Sector-wide or disease-specific? Implications of trends in development assistance for health for the SDG era

Sector-wide or disease-specific? Implications of trends in development assistance for health for... Abstract The record of the Millennium Development Goals broadly reflects the trade-offs of disease-specific financing: substantial progress in particular areas, facilitated by time-bound targets that are easy to measure and communicate, which shifted attention and resources away from other areas, masked inequalities and exacerbated fragmentation. In many ways, the Sustainable Development Goals reflect a profound shift towards a more holistic, system-wide approach. To inform responses to this shift, this article builds upon existing work on aggregate trends in donor financing, bringing together what have largely been disparate analyses of sector-wide and disease-specific financing approaches. Looking across the last 26 years, the article examines how international donors have allocated development assistance for health (DAH) between these two approaches and how attempts to bridge them have fared in practice. Since 1990, DAH has overwhelmingly favoured disease-specific earmarks over health sector support, with the latter peaking in 1998. Attempts to integrate system strengthening elements into disease-specific funding mechanisms have varied by disease, and more integrated funding platforms have failed to gain traction. Health sector support largely remains an unfulfilled promise: proportionately low amounts (albeit absolute increases) which have been inconsistently allocated, and the overall approach inconsistently applied in practice. Thus, the expansive orientation of the Sustainable Development Goals runs counter to trends over the last several decades. Financing proposals and efforts to adapt global health institutions must acknowledge and account for the persistent challenges in the financing and implementation of integrated, cross-sector policies. National and subnational experimentation may offer alternatives within and beyond the health sector. Sector-wide approaches, agenda setting, aid, Millennium Development Goals, donors, HIV, tuberculosis, non-communicable diseases, maternal and child health, health financing Key Messages Over the last quarter century, development assistance for health has disproportionately favoured specific diseases over sector support, with the latter peaking in 1998. More countries have received health sector support over time, but there is a weak relationship between country governance indicators and the proportion of assistance channelled in this way. Attempts to bridge disease-specific and sector-wide financing approaches have been limited, and new financing platforms have failed to gain traction. Recent estimates of the additional investment needed to achieve the health Sustainable Development Goal indicate that three-quarters of these resources need to be allocated to health system strengthening; yet, trends in financing persistently remain oriented otherwise. Introduction The field of global health is in transition from its heyday in the Millennium Development Goal (MDG) era, which saw exponential increases in attention and financing and the creation of new global institutions and initiatives, to plateauing funding within a much more crowded agenda space. These resources were channelled through two divergent funding strategies, reflecting distinct perspectives. Health sector and budget support aimed to better align development assistance with recipient country health priorities and financial systems. Earmarked resources targeted specific diseases, population groups or interventions. The trade-offs of disease-specific financing characterises the record of the MDGs more broadly: substantial progress in particular areas, facilitated by time-bound targets that are easy to measure and communicate, which shifted attention and resources away from other areas, masked inequalities within and between countries and exacerbated fragmentation, particularly at the national level (World Health Organization 2015). Scholars have noted swings in the orientation of global health over the decades between targeted and broader, systemic approaches (Mills 2005; Uplekar and Raviglione 2007; Bossert 2012; Hafner and Shiffman 2013). In many ways, the final set of Sustainable Development Goals (SDGs) reflect a profound shift of the pendulum back towards a more holistic, system-wide approach. Recent estimates of the additional investment needed to achieve the SDG health goal indicate that three-quarters of these resources need to be allocated to health system strengthening (Stenberg et al. 2017). At the dawn of the SDG era, the World Health Organisation (WHO) and others called for a deeper debate about the institutional architecture and financing needed to produce global public goods and to respond to country needs (Hill et al. 2014; UN Department of Economic and Social Affairs Technical Support Team in support of the General Assembly Open Working Group on Sustainable Development Group 2014; World Health Organization 2015). The Global Health Security Working Group on Health Financing has asserted: ‘it is clear that today’s global and national arrangements for health financing need to change’, noting high disability and premature mortality rates, high out of pocket spending and inadequate domestic spending for essential health services (Ottersen et al. 2017, p. 105). To advance these efforts, this article builds upon existing work on aggregate trends in donor financing, bringing together what have largely been disparate analyses of sector-wide and disease-specific financing approaches. Drawing on both primary and secondary, quantitative and qualitative data, it looks back across the last 26 years and asks: how have international donors allocated development assistance for health (DAH) between the health sector and specific diseases; to what extent does country-level governance and disease burden align with these allocations; and how have attempts to bridge these two approaches fared across funding platforms and diseases? Given these patterns, the article then discusses implications for the field moving forward and the ambitious agenda the global community has set for itself. These analyses present a clear picture: since 1990, DAH has overwhelmingly favoured the specific over the sector, with substantially more funding earmarked for individual diseases than channelled as health sector support. Most of the new global health institutions created in the new millennium followed this model. Attempts to integrate system strengthening elements into disease-specific funding mechanisms have been less successful in bridging the two approaches and more integrated funding platforms have failed to gain traction. Despite Paris Declaration commitments to align with national institutions and priorities, sector support largely remains an unfulfilled promise: proportionately low amounts which have been inconsistently allocated, and the overall approach inconsistently applied in practice. Thus, the expansive orientation and projected resource needs of the SDGs runs counter to the trends over the last several decades. Financing proposals and efforts to adapt global health institutions must acknowledge and account for the persistent challenges in the financing and implementation of integrated, cross-sector policies. Rationale for sector support and disease-specific approaches Scholars have debated for some time the relative merits of sector and disease-specific approaches, often characterized as horizontal/system strengthening versus vertical/targeted approaches; and, diagonal approaches, the elusive best of both worlds. Therefore, each is just briefly reviewed here. Although health sector support and earmarked funding for specific diseases or activities are not mutually exclusive, they represent distinct strategies, associated with particular perspectives on how best to improve development outcomes, with advantages and drawbacks that mirror one another. Sector-wide and budget support financing, which channels funding directly through national accounts, grew out of frustration with traditional project-based aid (Lawson et al. 2003; Koeberle and Stavreski 2006). It aims to reduce the existence of parallel delivery structures and increase the likelihood that programs will be sustained because they are integrated into government systems. Sector support is intended to enhance country ownership, enabling national governments to better target resources to their unique health profiles and address disease outbreaks and emergent health concerns. It can help develop workforce and infrastructure capacity that serves to support multiple disease control and prevention programs (Foster et al. 2000; Koeberle and Stavreski 2006; Eichler and Glassman 2008; Ooms et al. 2008; Peters et al. 2013). Some scholars argue that budget support is not a new financing mechanism; rather, that it represents the fourth generation of donor conditionality policies that begin in earnest in the 1980s with structural adjustment programs, and in the Paris Declaration era were framed in terms of country ownership (Hayman 2011; Langan 2015). Hafner and Shiffman (2013) document increased attention to health systems strengthening by multilateral and global health institutions from 2005 onwards, reflecting pressure to achieve the MDGs and concerns over health workforce shortages and unintended adverse effects of global health initiatives on national health systems. International donors have jointly signalled their commitment to align their assistance with recipient country priorities and financial and procurement systems as part of the 2005 Paris Declaration and subsequent high level fora on development cooperation, including the 2015 Addis Ababa Action Agenda.1 The flexible nature of sector support—its key advantage—makes it much more difficult to directly link financial investments to changes in specific outcomes. Moreover, the fungible nature of aid, the freeing up of government resources that can be used elsewhere, may have the perverse effect of reducing national expenditures on health (World Bank 1998; Deverajan and Swaroop 2000; Lu et al. 2010). Since funding is channelled directly through national accounts, donors are reliant on the government’s ability to manage finances and track funds in a transparent matter to reduce the likelihood that funds will be used for unauthorised expenditures. These capacities include managing budget processes which are governed by clear rules, with allocations linked to specific policies and expenditures linked to allocations, administering transparent procurement systems, ensuring timely and accurate financial reporting to track expenditures, and overseeing regular internal and external audits. Indeed, early evaluations identified low fiduciary risk (the extent to which actual expenditures differ from authorized expenditures) and government capacity to manage public finances as two key factors in effective budget support (USAID Development Information Services 2006; Lawson et al. 2006). In contrast, targeted funding is directed towards a specific disease (HIV), population group (children) or intervention (vaccinations). It can provide clearer links between financial inputs and health outputs and outcomes, facilitating results-based measurement. Targeted attention towards one disease can make it easier to garner initial public and political support, and with more visible results, easier to sustain support.2 The establishment of Gavi in 2000, the Global Fund to Fight AIDS, Tuberculosis and Malaria in 2002, and the President’s Emergency Plan for AIDS Relief (PEPFAR) in 2003 created a set of new institutions that embodied this targeted approach. At the same time, targeted funding for certain areas has also generated criticism. Earmarked funding has been critiqued for creating disease silos and siphoning resources away from lower profile issues (Behague and Storeng 2008; Levine 2008; Shiffman 2008; Epstein 2009; Shiffman, Berlan and Hafner 2009). Narrowly targeted funding can neglect health systems on which disease-specific interventions depend for successful implementation. Biesma et al. (2009) found that while the Global Fund, PEPFAR and World Bank Multi-country AIDS Program enabled swift scale-up and expanded stakeholder participation, they created parallel coordination structures, financial management, monitoring and reporting systems, and were not well aligned with overall national health priorities or harmonized with other donor efforts. These global health initiatives have increased the attrition rate of healthcare workers from the public sector to non-state providers, and efforts to scale-up responses to specific diseases has increased the burden on already overstretched health workforces (World Health Organization Maximizing Positive Synergies Collaborative Group 2009). While largely distinct approaches, particularly in their respective rationales, sector support and earmarked financing are not absolutely exclusive. Gavi, the Global Fund and PEPFAR have broadened their scope over time, allowing a portion to be used for health system strengthening (Goeman et al. 2010; Bowser et al. 2014; Moucheraud et al. 2016), discussed further below. Moreover, development assistance is not necessarily a zero-sum game among health conditions, health systems or across development issues. Indeed, until 2011 when DAH began to plateau, the 2000s witnessed a growing resource pie for the health sector overall and for nearly all health areas, including health system strengthening. That said, the 17 SDGs and 169 targets reflect a dramatically larger number of global ‘priorities’, in which health is much less prominent than its previous inclusion in three of the eight MDGs. The breadth in part reflects the mobilization of groups who learned from the HIV response and sought to raise the profile of their issue on the global agenda (Beaglehole et al. 2011; Bhutta et al. 2012; Smith and Rodriguez 2016). The establishment of the Partnership for Maternal, Newborn and Child Health (PMNCH) in 2005 and the non-communicable disease (NCD) Alliance in 2009—among many other new groups within and beyond the health sector—have created what the WHO (2015) characterizes as a competitive institutional landscape. In the next 25 years, DAH is expected to increase at a much more modest rate than in the past: 1.7% compared to 11.3% annually from 2000 to 2010 (Dieleman et al. 2016; Institute for Health Metrics and Evaluation 2017). Achieving the SDGs is expected to cost $1.4 trillion a year, with $274–371 billion annually in new resources needed for the health goal alone (Stenberg et al. 2017). Therefore, while the absolute volume of resources may continue to grow, the pace is slower and the global set of priorities and groups mobilized around them is much broader than in 2000. Methods Over the last decade, the availability and specificity of data on development assistance has improved substantially. This analysis draws on a number of sources: studies on sector and budget support, which are predominantly qualitative, and recent quantitative analyses of DAH, including associations between disease-specific funding and country burden of disease. Previous research on budget support in the international development literature was identified through Google Scholar and supplemented by articles on sector support published in previous issues of this journal and by references in articles on the topic. This article also presents new analyses to identify trends in system strengthening components of disease-specific funding and to examine the allocation of health sector support according to country-level governance indicators. These analyses are based on the Institute for Health Metrics and Evaluation (IHME) Development Assistance for Health data set, which tracks annual financial disbursements for health by source and channel to 166 recipient countries since 1990.3 It distinguishes funding to nine health focus areas: HIV, malaria, TB, maternal health, child and newborn health, NCDs, other infectious diseases, health sector strengthening/sector-wide approaches and other. The ‘focus’ of these nine areas varies considerably, with the latter four categories encompassing a much broader set of conditions and areas than the three named diseases and to a lesser extent, conditions affecting women, children and newborns. The 2017 data set for the first time disaggregates the amount of earmarked funding for a specific area that was dedicated to health systems strengthening. Therefore, it includes the total amount allocated to health sector support and the eight other areas, as well as the amount within each disease category supporting health systems strengthening (i.e. the amount of earmarked HIV funding dedicated to health systems strengthening). To examine the allocation of health sector support, the analyses use country-level World Bank Worldwide Governance Indicators (WGI), which group hundreds of variables from 32 sources, including Freedom House, the International Budget Project and the Afrobarometer. Selected for their comprehensive nature, the WGI cover six domains: government effectiveness, control of corruption, voice and accountability, political stability and absence of violence, regulatory quality, and rule of law.4 (Kaufmann et al. 2010). Quantitative analyses were conducted in SAS 9.4. These governance rankings provide a corollary to disability-adjusted life years (DALYs), in that they could help to inform donor decisions regarding how and where to allocate DAH. Given the importance of government capacity noted earlier, if donors are concerned about reducing fiduciary risk and maximizing effectiveness, the allocation of sector support should be positively associated with stronger governance indicators, particularly government effectiveness and control of corruption. By nature, countries with poorly performing institutions are less able to address the health concerns of their citizenry, who may suffer from a disproportionately higher burden of disease. This higher need may result in higher overall levels of DAH, including health sector support. However, one would hypothesize that donors would allocate a higher proportion of DAH channelled as sector support compared to earmarked aid to countries with stronger government capacity. Although the DAH data set represents a vast improvement over the data that was available just over a decade ago, it is not without its limitations. A large proportion of assistance, 15–44% of annual DAH, fell into the ‘other’ category and an additional 0.19–10% was unallocable. Categories were assigned based on project key words, which may not reflect all uses of the funds. For some countries and years, data were generated by statistical models to estimate disbursements. Moreover, the data set excludes government and private health spending, assistance from Gulf States, BRICS countries, charities outside of the US and the for-profit private sector (Dieleman et al. 2014, 2016; Nugent 2015). Results Trends in disease-specific funding Over the last 26 years, the majority of DAH has been allocated to specific diseases, ranging from 42% to 74% annually, from $3.1 billion in 1990 to $26.8 billion in 2015 (Figure 1). Previous studies have reported increases in both sector-wide and particularly disease-specific funding over time (Shiffman 2008; Shiffman, Berlan and Hafner 2009; Ravishankar et al. 2009; Dieleman et al. 2016), so only a brief overview is included here. Since 2000, DAH has prioritized the MDGs, with HIV, malaria, tuberculosis, maternal, newborn and child health growing substantially faster than other areas (Dieleman et al. 2016). Increases have been the most dramatic for HIV, rising from 4.4% in 1990 to 26% of DAH in 2015 and as high as 31% in 2010. Funding for malaria increased from 0.9% to 6.4% and for TB from 0.4% to 3.5%, still proportionally small amounts but large relative increases over the last two and a half decades. The relative proportion of DAH allocated to maternal and child health dropped from 35% in 1990 to 29% in 2015, but still constitutes a sizable share of the total. The proportion of funding allocated to NCDs has remained relatively constant, at its highest at 2.9% of DAH in 1993 (Institute for Health Metrics and Evaluation 2017). Figure 1. View largeDownload slide Amount of development assistance for health allocated to specific diseases, health sector support and other areas: 1990–2015. Note: ‘Other’ refers to cases with project-level information that does not fall into any of the health focus areas tracked by IHME. Figure 1. View largeDownload slide Amount of development assistance for health allocated to specific diseases, health sector support and other areas: 1990–2015. Note: ‘Other’ refers to cases with project-level information that does not fall into any of the health focus areas tracked by IHME. The overwhelming allocation of earmarked funding is also reflected in WHO finances, which have experienced a steady increase in the proportion of extra-budgetary funding over time, nearly 80%, with donors earmarking 93% of voluntary funding in the 2014–15 budget (Clinton and Sridhar 2017). This trend is not unique to the health sector. More broadly, earmarked funding from bilateral agencies that is channelled through multilateral organizations has increased 4-fold over the last two decades (Eichenauer and Reinsberg 2017). Country allocation of disease-specific funding varies by health area, and overall portrays an encouraging picture, with exceptions. Institute for Health Metrics and Evaluation reports (2016, 2017) indicate that alignment between DAH and DALYs has been the strongest for HIV, TB and child health. Funding for maternal health and malaria has been less well aligned with countries’ relative burden of disease. The relationship between the two has been the worst for NCDs. Allen (2016, 2017) posits that this mismatch could be explained by underestimates of NCD funding, their non-infectious nature, private sector producer interests, a weak evidence base and the lack of a compelling narrative. Country GDP and DALYs significantly predicted the amount of country-level DAH, with high-burden, low-income countries receiving the highest amount across all health areas. However, there are notable outliers where countries received disproportionately more or less funding than would be expected given their burden of disease (Dieleman et al. 2014).5 Across diseases, Figure 2 looks at trends in specific health areas prior to and following key global agenda setting moments. As mentioned earlier, there have been dramatic increases in funding for HIV, TB and malaria, with notable spikes following the MDGs and the creation of the Gates Foundation, Global Fund and PEPFAR—funding levels which have begun to slightly decline in recent years, but have not sharply dropped off over time. This provides a vivid example of the power of global norms and particular institutions in directing attention and resources that has been sustained over time. Figure 2. View largeDownload slide Global agenda setting milestones and trends in health funding over time: amount and proportion of overall development assistance for health allocated to different health areas. Figure 2. View largeDownload slide Global agenda setting milestones and trends in health funding over time: amount and proportion of overall development assistance for health allocated to different health areas. A similar response has not been true for other diseases. The NCD Alliance and subsequent High Level Summit have not shifted funding. Declines in funding for maternal and child health have reversed following the establishment of PMNCH, although the overall proportion of funding for this health area remains lower than in 1990 and 2000. These figures illustrate that absolute and relative changes in DAH has varied across diseases and sector support, and that there has been a staggered institutional response at the global level across diseases. The graphs should not be interpreted as causal relationships between global agenda setting moments and DAH. Taken together, these overall trends indicate that funding allocations strongly favour specific diseases and MDG areas. The alignment between disease burden and resource allocation has varied by disease, but overall reflects a positive association. Trends in health sector support In contrast, the record of health sector support largely remains an unfulfilled promise. Although sector and budget support was strongly promoted by some donors, including the European Commission, World Bank, Nordic countries, the Netherlands and the UK (Hayman 2011; Faust et al. 2012; Langan 2015), the scale of the shift has remained small. Paralleling overall increases in DAH since 1990, absolute funding levels for sector support have increased over time from $543 million in 1990 to $3.4 billion in 2015, peaking at $4.2 billion in 2011. The proportion of overall DAH for health sector support was at its highest level, 18.5%, in 1998 and has remained relatively constant at 9–12% over the last decade. Notably, the pattern of health sector support over time has differed compared to trends in financing for specific diseases, with the highest average annual gains seen in the 1990s at 11.4%, lower from 2000 to 2009 at 7.1% and tapering off to 2.3% since 2010 (Institute for Health Metrics and Evaluation 2017). Thus, unlike stated intentions, sector support has not substantially increased over time relative to other funding modalities, nor has budget support been implemented as an incremental process through which countries transition along a continuum of recipient country control from project aid towards general budget support. Donors have maintained hybrid portfolios which include a range of instruments that vary in the extent of their traceability and earmarking (Handley 2009, Ahsan et al. 2016). While small in terms of the overall relative amount, allocations of health sector support have become more widespread over time. In 1990, half of recipient countries received any health sector funding, rising to 76% 2 years later, to 85% in 1996 and to 94% in 2000. From 2003 onwards, nearly all countries have received a fraction of their DAH as sector support from at least one donor. Across sources of sector support, most donors have increased the absolute amount they are allocating over time, although levels fluctuate, peaking in the USA in 2004 ($941 million), Sweden in 2007 ($110 million), the Netherlands in 2011 ($141 million) and the UK in 2015 ($496 million). Among bilateral donors, South Korea has demonstrated the greatest absolute increases, initially allocating very low amounts of sector support but since 2008 exceeding individual Nordic countries. The amount channelled through multilateral agencies has increased overall but has varied across the time period as well: with $4 million channelled through the European Commission in 1996 and $140 million in 2015, peaking at $222 million in 2013. DAH channelled through the World Bank [International Bank for Reconstruction and Development (IBRD)] increased from $25 million in 1990 to $240 million in 2015, reaching as high as $1.4 billion in 2010. Across all channels of support, the proportion of sector support provided through the WHO has declined steadily from over 60% in 1990 to a low of 14% in 2013.6 The proportion channelled through World Bank has fluctuated from 5% in 1990 to 48% in 1998 to 32% in 2011 to 7% in 2015. Qualitative assessments indicate there is considerable variation among agencies regarding the purpose of budget and sector support, and how it is implemented across different countries. This variation includes differences in selection criteria, the extent to which recipients are required to fulfil specific conditions and how strictly these conditions are monitored, what constitutes a breach and how donors respond, and the relative emphasis between technical and political elements of policy dialogues. These differences send mixed signals to recipient governments; in some cases, donors have responded in opposite ways to the same incident7 (Handley 2009; Faust et al. 2012; Molenaers 2012; Tavakoli and Smith 2013). Examining the relationship between national governance rankings and the proportion of DAH allocated to sector support, correlations are positive but very weak, ranging from r = 0.08 for political stability to r = 0.14 for government effectiveness.8 To analyse trends in allocation among groups of countries with stronger and weaker governance indicators, we examined the median proportion and range of total DAH that has been allocated to sector support across country quintiles. This analysis did not demonstrate marked differences across country groups; Figure 3 illustrates these overall patterns through several examples. Figure 3. View largeDownload slide Proportion of development assistance for health allocated to health sector support across countries with different Worldwide Governance Indicator rankings. (a) Countries ranking at the bottom, median and top of the Worldwide Governance Indicators. This includes two examples of bottom ranked countries: Somalia, which has featured prominently in world politics, and low profile Equatorial Guinea. (b) Countries with the biggest variation in Worldwide Governance Indicator rankings over time. Note: These graphs use 3-year moving averages to account for large tranche payments in a particular year. Figure 3. View largeDownload slide Proportion of development assistance for health allocated to health sector support across countries with different Worldwide Governance Indicator rankings. (a) Countries ranking at the bottom, median and top of the Worldwide Governance Indicators. This includes two examples of bottom ranked countries: Somalia, which has featured prominently in world politics, and low profile Equatorial Guinea. (b) Countries with the biggest variation in Worldwide Governance Indicator rankings over time. Note: These graphs use 3-year moving averages to account for large tranche payments in a particular year. Looking at countries at the top, median and bottom of the governance rankings,9Figure 3a shows some variation across countries and over time, with common peaks in the mid-1990s and early 2000s—but not consistently ordered according to governance rankings or nearly as prominent as would be expected given the difference in institutional contexts. Figure 3b illustrates the proportion of DAH channelled as health sector support for two countries with some of the biggest changes (highest standard deviations) in government effectiveness and control of corruption over time. Again, like the cross-national trends, the proportion of health sector support in these countries was highest in the 1990s. In Rwanda, where governance has steadily improved over time, health sector support has been inversely related to this trend, with lower proportions of sector support as government effectiveness and control of corruption rankings have improved. In Cote d’Ivoire, declining health sector support paralleled declining governance from 1996 to 2010; however, sector support has remained low despite subsequent improvements in more recent years. The weak, inconsistent relationship between governance indicators and allocation of DAH is similar to findings regarding suspension of budget support. Molenaers et al. (2015) found that WGI voice and accountability trends were associated with a reduced probability of suspension; however, trends in government effectiveness, political stability and control of corruption were non-significant. Thus, despite donor concerns about fiduciary risk and government capacity, country governance indicators do not appear to guide allocation decisions of sector support to the extent that DALYs guide some disease-specific funds to countries in greatest need. In addition to the allocation of sector support, previous studies have also attempted to assess the effectiveness of this modality, finding that outcomes are highly context dependent. Overall, budget support has been most promising in improving the pro-poor nature of public expenditure, public financial management and donor harmonization (Woll 2008; Handley 2009; Tavakoli and Smith 2013; Ahsan et al. 2016). At the same time, it is transactionally intensive, and has been less influential in improving civil service functions or reducing corruption (Tavakoli and Smith 2013). For example, an assessment of sector-wide approaches in Bangladesh over the last 15 years found improvements in monitoring, procurement, supply chain and financial management, but challenges in supporting long-term training, rural physician retention and decentralizing functions to subnational levels (Ahsan et al. 2016). In terms of effects on government health spending, Fernandes Antunes et al. (2013) find that general budget support showed no effect, based on a panel data set of 82 low and middle-income countries. External funding for health was associated with declines in health expenditures from domestic sources, but it was not a full substitution effect, in that it was associated with increases in total government spending on health from all sources. Despite stated intentions to improve country ownership, budget support policy dialogues can still be dominated by donors, in some cases more powerfully than before when they act as a group rather than independently (Woll 2008, Handley 2009). Donors may use policy dialogues to increase their leverage for economic and trade liberationsation (Langan 2015). And, they have demonstrated that they are willing to suspend budget support, most often related to fraud, repression or human rights violations (Molenaers et al. 2015), thus reducing the consistency of this aid modality, a key rationale for its use (Hayman 2011; Tavakoli and Smith 2013). In sum, these trends in health sector financing indicate rather tepid support for this modality: amounts representing a relatively small proportion of overall DAH that have declined during the MDG period, with inconsistent allocation and application in practice, likely a result of and contributing to mixed results of its effectiveness. Efforts to bridge the vertical–horizontal divide As noted earlier, while the rationale for disease-specific and sector-wide funding approaches are distinct, they are not mutually exclusive. In response to critiques, institutions have expanded their scope to include system strengthening elements. This approach has the potential to combine the best of both—taking advantage of revenue generating opportunities that specific diseases facilitate and channelling a portion of those resources to support the broader health system, still benefiting but not limited to the particular disease. New disaggregated DAH data indicate that the proportion of disease-specific funding that is allocated towards systems strengthening varies across health areas. Figure 4 illustrates three predominant patterns: relatively larger increases over time, a very low and stable proportion increasing after 2003 and minor annual fluctuations with little overall change over the last 25 years. HIV and malaria represent the largest overall proportion and biggest increases in disease-specific funding allocated to system strengthening, from 1–20% and 2–19%, respectively. The patterns for maternal health and child health were nearly identical, ranging from 0.2% to 7% over the last several decades. Across all four of these health areas, 2003 marked the beginning of a period of sustained increases. In contrast, the proportion of funding for TB, NCDs and other infectious diseases allocated to systems strengthening has remained largely consistent over time, particularly during the MDG period. Notably, the patterns prior to 2003 for most health areas reflected those for health sector support overall, peaking in the late 1990s; however, since 2003 the trends have diverged, with declining proportions of overall health sector support and increasing proportions of HIV, malaria and maternal and child health dedicated to systems strengthening. Figure 4. View largeDownload slide Proportion of disease-specific funding allocated to health-systems strengthening. Note: Solid lines represent the proportion of disease-specific funding (as a subset of overall development assistance for health) allocated to health-systems strengthening. The dotted line represents the proportion of overall development assistance for health allocated to health sector support. Figure 4. View largeDownload slide Proportion of disease-specific funding allocated to health-systems strengthening. Note: Solid lines represent the proportion of disease-specific funding (as a subset of overall development assistance for health) allocated to health-systems strengthening. The dotted line represents the proportion of overall development assistance for health allocated to health sector support. To better understand how attempts to integrate system strengthening and disease-specific approaches have worked in practice, several studies have looked in greater depth at these efforts within Global Fund, Gavi and PEPFAR portfolios. Overall, they find that attempts have been less successful in bridging the gap between the two approaches than initially hoped. Global Fund grants that aimed to strengthen human resources were not coordinated with national plans and were invested in disease-focused in-service training, salary top-ups, performance incentives, and contracting part-time workers (Bowser et al. 2014). Gavi applications identified system constraints, but proposed interventions that were overwhelmingly oriented towards operational issues: purchase and delivery of equipment and medication and health workforce training, with little attention to barriers that affect citizen demand, and health worker behaviour and capacity (Goeman et al. 2010). The share of PEPFAR budgets allocated to governance and systems increased from 15% in 2004 to a high of 28% in 2013 but was inversely associated with the total budget (Moucheraud et al. 2016). In additional to health systems funding windows within disease-specific financing institutions, attempts have been made to create new funding platforms. The Health Systems Funding Platform, recommended in 2009 by the Task Force for Innovative International Financing for Health Systems (Hill et al. 2011; Brown et al. 2013) failed to gain traction.10 More recently, the Global Financing Facility attempts to strengthen health systems and increase funding for maternal and child health by mobilizing resources from donors, the World Bank, domestic sources and the private sector, heralded as ‘the future for the post-2015 agenda on diversifying financing for development’ (cited in Usher 2015, p. 1809). However, it remains to be see if this new facility is able to marshal the billions its aims to raise (Usher 2015; Horton 2016). More discouraging regarding the potential for synergies between disease-specific and sector-wide approaches is evidence of displacement effects between HIV funding and health sector funding and services. For countries with a high HIV burden, Lordan and colleagues (2011) find on average, a 1% increase in DAH for HIV was associated with a 0.09% decrease in funding to the health sector.11 Increases in HIV funding in sub-Saharan Africa were associated with reduced provision of childhood immunizations, have had no significant effect on maternal health services, but did have limited positive effects on prenatal blood testing (Grépin 2012). Discussion Taken together, these trends present a clear message. While the total volume and number of countries receiving health sector support has increased over time, it constitutes a small proportion of DAH—a far cry from the three-quarters deemed necessary to achieve the health SDG. Compared to sector support, earmarked funding has been much more substantive and allocated in more predictable ways. Across health areas, those included in the MDGs continue to benefit disproportionately. There is some evidence of displacement effects, indicating that even with a growing resource pie, funding one health area can have detrimental implications for others. Despite concerns regarding the potential misuse of budget support, its allocation has not been consistently associated with country governance indicators. Donors have implemented budget support differently and have reacted in contradictory ways to the same instance of corruption. The inconsistent allocation and application could reflect several factors. By nature, large reconstruction and trust funds established for fragile and conflict-affected states are allocated to countries with weak institutions. Donors may use policy dialogues associated with sector support for political reasons. Indeed, political economy factors may help to explain the behaviour of a particular donor in a particular country; however, these patterns appear to be widespread and not driven by a few outliers. Attempts to bridge the gap between earmarked funding and sector support have varied by disease and have been less successful in practice than intended. Using disease-specific funding mechanisms to strengthen health systems has not been used to address structural constraints. Even if the orientation shifts more in this direction, the level of this sub-stream of funding is likely to be insufficient to do so. The fundamental question remains: how, operationally, will the world advance the ambitious agenda it has set for itself? In terms of resource allocation, the improved alignment of DAH and DALYs for some diseases is an encouraging trend that should be reinforced, and corrected for NCDs. Like all global rankings, the Worldwide Governance Indicators are an imperfect measure; nevertheless, governance indicators appear to be underutilized and the inconsistent allocation of sector support suggests there remains great potential to considerably scale up the proportion of DAH provided in this way. At a minimum, donors could increase the use of this modality for countries with strong governance ratings, which would not pose a greater fiduciary risk than they are already taking with countries with weaker institutions. A synthesis report advises that health sector budget support be allocated on a large enough scale to expand public services to meet needs (Williamson and Dom 2010), threshold levels that do not appear to have been met. If universal health coverage, based on a primary care approach, is intended to be the cornerstone of the SDG health goal (Frenk 2009; World Health Organization 2015; Ottersen et al. 2017), sector support is a more direct way of advancing this goal. In terms of resource mobilization, it is not feasible to establish issue-specific funding mechanisms for each of the 169 targets. Some issue-specific financing will be part of the SDG response, but will not be able to cover these goals in the same way as disease-specific financing did for the MDGs. Moreover, many of the targets that are part of the health SDG are systems issues—access to quality services, workforce issues, early warning, risk reduction and management of health risks—and so would not fit well with a disease-specific approach. New platforms have not been implemented as anticipated, suggesting that the appetite for new initiatives may be waning. HIV may retain a larger relative share of development resources; however, the high profile of particular areas will continue to be threatened by an ever-expanding agenda space, both within and beyond the health sector. Undoubtedly, the way forward is more multi-dimensional than in the past. The SDGs offer the opportunity to address structural drivers and social determinants of health (Taylor et al. 2015; Horton 2016). A ‘health in all’ approach (Hill et al. 2014; Buse and Hawkes 2015; World Health Organization 2015) could improve health outcomes through efforts to address other issues, many of which are underlying contributors to poor health. The SDGs offer a leadership opportunity for intergovernmental organizations with a broad mandate, like the WHO,12 to whom states and key actors turn for guidance and norms when they are unable to resolve issues themselves. Attention is shifting from global to national, and in many cases subnational levels (World Health Organization 2015; Schwartländer et al. 2011). More recent discussions of development financing place a much stronger emphasis on domestic resources (UN 2015; Institute for Health Metrics and Evaluation 2017; Ottersen et al. 2017), although low income countries face an estimated $17–35 billion annual financing gap and will continue to rely on international funds (Stenberg et al. 2017). Both in terms of financing and service provision, there may be greater opportunities for sector-wide and cross-sector integration at subnational and national levels. Buffardi (2014), for instance, identified targeted aspects of sector-wide programs and system strengthening elements of targeted programs, observing that the tension between the two was less marked at the national level. More research is needed on integration strategies in different contexts to understand how countries and donors are adapting. Conclusion The last quarter century, in which health featured prominently on the global agenda, has revealed clear funding patterns, overwhelmingly in favour of specific diseases rather than health systems. Proposals to finance public goods, including health, and to shape global institutions must account for these trends. This mismatch between the many, diverse and interconnected global goals and the way in which programs are financed threatens the achievement of the SDGs. National experimentation may offer alternatives, both within and beyond the health sector, that warrant closer attention and systematic investigation. Acknowledgements This work was initiated when the author was a doctoral candidate at the Evans School of Public Policy and Governance at the University of Washington. The author would like to thank Mary Kay Gugerty, Sanjeev Khagram, Peter May and Joe Zunt for providing valuable comments on an early version of this article, and to Katie Banks, Łukasz Marć, Anne Roemer-Mahler, participants at the 2017 International Studies Association conference in Baltimore and two anonymous reviewers for their insightful contributions to the current version. Funding Partial support for this research came from a Eunice Kennedy Shriver National Institute of Child Health and Human Development research infrastructure grant, P2C HD042828, to the Center for Studies in Demography & Ecology at the University of Washington. Footnotes 1 The 2005 Paris Declaration on Aid Effectiveness and the 2008 Accra Agenda for Action, 2011 Busan Partnership for Effective Development Cooperation, 2014 First High-Level Meeting of the Global Partnership for Effective Development Cooperation: Building Towards an Inclusive Post-2015 Development Agenda, 2015 Addis Ababa Action Agenda of the Third International Conference on Financing for Development, 2016 Global Partnership for Effective Development Cooperation: Nairobi Outcome Document. 2 Some donors may also have existing commitments to particular issues or health conditions. 3 Sources cover 23 bilateral donors (Australia, Austria, Belgium, Canada, Switzerland, Germany, Denmark, Spain, Finland, France, Greece, Ireland, Italy, Japan, Korea, Luxembourg, The Netherlands, Norway, New Zealand, Portugal, Sweden, UK and USA), non-OECD DAC countries, the Bill and Melinda Gates Foundation (BMGF), corporate donors and debt repayments. Channels include the same bilateral donors and BMGF, as well as the European Commission, Gavi, the Global Fund, five multilateral development banks (World Bank International Bank for Reconstruction and Development, World Bank International Development Assistance, InterAmerican Development Bank, African Development Bank, Asian Development Bank), five UN agencies (PanAmerican Health Organisation, UNAIDS, UNFPA, UNICEF, WHO), US non-governmental organization (NGOs), international NGOs and US foundations. The IHME data set uses financial data from the Organization for Economic Cooperation and Development Creditor Reporting System as a base, with corrections for underreporting and double counting. A detailed explanation of the methods used to construct the data set is available as an online supplementary appendix to: Dieleman et al. (2016). 4 Government effectiveness captures perceived quality of public services, policy formulation and implementation, and independence of civil service; control of corruption, the extent to which public power is used for private gain; voice and accountability, freedom of expression and association, free media and democratic elections; political stability and absence of violence; regulatory quality, ability of the government to formulate and implement policies promoting private sector development; and rule of law, quality of contract enforcement, property rights, policy and the courts. These indicators were available biannually from 1996 to 2002 and annually thereafter. 5 Botswana, Namibia, South Africa, Mexico and the Dominican Republic received more than five times the total DAH predicted based on DALYs and GDP. Iran, Chile, Venezuela, Algeria, Malaysia, North Korea and the Central African Republic received less than 20%. 6 As mentioned in the methods section, the data set distinguishes between source (bilateral donors, BMGF, corporate donors and debt repayments) and channel (bilateral donors, multilaterals organizations, UN agencies, US foundations and NGOs). It is also important to note that sector support may vary more on an annual basis, compared to project assistance, as a result of large tranche payments in a particular year. 7 Faust et al. (2012) give the example of Zambia, where Sweden and the Netherlands suspended budget support distributions in response to fraud in the health sector in 2009. The EC released funds, interpreting the detection and prosecution as strengthened government regulation and oversight. 8 Correlations are based on DAH and country governance rankings since 1996. The number of observations varied slightly by governance indicator: political stability n = 2344, regulatory quality and government effectiveness n = 2362, control of corruption n = 2368, voice and accountability n = 2396, rule of law n = 2397. 9 The 10 top, median and bottom ranked countries in control of corruption and government effectiveness who received development assistance for health were identified by country rankings in 1998, the year when health sector support was at its peak, and countries’ median rankings from 1996 to 2014, the period for which Worldwide Governance Indicators and country-level DAH are available. 10 Had the Health Systems Funding Platform been instituted, it may still have been pulled towards shorter-term, discrete, operational issues rather than longer-term, more systematic concerns, similar to the experiences of the Global Fund, Gavi and PEPFAR described above. 11 The displacement effect was much more pronounced for malaria, and not found for TB or an aggregate grouping of all other health areas. These regression analyses are also based on IHME data. 12 The WHO (2015) has asserted that a pragmatic middle ground is possible. At the same time, they acknowledge that previous trends do not suggest that necessary shifts will take place and conclude that ‘the jury is still out on how the SDGs will influence financing for health’ (p. 198). References Allen L. 2017. Non-communicable disease funding. The Lancet Diabetes & Endocrinology  5: 92. Google Scholar CrossRef Search ADS PubMed  Allen L. 2016. Why is there no funding for non-communicable diseases? Journal of Global Health Perspectives . https://ora.ox.ac.uk/objects/uuid:3fe582f2-f4c7-46ad-9e4d-26ac33add7bb. Ahsan KZ, Streatfield PK, Ijdi RE et al.   2016. Fifteen years of sector-wide approach (SWAp) in Bangladesh health sector: an assessment of progress. Health Policy and Planning  31: 612– 23. Google Scholar CrossRef Search ADS PubMed  Beaglehole R, Bonita R, Horton R et al.   2011. Priority actions for the non-communicable disease crs. The Lancet  377: 1438– 47. Google Scholar CrossRef Search ADS   Behague DP, Storeng KT. 2008. Collapsing the vertical-horizontal divide: an ethnographic study of evidence-based policymaking in maternal health. American Journal of Public Health  98: 644– 9. Google Scholar CrossRef Search ADS PubMed  Bhutta ZA, Cabral S, Chan CW, Keenan WJ. 2012. Reducing maternal, newborn, and infant mortality globally: an integrated action agenda. International Journal of Gynecology & Obstetrics  119: S13– 7. Google Scholar CrossRef Search ADS   Biesma RG, Brugha R, Harmer A et al.   2009. The effects of global health initiatives on country health systems: a review of the evidence from HIV/AIDS control. Health Policy and Planning  24: 239– 52. Google Scholar CrossRef Search ADS PubMed  Bossert TJ. 2012. Health systems. Health Policy and Planning  27: 8– 10. Google Scholar CrossRef Search ADS PubMed  Bowser D, Sparkes SP, Mitchell A et al.   2014. Global fund investments in human resources for health: innovation and missed opportunities for health systems strengthening. Health Policy and Planning  29: 986– 97. Google Scholar CrossRef Search ADS PubMed  Brown SS, Sen K, Decoster K. 2013. The health systems funding platform and World Bank legacy: the gap between rhetoric and reality. Globalization and Health  9: 9. Google Scholar CrossRef Search ADS PubMed  Buffardi AL. 2014. From polarisation to practice: puzzles and insights on integrated approaches from public health professionals. Global Public Health  9: 741– 51. Google Scholar CrossRef Search ADS PubMed  Buse K, Hawkes S. 2015. Health in the sustainable development goals: ready for a paradigm shift? Globalization and Health  11: 13. Google Scholar CrossRef Search ADS PubMed  Clinton C, Sridhar D. 2017. Who pays for cooperation in global health? A comparative analysis of WHO, the World Bank, the Global Fund to Fight HIV/AIDS, Tuberculosis and Malaria, and Gavi, the Vaccine Alliance. The Lancet 390: 324– 32. Dieleman JL, Graves CM, Templin T et al.   2014. Global health development assistance remained steady in 2013 but did not align with recipients’ disease burden. Health Affairs  33: 878– 86. Google Scholar CrossRef Search ADS PubMed  Dieleman JL, Schneider MT, Haakenstad A et al.   2016. Development assistance for health: past trends, associations, and the future of international financial flows for health. The Lancet  387: 2536– 44. Google Scholar CrossRef Search ADS   Deverajan S, Swaroop V. 2000. The Implications of Foreign Aid Fungibility for Development Assistance . Washington, DC: World Bank. Google Scholar CrossRef Search ADS   Eichenauer VZ, Reinsberg B. 2017. What determines earmarked funding to international development organizations? Evidence from the new multi-bi aid data. Review of International Organization  12: 171. Google Scholar CrossRef Search ADS   Eichler R, Glassman A. 2008. Health Systems Strengthening via Performance-based Aid: Creating Incentives to Perform and to Measure Results . Brookings Global Economy and Development Working Paper Series. Washington, DC: Brookings Institution. Epstein H. 2009. Towards a global fund for the health MDGs? The Lancet  373: 2110– 1. Google Scholar CrossRef Search ADS   Faust J, Koch S, Molenaers N, Tavakoli H, Vanheukelom J. 2012. The Future of EU Budget Support: Political Conditions, Differentiation and Coordination . London: European Think Tanks Group. Fernandes Antunes A, Xu K, James CD et al.   2013. General budget support: has it benefited the health sector? Health Economics  22: 1440– 51. Google Scholar CrossRef Search ADS PubMed  Foster M, Brown A, Conway T. 2000. Sector-wide Approaches for Health Development: A Review of Experience . Geneva, Switzerland: World Health Organization. Frenk J. 2009. Reinventing primary health care: the need for systems integration. The Lancet  374: 170– 3. Google Scholar CrossRef Search ADS   Grépin KA. 2012. HIV donor funding has both boosted and curbed the delivery of different non-HIV health services in sub-Saharan Africa. Health Affairs  31: 1406– 14. Google Scholar CrossRef Search ADS PubMed  Goeman L, Galichet B, Porignon DG et al.   2010. The response to flexibility: country intervention choices in the first four rounds of the Gavi Health Systems Strengthening applications. Health Policy and Planning  25: 292– 9. Google Scholar CrossRef Search ADS PubMed  Hafner T, Shiffman J. 2013. The emergence of global attention to health systems strengthening. Health Policy and Planning  28: 41– 50. Google Scholar CrossRef Search ADS PubMed  Handley G. 2009. Sector Budget Support in Practice: Literature Review . London: Overseas Development Institute. Hayman R. 2011. Budget support and democracy: a twist in the conditionality tale. Third World Quarterly  32: 673– 88. Google Scholar CrossRef Search ADS   Hill PS, Vermeiren P, Miti K, Ooms G, Van Damme W. 2011. The health systems funding platform: is this where we thought we were going? Globalization and Health  7: 16. Google Scholar CrossRef Search ADS PubMed  Hill PS, Buse K, Brolan CE, Ooms G. 2014. How can health remain central post-2015 in a sustainable development paradigm? Globalization and Health  10: 18. Google Scholar CrossRef Search ADS PubMed  Horton R. 2016. Offline: global health—tipping into irrelevance. The Lancet  388: 1362. Google Scholar CrossRef Search ADS   Institute for Health Metrics and Evaluation. 2016. Financing Global Health 2015: Development Assistance Steady on the Path to New Global Goals . Seattle, WA: Author. Institute for Health Metrics and Evaluation. 2017. Financing Global Health 2016: Development Assistance, Public and Private Health Spending for the Pursuit of Universal Health Coverage . Seattle, WA: Author. Kaufmann D, Kraay A, Mastruzzi M. 2010. The Worldwide Governance Indicators: A Summary of Methodology, Data and Analytical Issues . World Bank Policy Research Working Paper No. 5430. Washington, DC: The World Bank. Google Scholar CrossRef Search ADS   Koeberle S, Stavreski Z. 2006. Budget support: concept and issues. In: Koeberle S, Stavreski Z, Walliser J (eds). Budget support as more effective aid? Recent experiences and emerging lessons . Washington, DC: The World Bank, 3– 28. Google Scholar CrossRef Search ADS   Langan M. 2015. Budget support and Africa–European Union relations: free market reform and neo-colonialism? European Journal of International Relations  21: 101– 21. Google Scholar CrossRef Search ADS   Lawson A, Booth D, Harding A, Hoole D, Naschold F. 2003. General Budget Support Evaluability Study Phase 1: Synthesis Report, Vol. I. London: DFID Evaluation Report EV 643. Lawson A, Booth D, Msuya M, Wangwe, Williamson T. 2006. Does general budget support work? Evidence from Tanzania. In: Koeberle S, Stavreski Z, Walliser J (eds). Budget Support as More Effective Aid? Recent Experiences and Emerging Lessons . Washington, DC: The World Bank, 121– 36. Levine R. 2008. Healthy foreign policy: bringing coherence to the global health agenda. In: Birdsall (ed). The White House and the World: A Global Development Agenda for the Next US President . Washington, DC: Center for Global Development, 43– 62. Lordan G, Tang KK, Carmignani F. 2011. Has HIV/AIDS displaced other health funding priorities? Evidence from a new dataset of development aid for health. Social Science & Medicine  73: 351– 5. Google Scholar CrossRef Search ADS   Lu C, Schneider MT, Gubbins P et al.   2010. Public financing of health in developing countries: a cross-national systematic analysis. The Lancet  375: 1375– 87. Google Scholar CrossRef Search ADS   Mills A. 2005. Mass campaigns versus general health services: what have we learnt in 40 years about vertical versus horizontal approaches? Bulletin of the World Health Organization  83: 315– 6. Google Scholar PubMed  Molenaers N. 2012. The great divide? Donor perceptions of budget support, eligibility and policy dialogue. Third World Quarterly  33: 791– 806. Google Scholar CrossRef Search ADS   Molenaers N, Gagiano A, Smets L, Dellepiane S. 2015. What determines the suspension of budget support? World Development  75: 62– 73. Google Scholar CrossRef Search ADS   Moucheraud C, Sparkes S, Nakamura Y et al.   2016. PEPFAR investmetns in governance and health systems were one-fifth of countries’ budgeted funds, 2004–14. Health Affairs  35: 847– 55. Google Scholar CrossRef Search ADS PubMed  Nugent R. 2015. Bilateral and Multilateral Financing for Noncommunicable Diseases. Policy Brief Prepared for the WHO Global Coordination Mechanism on the Prevention and Control of Noncommunicable Diseases, Working Group on How to Realise Governments' Commitment to Provide Financing for NCDs. Geneva: WHO. http://www.who.int/nmh/ncd-coordination-mechanism/Policybrief5.2docx.pdf, accessed 14 September 2017. Ooms G, Van Damme W, Baker BK, Zeitz P, Schrecker T. 2008. The ′diagonal′ approach to Global Fund financing: a cure for the broader malaise of health systems? Global Health  4: 6. Google Scholar CrossRef Search ADS PubMed  Ottersen T, Evans DB, Mossialos E, Røttingen JA. 2017. Global health financing towards 2030 and beyond. Health Economics, Policy and Law  12: 105– 11. Google Scholar CrossRef Search ADS   Peters DH, Paina L, Schleimann F. 2013. Sector-wide approaches (SWAps) in health: what have we learned? Health Policy and Planning  28: 884– 90. Google Scholar CrossRef Search ADS PubMed  Ravishankar N, Gubbins P, Cooley RJ et al.   2009. Financing of global health: tracking development assistance for health from 1990 to 2007. The Lancet  373: 2113– 24. Google Scholar CrossRef Search ADS   Schwartländer B, Stover J, Hallett T et al.   2011. Towards an investment approach for an effective response to HIV/AIDS. The Lancet  377: 2031– 41. Google Scholar CrossRef Search ADS   Shiffman J. 2008. Has donor prioritization of HIV/AIDS displaced aid for other health issues? Health Policy and Planning  23: 95– 100. Google Scholar CrossRef Search ADS PubMed  Shiffman J, Berlan D, Hafner T. 2009. Has aid for AIDS raised all health funding boats? Journal of Acquired Immune Deficiency Syndromes  52: S45– 8. Google Scholar CrossRef Search ADS PubMed  Smith SL, Rodriguez MA. 2016. Agenda setting for maternal survival: the power of global health networks and norms. Health Policy and Planning  31: i48– 59. Google Scholar CrossRef Search ADS PubMed  Stenberg K, Hanssen O, Edejer TTT et al.   2017. Financing transformative health systems towards achievement of the health Sustainable Development Goals: a model for projected resource needs in 67 low-income and middle-income countries. The Lancet Global Health  5: e875– 87. Google Scholar CrossRef Search ADS PubMed  Tavakoli H, Smith G. 2013. Back under the microscope: insights from evidence on budget support. Development Policy Review  31: 59– 74. Google Scholar CrossRef Search ADS   Taylor S, Williams B, Magnus D, Goenka A, Modi N. 2015. From MDG to SDG: good news for global child health? The Lancet  386: 1213. Google Scholar CrossRef Search ADS   UN Department of Economic and Social Affairs Technical Support Team in support of the General Assembly Open Working Group on Sustainable Development Group. October 2014. Compendium of TST Issue Briefs: Health and Sustainable Development. https://sustainabledevelopment.un.org/content/documents/1554TST_compendium_issues_briefs_rev1610.pdf, accessed 16 February 2017. United Nations. 2015. Addis Ababa Action Agenda of the Third International Conference on Financing for Development. http://www.un.org/esa/ffd/wp-content/uploads/2015/08/AAAA_Outcome.pdf, accessed 14 September 2017. Uplekar M, Raviglione MC. 2007. The “vertical-horizontal” debates: time for the pendulum to rest (in peace)? Bulletin of the World Health Organization  85: 413– 4. Google Scholar CrossRef Search ADS PubMed  USAID Development Information Services. 2006. General budget support: key findings of five USAID studies. In: Koeberle S, Stavreski Z, Walliser J (eds). Budget Support as More Effective Aid? Recent Experiences and Emerging Lessons . Washington, DC: The World Bank, 105– 16. Google Scholar CrossRef Search ADS   Usher AD. 2015. Global Financing Facility: where will the funds come from? The Lancet 386: 1809. Woll B. 2008. Donor harmonisation and government ownership: multi-donor budget support in Ghana. The European Journal of Development Research  20: 74– 87. Google Scholar CrossRef Search ADS   Williamson T, Dom C. 2010. Sector Budget Support in Practice: Synthesis Report . London: Overseas Development Institute. World Bank. 1998. Assessing Aid: What Works, What Doesn’t, and Why . New York, NY: Oxford University Press. World Health Organization. 2015. Health in 2015: From MDGs, Millennium Development Goals to SDGs. Sustainable Development Goals . Geneva: Author. World Health Organization Maximizing Positive Synergies Collaborative Group. 2009. An assessment of interactions between global health initiatives and country health systems. The Lancet  373: 2137– 69. CrossRef Search ADS   © The Author(s) 2018. Published by Oxford University Press in association with The London School of Hygiene and Tropical Medicine. All rights reserved. For permissions, please e-mail: journals.permissions@oup.com http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Health Policy and Planning Oxford University Press

Sector-wide or disease-specific? Implications of trends in development assistance for health for the SDG era

Loading next page...
 
/lp/ou_press/sector-wide-or-disease-specific-implications-of-trends-in-development-y1XtI6wlzr
Publisher
Oxford University Press
ISSN
0268-1080
eISSN
1460-2237
D.O.I.
10.1093/heapol/czx181
Publisher site
See Article on Publisher Site

Abstract

Abstract The record of the Millennium Development Goals broadly reflects the trade-offs of disease-specific financing: substantial progress in particular areas, facilitated by time-bound targets that are easy to measure and communicate, which shifted attention and resources away from other areas, masked inequalities and exacerbated fragmentation. In many ways, the Sustainable Development Goals reflect a profound shift towards a more holistic, system-wide approach. To inform responses to this shift, this article builds upon existing work on aggregate trends in donor financing, bringing together what have largely been disparate analyses of sector-wide and disease-specific financing approaches. Looking across the last 26 years, the article examines how international donors have allocated development assistance for health (DAH) between these two approaches and how attempts to bridge them have fared in practice. Since 1990, DAH has overwhelmingly favoured disease-specific earmarks over health sector support, with the latter peaking in 1998. Attempts to integrate system strengthening elements into disease-specific funding mechanisms have varied by disease, and more integrated funding platforms have failed to gain traction. Health sector support largely remains an unfulfilled promise: proportionately low amounts (albeit absolute increases) which have been inconsistently allocated, and the overall approach inconsistently applied in practice. Thus, the expansive orientation of the Sustainable Development Goals runs counter to trends over the last several decades. Financing proposals and efforts to adapt global health institutions must acknowledge and account for the persistent challenges in the financing and implementation of integrated, cross-sector policies. National and subnational experimentation may offer alternatives within and beyond the health sector. Sector-wide approaches, agenda setting, aid, Millennium Development Goals, donors, HIV, tuberculosis, non-communicable diseases, maternal and child health, health financing Key Messages Over the last quarter century, development assistance for health has disproportionately favoured specific diseases over sector support, with the latter peaking in 1998. More countries have received health sector support over time, but there is a weak relationship between country governance indicators and the proportion of assistance channelled in this way. Attempts to bridge disease-specific and sector-wide financing approaches have been limited, and new financing platforms have failed to gain traction. Recent estimates of the additional investment needed to achieve the health Sustainable Development Goal indicate that three-quarters of these resources need to be allocated to health system strengthening; yet, trends in financing persistently remain oriented otherwise. Introduction The field of global health is in transition from its heyday in the Millennium Development Goal (MDG) era, which saw exponential increases in attention and financing and the creation of new global institutions and initiatives, to plateauing funding within a much more crowded agenda space. These resources were channelled through two divergent funding strategies, reflecting distinct perspectives. Health sector and budget support aimed to better align development assistance with recipient country health priorities and financial systems. Earmarked resources targeted specific diseases, population groups or interventions. The trade-offs of disease-specific financing characterises the record of the MDGs more broadly: substantial progress in particular areas, facilitated by time-bound targets that are easy to measure and communicate, which shifted attention and resources away from other areas, masked inequalities within and between countries and exacerbated fragmentation, particularly at the national level (World Health Organization 2015). Scholars have noted swings in the orientation of global health over the decades between targeted and broader, systemic approaches (Mills 2005; Uplekar and Raviglione 2007; Bossert 2012; Hafner and Shiffman 2013). In many ways, the final set of Sustainable Development Goals (SDGs) reflect a profound shift of the pendulum back towards a more holistic, system-wide approach. Recent estimates of the additional investment needed to achieve the SDG health goal indicate that three-quarters of these resources need to be allocated to health system strengthening (Stenberg et al. 2017). At the dawn of the SDG era, the World Health Organisation (WHO) and others called for a deeper debate about the institutional architecture and financing needed to produce global public goods and to respond to country needs (Hill et al. 2014; UN Department of Economic and Social Affairs Technical Support Team in support of the General Assembly Open Working Group on Sustainable Development Group 2014; World Health Organization 2015). The Global Health Security Working Group on Health Financing has asserted: ‘it is clear that today’s global and national arrangements for health financing need to change’, noting high disability and premature mortality rates, high out of pocket spending and inadequate domestic spending for essential health services (Ottersen et al. 2017, p. 105). To advance these efforts, this article builds upon existing work on aggregate trends in donor financing, bringing together what have largely been disparate analyses of sector-wide and disease-specific financing approaches. Drawing on both primary and secondary, quantitative and qualitative data, it looks back across the last 26 years and asks: how have international donors allocated development assistance for health (DAH) between the health sector and specific diseases; to what extent does country-level governance and disease burden align with these allocations; and how have attempts to bridge these two approaches fared across funding platforms and diseases? Given these patterns, the article then discusses implications for the field moving forward and the ambitious agenda the global community has set for itself. These analyses present a clear picture: since 1990, DAH has overwhelmingly favoured the specific over the sector, with substantially more funding earmarked for individual diseases than channelled as health sector support. Most of the new global health institutions created in the new millennium followed this model. Attempts to integrate system strengthening elements into disease-specific funding mechanisms have been less successful in bridging the two approaches and more integrated funding platforms have failed to gain traction. Despite Paris Declaration commitments to align with national institutions and priorities, sector support largely remains an unfulfilled promise: proportionately low amounts which have been inconsistently allocated, and the overall approach inconsistently applied in practice. Thus, the expansive orientation and projected resource needs of the SDGs runs counter to the trends over the last several decades. Financing proposals and efforts to adapt global health institutions must acknowledge and account for the persistent challenges in the financing and implementation of integrated, cross-sector policies. Rationale for sector support and disease-specific approaches Scholars have debated for some time the relative merits of sector and disease-specific approaches, often characterized as horizontal/system strengthening versus vertical/targeted approaches; and, diagonal approaches, the elusive best of both worlds. Therefore, each is just briefly reviewed here. Although health sector support and earmarked funding for specific diseases or activities are not mutually exclusive, they represent distinct strategies, associated with particular perspectives on how best to improve development outcomes, with advantages and drawbacks that mirror one another. Sector-wide and budget support financing, which channels funding directly through national accounts, grew out of frustration with traditional project-based aid (Lawson et al. 2003; Koeberle and Stavreski 2006). It aims to reduce the existence of parallel delivery structures and increase the likelihood that programs will be sustained because they are integrated into government systems. Sector support is intended to enhance country ownership, enabling national governments to better target resources to their unique health profiles and address disease outbreaks and emergent health concerns. It can help develop workforce and infrastructure capacity that serves to support multiple disease control and prevention programs (Foster et al. 2000; Koeberle and Stavreski 2006; Eichler and Glassman 2008; Ooms et al. 2008; Peters et al. 2013). Some scholars argue that budget support is not a new financing mechanism; rather, that it represents the fourth generation of donor conditionality policies that begin in earnest in the 1980s with structural adjustment programs, and in the Paris Declaration era were framed in terms of country ownership (Hayman 2011; Langan 2015). Hafner and Shiffman (2013) document increased attention to health systems strengthening by multilateral and global health institutions from 2005 onwards, reflecting pressure to achieve the MDGs and concerns over health workforce shortages and unintended adverse effects of global health initiatives on national health systems. International donors have jointly signalled their commitment to align their assistance with recipient country priorities and financial and procurement systems as part of the 2005 Paris Declaration and subsequent high level fora on development cooperation, including the 2015 Addis Ababa Action Agenda.1 The flexible nature of sector support—its key advantage—makes it much more difficult to directly link financial investments to changes in specific outcomes. Moreover, the fungible nature of aid, the freeing up of government resources that can be used elsewhere, may have the perverse effect of reducing national expenditures on health (World Bank 1998; Deverajan and Swaroop 2000; Lu et al. 2010). Since funding is channelled directly through national accounts, donors are reliant on the government’s ability to manage finances and track funds in a transparent matter to reduce the likelihood that funds will be used for unauthorised expenditures. These capacities include managing budget processes which are governed by clear rules, with allocations linked to specific policies and expenditures linked to allocations, administering transparent procurement systems, ensuring timely and accurate financial reporting to track expenditures, and overseeing regular internal and external audits. Indeed, early evaluations identified low fiduciary risk (the extent to which actual expenditures differ from authorized expenditures) and government capacity to manage public finances as two key factors in effective budget support (USAID Development Information Services 2006; Lawson et al. 2006). In contrast, targeted funding is directed towards a specific disease (HIV), population group (children) or intervention (vaccinations). It can provide clearer links between financial inputs and health outputs and outcomes, facilitating results-based measurement. Targeted attention towards one disease can make it easier to garner initial public and political support, and with more visible results, easier to sustain support.2 The establishment of Gavi in 2000, the Global Fund to Fight AIDS, Tuberculosis and Malaria in 2002, and the President’s Emergency Plan for AIDS Relief (PEPFAR) in 2003 created a set of new institutions that embodied this targeted approach. At the same time, targeted funding for certain areas has also generated criticism. Earmarked funding has been critiqued for creating disease silos and siphoning resources away from lower profile issues (Behague and Storeng 2008; Levine 2008; Shiffman 2008; Epstein 2009; Shiffman, Berlan and Hafner 2009). Narrowly targeted funding can neglect health systems on which disease-specific interventions depend for successful implementation. Biesma et al. (2009) found that while the Global Fund, PEPFAR and World Bank Multi-country AIDS Program enabled swift scale-up and expanded stakeholder participation, they created parallel coordination structures, financial management, monitoring and reporting systems, and were not well aligned with overall national health priorities or harmonized with other donor efforts. These global health initiatives have increased the attrition rate of healthcare workers from the public sector to non-state providers, and efforts to scale-up responses to specific diseases has increased the burden on already overstretched health workforces (World Health Organization Maximizing Positive Synergies Collaborative Group 2009). While largely distinct approaches, particularly in their respective rationales, sector support and earmarked financing are not absolutely exclusive. Gavi, the Global Fund and PEPFAR have broadened their scope over time, allowing a portion to be used for health system strengthening (Goeman et al. 2010; Bowser et al. 2014; Moucheraud et al. 2016), discussed further below. Moreover, development assistance is not necessarily a zero-sum game among health conditions, health systems or across development issues. Indeed, until 2011 when DAH began to plateau, the 2000s witnessed a growing resource pie for the health sector overall and for nearly all health areas, including health system strengthening. That said, the 17 SDGs and 169 targets reflect a dramatically larger number of global ‘priorities’, in which health is much less prominent than its previous inclusion in three of the eight MDGs. The breadth in part reflects the mobilization of groups who learned from the HIV response and sought to raise the profile of their issue on the global agenda (Beaglehole et al. 2011; Bhutta et al. 2012; Smith and Rodriguez 2016). The establishment of the Partnership for Maternal, Newborn and Child Health (PMNCH) in 2005 and the non-communicable disease (NCD) Alliance in 2009—among many other new groups within and beyond the health sector—have created what the WHO (2015) characterizes as a competitive institutional landscape. In the next 25 years, DAH is expected to increase at a much more modest rate than in the past: 1.7% compared to 11.3% annually from 2000 to 2010 (Dieleman et al. 2016; Institute for Health Metrics and Evaluation 2017). Achieving the SDGs is expected to cost $1.4 trillion a year, with $274–371 billion annually in new resources needed for the health goal alone (Stenberg et al. 2017). Therefore, while the absolute volume of resources may continue to grow, the pace is slower and the global set of priorities and groups mobilized around them is much broader than in 2000. Methods Over the last decade, the availability and specificity of data on development assistance has improved substantially. This analysis draws on a number of sources: studies on sector and budget support, which are predominantly qualitative, and recent quantitative analyses of DAH, including associations between disease-specific funding and country burden of disease. Previous research on budget support in the international development literature was identified through Google Scholar and supplemented by articles on sector support published in previous issues of this journal and by references in articles on the topic. This article also presents new analyses to identify trends in system strengthening components of disease-specific funding and to examine the allocation of health sector support according to country-level governance indicators. These analyses are based on the Institute for Health Metrics and Evaluation (IHME) Development Assistance for Health data set, which tracks annual financial disbursements for health by source and channel to 166 recipient countries since 1990.3 It distinguishes funding to nine health focus areas: HIV, malaria, TB, maternal health, child and newborn health, NCDs, other infectious diseases, health sector strengthening/sector-wide approaches and other. The ‘focus’ of these nine areas varies considerably, with the latter four categories encompassing a much broader set of conditions and areas than the three named diseases and to a lesser extent, conditions affecting women, children and newborns. The 2017 data set for the first time disaggregates the amount of earmarked funding for a specific area that was dedicated to health systems strengthening. Therefore, it includes the total amount allocated to health sector support and the eight other areas, as well as the amount within each disease category supporting health systems strengthening (i.e. the amount of earmarked HIV funding dedicated to health systems strengthening). To examine the allocation of health sector support, the analyses use country-level World Bank Worldwide Governance Indicators (WGI), which group hundreds of variables from 32 sources, including Freedom House, the International Budget Project and the Afrobarometer. Selected for their comprehensive nature, the WGI cover six domains: government effectiveness, control of corruption, voice and accountability, political stability and absence of violence, regulatory quality, and rule of law.4 (Kaufmann et al. 2010). Quantitative analyses were conducted in SAS 9.4. These governance rankings provide a corollary to disability-adjusted life years (DALYs), in that they could help to inform donor decisions regarding how and where to allocate DAH. Given the importance of government capacity noted earlier, if donors are concerned about reducing fiduciary risk and maximizing effectiveness, the allocation of sector support should be positively associated with stronger governance indicators, particularly government effectiveness and control of corruption. By nature, countries with poorly performing institutions are less able to address the health concerns of their citizenry, who may suffer from a disproportionately higher burden of disease. This higher need may result in higher overall levels of DAH, including health sector support. However, one would hypothesize that donors would allocate a higher proportion of DAH channelled as sector support compared to earmarked aid to countries with stronger government capacity. Although the DAH data set represents a vast improvement over the data that was available just over a decade ago, it is not without its limitations. A large proportion of assistance, 15–44% of annual DAH, fell into the ‘other’ category and an additional 0.19–10% was unallocable. Categories were assigned based on project key words, which may not reflect all uses of the funds. For some countries and years, data were generated by statistical models to estimate disbursements. Moreover, the data set excludes government and private health spending, assistance from Gulf States, BRICS countries, charities outside of the US and the for-profit private sector (Dieleman et al. 2014, 2016; Nugent 2015). Results Trends in disease-specific funding Over the last 26 years, the majority of DAH has been allocated to specific diseases, ranging from 42% to 74% annually, from $3.1 billion in 1990 to $26.8 billion in 2015 (Figure 1). Previous studies have reported increases in both sector-wide and particularly disease-specific funding over time (Shiffman 2008; Shiffman, Berlan and Hafner 2009; Ravishankar et al. 2009; Dieleman et al. 2016), so only a brief overview is included here. Since 2000, DAH has prioritized the MDGs, with HIV, malaria, tuberculosis, maternal, newborn and child health growing substantially faster than other areas (Dieleman et al. 2016). Increases have been the most dramatic for HIV, rising from 4.4% in 1990 to 26% of DAH in 2015 and as high as 31% in 2010. Funding for malaria increased from 0.9% to 6.4% and for TB from 0.4% to 3.5%, still proportionally small amounts but large relative increases over the last two and a half decades. The relative proportion of DAH allocated to maternal and child health dropped from 35% in 1990 to 29% in 2015, but still constitutes a sizable share of the total. The proportion of funding allocated to NCDs has remained relatively constant, at its highest at 2.9% of DAH in 1993 (Institute for Health Metrics and Evaluation 2017). Figure 1. View largeDownload slide Amount of development assistance for health allocated to specific diseases, health sector support and other areas: 1990–2015. Note: ‘Other’ refers to cases with project-level information that does not fall into any of the health focus areas tracked by IHME. Figure 1. View largeDownload slide Amount of development assistance for health allocated to specific diseases, health sector support and other areas: 1990–2015. Note: ‘Other’ refers to cases with project-level information that does not fall into any of the health focus areas tracked by IHME. The overwhelming allocation of earmarked funding is also reflected in WHO finances, which have experienced a steady increase in the proportion of extra-budgetary funding over time, nearly 80%, with donors earmarking 93% of voluntary funding in the 2014–15 budget (Clinton and Sridhar 2017). This trend is not unique to the health sector. More broadly, earmarked funding from bilateral agencies that is channelled through multilateral organizations has increased 4-fold over the last two decades (Eichenauer and Reinsberg 2017). Country allocation of disease-specific funding varies by health area, and overall portrays an encouraging picture, with exceptions. Institute for Health Metrics and Evaluation reports (2016, 2017) indicate that alignment between DAH and DALYs has been the strongest for HIV, TB and child health. Funding for maternal health and malaria has been less well aligned with countries’ relative burden of disease. The relationship between the two has been the worst for NCDs. Allen (2016, 2017) posits that this mismatch could be explained by underestimates of NCD funding, their non-infectious nature, private sector producer interests, a weak evidence base and the lack of a compelling narrative. Country GDP and DALYs significantly predicted the amount of country-level DAH, with high-burden, low-income countries receiving the highest amount across all health areas. However, there are notable outliers where countries received disproportionately more or less funding than would be expected given their burden of disease (Dieleman et al. 2014).5 Across diseases, Figure 2 looks at trends in specific health areas prior to and following key global agenda setting moments. As mentioned earlier, there have been dramatic increases in funding for HIV, TB and malaria, with notable spikes following the MDGs and the creation of the Gates Foundation, Global Fund and PEPFAR—funding levels which have begun to slightly decline in recent years, but have not sharply dropped off over time. This provides a vivid example of the power of global norms and particular institutions in directing attention and resources that has been sustained over time. Figure 2. View largeDownload slide Global agenda setting milestones and trends in health funding over time: amount and proportion of overall development assistance for health allocated to different health areas. Figure 2. View largeDownload slide Global agenda setting milestones and trends in health funding over time: amount and proportion of overall development assistance for health allocated to different health areas. A similar response has not been true for other diseases. The NCD Alliance and subsequent High Level Summit have not shifted funding. Declines in funding for maternal and child health have reversed following the establishment of PMNCH, although the overall proportion of funding for this health area remains lower than in 1990 and 2000. These figures illustrate that absolute and relative changes in DAH has varied across diseases and sector support, and that there has been a staggered institutional response at the global level across diseases. The graphs should not be interpreted as causal relationships between global agenda setting moments and DAH. Taken together, these overall trends indicate that funding allocations strongly favour specific diseases and MDG areas. The alignment between disease burden and resource allocation has varied by disease, but overall reflects a positive association. Trends in health sector support In contrast, the record of health sector support largely remains an unfulfilled promise. Although sector and budget support was strongly promoted by some donors, including the European Commission, World Bank, Nordic countries, the Netherlands and the UK (Hayman 2011; Faust et al. 2012; Langan 2015), the scale of the shift has remained small. Paralleling overall increases in DAH since 1990, absolute funding levels for sector support have increased over time from $543 million in 1990 to $3.4 billion in 2015, peaking at $4.2 billion in 2011. The proportion of overall DAH for health sector support was at its highest level, 18.5%, in 1998 and has remained relatively constant at 9–12% over the last decade. Notably, the pattern of health sector support over time has differed compared to trends in financing for specific diseases, with the highest average annual gains seen in the 1990s at 11.4%, lower from 2000 to 2009 at 7.1% and tapering off to 2.3% since 2010 (Institute for Health Metrics and Evaluation 2017). Thus, unlike stated intentions, sector support has not substantially increased over time relative to other funding modalities, nor has budget support been implemented as an incremental process through which countries transition along a continuum of recipient country control from project aid towards general budget support. Donors have maintained hybrid portfolios which include a range of instruments that vary in the extent of their traceability and earmarking (Handley 2009, Ahsan et al. 2016). While small in terms of the overall relative amount, allocations of health sector support have become more widespread over time. In 1990, half of recipient countries received any health sector funding, rising to 76% 2 years later, to 85% in 1996 and to 94% in 2000. From 2003 onwards, nearly all countries have received a fraction of their DAH as sector support from at least one donor. Across sources of sector support, most donors have increased the absolute amount they are allocating over time, although levels fluctuate, peaking in the USA in 2004 ($941 million), Sweden in 2007 ($110 million), the Netherlands in 2011 ($141 million) and the UK in 2015 ($496 million). Among bilateral donors, South Korea has demonstrated the greatest absolute increases, initially allocating very low amounts of sector support but since 2008 exceeding individual Nordic countries. The amount channelled through multilateral agencies has increased overall but has varied across the time period as well: with $4 million channelled through the European Commission in 1996 and $140 million in 2015, peaking at $222 million in 2013. DAH channelled through the World Bank [International Bank for Reconstruction and Development (IBRD)] increased from $25 million in 1990 to $240 million in 2015, reaching as high as $1.4 billion in 2010. Across all channels of support, the proportion of sector support provided through the WHO has declined steadily from over 60% in 1990 to a low of 14% in 2013.6 The proportion channelled through World Bank has fluctuated from 5% in 1990 to 48% in 1998 to 32% in 2011 to 7% in 2015. Qualitative assessments indicate there is considerable variation among agencies regarding the purpose of budget and sector support, and how it is implemented across different countries. This variation includes differences in selection criteria, the extent to which recipients are required to fulfil specific conditions and how strictly these conditions are monitored, what constitutes a breach and how donors respond, and the relative emphasis between technical and political elements of policy dialogues. These differences send mixed signals to recipient governments; in some cases, donors have responded in opposite ways to the same incident7 (Handley 2009; Faust et al. 2012; Molenaers 2012; Tavakoli and Smith 2013). Examining the relationship between national governance rankings and the proportion of DAH allocated to sector support, correlations are positive but very weak, ranging from r = 0.08 for political stability to r = 0.14 for government effectiveness.8 To analyse trends in allocation among groups of countries with stronger and weaker governance indicators, we examined the median proportion and range of total DAH that has been allocated to sector support across country quintiles. This analysis did not demonstrate marked differences across country groups; Figure 3 illustrates these overall patterns through several examples. Figure 3. View largeDownload slide Proportion of development assistance for health allocated to health sector support across countries with different Worldwide Governance Indicator rankings. (a) Countries ranking at the bottom, median and top of the Worldwide Governance Indicators. This includes two examples of bottom ranked countries: Somalia, which has featured prominently in world politics, and low profile Equatorial Guinea. (b) Countries with the biggest variation in Worldwide Governance Indicator rankings over time. Note: These graphs use 3-year moving averages to account for large tranche payments in a particular year. Figure 3. View largeDownload slide Proportion of development assistance for health allocated to health sector support across countries with different Worldwide Governance Indicator rankings. (a) Countries ranking at the bottom, median and top of the Worldwide Governance Indicators. This includes two examples of bottom ranked countries: Somalia, which has featured prominently in world politics, and low profile Equatorial Guinea. (b) Countries with the biggest variation in Worldwide Governance Indicator rankings over time. Note: These graphs use 3-year moving averages to account for large tranche payments in a particular year. Looking at countries at the top, median and bottom of the governance rankings,9Figure 3a shows some variation across countries and over time, with common peaks in the mid-1990s and early 2000s—but not consistently ordered according to governance rankings or nearly as prominent as would be expected given the difference in institutional contexts. Figure 3b illustrates the proportion of DAH channelled as health sector support for two countries with some of the biggest changes (highest standard deviations) in government effectiveness and control of corruption over time. Again, like the cross-national trends, the proportion of health sector support in these countries was highest in the 1990s. In Rwanda, where governance has steadily improved over time, health sector support has been inversely related to this trend, with lower proportions of sector support as government effectiveness and control of corruption rankings have improved. In Cote d’Ivoire, declining health sector support paralleled declining governance from 1996 to 2010; however, sector support has remained low despite subsequent improvements in more recent years. The weak, inconsistent relationship between governance indicators and allocation of DAH is similar to findings regarding suspension of budget support. Molenaers et al. (2015) found that WGI voice and accountability trends were associated with a reduced probability of suspension; however, trends in government effectiveness, political stability and control of corruption were non-significant. Thus, despite donor concerns about fiduciary risk and government capacity, country governance indicators do not appear to guide allocation decisions of sector support to the extent that DALYs guide some disease-specific funds to countries in greatest need. In addition to the allocation of sector support, previous studies have also attempted to assess the effectiveness of this modality, finding that outcomes are highly context dependent. Overall, budget support has been most promising in improving the pro-poor nature of public expenditure, public financial management and donor harmonization (Woll 2008; Handley 2009; Tavakoli and Smith 2013; Ahsan et al. 2016). At the same time, it is transactionally intensive, and has been less influential in improving civil service functions or reducing corruption (Tavakoli and Smith 2013). For example, an assessment of sector-wide approaches in Bangladesh over the last 15 years found improvements in monitoring, procurement, supply chain and financial management, but challenges in supporting long-term training, rural physician retention and decentralizing functions to subnational levels (Ahsan et al. 2016). In terms of effects on government health spending, Fernandes Antunes et al. (2013) find that general budget support showed no effect, based on a panel data set of 82 low and middle-income countries. External funding for health was associated with declines in health expenditures from domestic sources, but it was not a full substitution effect, in that it was associated with increases in total government spending on health from all sources. Despite stated intentions to improve country ownership, budget support policy dialogues can still be dominated by donors, in some cases more powerfully than before when they act as a group rather than independently (Woll 2008, Handley 2009). Donors may use policy dialogues to increase their leverage for economic and trade liberationsation (Langan 2015). And, they have demonstrated that they are willing to suspend budget support, most often related to fraud, repression or human rights violations (Molenaers et al. 2015), thus reducing the consistency of this aid modality, a key rationale for its use (Hayman 2011; Tavakoli and Smith 2013). In sum, these trends in health sector financing indicate rather tepid support for this modality: amounts representing a relatively small proportion of overall DAH that have declined during the MDG period, with inconsistent allocation and application in practice, likely a result of and contributing to mixed results of its effectiveness. Efforts to bridge the vertical–horizontal divide As noted earlier, while the rationale for disease-specific and sector-wide funding approaches are distinct, they are not mutually exclusive. In response to critiques, institutions have expanded their scope to include system strengthening elements. This approach has the potential to combine the best of both—taking advantage of revenue generating opportunities that specific diseases facilitate and channelling a portion of those resources to support the broader health system, still benefiting but not limited to the particular disease. New disaggregated DAH data indicate that the proportion of disease-specific funding that is allocated towards systems strengthening varies across health areas. Figure 4 illustrates three predominant patterns: relatively larger increases over time, a very low and stable proportion increasing after 2003 and minor annual fluctuations with little overall change over the last 25 years. HIV and malaria represent the largest overall proportion and biggest increases in disease-specific funding allocated to system strengthening, from 1–20% and 2–19%, respectively. The patterns for maternal health and child health were nearly identical, ranging from 0.2% to 7% over the last several decades. Across all four of these health areas, 2003 marked the beginning of a period of sustained increases. In contrast, the proportion of funding for TB, NCDs and other infectious diseases allocated to systems strengthening has remained largely consistent over time, particularly during the MDG period. Notably, the patterns prior to 2003 for most health areas reflected those for health sector support overall, peaking in the late 1990s; however, since 2003 the trends have diverged, with declining proportions of overall health sector support and increasing proportions of HIV, malaria and maternal and child health dedicated to systems strengthening. Figure 4. View largeDownload slide Proportion of disease-specific funding allocated to health-systems strengthening. Note: Solid lines represent the proportion of disease-specific funding (as a subset of overall development assistance for health) allocated to health-systems strengthening. The dotted line represents the proportion of overall development assistance for health allocated to health sector support. Figure 4. View largeDownload slide Proportion of disease-specific funding allocated to health-systems strengthening. Note: Solid lines represent the proportion of disease-specific funding (as a subset of overall development assistance for health) allocated to health-systems strengthening. The dotted line represents the proportion of overall development assistance for health allocated to health sector support. To better understand how attempts to integrate system strengthening and disease-specific approaches have worked in practice, several studies have looked in greater depth at these efforts within Global Fund, Gavi and PEPFAR portfolios. Overall, they find that attempts have been less successful in bridging the gap between the two approaches than initially hoped. Global Fund grants that aimed to strengthen human resources were not coordinated with national plans and were invested in disease-focused in-service training, salary top-ups, performance incentives, and contracting part-time workers (Bowser et al. 2014). Gavi applications identified system constraints, but proposed interventions that were overwhelmingly oriented towards operational issues: purchase and delivery of equipment and medication and health workforce training, with little attention to barriers that affect citizen demand, and health worker behaviour and capacity (Goeman et al. 2010). The share of PEPFAR budgets allocated to governance and systems increased from 15% in 2004 to a high of 28% in 2013 but was inversely associated with the total budget (Moucheraud et al. 2016). In additional to health systems funding windows within disease-specific financing institutions, attempts have been made to create new funding platforms. The Health Systems Funding Platform, recommended in 2009 by the Task Force for Innovative International Financing for Health Systems (Hill et al. 2011; Brown et al. 2013) failed to gain traction.10 More recently, the Global Financing Facility attempts to strengthen health systems and increase funding for maternal and child health by mobilizing resources from donors, the World Bank, domestic sources and the private sector, heralded as ‘the future for the post-2015 agenda on diversifying financing for development’ (cited in Usher 2015, p. 1809). However, it remains to be see if this new facility is able to marshal the billions its aims to raise (Usher 2015; Horton 2016). More discouraging regarding the potential for synergies between disease-specific and sector-wide approaches is evidence of displacement effects between HIV funding and health sector funding and services. For countries with a high HIV burden, Lordan and colleagues (2011) find on average, a 1% increase in DAH for HIV was associated with a 0.09% decrease in funding to the health sector.11 Increases in HIV funding in sub-Saharan Africa were associated with reduced provision of childhood immunizations, have had no significant effect on maternal health services, but did have limited positive effects on prenatal blood testing (Grépin 2012). Discussion Taken together, these trends present a clear message. While the total volume and number of countries receiving health sector support has increased over time, it constitutes a small proportion of DAH—a far cry from the three-quarters deemed necessary to achieve the health SDG. Compared to sector support, earmarked funding has been much more substantive and allocated in more predictable ways. Across health areas, those included in the MDGs continue to benefit disproportionately. There is some evidence of displacement effects, indicating that even with a growing resource pie, funding one health area can have detrimental implications for others. Despite concerns regarding the potential misuse of budget support, its allocation has not been consistently associated with country governance indicators. Donors have implemented budget support differently and have reacted in contradictory ways to the same instance of corruption. The inconsistent allocation and application could reflect several factors. By nature, large reconstruction and trust funds established for fragile and conflict-affected states are allocated to countries with weak institutions. Donors may use policy dialogues associated with sector support for political reasons. Indeed, political economy factors may help to explain the behaviour of a particular donor in a particular country; however, these patterns appear to be widespread and not driven by a few outliers. Attempts to bridge the gap between earmarked funding and sector support have varied by disease and have been less successful in practice than intended. Using disease-specific funding mechanisms to strengthen health systems has not been used to address structural constraints. Even if the orientation shifts more in this direction, the level of this sub-stream of funding is likely to be insufficient to do so. The fundamental question remains: how, operationally, will the world advance the ambitious agenda it has set for itself? In terms of resource allocation, the improved alignment of DAH and DALYs for some diseases is an encouraging trend that should be reinforced, and corrected for NCDs. Like all global rankings, the Worldwide Governance Indicators are an imperfect measure; nevertheless, governance indicators appear to be underutilized and the inconsistent allocation of sector support suggests there remains great potential to considerably scale up the proportion of DAH provided in this way. At a minimum, donors could increase the use of this modality for countries with strong governance ratings, which would not pose a greater fiduciary risk than they are already taking with countries with weaker institutions. A synthesis report advises that health sector budget support be allocated on a large enough scale to expand public services to meet needs (Williamson and Dom 2010), threshold levels that do not appear to have been met. If universal health coverage, based on a primary care approach, is intended to be the cornerstone of the SDG health goal (Frenk 2009; World Health Organization 2015; Ottersen et al. 2017), sector support is a more direct way of advancing this goal. In terms of resource mobilization, it is not feasible to establish issue-specific funding mechanisms for each of the 169 targets. Some issue-specific financing will be part of the SDG response, but will not be able to cover these goals in the same way as disease-specific financing did for the MDGs. Moreover, many of the targets that are part of the health SDG are systems issues—access to quality services, workforce issues, early warning, risk reduction and management of health risks—and so would not fit well with a disease-specific approach. New platforms have not been implemented as anticipated, suggesting that the appetite for new initiatives may be waning. HIV may retain a larger relative share of development resources; however, the high profile of particular areas will continue to be threatened by an ever-expanding agenda space, both within and beyond the health sector. Undoubtedly, the way forward is more multi-dimensional than in the past. The SDGs offer the opportunity to address structural drivers and social determinants of health (Taylor et al. 2015; Horton 2016). A ‘health in all’ approach (Hill et al. 2014; Buse and Hawkes 2015; World Health Organization 2015) could improve health outcomes through efforts to address other issues, many of which are underlying contributors to poor health. The SDGs offer a leadership opportunity for intergovernmental organizations with a broad mandate, like the WHO,12 to whom states and key actors turn for guidance and norms when they are unable to resolve issues themselves. Attention is shifting from global to national, and in many cases subnational levels (World Health Organization 2015; Schwartländer et al. 2011). More recent discussions of development financing place a much stronger emphasis on domestic resources (UN 2015; Institute for Health Metrics and Evaluation 2017; Ottersen et al. 2017), although low income countries face an estimated $17–35 billion annual financing gap and will continue to rely on international funds (Stenberg et al. 2017). Both in terms of financing and service provision, there may be greater opportunities for sector-wide and cross-sector integration at subnational and national levels. Buffardi (2014), for instance, identified targeted aspects of sector-wide programs and system strengthening elements of targeted programs, observing that the tension between the two was less marked at the national level. More research is needed on integration strategies in different contexts to understand how countries and donors are adapting. Conclusion The last quarter century, in which health featured prominently on the global agenda, has revealed clear funding patterns, overwhelmingly in favour of specific diseases rather than health systems. Proposals to finance public goods, including health, and to shape global institutions must account for these trends. This mismatch between the many, diverse and interconnected global goals and the way in which programs are financed threatens the achievement of the SDGs. National experimentation may offer alternatives, both within and beyond the health sector, that warrant closer attention and systematic investigation. Acknowledgements This work was initiated when the author was a doctoral candidate at the Evans School of Public Policy and Governance at the University of Washington. The author would like to thank Mary Kay Gugerty, Sanjeev Khagram, Peter May and Joe Zunt for providing valuable comments on an early version of this article, and to Katie Banks, Łukasz Marć, Anne Roemer-Mahler, participants at the 2017 International Studies Association conference in Baltimore and two anonymous reviewers for their insightful contributions to the current version. Funding Partial support for this research came from a Eunice Kennedy Shriver National Institute of Child Health and Human Development research infrastructure grant, P2C HD042828, to the Center for Studies in Demography & Ecology at the University of Washington. Footnotes 1 The 2005 Paris Declaration on Aid Effectiveness and the 2008 Accra Agenda for Action, 2011 Busan Partnership for Effective Development Cooperation, 2014 First High-Level Meeting of the Global Partnership for Effective Development Cooperation: Building Towards an Inclusive Post-2015 Development Agenda, 2015 Addis Ababa Action Agenda of the Third International Conference on Financing for Development, 2016 Global Partnership for Effective Development Cooperation: Nairobi Outcome Document. 2 Some donors may also have existing commitments to particular issues or health conditions. 3 Sources cover 23 bilateral donors (Australia, Austria, Belgium, Canada, Switzerland, Germany, Denmark, Spain, Finland, France, Greece, Ireland, Italy, Japan, Korea, Luxembourg, The Netherlands, Norway, New Zealand, Portugal, Sweden, UK and USA), non-OECD DAC countries, the Bill and Melinda Gates Foundation (BMGF), corporate donors and debt repayments. Channels include the same bilateral donors and BMGF, as well as the European Commission, Gavi, the Global Fund, five multilateral development banks (World Bank International Bank for Reconstruction and Development, World Bank International Development Assistance, InterAmerican Development Bank, African Development Bank, Asian Development Bank), five UN agencies (PanAmerican Health Organisation, UNAIDS, UNFPA, UNICEF, WHO), US non-governmental organization (NGOs), international NGOs and US foundations. The IHME data set uses financial data from the Organization for Economic Cooperation and Development Creditor Reporting System as a base, with corrections for underreporting and double counting. A detailed explanation of the methods used to construct the data set is available as an online supplementary appendix to: Dieleman et al. (2016). 4 Government effectiveness captures perceived quality of public services, policy formulation and implementation, and independence of civil service; control of corruption, the extent to which public power is used for private gain; voice and accountability, freedom of expression and association, free media and democratic elections; political stability and absence of violence; regulatory quality, ability of the government to formulate and implement policies promoting private sector development; and rule of law, quality of contract enforcement, property rights, policy and the courts. These indicators were available biannually from 1996 to 2002 and annually thereafter. 5 Botswana, Namibia, South Africa, Mexico and the Dominican Republic received more than five times the total DAH predicted based on DALYs and GDP. Iran, Chile, Venezuela, Algeria, Malaysia, North Korea and the Central African Republic received less than 20%. 6 As mentioned in the methods section, the data set distinguishes between source (bilateral donors, BMGF, corporate donors and debt repayments) and channel (bilateral donors, multilaterals organizations, UN agencies, US foundations and NGOs). It is also important to note that sector support may vary more on an annual basis, compared to project assistance, as a result of large tranche payments in a particular year. 7 Faust et al. (2012) give the example of Zambia, where Sweden and the Netherlands suspended budget support distributions in response to fraud in the health sector in 2009. The EC released funds, interpreting the detection and prosecution as strengthened government regulation and oversight. 8 Correlations are based on DAH and country governance rankings since 1996. The number of observations varied slightly by governance indicator: political stability n = 2344, regulatory quality and government effectiveness n = 2362, control of corruption n = 2368, voice and accountability n = 2396, rule of law n = 2397. 9 The 10 top, median and bottom ranked countries in control of corruption and government effectiveness who received development assistance for health were identified by country rankings in 1998, the year when health sector support was at its peak, and countries’ median rankings from 1996 to 2014, the period for which Worldwide Governance Indicators and country-level DAH are available. 10 Had the Health Systems Funding Platform been instituted, it may still have been pulled towards shorter-term, discrete, operational issues rather than longer-term, more systematic concerns, similar to the experiences of the Global Fund, Gavi and PEPFAR described above. 11 The displacement effect was much more pronounced for malaria, and not found for TB or an aggregate grouping of all other health areas. These regression analyses are also based on IHME data. 12 The WHO (2015) has asserted that a pragmatic middle ground is possible. At the same time, they acknowledge that previous trends do not suggest that necessary shifts will take place and conclude that ‘the jury is still out on how the SDGs will influence financing for health’ (p. 198). References Allen L. 2017. Non-communicable disease funding. The Lancet Diabetes & Endocrinology  5: 92. Google Scholar CrossRef Search ADS PubMed  Allen L. 2016. Why is there no funding for non-communicable diseases? Journal of Global Health Perspectives . https://ora.ox.ac.uk/objects/uuid:3fe582f2-f4c7-46ad-9e4d-26ac33add7bb. Ahsan KZ, Streatfield PK, Ijdi RE et al.   2016. Fifteen years of sector-wide approach (SWAp) in Bangladesh health sector: an assessment of progress. Health Policy and Planning  31: 612– 23. Google Scholar CrossRef Search ADS PubMed  Beaglehole R, Bonita R, Horton R et al.   2011. Priority actions for the non-communicable disease crs. The Lancet  377: 1438– 47. Google Scholar CrossRef Search ADS   Behague DP, Storeng KT. 2008. Collapsing the vertical-horizontal divide: an ethnographic study of evidence-based policymaking in maternal health. American Journal of Public Health  98: 644– 9. Google Scholar CrossRef Search ADS PubMed  Bhutta ZA, Cabral S, Chan CW, Keenan WJ. 2012. Reducing maternal, newborn, and infant mortality globally: an integrated action agenda. International Journal of Gynecology & Obstetrics  119: S13– 7. Google Scholar CrossRef Search ADS   Biesma RG, Brugha R, Harmer A et al.   2009. The effects of global health initiatives on country health systems: a review of the evidence from HIV/AIDS control. Health Policy and Planning  24: 239– 52. Google Scholar CrossRef Search ADS PubMed  Bossert TJ. 2012. Health systems. Health Policy and Planning  27: 8– 10. Google Scholar CrossRef Search ADS PubMed  Bowser D, Sparkes SP, Mitchell A et al.   2014. Global fund investments in human resources for health: innovation and missed opportunities for health systems strengthening. Health Policy and Planning  29: 986– 97. Google Scholar CrossRef Search ADS PubMed  Brown SS, Sen K, Decoster K. 2013. The health systems funding platform and World Bank legacy: the gap between rhetoric and reality. Globalization and Health  9: 9. Google Scholar CrossRef Search ADS PubMed  Buffardi AL. 2014. From polarisation to practice: puzzles and insights on integrated approaches from public health professionals. Global Public Health  9: 741– 51. Google Scholar CrossRef Search ADS PubMed  Buse K, Hawkes S. 2015. Health in the sustainable development goals: ready for a paradigm shift? Globalization and Health  11: 13. Google Scholar CrossRef Search ADS PubMed  Clinton C, Sridhar D. 2017. Who pays for cooperation in global health? A comparative analysis of WHO, the World Bank, the Global Fund to Fight HIV/AIDS, Tuberculosis and Malaria, and Gavi, the Vaccine Alliance. The Lancet 390: 324– 32. Dieleman JL, Graves CM, Templin T et al.   2014. Global health development assistance remained steady in 2013 but did not align with recipients’ disease burden. Health Affairs  33: 878– 86. Google Scholar CrossRef Search ADS PubMed  Dieleman JL, Schneider MT, Haakenstad A et al.   2016. Development assistance for health: past trends, associations, and the future of international financial flows for health. The Lancet  387: 2536– 44. Google Scholar CrossRef Search ADS   Deverajan S, Swaroop V. 2000. The Implications of Foreign Aid Fungibility for Development Assistance . Washington, DC: World Bank. Google Scholar CrossRef Search ADS   Eichenauer VZ, Reinsberg B. 2017. What determines earmarked funding to international development organizations? Evidence from the new multi-bi aid data. Review of International Organization  12: 171. Google Scholar CrossRef Search ADS   Eichler R, Glassman A. 2008. Health Systems Strengthening via Performance-based Aid: Creating Incentives to Perform and to Measure Results . Brookings Global Economy and Development Working Paper Series. Washington, DC: Brookings Institution. Epstein H. 2009. Towards a global fund for the health MDGs? The Lancet  373: 2110– 1. Google Scholar CrossRef Search ADS   Faust J, Koch S, Molenaers N, Tavakoli H, Vanheukelom J. 2012. The Future of EU Budget Support: Political Conditions, Differentiation and Coordination . London: European Think Tanks Group. Fernandes Antunes A, Xu K, James CD et al.   2013. General budget support: has it benefited the health sector? Health Economics  22: 1440– 51. Google Scholar CrossRef Search ADS PubMed  Foster M, Brown A, Conway T. 2000. Sector-wide Approaches for Health Development: A Review of Experience . Geneva, Switzerland: World Health Organization. Frenk J. 2009. Reinventing primary health care: the need for systems integration. The Lancet  374: 170– 3. Google Scholar CrossRef Search ADS   Grépin KA. 2012. HIV donor funding has both boosted and curbed the delivery of different non-HIV health services in sub-Saharan Africa. Health Affairs  31: 1406– 14. Google Scholar CrossRef Search ADS PubMed  Goeman L, Galichet B, Porignon DG et al.   2010. The response to flexibility: country intervention choices in the first four rounds of the Gavi Health Systems Strengthening applications. Health Policy and Planning  25: 292– 9. Google Scholar CrossRef Search ADS PubMed  Hafner T, Shiffman J. 2013. The emergence of global attention to health systems strengthening. Health Policy and Planning  28: 41– 50. Google Scholar CrossRef Search ADS PubMed  Handley G. 2009. Sector Budget Support in Practice: Literature Review . London: Overseas Development Institute. Hayman R. 2011. Budget support and democracy: a twist in the conditionality tale. Third World Quarterly  32: 673– 88. Google Scholar CrossRef Search ADS   Hill PS, Vermeiren P, Miti K, Ooms G, Van Damme W. 2011. The health systems funding platform: is this where we thought we were going? Globalization and Health  7: 16. Google Scholar CrossRef Search ADS PubMed  Hill PS, Buse K, Brolan CE, Ooms G. 2014. How can health remain central post-2015 in a sustainable development paradigm? Globalization and Health  10: 18. Google Scholar CrossRef Search ADS PubMed  Horton R. 2016. Offline: global health—tipping into irrelevance. The Lancet  388: 1362. Google Scholar CrossRef Search ADS   Institute for Health Metrics and Evaluation. 2016. Financing Global Health 2015: Development Assistance Steady on the Path to New Global Goals . Seattle, WA: Author. Institute for Health Metrics and Evaluation. 2017. Financing Global Health 2016: Development Assistance, Public and Private Health Spending for the Pursuit of Universal Health Coverage . Seattle, WA: Author. Kaufmann D, Kraay A, Mastruzzi M. 2010. The Worldwide Governance Indicators: A Summary of Methodology, Data and Analytical Issues . World Bank Policy Research Working Paper No. 5430. Washington, DC: The World Bank. Google Scholar CrossRef Search ADS   Koeberle S, Stavreski Z. 2006. Budget support: concept and issues. In: Koeberle S, Stavreski Z, Walliser J (eds). Budget support as more effective aid? Recent experiences and emerging lessons . Washington, DC: The World Bank, 3– 28. Google Scholar CrossRef Search ADS   Langan M. 2015. Budget support and Africa–European Union relations: free market reform and neo-colonialism? European Journal of International Relations  21: 101– 21. Google Scholar CrossRef Search ADS   Lawson A, Booth D, Harding A, Hoole D, Naschold F. 2003. General Budget Support Evaluability Study Phase 1: Synthesis Report, Vol. I. London: DFID Evaluation Report EV 643. Lawson A, Booth D, Msuya M, Wangwe, Williamson T. 2006. Does general budget support work? Evidence from Tanzania. In: Koeberle S, Stavreski Z, Walliser J (eds). Budget Support as More Effective Aid? Recent Experiences and Emerging Lessons . Washington, DC: The World Bank, 121– 36. Levine R. 2008. Healthy foreign policy: bringing coherence to the global health agenda. In: Birdsall (ed). The White House and the World: A Global Development Agenda for the Next US President . Washington, DC: Center for Global Development, 43– 62. Lordan G, Tang KK, Carmignani F. 2011. Has HIV/AIDS displaced other health funding priorities? Evidence from a new dataset of development aid for health. Social Science & Medicine  73: 351– 5. Google Scholar CrossRef Search ADS   Lu C, Schneider MT, Gubbins P et al.   2010. Public financing of health in developing countries: a cross-national systematic analysis. The Lancet  375: 1375– 87. Google Scholar CrossRef Search ADS   Mills A. 2005. Mass campaigns versus general health services: what have we learnt in 40 years about vertical versus horizontal approaches? Bulletin of the World Health Organization  83: 315– 6. Google Scholar PubMed  Molenaers N. 2012. The great divide? Donor perceptions of budget support, eligibility and policy dialogue. Third World Quarterly  33: 791– 806. Google Scholar CrossRef Search ADS   Molenaers N, Gagiano A, Smets L, Dellepiane S. 2015. What determines the suspension of budget support? World Development  75: 62– 73. Google Scholar CrossRef Search ADS   Moucheraud C, Sparkes S, Nakamura Y et al.   2016. PEPFAR investmetns in governance and health systems were one-fifth of countries’ budgeted funds, 2004–14. Health Affairs  35: 847– 55. Google Scholar CrossRef Search ADS PubMed  Nugent R. 2015. Bilateral and Multilateral Financing for Noncommunicable Diseases. Policy Brief Prepared for the WHO Global Coordination Mechanism on the Prevention and Control of Noncommunicable Diseases, Working Group on How to Realise Governments' Commitment to Provide Financing for NCDs. Geneva: WHO. http://www.who.int/nmh/ncd-coordination-mechanism/Policybrief5.2docx.pdf, accessed 14 September 2017. Ooms G, Van Damme W, Baker BK, Zeitz P, Schrecker T. 2008. The ′diagonal′ approach to Global Fund financing: a cure for the broader malaise of health systems? Global Health  4: 6. Google Scholar CrossRef Search ADS PubMed  Ottersen T, Evans DB, Mossialos E, Røttingen JA. 2017. Global health financing towards 2030 and beyond. Health Economics, Policy and Law  12: 105– 11. Google Scholar CrossRef Search ADS   Peters DH, Paina L, Schleimann F. 2013. Sector-wide approaches (SWAps) in health: what have we learned? Health Policy and Planning  28: 884– 90. Google Scholar CrossRef Search ADS PubMed  Ravishankar N, Gubbins P, Cooley RJ et al.   2009. Financing of global health: tracking development assistance for health from 1990 to 2007. The Lancet  373: 2113– 24. Google Scholar CrossRef Search ADS   Schwartländer B, Stover J, Hallett T et al.   2011. Towards an investment approach for an effective response to HIV/AIDS. The Lancet  377: 2031– 41. Google Scholar CrossRef Search ADS   Shiffman J. 2008. Has donor prioritization of HIV/AIDS displaced aid for other health issues? Health Policy and Planning  23: 95– 100. Google Scholar CrossRef Search ADS PubMed  Shiffman J, Berlan D, Hafner T. 2009. Has aid for AIDS raised all health funding boats? Journal of Acquired Immune Deficiency Syndromes  52: S45– 8. Google Scholar CrossRef Search ADS PubMed  Smith SL, Rodriguez MA. 2016. Agenda setting for maternal survival: the power of global health networks and norms. Health Policy and Planning  31: i48– 59. Google Scholar CrossRef Search ADS PubMed  Stenberg K, Hanssen O, Edejer TTT et al.   2017. Financing transformative health systems towards achievement of the health Sustainable Development Goals: a model for projected resource needs in 67 low-income and middle-income countries. The Lancet Global Health  5: e875– 87. Google Scholar CrossRef Search ADS PubMed  Tavakoli H, Smith G. 2013. Back under the microscope: insights from evidence on budget support. Development Policy Review  31: 59– 74. Google Scholar CrossRef Search ADS   Taylor S, Williams B, Magnus D, Goenka A, Modi N. 2015. From MDG to SDG: good news for global child health? The Lancet  386: 1213. Google Scholar CrossRef Search ADS   UN Department of Economic and Social Affairs Technical Support Team in support of the General Assembly Open Working Group on Sustainable Development Group. October 2014. Compendium of TST Issue Briefs: Health and Sustainable Development. https://sustainabledevelopment.un.org/content/documents/1554TST_compendium_issues_briefs_rev1610.pdf, accessed 16 February 2017. United Nations. 2015. Addis Ababa Action Agenda of the Third International Conference on Financing for Development. http://www.un.org/esa/ffd/wp-content/uploads/2015/08/AAAA_Outcome.pdf, accessed 14 September 2017. Uplekar M, Raviglione MC. 2007. The “vertical-horizontal” debates: time for the pendulum to rest (in peace)? Bulletin of the World Health Organization  85: 413– 4. Google Scholar CrossRef Search ADS PubMed  USAID Development Information Services. 2006. General budget support: key findings of five USAID studies. In: Koeberle S, Stavreski Z, Walliser J (eds). Budget Support as More Effective Aid? Recent Experiences and Emerging Lessons . Washington, DC: The World Bank, 105– 16. Google Scholar CrossRef Search ADS   Usher AD. 2015. Global Financing Facility: where will the funds come from? The Lancet 386: 1809. Woll B. 2008. Donor harmonisation and government ownership: multi-donor budget support in Ghana. The European Journal of Development Research  20: 74– 87. Google Scholar CrossRef Search ADS   Williamson T, Dom C. 2010. Sector Budget Support in Practice: Synthesis Report . London: Overseas Development Institute. World Bank. 1998. Assessing Aid: What Works, What Doesn’t, and Why . New York, NY: Oxford University Press. World Health Organization. 2015. Health in 2015: From MDGs, Millennium Development Goals to SDGs. Sustainable Development Goals . Geneva: Author. World Health Organization Maximizing Positive Synergies Collaborative Group. 2009. An assessment of interactions between global health initiatives and country health systems. The Lancet  373: 2137– 69. CrossRef Search ADS   © The Author(s) 2018. Published by Oxford University Press in association with The London School of Hygiene and Tropical Medicine. All rights reserved. For permissions, please e-mail: journals.permissions@oup.com

Journal

Health Policy and PlanningOxford University Press

Published: Apr 1, 2018

There are no references for this article.

You’re reading a free preview. Subscribe to read the entire article.


DeepDyve is your
personal research library

It’s your single place to instantly
discover and read the research
that matters to you.

Enjoy affordable access to
over 12 million articles from more than
10,000 peer-reviewed journals.

All for just $49/month

Explore the DeepDyve Library

Unlimited reading

Read as many articles as you need. Full articles with original layout, charts and figures. Read online, from anywhere.

Stay up to date

Keep up with your field with Personalized Recommendations and Follow Journals to get automatic updates.

Organize your research

It’s easy to organize your research with our built-in tools.

Your journals are on DeepDyve

Read from thousands of the leading scholarly journals from SpringerNature, Elsevier, Wiley-Blackwell, Oxford University Press and more.

All the latest content is available, no embargo periods.

See the journals in your area

DeepDyve Freelancer

DeepDyve Pro

Price
FREE
$49/month

$360/year
Save searches from Google Scholar, PubMed
Create lists to organize your research
Export lists, citations
Read DeepDyve articles
Abstract access only
Unlimited access to over
18 million full-text articles
Print
20 pages/month
PDF Discount
20% off