead, stakeholders (including representatives of government, funders, civil society and other religious groups) were quick to point out what they understood to be the characteristic strengths of CREs and their potential for engagement in HIV/AIDS response: ⋅ The reach and credibility of CREs, especially in rural areas. This was recognized as being based on their durable and long-term location in communities (not ‘drop-in’ NGOs with a limited timeframe), and their well-organized structures in rural or 60 This was confirmed by study participants in 2007 and 2008. Since then, however, there are some indications that this ‘boom’ might now be slowing. 170 hardship areas where, for example, a government participant from Kenya noted “...even the government does not have (these) structures in the rural areas.� In Malawi stakeholders also noted that CREs are one of the few organs of civil society that were accessible and trusted by ‘grassroots people’. DRC stakeholders understood the credibility of CREs to be linked to their possessing a powerful moral voice. ⋅ The provision of psychosocial and spiritual support. Was recognized as a particular strength that was not being provided by others. ⋅ Their resources, both human and material. For example, access to volunteers and committed staff, and well developed facilities and structures in areas in which there were otherwise none. Stakeholders in all three countries perceived that CREs had access to particular regional and international networks which provided resources that were otherwise not accessible to other civil society organizations. Of course, stakeholders did raise concerns about certain kinds of CREs, and some characteristic weaknesses which created obstacles to more effective multisectoral collaboration: ⋅ CREs often display an ethos of dogmatism and conservative belief system. This manifests through patriarchal and hierarchical structures, and sometimes results in an unwillingness to deal with high risk groups such as commercial sex workers. There were also concerns that some CREs were misrepresenting the epidemic as a result of ‘inappropriate theological understandings’, for example, being unwilling to promote condoms due to a conservative theological position. ⋅ Characteristic lack of organisational capacity relating to the management of human and financial resources. This was frequently described as resulting in a lack of accountability and reporting to funders. A perceived over-emphasis on valuing of religious human resources was also noted, for example, when staff appointments were made on the basis of the person being a ‘good Christian’ rather than a good administrator. “They hold to a theological value that says the expression of faith is more important than skill� (Kenya, stakeholder participant). 61 ⋅ Networking, competition and inter-faith structures. Stakeholders were also concerned by the competition witnessed between CREs, and the lack of strong interfaith bodies and sustained leadership representation, making it challenging to know who to work with. However, even when stakeholders raised such concerns, this was done in a balanced manner. For example, stakeholders noted that a conservative theology or ethos might create difficulties, but it also gave weight to a (needed) moral authority and was balanced by a theological emphasis on the poor and vulnerable. Stakeholders also noted that the selection of human resources based on ‘spiritual’ criteria created difficulties – but also resulted in a theological commitment and a stronger work ethic. Indeed, the weaknesses and challenges identified by stakeholders were made in a constructive tone. They were reported mainly as operational difficulties to be overcome which was felt to be a worthwhile exercise. 61 Ironically, Christian entity participants, however, saw this same concern as a result of ‘lack of trust’ in their ability to deal appropriately with large sums of money. 171 Broadly-speaking, CREs were similarly positive about, and committed to improving collaboration with government, donors and other stakeholders. However, there were also deep frustrations, the main ones being: ⋅ Lack of proper collaborative processes and proper representation on some forums. We discuss this in more detail below, but in all three countries, CREs did not feel that the inter-faith collaborative bodies that had been set up to represent the ‘faith sector’ were adequately representative. For example, a Kenyan CRE participant notes, “The whole perception that KIRAC [The Kenya Inter-Religious AIDS Consortium] represents the interfaith community is a mis-representation…it is referred to all the time like that, but [it is] not.� ⋅ Increasing bureaucracy around reporting and lack of stakeholder coordination. CREs were deeply frustrated with the multiple reporting demands that they faced. This was not expressed as unwillingness to report, but rather a frustration with weak information systems and frequently conflicting demands. For example, CREs described situations of donors ‘fighting over’ results as a DRC participant noted “Funders want results for themselves…there is competition between funders…they don’t like [to give to organizations with] multiple funders.� Also noted was a serious lack of donor coordination, so that M&E requirements were different and time- consuming; questions about the necessity of providing information to those who they were not receiving any support from; and a significant lack of feedback from those to whom they were providing information (including government). As one participant from Kenya said “...we have given accounts three times, we give them the documents and after three months they ask again. I have actually told my people to stop accounting, because you don’t know whether we are accounting for other organizations.� And another from Malawi “...they are not networking at the donor level…and this is trickling down and creating competition and fighting...� Finally, several CREs noted that there was a lack of clarity where some partners began as donors, and then shifted to become implementers (in competition) once the funds were made available. ⋅ The lack of financial commitment from government and over-stretched expectations. CREs also noted that there was a lack of financial commitment from government (for example, that budgets attached to national policies lacked conviction). Furthermore, there were concerns that government and donors had too-high expectations of the capacity of faith communities to carry the burden of some kinds of care and support. For example, in all three countries it was noted that CREs were increasingly overwhelmed with the responsibility they carried in relation to the care of orphans and vulnerable children. ⋅ Donor-driven agendas and short-term plans. CREs were grateful for the support they had, but felt that HIV/AIDS funding was largely driven by international agendas. One Kenyan participant described this situation as “...another form of ‘colonial intervention.� Most CREs felt frustration at the way their actions were dictated, without a proper appreciation of local realities. They also expressed concern with the way agendas (and therefore funding allocations) were often changed mid-project, without any recourse. As another Kenyan participant noted, “One minute they are talking about a thing, and then they shift to another and do not even tell us...(they) 172 say what needs to be done and give no room for other options…they threaten to pull out if what they want is not allowed.� It was felt that the withdrawal of funding was used as a tool of power that was constantly felt, and often paralyzed local stakeholders. A further concern was the way funds often have to be utilized within a short time-period which inhibited long-term planning. Nevertheless, balancing these concerns and poor experiences, CREs were grateful for the funds that allowed them to do their work – including new funding opportunities (such as access to PEPFAR and GFTAM support). They were particularly positive about long- term donor partnerships, expressed as trusted relationships, which enabled them to develop a long-term vision for their work without getting tied into short-term interventions. CREs were also particularly appreciative of donor partners who were ‘flexible’ which enabled a more timely response to urgent needs, and did not hold them too rigidly to donor agendas or vertical program requirements. In addition to financial resources, participants equally valued the capacity-building opportunities offered by partners as well as the technical expertise they gain from these. Participants thus felt that collaborative relationships were strengthened by partnerships that did not only involve funding, but included mutual sharing of resources, experiences and ideas. MULTIPLE MECHANISMS FOR COLLABORATION It was found in all three countries that there were multiple mechanisms for collaboration in place including a variety of councils (interfaith AIDS councils, interfaith councils, national HIV/AIDS councils, Global Fund councils, district AIDS task forces, and the like), different policies, and monitoring mechanisms. Specific to HIV/AIDS multisectoral collaboration, this study confirmed that CREs collaborate with government through two primary avenues. Firstly governments’ collaboration with religious health (service) providers such as national faith-based health networks (NFBHNs), or the Christian Health Associations (CHAs), most commonly managed through Ministries of Health. The focus on HIV/AIDS through this avenue is usually medical in nature – especially relating to the provision of ART. Secondly, governments tend to collaborate with REs that are perceived to be part of ‘civil society’: this relationship managed (on the governments’ side) through national AIDS Commissions or Councils and multisectoral committees. The comparative success of these very different mechanisms is not well documented – but of interest here is the way collaboration with CREs in response to HIV/AIDS is carried out in both ways. 173 Table 1: Main mechanisms for HIV/AIDS collaboration of CREs with government 62 Health-services collaboration with government Civil society collaboration with government Kenya Ministry of Health (MOH): direct relationship, National AIDS Control Council (NACC) is including MOUs, with Christian Health responsible for ensuring civil society Association of Kenya (CHAK) and Kenya engagement in the national HIV/AIDS Episcopal Council (KEC). Collaboration response. CREs are represented in NACC mainly organized through working groups, collaboration – mainly organised through such as the Technical Working Group, representation on the Kenya Inter-religious Ministry of Health-Faith Based Health AIDS Consortium (KIRAC). Services (MOH-FBHS-TWG) Malawi Ministry of Health and Population (MOHP): National AIDS Commission (NAC) is direct relationship, including MOU, with responsible for ensuring civil society Christian Health Association of Ghana engagement in the national response. CREs are (CHAG) represented in NAC collaboration, mainly represented by the Malawi Interfaith AIDS Association (MIAA), set up with government funds as a representative body, making funding recommendations to government. DRC Ministry of Health: direct relationship with Programme National Multisectoriel de Lutte Eglise du Christ au Congo�Direction des contre le VIH/SIDA (PNMLS: National Multi- Oeuvres Médicales (Protestant Church of Sector Program against HIV and AIDS) - Congo Department of Medical Works) and established in 2004 and is tasked with SANRU. HIV/AIDS collaboration is split coordinating all sectors involved in HIV/AIDS between PNLS and PNMLS: Programme response (public sector, MOH, private and National de Lutte contre le SIDA (PNLS - enterprise sectors, NGO and faith National AIDS Control Programme): communities, as well as provincial and Established early in the epidemic (around community representatives). The National 1987) holds national responsibility for the HIV Council of Interfaith-based Alliance (CIC): has response, especially ART scaleup, health been set up to integrate religious and sector response and surveillance. government responses to HIV/AIDS but is not widely representative. Source: authors’ compilation This study also verifies the increasingly important role of the NFBHNs in all three countries. These country-level umbrella networks of health providers clearly play a central role in coordinating and strengthening collaboration at a number of levels: forging collaboration between their own members, between their members and the government, between other NFBHNs in their own countries and regionally as well as the broader spectrum of secular and non-profit actors. They also create a stronger platform for advocacy and technical support (see Schmid et al 2008). What is also noticeable, is that NFBHNs in these three countries hold a collaboratively strong position with government, in comparison with other CREs who are mainly grouped as civil society. For example, the health providers represented by NFBHNs (CHAK, KEC, CHAM, ECC- DOM/SANRU) collaborate with government in relation to HIV/AIDS response through both avenues described above: directly with the MOH and also through NACs and represented by interfaith councils. They generally enjoy a more direct relationship with government, as well a more varied relationship with more ‘parts’ of government. 62 Note that there are many other collaborative mechanisms in place in each of these countries – different government task forces and councils – we only highlight here what were identified as the two main mechanisms for CREs to collaborate with government (see Annex) 174 In addition, although this study was focused on national-level collaboration, participants in all three countries noted the different nature of multisectoral HIV/AIDS collaboration at a district level. In previous research, ARHAP (2006) also found strong district level multisectoral collaboration in Zambia (for example through District AIDS Task Forces) - although weak interfaith collaboration at the same level, especially with traditional religious leaders, between Christian and Muslim initiatives, and between mainstream and charismatic Christian groups. In Malawi, CRE participants agreed that multisectoral collaboration worked best at a district level (the same level that stakeholders noted that CREs provided a particular contribution to HIV/AIDS response). However, participants in all countries felt that district-level collaboration still had challenges such as the lack of capacity to drive collaboration at this level, the lack of legitimate and recognized intermediary bodies to support this work, and the difficulty of ensuring that communication and information filtered down to community levels. Broadly speaking, CREs were recognized as having characteristic collaborative methods and mechanisms which were different to other civil society organizations. In the first place, CREs tend to have more complex collaborative relations – with more and different collaborative partnerships and networks to manage. Stakeholders also felt that CREs tended to collaborate more broadly, with access to regional and international networks that other civil society organizations did not. Many CREs have historic relationships with denominational and associational networks. In all three countries, the Catholic Church provides a good example of this, with links between local initiatives and denominational bodies, episcopal councils and international agencies (such as Caritas, CAFOD or Catholic Relief Services). Others have noted the strong associational infrastructure existing at the national and local levels between FBOs involved in HIV/AIDs in other SSA countries. For example, Liebowitz (2002) gives the example of the Anglican Church of Uganda, which forms an associational infrastructure that includes the Mothers’ Union, Fathers’ Union, youth groups and other church-related associations, and argues that this provides an advantage to organizations within this umbrella, particularly in rural areas where religious groups are often based. Participants in this study (Haddad et al 2008) noted that funding partners are also drivers of multisectoral collaboration in the way they act as ‘subregional networks’, building relationships between recipients. Others have demonstrated that CREs tend to operate in regional networks that spread across national borders. Parry (2002) for example, notes that Kenya acts as a regional hub for East African HIV/AIDS response: “Kenya hosts numerous faith based Associations, Councils, Organizations, Secretariats, Consortia and Networks that are not only national but also regional and international. Because they are umbrella organizations, or have chapters in many countries, they represent a huge constituency of some millions of believers and as such have the potential for enormous influence.� However, not enough is known about the way these networks operate, or the precise way that denominational bodies and international FBOs operate as sub-regional networks, and certainly more research would be useful here. 175 STRENGTHENING MECHANISMS FOR ‘INTERNAL COLLABORATION’ Another key finding of this study was that ‘internal collaboration’ within sectors tends to be poor, and needs to be strengthened as a precursor to multisectoral collaboration. By ‘internal collaboration’ we mean the collaboration within sectors (between REs, between donors and between government departments and bodies). For example, CREs noted significant frustrations with what they perceived to be a lack of coordination between stakeholders (donors and government). This extended beyond the issue of basket funding, to operational concerns about conflicting agendas, mismatching reporting requirements and the like. While all three countries had national HIV/AIDS coordinating mechanisms in place (see Table 1 above), they also noted that in practice they collaborated separately with a wide range of different government departments. Both CREs and stakeholders noted significant challenges to interfaith and ecumenical collaboration in all three countries. None of the coordinating structures that had been set up to draw the ‘faith sector’ together were functioning optimally. While government stakeholders were assuming that these structures were the legitimate representative body through which to channel resources and information to CREs, the CREs did not recognize these as (currently) having that authority or being authentically representative. This was especially the case for those interfaith HIV/AIDS coordinating structures that had been set up by government, which were perceived as an externally imposed collaborative structure. There were several lessons from this, a primary one being that governments need to ensure that when electing representatives to such bodies, it must adopt a strictly participatory process in order to avoid relating to particular individuals who prove unhelpful to an overall strategic response to the HIV/AIDS epidemic. That is, the establishment of a coordinating interfaith HIV/AIDS structure should not be the culmination of, or replace a broad-scale participatory process, but should continue to be (re)examined, especially in terms of its representivity. This is particularly the case with the ‘faith sector’ which as a category contains a broad and complex range of organizations and individuals – including some with difficult relational histories. Participants clearly demonstrated that while there are many cases of historic ecumenical cooperation (for example between Kenyan Protestant and Catholic health providers), other interfaith and ecumenical relations were marred by suspicion and historic disassociation (for example, between ‘mainline’ churches and the charismatic movements). Some stakeholders saw the major weakness of CREs as the competition that exists between them, which in turn leads to a lack of credibility. Indeed, fragmentation within ‘sectors’ (faith/donor/government) was in each case seen to lead directly to competition within that sector and also indirectly to competition between other related partners. The researchers felt strongly that this context of competition significantly hindered effective collaboration, limiting commitment to the Three Ones, particularly in relation to M&E. It was therefore strongly recommended, based on this study, that for multisectoral HIV/AIDS collaboration to be strengthened, the collaborative relationship within each sectoral grouping needed to be strengthened. Or said differently, for multisectoral 176 collaboration to be effective, stakeholders first need to have good relationships and cooperation within their own groupings. DIFFERENT ‘STAGES’ OF MULTISECTORAL HIV/AIDS COLLABORATION Finally, we considered the country-specific findings in Kenya, Malawi and the DRC, and matched them against key components of the Three Ones policy. Through this process, while it is apparent that the collaborative contexts are specific to each country, it is also possible to describe these countries as being at three different ‘stages’ of multisectoral collaboration as a result of internal and external factors, which suggests that some pattern for policy-level engagement exists. DRC: Early and fragile collaborative stage The DRC case-study highlighted just how difficult multisectoral collaboration (and adherence to the Three Ones principles) can be in fragile and post-conflict contexts. Categorized as one of the poorest nations in the world, it was overtly apparent how the socio-political situation (of conflict and limited resources) in the last decade had hampered a coordinated HIV/AIDS response. The DRC health system is fragile and over- burdened. Health services are also significantly provided by CREs, who currently manage 35 percent of the 515 health zones in the country – and are said to provide around half of the national health service (see Dimmock 2007 and Kintaudi 2006). The presence and agenda-setting power of international donors is also much stronger in the DRC. All of these factors have an impact on multisectoral HIV/AIDS collaboration. For one thing, HIV/AIDS is prioritized and addressed differently here – as one priority among many, and one activity among a broader portfolio of urgent activities. As one DRC participant said, “It is a problem of priority…where do you start? Everything is a priority...that is the challenge…a chicken and egg situation…a question of how the little we have can be used efficiently.� Significantly, both the CRE and stakeholder participants in the DRC characterized HIV/AIDS mainly as a ‘medical problem’ rather than a problem of development (characterized by the fact that the CRE HIV/AIDS program coordinators who participated were mainly medical doctors). Said differently, activities, partnerships and potentials were primarily focused on the medical response (such as delivery of ART) and rarely framed as a challenge within a broader development crisis. This, of course, shifts most of the collaborative focus away from civil society multisectoral response – towards the negotiation between the faith-based health providers and the MOH and PNLS (see Table 1). The DRC collaborative context generally demonstrated an uncertain division of labour between collaborative mechanisms. Participants confirmed what has been noted by others such as the WHO, that the division of responsibility between PNLF and PNMLS is also problematic “this situation is responsible for a weak national leadership which is detrimental to coordination of programmes� (UNAIDS 2006). The DRC collaborative context is particularly complex, with CREs each maintaining a web of individual relationships. For example, a Catholic Church representative described a complex HIV/AIDS intervention strategy, which is done in close collaboration with international and national partners such as CORDAID, Trocaire, CAFOD, CRS, local 177 dioceses, Bureau Diocésain des Oeuvres Médicales (Diocesan Office for Medical Works - BDOM) and Caritas – among others. A second example of this complexity is the unique SANRU project which began in the 1980s as the Basic Rural Health Project (later known as SANRU - Projet Santé Rurale) and is itself a complex collaborative structure which coordinates ‘health zones’ around Protestant, Catholic, governmental, and other NGO- managed hospitals in partnership with USAID. This collaborative relationship is built on multiple layers of partnership: between this project and the government (in relation to health services provision generally as well as HIV/AIDS activities), interfaith collaboration (between Catholic and Protestant groups, as well as others such as the Kimbanguists), between CREs and other NGOs, and of course with donors. As Schmid et al (2008) note, “the fragile and sometimes volatile political situation in the DRC has created an unusual, and by all appearances, effective collaborative partnership in a time of crisis.� However, despite (or because of) these complex collaborative relations, of the three countries, the collaborative HIV/AIDS mechanisms were least effective here, and sector groups the most fragmented. For example, there was only limited interfaith or ecumenical cooperation demonstrated, “we know each other, but we are not working together� (DRC participant). The Interfaith AIDS Council (CIC) was formed by the mainline churches, but it was felt that it “doesn’t really bring everyone together�. There was also limited cooperation demonstrated between donors. Donors’ priorities and agendas carried more weight in the DRC, and CREs demonstrated a stronger vulnerability, caused by a greater reliance on external donors. The relationship between CREs and donors was seen as more significant than the relationship between CREs and the government. In addition, the collaborative relationship between CREs and stakeholders was permeated with ‘suspicion’ rather than ‘trust’. Under condition of anonymity, participants cautiously described a lack of political will being a greater obstacle than lack of funds, and that the corrupt use of funds first needed to be rooted out of government before multisectoral collaboration could become a reality. Kenya: Moving forward strongly The Kenyan case study demonstrated that of the three countries, it was the most committed to the Three Ones principles. There is a clear National AIDS policy, a National AIDS committee, and a Monitoring and Evaluation system is well in place. CREs seemed to feel valued for their work in mitigating the epidemic by government stakeholders and generally there was a positive context of openness to collaboration, demonstrated by CHAK (2008) who reports that “…the present Minister of Health, Hon. Mrs. Charity Ngilu…recognized the work of FBOs and singled out the Catholic Church as contributing up to 40 percent of the national struggle against HIV/AIDS.� This openness was related to the efficient way in which the principles have been targeted and applied, but also about the manner in which civil society had been drawn into the process. Until just before this research was conducted, the socio-political context had been stable and Kenyan society had enjoyed a proactive response to the epidemic on the part of all the stakeholders. More specifically, government had overtly embraced the services of REs. There was clearly still room for improvement in terms of multisectoral collaboration, and the engagement of CREs. Although there was adherence to the principles of the Three 178 Ones, the coordinating mechanism for ‘faith sector’ collaboration was not yet functioning optimally. For example, government stakeholders assumed that these structures were a legitimate vehicle through which to channel resources and information to CREs, but the CRE representatives argued that this was not their experience. These bodies were not seen to be representative of all stakeholders, and CREs requested that an improved system be put in place when representatives are elected to such bodies (for example, there was some concern that KIRAC had been established by the government, was not yet properly representative, and that there was some overlap between KIRAC and other collaborative networks such as IRCK). The common concerns about competition and unbalanced power relations between CREs and stakeholders were also raised. However, these were presented as operational challenges. Overall, Kenyan participants demonstrated an obvious commitment to strengthening the partnership between CREs and collaborative stakeholders, with actionable recommendations made by all parties. It was notable in the Kenyan research workshops that participants and representatives of key coordinating or collaborative bodies were willing to admit weaknesses within their organizations, and in a non-defensive manner, which encouraged comment and critique from fellow CRE participants. The researchers felt that this attitude was itself a strong force in enhancing collaboration, and spoke to Kenya’s collaborative environment and also the importance of having the right individuals heading the collaborative efforts. Malawi: An established collaborative ‘AIDS industry’ The researchers felt that the Malawian case study demonstrated a further ‘stage’ of collaboration – where despite all the right policies and mechanisms and a commitment on the part of government to the Three Ones principles, there was a great deal of fragmentation amongst the stakeholders in their response to the epidemic. In comparison with Kenya, CREs were not collaborating well together, largely as a result of intensive (and extensive) donor involvement. There was a stronger sense of competition between CREs than there was in Kenya, more acutely felt conditionalities imposed on partnerships, and a lack of open engagement between CREs and national collaborative mechanisms. Another observation was that there was a general reluctance to attempt to bridge differences in perspectives (such as on the meaning of ‘healing’) which researchers felt had an impact on the potential for intervention through intercultural dialogue. Furthermore, while Malawi had a culturally sensitive National AIDS policy in place, there was a lack of ownership of this by stakeholders, including the CREs, also demonstrated as a reluctance to engage critically with the policy or policy environment. The Malawian collaborative context was one of an established and saturated ‘AIDS industry’, with large numbers of diverse funders in a geographically small country. Although the ‘faith sector’ had a coordinating body in place (MIAA), it was not seen as representative of the interests of all involved. Limited relations were observed between the CRE participants (and between their programs) – demonstrated by virtually no ecumenical or interfaith funding relations between the CRE participants. CONCLUSION Most significantly, this research highlighted the importance of ‘trust’. Although there are all sorts of policies, mechanisms, meetings and collaborative spaces being opened up – 179 the possession of one national action framework, one coordinating body, and one monitoring and evaluation system, in and of itself, does not promote better collaboration between government, donors, and CREs. What we would describe as ‘trust’ is the element that makes the Three Ones work in some contexts and not in others – determining to what degree CREs own the principles of the Three Ones. The DRC case- study demonstrated the least commitment on the part of the national government to these principles, while Malawi had a well-structured policy, national body, and a monitoring and evaluation system. However, there was not substantially more commitment to these structures by CREs in Malawi than in the DRC. In both countries, the relationship between government and CREs was not as trusting as is the case of Kenya. 63 All countries demonstrated that the trusted relationship was far more valued, and had a greater impact on collaboration than any number of policies, structures or multisectoral mechanisms. Trust was demonstrated through mutual openness to criticism, as mutual appreciation of contributions, and long-term sustained relationships. Of course, trust cannot be forced, and neither CREs nor stakeholders can be ‘blamed’ for lacking trust, especially in some contexts where there may be very good reasons for reacting otherwise. The wider framework of transparency, accountability and good governance provides an important context in which trust is to be understood. Trust also plays a direct role in the strength of collaborative partnerships between CREs and between donors and CREs – and in lessening competitive tensions. This paper has briefly reported on partial findings from a recent study – and has compared three contexts of multisectoral collaboration – focusing on the relationship between CREs and national stakeholders. Broadly speaking, Christian Religious Entities are engaged in a complex collaborative environment, negotiating multiple relationships and power dynamics. Many new collaborative mechanisms have been put in place over the last decade in relation to the push for multisectoral HIV/AIDS collaboration: some of which are working, and others that are not. The practice of collaboration remains a challenge – with representation a constant problem which requires consistent attention. This is especially the case if a diverse array of CREs are being clustered together as a ‘faith sector’. However, although we have described several challenges, the broad perspective was positive, and participants who were otherwise stretched to capacity, still saw the value in putting energy towards improving collaboration with and between CREs. There are clearly a number of operational steps that can and must be taken. Further (and ongoing) research is needed to keep mapping multisectoral collaboration, in the context of HIV/AIDS, but also with regards to development and health systems strengthening more broadly. Social network mapping may be one way to display the way relationships are in place, especially if this can be kept up to improve our understanding of how they evolve over time. Longitudinal work is clearly needed to understand how activities change and some relationships endure over time. Certainly this work need not be solely 63 We are addressing findings based on overall analysis – even in Kenya there was evidence of ‘distrust’ – for example some CRE participants described some parts of the national structures to be corrupt. 180 focused on HIV/AIDS: as can be seen in this study, the work and relationships between CREs and stakeholders is rarely neatly confined to a particular area. There is also substantial work that remains to improve collaborative mechanisms and to ensure better representation. REs that are not part of the ‘mainstream’ (such as the charismatic Christian groups, traditional religious leaders and Islamic communities) need to be more conscientiously drawn into collaboration. The national NFBHNs such as the CHAs could play stronger leadership roles in driving forward multisectoral collaboration, given their position and experience. Further research and collaborative engagement would be useful at a district and community level in order to understand how this operates differently than national-level engagement. In addition, researchers in this study recognized that not enough focus was placed on the collaborative relationships between CREs and fellow NGOs engaged in the same work. Finally, we are reminded that partnership and trust develops slowly – and indeed, collaboration needs to be built gradually over time. As UNFPA (2004) authors say, “Collaboration, dialogue and partnership should be on an ongoing basis, rather than for a single program or event. Mature relationships and partnerships would then mature, and create possibilities for other joint activities.� REFERENCES African Religious Health Assets Programme (ARHAP). 2006. “Appreciating Assets: The Contribution of Religion to Universal Access in Africa.� Cape Town. Benn, C. 2003. “Why Religious Health Assets Matter.� In ARHAP: Assets and Agency Colloquium, 3-11, Pietermaritzburg: African Religious Health Assets Programme. Birdsall, K., and K. Kelly. 2007. “Pioneers, Partners, Providers: The Dynamics of Civil Society and AIDS Funding in Southern Africa.�, Johannesburg: Centre for AIDS Development, Research and Evaluation and the Open Society Initiative for Southern Africa. Christian Health Association of Kenya (CHAK). 2008. CHAK website. [Accessed 04/02/2008]. Dimmock, F. 2007. “Faith-Based Health Networks in Africa, Partnerships for Health and Wholeness.� Christian Connections for International Health Conference, Bagamoyo, Tanzania. The Global Fund to Fight AIDS, Tuberculosis and Malaria (GFTAM). 2008. “Report on the Involvement of Faith-Based Organizations in the Global Fund.� Geneva. Haddad, B., J. Olivier, and S. De Gruchy. 2008. �The Potential and Perils of Partnership: Christian Religious Entities and Collaborative Stakeholders Responding to HIV and 181 AIDS in Kenya, Malawi and the DRC.� Cape Town, South Africa: African Religious Health Assets Programme. Kintaudi L. 2006. “Churches and Health Care in DR Congo.� SANRU presentation online [Accessed 15/08/2011] Liebowitz, J. 2002. “The Impact of Faith�Based Organisations on HIV/AIDS Prevention and Mitigation in Africa.� KwaZulu Natal: Health Economics and HIV/AIDS Research Division, University of KwaZulu�Natal Mandi, F. 2006. “Planning, Developing and Supporting the Faith�Based Health Workforce.� African Christian Health Associations’ Human Resources for Health Mini� Forum, Nairobi: The Capacity Project and Medicus Mundi International. National AIDS Control Council (NACC). 2005. “Kenya National HIV/AIDS Strategic Plan 2005/6–2009/10: A call to Action.� [KNASP 2005/6–2009/10]. Nairobi: NACC, Office of the President. Olivier, J., J. R. Cochrane, and B. Schmid. 2006. “ARHAP Literature Review: Working in a Bounded Field of Unknowing.� Cape Town: African Religious Health Assets Programme. Parry, S. 2002. “Responses of the Churches to HIV/AIDS: Three Southern African Countries.� Harare and Geneva: World Council of Churches, Ecumenical HIV/AIDS Initiative in Africa, Southern Africa Regional Office. Schmid, B., E. Thomas, J. Olivier, and J. R. Cochrane. 2008. “The Contribution of Religious Entities to Health in sub�Saharan Africa.� Cape Town: African Religious Health Assets Programme. United Nations AIDS Programme (UNAIDS). 2006. Report on the Global AIDS Epidemic. Geneva. United Nations Population Fund (UNFPA). 2004. Culture Matters: Working with Communities and Faith-Based Organisations: Case Studies from country Programmes. New York. 182 About this series... This series is produced by the Health, Nutrition, and Population Family (HNP) of the World Bank’s Human Development Network. The papers in this series aim to provide a vehicle for publishing preliminary and unpolished results on HNP topics to encourage discussion and debate. The findings, interpretations, and conclusions expressed in this paper are entirely those of the author(s) and should not be attributed in any manner to the World Bank, to its affiliated organizations or to members of its Board of Executive Directors or the countries they represent. Citation and the use of material presented in this series should take into account this provisional character. For free copies of papers in this series please contact the individual authors whose name appears on the paper. Enquiries about the series and submissions should be made directly to the Editor Martin Lutalo (mlutalo@worldbank.org) or HNP Advisory Ser- vice (healthpop@worldbank.org, tel 202 473-2256, fax 202 522-3234). For more information, see also www.worldbank.org/hnppublications. The world bank 1818 H Street, NW Washington, DC USA 20433 Telephone: 202 473 1000 Facsimile: 202 477 6391 Internet: www.worldbank.org E-mail: feedback@worldbank.org