What works for Africa's poorest?

David Hulme and David Lawson

This chapter presents both the context from which this volume evolved as well as the situational backdrop for the chapters that follow. In so doing, it asks who are the very poorest in sub-Saharan Africa (SSA), how do they survive, and establishes the precariousness of their situation in relation to the steadily, if slowly, improving development picture experienced by many more of the world's poorest. Rather than focus on macro-level conceptualisations of 'being poor' this volume is partially structured around a micro and meso-level lifecycle analysis of extreme poverty that focuses on evidence showing people can be extremely poor at different points in their life. The volume, thus, hopes to: (i) raise awareness of the problems faced by the very poorest people in SSA; (ii) encourage policy-makers and development practitioners to identify projects, programmes, and policies that can assist the very poorest, in the next few years; and (iii) provide practical examples of 'what works for SSA's poorest'.

Keywords: Sub-Saharan Africa; extreme poverty; micro and meso analysis; targeting; survival; lifecycle

Introduction

Over the past 20 years, the world has experienced an unparalleled period of economic growth and dramatic reductions in income/consumption (and multidimensional) poverty. This is a cause for optimism, but at the same time, we must note that the benefits of this contemporary growth have been very unevenly spread. The world's richest 62 people now own assets that have the same value as those held by the poorest 3.6 billion; the bottom half of humanity (Oxfam, 2016).

At the global level, the benefits have been concentrated within China, India, and Southeast Asia, while other regions – sub-Saharan Africa (SSA), much of South Asia, the Andean region, the states of the former Soviet Union, and the Pacific – have improved at much slower rates, and in some countries (including Egypt, Iraq, Somalia, Syria, and Zimbabwe) life has got harder. SSA countries, long seen as being locked in poverty traps (Collier, 2007) that kept the well-being levels of the majority of their population very low (a more accurate concept might be ill-being), have recently escaped this 'Afro-pessimism' image. Across the continent, economic growth rates have improved and human development indicators have risen. In 2015 Ethiopia, for example, was the world's fastest growing economy (World Economic Forum, 2016), with Côte d'Ivoire predicted to be the world's second fastest in 2016. Additionally, since 1990, infant mortality for SSA children under the age of five years has been reduced by 43 per cent, and the number of maternal deaths has also declined by 47 per cent (Save the Children, 2013; see Ado-Kofie and Lawson (2016) for further discussion).

However, things vary greatly from country to country and at the sub-national level. A common pattern is that many urban areas have seen dramatic improvements in their average levels of income and in human development indicators. By contrast, rural areas, and particularly remote rural areas and conflict zones, have seen little or no benefits. It is evident that the types of broadly based growth that generated the relatively egalitarian East Asian Miracle of the mid- and late-twentieth century – growth with mass poverty reduction – are different from contemporary growth processes in Africa. SSA's dependence on hydrocarbons and minerals for its recent growth, combined with the region starting from a much higher level of socioeconomic inequality, have meant that Africa's growth has been concentrated in fewer hands than in East Asia. As Kofi Annan, former Secretary-General of the United Nations, put it: 'Africa is a rich continent with a lot of poor people' (BBC World, 27 July 2015).

This extreme and chronic poverty has spatial and social relational dimensions. Spatially, it is often concentrated in particular areas, for example, mountainous regions, arid/semi-arid lands and landlocked regions across Africa, and informal 'settlements' in the major cities. Socially, it is concentrated within specific groups, such as indigenous or 'tribal' groups (in Botswana, DRC, and elsewhere), ethnic and religious minorities, internally displaced people, and refugees. At the micro level – the household, village, or community – particularly vulnerable individuals, including older people, widows, orphans, and disabled people, are likely to find that they can barely maintain their lives and that they have minimal or no prospects for improvement. At the extreme, the poorest simply disappear, dying unregistered but easily preventable deaths.

This volume arises out of the concerns of a group of development practitioners, policymakers, and researchers about the poorest. It builds on our earlier work (Lawson et al., 2010), and seeks to provide ideas and guidance about how the poorest might be assisted in their personal efforts to sustain themselves and improve their prospects. There have been significant breakthroughs in assisting the poor – from jobs in garment and shoe factories, to cash transfers, to microfinance – but there is also a growing recognition that these have rarely reached the poorest. At the global level, the new UN Sustainable Development Goals promise that 'no one will be left behind' – a grand aspiration, but at the same time, an acknowledgement that hundreds of millions have been left behind during an exceptional period of prosperity in the developing world. The book's orientation is explicitly practical and policy-focused. It does not seek to propound general theories about causes of extreme poverty or universal generalizations about how to tackle it. The focus is on the micro- and meso-level projects, programmes, and policies that seek to improve the condition of Africa's poorest people in the short and medium terms.

We realize that this leaves us open to criticisms. Neo-classical macroeconomists can argue that we have neglected the analysis of international trade, economic liberalization, and macro-economic policies. From the opposite end of the ideological spectrum, radical political economists can argue that we fail to examine the social relations engendered by contemporary global capitalism, ranging from worker exploitation to violent conflicts. We acknowledge such lacunae in this volume and refer the reader to other sources (Collier, 2007; CPRC, 2008; Greig et al., 2007; Hulme, 2015) that address such 'macro' issues. In our defence, we argue that there has been a relative neglect in the literature on how to assist the poorest in operational terms. This volume seeks to partially fill this gap, complementing and extending upon the previous volume of What Works for the Poorest (2010), and to provide practitioners and policymakers with ideas and experiences that may provide insights and guidance.

In this volume we seek to:

1. Raise awareness of the problems faced by the poorest people in SSA.

3. Encourage policymakers and development practitioners to identify projects, programmes, and policies that will assist the poorest in the next few years. The poorest cannot afford to wait for growth to trickle down or good governance to emerge: they may die or have their capabilities (e.g. physical and mental health, cognitive abilities, reproductive capacity) disabled or destroyed while 'waiting'.

4. Provide a number of practical examples from 12 countries and four regions across SSA, and draw some policy conclusions for practitioners and policymakers of 'what works for the poorest'. These may serve as 'models' for action, but our preference is for them to inspire new experiments based on a mix of inspiration and the nitty-gritty routines of planning, implementing, monitoring, learning, and strengthening projects and programmes in the field. Ultimately, we have to build networks of institutions that can support the efforts of Africa's poor people.

We are fully aware that writing and editing papers about assisting the poorest is the easiest part of such efforts. Planning, managing, and implementing these initiatives on a day-to-day basis, and running the public agencies and NGOs that take the lead in such work, are the critical ingredients, of which the world needs more.

Who are the poorest?

Defining and/or identifying 'the poorest' is challenging in terms of both concepts and methods. There are five main ways in which extreme poverty can be conceptualized. The most common are income- or consumption-based poverty lines based on an assessment of the amount of money needed to acquire food that meets the minimum calorific needs of a human being; these are used by most economists and policymakers. This is commonly estimated to be around 2,100 to 2,500 calories per day for an adult, but there are many assumptions and technical issues behind such estimations. In many developing countries the extreme poverty line is defined as a percentage of the overall poverty line.

Since 1990, a global 'extreme poverty' line has been identified: the widely used US$1/day measure (which became the $1.25/day measure in 2008 and the $1.90/day measure in 2015). This was computed by World Bank economists for the World Development Report 1990, which argued that in 1985, anyone with consumption valued at less than $1/day (in purchasing power parity (PPP) terms) was extremely poor. The most recent detailed application of this yardstick to the global population estimates that, in 2010, 1.2 billion people were extremely poor, and that 410 million (an increase from 205 million in 1981) of these were sub-Saharan Africans (World Bank, 2013). As a result, and despite a percentage decline in extreme poverty in SSA (from 57 per cent in 1990 to 43 per cent in 2012; Beegle et al., 2016), the extreme poor in SSA now represent a third of the world's extreme poor (11 per cent in 1981) (World Bank, 2013).

However, there are big questions about the arbitrariness of the $1.90/day line and about the quality of data available. A particular disadvantage of this concept is that it cannot be used in the field to practically identify the poorest people, as collecting accurate data on the income of very poor people is costly, time-consuming, and beyond the capacity (and the cost structures) of most service delivery agencies. Large numbers of development specialists (managers, policymakers, researchers) believe that assessing poverty purely in terms of income and consumption is logically flawed, and prefer the concept of 'human development'. This developed out of the work of Paul Streeten, Mahbub ul Haq, and Amartya Sen and has been popularized by the United Nations Human Development Reports. Human development views poverty and extreme poverty as multi-dimensional deprivation: lack of food and income, illiteracy, inadequate access to health services and potable water, and other factors.

For almost 20 years such analysts used the UNDP's Human Development Index (HDI), but this was of no use in examining the poverty headcount number, as it is an aggregate national measure. This shortcoming has been tackled by Sabina Alkire and colleagues at the Oxford Poverty and Human Development Initiative through the computation of a global Multi-dimensional Poverty Index (MPI). The MPI assesses levels of severe deprivation in health, education, and living standards at the individual level. It has ten weighted indicators: nutrition; child mortality; years of schooling; school attendance; access to cooking fuel; sanitation; potable water; electricity; quality of flooring; and household ownership of basic assets. If someone is deprived in more than one-third of these ten basic needs indicators, they are identified as 'MPI poor'. If they experience extreme deprivation in these indicators (for example, they have experienced the deaths of two of their children), then they are 'MPI destitute'. The global MPI measure has made great progress since its 2010 launch, and by 2014 it was covering 108 countries and 78 per cent of the world's population. According to the MPI around 460 million people in SSA are multi-dimensionally deprived and the vast majority of them, more than 85 per cent, live in rural areas.

So what difference would the use of the MPI – with its advantages of more closely mirroring the reported experiences of poor people, and its not needing PPP adjustments – make, when compared to using the $1.25/day measure? In terms of static comparisons, MPI poverty headcounts are significantly higher than $1.25/day counts in some countries, such as Côte d'Ivoire, Ethiopia, Senegal, and Uganda. Very worryingly, for Ethiopia the MPI headcount is 86 per cent against a $1.25/day headcount of 32 per cent; that is, tens of millions of 'additional' poor people. By contrast, in countries such as Nigeria and Zambia the MPI headcounts are much lower. For Nigeria, an MPI headcount of 42 per cent contrasts with a $1.25/day poverty headcount of 68 per cent; from an MPI perspective, poverty in Nigeria appears less serious than from an income perspective. The MPI and income poverty perspectives would also suggest different dynamics of poverty reduction in different countries. The MPI indicates that poverty reduction has been faster in Rwanda and Ghana than does the income poverty measure. But for Ethiopia, poverty reduction appears to have been significantly slower than the $1.25/day measure suggests. But, just like the $1.25/day measure, the MPI has experienced great criticism. This includes the many assumptions that lie behind the lists, the choice of cut-off points for deprivation, and the number of deprivations needed to be classed as multi-dimensionally poor or destitute.

A third possible means of identifying the poorest is the duration of poverty. The Chronic Poverty Research Centre (CPRC) has been spearheading such an approach (CPRC, 2004, 2008; Hulme and Shepherd, 2003) that has used the $1/day measure to estimate that over 400 million people are chronically poor: they had been in extreme poverty for five years or more. CPRC has encouraged the use of human development measures, but these have only been taken up to a very limited degree (Gunter and Klasen, 2009). Interestingly, Gunter and Klasen's (2008) work suggests that the correlation between chronic poverty in the income and non-income dimensions is very low, and non-income poverty is certainly more stable over time than income poverty. Calvo and Dercon (2009) have argued that analytically and practically the challenge is to identify who is likely to remain poor in the future. These are the people whom the policy must try to prioritize.

A particularly important component of CPRC's work has focused on the intergenerational transmission of poverty. Arguably, those households in which the poverty of the parents is likely to be transferred to children – through the blocking of their capabilities by poor nutrition, ill health, lack of education, and other factors – can be regarded as among the poorest.

The fourth means of identifying the poorest is more intuitive, and is an approach used by organizations and practitioners at the field level. It is to identify an indicator that can be easily assessed and is believed to reveal that an individual or a household is experiencing extreme poverty. The most common such indicator is access to food: 'have you gone without food this year?' or 'how many meals a day do you eat?' Kabeer (2010) found that the simple question, 'what category best describes the food situation in your household over the last year: chronic food shortage, occasional food shortage, 'breakeven' food supply, or food surplus?' could identify the consumption poor with great accuracy. Combinations of indicators can be used, as shown by Sen and Begum (2010). Data on housing quality, land holding, and occupation can be used to rapidly and accurately identify the poorest people in Bangladesh.

The fifth approach is participatory. This includes a variety of different methods ranging from the widely used Participatory Rural Appraisal techniques described by Robert Chambers (1994) to the more Freirian and dialogical methods proposed by Xavier Godinot (2000). This approach has received much attention in recent years with Narayan et al.'s (2000) Voices of the Poor and Krishna's (2009) 'Stages of Progress' methodology. The BRAC Targeting the Ultra-Poor Programme analyzed by Hulme and Moore (2010) and its many 'replicas' use the classic PRA technique of wealth ranking to initially identify the poorest people in rural Bangladesh before proceeding to use more 'objective' methods.

The volume is partially structured around a lifecycle analysis in recognition of the evidence that people can be extremely poor at different stages in their life, that extreme poverty early on in the lifecycle raises the probability of experiencing poverty at later stages and/or for a whole lifetime, and that strategies to tackle extreme poverty should incorporate a lifecycle analysis to be effective. Evidence continues to reveal the ways in which 'the first 1,000 days' of a child's life – from inception to the age of two – profoundly affect lifetime capabilities and achievements. The extreme poverty of many people commences in utero, when a mother (often experiencing poverty) is unable to gain access to an adequate diet and basic medical care. Inadequate access to nutrients, including micro-nutrients such as iodides, can limit the physical and cognitive ability of people in later life, and may predispose them to higher levels of health problems as adults. Mothers who do not receive basic maternal health screening – in some countries defined as four check-ups during pregnancy – are more likely to deliver babies with medical problems that could have been avoided, miss an opportunity for learning about caring for infants, and face higher levels of personal health risk during their pregnancy (especially preeclampsia).