Jump to ContentJump to Main Navigation
Child poverty, evidence and policyMainstreaming children in international development$

Nicola A. Jones

Print publication date: 2011

Print ISBN-13: 9781847424464

Published to Policy Press Scholarship Online: March 2012

DOI: 10.1332/policypress/9781847424464.001.0001

Show Summary Details
Page of

PRINTED FROM POLICY PRESS SCHOLARSHIP ONLINE (www.policypress.universitypressscholarship.com). (c) Copyright Policy Press, 2021. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in PPSO for personal use.date: 25 July 2021

Child poverty, knowledge and policy in Africa

Child poverty, knowledge and policy in Africa

(p.88) (p.89) Four Child poverty, knowledge and policy in Africa
Child poverty, evidence and policy

Nicola Jones

Andy Sumner

Policy Press

Abstract and Keywords

This chapter deals with children and the knowledge–policy interface in sub-Saharan Africa. It briefly outlines the extent and nature of child poverty and well-being across Africa using the 3D approach. The chapter reflects on the characteristics of the knowledge-generation process in Africa and discusses the knowledge–policy interface surrounding child well-being in Africa. It also focuses on a case study of evidence-informed policy change in the context of an expert-led initiative to promote a more child-sensitive PRSP during the revision process of Ethiopia's second-generation PRSP.

Keywords:   children, knowledge–policy interface, sub-Saharan Africa, child poverty, 3D approach, knowledge-generation process, child-sensitive PRSP, second-generation PRSP

4.1 Introduction

This chapter is about children and the knowledge–policy interface in sub-Saharan Africa, and is structured as follows: Section 4.1 briefly outlines the extent and nature of child poverty and well-being across Africa using the 3D approach. Section 4.2 reflects on the characteristics of the knowledge-generation process in Africa. Section 4.3 discusses the knowledge–policy interface surrounding child well-being in Africa. Section 4.4 focuses on a case study of evidence-informed policy change in the context of an expert-led initiative to promote a more child-sensitive PRSP during the revision process of Ethiopia’s second-generation PRSP1 and Section 4.5 concludes.

4.2 Child poverty and well-being in Africa

In this section we provide an overview of the extent and nature of child poverty and well-being in Africa across material, relational and subjective dimensions.

Material child well-being

To recap Chapter 1, the material dimension of child well-being concerns practical welfare and standards of living and the objectively observable outcomes that children and adults are able to achieve, for example, income, wealth and assets; employment and livelihood activities; education and skills; physical health and (dis)ability; access to services and amenities; and environmental quality. A brief overview of child nutrition, child education and child health using the MDGs as a barometer reveals an uneven picture of children’s material well-being in Africa.

The child-relevant MDGs – 1 (underweight children) and 4 (under-five mortality) – are unlikely to be met in sub-Saharan Africa. The proportion of underweight children at birth is currently 28% in sub-Saharan Africa, and is expected to fall to 26% in 2015 compared (p.90)

Child poverty, knowledge and policy in Africa

Figure 4.1: MDG 1 – underweight children in sub-Saharan Africa

Source: UNDESA (2009).

Child poverty, knowledge and policy in Africa

Figure 4.2: MDG 2 – net primary enrolment in sub-Saharan Africa

Source: UNDESA (2009).

to the 15% target. Under-five mortality in the region is not on track either, currently standing at 145 per 1, 000 live births and likely to fall to 123 in 2015 compared to the target of 61. In contrast, child MDG 2 (primary education) is more progressed at 74% and estimated to reach (p.91)
Child poverty, knowledge and policy in Africa

Figure 4.3: MDG 4 – under-five mortality in sub-Saharan Africa

Source: UNDESA (2009).

Table 4.1: Children in sub-Saharan Africa: selected material well-being indicatorsm




2015 estimate on current trend

MDG target

Underweight children (%)





Net primary enrolment (%)





Under-five mortality (per 1, 000 live births)





Source: UNDESA (2009).

91% in 2015 compared to the 100% target (reflecting the policy focus given to it by many governments and donors).

The above picture has been further complicated by the Lancet review of MDG 4 under-five mortality data (You et al, 2009).2 Changes made following recent re-estimates (see Table 4.2) include substantial changes in child mortality levels and trends for some countries; adjusting estimates for countries with a high HIV prevalence where there was a reporting bias associated with maternal health; as well as the inclusion of new methods of estimating absolute numbers of deaths. You et al (2009) conclude that although under-five mortality in sub-Saharan Africa has declined by 22% since 1990, the rate of improvement in child survival is still insufficient to reach MDG 4.


Table 4.2: Africa: Levels and trends in MDG 4 – under-five mortality, 1990–2008 (mortality rate per 1, 000 live births)





Decrease 1990–2008 (%)

Average annual rate of reduction 1990–2008 (%)








Sub-Saharan Africa







Eastern and Southern Africa







West and Central Africa







Middle East and North Africa







Source: You et al (2009:1–2).

Relational child well-being

To recap Chapter 1, the relational dimension of child well-being concerns the extent to which children and adults are able to engage with others in order to achieve their particular needs and goals. This includes, for example, relations of love and care (networks of support and obligation) and relations with the state (law, politics, welfare, social, political and cultural identities and inequalities; violence, conflict and (in)security; and scope for personal and collective action and influence).

Child poverty, knowledge and policy in Africa

Figure 4.4: MDG 3 – gender equality in primary education in sub-Saharan Africa

Source: UNDESA (2009).

Child poverty, knowledge and policy in Africa

Figure 4.5: MDG 5 – maternal mortality ratio in sub-Saharan Africa

Source: UNDESA (2009).

Such relational well-being data is difficult to find for Africa, Asia and Latin America, but there are growing data-collection efforts.

First, relational well-being indicators are evident in a crude sense in the MDG indicators that relate to children and inequality such as gender (in)equality in education, inequality in under-five mortality and maternal mortality.3 A quick review of these, again, suggests a very mixed picture in Africa.

At 90%, MDG 3 (gender equality in education) is almost on track in sub-Saharan Africa, and is estimated to reach 96% in 2015 compared with the target of 100%. However, at 900, MDG 5 (maternal mortality) is very badly off track, and is estimated to reach 887 in 2015 compared to a target of 230.

Hogan et al (2010) recently re-estimated maternal mortality data in The Lancet using all available data for maternal mortality from 1980 to

Table 4.3: Children in sub-Saharan Africa: selected relational well-being indicators




2015 estimate on current trend

MDG target

Gender equality in education (%)





Maternal mortality (per 100, 000 live births)





Source: UNDESA (2009).

(p.94) 2008 for 181 countries (from registration data, censuses, surveys and verbal autopsy studies – see Table 4.4). Hogan et al argue that in the absence of HIV, progress in reducing sub-Saharan Africa’s maternal mortality ratio (MMR) would have been much more extensive than was recorded.4 Their estimates are lower than the UN Department for Economic and Social Affairs (UNDESA) but show MMR increasing in Southern, East and West sub-Saharan Africa in the 1990s.

An important part of relational well-being is inequality. The poorest children are considerably poorer than the average, and trends are not good for the poorest children. If we take the under-five mortality Demographic and Health Survey (DHS) country average (not even the richest 20%), there is often a difference of approximately 20–30% between the poorest and the average for a country. Furthermore, there are few countries where the trend is towards convergence.

We can also examine data on the poorest children and under-five mortality rates (see Tables 4.5 and 4.6). The World Bank’s DHS provides

Table 4.4: MMR per 100, 000 live births by region, sub-Saharan Africa





North Africa/Middle East





Sub-Saharan Africa, Central





Sub-Saharan Africa, East





Sub-Saharan Africa, Southern





Sub-Saharan Africa, West





Source: Hogan et al (2010)

Table 4.5: Under-five mortality rates: average versus poorest quintiles in selected sub-Saharan African countries

Average 2000–05 (year)

Poorest 20% 2000–05 (year)

Poorest as % of average

Benin (2001)




Burkina Faso (2003)




Ethiopia (2005)




Ghana (2003)




Kenya (2003)





156 (2004)

184 (2004)



233 (2001)

248 (2001)



172 (2003)

196 (2003)



130 (2004)

137 (2004)



146 (2006)

172 (2006)


Source: Vandemoortele and Delamonica (2010), based on DHS.


Table 4.6: Under-five mortality rates: trend data of average versus poorest quintiles in selected sub-Saharan African countries


Poorest 20%

1995–2000 (year)

2000–05 (year)

1995–2000 (year)

2000–05 (year)

Benin (1996–2001)





Burkina Faso (1998/99–2003)





Ethiopia (2000–05)





Ghana (1998–2003)





Kenya (1998–2003)





Malawi (2000–04)





Mali (1995/96–2001)





Mozambique (1997–2003)





Tanzania (1999–2004)





Uganda (2000/01–2006)





Source: Vandemoortele and Delamonica (2010), based on DHS.

strong and reliable data for the poorest 20% of the population because they have data-by-wealth quintiles (i.e. fifths of the population) for numerous countries.5 The poorest 20% are generally worse off but the degree varies considerably across countries: Ethiopia, Tanzania and Mali have smaller differences; Benin, Ghana, Kenya, Malawi, Mozambique and Uganda have differences in the range of 20–30% or more.

The UNICEF-supported MICS data set contains a limited number of indicators that relate to children’s relational well-being, albeit for a limited number of sub-Saharan African countries. The indicators include birth registration; the prevalence of orphans, children with inadequate care and vulnerable children; whether young women have comprehensive knowledge about HIV prevention and transmission; rates of marriage for young adolescents and all adolescents; and the rate of female genital mutilation/cutting (FGM/C).6 In Africa, as Table 4.7 below illustrates, a variety of patterns emerge.

While the percentage of births registered varies widely, from a low of 3% in Somalia to a high of nearly 80% in Togo, overall only one in two African children is registered at birth. Children remain highly vulnerable, regardless of whether they are orphans. In Togo, nearly one in three children is without adequate care, while nearly one fifth of children are orphans in Burundi and are judged to be vulnerable in Sierra Leone. Regionally, approximately 10% of all children have lost their parents.


Table 4.7: Selected sub-Saharan African countries and relational well-being data in UNICEF MICS

Marriage before age 15 among women (%)

Marriage before age 18 among women (%)

Preva-lence of orphans (%)

Children left with inadequate care (%)

Prevalence of FGM/C (%)

Prevalence of vulnerable children (%)

Burkina Faso (2006)







Burundi (2005)







Gambia (2005–06)







Guinea Bissau (2006)







Malawi (2006)







Sierra Leone (2005)







Somalia (2006)







Togo (2006)







Source: UNICEF MICS3 downloaded from www.micscompiler.org.MICS.html.

The incidence of adolescent marriage, while dropping, remains alarmingly high in some countries. In Sierra Leone, for example, over one quarter of all girls are married before the age of 15 and nearly two thirds by the age of 18. In Burundi, on the other hand, marriage in early adolescence is extremely rare, at less than 3%, and only 20% of women had married before the age of 18. Overall, approximately one in two African women married as an adolescent. The prevalence of FGM/C also shows extreme variation, from nearly 100% in Sierra Leone and Somalia to less than 5% in Togo. Burkina Faso and Gambia also had high rates of FGM.

Children and subjective well-being

To recap Chapter 1, the subjective dimension of child well-being concerns meanings that children and adults give to the goals they achieve and the processes in which they engage. For example, understandings of the sacred and the moral order, self-concept and personality, hopes, fears and aspirations, sense of meaning/ meaninglessness, levels of (dis) satisfaction, and trust and confidence. Such subjective well-being data are difficult to find for Africa, Asia and Latin America. However, a relatively small number of studies have asked children in Africa about their perceptions of poverty and well-being, as shown in Table 4.8.

Overall, these studies suggest that children view poverty not only in terms of material deprivation, but also in terms of social marginalisation, insecurity, vulnerability and distress. Bethlehem et al (2009), for instance, (p.97)

Table 4.8: Children in sub-Saharan Africa: selected subjective well-being studies of children’s perceptions of poverty and well-being

Country and reference


Ethiopia Bethlehem et al (2009)

This article draws on a small field study to discuss the advantages and disadvantages of using qualitative methods to ascertain poor children’s perspectives on their own well-being. The author used diaries, drawings, and interviews to identify what children saw as threats to and positive influences on themselves. For example, children reported that it made them happy to play football or jump rope with their friends and to be in a clean environment. They reported that teachers who hit, being sent on errands at night, and the school toilet were things that they did not like.

South Africa Bonn et al (1999)

This interview-based study of rural and urban children found that children’s environment was instrumental in forming their beliefs about poverty and unemployment. Children with regular exposure to unemployment, for example, believed it to be linked to poverty. Age and developmental stage, however, were more important than environment in terms of children processing causality.

Ethiopia Camfield and Tafere (2009)

Using qualitative data from Young Lives, this study explores what children perceive to be ‘the good life’ – and how to get it. The study also addresses how these perceptions vary by the child’s place in social relationships.

Ethiopia Camfield (2010)

Using mixed-methods data from Young Lives, the author found that even young children were able to identify poverty indicators. Personal appearance, clothing, education, food, and housing were salient markers of poverty. Children reported shame and stigma as the result of poverty.

Ethiopia, Peru, Vietnam and India Crivello et al (2009)

This article reviews Young Lives’ work on developing childfocused,participatory, qualitative methods that capture what children understand vis-à-vis their own well-being and how that understanding changes over time. These methods, which included timelines and body mapping, allow research to move beyond simple quantitative measures of child poverty. They also showed that, despite the fact that family well-being was crucial to child well-being, children had unique perspectives that needed to stand on their own. Children, for example, wanted playtime, which adults rarely saw as important. They also wanted to focus on their education, which adults often considered secondary to work. Furthermore, children simply saw the world through a different lens. For example, in India adults saw sickness and dirty appearance as two key indicators of child ill-being. Children, on the other hand, chose ‘plays in drainage ditch’ and ‘kills birds’ as their top indicators.

Uganda Whitter and Bukokhe (2004)

This field study asked Ugandan children about their experiences of poverty and their opinions about the government’s policies addressing it. It found that adults and children have different perspectives on poverty, that children see a positive role for themselves in terms of mitigating poverty and that they are highly critical of the governmental response.

(p.98) found that poor Ethiopian children wanted more playtime and to be able to concentrate more on their studies. Camfield (2010) reports that even young children in that country are aware of the stigma of poverty and feel shame over their appearance and housing. While these poverty indicators are familiar to us as adults, Crivello et al (2009), drawing on the Young Lives study, found that poor children often have other, less obvious, ways of understanding poverty. For example, they saw playing in drainage ditches and killing birds as indicators of child ill-being. Bonn et al (1999) found that as children age they are more able to understand what causes poverty. When they are young they have only their own experience to draw upon, meaning that, for example, only poor South African children who have experience with parental unemployment necessarily see unemployment as a root cause of poverty. Whitter and Bukokhe (2004) found that children in Uganda were highly critical of the government’s response to poverty, though they saw a positive role for themselves.

4.3 Knowledge generation and child well-being in Africa

As discussed in Chapter 2, the generation of research-based knowledge on child poverty and well-being is a relatively new phenomenon, especially in developing countries. As Wells (2009: 9) notes in her discussion of the history of childhood, ‘what we know of children’s experiences and society’s concepts of childhood in the history of Africa is very limited … . There are no general surveys that form part of a coherent narrative of children’s world in Africa as there is for North American and European history’. This limited knowledge base persists today, in part because, with the partial exception of South Africa, there are few institutions or research programmes focusing on the study of child poverty and well-being in African contexts. As Table 4.9 highlights, the African Child Policy Forum (ACPF) is the only pan-African initiative to monitor and evaluate progress on children’s rights to well-being. A small number of national programmes on various parts of the continent focus predominantly on child rights, education and social policy reforms.

Given the high level of international aid to the African region, it could be expected that this research gap would be partially filled by donor agencies, especially as growing attention is being accorded to the linkages between knowledge and development and the importance of participation in the ‘knowledge economy’ (World Bank, 2007).

(p.99) (p.100)

Table 4.9: Selected institutions with a research focus on child poverty and well-being in Africa


Home location

Affiliations and partnerships

Thematic focus

African Child Policy Forum (ACPF)

Addis Ababa, Ethiopia

  • Supports projects related to children across Africa

  • Runs the African Child Observatory and publishes the African Report on Child Well-being

Established in 2003, the ACPF is an independent, not-for-profit, pan-African institution of policy research and dialogue on the African child. ACPF was established with the conviction that putting children first on the public agenda is fundamental for the realisation of their rights and well-being and for bringing about lasting social and economic progress in Africa. It has also developed a child friendliness index against which it ranks governments in the region. http://www.africanchildforum.org/site/

Associates for Change

Accra, Ghana

Research Consortium on Educational Outcomes and Poverty (RECOUP) partner in Ghana

Associates for Change is a research and consulting firm providing social science and policy advice in Africa. http://www.associatesforchange.org/

Department of Educational Psychology, Kenyatta University


Childwatch partner RECOUP partner

Kenyatta University takes pride in the fact that it is home to some of the world’s top scholars, researchers and educationalists. The university is especially renowned for the high standards of its education department, its flagship department since inception. http://www.ku.ac.ke/schools/education/index.php/departments/educationalpsychology.html

Child Rights Information and Documentation Centre (CRIDOC)

Lilongwe, Malawi

CRIDOC’s objectives are to initiate research on child rights and be a singlepoint information resource on issues related to children and their rights. Areas of expertise include child labour, education, health, violence and gender. http://www.cridoc.net

Council for the Development of Social Science Research in Africa (CODESRIA)


Childwatch partner working throughout Africa

CODESRIA was established in 1973 as an independent pan-African organisation focused on the social sciences, broadly defined. It is recognised not only as the pioneer African social research organisation, but also as a leading nongovernmental centre of social knowledge production. Many of the council’s projects focus on children and their needs and rights. In 2006, they organised a colloquium to discuss research capacity in Africa, especially for the promotion of child rights. It called for the development of a Childwatch International Research Network in sub-Saharan Africa as a way to close the gap between research and policy. http://codesria.net/

Child and Youth Research and Training Programme, University of the Western Cape

Cape Town, South Africa

Childwatch partner

The Child and Youth Research and Training Centre operates out of the Faculty of Community and Health Sciences. The centre’s primary research areas are aligned with strategic social policy concerns for children and youth within the transformation and development agenda of post-apartheid South Africa.

The Children’s Institute

Cape Town, South Africa

Childwatch partner; publishes the South African Child Gauge and manages the Children Count website (http://www.childrencount.ci.org.za/)

The Children’s Institute aims to bring research to bear on the development of laws, policies, programmes, and service interventions for children across a number of disciplines. Key focus areas are in line with major issues that impact on children’s lives significantly, namely poverty and HIV/AIDS. These are addressed within a rights-based framework underpinned by the UNCRC and the Bill of Rights in the South African Constitution. http://www.ci.org.za/

National Research Foundation Unit for Child Development

Pietermarizburg, South Africa

Housed at the University of KwaZulu Natal

The unit works to build scientific knowledge relevant to children in South Africa. Its aim is to produce practical benefits for disadvantaged and underserved children around the country. http://www.psychology.unp.ac.za/nrfunit.htm

Institute for Childhood Education

Soweto, South Africa

Housed at the University of Johannesburg

Launched in 2010, the institute is dedicated to finding out what promotes or hinders the attainment of educational skills in young grade school children. http://www.uj.ac.za/en/faculties/edu/centresandinstitutes/ujice/pages/home.aspx

Children’s Rights Project

Western Cape, South Africa

Based at the University of the Western Cape

The project researches children’s rights and works towards their recognition and protection. http://www.communitylawcentre.org.za/clc-projects/childrens-rights/

However, as Table 4.10 highlights, while most European and North American governmental donor agencies are heavily involved in development initiatives on the African continent, investment in child well-being-related research is limited. Partial exceptions include several large, multi-year, DFID-funded projects – the Young Lives project on childhood poverty includes Ethiopia as one of four focus countries globally; the CHIP initiative on children and chronic poverty, which carried out research in Tanzania; and the Consortium for Research on Educational Access, Transitions and Equity (CREATE), which (p.101) (p.102)

Table 4.10: International donor agencies and investment in research on child poverty and well-being

Name of agency

Investment in children’s research over time

Focus of children’s research

Geographical focus

Austrian Development Cooperation (ADC)

The systematic protection of the rights of the child is explicitly stated in the Austrian Development Cooperation (ADC) Act, and is reiterated in Austria’s 2006–08 Three-year Development Policy Programme.

While the collection of child-sensitive information, its analysis and assessment of the consequences is one of the six key principles that underpin work for and with children in the ADC guidelines, evidence suggests that it is a weak cornerstone. Child rights issues may be included on a case-by-case basis if relevant or requested by the Austrian government or an NGO, but there is no broader systematic approach to the generation and management of knowledge about children’s rights.

Programmes and projects are largely located in Nicaragua, Bhutan, Cape Verde, Burkina Faso, Uganda, Ethiopia, Mozambique, the Palestinian Territories, Albania, Bosnia and Herzegovina, Macedonia, Montenegro, Kosovo, Serbia, and Moldova.

Directorate-General for Development Cooperation and Belgian Technical Cooperation (BTC)

Child rights became a cross-cutting theme for the ministry in 2005, and in 2008 a formal document was issued conceptualising child rights.

At present, child rights issues do not appear to be the subject of the ministry’s research agenda. No evaluations have been carried out on programmes that target children, hence there is no knowledge management system based around these issues.

Since 2003, Belgium has concentrated its governmental cooperation on 18 countries: Algeria, Benin, Bolivia, Burundi, the Democratic Republic of Congo (DRC), Ecuador, Mali, Morocco, Mozambique, Niger, the Palestinian Territories, Peru, Rwanda, Senegal, South Africa, Tanzania, Uganda and Vietnam.

Canadian International Development Agency (CIDA)

CIDA had a strong research team from 2001 to 2007, but today its research capacity is precarious. There has been little substantive effort to reinvigorate research into child protection and rights, and key findings have not been effectively disseminated throughout the agency.

The research team historically focused on child protection and rights, but since 2007 this work has been largely discontinued.

Policies and services specifically directed towards child wellbeing and child rights outcomes were visible in plans for Colombia, Haiti, Mali, Burkina Faso, and Côte d’Ivoire. CIDA also works in Bolivia, Honduras, Peru, Afghanistan, Bangladesh, Indonesia, Pakistan, Vietnam, Ethiopia, Ghana, Mozambique, Senegal, Sudan, and Tanzania.

German Development Cooperation (GTZ)

A specific focus on how children’s rights are unique compared to those of other social groups is largely missing from the 2008–10 Human Rights Action Plan.

There is no specific research strategy on children and key informants within the Federal Ministry for Economic Cooperation and Development (BMZ) and the GTZ. Instead, children and youth remain seen as ‘just another target group’ and a ‘soft issue’. Nonetheless, considerable priority has been accorded to documentation of good practice in child and youth rights programming and practice.

GTZ works worldwide, serving 16 countries in Asia, 41 countries in sub-Saharan Africa, and 19 countries in Latin America.

Swedish International Development Cooperation Agency (Sida) and Ministry for Foreign Affairs

As early as 1998, Sida produced a handbook on integrating child rights into diverse policy and programming areas. Swedish development assistance has an explicit commitment to child rights, as laid out in a 2001 Government Communication on the Rights of the Child as a Perspective in Development Cooperation.

The child rights team does not manage a portfolio of childfocused research or policy analysis, and was unaware of country offices undertaking similar work. The team has limited contact with the Sida Research Department. Furthermore, with the exception of a 2005 report, no systematic lesson learning initiative has been undertaken on child rights programming.

Sida works worldwide, serving 19 countries in sub-Saharan Africa, 10 countries in Asia, and four countries in Latin America.

UK Department for International Development (DFID)

In 2007, a substantial (30-day) review of DFID policy and practice on child rights was commissioned by NGOs and paid for by DFID (Maguire, 2007). DFID recognised its programming was not especially oriented to child rights, but in subscribing to the research it was responding to both an external demand for clarity and an internal desire to envision a sharper focus.

The organisational imperative to address children’s well-being is thus oriented to the delivery of the MDGs, with children’s well-being achieved as an indirect outcome of good governance, economic growth, trade, and access to services. While DFID has a strong research programme, children, rights and child rights are not specifically mentioned as core themes in its 2008–13 research strategy. However, DFID has funded research on childhood and rights, in particular Young Lives, CHIP and CREATE. Childrelated DFID research outputs include projects on HIV/AIDS and malaria. DFID also co-funds with Nike the Girl Hub.

DFID works worldwide, with programmes in five countries in Latin America and the Caribbean, 23 countries in sub-Saharan Africa, and 14 countries in Asia (DFID’s bilateral and multilateral aid programmes are under review as of 2010).

European Union (EU)

A Special Place for Children in EU External Action (2008 communication) represents a significant EU policy document on children. It sets out EU policy for incorporating children into its external policies – its political dialogue, development, and humanitarian programming.

The EU’s research agenda includes children, but there is no clear focus. The first edition of the European Report on Development, a joint initiative between a variety of European research centres and several EU member states, was published in 2009. A report on mid-term progress towards the MDGs, many of which are focused on child-related indicators, was launched in 2008.

The Directorate-General for Development is in charge of policy and relations with African, Caribbean and Pacific (ACP) countries. The Crisis Section of the Directorate-General Relex works across Asia and Africa.

Source: Harper et al (forthcoming).

has partner institutions in Ghana and South Africa – and German Development Cooperation’s (GTZ) documentation of good practices for working with children and youth.

In 2007, the Overseas Development Institute (ODI) undertook a DFID-funded review of the top-20 international research donors. On the basis of existing but limited data on research funding, this study (p.103) found that issues related to rights and social justice, including on children and youth-related issues, tended to be priorities for smaller bilateral donors (especially Scandinavia, but also Germany) and some private foundations (Ford, the Rockefeller Centre, the Open Society Institute and the Leverhulme Trust) (Jones and Young, 2007).

On a more positive note, however, UNICEF and the World Bank are increasingly carrying out important research into childhood poverty and well-being in Africa. UNICEF country and subregional offices on the African continent have partnered with a range of research institutions (e.g. Institute for Development Studies [IDS], Maastricht University [ODI], Oxford Policy Management [OPM] and the University of Bristol’s Townsend Centre) to undertake research on children and PRSPs, children and social protection, children and public finance management, and multidimensional poverty and deprivation. In addition, there are two major international data-collection initiatives funded and implemented by UNICEF in partnership with governmental statistical offices and local research institutes: the MICS and the Global Childhood Poverty Study, both of which include a significant number of African country cases. As we note in more detail later, many of these research initiatives have led to major policy advocacy and influencing efforts at the national and regional levels. The World Bank has also developed its own website on children and youth, providing a range of publications on children and young people, especially related to early childhood education, child nutrition and youth employment and vocational training.7

Finally, there is a growing, albeit still fledgling, body of research about children covering a small subset of African countries that involves children’s participation (see Table 4.11).8 This work highlights the necessity of considering context-specific dynamics such as age hierarchies, social stigma and the location of research interactions in order to effectively involve children in research about sensitive issues in their lives and to employ innovative methodological approaches to recognise children’s voices. It also underscores new perspectives that children may bring to development initiatives – for instance, children pointed to unrecognised hazards in a transport planning initiative in Ghana (Porter and Albane, 2008), prioritised a school over a planned youth centre in a refugee camp (Guyot, 2007) and highlighted school violence and inadequate toilets as key education-related concerns (Bethlehem et al, 2009).

(p.104) (p.105)

Table 4.11: Selected research in peer-reviewed journals about participatory research with children in Africa




Key findings

Abebe (2009)

Children’s participation


This article uses two field studies with disadvantaged children in Ethiopia to discuss the methods and ethics of participatory research with children. It argues that notions of research ethics formed in the ‘global North’ are difficult to apply in other contexts. For example, parental consent is difficult when a child has no clear guardian; privacy is impossible in the context of communal living spaces in which children occupy subordinate positions (adults often interrupted to clarify children’s positions); and children often mentioned involvement in dangerous activities, such as selling or using drugs, which raised confidentiality issues. The article concludes that it is vital for researchers to recognise the complexity of children’s lives.

Bethlehem et al (2009)

Use of qualitative methods in research with children


This article draws on a small field study to discuss the advantages and disadvantages of using qualitative methods to ascertain poor children’s perspectives on their own well-being. The author used diaries, drawings, and interviews to identify what children saw as threats to and positive influences on themselves. For example, children reported that it made them happy to play football or jump rope with their friends and to be in a clean environment. They reported teachers who hit, being sent on errands at night, and the school toilet as things that they did not like. While all methods produced valuable data, the cultural context of oral tradition led to some challenges.

Camfield and Tafere (2009)

Children’s perceptions of poverty


Using qualitative data from Young Lives, this study explores what children perceive to be ‘the good life’ – and how to get it. The study also addresses how these perceptions vary based on the child’s place in social relationships.

Clacherty and Donald (2007)

Ethical considerations of research with children


This article highlights the ethical challenges that underlie work with children, all of which fundamentally stem from the unequal positions of power occupied by children and adults. Care needs to be taken in understanding context for participation – for example, in Africa strong cultural norms of obedience to adults often preclude children speaking their minds. Furthermore, children need to be treated as experts on their lives and given the space to explain their own data. Finally, care needs to be taken around the issues of consent and confidentiality. This is particularly tricky in the context of extremely rural homes, often migratory parents, and the pervasive stigma of HIV. For example, researchers should take pains to ensure that neither they nor their vehicles are in any way marked with symbols or words that would cause neighbours to think that HIV is a research focus. Furthermore, allowing children to tell stories without naming names (using silhouette cut outs for example) allows them the distance they need to talk.

Cooper (2007)

Refugee youth participation


This article presents a case study in which participatory action research methods were used with long-term youth refugees in Kenya. It found that, despite the top-down nature of the setting, the project engendered new capacity and psychological growth in the teens. They reported learning new skills, such as listening without leading, and finding new areas of expertise in their own lives – one young woman became a spokesperson against forced marriage.

Guyot (2007)

Refugee youth participation


This article proposes asking child refugees what they think they need. Covering literature from several disciplines, the author tells of one humanitarian organisation that came to build a youth centre in a camp. Upon hearing that a school was actually preferred, they were unable to shift set and accommodate the children’s wishes. The author speculates that in refugee situations, where traditions have been fractured and customs abandoned, it may be easier to work with the idea of children’s voice.

Keenan (2007)

Children’s participation


This study used participatory methods to ascertain whether youth participation in urban agriculture projects contributed to their own development and that of their communities. The youths in question were former street children who were now living in centres. It found that the children were proud of their work and felt connected through it to their rural home communities. They felt the skills they learned were vital to their futures – both in terms of farming and speaking for themselves.

Naker (2007)

Children’s participation


Drawing on a field study in Uganda on violence against children, the author discusses complications inherent in representing children’s perspectives. These include raising children’s expectations and a need to account for the heterogeneous nature of children’s experiences, which vary by age, gender, and school enrolment, for example. Care needs to be taken to allow time for trust to develop. Location may also have different meanings for children. For example, the researcher found that meeting with children in a school building was not nearly as productive as meeting with them at a playground. Moving beyond rhetoric to genuine inclusiveness requires a fundamental power shift that may threaten cultural norms, which in Africa often make it difficult for even practitioners to release power to children.

Porter and Albane (2008)

Child-led research


This article discusses the differences between child-focused and child-centred research, and lays out the evolution of, and concerns inherent to, truly child-led research. Children have never been included in transport planning before, and this pilot demonstrated that they were capable contributors bringing new perspectives to the project. For example, children saw open drains beside roads, which they could be forced into when cars attempted to pass one another, as a significant risk. Lack of street lights, danger crossing the road, and molestation by taxi drivers were all seen as issues by children but not adults. The author notes that a risk that must be addressed is the possibility that addressing children as agents could increase conflict by undermining local customs.

Robson et al (2009)

Children’s participation


This article discusses ethical issues surrounding a child-led transport study. The children’s research included many spaces that made them afraid, such as graveyards, bushy places, and bridge underpasses. They were also afraid of dangerous animals and people. The article concludes that while children’s perspectives are vitally important and often different from those of adults, it is also important to understand that child researchers do not necessarily represent all children. Differences in age, class, gender, education and religion make the complexity of childhood difficult to understand – even for children.

Twum-Danso (2009)

Cultural impact on the ethics of child participation


This discussion of cultural impact on children’s participation raises questions about the balance of power between children and adults, informed consent, and volunteerism. Children, for example, are used to saying what they believe adults want to hear, regardless of whether it is what they believe. Adults, on the other hand, even those who work with children, are used to believing that children either do not know what they need or that they should not ask for it. Balancing the cultural realities of ‘Southern’ children’s lives with the ‘North’s need to ensure methodological and ethical correctness is difficult.

Young and Barrett (2001)

Children’s participation in research


This review of four visual methods of encouraging children’s participation found that all had positive effects. Photographs, drawings, maps, and daily timelines all allowed for a multiplicity of life experiences and permitted children to be active researchers of their own lives. It was found, for example, that older children used more ‘space’ in the city than did younger children; that leisure, work, eating, and stealing were important daily activities; and that survival is always at the front of street children’s minds.


4.4 Knowledge–policy interactions in Africa

As discussed in Chapter 3, the ways in which knowledge can influence the policy process are complex and context-dependent. Sub-Saharan Africa is obviously a very diverse continent with considerable variation in terms of types and quality of governance, the nature of state–civil society interactions, levels of economic development, social policy regimes and so on. By the same token, the opportunities and challenges for engaging in evidence-informed policy influencing in the region are also quite distinct from those in the North, as well as in other developing world regions (see also Chapters 5 and 6). Therefore, this section provides a brief overview of some of the key contours that shape the knowledge–policy interface in the sub-Saharan African region.

First, it is important to consider the nature of interaction between the state and civil society – is it one of opposition and confrontation, critical engagement or co-optation? Is there an open exchange of knowledge or is access to knowledge and its expression restricted? The existence of modern civil society in Africa can be traced back to political movements that rallied against the colonial powers for independence, but it is largely in the past 20 years that CSOs have been able to participate openly in political and development processes. The transition of many African countries to multiparty systems in the 1990s meant that CSOs were afforded a larger platform and accorded more legitimacy by ruling governments (e.g. Makumbe, 1998). In this environment, CSOs, including research-oriented organisations such as the Civil Society for Poverty Reduction in Zambia, the Malawi Economic Justice Network in Malawi, Research on Poverty Alleviation (REPOA) in Tanzania and the Ghana Centre for Democratic Development burgeoned and began participating in the policy process at a number of different levels.

The efficacy of African CSOs’ involvement in the policy process is, however, highly contested (Jones and Tembo, 2008). Concerns include the strong influence that donors have played in shaping civil society formation at the potential expense of independence and longer-term sustainability (Mamdani, 1996; Buhler, 2002); insufficient coordination and funding (Fatton, 1995, 1999); and inadequate linkages with governmental or CSO actors to develop credible policy alternatives (Nasong’o, 2007). Admittedly, this pattern has changed somewhat over the past decade, due in part to several cycles of multiparty elections and rapid growth in CSO activities. Some CSOs in newly democratising and post-conflict states have sought to establish a set of minimum ‘engagements’ between civil society and executive and legislative branch actors, promoting government transparency and respect for human (p.107) rights (Amundsen and Abreu, 2006). Nevertheless, the relationship remains uneasy in many contexts in the region, as is highlighted by a recent trend towards tighter regulation of NGOs’ roles in policy advocacy (for example, Ethiopia’s recent NGO law – see discussion later this chapter).9

Another key challenge faced at the knowledge–policy interface in the region is widespread neglect of investment in higher education (Bloom et al, 2005). In 2000, sub-Saharan Africa’s higher education Gross Enrolment Rate was just 3.5% (Varghese, 2004). Contributing factors include resource-constrained governments, low prioritisation of education spending (until recently), donor focus on primary education and stringent structural adjustment policies in the 1980s/90s that further curtailed social sector spending. Moreover, with the exceptions of Sida and the International Development Research Centre (IDRC), very few donors have invested in long-term research capacity, limiting the efficacy of investment in Southern research capacities (Jones et al, 2007). This underinvestment is particularly pronounced in the case of social sciences, with the exception of economics (Jones et al, 2007). As a result, both the quantity and research capacities of local researchers have been seriously impacted (Sawyerr, 2004).

These effects have not been limited to the research community, but have spilled over into the research literacy, or research uptake capacities, of policymakers. Many government officials have limited research knowledge and, in particular, limited ability to interpret research findings, especially qualitative research (Chowdhury et al, 2006). These challenges have been exacerbated by weak knowledge management practices and extractive models of research undertaken by Northern researchers (Touré, quoted in Jones et al, 2007).

Finally, evidence-informed policy advocacy initiatives are often constrained by a dearth of well-positioned intermediaries to facilitate the uptake of new knowledge in policy debates. As Table 4.12 highlights, there are very few communities of practice concerned with child well-being in sub-Saharan Africa that have either a regional or national focus, and are able to straddle research, policy and practitioner communities. While international initiatives such as the MDGs and the Education for All Campaign have had an important impact on children’s access to health and education services (Sumner and Melamed, 2010), local champions of children’s rights to protection from abuse, neglect and violence and participation in decision-making are less prominent and lack resources (see, for instance, the Inter-African Committee on Traditional Practices, the African Network for the Prevention and Protection against Child Abuse and Neglect [ANPPCAN] and (p.108) (p.109) (p.110)

Table 4.12: Selected communities of practice on child well-being in Africa


Time frame


Examples of policy impact

Inter-African Committee on Traditional Practices (IAC) http://www.iac-ciaf.net/

The IAC is a membership organisation with national committees in 28 African countries working to eliminate FGM and other harmful traditional practices, including child marriage, abduction, nutritional taboos, widow inheritance, wife sharing, practices associated with childbirth, and skin-cutting practices like scarification and tattooing. These not only constitute serious health risks but also violate the basic human rights of women and girls.

The IAC educates and empowers women and girls, undertakes research on FGM and advocates and lobbies for an end to the practice. The IAC has had a variety of policy impacts, including:

  • inclusion of FGM in the Maputo Protocol;

  • networking with Solidarity for African Women’s Rights (SOAWR) for ratification of the Maputo Protocol (currently there are 45 signatories and 26 ratifications in force);

  • declaration on gender equality by African heads of state;

  • declaration of 6 February as the International Day of Zero Tolerance to FGM;

  • adoption of Common Agenda for Action to eliminate FGM, with emphasis on integrated approach.

The African Network for the Prevention and Protection against Child Abuse and Neglect (ANPPCAN) http://www.childtraffickinginafrica.org/


Based in Kenya, the network works in partnership with other organisations against child maltreatment in Africa, and has undertaken a limited number of researchbased reports to raise awareness of child trafficking.

ANPPCAN’s anti-trafficking programme aims to raise awareness and build capacity for service providers towards eliminating child trafficking in Ethiopia, Kenya, Tanzania, and Uganda.

African Child Policy Forum (ACPF) http://www.africanchildforum.org/site


A pan-African policy advocacy centre based in Addis Ababa, the forum aims to put African children on the public agenda by providing a forum for dialogue, improving knowledge of African children, and promoting African action to develop and implement effective policies and programmes.

The Forum runs the African Child Observatory and the African Child Information Hub (http://www.africanchildinfo.net/site/). It also publishes a variety of reports, including the African Report on Child Well-being (http://www.africanchildinfo.net/africanreport08/index.php), which is groundbreaking in that it scores the performance of African government efforts to improve child well-being. Governments are ranked as ‘most child friendly’ (e.g. Namibia, Tunisia, South Africa), ‘child-friendly’, ‘fairly childfriendly’, ‘less child-friendly’ and ‘least child-friendly’ (including Chad, Liberia, Eritrea). The report also stresses the dissonance between countries’ formal acceptance of an international treaty and practice, while decreasing government budgetary commitments.

Southern African Network to End Corporal and Humiliating Punishment of Children http://www.rapcan.org.za/sanchpc/default.asp


The network is an alliance of Southern African organisations working independently in their own countries to prevent and address child abuse and neglect and ensure the protection of the rights of children.

The network brings together interested and committed organisations in Southern Africa to work towards prohibiting corporal punishment through capacity building, information dissemination, and joint regional advocacy initiatives.

Committee for Liaison between Social Organisations for the Defence of Child Rights (CLOSE), Benin


CLOSE is a network of more than 30 NGOs in Benin concerned with protecting children from sexual abuse and commercial sexual exploitation.

CLOSE is involved in the implementation of a bilateral agreement with Nigeria to combat trafficking.

Child Protection Alliance (CPA), Gambia http://www.cpagambia.gm/


CPA has 63 member organisations. Its objectives include: raising awareness of child abuse and exploitation; building national and institutional capacity to prevent child abuse and exploitation and protect victims; promoting networking and alliance building; and empowering children.

CPA runs sensitisation workshops for teachers, religious and community leaders, parents, protection service providers, and security officers. It also produces quarterly newsletters and position papers on child protection issues and promotes research on children’s issues.

Ghana NGO Coalition on the Rights of the Child (GNCRC) http://www.smeghana.com/mysite/index.cfm?CompanyID=147


The GNCRC is an umbrella organisation that aims to: build capacity on good models of law enforcement practice to prevent the commercial sexual exploitation of children; build the capacity of local youth groups, the media, and lobbyists; raise awareness and provide training on children and young people’s participation for groups; and identify areas for regional collaboration.

The GNCRC collaborates with ministries and agencies, including the Ministry for Women and Children Affairs, lobbies government to establish temporary shelters for victims in each region, and ensures implementation of the make-IT-safe campaign. It also develops structures to protect child welfare (foster homes, hospitals, etc), collects data, develops child-friendly material, and disseminates information across regions to strengthen networks.

Uganda Child Rights NGO Network (UCRNN) http://www.ucrnn.net/ucrnn/index.php


UCRNN is a network of more than 60 childfocused organisations that engage in direct programme implementation and service delivery throughout Uganda.

UCRNN is engaged in promoting and popularising national, regional, and international instruments on the rights of children; monitoring the implementation of the UNCRC and its optional protocols; carrying out national level advocacy on key child rights issues; engaging in policy development and review processes; facilitating information sharing through networking; researching key child rights issues; capacity building with member organisations and other actors; and piloting initiatives to increase access to services for vulnerable children.

End Child Prostitution, Child Pornography, Child Trafficking for Sexual Purposes (ECPAT) Uganda http://www.ecpatuganda.net/index.php


ECPAT Uganda is a six-member coalition working to create vigilant communities and stakeholders who safeguard children from trafficking and sexual exploitation.

ECPAT undertakes research on commercial sexual exploitation of children (CSEC) to provide up-todate information on its nature and magnitude. It lobbies and runs media awareness campaigns, networks with other actors, and works with children. It is also working to strengthen legal and policy frameworks to ensure better and sustainable protection mechanisms for children.

Child Welfare South Africa (CWSA) http://www.childwelfaresa.org.za/


CWSA has 263 member organisations and is the largest non-profit, nongovernmental, and volunteer -driven organisation providing child protection services in the country. CWSA works to prevent child abuse, serve orphans and vulnerable children (OVC) – including those made vulnerable by HIV/AIDS, build organisational capacity, and advocate for children on a national and regional level.

CWSA develops forums to create awareness of the specialised services needed for child victims of trafficking and commercial exploitation; runs workshops and training sessions countrywide; ensures that legislation is in place to protect children; and forges national and international linkages that develop Codes of Conduct and a good practice model for children affected by violence.

Children in Need Network (CHIN), Zambia http://www.chin.org.zm/


CHIN is a consortium of over 240 organisations working to promote the rights and welfare of children in Zambia, including child protection and resource tracking with regard to orphans.

CHIN’s activities include awarenessraising, such as consensus-building workshops to identify issues and strategies, community meetings, group discussions, drama performances, and film screenings. It has recently undertaken research to ascertain the prevalence of CSEC in Zambia, and is now undertaking an awareness raising campaign on national TV highlighting issues of child trafficking, pornography, prostitution and child sex tourism (CST).

(p.111) the Southern African Network to End Corporal and Humiliating Punishment of Children in Table 4.12). Moreover, these communities have generally had greater success in agenda-setting and securing discursive commitments from governments than in impacting on substantive policy change and influencing behavioural shifts.

International agencies such as Save the Children, Plan, World Vision and UNICEF are all helping to plug this gap, as their work increasingly shifts away from direct service delivery, and towards evidence-based policy advocacy (see Table 4.13 for examples). However, they still face considerable challenges in establishing dialogue with policy actors outside ministries with child-related mandates (i.e. ministries of children and youth, health, education) and with International Financial Institutions (IFIs). (p.112)

Table 4.13: Selected examples of evidence-informed policy influencing child well-being in Africa by international agencies




Evidence base

Policy outcomes

UNICEF West and Central Africa and ODI


Child-sensitive social protection

  • Contributed to the development of a five-year social protection plan

  • Initiated a national cash transfer programme in Senegal

  • Secured formal commitments to invest in social protection for children in Equatorial Guinea (the Malabo Declaration) and in Congo (adoption of a White Paper on social protection)

  • Changed UNICEF’s historic support for the Bamako Initiative, which endorses user fees for health care

Save the Children UK, Help Age International, and IDS


Cash transfers to promote human development

Report called ‘Making Cash Count: Lessons from Cash Transfer Schemes in East and Southern Africa for Supporting the Most Vulnerable Children and Households’ www.ids.ac.uk/go/idsproject/making-cashcount

Invested in cash transfer programmes by international agencies and governments in the region

World Vision and Johns Hopkins University Centre for Refugee and Disaster Response


Children’s postconflict mental health

Impact evaluation on mental health programme in refugee camps in northern Uganda http://www.certi.org/publications/policy/ugandafinahreport.htm

Contributed to successful fund-raising to undertake similar work in other contexts on the basis of the rigorous evidence base developed (Jones et al, 2008)

Plan, ODI, and the International Observatory on Violence in Schools


School-based violence

Reports on the prevalence of and underlying causes of school-based violence in developing and OECD countries http://plan-international.org/learnwithoutfear/resources/publications/campaign-report

This research underpinned Plan’s Learn Without Fear Campaign. Initial outcomes in Africa include: contribution to UNICEF and World Health Organization (WHO) complementary initiatives for developing health-related and behavioural indicators to improve international violence against children monitoring and evaluation standards; the adoption of the Togolese Children’s Code and community training on the UNCRC and its child protection-related provisions; and the establishment of a free 24-hour telephone helpline facilitating preventive and support services through referrals and school outreach services.

4.5 Case study: children, expert-led policy advocacy and the Ethiopian PRSP

As the central, country-led strategy for achieving the Millennium Development Goals, PRSPs must include a strong focus on children’s rights. The Committee urges governments, donors and civil society to ensure that children are a prominent priority in the development of PRSPs (Committee on the Rights of the Child, 2003, p 14, para 62)


We now turn to a case study on evidence-informed policy advocacy efforts in Ethiopia, a country that shares many of the broader knowledge–policy–practice characteristics outlined in the previous sections. The case study concerns an initiative led by an academic–NGO partnership to mainstream a child-sensitive perspective into Ethiopia’s second PRSP and provides a useful lens for exploring linkages between knowledge on child poverty and policy change in sub-Saharan Africa.10

(p.113) Policy change objective(s) and children’s 3D well-being

The Ethiopia children and PRSP project – an IDRC-funded policy research initiative involving a North–South academic consortium and an NGO, Save the Children UK – sought to examine the impacts of Ethiopia’s first PRSP (2002–05) on children’s experiences of poverty and well-being, and to draw lessons for the formulation of its second PRSP, the 2006–10 Plan for Accelerated and Sustained Development to End Poverty (PASDEP). Given that the first-generation PRSP did not include a specific analysis of childhood poverty and vulnerability, had limited child-related policy commitments (only those related to the MDG commitments on child survival, health and education) and had even fewer measurable indicators (Heidel, 2005), the aim of the project was to explore the national poverty-reduction strategy’s broader impacts on children’s multidimensional well-being and to raise awareness among civil society and state actors in order to influence the second-generation PRSP revision process. The aim was thus to challenge the normalisation of the exclusion of children and children’s rights from mainstream policy processes, and to enhance the visibility of children’s well-being not only in policy sectors that are child-focused but also in other areas of development that may have important intended or unintended impacts on children.

In response to growing criticism of the top-down, technocratic approach of structural adjustment programmes of the late 1980s/ early 1990s, public participation and consultation were introduced as defining characteristics of the PRSP process in the late 1990s/early 2000s (Piron and Evans, 2004). The recognition that civil society groups (from peasant associations to women’s groups and labour unions to traditional authorities and religious groups) had important insights to offer on poverty and vulnerability marked a critical departure from prior international development community thinking (McGee, 2004). The integration of a child rights perspective, however, proved more challenging. Especially in the interim and first-generation PRSPs, children and young people were routinely excluded from civic consultation processes, and civil society umbrella organisations were often equally remiss at including child rights issues in their recommendations papers (e.g. Marcus and Wilkinson, 2002; Jones et al, 2005).

Moreover, where child rights issues were incorporated, this tended to be in terms of child-targeted programmes in traditional social policy areas such as education and health. While important, this approach overlooked a growing body of evidence regarding the ways in which (p.114) children living in poverty may also be deeply affected by broader economic development and poverty-reduction policies (Marcus and Wilkinson, 2002). In addition, data constraints hampered situational analyses and the development of measurable child-sensitive indicators, while officials’ and donors’ lack of awareness of the importance of including children in policy processes limited uptake of the knowledge generated by children’s participatory poverty assessments (PPAs). Recognition of the multidimensional nature of child poverty and well-being and the ways social and power relations often perpetuate children’s vulnerability to exploitation, abuse and neglect were also overlooked (Harper et al, 2009).

3D evidence generation on 3D child well-being

In order to overcome some of these political and data constraints, the case study project sought to generate evidence on childhood poverty and well-being and its reflection in poverty-reduction frameworks in Ethiopia. A mixed-methods approach was employed, focusing on: (i) a child-focused content analysis of the first-generation PRSP and related sector policy frameworks; (ii) a review of child-related indicators in select international PRSPs identified as child-friendly by a 2002 comparative review (see Marcus and Wilkinson, 2002); and (iii) primary research on children’s time use, education experiences and nutritional health over the course of the implementation of the first PRSP. There was thus a mix of primary and secondary research using both qualitative and quantitative approaches that included children’s own vision of their poverty experiences.

The project was embedded within the broader Young Lives research programme, a multi-year UK DFID-funded project on childhood poverty over the course of the MDGs, and was thus able to draw on the Young Lives data set, which involves a relatively large (3, 000 households) sample spanning five of the most populated regions of the country. The project combined quantitative analysis of this survey data with community dialogues, focus group discussions and semi-structured interviews undertaken with government officials, caregivers and children. The latter allowed for greater insight into causal processes and the complex dynamics behind quantitative findings.

Equally importantly, the project’s research team combined multiple academic disciplines: economics, political science, public health, sociology and gender studies. Although doubtless more time- and labour-intensive than mono-discipline research, the combined perspectives facilitated an analytical approach that was convincing (p.115) for multiple audiences. Econometric analysis provided currency in the language of power: not only are economists highly respected in Ethiopian society, but they also constitute the majority of officials in the Ministry of Finance and Economics and were key players in the donor community. Meanwhile, contextual sociological analysis and in-depth case studies and participatory approaches allowed technical analysis to be translated into a compelling, human-centred narrative about the implications of the PRSP on child well-being, and broadened the project’s reach to diverse civil society and public audiences. Moreover, where possible, resulting policy recommendations were informed by international good practice, especially with regard to progress indicators (see Jones et al, 2005).

3D approaches to knowledge–policy interaction

The project’s approach to knowledge–policy interaction was embedded within the project design from the outset. It included a multi-pronged dissemination and communications strategy that sought to engage with the dynamics of policy narratives/messages, actors/networks and context/institutions.

In terms of actors and networks, seminars with key policy and academic stakeholders were held to discuss working paper findings, the development of video documentary and photography projects to raise awareness of the urgency of tackling childhood poverty, and capacity-building workshops with national- and state-level officials and civil society practitioners to foster a better understanding of linkages between child well-being and macro-level poverty and economic development policies.

The political context and analysis of it formed the rationale for this choice of approach (see Jones et al, 2008), which, as Piron and Evans (2004: 10) argue, cannot be underestimated in PRSP processes:

The process interacts with institutional constraints, in particular those which originate from the nature of the state, its historical antecedents, and the way its power is exercised. Formal aspects of the political system matter as well as the informal rules by which they operate … . The PRSPs are significantly affected by the degree to which poverty is politically salient and to which there is ‘political capital’ to be derived from poverty reduction efforts. This is affected in turn by the nature of the nation-building project and associated political ideologies.

(p.116) In this case, a situation analysis underscored the Ministry of Finance and Economy’s (MOFED’s) dominant role in shaping the PRSP formulation process, supported by a select group of academic and IFI consultants and advisors. By contrast, civil society engagement with the first-generation PRSP process had been relatively superficial and restricted, with more substantive negotiations taking place between the donor community (responsible for significant aid flows) and MOFED (Tefera, 2003). In order to manoeuvre within this complex constellation of actors and institutions, it would therefore be important to liaise closely with MOFED officials and their networks on a formal and informal basis to understand their knowledge demands and constraints, but also to keep civil society actors abreast of emerging evidence on child-specific PRSP outcomes. Given Ethiopia’s federal political structure, it would also be necessary to ensure that policy-influencing efforts were directed at both the national and sub-national regional state levels.

The accumulated learning by researchers and activists alike has shown that a sense of government and community ‘ownership’ of a research project is likely to facilitate the acceptance and recognition of research findings (e.g. Pham, 2003). Accordingly, one of the central aims of the project was to promote government buy-in to the research objectives from the outset. An important component of this strategy involved building strong relationships with key government champions – this involved housing the research component of the project within the Ethiopian Development Research Institute (EDRI), which was headed by the Prime Minister’s Chief Economic Advisor; securing approval from the Disaster Prevention and Preparedness Committee for the dissemination and advocacy components carried out by Save the Children; and forming a project advisory panel comprised of key sector ministry officials and donor and international organisation representatives. In addition, while structuring seminars with donors and government officials where project research findings were launched, key players in the PRSP development process were invited to give presentations on how they were seeking to incorporate children’s rights. In this regard, rather than seeking to criticise existing government policies or to embarrass prominent officials into action, the format provided space for officials (from EDRI, MOFED and the Ministry of Labour and Social Affairs) to reflect on the relationship between broad development strategies and children’s rights, and to develop an approach to strengthen these linkages.

The project’s relationship with civil society actors was less clear-cut. Although project members enjoyed links with the two major civil (p.117) society umbrella groups, the Christian Relief Development Agency (CRDA) and the Poverty Action Network Ethiopia (PANE), closer networking with MOFED suggested that government officials were somewhat wary of these groups. The roles of these umbrella bodies had not been clearly communicated to the PRSP committee, and there were questions about the rigour of their analysis and evidence base. Somewhat disappointingly, opportunities for alliances with child rights-focused organisations also proved limited as their organisational strategies were sector-focused and had minimal space for engaging with broader macroeconomic issues. This experience, therefore, underscored the need for partnership projects of this nature to be cognisant of and flexible regarding who delivers policy research messages. As Start and Hovland (2004) argue, the messenger matters in facilitating the translation of ideas into policy action.

Finally, in terms of policy ideas/narratives, message-framing was also of critical importance. Given limited awareness about the nature, severity and underlying causes of childhood poverty, the construction of policy narratives that drew the attention of donor, civil society and government audiences ordinarily unfamiliar with child rights issues was a critical part of the project’s knowledge interaction endeavour. Framing policy messages requires a mixture of skills, including the use of culturally and audience-appropriate discourses, the construction of pithy narratives that do not unnecessarily ‘dumb down’ what are often complex messages, and the development of specific concrete policy recommendations. As Tarrow (1995) argues, collective action does not result from a simple conversion of objective socio-economic conditions into demands for change, but rather depends on subjective perceptions of injustice and how political discourses are framed in culturally resonant ways.

Two examples suffice to illustrate this point. First, while international conventions and standards hold some sway at least in part because of Ethiopia’s heavy reliance on international aid, there is simultaneously a strand of political culture that is weary of accepting international norms without first assessing their feasibility in a low-income, multi-ethnic society with a recent history of political turmoil. Accordingly, there tends to be a strong emphasis on ensuring that international frameworks are ‘localised’: for example, rather than speak of the UNCRC, officials often prefer to look towards the ‘National Action Plan for Ethiopian Children’. In this vein, the project’s partnership with Southern institutions, which are more attuned to such cultural sensitivities, was a key ingredient of successful message-framing.

(p.118) King et al (2005) argue that ‘skilful narratives’ and ‘pithy summaries’ are needed to encapsulate the key elements of research conclusions. Given the public’s and, in particular, the media’s penchant for messages in sound-bite format, there is a frequent danger that the impact of findings will be diluted or even misinterpreted if they are stripped of context. Mindful of these considerations, the project sought to reframe common assumptions about the nature and causes of childhood poverty. Findings were presented around a core message that children are not only impacted by education and health sector policies, but that their well-being may be critically shaped by broader development and poverty-reduction policies. As such, children’s rights need to be ‘mainstreamed’ into national development and economic policy frameworks. By adapting the language of gender mainstreaming, which has been widely adopted throughout development circles in the country, the project sought to highlight that all sectoral ministries need to consider the direct and indirect impact of their policies on children and that policymakers need to pay attention to the potential synergies or contradictions among policies on child outcomes. For example, research on the impacts of the core economic pillar of Ethiopia’s first PRSP – agricultural-led industrial development – on child well-being highlighted unintended negative spillover impacts on children. The PRSP’s agricultural extension policy’s heavy reliance on subsistence agriculture had increased child involvement in work activities, particularly animal herding, to the detriment of their school attendance and/or time available to invest in homework and study (Woldehanna et al, 2005a). Similarly, food or cash for work programmes in the absence of affordable and available childcare services had been found to encourage women and children’s participation in public work activities at the cost of caring time for children and/or children’s education (Woldehanna et al, 2005b). In other words, household-level poverty-reduction strategies risked indirectly undermining education sector initiatives to increase child enrolment and, as such, threatened achievement of the MDG for Universal Education for All. A more child-sensitive approach to poverty-alleviation policies was therefore needed, including piloting alternative policy measures such as community childcare mechanisms and communal grazing policies.

Outcomes of a 3D approach

In what way – if at all – did the 3D approach to knowledge–policy interaction shape child well-being outcomes? It is exceedingly difficult to measure the impact of research knowledge on policy change, (p.119) especially given the multiplicity of factors that shape policy processes and the potential for research to directly and indirectly affect actors involved in the knowledge–policy interface (see Chapter 3 for further discussion of this issue). In this case, given the project’s close links to officials drafting the second PRSP, we can surmise that the project’s research findings did contribute to greater visibility of children’s well-being in policy circles as reflected in the inclusion of a specific section on children’s poverty and well-being in the PRSP’s vulnerability analysis. The section explicitly mentioned that the PRSP should be implemented in such a way as to be compatible with the National Action Plan for Ethiopian Children, which incorporates a strong child rights perspective. Accordingly, there was a window through which recommendations related to children’s multidimensional rights could subsequently be taken up. However, while this marked an important advance vis-à-vis the country’s first-generation PRSP, the analysis was relatively brief and lacked specific recognition of children’s rights to protection from violence and abuse, including exposure to harmful forms of child labour, which had been a key recommendation. Progress indicators against which governments are held responsible by donors were also not expanded to include many of the child-sensitive indicators that the project had recommended based on lessons learned from other low-income contexts.

To understand the broader context in which the PRSP revision process took place, it is also important to be mindful of the dramatic shift the May 2005 national elections represented in Ethiopian politics. Prior to the elections, Ethiopia was generally perceived to be on a positive development trajectory, with stable economic growth and a pro-poor development agenda, the cessation of conflict with its neighbour Eritrea, and improving governance conditions, including growing openness to civil society engagement in policy debates. Indeed, President Zenawi Meles had been heralded by Tony Blair as emblematic of the ‘New Africa’. Accordingly, few analysts predicted the speed with which the political context would shift. Unexpectedly high voter turnout and a surprisingly strong showing by the new coalition of opposition forces had two major implications: (i) discussions about (and media coverage of) the PRSP and general development issues were overshadowed by highly charged debates about the election, election fraud and violent unrest in Addis Ababa; and (ii) the credibility of the CRDA, the main civil society umbrella group, was eroded in the eyes of the government due to its alleged link to the opposition. Overall, this resulted in the demise of an already fragile (but previously thawing) relationship between civil society and the ruling party/government, with the latter (p.120) conceding little political space. It is perhaps not surprising, then, that the eventual outcome of the PRSP revision process was limited in scope as more pressing political-economy dynamics came to dominate the political stage.

4.6 Conclusions

Three general insights emerge from the discussion on child well-being, the regional overview of knowledge and policy actors and processes working on policy change to improve child well-being, and the empirical study of knowledge–policy interactions in one context in Africa.

Policy ideas and narratives

First, research needs to be credible and framed around narratives that are culturally resonant. Owing to what Ahmed (2005: 767) calls the ‘multiplier effect’, research is likely to prove more persuasive to a broader audience if it includes interdisciplinary perspectives and mixed methodologies. For example, econometric analysis can be powerful in persuading poverty-reduction strategists of the importance of incorporating children’s rights into national development strategies as it resonates with the disciplinary perspective of officials in ministries of economics and finance and IFIs. But equally importantly, in-depth qualitative findings can enable research about child well-being to be framed in human terms and to make sense of sometimes seemingly counter-intuitive survey findings. This is critical in reaching broader civil society and sub-national-level audiences.

Framing messages in succinct, easily remembered and culturally resonant ways provides a linguistic bridge between often complex academic texts and policy action. Such packaging needs to take into account politico-cultural and ideological sensitivities, and for this the insights from Southern partners are essential. Reference to best practices elsewhere can also strengthen policy recommendations, but only if care is taken to ensure that ideas are adapted to the local context.

Policy actors and networks

Second, the importance of securing strong relationships with key players or policy entrepreneurs – many of whom may be outside child-focused ministries or agencies – cannot be overestimated. Such links provide vital information on officials’ knowledge demands, decision-making (p.121) hierarchies, processes and timelines. Moreover, research findings are unlikely to be accorded the necessary credibility if stakeholder support for a project’s objectives has not been previously established, and this is arguably particularly the case with many child-related issues that have yet to be mainstreamed. In some political contexts, policy entrepreneurship may also entail investing significant energy and resources in forging strategic partnerships. For instance, locating a research project within a government-affiliated agency may be critical to promote buy-in. In short, the credibility of the ‘messenger’ needs to be taken as seriously as the development of the actual messages.

Policy context and institutions

Third, there are ways of approaching policy advocacy that are more likely to lead to policy change. Given the complexities of the policy process and the role of knowledge in shaping policy trajectories, intent to shape policy change is significant.

Although research-based knowledge may have an impact on policy practitioners’ thinking and practice through the process of ‘knowledge creep’, whereby ideas gradually filter through to a broader array of policy stakeholders (Crewe et al, 2005), there is growing consensus that research explicitly designed to influence policy has a better chance of success than research that relies on chance or accident to shape policy (Saxena, 2005). Policymakers’ demands for research findings to be translated into specific, context-appropriate indicators and policy recommendations means that if policy influencing is not an explicit aim, it is unlikely that the effort required for this interpretive task will be taken. For example, the effort required to package an academic-style research paper into readily accessible policy-relevant messages is considerable and cannot be left to chance. Not only do research outputs have to be produced under tight and changing deadlines to meet governmental drafting deadlines and to hold stakeholder workshops, but there often has to be considerable flexibility in approaching the way in which research findings are framed and communicated over the course of a project as politico-institutional contexts change. In light of this fluidity, a dual strategy of engagement may be most conducive to ensuring social change whereby independent dialogue with officials is balanced with networking and awareness-raising with civil society coalitions in order to develop a broader support base.


(1) This section draws on Jones et al (2008).

(2) In response to the need to generate accurate estimates for under-five mortality (due to the challenge posed by lack of data in developing countries), experts at UNICEF, WHO, the World Bank, the United Nations Population Division (UNPD) and members of the academic community formed the Inter-agency Group for Child Mortality Estimation (IGME). The group aims to source and share data on child mortality, to improve and harmonise estimation methods across partners, and to produce consistent estimates on the levels and trends in child mortality worldwide. It does this by compiling national-level estimates, including data from vital-registration systems, population censuses and household surveys. A regression curve is then fitted to these points and extrapolated to a common reference year. In addition, to increase the transparency of the estimation process, the IGME developed a publicly accessible database containing full details of country-specific estimates and their underlying source data.

(3) We include maternal mortality here because a high number of maternal deaths are suffered by girls. Adolescents aged 15 through 19 are twice as likely to die during pregnancy or childbirth as those aged over 20, and girls under 15 are five times more likely to die (WHO and UNFPA, 2006). The foundations for maternal risk are often laid in girlhood. Girls and women whose growth has been stunted by chronic malnutrition are vulnerable to obstructed labour. Anaemia predisposes women to haemorrhage and sepsis during delivery and has been implicated in at least 20% of post-partum maternal deaths in Africa and Asia. The risk of childbirth is even greater for girls and women who have undergone female genital mutilation (FGM) – there are an estimated two million girls each year who are subject to this procedure. Child marriage is another important risk factor – young women report being less able to discuss contraceptive use with their husband, and thus child marriage is associated with higher prevalence of early childbearing. Early marriage also puts young women at greater risk of HIV (ICRW, 2006). In addition, at least two million young women in developing countries undergo unsafe abortions, which can have devastating health consequences (ICRW, 2006). Indeed, in Nigeria, abortion complications account for 72% of all deaths in women under the age of 19 and 50% of all maternal deaths among Nigerian adolescents (Airede and Ekele, 2003). See also: http://www.childinfo.org/maternal_mortality.html

(p.123) (4) The new study narrows the uncertainty around global and national MMR estimates compared to previous assessments. The improved accuracy is a result of an extensive database coupled with the use of analytical methods with increased explanatory power and improved out-of-sample predictive validity.

(5) This is more reliable than income or consumption inequality survey data because it is based on household assets that can be readily observed – such as the possession of a bicycle or a radio, electricity or water connections, and the size and construction quality of a dwelling (Filmer and Pritchett, 2001). Country differences are much clearer based on this criteria. Bolivia and Namibia have similar under-five mortality, but children in the bottom quintile in Bolivia are considerably worse off than what the national under-five mortality rate statistic suggests. Children in the top quintile in Bolivia, on the other hand, face a much smaller risk of premature death than their counterparts in Namibia. Vandemootele and Delamonica (2010) also note that of the 63 countries in their under-five mortality sample, 46 present trend data – of those, the majority display either widening disparities over time or no consistent trend. Only two countries (Bolivia and Ghana) show a distinct tendency towards less inequity (see Vandemoortele and Delamonica, 2010, Table 4).

(6) Note that these indicators are only currently available for the MICS 3 data set.

(8) Note that there is some overlap with Table 4.8, which focuses on children’s perceptions of poverty and generally employed participatory methods.

(9) This law was passed ahead of the 2010 elections and prohibits NGOs that receive more than 10% of their funding externally from participating in policy debates on a far-reaching list of thematic areas, including governance, rights and gender equality.

(10) This case study draws on Jones et al (2008).