Effectiveness of existing health programmes at reaching the poor
Von Veronica Walford
Do we know how effective existing programmes are at reaching the poor? What does the evidence show? An overview and some points to consider in addressing this issue.
Do we know how effective existing health programmes are at reaching the poor? - Routine health information does not tell us who is benefiting from health programmes in terms of whether they are poor. There may be data that shows use by gender and geographic location which can give some indication, when combined with data on where the poor live (such as immunisation coverage by district). But routine data does not show who is actually using the services within those districts.
Several types of research and studies are available that do try to capture the issue of who benefits from health programmes. Major types are:
Population based household surveys that correlate service use with household assets. The Demographic and Health Surveys (DHS) are the largest that focus on health and tend to be repeated every five years or so, allowing for monitoring of trends. Living Standards Measurement Surveys (LSMS) are another type of large survey although their coverage of health tends to be limited. Another useful approach is regular service delivery surveys that look at service use and views on services, such as those in Bangladesh. There may also be efforts to monitor whether the poor and vulnerable are benefiting from different services (e.g. education, health and security) as part of the monitoring of implementation of poverty reduction strategies.
Benefit incidence analysis – looking at who within the population benefits from public subsidies to health services. This combines information on who uses services with the costs of services to see who is getting the benefit of public spending.
Service user surveys comparing users to the general population. This was typically the design of a recent series of studies under the ‘Reaching the Poor Program’ supported by World Bank and others, and presented at a meeting in February 2004.
There are also qualitative studies that look at who uses services and why. Issues however remain in trying to assess whether the poor are benefiting:
How to assess who is poor? One question is whether the measures are valid, especially given thinking on definitions of poverty that include access to services and security. There is also a danger of circularity here if we are partly defining the poor by whether they have access to services. In practice, the most common approach has been to use an asset index approach, which is devised within the country as this is easier to measure than income. Studies can also select particular groups to cover e.g. slum dwellers; ethnic minority.
Do surveys really reach the poorest? Surveys such as DHS or LSMS may miss out the poorest, as the sampling is based on the census and may be out of date (e.g. missing recent rural migrants). (1)
Even if there is equality in use of services, this may not reflect needs – the poor may have more ill health so to achieve equity they would need more of some services. Note this is not the case for everything, e.g. immunisation should be the same for all children.
Whilst the surveys are useful at showing what is going on, key questions remain on why - What is reason for differences? What are the barriers to more use by the poor? How to tackle them? Qualitative studies are useful to answer this.
However, let us look at the evidence there is.
Reaching the poor
Starting at global level, let us look crudely at whether health services are reaching the poor by looking at expenditure. [Table 1] These figures on spending per capita show that spending in rich countries is more than 100 times per capita what it is in low income (poor) countries.
Health Expenditure per Capita
Table 1: Health Spending per Capita.
Source: World Bank, World Development Report, 2001
How does this stark differential compare with the needs for health care? Consider the high per capita burden in lower income countries, especially in Africa. A high proportion of the burden is due to communicable disease and accidents; many of these are amenable to relatively cheap preventive and treatment measures.
If we look at global programmes on health, recent analysis by IHSD staff shows that the global initiatives are favouring the poorer countries to a greater degree than development aid in general. For example, the Global Alliance for Vaccines and Immunisation (GAVI) and Polio Eradication Initiative are allocating some 98% to low income countries, and the Global Fund is doing well with 78%. This compares with less than two thirds (some 64%) for development aid in general.
Apart from funding, are the global health programmes tackling the diseases that affect the poor? The answer is yes - the poor are likely to suffer from HIV, malaria and TB, measles and polio, so programmes to address these are relevant. However, the balance between them may not reflect their impact on the poor. In rough terms, HIV brings about twice the disease burden as malaria, yet the level of funding from global programmes is some 25 times that for malaria.
At country level, the evidence shows better health outcomes among the richer groups as well as better service uptake. See for example graph 1 that summarises infant and child mortality rate data for 21 countries in Africa. However there is considerable variation between countries in the extent of the gap. Graph 2 shows figures for selected countries, and shows that the gap between quintiles (20% of the population) varies widely.
Graph 1: Health Outcomes according to socio economic status. Sample of 21 sub Saharan African countries. Source: Demographic and Health Surveys various years
Graph 2: Child Mortality Rate Differentials for Selected Countries. Source: D. Gwatkin et al., Initial Country-Level Information About Socio-Economic Differences in Health, and Population, Volumes I and II (November 2003)
Thus health status varies – how about access to services? Graph 3 shows that access in terms of utilisation, which incorporates both financial and geographic access, rises with wealth looking across 21 countries. This is the case for two indicators – immunisation coverage and the proportion of deliveries attended by skilled health workers.
Graph 3: Access to Essential Health Services in 21 sub Saharan African Countries by Socio-Economic Group. Source: DHS various years
However the extent of the gap is variable within countries – see for example the figures (Graphs 4 and 5) for two states in India. Kerala is much more equitable in terms of access than Orissa. Kerala also shows that the gap in access between the poor and better off can be pretty small – a huge gap is not inevitable.
Graph 4: Kerala: Distribution of Institutional Deliveries in Public and Private Facilities. Source: National Council of Applied Economic Research: Benefit Incidence Studies 2001 – figures relate to 1995/6
Graph 5: Orissa: Distribution of Institutional Deliveries in Public and Private Facilities. Source: National Council of Applied Economic Research: Benefit Incidence Studies 2001 – figures relate to 1995/6
Looking at specific programmes and diseases, Filmer’s study on 22 African countries shows that while there was only a slightly higher rate of illness among the poorer groups, they were much less likely to obtain suitable treatment (2). A Tanzanian study found similarly that within a rural area, there was little difference in incidence but lower treatment uptake by the poorer families (3).
The recent studies that were presented at the Reaching the Poor conference showed that health initiatives had some success in improving the health of the poorer groups, either more, or to a similar extent as the better off (4). However they could still do better particularly at reaching the poorest 20%. One of the reviewers, Abbas Bhuiya from Bangladesh, said: “Programmes/projects seemed to have greater potential … in reducing the rich-poor gap are those targeted to the poor and/or with built in special attention to the needs of the poor”. He goes on to advocate for genuine community participation to hold health workers accountable (5).
As for the global initiatives, there is not much evidence on who is benefiting. Where coverage is limited, e.g. where the country is introducing antiretroviral treatment, the policies for who will get treatment are not often explicit, but it seems unlikely that the poor will get priority.
Looking at benefit incidence – who benefits from public spending – then table 2 indicates that there is variation between countries, but often it is the better off who get most from public services, especially hospitals. In primary health services, the poor are more likely to get a larger share of the benefits. However this needs to be viewed in context e.g. in Latin America countries where the public sector tends to serve the poor while others are covered by social security arrangements.
to lowest 20%
to richest 20%
Countries with subsidy to poorest > to richest 20%
all health spend
4 (of 21)
9 (of 21)
Table 2: PHC spend more likely to benefit the poor. Source: Gwatkin, D., Bhuiya, A., Victora, C., Making health systems more equitable
Note also that use of health services can contribute to poverty. In all income groups, people resort to borrowing or selling assets, and this can drive them into poverty.
Summary and issues
The findings demonstrate that
The poor experience worse health outcomes than the better off (this is true for most well off countries too, such as the UK where the differentials between socio-economic groups have been getting worse over time);
There is evidence that the poor do not benefit as much from private or public health services as the better off. There is little data on access to NGO services though we tend to assume they are reaching the poor.
Public spending on health tends to benefit the better off in many cases; although the extent of the difference varies widely even within a country (e.g. between Indian states). The poor are more likely to get a fair share of subsidies to primary care.
Some programmes do better than others in reaching the poor. It seems to help for them to have explicit objectives and provisions to reach the poor.
At global level, the global health initiatives are targeting diseases of the poor and lower income countries.
What can be done about the inequities? One of the leading thinkers on this is Davidson Gwatkin who used to be at the World Bank. He has argued that a focus on universal coverage appears to promote equity; however experience suggests that unless there is an explicit focus on reaching the poorer groups, then programmes will expand to reach the better off first – they are often easier to reach than the very poor.(6)
Too much focus on meeting the Millennium Development Goals (MDGs) also risks this sort of response. Gwatkin has shown that the MDGs can be reached with little impact on the poorest.
Another issue is the importance of monitoring who benefits from programmes, and understanding why there is success or not in reaching target groups, so that this can feed back into service design. This requires monitoring and studies of these factors built into our health projects and programmes, or ensuring that they are covered by broader poverty and Poverty Reduction Strategies monitoring efforts.
*Veronica Walford is the Director of the London-based Institute for Health Sector Development. She has worked on diverse aspects of health policy and planning, public sector management and reform, programme design and evaluation. This includes four years with the Uganda Ministry of Health as senior economic adviser. She has also been a regional manager for south East Asia with the DFID funded Health Systems Resource Centre, and has participated in numerous consultancies, both in the UK and internationally. Contact: firstname.lastname@example.org.
(1) For more on this, see Diamond, Matthews & Stephenson, Assessing the health of the poor: towards a pro-poor measurement strategy, DFID Health Systems Resource Centre, 2001.
(2) Filmer, D., Fever and its treatment among the more and less poor in Sub-Saharan Africa, World Bank Development Economics Research Group working paper 2789, March 2002.
(3) Armstrong Schellenberg, Victora, Mushi, de Savigny, Schellenberg, Mshinda, Bryce, Inequalities among the very poor: health care for children in rural southern Tanzania, Tanzania IMCI MCE baseline household survey study group, The Lancet, Vol 361: 561-66, February 2003.
(6) Gwatkin, D., Bhuiya, A., Victora, C., Making health systems more equitable, The Lancet, Vol 364: 1273-80. October 2, 2004.
Armstrong Schellenberg, J. et al, Inequalities among the very poor: health care for children in rural southern Tanzania, Tanzania IMCI MCE baseline household survey study group, The Lancet, Vol 361: 561-66, February 2003.
Bhuiya, A. Concluding Remarks on Conference on Reaching the Poor with Effective Health, Nutrition and Population Services: What works? What does not? And why? Washington DC, 18-20 February 2004.
Carr, D. Improving the Health of the World’s Poorest People, Health Bulletin –A Publication of the Population Reference Bureau, Number 1: 1-8, February 2004.
Diamond, I., Matthews, Z. & Stephenson R. Assessing the health of the poor: towards a pro-poor measurement strategy’, DFID Health Systems Resource Centre, 2001.
Filmer, D. Fever and its treatment among the more and less poor in Sub-Saharan Africa, World Bank Development Economics Research Group working paper 2789, March 2002.
Gwatkin, D. Who Would Gain most from Efforts to Reach the Millennium Development Goals for Health? An inquiry in to the possibility of progress that fails to reach the poor, Vol. 1, Health, Nutrition and Population Family (HNP) of World Bank’s human Development Network, December 2002
Gwatkin, D., Bhuiya A. & Victora, C. Making health systems more equitable, The Lancet, Volume 364: 1273-80. October 2, 2004.
Gwatkin, D. How well do health programmes reach the poor?, The Lancet, Vol 361: 540-41, February 4, 2003.
Independent Inquiry into Inequalities in Health Report, London: Stationery Office 1998.
Nyonator, F. Conference on Reaching the Poor with Effective Health, Nutrition and Population Services: What works? What does not? And why? comments from Frank Nyonator – Member of Conference Expert Panel’, Washington DC, February 20, 2004.
Population Services International, Reaching the Poor: PSI Provides Health to Low-Income People, July 2003.