Authors : Catherine Blampied, Soumya Chattopadhyay, Romilly Greenhill, Imran Aziz, Christine Ellison, Rachel Thompson, Abdul-Gafaru Abdulai and Adam Salifu
Publication Site : ALNAP
Publication Type : Report
Date of publication : April 2018
The Sustainable Development Goals (SDGs) paint an inspiring vision of what the world could look like in 2030 (UN, 2015). A fundamental tenet of the SDGs is the concept of ‘leaving no one behind’. Most notably, in signing up to these goals as part of Agenda 2030, governments have pledged to ‘reach those furthest behind first’ (Ibid.) – a quite different paradigm to that of the Millennium Development Goals (MDGs), which focused on average or aggregate progress.
Yet, despite multiple references to the concept in Agenda 2030 and in other policy discourse at international and national levels, there is a risk that the concept is left open to a very broad range of interpretations. While its achievement is pivotal to the success of Agenda 2030, the concern is that ‘leaving no one behind’ will not be implemented or monitored in the same way as other aspects of the SDGs.
The SDG that directly relates to health is Goal 3: ‘Ensure healthy lives and promote well-being for all at all ages’, which has 13 associated targets. In this report, we focus our attention on target 3.8, universal health coverage (UHC). This was chosen because it underpins efforts at leaving no one behind across the other targets.
Education is recognised in the SDGs as both a fundamental human right and an enabling right that is central to the realisation of the whole Agenda 2030
In addition, one of the most challenging aspects of achieving the goal is delivering services to those most in need and more specifically to the poorest and most marginalised groups (Wong, 2015). Furthermore, it is now well documented that expanding health coverage towards universality, and diminishing health inequities, is associated with improved health outcomes for the overall population (with the largest gains accruing to poor people) (Moreno-Serra and Smith, 2012).
Education is recognised in the SDGs as both a fundamental human right and an enabling right that is central to the realisation of the whole Agenda 2030. The education goal is SDG 4 : ‘Ensure inclusive and equitable quality education and promote lifelong learning opportunities for all’, but it is also mentioned as a key enabler under many other goals. SDG4 covers learning from early childhood through adulthood, while stressing the key themes of education quality, learning, inclusion and equity.
In this report, we focus on targets 4.1 (‘ensure that all girls and boys complete free, equitable and quality primary and secondary education leading to relevant and effective outcomes’) and 4.5 (‘eliminate gender disparities in education and ensure equal access to all levels of education and vocational training for the vulnerable, including persons with disabilities, indigenous peoples and children in vulnerable situations’).
In this study, we focus predominantly on basic education, i.e. pre-secondary level. This is because a good-quality basic education lays the necessary foundation for a child’s progression through the rest of the education system, and for myriad life opportunities beyond. It will be impossible to ensure that all children complete free, equitable and quality secondary education – as pledged in SDG 4.1 – unless all children can first complete free, equitable and quality primary education (which also facilitates a smooth transition to lower secondary school, at a juncture when some children are particularly vulnerable to dropping out).
Moreover, evidence shows that inequalities in educational attainment generally emerge even before the end of primary school; improving equity and inclusion rests on addressing these as early on as possible (Rose and Alcott, 2015). Children’s learning outcomes and the quality of education are increasingly recognised as critical issues, including for equity (World Bank, 2017d).
Owing to issues relating to data availability, we focus primarily on indicators of coverage and access; however, we do examine some dimensions of quality and learning, such as examination pass rates, where the data allows this at a disaggregated level.
Knowing who is left behind in a country is one thing, actually doing something about it is another. Insofar as politics is about ‘who gets what, when and how’ (Laswell, 1936), creating policies to improve the lives of marginalised people is an inherently political process. For instance, how should governments go about financing health care that is available to and affordable for all? What package of services should be offered? Sophisticated curative care may be of significant interest to doctors and the middle classes, but preventive and primary health services are likely to be more important to the poor.
In education, if fees are to be charged, in which levels of the health system should this be implemented and how much should be charged? How should public funding be split between pre-primary, primary, secondary and tertiary levels? Should a new hospital or school be built in this or that location? If the overall percentage of national resources spent on health or education is to increase, resources must be taken away from something else. Prioritisation and difficult trade-offs need to be decided – and all are political.
Furthermore, there is no guarantee that money will be spent effectively or equitably. Different institutions
and actors need to coordinate effectively. Finances must be managed efficiently, and make their way through the system to the right people. Human resources must be distributed to the right places and staff motivated to do their jobs well.
Indeed, getting health workers and teachers to take up posts in deprived or rural areas where the marginalised people often live is frequently a challenge in developing countries (World Bank, 2003) and, indeed, in wealthier countries as well (Ono et al., 2014). Institutional coordination, monitoring and supervising, and incentivising demand – activities inextricably bound up with health and education governance – are all inherently political.
Whether a pro-poor policy is actually implemented relies both on favourable political dynamics and accountability to avoid elite capture or blockage by adverse interests, and a strong administrative system that can allocate resources efficiently
Our conceptual framework recognises that there are both political and technical requirements at each stage of the causal chain outlined above. For example, there must be political will to define and monitor who is being left behind, and there must also be the technical capacity to implement effective data systems, centrally and locally.
Whether a pro-poor policy is actually implemented relies both on favourable political dynamics and accountability to avoid elite capture or blockage by adverse interests, and a strong administrative system that can allocate resources efficiently. We identify bottlenecks along the way in each of the areas of the conceptual model, and conclude with a set of policy recommendations (Chapter 7).
Another important aspect of the stocktake framework is to listen to the perspectives of poor and marginalised communities about their own needs and priorities, and what they perceive as the main points of progress and challenges in accessing health and education services.
Such information is rarely gathered with respect to official policy-making, planning and evaluation processes, especially for typically vulnerable and left-behind groups, but it is important for guiding efforts to improve service delivery.
The two regions selected for sub-national investigation were Northern and Brong Ahafo. These were identified because they had different levels of health access (as measured by the CCI), but similar levels of financing and facilities (see Table 3). Moreover, Northern region forms part of the north of Ghana and is predominantly Muslim, while Brong Ahafo is considered part of the south and is predominantly Christian. Although the study was not set up to be explicitly and exclusively comparative, we considered that such differences would be useful to explore possible supply- and demand-side barriers beyond the apparently similar levels of financing and facilities.
Our hypothesis was that the quality and effectiveness of the supply-side inputs would be found to vary across regions, and/or there would be context-specific demand-side barriers undermining access. We made the case selection based on health criteria, rather than education, to keep the analysis simple and for consistency with the approach used in the previous case studies. However, as we will see in Chapter 3, education access is also much better in Brong Ahafo than in Northern.
After selecting two regions, we also chose one district in each region for more detailed investigation: Banda in Brong Ahafo region and Zabzugu in Northern region. The main criteria for selecting these districts was the existence of reputable local partners who would be able to facilitate access to decision-makers and arrange FGDs. We also checked that financial, economic and health access indicators in each district were broadly representative of their respective regions as a whole.
A major divergence concerns health worker distribution. Brong Ahafo has 33% more doctors per capita than Northern region, and 38% more midwives per woman of fertile age. The other big difference between the regions is their territorial size.
Overall, Brong Ahafo and Northern region have a very similar population size, and a similar level of per capita spending, just below the national average. GoG facilities are only slightly more numerous in Brong Ahafo. However, the total number of health facilities is significantly higher in Brong Ahafo, partly as a result of a larger number of facilities run by the quasi-public, faith-based provider, the Christian Health Association of Ghana (CHAG). CHAG does not operate in the predominantly Muslim Northern region.
A major divergence concerns health worker distribution. Brong Ahafo has 33% more doctors per capita than Northern region, and 38% more midwives per woman of fertile age. The other big difference between the regions is their territorial size. Northern region covers an area of 70,384 km2, almost twice the size of Brong Ahafo, and thus the population density of the former is significantly lower.
This typically makes providing equivalent services more expensive and logistically challenging. For example, to ensure health facilities are available within a given distance of each community, Northern region would need considerably more health facilities per capita than Brong Ahafo.
Inequality in CCI between groups is considerably higher in Northern region than Brong Ahafo. Whereas Brong Ahafo seems to have achieved remarkably similar coverage for the bottom wealth quintile as the top four quintiles, Northern is more similar to (and slightly worse than) Ghana’s national gap, with an 11.4% gap in CCI between the bottom quintile and the rest. In fact, the average CCI of the top four wealth quintiles in Northern is just marginally higher than the bottom quintile of the national average – indicating poor overall health service access in the entire region.
In Northern region, CCI coverage also varies more significantly than in Brong Ahafo across urban versus rural groups, ethnic minorities versus non-minorities and those with more versus those with less education. This may be because the larger distances between communities and health facilities cause greater direct and/or indirect barriers to the poorest households, which, for example, are less able to afford to pay for transport. Rural households and ethnic minorities are also more likely to live in remote locations further from health facilities, on average.
In education, we see large divergences between Northern and Brong Ahafo regions. Brong Ahafo falls close to the national average in each case, while Northern falls far below. The gross primary enrolment rates (at least on paper) appear quite similar – 113% in Brong Ahafo versus 115% in Northern. However, nearly a fifth (18%) of 13-15 year olds in Northern have no formal education, compared to just 2% in Brong Ahafo.
While the total size of the populations of Brong Ahafo and Northern regions are fairly similar, the school-age population is much larger in Northern region, which is consistent with the higher fertility rates there (6.6 in Northern, compared to 4.8 in Brong Ahafo).17 Thus, while the overall budget for education is slightly higher in Northern region, public spending per school-age child is lower.
Another factor is that there appear to be many more non-public (i.e. private or non-governmental) primary schools in Brong Ahafo region, meaning the number of children per school is 40% lower. Finally, the much larger physical area of Northern region also means that the average distance between schools and children’s households will also be greater, combined with poorer road infrastructure and transport options.
Ghana’s national income poverty rate declined from 31.9% in 2005-2006 to 24.2% in 2012-2013. The extreme poverty rate declined even more quickly, dropping from 16.5% in 2005-2006 to 8.4% in 2012-2013 (GSS, 2014b).
However, while overall levels of poverty have declined, they remain high in some places and among certain groups; for example, ‘women and girls [are] performing worse across all the main social indicators’ (GoG, 2014: 3). On the Multidimensional Poverty Index, 49% of Ghana’s rural population is estimated to live in multidimensional poverty, compared to less than 18% in urban areas.
The north of Ghana has also continued to lag far behind the rest of the country in educational attainment and health outcomes
The three northern regions of the country (i.e. Northern, Upper East and Upper West regions) experience significantly higher levels of poverty than the rest of the country. In the most recent Ghana Living Standards Survey (GLSS), undertaken in 2012-2013, 45.1% of the population in the Upper West region was deemed as extremely income-poor compared to a national average of 8.4% (GSS, 2014b). Nearly 74% of people in Northern region are estimated to live in multidimensional poverty, compared to 14% in Greater Accra, for example. The GoG acknowledges high levels of inequality as a major problem, describing it as ‘a dangerous sign that the poverty reduction effort is not being properly targeted at those who need it most’ (GoG, 2012).
The north of Ghana has also continued to lag far behind the rest of the country in educational attainment and health outcomes (GSS et al., 2015; MoE, 2016). The under-five mortality rate (per 1,000 live births) is 128 in Northern region, compared to a national average of 82. Chronic malnutrition (stunting) affects 22% of children under five nationwide, but 37% of children in Northern region (World Bank, 2016).
In the 2016 National Education Assessment, in Greater Accra, only about 5% of students recorded less than minimum competency rates in mathematics and English, but the corresponding figures for the Upper East and Upper West regions were 47% and 51% respectively (MoE and GES, 2016). These inequalities have their roots in British colonial policies that subordinated the interests of the north to those of the south (see Box 1).
Assessing health coverage – the Composite Coverage Index
The next section of analysis uses the CCI of access to eight key reproductive, maternal, newborn and child health services (see Chapter 1 for a fuller explanation) as the main metric to examine who is being left behind in health service coverage. We analyse patterns of CCI by region and population group, both for the most recent data point available (2014) and over time, referring back to the results of the previous GDHS (2008).
Ghana’s overall CCI performance in 2014 was 65.9% – this falls below the median among all LICs and MICs of 70.6%
Ghana’s overall CCI performance in 2014 was 65.9% – this falls below the median among all LICs and MICs of 70.6%. Comparing Ghana’s score with the results from our earlier stocktakes for Kenya and Nepal also suggests poor performance. Ghana’s CCI score is considerably below that of Kenya, another sub-Saharan African LMIC, at 76.0%, and virtually the same as Nepal’s score, despite the latter being a least developed country with GDP per capita equal to roughly half that of Ghana.
The biggest differences in health access in Ghana are associated with regional variation. Northern region’s average CCI is just 49.5%, 16 percentage points behind the national average. There is currently a 22 percentage point gap between the best-performing (Upper East) and worst-performing regions (Northern). This is a lower level of geographic inequality than in Kenya, where the gap between best and worst-performing regions was 38.
On a national basis, the highest level of education attended in the household is the next most significant factor associated with health service coverage, followed by household wealth
Figure 15: CCI by region, 2014 percentage points; but more unequal than Nepal, where the gap was 16 percentage points. Ghana’s best-performing region, Upper East, is one of the three that comprise the historically marginalised north of Ghana and is home to high levels of deprivation. However, there is a particular story of exceptional leadership and local innovation in Upper East (see Chapter 4) and, as shown in Chapter 5, it has in recent years received an above-average per capita health budget.
Even groups we posited as ‘non-marginalised’ fare worse in Northern region than most marginalised groups in other regions, suggesting that exclusion from health coverage in Ghana is predominantly a regional phenomenon. In this, our findings reinforce a well-established literature that emphasises the historical deprivation and marginalisation of Ghana’s north (Abdulai and Hulme, 2014; Saleh, 2013).
On a national basis, the highest level of education attended in the household is the next most significant factor associated with health service coverage, followed by household wealth. In households where the highest level of education attended was primary (or none), the average CCI was 61%, compared to 79% in households where the highest level attended was secondary or above.49 In the poorest quintile of households, the average CCI was 58%, compared to 68% for the rest of the population.
The lack of policies and measures to incentivise and enforce the distribution of teachers across deprived areas has a large impact on the functioning and quality of education among marginalised communities, especially in the north of Ghana.
Even groups we posited as ‘non-marginalised’ fare worse in Northern region than most marginalised groups in other regions, suggesting that exclusion from health coverage in Ghana is predominantly a regional phenomenon
The final major set of issues in financing Ghana’s health system is the raft of challenges concerning the NHIS, which are disproportionately affecting the poorest and most vulnerable. While these extend far beyond the specific focus of this study and have been well covered in other literature, the current crisis facing the NHIS not only overwhelms debate about health policy in Ghana, but also disproportionately affects the poorest and most vulnerable people who rely on functioning, free, public health care and, in many cases, NHIS exemptions.
The National Health Insurance Fund (NHIF) constitutes 3% of Ghana’s total government spending (World Bank, 2017b). As outlined in Chapter 4, the NHIS was established as the key tool to bring the country towards UHC, with certain categories of vulnerable people exempted from paying premiums so as to subsidise their health care demand. These exempted groups now account for two-thirds of enrolments.84 However, only 38% of the total population is enrolled, and many of these people are not actually receiving the benefits they should (Atim and Amporfu, 2016).
Two recent reviews of the NHIS – one by the World Bank focused on efficiency and financial sustainability, and another commissioned by the government to look at the whole NHIS system – have concluded that there are multiple sources of inefficiency (World Bank, 2017b; Atim and Amporfu, 2016).
One important finding of these reviews is that, while the NHIS offers an (overly) broad and generous package of services, it excludes many cost-effective preventive services and does not adequately focus on quality of care, thus the NHIS is ‘essentially paying for the consequences of under-performance of public health programmes’ (Atim and Amporfu, 2016). There is a real risk that a well-meaning and progressive policy is producing regressive effects owing to problems in its financing and implementation.
Les Wathinotes sont des extraits de publications choisies par WATHI et conformes aux documents originaux. Les rapports utilisés pour l’élaboration des Wathinotes sont sélectionnés par WATHI compte tenu de leur pertinence par rapport au contexte du pays. Toutes les Wathinotes renvoient aux publications originales et intégrales qui ne sont pas hébergées par le site de WATHI, et sont destinées à promouvoir la lecture de ces documents, fruit du travail de recherche d’universitaires et d’experts.
Wathinotes are excerpts from publications chosen by WATHI and conform to the original documents. The reports used for the development of the Wathinotes are selected by WATHI based on their relevance to the country context. All Wathinotes refer to original and integral publications that are not hosted by the WATHI website, and are intended to promote the reading of these documents, which are the result of the research work of academics and experts.