UNIVERSAL HEALTH COVERAGE STUDY SERIES NO. 39 2 UNIVERSAL HEALTH COVERAGE STUDY SERIES No. 39 Universal Health Coverage in Low-Income Countries: Tanzania’s Efforts to Overcome Barriers to Equitable Health Service Access Huihui Wang and Nicolas Rosemberg The World Bank, Washington, DC, 2018 3 4 Table of Contents Preface to the second round of the Universal Health Coverage Study Series ................................ 9 About the Authors ......................................................................................................................... 11 Executive Summary ...................................................................................................................... 12 1. Introduction ........................................................................................................................ 13 2. Service Provision ............................................................................................................... 16 Devolution of Functions in the Health Sector ........................................................................... 18 Strengthening Primary Health Care ......................................................................................... 19 Results-Based Financing ........................................................................................................... 20 3. Social Protection, User Fees, and Financial Protection in Health ..................................... 21 4. Unpacking Community Health Funds ............................................................................... 24 Revenue Collection, Pooling, and Purchasing of Services ....................................................... 25 Enrollment ................................................................................................................................. 26 CHF’s Management Reform ..................................................................................................... 28 5. Implementation Challenges and Opportunities .................................................................. 28 6. Looking Ahead................................................................................................................... 31 Annex ............................................................................................................................................ 33 Box Box 1 Tanzanian Social Action Fund ........................................................................................... 23 Table of Figures Figure 1 Changes in Under-5 Mortality between 2000 and 2015 ................................................. 14 Figure 2 Reductions in Maternal Mortality between 2000 and 2015 ........................................... 15 Figure 3 Percentage of Births Delivered by Skilled Provider....................................................... 16 Figure 4 Health Care Provision in Tanzania ................................................................................. 17 Figure 5 Governance of the Health Sector .................................................................................... 18 Figure B1.1 TASAF’s Coverage, by Income Decile .................................................................... 24 Figure A.1 Trends in Total Health Expenditure (THE) and THE Per Capita............................... 33 Figure A.2 Trends in Financing Sources ...................................................................................... 33 Figure A.3 Targeting Performance of CCTs in Selected Countries ............................................. 36 5 Table of Tables Table 1 Service Coverage ............................................................................................................. 14 Table 2 Distribution of Health Care Providers by Ownership in 2014......................................... 17 Table 3 Prepayment Schemes and Health Insurance in Tanzania ................................................ 22 Table A.1 Sources of Revenue and Financing Schemes............................................................... 34 Table A.2 Quality Assessment Framework .................................................................................. 35 6 ABBREVIATIONS AND ACRONYMS CCT Conditional Cash Transfers CHF Community Health Fund CHSB Council Health Service Board GDP Gross Domestic Product HFS Health Financing Strategy HIV/AIDS Human Immunodeficiency Virus/Acquired Immune Deficiency Syndrome HRH Human Resources for Health LGA Local Government Authority MDG Millennium Development Goal MHCDGEC Ministry of Health, Community Development, Gender, Elderly and Children MMAM Primary Health Services Development Program, Mpango wa Maendeleo wa Afya ya Msingi NGO Nongovernmental Organization NHA National Health Accounts NHIF National Health Insurance Fund PHC Primary Health Care PO-RALG President’s Office Regional Administration and Local Government RBF Results-based Financing SDG Sustainable Development Goals SNHI Single National Health Insurance TASAF Tanzanian Social Action Fund THE Total Health Expenditure TIKA Tiba kwa Kadi T Sh Tanzanian Shilling UHC Universal Health Coverage US$ United States dollar 7 8 Preface to the second round of the Universal Health Coverage Study Series All over the world countries are implementing pro-poor reforms to advance universal health coverage. The widespread trend to expand coverage resulted in the inclusion of the “achieving universal health coverage by 2030” target in the Sustainable Development Agenda. Progress is monitored through indicators measuring gains in financial risk protection and in access to quality essential health-care services. The Universal Health Coverage (UHC) Studies Series was launched in 2013 with the objective of sharing knowledge regarding pro-poor reforms advancing UHC in developing countries. The series is aimed at policy-makers and UHC reform implementers in low- and middle-income countries. The Series recognizes that there are many policy paths to achieve UHC and therefore does not endorse a specific path or model. The Series consists of country case studies and technical papers. The case studies employ a standardized approach aimed at understanding the tools –policies, instruments and institutions- used to expand health coverage across three dimensions: population, health services and affordability. The approach relies on a protocol involving around 300 questions structured to provide a detailed understanding of how countries are implementing UHC reforms in the following areas: • Progressive Universalism: expanding population coverage while ensuring that the poor and vulnerable are not left behind; • Strategic Purchasing: expanding the statutory benefits package and developing incentives for its effective delivery by health-care providers; • Raising revenues to finance health care in fiscally sustainable ways; • Improving the availability and quality of health-care providers; and, • Strengthening accountability to ensure the fulfillment of promises made between citizens, governments and health institutions. By 2017, the Series had published 24 country case studies and conducted a systematic literature review on the impact of UHC reforms. In 2018 the Series will publish an additional15 case studies, A book analyzing and comparing the initial 24 country case studies is also available: Going Universal: How 24 Developing Countries are Implementing UHC Reforms from the Bottom Up . Links to the Series and the book are included below. Daniel Cotlear, D. Phil. Manager and Editor Universal Health Coverage Study Series Links: http://www.worldbank.org/en/topic/health/publication/universal-health-coverage-study-series http://www.worldbank.org/en/topic/universalhealthcoverage/publication/going-universal-how- 24-countries-are-implementing-universal-health-coverage-reforms-from-bottom-up 9 Acknowledgments The team wishes to express thanks to Daniel Cotlear and Magnus Lindelow for their continuous guidance, and to the peer reviewers who kindly shared their feedback: Pia Schneider, Gayle Martin, Chiho Suzuki, Son Nam Nguyen, and Yi-Kyoung Lee. The team would like to acknowledge the valuable contributions and support of the Ministry of Health, Community Development, Gender, Elderly and Children, in particular Mariam Ally and Kuki Tarimo. Furthermore, inputs from CHF and NHIF, and development partners such as GIZ, KfW, P4H and USAID helped shape this report. The team is grateful to Nina Siegert for her comment and insights. 10 About the Authors Huihui Wang is a Senior Economist in the Health, Nutrition and Population Global Practice in the World Bank Group. She has almost 20 years of experience working in low-, middle-, and high- income countries with a focus on supporting them to achieve universal health coverage. Huihui leads the World Bank’s technical support in health system reforms of various countries, including public financing, health insurance, service delivery, and quality of care. She was appointed by the President of Ghana as a member of Technical Review Committee for the reform of the Ghana National Health Insurance Scheme. Her responsibilities at the World Bank also include preparing, supervising, and evaluating lending operations at the country level. Prior to joining the World Bank, she worked on health sector reform initiatives in both China and the United States as a researcher. Huihui has a multidisciplinary background that makes her a competent player in the global health field. She holds a medical degree from Beijing Medical University, and a Master’s degree in economics and a PhD degree in health services and policy analysis from the University of California, Berkeley. She has extensive experience teaching executives, midcareer, graduate students, and undergraduates through lectures at Peking University; the University of California, Berkeley; and the World Bank. Nicolas Rosemberg is a Consultant to the World Bank. He focuses on health financing topics and on the promotion of equity in health. His work includes the review of universal health care policies in 46 African countries, a study on propoor health coverage programs in Tanzania, and a Public Expenditure Review of the health sector in Zambia. Before joining the Bank, Nicolas worked as a Consultant at the Inter-American Development Bank and the GIZ. He holds a Master’s in Public Policy from the Hertie School of Governance in Berlin, Germany, and a Bachelor’s degree in Economics from the University of Buenos Aires. 11 Executive Summary In the last two decades, the Government of Tanzania has embarked on many health sector reforms that promote universal health coverage (UHC). This case study focuses primarily on the initiatives that foster UHC inclusive of the poor. Tanzania’s efforts that can potentially benefit the poor can be classified into two main groups: supply-side efforts and demand-side efforts. Supply-side efforts include the devolution of functions to local government authorities, strengthening primary health care, and incentives to improve performance through implementation of a results-based financing program. Demand-side efforts consist of a prepayment scheme for the informal sector in rural areas called Community Health Funds. To judge the effectiveness of these programs to provide coverage to the poor, this case study identifies the major barriers to equitable access and utilization of health services in Tanzania. These are the poor quality of health services in rural areas and the limited financial protection resulting from weak implementation of the user fee waiver policy. Subsequently, the case study analyzes the effectiveness of the above-mentioned efforts to overcome those challenges and achieve UHC inclusive of the poor, and finds that there have been important improvements (for example, the increase in the number of trained health personnel and the improvements in the quality of health services favored by results-based financing). However, progress has been slow, and important inequities remain in the access to health services between the rich and the poor and between urban and rural areas. The Health Financing Strategy is expected to address some of these challenges to UHC inclusive of the poor. Furthermore, the reforms promoted by the strategy present an outstanding opportunity to improve data collection systems that would allow for better assessments of the pro-poor nature of health policies and programs. 12 1. Introduction During the past decade, Tanzania has experienced steady economic growth, with average annual growth rates of between 6 and 7 percent. Despite this positive trend in the economy, poverty rates have not decreased accordingly1; more than one-fourth of Tanzania’s 53 million inhabitants live below the poverty line and almost 10 percent live in extreme poverty. Furthermore, 80 percent of the population living below the poverty line resides in rural areas,2 and the fertility rates in these areas are almost double the fertility rates of urban dwellers. The health sector has been identified as a policy priority area in Tanzania.3 The share of total health expenditure (THE) as a percentage of nominal GDP has increased from 5 percent in 2002 to 9.2 percent in 2014. Despite high population growth rates, THE per capita has more than tripled in the same period and reached T Sh 73,365 (US$404) in 2014.5 The share of external funding is high, and represented 48.3 percent of THE in 2012, after which it decreased to 37 percent in 2014. The contribution of government spending to THE is low (28 percent in 2014), and the funds allocated to health as a percentage of government’s total expenditure have increased at a modest pace, from 6.1 percent in 2003 to 6.8 percent in 2014 (see Annex). Public funding for health comes mainly from the general budget, and private sources—including user fees—can represent up to one-third of districts’ recurrent budgets. The limited pooling of resources in the health sector and the multiplicity of sources of revenue translate into a highly-fragmented health financing system. Public sources flow through various financial streams (such as Personal Emoluments Block Grants, Other Charges Block Grants, and Development Grants), prepayment schemes are managed in separate pools at the district level and, with the exception of the Basket Fund, external sources are channeled through vertical programs. In line with the recommendations of the Bamako Initiative,6 Tanzania introduced user fees in the health sector in 1993. A policy to exempt the poor was designed, but implementation of this policy has been limited. Furthermore, despite recent improvements in the coverage of prepayment schemes, contributions to mandatory and voluntary programs are still low, and out-of-pocket payments financed a high share of THE.7 Out-of-pocket expenditures is the most regressive and inequitable source of financing. The largest prepayment schemes are the Community Health Funds (CHF). Community Health Funds are managed by district authorities and are designed to increase coverage of the informal sector in rural areas and to raise districts’ resource mobilization. The National Health Insurance Fund (NHIF) is administered by a parastatal organization and is mandatory for civil servants, although all citizens can enroll (including the informal sector). Coverage of non-civil servants is extremely low. In the past two decades, important achievements in the health sector have contributed to the improvement of the health status of the population. Life expectancy increased from 49 in 1995 to 65 in 2014. Furthermore, Tanzania is committed to the promotion of UHC and has made important efforts to reach different sectors of the population. The progress in the reduction of under-five mortality has been significant and Tanzania achieved the fourth Millennium Development Goal (MDG) in 2012. The under-five mortality rate in 2015 was 48.7 deaths per 1,000 live births, which is significantly below other low-income countries and other countries in the region8 (see figure 1). 13 Figure 1 Changes in Under-5 Mortality between 2000 and 20159 Source: World Development Indicators 2017. Note: LIC = low-income country; SSA = Sub-Saharan Africa. Reductions in under-five mortality have been a clear manifestation of Tanzania’s general effort to ensure the supply of a basic package of health services. Child health has been an important topic in Tanzania’s health agenda since the 1980s and it has received steady financial support, with a significant increase in funding in 2000. The increase in coverage of strategic interventions such as integrated management of childhood illnesses, immunization, malaria prevention, and prevention and treatment of HIV/AIDS was among the most salient protective determinant of under-five mortality. Moreover, around two-thirds of the children with fever or diarrhea sought treatment from a health facility. In the last decade, coverage of maternal health services has also increased, as evidenced by the higher rates of skilled birth attendance, the higher share of pregnant women receiving prenatal care from a skilled provider, and the higher share of women delivering in health facilities (see table 1). Table 1 Service Coverage Service 2005 2010 2015 Women who had a live birth in the 5 years preceding the survey 94.3 95.9 98 – percentage with prenatal care from a skilled provider Live births in the 5 years preceding the survey – percentage 54.3 50.6 63.7 delivered by a skilled provider Live births in the 5 years preceding the survey – percentage 47.1 50.2 60.2 delivered in a health facility Children who received all basic vaccinations 71.1 75.2 75.2 Children with diarrhea – percentage for whom treatment was 47 52.6 67.1 sought from a health facility Source: Demographic and Health Survey 2005, 2010, 2015. 14 Despite significant progress, Tanzania’s performance in the reduction of maternal and neonatal mortality rates reflects difficulties in expanding the provision of services beyond the basic package. Even though Tanzania has significantly reduced the maternal mortality rate and has a lower rate than other low-income countries, it did not achieve MDG 5 (see figure 2). In 2015, the maternal mortality rate was 398 deaths per 100,000 live births and the neonatal morality rate was 18.8 per 1,000 live births. Figure 2 Reductions in Maternal Mortality, 2000–1510 Source: World Development Indicators 2017. Note: DRC = Democratic Republic of Congo; LIC = low-income country; SSA = Sub-Saharan Africa. In addition, important disparities in service coverage can be observed, both by income level and geographic area (urban compared to rural). While 95.1 percent of the deliveries by women in the richest quintile were done by a skilled provider, only 41.8 percent of the women in the poorest quintile met these standards. Inequities in service coverage between rural and urban residents is also stark, and the access rate to skilled birth deliveries in urban areas is twice the access rate in rural settings (see figure 3). Similar trends were observed in child health indicators.11 Several factors affect poor people’s access to health services. Two salient challenges to an equitable access to health services in Tanzania are the poor quality of health services (particularly in rural areas) and the limited financial protection in health. 15 Figure 3 Percentage of Births Delivered by Skilled Provider Panel A By Geographic Area Panel B By Wealth Quintile 100 100 80 80 60 60 % % 40 40 20 20 0 0 Rural Urban Q1 Q2 Q3 Q4 Q5 2005 2010 2015 2005 2010 2015 Source: Demographic and Health Survey 2005, 2010, and 2015. The main purpose of this case study is to describe Tanzania’s efforts to promote UHC inclusive of the poor, and to identify challenges and opportunities for the health system to advance on this path in a coherent and integrated fashion. Given the large number of interventions implemented in the health sector, efforts were selected based on their potential to address the above-mentioned challenges to the equitable access to health services, namely the poor quality of health services and the limited financial protection. The rest of the paper is organized as follows. Section 2 describes supply-side efforts and features the devolution of health services, the primary health care (PHC) strengthening program, and results-based financing. Section 3 analyzes Tanzania’s social protection program and synthesizes the country’s experience with user fees, as well as their impact on financial protection. Section 4 features the Community Health Funds, Tanzania’s most explicit effort to increase financial protection in health. Section 5 assesses the implementation of these initiatives. Section 6 highlights some opportunities to include the poor and address the challenges to pro-poor UHC in Tanzania. 2. Service Provision Service provision in Tanzania is organized following a pyramidal model, where health promotion and primary health care constitute the base of the pyramid (see figure 4). Primary health care is provided at dispensaries and health centers. Dispensaries provide outpatient services and engage in a limited number of outreach activities.12 Health centers, in turn, offer outpatient and basic inpatient services. They are also responsible for the collection of data and the management of health information systems at the primary level. Hospitals are organized based on the complexity of the services provided (specialized care or general services) and the level of provision (district, regional, zonal, or national). Of the 125 public hospitals, 80 percent are district hospitals, 16 percent are regional hospitals, and 4 percent are national hospitals. There are National Treatment Guidelines that determine which services are provided at which facility type. 16 Figure 4 Health Care Provision in Tanzania Source: Health Sector Strategic Plan IV 2015. Note: ADDOs = accredited drug dispensing outlets; APHFTA = association of private health facilities in Tanzania; CBO = community cased organizations; CCBRT = comprehensive community based rehabilitation in Tanzania; CSSC = Christian social services commission; FBO = faith-based organization; KCMC = Kilimanjaro Christian medical centre; MOI = Muhimbili orthopaedic institute; NGO = nongovernmental organization; ORCI = Ocean Road cancer institute. Overall, 73 percent of all facilities are public, 1 percent are parastatal, 13 percent are run by faith- based organizations, and 14 percent are private for-profit13 (see table 2). Moreover, hospital beds at private for-profit providers represent 4 percent of all hospital beds. While hospital-level services were financed mostly through government funding, ambulatory and preventive care were financed primarily through nongovernmental organizations (NGOs) and out-of-pocket expenditure. Table 2 Distribution of Health Care Providers by Ownership, 2014 Public Parastatal FBO Private Total Clinic 7 1 10 158 176 Dispensary 4,521 59 648 682 5,910 Health center 478 7 138 93 716 Hospital 120 4 104 35 263 Total 5,126 71 900 968 7,065 % 73% 1% 13% 14% 100% Source: Government Open Data Portal. Note: FBO = faith-based organization. The system was conceived to operate as an integrated network of health providers. Patients that require more specialized care are supposed to be referred from the dispensary to the health center, 17 and from there—if required—to the district hospital. However, the referral system is poorly enforced. Bypassing the primary level is partly explained by the administrative burden, and the unavailability or poor quality of PHC services. The inadequate enforcement of the referral system, combined with the absence of effective gatekeeping mechanisms, represents a significant source of inefficiencies in the health sector. Devolution of Functions in the Health Sector The management and governance of the health sector in Tanzania is heavily influenced by the process of the devolution of functions to local government authorities (LGAs), which entrusts important responsibilities to regional and district governments. The process follows the principle of subsidiarity, according to which the lowest levels of government are more efficient in the delivery of services. The devolution of health services has been devised as a strategy to reach the poor, provided government authorities at lower levels are in a better position to identify them, understand their needs, and design programs that are in line with the satisfaction of their needs. The main stakeholders in the health sector are the Ministry of Health, Community Development, Gender, Elderly and Children (MHCDGEC); the President’s Office Regional Administration and Local Government (PO-RALG); the Ministry of Finance and the Ministry of Education and Vocational Training; and the President’s Office Public Service Management (see figure 5). Figure 5 Governance of the Health Sector Government of Tanzania Ministry of Ministry of MHCDGEC PO-RALG Finance and Education, Science Planning and Technology Departments and National, Regional (Referral) Government and Universities and Agencies, Training Specialized and Hospitals and LGAs Holding account Colleges Centers Zonal Hospitals RHMT Support functions Provision of PHC Disbursement of Provision of health Training of health in the health and hospital funds, financial services staff sector services reporting Training of health Contract private staff health providers Source: Based on Human Resources for Health and Social Welfare Strategic Plan 2014–2019. Note: LGA = local government authority; MHCDGEC = Ministry of Health, Community Development, Gender, Elderly and Children; PHC= primary health care; PORALG = President’s Office Regional Administration and Local Government; RHMT = Regional Health Management Team. The MHCDGEC plays an important role in the stewardship and regulation of the sector and is in charge of the provision of health services at national-level hospitals. In collaboration with the Ministry of Education and Vocational Training, the MHCDGEC is responsible for the training of medical staff.14 The Ministry of Finance is responsible for the disbursement of funds. 18 There are 26 regions and 184 LGAs. The highest authority at the regional level is the Regional Medical Officer. The Regional Medical Officers, together with their Regional Health Management Team, are responsible for the provision of health services in regional hospitals. At the LGA level, a Council Health Management Team, headed by the District Medical Officer, is responsible for the provision of services at dispensaries, health centers, and district hospitals. The District Medical Officer participates in the development of district’s health budgets, plans the activities to execute the budget, and contracts private providers, when needed. While the devolution of health service provision has been successful, health financing responsibilities remain highly centralized. The central level has the biggest responsibility for the mobilization of public funds in the sector. Funds raised at the local level represented only 0.4 percent of THE, so the provision and administration of health services at the local level is highly dependent on central government’s transfers. In terms of the administration of the health budget, Government authorities managed 39.1 percent of THE in 2012. The largest share (52 percent) of public funds was administered at the central level. This includes the Global Fund’s disbursements, at almost T Sh 200 billion (US$90 million) in 2012. Despite their low revenue generation, local authorities (mostly at the district level) managed 18.8 percent of THE. LGA budgets consist of cost-sharing funds (user fees and Community Health Funds and National Health Insurance Fund contributions), transfers from the central government, and transfers from development partners. Central government transfers represented 92 percent of LGA budgets in 2012. Delays in inter- fiscal transfers represent a major challenge for the financial management of district health services. A small share of the funds allocated from the central to the district level follow equity-based allocation formulas, which take into account a district’s characteristics.15 In practice, however, the allocation of these funds does not always follow the formulas. Council Health Planning Guidelines set a ceiling on the allocation of LGA funds to the District Medical Officer, the Council Hospital, the Voluntary Agency Hospital, health centers, dispensaries, and community initiatives. LGA capacity to use these funds is limited: funds are not disbursed in a timely manner, knowledge of rules and procedures for the disbursement of funds is low, project management skills are low, and different stakeholders use separate reporting instruments, which puts pressure on the reduced number of employees. Little is known about the flow of resources from the LGAs to the health facilities. There is some evidence that insufficient funds are reaching the frontline workers in dispensaries and health centers, where the greatest volume of patients is being handled.16 With the exception of Community Health Fund contributions, facilities have relative autonomy in the use of funds. Community Health Fund contributions are subject to cost-sharing guidelines, which require local authorities to use at least 67 percent of these funds to purchase medicines. Strengthening Primary Health Care In 2007, the Government of Tanzania launched the Primary Health Services Development Program (Mpango wa Maendeleo wa Afya ya Msingi, MMAM). MMAM aims to reduce barriers to access to health services and to improve the quality of PHC, with a particular focus on rural areas. The government identified 17 strategic interventions, including the increase in availability of human 19 resources for health and medicines and the improvement of health infrastructure. Implementation of MMAM has led to a considerable increase in the number of health facilities. Since 2007, more than 1,500 facilities have been built or rehabilitated, out of which 87 percent are primary health care facilities (dispensaries and health centers). This represented an increase of 31 percent in the number of health facilities in the country. The construction of new health facilities creates an additional demand for health professionals. The main guidelines for human resources for health (HRH) reforms are summarized in the third Human Resources for Health and Social Welfare Strategic Plan 2014–2019. HRH planning, monitoring, and evaluation is a responsibility of LGAs, and the plan was developed in close collaboration between the national and lower-level government authorities. The first two Human Resources for Health and Social Welfare strategic plans successfully increased the number of students that enroll in training institutions and the provision of loans has enabled students to register in these degrees. The third plan intends to continue these efforts and ensure that the profiles of health professionals trained matches those required by the health sector. It has the objective of lowering the HRH gap17 from 52 percent in 2014 to 30 percent in 2019, increasing the number of health professionals from 7,000 in 2014 to 10,000 in 2019, and retaining 70 percent of the staff within the sector. In addition, the Health Sector Strategic Plan IV (2015–2020) introduced a quality assurance tool to monitor the progress made in these areas. It consists of a star rating system that evaluates primary level health facilities. The indicators used to evaluate the quality of the services provided can be found in Annex table A.2. This tool is being used by an independent body to gradually accredit facilities. Moreover, facilities that fail to obtain three stars and get accredited are identified by the district authorities and receive special support (including mentoring and supportive supervision) to help them reach the desired quality standards. This support is included in the Strategic Plan’s quality improvement framework, for which the MHCDGEC has developed standards and trained health personnel. Results-Based Financing The MHCDGEC and the President’s Office Regional Administration and Local Government (PO- RALG) are co-implementing a Results-Based Financing18 (RBF) program in mainland Tanzania. It was piloted in Shinyanga region in 2013 and, as of 2017, covers seven regions. The goal of RBF is the improvement of the quality of health services, particularly those related to the MDG/SDG targets in the health sector. The program also focuses on areas with high poverty rates and poor health outcomes. Even though performance indicators have a strong focus on MDG/SDG-related health services, RBF in Tanzania is devised as an instrument to improve overall health system performance. For that purpose, the program establishes incentives for stakeholders at different stages of the health system; while most similar interventions pay special attention to primary health care providers, Tanzania’s RBF program also measures the performance of community health workers, of health managers at the district and regional level,19 and the Medical Store Department. The Medical Store Department is a semi-autonomous organization responsible for the procurement, storage, and distribution of medical supplies and medicines. The payment of incentives for achieving 20 performance targets can induce behavior changes that catalyze improvements in the quality of health services. In addition to the incentives scheme, the program seeks to strengthen management skills by introducing certain reforms to standard procedures. For example, facilities participating in RBF can purchase medical supplies from a wider range of certified providers. To enroll in the RBF program, facilities must achieve, among other requirements, at least a one- star rating in the human resources dimension of the Big Results Now’s quality assessment framework, that is, have at least one skilled personnel at the dispensary level. Public facilities that fulfill this criterion are eligible for a readiness fund that must be used to improve the quality of the facility’s infrastructure, and to ensure the availability of essential medical equipment and supplies. Facilities that receive less than one star must work with the Regional Health Management Team and the Council Health Management Team to achieve the required readiness level for accreditation in RBF. Even though RBF has been presented separately from devolution and PHC strengthening, it is not a stand-alone intervention, and there are important links to the broader health system. RBF has played an instrumental role in the organization of the health sector, by clearly defining key functions, such as regulator (MHCDGEC), facilitator (PORALG and LGAs), and fund holder (Ministry of Finance and Planning). Furthermore, RBF increases a facility’s level of autonomy and builds financial management skills at the LGA level. This, together with governance improvements due to routine monitoring and better coordination among government levels, creates a favorable environment for the adequate devolution of health services. In addition, RBF’s incentives have successfully improved key areas of PHC service delivery, like doctors’ absenteeism rates and implementation of medical protocols. 3. Social Protection, User Fees, and Financial Protection in Health Tanzania introduced user fees in the health sector in 1993. User fees for primary care services are lump-sum amounts, which oscillate between T Sh 1,000 (US$0.4) and T Sh 3,000 (US$1.30) per visit. At the hospital level, fees vary based on the type of services provided, and they are determined at the national level by the MHCDGEC. In the public sector, amounts are not calculated to fully cover the cost of the services provided. This is explained by the concern that user fees that cover the full cost could negatively affect the utilization of health services by the poor. Furthermore, user fees were mainly devised as a mechanism to raise complementary resources for the health sector. Additional resources could, in turn, help improve the quality in the delivery of health services. The policy was initially implemented in public hospitals and then was rolled out to public dispensaries and health centers. To offset the potential negative effect of user fees, the Government of Tanzania gradually implemented prepayment schemes and a mandatory health insurance scheme for civil servants (table 3),20 and designed a waiver and exemption policy. The prepayment schemes were originally launched in rural areas (Community Health Funds) and then rolled out to urban areas (TIKA). These schemes target the informal sector (mostly in rural areas) and they provide a limited benefits package. Enrolment in these schemes is voluntary and the premium varies by district, ranging between T Sh 5,000 (US$2.20) and T Sh 30,000 (US$13.40). Contributions are transferred to LGA 21 health budgets and there is no separate purchaser-provider split. The health insurance scheme for civil servants is the National Health Insurance Fund (NHIF), which is funded via a 6 percent payroll deduction, equally distributed between employers and employees. The NHIF covers 7.2 percent of the population, and even though it started as a scheme for civil servants, private sector employees and self-employed can voluntarily join. The yearly fee for non-civil servants is a lump- sum contribution of T Sh 1,501,200 (US$672.60) per household. The benefits package offered by the NHIF is more comprehensive and includes outpatient and inpatient care. Service providers are reimbursed for the services offered to NHIF beneficiaries on a fee-for-service basis. Table 3 Prepayment Schemes and Health Insurance in Tanzania National Health Insurance Community Health Fund/TIKA Fund Target Population Civil servants Informal workers in rural areas Population Coverage21 7.2% 19.8% Enrolment Mandatory Voluntary Benefits Package Medium range Mostly Primary Health Care level Premium Range 6% of payroll T Sh 5,000 (2.2) – T Sh 30,000 (13.4) * Source: National Health Financing Strategy 2015; *plus the matching grant paid by the central government. Population groups exempted from the payment of user fees include pregnant women, children under five years of age, and patients with priority health conditions such as HIV/AIDS and tuberculosis. Waivers, in turn, apply to the elderly and the poor. A study commissioned by the MHCDGEC in 2003 recommended that the waiver system should be refined to cover the poorest segments of the population. While the exemption policy is widely implemented, the waiver for the poor is often not applied. This can be explained by the perverse incentives created by the process of identification of the poor and by the absence of a reimbursement system to compensate facilities for the services provided. In addition, studies have shown that district authorities and service providers have limited knowledge about the existence of a waiver system in the first place. Unlike the exempted groups, the identification of the poor is more complex and costly. The Tanzanian Social Action Fund developed a sophisticated instrument to identify the poor to participate in cash transfer programs (see box 1). The instrument, however, is only applied for the purpose of that program, and there are no clearly defined, national-level criteria that guide the identification process of the poor in the health sector. The responsibility to identify the poor is borne by local-level authorities and, in some districts, by service providers. This process relies on the active participation of the poor: at the primary care level, the poor need to approach village authorities (village executive officers) and request a letter in order to receive a waiver. For referral to district hospital level, an additional letter from the Social Welfare Officer is required. In some districts (for example, Iramba district), facilities receive a list of households eligible for a waiver and this list is displayed at the facility. The low level of awareness of the waiver policy— presumably lower among the poor—therefore represents an important challenge to its implementation. In addition, the assessment by which village authorities appraise individual’s financial situation is rarely documented, and concerns have been raised regarding the delivery of letters to nonpoor families. When the waiver policy is effectively implemented, the poor receive services for free, but facilities are not reimbursed by the government for the provision of these services. The guidelines that 22 introduced the waiver policy did not provide for a mechanism to compensate providers for foregone revenues from nonpaying patients. The absence of a reimbursement system engenders perverse incentives for providers to grant waivers, particularly in those cases where service providers are responsible for the identification of the poor. Furthermore, given that the exempted population represents an important share of the patient load,22 providers are more reluctant to implement the waiver scheme, as this represents an additional reduction in the collection of funds. The impact of the waiver policy on facilities’ financial situation has been insufficiently explored. Box 1 Tanzanian Social Action Fund In 2000, the Government of Tanzania, with the support of the World Bank (through the International Development Association), established the Tanzanian Social Action Fund (TASAF). It is the largest social protection program in the country, reaching 1.1 million poor households (about 6 million individuals), representing around 24.1 percent of the extreme poor.a TASAF’s main goal is to empower communities to foster economic growth and reduce poverty. It was originally piloted in 8 districts. The first phase of the program rolled out the scheme to 40 districts, and since the second phase, the program has reached all LGAs (but not every village). It is currently on its third phase. TASAF III has four main pillars, which include a conditional cash transfer (CCT) program, a public works program focused on green jobs, a capacity-building component targeted at livelihood skills, and an institutional strengthening component. The CCT program includes one health-related conditionality related to the number of times beneficiaries visit health centers per year. An impact evaluation of the program concluded that the scheme had led to a reduction in the number of visits. This can be partly explained by the fact that the likelihood that beneficiaries reported being sick in the past four weeks was smaller than in the control group. Furthermore, the number of visits required by the program was lower than the baselines value, so the condition was binding for a small share of the beneficiaries (World Bank 2013). TASAF has developed a sophisticated tool to identify the poor. It is a multiple targeting mechanism that consists of four steps.b The first step is geographic targeting. Poverty maps based on the Household Budget Survey (2011/12) are used to select the poorest villages. Resources are allocated to each village, first following an equity component and then using an allocation formula that takes into account the population (40 percent), poverty rates (40 percent), and the geography (20 percent) of each district.c The second step is the participatory community-based targeting. In this part of the process, members of the community, organized in a CCT Management Committee, define village-specific poverty criteria, identify the households that fulfill these criteria, and collect their data. The third step is a proxy means test, where TASAF authorities assign a welfare score to each household identified by the CCT Management Committee based on the data collected. Households whose score is below the extreme poverty line qualify for the program. The list of candidates is entered into a Unified Registry of Beneficiaries, which is benchmarked against national-level indicators. The fourth and final step consists of validation of the results of community-based targeting and proxy means test targeting in a village assembly. The inclusion error of TASAF’s poverty identification tool is low, and 60 percent of those targeted by the program fall in the first two income deciles (see figure B1.1). Compared to similar instruments in other countries, the precision of the tool is high (see Annex). In addition, households that have not been included in the Unified Registry of Beneficiaries can file a complaint and request an evaluation of their case. TASAF’s poverty identification tool is the most sophisticated poverty identification mechanism in the country. It is therefore surprising that the use of this tool is limited to TASAF-related activities only. Sources: a. World Bank 2015a. b. Stoerner et al. 2013. c. Government Project Preparation Team 2013b. 23 Figure B1.1 TASAF’s Coverage, by Income Decile Source: Productive Safety Net Program Impact Evaluation Baseline. 4. Unpacking Community Health Funds Community Health Funds (CHFs) were piloted in Igunga district in 1996 and then progressively rolled out throughout the country. Of 187 districts, 155 have already introduced CHFs.23 The scheme was designed with three main objectives. The first objective was to increase local revenue generation earmarked to health. As with user fees, the introduction of prepayment mechanisms was part of the government’s effort to raise additional resources to address the important financial constraints faced by the health sector. The second objective was to improve the quality of health services. As mentioned, the unavailability of funds at the facilities was identified as a major driver of the low quality of service delivery. CHF funds would allow providers to respond to stock-outs in a timely fashion and therefore improve the quality of services. The third objective was to improve the management of health services and to empower communities to take an active role in health policy making.24 An assessment of the decentralization process had concluded that local authorities were performing poorly in the management of social services, and CHFs were devised as a way of strengthening their managerial skills. An implicit goal of CHFs was to provide financial protection to the most vulnerable. The Community Health Fund Act of 2001 defines the organizational structures required for the Fund to operate and describes CHFs’ institutional setup. At the district level, the Council Health Service Board (CHSB) is responsible for the management of the scheme. The Council includes representatives from service providers, members of the community and local authorities. It oversees the activities of the schemes in the district and it monitors the use of funds. The district CHF coordinators supervise the operations of the scheme, and they report on the number of households enrolled and the contributions collected. At the ward and village level, the Ward Development Committee and the Village Health Committee, together with the Health Facility Governing Committee and facilities’ health workers, mobilize the community to enroll in CHFs. Besides the organizational structures required by the Community Health Fund Act to operate, there are no clear guidelines for implementation of the schemes, and LGAs have autonomy to decide on the design of CHF in their districts. Important implementation aspects such as the definition of 24 CHF premiums, of the criteria applied to exempt poor populations, and of the benefits package, are determined by CHF’s bylaws. CHF bylaws are written by district authorities, which has led to significant heterogeneity in CHF schemes across the country and poses challenges to making generalizations about CHF’s features. Revenue Collection, Pooling, and Purchasing of Services CHF has two main sources of income—member contributions and matching grants paid by the central government. CHF contribution rates are determined at the district level and range between T Sh 5,000 (US$2.20) and T Sh 30,000 (US$13.40) per household per year. There is a CHF fee waiver system for poor people. The national average contribution is T Sh 10,752 (US$4.80) per household per year. In some districts, contributions are paid per individual. The central government matches all member contributions through the matching grant. The main purpose of the matching grant is to improve CHF’s balance sheets and to target health spending to the most vulnerable. In most districts, CHF revenues are collected at the facilities and pooled at the district level in the district’s cost-sharing account. Cost-sharing funds include out-of-pocket payments, CHF contributions, and NHIF reimbursements. Matching grants are then transferred to this account. There is no separate pool of funds for the poor. When facilities want to use the funds in the district’s cost-sharing account, facility managers need to request them from the CHSB. If the facility has a bank account, the CHSB can transfer the funds back to the facility. Otherwise, the CHSB purchases the items identified by the facilities and distributes them. The allocation of cost-sharing funds is not necessarily linked to facilities’ premium collection, and it can be done based on general needs. In this sense, CHF represents a redistribution mechanism among facilities within a district. As allocations are not always linked to activity, facilities do not perceive district’s transfers of cost-sharing funds as a reimbursement for the services provided. For facilities that receive less funding than what they had originally collected, this can create adverse incentives for the enrolment of new beneficiaries. Cost-sharing funds represent a small share of the resources available at the facility level, and CHF premiums often represent an important percentage of these funds. In districts with high CHF coverage, the share of CHF to total cost-sharing funds is higher and copayments are rare. A World Bank study analyzed the flow of funds in 30 health facilities in 9 LGAs in Geita and Shinyanga regions and found that cost-sharing funds represented only 1.6 percent of LGA resources, and CHF contributions represented one-third. In most districts, CHF membership only covers preventive and curative services at the primary health care level (dispensaries and health centers), including all diagnostics and medicines. Portability of coverage is limited and beneficiaries can often access services only in the facility where they are registered (first point of service). In some districts, beneficiaries can get treatment in all public dispensaries and health centers within the district. In addition, LGAs have service agreements with some private providers (faith-based organizations and private for-profit). In districts with higher membership fees, coverage also includes outpatient care in district and regional hospitals. A referral letter is required to access these services. Only a few districts have 25 managed to offer inpatient services, and these require copayments. The revision and update of CHFs’ benefits package has not been institutionalized, and there is no evidence of changes in the scope of services offered by CHF. Even though CHF premiums are higher in districts offering a more comprehensive benefits package, there is no evidence that premiums are determined on a cost-recovery basis. Furthermore, there has been no systematic assessment by the government of the costs and revenues of the scheme. Independent research shows that at the current level of CHF premiums, the combined cost of running CHFs and providing services to beneficiaries exceeds the revenue generated. While this deficit challenges the financial position of health facilities providing services to CHF beneficiaries, it is important to consider that CHF is a fund-raising mechanism, but it is the government’s responsibility to ensure the financial sustainability of the health sector. Enrollment CHF was envisaged to serve rural communities and informal workers. There are, however, no eligibility criteria to enroll in the scheme. In most districts, the household is the unit of enrolment. Households might comprise up to six members (main recipient, spouse, and up to four additional dependents). Beneficiaries are registered at the facilities when they seek care (passive enrolment). There are concerns that this is leading to problems of adverse selection. Some districts practice group enrolment (for example, through public-private partnerships with local associations and cooperatives) and mass enrolment campaigns (for example, CHF day in Kahama district) to facilitate the registration process and mitigate adverse selection. In Bariadi district, for example, enrolment campaigns are organized during school registration in two local secondary schools. In Igunga district, in turn, sensitization campaigns take place during the harvesting season to reduce financial barriers to access to CHF. Upon registration, beneficiaries must pay the annual fee in exchange for which they receive a CHF card. Facilities are responsible for collecting household data (name of contributors, name of dependents, age of beneficiaries, photo of beneficiaries, date of enrolment and date of renewal), and local authorities aggregate the database of beneficiaries at the district level. Coverage is yearly, and only few districts have explicit mechanisms in place to promote automatic re-enrolment after coverage expires (for example, Lushoto district has systematic renewals for group enrolment. CHF coverage has increased significantly in recent years, from 6.9 percent of the population in 2011 to 19.8 percent in 2015. Recent estimates conclude that there are 1,452,855 households (8,717,130 individuals) enrolled in CHF schemes. In this regard, CHF has outperformed similar schemes in the region and reached a high level of coverage for a voluntary scheme targeting the informal sector. There are, however, salient differences in CHF coverage among districts. Local authorities’ engagement in awareness raising has been identified as an important determinant of coverage. Furthermore, nonrenewal and dropout rates are high, especially in years with low agricultural yields. The limited number of services included in the benefits package represents a challenge for the uptake of CHF, and is often mentioned as a determinant of dropout rates. In addition to the limitations in the comprehensiveness of the benefits package, beneficiaries complain about the unavailability of services covered by CHF, particularly medicines. 26 Poor households are identified using a community-based approach. This means that community leaders and authorities, often in consultation with the community, determine who is poor according to their own definitions and perceptions of poverty. In Muheza district, for example, the criteria include aspects of household composition (for example, presence of elders or orphans) and housing (for example, access to water). This approach has the advantage of being comparatively inexpensive and of taking into account context-specific factors. Community-based approaches also tend to be more successful at identifying marginalized groups. An important limitation of this approach is that it is hard to compare poverty levels among communities, given that the criteria applied by each community might defer. There is no national guideline that regulates which criteria can be applied for the CHF waiver policy. In some districts, CHF authorities collaborate with other institutions (Social Work Department, Social Welfare Department) and rely on them to identify the poor. After the Village Council has identified the poor, ward authorities must validate the list of poor people. CHF cards are issued by the Council Health Service Board to those endorsed by ward authorities. District authorities are responsible for seeking alternative sources of funds to compensate facilities for the services provided to exempted CHF members. Even though the exemption process is well defined, its implementation is defective. Only a small number of districts provide systematic financial support to cover the poor. During a speech, former President Mkapa urged districts to cover CHF premiums for the poor. Since CHF schemes are managed at the district level, the response has been diverse. In some districts, authorities rely on NGOs and private for-profit companies to subsidize CHF premiums. This is often a one-time support (for example, Kahama district), and is therefore considered unsustainable. A good practice fostered by the MHCDGEC is the introduction of a line in the budget of LGAs to ensure the availability of funds to cover the poor. Anecdotal evidence suggests that 1 in 10 districts allocate their own resources to pay for CHF premiums for the poor. Problems in implementation of CHF waiver policies are closely related to the ineffective implementation of user fee waivers. Districts are reluctant to use their own funds to subsidize CHF premiums, considering that the central government should be covering the cost of health services for this group. Similarly, in the districts where local authorities do not provide financial support to subsidize CHF coverage to the poor, facilities have adverse incentives to apply the exemption, as they are not reimbursed for the costs of treating exempted patients. The absence of adequate reimbursement mechanisms, and clear responsibilities for covering the costs of health services for the poor, results in the deprivation of access to health care. There are few studies assessing the impact of CHF coverage on the effective utilization of health care services. There is no evidence, however, that contributing households access services different from those members whose premium is subsidized. 27 CHF’s Management Reform In 2009, the NHIF signed a memorandum of understanding with the MHCDGEC and the PO- RALG to take over CHF’s management responsibilities from the MHCDGEC.25 The main goals of the reform were to improve efficiency and increase coverage. To achieve the first goal, the NHIF supports district authorities to prepare matching grant applications and advise them on efficient ways to allocate these funds. Furthermore, the NHIF monitors that LGAs’ use of matching grants following cost-sharing guidelines and assists districts that are failing to use CHF funds to improve quality indicators. This has started a community of knowledge-sharing among CHF- implementing LGAs that seeks to close the quality gap in underperforming districts. The NHIF is additionally helping with data collection, by providing the books facilities need to use to gather relevant information. To achieve the second goal, the NHIF is providing support during sensitization campaigns. Assessments had shown that CHF awareness was low, and this was partly attributed to CHF’s reduced sales force. Since 2009, the NHIF—often with the support of development partners—has introduced CHF awareness raising to its own sensitization campaigns.26 Awareness of CHF has increased, and a recent study shows that sensitization campaigns are among the most frequently cited sources of information about CHFs. The same study reports, however, that even though beneficiaries are aware of CHFs, their knowledge about CHFs is poor. 5. Implementation Challenges and Opportunities This case study has identified two major limitations to the provision of health coverage to the poor. The first is the low quality of services in rural areas, and the second is the limited financial protection. To what extent have the efforts described in the previous sections contributed to overcoming these limitations? Have supply-side efforts improved the quality of health services in rural areas? Have CHFs offered financial protection to the poor? While there has been no clear prioritization of the poor guiding the implementation of devolution, it has been argued that this process would contribute to the promotion of a health system that is more responsive to the needs of the poor. This is based on the understanding that LGAs are in a better position to observe the needs of the poor and use this information to design programs that meet these needs. An evaluation of the devolution of health services, however, found that citizens do not feel like their voices are, in practice, being heard. This can limit LGA effectiveness in addressing the challenges faced by the poor. Furthermore, in the current distribution of functions in the health sector, lower-level officials do not have the authority (or the responsibility) to design policies and programs. In terms of quality improvement, the devolution of health services could entail that efforts are not just concentrated in selected areas and, instead, all LGAs implement quality improvement measures. An important shortcoming of the implementation of these policies and programs in rural and remote areas, however, is the limited capacity of government officials at the LGA level. Strengthening PHC, in turn, is considered an effective way to reach the poor and improve equity in health service delivery. Other countries in the region, like Ethiopia and Kenya, have adopted 28 similar approaches. While many PHC strengthening initiatives include quality improvement measures, general PHC strengthening interventions can increase health services quality by increasing the availability of health services. In Tanzania, PHC strengthening focused primarily on the construction and rehabilitation of health centers through MMAM, and on closing the HRH gap. In both cases, prioritization of the poor has been based on geographic criteria, and pro- poorness has been conceived as disproportionally benefiting residents of rural areas. In terms of physical infrastructure, only 36 percent of all rural facilities had electricity, water, and sanitation, compared to 79 percent of urban facilities. Even though Tanzania performs better than other countries in the region, the inadequacy of the physical infrastructure in rural health facilities represents a challenge for the provision of quality health services. The investment required to provide electricity, water, and sanitation to these facilities is, nevertheless, significantly lower than the investment required to build these centers. In terms of HRH, there are insufficient human resources in most cadres, including doctors, nurses, and pharmacists. The civil service reforms of the 1990s had a severe impact on the number of health personnel, and improvements in this area have been slow. The interventions described in section 2 contributed to important progress in terms of the training of health professionals, but this has disproportionally benefited the rich. This is explained by problems with the geographic distribution of HRH, which is extremely unequal. In theory, HRH policies can be deemed as pro- poor, to the extent that most policies (including the Human Resources for Health and Social Welfare Strategic Plan) focus on the relative increase in HRH allocated to rural areas. In practice, however, HRH are not deployed to the areas where they are needed the most: unlike policies regulating the allocation of school teachers, HRH policies are designed to close HRH gaps. This means that HRH are deployed to areas that fail to meet the required staffing levels, and priority is not given to those areas where the situation is the most critical. Furthermore, new recruits assigned to rural and hard-to-reach areas often do not take the job or quit: of 4,812 new positions offered in 2014, 63 percent were taken and 13 percent quit. As a result, the distribution between urban and rural areas is extremely unequitable, and only 28 percent of all health professionals (and 9 percent of all doctors) reside in rural areas.27 In addition, health personnel in rural areas are less qualified than their colleagues in urban settings. This is reflected in lower diagnostic accuracy, lower adherence to clinical guidelines, and greater difficulty managing maternal and neonatal complications.28 RBF is possibly the intervention that most explicitly targets quality improvements. It is difficult to assess the extent to which RBF has, in practice, improved the quality of health services, as it has only been implemented since 2015. Initial assessments of facilities participating in the RBF program show promising results, with improvements in most quality dimensions. These results were discussed during the RBF Week and RBF National Steering Committee meetings in 2016, and plans to extend the program to other regions are under preparation. An important requisite for quality health services is the availability of essential drugs. RBF intends to improve the availability of medical supplies across the country by providing adequate incentives to Medical Store Department management. Nevertheless, essential medicines were out of stock in 50 percent of the facilities and, despite the accreditation of private pharmacies to serve health 29 insurance beneficiaries, the availability of medicines is still low. The service delivery indicators show important differences in the availability of 14 tracer medicines between rural and urban facilities. At the facility level, the main cause of stock-outs is the unavailability of funds with which to order medicines. By design, however, RBF can be considered to be pro-poor: it has been introduced in areas with high poverty rates and poor health outcomes, and it is also intended to be scaled up following a similar logic. In addition, PHC facilities implementing RBF are incentivized financially to provide services to TASAF recipients, and the readiness fund should help underperforming facilities (most likely those serving the poorest groups) to improve the quality of the services provided. In this case, the prioritization of the poor is explicit and based on income. There are only a limited number of interventions designed to cover the poor for catastrophic and impoverishing health expenditures. This might be explained by the fact that these interventions would not be required if the national policy to waive user fees for the poor had been effectively implemented. The government’s most explicit effort to compensate for its poor enforcement are CHFs. But to what extent do CHF schemes provide financial protection and how pro-poor are they? CHF beneficiaries are mostly low-income households in rural areas. There are no studies that provide a complete picture of the socioeconomic characteristics of CHF beneficiaries.29 This hampers the evaluation of the extent to which CHF covers the poor. Unlike RBF, by design no clear priority has been given to the poor, and only a few districts offer free CHF coverage to the poor. An important element of pro-poor financial protection is a progressive and equitable financing mechanism. Lump-sum prepayments that are not linked to individuals’ capacity to pay, such as CHF, are regressive. Using two measures of health financing progressivity (the Kakwani and the concentration indexes), CHF contributions were found to be the most regressive sources of health financing. In fact, CHF contributions were more regressive than out-of-pocket expenditures. The concentration index for CHF contributions was, however, insignificant. This might be explained by the small share of households with CHF coverage at the time of the study, and the low share of CHF contributions to total health expenditure. Matching grants, in turn, favor richer districts. While districts with more beneficiaries receive more money in absolute terms, transfers per beneficiary are higher in districts where premiums are higher. Even though there are no studies showing that premiums are positively correlated to the relative socioeconomic status in each district, it can be expected that premiums are higher in wealthier districts. In this regard, matching grants are pro-rich. In addition to being regressive in terms of how CHF contributions are paid, some questions have been raised about the extent to which CHF offers financial protection to its beneficiaries (depth of coverage). A recent estimation of catastrophic30 and impoverishing health expenditures suggests that CHF beneficiaries are more likely to experience catastrophic health expenditures than households with no health insurance, but impoverishing health expenditures are less likely among CHF beneficiaries than among those without affiliation. The share of households covered by CHF that incurred catastrophic health expenditure is, nevertheless, low (4.5 percent). 30 Two interrelated reasons might explain the higher share of catastrophic health expenditure among CHF beneficiaries. The first relates to the issue of adverse selection. Passive enrolment entails that it is mostly the sick who purchases CHF coverage. The second reason relates to the limited scope of services covered by CHF. Services beyond PHC are covered by CHF in only a few districts, and even PHC services are not always available, forcing beneficiaries to purchase them from private facilities. Defenders of CHF argue that its objectives can be reached only if the scheme reaches a critical level of coverage. If higher levels of coverage would be reached, districts would have more funds available to cross-subsidize the poor and cover a wider range of health services. It is also important to highlight that the coverage of CHF has increased significantly in the last three years. Most assessments of financial protection use data collected up to 2012. It is difficult to ascertain whether the current levels of coverage allow districts to mobilize sufficient resources to provide financial protection for the poor. Given the variations in CHF features among districts, there are examples of districts that have managed to overcome some of the above-mentioned challenges. The subsidization of CHF premiums for the poor using districts budgets, and the expansion of CHF benefits have been identified as best practices for financial protection of the poor. Challenges in the provision of financial protection have led to the persistence of financial barriers to access to health services. The percentage of women who face financial barriers to health services increased between 2005 and 2015, and these disproportionally affect the poor: while 62.8 percent of women in the poorest quintile faced financial barriers, 33.1 percent of the women in the richest quintile had similar difficulties. Moreover, financial barriers are more common in rural areas, where 53.8 percent of the women had problems getting the money to pay for medical treatment, compared to 41.9 percent of the women in urban areas. 6. Looking Ahead Due to the complexity and multiplicity of challenges to cover the poor in rural areas, it is unlikely that a single intervention will be capable of closing the quality gap. It is thus important to assess the coherence among the several policies and programs implemented by the Government of Tanzania, and to understand whether these approaches produce synergies that induce a multiplicative effect on the coverage of the poor. The high fragmentation in the implementation of pro-poor policies and programs is evidenced by the lack of a consistent approach to the prioritization of services for the poor. As mentioned, efforts to address the challenges to equitable access to health services either lack a prioritization criterion (like the devolution of health services) or they are guided by different criteria. In this context, the Tanzania’s Health Financing Strategy (HFS), expected to be approved by late 2017, can play a key role in advancing UHC inclusive of the poor. The HFS was conceived with the objective of improving the health sector’s responsiveness to the needs of the poor, among other objectives. In fact, equity is one of the strategy’s guiding principles, showing the political will to 31 close the access gap in the health sector. Furthermore, the HFS provides a strategic direction that can help harmonize the different efforts described in this case study. The core of the reforms proposed by the HFS consists of the creation of a mandatory Single National Health Insurance (SNHI) for all Tanzanians. The SNHI will have a single risk and financial pool, which allows for cross-subsidization between the rich and the poor (the fee for the poor will be subsidized). The HFS sets an explicit strategic objective of establishing a pro-poor financing mechanism, and suggests adapting TASAF’s tool to identify the poor in order to better target the lowest income quintiles. In addition, the scheme will cover a standard benefits package for all Tanzanians that includes health services in dispensaries, health centers, and district hospitals. A referral letter will be required to access services in district hospitals. If the HFS is approved and implemented, the SNHI is expected to significantly increase the financial protection of the poor. The focus on equity introduced by the HFS provides an excellent opportunity to work on some of the challenges to appraise and implement reforms that are inclusive of the poor in Tanzania. Successful implementation of SNHI requires not only covering the poor, but also having good- quality services available for the poor as a prerequisite. The lack of information about the sociodemographic profile of the users of health services hinders the evaluation of the pro-poorness of these schemes. Monitoring the performance of the SNHI will therefore require major improvements in the data collection process, to capture differences in coverage, access, and utilization by wealth quintiles and geographic area. The HFS recommends the development of a Health Management Information System, which presents an opportunity to introduce socioeconomic information about the beneficiaries that would allow for evaluations of the pro-poorness of the scheme. Furthermore, the information provided by the star rating quality assessment tool could be paired with socioeconomic data to monitor the performance of those facilities that serve the poor. 32 Annex Figure A.1 Trends in Total Health Expenditure (THE) and THE Per Capita 90,000 10% 78,326 80,000 71,295 73,365 70,000 8% 60,000 54,529 47,467 6% Bn Tshs 50,000 % 40,000 4% 30,000 22,634 20,000 2% 10,000 0 0% 2002/3 2005/6 2009/10 2011/12 2013/14 2014/15 THE as a % of Nominal GDP THE per Capita (Tshs) Source: NHA 2002/3; NHA 2005/6; NHA 2009/10; NHA 2011/12; NHA 2013/14; NHA 2014/15. Figure A.2 Trends in Financing Sources 100% 90% 27.4% 80% 39.6% 41.0% 37.0% 44.1% 48.3% 70% 60% 50% 47.1% 34.4% 35.0% 40% 27.8% 33.0% 29.7% 30% 20% 25.4% 28.1% 26.0% 26.0% 28.0% 10% 20.7% 0% 2002/3 2005/6 2009/10 2011/12 2013/14 2014/15 Public Private Donors Source: NHA 2002/3; NHA 2005/6; NHA 2009/10; NHA 2011/12; NHA 2013/14; NHA 2014/15. 33 Table A.1 Sources of Revenue and Financing Schemes Mandatory Private Tanzanian Voluntary Household Other social sector Rest of shilling Government health out-of- financing health agents the Total (TZS), schemes insurance pocket schemes insurance (NGOs and world Million schemes payment (n.e.c.) schemes enterprises) Transfers from government domestic revenue 100% 647,710 (allocated to health purposes) Transfers distributed by government 99.2% 0.2% 0.6% 537,403 from foreign origin Social insurance 100% 86,794 contributions Voluntary 100% 69,386 prepayment Out-of-pocket 100% 772,070 Direct foreign 0.2% 90.4% 3.7% 5.8% 972,209 transfers Other revenues 100% 41,647 (n.e.c.) Total 1,224,091 86,794 69,386 879,660 772,070 35,657 59,560 3,127,220 Source: NHA 2011/12. 34 Table A.2 Quality Assessment Framework Assessment Area Characteristics of Facility at 5-Star Level ▪ Strong governance structure with decentralization of fiscal management ▪ Implements best practices for managing resources Facility Management & ▪ Appropriate skilled staffing complement per staffing Governance establishment ▪ Excellent working conditions and environment for staff including housing and appropriate incentives ▪ Accurate and comprehensive data Use of Facility Data ▪ Staff who can perform data analysis ▪ Staff who use data for service improvement ▪ Functioning performance system for staff Staff Performance Assessment ▪ Staff who have met over 80% of their performance targets ▪ Well-organized setup for service delivery Organization of Services ▪ Well-organized and efficient process for maintaining and accessing records ▪ Fully trained staff and a strong functioning system to triage, Emergency Care and Referral refer, if needed Mechanism ▪ Successfully handle emergency cases per the norms for the facility type ▪ Strong functioning Health Facility Governing Committee/Council Health Service Board that is responsive Client Focus and Social to the needs of the community Accountability ▪ Facilitates an inclusive planning process for health financing plans and by fully executing these plans ▪ Consistently available power, running water, and functional equipment Health Infrastructure & ▪ Infection prevention and control and waste management Infection Prevention Control systems that are implemented according to national guidelines ▪ Reproductive, maternal, neonatal and child health, family planning, outpatient and inpatient, and specialist services are Clinical Services fully provided according to standard protocols ▪ Minimal patient waiting times ▪ Continuous availability of medicines that are appropriately stored and rationally used ▪ Availability of quality diagnostic services according to the Clinical Support Services standards of the facility type ▪ For health centers and hospitals, they must have fully functioning operating theatre with measures in place to prevent sepsis Source: Health Sector Strategic Plan IV. 35 Figure A.3 Targeting Performance of CCTs in Selected Countries Source: Productive Safety Net Program Impact Evaluation Baseline. 36 Notes 1. During 2007–2011/12, growth became more inclusive than during 2001–07, and the growth elasticity of poverty, defined as the rate by which poverty declines for each percent of GDP per capita growth, was 1.02 during 2007– 2011/12. Overall, however, growth, which has been faster in capital-intensive sectors, and agriculture—where most of the poor are employed—grew at a slower pace (World Bank 2015a). 2. On average, over two-thirds of the population lives in rural areas (World Development Indicators). Given the high share of poor people living in rural areas, this study will often link these two dimensions. 3. The government has identified certain priority sectors for investment in its Big Results Now strategy. 4 Unlike the other conversions to U.S. dollars in this case study, this figure uses the exchange rate of Tanzania’s National Health Accounts 2014/15. 5. THE per capita increased between 2002 and 2012 and dropped in 2013. In 2014, THE per capita began a positive trend but has not reached 2012 levels. 6 The Bamako Initiative was adopted in 1988. It fostered the introduction of user fees in order to improve the quality of health services and ensure an equitable access to health care (Ridde 2003). 7. In 2015, 26 percent of THE was financed through out-of-pocket payments (National Health Accounts 2014/15). 8. In 2014, the average under-five mortality rate in low-income countries was 76.1 deaths per 1,000 live births, and the average rate in Sub-Saharan Africa was 83 deaths per 1,000 live births (World Development Indicators). 9. The graph compares Tanzania with other countries that had similar under-five mortality rates in 2000. 10. The graph compares Tanzania with other countries that had similar maternal mortality rates in 2000. 11. While 83 percent of children in the top wealth quintile had all basic vaccines, 65.2 percent of the children in the bottom wealth quintile completed the basic vaccination scheme. Furthermore, 82.2 percent of children in urban areas received all vaccinations, compared to 72.6 percent of children in rural areas. 12. In Tanzania, community health workers support outreach activities, but they are not government employees and they work on a voluntary basis. 13. Utilization rates follow a similar trend. In 2012, 71 percent of all consultations took place at a public facility, 21 percent at a private facility, and 8 percent at a faith-based organization (Household Budget Survey 2014). 14. MHCDGEC’s mandate is to train all cadres below the undergraduate level, including certificates, diplomas, and advanced diplomas. The Ministry of Education and Vocational Training trains all cadres above that level. 15. For the Health Basket Fund, the allocation is done based on population (60 percent), size of the council (20 percent), poverty rates (10 percent), and under-five mortality rates (10 percent). The Health Service Development Grant—an earmarked grant to support a government program to strengthen primary health care —combines an allocation formula with a performance-based allocation mechanism. 16. Boex, Fuller, and Malik (2015) found that only 25 percent of nonwage funds are reaching the dispensary and 18 percent the health centers, compared to 36.3 percent reaching the district hospitals and 20.5 percent retained by the Council Health Management Team. 17. The gap is calculated based on HRH requirements according to international guidelines and actual staffing levels. 18. RBF has been defined as “a cash payment or nonmonetary transfer made to a national or sub-national government, manager, provider, payer or consumer of health services after predefined results have been attained and verified. Payment is conditional on measurable actions being undertaken” (rbfhealth.org). 19. The performance of district authorities depends on, among other factors, the mobilization of households to enroll in CHF schemes. 20. In Tanzania, there are also other forms of health insurance, all of which have limited coverage. These include a health insurance scheme for the private sector (National Social Security Fund–Social Health Insurance Benefits), private health insurances, and number of community-based health insurances. 21. Coverage figures are not cumulative, and they represent the annual coverage of the schemes in 2015. 22. In Iramba district, exempted groups represent about 50 percent of the patient load (Maluka 2013). 23. The districts that have not introduced CHF yet are the recently created districts. 24. Studies found that CHF had no impact on the participation and involvement of the community in the decision- making process (Mtei and Mulligan 2007). 25. According to the Community Health Fund Act, the MHCDGEC is responsible for the provision of technical support and for monitoring the use of Community Health Funds in the districts. 26. There is no evidence that sensitization campaigns aim at reaching the poor. 27. The distribution of health professionals among dispensaries in the same districts is also high and ranges from 1 to 9 (Coffey 2015). 37 28. The unavailability of inputs and medical equipment might also explain the differences in these indicators. 29. An impact evaluation of TASAF’s CCT program (World Bank 2013) describes the socioeconomic characteristics of CHF beneficiaries that are TASAF recipients, but the literature around CHF beneficiary profiles is limited. 30 The study uses the 40 percent threshold. 38 39 40