HOW COLLABORATION, EARLY ENGAGEMENT AND COLLECTIVE OWNERSHIP INCREASE RESEARCH IMPACT:How collaboration, early engagement and collective ownership increase research impact
Strengthening community-based child protection mechanisms in Sierra Leone*
Using inter-agency action research in Sierra Leone, this chapter provides a case study on how a highly collaborative approach can enable child protection research to achieve a significant national impact. The chapter describes how the inter-agency research facilitated a community-driven approach to addressing teenage pregnancy. The promising results obtained before the Ebola crisis helped shape a new Child and Family Welfare Policy that featured the role of families and communities rather than formal structures. Then it examines how the social process of the research enabled it to have a national impact.
A strategic partnership with UNICEF, a collaborative, dialogue-oriented approach to finalising the methodology and site selection, and ongoing learning enabled a spirit of collective ownership. Key lessons include the importance of using a collaborative, inter-agency approach at all stages; promoting early engagement with diverse actors; having ongoing engagement with the relevant government ministry at multiple levels; and working with a broker that helps to understand and manage power dynamics. Although the process described may not be possible in all settings, a collaborative, collectively owned approach is a promising approach for boosting research impact.
Research is typically designed and conducted with an eye towards technical considerations such as robustness, validity and reliability. However, research that meets stringent technical standards frequently fails to achieve the desired impact in enabling changes in practice, policy, or both. Not uncommonly, this situation leaves researchers scratching their heads and asking ‘Why aren’t the leaders listening?’
The purpose of this chapter is to help illuminate how a collaborative, partnership approach can enable research to have a greater impact on policy and practice at a national level. Telling the story of inter-agency action research on child protection in Sierra Leone, it features the human side of research and the importance of collective ownership. First the chapter outlines the origins of the action research on child protection and wellbeing and discusses its methodology, key findings and contribution to a new national Child and Family Welfare Policy. It then analyses the key lessons learned about how the action research process enabled it to have national impact.
Background on the inter-agency action research
Many of the world’s most vulnerable people are children, defined under international law as people under 18 years of age. In many developing countries, children frequently comprise half or more of the population. In both emergencies and development settings, children’s vulnerability owes in no small part to the myriad threats or child protection risks in their environment, that is, in their social ecologies (Bronfenbrenner 1979) such as families, schools, communities, and the wider social system. These may include risks such as violence, rape and other forms of sexual assault, armed attack, mass displacement, separation from caregivers, loss of loved ones, trafficking, HIV and AIDS, child labour, and recruitment into armed forces or armed groups (Boothby, Strang and Wessells 2006; Fernando and Ferrari 2013). As these risks accumulate, children may be likely to experience intense suffering, mental health issues, developmental delays and difficulties functioning well in tasks such as education. Collectively, these issues make it a high priority to provide child protection, defined as ‘the prevention of and response to abuse, neglect, exploitation, and violence against children’ (Child Protection Working Group 2012: 13).
A largely unanswered question, however, is ‘What are the best means of protecting children?’ At present, the emphasis is on the strengthening of national child protection systems (African Child Policy Forum et al. 2013; Davis, McCaffery and Conticini 2012; Krueger, Thompstone and Crispin 2013; UNICEF et al.2013; Wulczyn et al. 2010). Key to system strengthening is the work of formal actors such as police, government social workers and magistrates and also non-formal actors such as families, communities and leaders including religious leaders, elders and teachers.
In strengthening child protection, international non-governmental organisations (NGOs) have used a predominantly top-down approach in which outside child protection experts impose their approaches on local people (Freymond and Cameron 2006; Wessells 2009, 2015). For example, at grassroots level, international NGOs have made widespread use of community-based mechanisms such as Child Welfare Committees (CWCs; also called Child Protection Committees) to help protect children (Wessells 2009). Typically, a CWC consists of 10–15 people and includes women and men and several teenage boys and girls. Having been trained in child rights and child protection, the CWC members monitor their village or neighbourhood for violations against children and report the violations to appropriate authorities such as the police. CWC members also work to prevent violations through community discussions and education. This is a top-down approach in that the idea for the CWC came from the NGO, which then persuaded and led the community to accept it. As explained below, there are reasons to question this approach.
1.1 A global review
In 2009, however, a global, inter-agency review of community-based child protection mechanisms reported that the evidence base showing the effectiveness of CWCs is quite weak (Wessells 2009). Few evaluations used robust designs that enable one to make causal attributions about the intervention’s effectiveness, and most evaluations focused more on outputs such as the number of trainings conducted for CWCs than on the actual outcomes for children. Also, NGO-facilitated CWCs typically had low to moderate levels of community ownership, as local people tended to view them as NGO projects rather than as processes that communities themselves had constructed to fulfil their obligations to protect children. This finding was problematic for the use of CWCs because the review found that community ownership was the most important determinant of effectiveness and sustainability. In many settings, when the funding for the CWCs dried up, the CWCs typically languished or collapsed. In addition, NGO-facilitated CWCs were found in some cases to compete with and undermine indigenous community mechanisms such as action by chiefs and elders on behalf of vulnerable children. This is unfortunate because the latter mechanisms frequently enjoy high levels of community ownership and are more likely to be sustainable. Further, community-based mechanisms were more likely to be effective and sustainable when they were linked with and supported by formal actors at higher levels (e.g. district level) within the wider child protection system. Together, these results indicated the need to develop a different child protection approach that enables higher levels of community ownership.
1.2 Designing the action research
Subsequently, Save the Children and the United Nations Children’s Fund (UNICEF) convened a meeting of inter-agency stakeholders, including government and community stakeholders, in Nairobi to plan appropriate next steps. Through a highly participatory process, the participants decided to develop and test, using mixed methods, an alternate approach to community-based child protection mechanisms that would feature high levels of community ownership and also appropriate links with district-level child protection stakeholders. In contrast to the top-down approach, the non-formal–formal links were to be decided by the community in a grassroots-driven or bottom-up approach.
Broadly, the design included elicitive learning (i.e. not using preconceived questions and categories) about harms to children and community mechanisms for supporting vulnerable children, followed by use of a robust design that permits one to make causal inferences about the effects of the intervention. To achieve high levels of community ownership, the intervention was to be community-driven rather than NGO- or expert-driven. Following an action research approach, communities themselves would select which issue to address, develop an intervention, implement the intervention and help to evaluate it. To help strengthen the evidence base, the design included the use of baseline, mid-point, and endline measures of actual outcomes for children.
The group also decided to form an Inter-Agency Learning Initiative on Community-Based Child Protection Mechanisms and Child Protection Systems. Save the Children (via Sarah Lilley) was selected to coordinate a global Reference Group, with one of its members – the Columbia Group for Children in Adversity (via Michael Wessells) – serving as the technical arm for the research. This collaborative approach was not incidental but grounded in the belief that no single agency by itself can protect children and that mutual learning and collaboration are at the heart of strengthening the child protection sector both nationally and globally.
To increase the generalisability of the approach and findings, the group decided to conduct the action research in two different regions of sub-Saharan Africa: West Africa, and East and Southern Africa. The selection of one country in each region as a site for the research was guided by multiple criteria, including the willingness of the UNICEF country office to help support the research. This criterion proved to be of pivotal importance since UNICEF has a mandate to work closely with and support governments and is well positioned to influence policies relating to children. Also, since UNICEF is the global standard bearer in regard to child protection, UNICEF involvement and support are key for influencing practice.
1.3 The action research in Sierra Leone
Sierra Leone was selected in 2010 through a consultative process that had both national and international dimensions. A key consideration was the keen interest of UNICEF Sierra Leone in participating in and supporting the research. Sierra Leone had a plethora of child protection issues in 2010, some eight years following the end of its brutal decade-long war. UNICEF was concerned about addressing these issues not only because of their magnitude but also because they had reason to question the dominant approach then used to protect children. In 2007, the Government of Sierra Leone (GoSL) had enacted a Child Rights Act that had mandated the establishment of a CWC in each village. Even in 2010, however, UNICEF had preliminary evidence that this approach was not functioning as had been intended (Child Frontiers 2010). This realisation made the research of keen interest to UNICEF. Internationally, Sierra Leone was of interest because many CWCs had been established during the war but had collapsed afterwards, suggesting the need for an alternative approach. Also, the Principal Investigator (PI) (Wessells) had worked in Sierra Leone off and on during the war and was familiar with the context. From experience, he knew that Sierra Leone had many talented child protection workers, chief among whom was UNICEF worker David Lamin.
The action research was led by a mixture of national and international researchers. The Lead National Researcher, Dora King, oversaw national teams of trained, female and male Sierra Leonean researchers who spoke the local languages and understood the local contexts. The data collectors were backstopped and mentored by Team Leaders in Moyamba (Lamin) and Bombali (King). In turn, the Team Leaders were supported by international researchers, primarily Lindsay Stark (Lead Methodologist) and the PI.
The research was conducted in multiple phases, beginning in January 2011. The initial ethnographic phase aimed to establish trust with and learn deeply about communities and their views of who were children, what were the main harms to children, and what happened when particular harms to children occurred. Living and working in villages, the researchers used methods such as participant observation, in-depth interviews, group discussions and body mapping to learn from different sub-groups such as girls, boys, women and men. The main harms to children that local people identified were: children being out of school, teenage pregnancy out of wedlock, heavy work, and maltreatment of children not living with their biological parents. Surprisingly, among the top ten harms was ‘child rights’, which adults said had undermined their authority as parents since child rights workers had taught that parents should not discipline their children by corporal punishment. Further, the participants reported overwhelmingly that people did not report violations against children, even criminal offences, to the CWCs or government officials such as the police (Wessells 2011; Wessells et al. 2012). These findings, which resonated with others (e.g. Behnam 2011; Child Frontiers 2010), raised strong questions about the effectiveness of top-down approaches (Wessells et al. 2012, 2015), including CWCs and the Child Rights Act itself.
In the second phase (2012), the research team used a free listing methodology to learn how local adults and teenagers (13–18 years of age) understand children’s wellbeing. They consistently identified aspects such as participation in education, contributing to one’s family, respect for elders and obedience as key signs that children are doing well (Stark et al. 2012). These outcome areas, together with those derived from the ethnographic research, were used to define key outcome areas for children’s risks and wellbeing. Subsequently, these outcome areas were used to define specific indicators and to construct a survey that measured children’s risks and wellbeing outcomes. In this manner, local views regarding important outcomes for children were incorporated into systematic measures. The population-based survey that was developed also reflected a balance of outcomes for children that were based on international child rights standards.
In the third phase (2013–15), the research used a quasi-experimental design in which clusters of communities were assigned randomly to an intervention condition or to a comparison condition. To enable community ownership of work to support vulnerable children, the approach taken was that of participatory action research, which both reflects and enables community resilience (Mckay et al. 2011; Wessells 2012). In participatory action research, local groups of people collectively identify a problem of concern and then mobilise themselves to plan, implement and evaluate an intervention to address the problem. This approach generates high levels of community ownership since it is the community that holds the power and makes key decisions, defines the problem and manages or runs the intervention.
The idea was to have communities select a harm to children and then implement a self-designed intervention to address it. To promote bottom-up system strengthening, the communities were to choose and collaborate with formal (government) actors in the child protection system. Living within each intervention cluster was a trained facilitator who was highly process-focused and enabled inclusive participation, slow dialogue and group problem solving, and decision-making by the communities, without excessive guidance by their chiefs. The plan was to collect baseline, mid-line, and endline survey data and to collect qualitative data as well. Towards the end of the planning process, baseline measures were collected (Stark et al. 2013) using the survey and also intervention-specific measures.
In both districts, the intervention cluster chose to address teenage pregnancy through a mixture of family planning, sexual and reproductive health education and life skills. To build community capacities for the intervention, trainings were provided by Marie Stopes and Restless Development in Bombali and by Restless Development in Moyamba. High levels of ownership were achieved because the communities themselves created an inclusive planning process, defined the problem to address, chose how to address it and implemented the intervention. Government collaboration occurred through the District Ministry of Health providing contraceptives, training health post staff how to use implants, and having health staff contribute to education around issues of puberty, sexuality, pregnancy and pregnancy prevention. The foundation of the intervention was community action, including: role plays by teenage girls and boys followed by discussions; parent–child discussions of puberty, sex and pregnancy; creation of and transmission by teenagers of youth-oriented messages about preventing teenage pregnancy; ongoing community dialogues and reflection about teenage pregnancy; and support from health workers and authorities (Wessells 2015; Wessells, Manyeh and Lamin 2014). Table 1 provides an overview of the intervention elements.
Table 1: The main components of the community-driven intervention to reduce teenage pregnancy
|Collective dialogue, awareness raising and negotiation||In village meetings and sub-groups such as teenage girls, teenage boys, adult women, adult men, and elders discussed the main harms to children, which issue should be addressed, how to address the issue, and diverse aspects of teenage pregnancy. These dialogues raised collective awareness and created readiness to receive various messages associated with teenage pregnancy.|
|Collective decision-making, empowerment and responsibility||The communities made their own decisions about which issues to address, how to address it, etc. As a result, they saw the decisions and intervention process as 'theirs', and they took responsibility for insuring its success.|
|Linkage of communities with health services||The District Medical Office agreed to keep up the supply of contraceptives and train health post nurses to do procedures such as implants. Feeling supported by health staff, people visited the health post for contraceptives and invited nurses to visit the villages and help to educate people about puberty, reproductive health and pregnancy.|
|Peer education||Community-selected Peer Educators (including teenage girls and boys), trained by NGOs, helped to educate their peers on an ongoing basis. Informal peer education occurred also through everyday discussions in the community.|
|Use of culturally relevant media||Using song and drama, peer educators conducted culturally appropriate educational activities such as role plays followed by group discussions in which teenagers and adults discussed the benefits of good decisions made by young people, and the problems associated with bad decisions.|
|Child leadership and messaging||Girls and boys played leadership roles. In light of the fact that children talk in distinctive ways, children created their own messages based on what had been learned in NGO-led workshops and discussions with health workers.|
|Inclusion and outreach||Representatives of diverse sub-groups took part on a task force that facilitated much of the work to prevent teenage pregnancy. To include marginalised people such as children with disabilities, the task force members and peer educators made home visits on a regular basis.|
|Parent–child discussions||Rejuvenating an older practice that the war had disrupted, parents and children discussed issues of puberty, sexual and reproductive health, sex, and teenage pregnancy prevention. In some cases, the children were better informed than adults and helped to correct parental misconceptions.|
|Role modelling||By taking part in activities such as drama and singing songs, young people, including teenage boys, signalled that they wanted to prevent teenage pregnancy. Similarly, parents provided role models for each other in talking constructively with their children about teenage pregnancy.|
|Legitimation by authority||The paramount chiefs publicly supported the importance of preventing teenage pregnancy and encouraged people to get involved in the intervention. Other community leaders such as teachers and religious leaders also encouraged support for preventing teenage pregnancy.|
The intervention began in March–April 2013, and the mid-line effects of the intervention were assessed in 2014 using the quantitative survey (Stark et al. 2014) and qualitative findings from key informant interviews and a community self-assessment (Wessells et al. 2014). As shown in Box 1, the midline results were promising and featured high levels of community ownership and diverse signs of the intervention effects in addressing teenage pregnancy. However, the results were preliminary in that more time was needed to see fully the effects of the intervention. Also, some of the effects visible from the community descriptions and qualitative data were not triangulated fully with the quantitative data. It was hoped that the subsequent endline measures would allow full triangulation and analysis of the results, including systematic comparisons with the control clusters.
Unfortunately, the eruption of the Ebola crisis in Sierra Leone in August 2014 disrupted the intervention and also made it impossible to collect the endline survey data as had been planned. Reports from the field indicated that the Ebola crisis had introduced a host of confounding variables and threats to children, including increases in teenage pregnancy. Without knowing more about the status and characteristics of the intervention, what the confounding variables were, and how conditions that affect teenage pregnancy had changed, it would have been impossible to interpret the endline survey data in a meaningful way. For these reasons, a decision was taken to replace the endline survey with open, ethnographic learning that would illuminate the situation in the communities. This ethnographic learning took place in November–December 2015 (Kostelny et al. 2016) and is not discussed here since the focus is on the community-driven intervention and its wider impact.
1.4 National impact
Notwithstanding the impact of the Ebola crisis, the inter-agency research approach and findings, which converged with the findings of other studies (e.g. Child Frontiers 2010), enabled the action research to have a significant influence on the national policy to support vulnerable children in Sierra Leone. The findings that local people relied mostly on family and community mechanisms and that community-owned processes were effective even in addressing challenging issues such as teenage pregnancy argued in favour of a policy that emphasised the importance of supporting existing family and community mechanisms. At the same time, research conducted by Harvard University with UNICEF indicated that local people were more likely to report severe violations against children to two people acting as focal points who had been chosen by the community and trained for their work.
The preliminary results included positive outcomes related to child protection, the community process and system strengthening.
Community ownership. High levels of community ownership were evident in how many people volunteered their time and work, without material compensation, and regularly referred to the intervention as ‘ours’, stating that NGOs and the government support them but do not lead the intervention.
Non-formal–formal linkage and collaboration. The intervention process significantly improved communities’ collaboration and linkage with the local health posts. In contrast to previous low use of health posts, many teenagers and/or their parents visited the health posts regularly for contraceptives or advice. Villages frequently invited nurses and other health staff to visit in order to educate villagers about puberty, sex and preventing teenage pregnancy.
Contraception. The district medical officers fulfilled their promise to supply the contraceptives and train the health staff. Relative to the comparison condition, teenagers in the intervention communities reported increased intent to use condoms regularly and increased willingness to ask their partners to use a condom. These can be precursors of wider changes in behaviour and social norms related to sex.
Life skills. Teenage girls reported that because of the intervention, they said ‘No’ more frequently to unwanted sex. Both girls and boys said that they had learned how to discuss and negotiate with their partners about sex, and also how to plan their sexual activities in light of wider life goals. In addition, boys said openly that they had a responsibility to prevent teenage pregnancy, which contrasted sharply with the boys’ previous behaviour.
Teenage pregnancy. Participant observations and interviews with health post staff, monitors, teenagers and adults indicated a significant decrease in teenage pregnancies. In the intervention communities in both districts, participants reported that in an average school year (September–June) before the intervention had begun there were five or six teenage pregnancies per village. In contrast, in the 2013/14 school year, half the communities reported no new teenage pregnancies, and the other half reported only one new teenage pregnancy. Grandmothers, who are respected community figures, assured that it was impossible to hide pregnancies in the villages.
Spin-offs. Participants said that school dropouts had decreased. Also, some villages had spontaneously begun to discuss the problem of early marriage. Having learned more about the adverse effects of teenage pregnancy, villagers had begun to question the appropriateness of any teenage pregnancy and also of early marriage.
Encouraged by these findings, the GoSL and UNICEF decided to develop a new policy that placed support for families and communities at the centre and avoided the ‘add a structure’ approach that governments frequently take in addressing problems. To support the drafting and development of a new policy, UNICEF hired Child Frontiers, the consulting group that had led the initial mapping of the child protection system in West Africa. However, the development of the new policy was interrupted by the Ebola crisis beginning in July 2014 and also hampered by turnover in the Ministry of Social Welfare Gender and Children’s Affairs. Nevertheless, the GoSL enacted in December 2015 a new Child and Family Welfare Policy that embodied the insights from the inter-agency action research. Ultimately, the GoSL listened to the research because it saw it as its own and as addressing the questions that were at the heart of its efforts to support vulnerable children.
The implementation of the new policy faces challenges related to scale, cost and the capacities of different partners to enable effective implementation. Via UNICEF, a technical unit of four agencies that had been very active in the research has been convened to plan and prepare for the roll-out of the new policy using the methods and approach of the research. The plan is to go to scale in a measured approach that enables learning about capacity building and implementation on a continuing basis. Initially, the approach will be extended throughout Moyamba and Bombali Districts through partners that have been trained in how to facilitate the community-driven approach. Subsequently, the community-driven approach will be extended to cover all 14 districts. In this manner, UNICEF, the GoSL and the research team hope to address the frequently expressed concern that bottom-up approaches have difficulty going to scale. Collectively, this work will transform the strictly top-down approach to child protection system strengthening towards the mixture of top-down and bottom-up approaches that are needed for building a system that effectively enables children’s protection and wellbeing.
Lessons learned on how to move from research to impact
A key question is what enabled the research to achieve a national impact? Broadly, five key factors are discernible: a collaborative inter-agency approach, early engagement with diverse actors, vertical engagement with the key government ministry, collective reflection on the implications of the research, and the management of power dynamics. These factors are examined below in the form of lessons learned that aim to highlight the practical implications for other research projects. Although the lessons are discussed individually, their interconnectedness should also be recognised.
2.1 Lesson 1: Use a collaborative, inter-agency approach at all stages of the research
A key lesson was that inter-agency collaboration contributed to a sense of collective ownership in the research. As discussed above, even the idea behind the action research originated in collaboration and discussion between key international agencies that work on child protection. The fact that the question being asked and the broad methodology had been worked out collectively meant that some of the main NGOs in the global child protection sector (e.g. UNICEF, Save the Children, Plan International, World Vision, ChildFund) saw the importance of the research and experienced a sense of collective ownership for it. This collective backing for the research probably increased its salience and perceived importance in Sierra Leone. Also, the sense of ownership felt by individual agencies such as Save the Children internationally probably trickled down to their Sierra Leone offices. The fact that the research was collective in nature may also have helped to calm the inter-agency rivalries that might otherwise have impeded the research.
An important part of the collaborative process early on was the establishment of an in-country Reference Group, which was coordinated by Save the Children in Sierra Leone. Its purpose was to develop a collaborative, interagency approach in guiding and supporting the research. Different members such as UNICEF, Plan International, World Vision, ChildFund, Action Aid and Goal supported the research in diverse ways. Save the Children, for example, seconded one of its staff to the research team for the initial ethnographic phase of the research. Plan International provided financial support for the field testing of the survey instrument, and it provided the use of its guest house and office in Moyamba at various times. World Vision helped to support the intervention planning and development process. As agencies invested in the research, they also developed a sense of ownership for it.
Over time, the collaboration with UNICEF became increasingly important in areas such as resource sharing, door opening and logistics. At the request of the PI, UNICEF agreed to enable David Lamin to serve as one of the leaders of the national research team. Lamin went on to become the main strategist and actor who orchestrated the national team building and collaboration with government and civil society partners that underpinned the impact of the research. UNICEF also provided financial support, for example, for the initial baseline survey. In addition, UNICEF opened the doors with the government actors at district level and with UNICEF regional staff who advised on how to contextualise the research methodology. With respect to operations, an ongoing financial challenge was the rising costs of fuel and vehicle rentals. Fortunately, UNICEF provided on multiple occasions the vehicles and drivers that were needed to transport research teams during the research.
An important part of the collaborative process that enabled national impact was a collaborative approach to co-learning with and influencing multiple stakeholders on an ongoing basis. As described below, the process of ongoing engagement, mutual learning and reflection on how to improve policy and practice was probably as important as were particular events and decisions. This co-learning orientation helped to bring forward the insights of different actors and also avoided making agencies or policy leaders feel that somehow an outside group of researchers was imposing its own views.
2.2 Lesson 2: Promote early engagement with diverse actors on a national level
A second lesson learned was the importance of early engagement, both horizontal and vertical, with diverse actors in Sierra Leone. Typically, engagement strategies are guided by a stakeholder analysis that identifies relevant stakeholders, their relative power, and appropriate means of engaging with them. Fortunately, in Sierra Leone UNICEF had in practice already conducted a stakeholder analysis, was highly knowledgeable about various stakeholders, and was engaged with different key actors on an ongoing basis. Via David Lamin, UNICEF helped to develop an early engagement strategy that would help to contextualise the research and cultivate collective ownership for it.
As recommended by UNICEF, two key actors to engage with were the national Child Protection Committee (CP Com) and the Ministry of Social Welfare Gender and Children’s Affairs (MSWGCA), which serves as Chair of the CP Com. The CP Com is an influential group because its members include UNICEF, international NGOs and Sierra Leonean groups that lead the national work on child protection and could help to contextualise and guide the research. In many respects, the CP Com is the national ‘brains trust’ on child protection in Sierra Leone, and it offers influential advice on policy and practice issues. The MSWGCA is the lead government agency in Sierra Leone on issues related to child protection. It oversees work on strengthening the national child protection system, including steps to improve policy and practice regarding children’s protection and wellbeing.
With UNICEF support, Sarah Lilley and Michael Wessells had an initial two-hour meeting with the CP Com in late 2010 that was chaired by the Minister of the MSWGCA and was important both for its process and its outcomes. The process included a UNICEF briefing on the research with the minister and key actors on the CP Com. This pre-briefing was essential in helping key people to understand the research and see its potential relevance to the agenda of the CP Com. The process of the CP Com meeting itself was characterised by participatory dialogue and a spirit of mutual learning. Important elements included exploration of the potential value of the research, critical dialogue about why the research would focus only on a few areas when the needs in the entire country were severe, and how the approach could be adapted to the Sierra Leone context.
A significant outcome was that the CP Com, including the minister himself, expressed support for the research, saying it would help them to learn more deeply and find better ways of protecting children. In addition, the CP Com members agreed that the research should be conducted in two areas: Moyamba District within the Mende-speaking southern area and Bombali District within the predominantly Temne-speaking northern area. These two areas were regarded as broadly typical of Sierra Leone, which remains a primarily agricultural society. Further, CP Com members whose agencies worked in Moyamba and Bombali districts agreed to provide advice and operational support for the effort to identify within each district two non-contiguous, similar chiefdoms where the research would be conducted. Overall, the early discussion about sites and methodology planted the seeds for ongoing collaboration and also built a sense of collective ownership of the action research.
2.3 Lesson 3: Organise ongoing engagement with the relevant government ministry at multiple levels
A key strategy for achieving a national impact was to influence the MSWGCA, since it played such an important role in the protection and wellbeing of vulnerable children. A critical lesson learned in this research was the importance of ongoing engagement at multiple levels of the Ministry. Not infrequently, researchers focus primarily on getting the attention of the minister, yet it can be equally important to cultivate strong relations with mid-level managers in the ministry.
The engagement with the MSWGCA regarding the research took two forms, the first of which consisted of meetings between the minister and David Lamin and the PI, Michael Wessells. UNICEF brokered an early meeting with the minister that explained the research in greater depth, discussed its potential significance for policy and practice, and suggested how the MSWGCA could help to support it at both district and national levels. Similar meetings of the minister with Lamin and the PI continued to occur every six months.
Inter-agency larger group meetings were the second means of vertical engagement with the MSWGCA regarding the research. Following each main stage of the research, members of the research team met with the national CP Com, chaired by the Minister of the MSWGCA. Typically, the PI made a brief research update, sometimes accompanied by a slide presentation, followed by open discussion of the implications of the findings. A primary example was a meeting of Wessells and Lamin with the CP Com following the ethnographic phase in 2011. A slide presentation on the findings stimulated animated discussion about the value of open-ended learning and the importance of trust between researchers and community members. The discussion of the findings could have been a very tense moment since the findings indicated significant limitations of the Child Rights Act and the CWCs. However, the minister himself commented that they had received some reports of problems in the child protection system at community level and that this research had confirmed those reports and provided a more systematic understanding of the problem. The minister’s words were pivotal in establishing respect for evidence, even when the findings contravene existing policies. They also opened the door to trying new approaches for supporting vulnerable children.
A challenge, however, was the rapid turnover in the post of minister (three different ministers served during the period 2011–14). If the effort to gain the MSWGCA’s support had focused exclusively on the minister, a change in ministers could have caused serious discontinuities or loss of support for the research. In addition, the minister had too many responsibilities to be able to do the follow-up work needed to fully take on board the approach and its implications. For both these reasons, emphasis was placed also on meeting regularly with and cultivating the support of the senior managers within the MSWGCA who enjoyed longer-term posts and were positioned to help it to achieve its intended impact.
When a new minister arrived, David Lamin met with him and briefed him on the research, what it was finding, and how it had been helping the ministry, UNICEF and partners achieve their goals related to child protection. When Wessells visited, Lamin organised a meeting with the minister that helped to enable understanding of and support for the research and to consider its policy implications. The PI deliberately avoided pushing too hard on the latter because UNICEF was best positioned to use the research to advocate for policy changes and improvements. Nevertheless, a key message was that communities themselves are significant, functional actors in the child protection system who need support in their work on behalf of vulnerable children. This message pointed in a different direction than the extant GoSL priorities of forming and capacitating CWCs. On an ongoing basis, UNICEF reinforced this message and influenced the GoSL to make more space for and to prioritise community action in support of vulnerable children.
In addition, Lamin and the research team worked closely with senior managers within the MSWGCA to cultivate understanding of and support for the research, including a willingness to identify ways of enabling the GoSL and the MSWGCA to prioritise community action and also provide more space for it in practice. A senior management team from the GoSL and UNICEF participated in a key regional meeting on national child protection systems in West Africa (Davis et al. 2012). This meeting provided a platform for discussing community-driven action and helped other African governments learn about the bottom-up approach pioneered in Sierra Leone. As the senior managers became supporters, they provided a valued source of continuity within the MSWGCA and a consistent voice for the importance of community-led action on behalf of children.
2.4 Lesson 4: Support multiple partners in learning about and taking on board the approaches, methods and tools of the research
The impact of the action research was due also to the fact that multiple agencies – not just the MSWGCA – had developed a solid understanding of and support for the research approach, methods and tools. This understanding and support was achieved by means of inter-agency workshops, sharing of findings, tools and approaches and, above all, the creation of reflective space.
Regular meetings with the CP Com served these functions of sharing and collective learning and reflection. For example, the meeting with the CP Com following the baseline data collection sparked discussion of the value of blending qualitative and quantitative work, and also of using population-based approaches to measurement like those found in the field of public health (Wessells 2014). Also, the CP Com meetings provided reflective space in which busy practitioners were able to step back and reflect on the advantages on a community-driven approach. The group reflections whetted the appetites of different agencies to delve more deeply into the methods, tools, and approaches of the action research.
To meet the desires for ongoing learning, single agency and multi-agency reflective workshops were conducted to help different stakeholders understand more fully the tools, approaches and current findings of the research. Taking a non-didactic approach, these workshops provided space for reflection on the implications for how the agency partners conducted their work. For example, in August 2013, Save the Children and UNICEF convened an inter-agency workshop for 35 people that reviewed the ethnographic outcomes study and baseline phase approaches, tools and findings. The emphasis, however, was on the implications for the work of the eight partner agencies that participated. Different agencies were asked to describe how they typically evaluated child protection programmes and to identify how the action research work differed from their usual mode of doing assessments or baseline studies. Together they reflected on what they would like to change in their own work based on the more grounded, participatory approach to learning inherent in the action research. Animated discussions centred on the importance of building trust in learning deeply from communities, taking a non-judgemental position early on, learning and asking about local categories and understandings rather than asking only outsider-constructed questions, and the value of using mixed methods.
Because UNICEF was such a central partner in the action research, visits to Sierra Leone by the PI were frequently used as occasions for him to make a presentation or give an update on the research. The discussions that followed frequently involved group reflection on the current approaches to child protection systems strengthening in Sierra Leone and their limits, and how the action research and approaches could help to lead to different approaches that would strengthen child protection systems in a more effective, sustainable manner. The topics discussed included the limits of top-down approaches, the value of linking community mechanisms and with formal stakeholders, the importance of building local ownership at all levels, and the value of an intersectoral approach that engaged not only the social welfare sector but also health, education and other sectors. These meetings also enabled reflection on the value of policies that placed less emphasis on structures such as CWCs and greater emphasis on supporting the families and communities that did the ‘heavy lifting’ in regard to children’s protection and wellbeing.
Collectively these workshops and discussions both embodied and enhanced the spirit of collective ownership and mutual learning that were at the heart of the action research. Because they were part of an ongoing collaborative process, different agencies saw the methods, findings, tools and approaches as relevant and as having implications for how to reorient or enrich child protection work. As participants reflected together, they took on board particular findings or approaches and thought critically about how to strengthen and transform not only their work but also the collective work on child protection in Sierra Leone.
2.5 Lesson 5: Work with a broker to understand and manage power dynamics
The analysis and management of power dynamics was key to the success of this action research. To achieve a positive impact, it was essential to understand the different stakeholders and focus limited human resources on engaging with the appropriate actors in ways that would most likely contribute to a positive impact. UNICEF played a pivotal role in regard to both points. UNICEF knew that the government and the CP Com would probably be receptive to learning more about and strengthening community-driven approaches because previous work on mapping the national child protection system had suggested the existence of a gap between community processes and the government-led aspects of the child protection system (Child Frontiers 2010). In particular, local people preferred to rely on community processes rather than on CWCs. To follow up on that finding, a logical next step was to conduct more systematic enquiry into the nature and origins of the gap, which this research was able to provide. As one UNICEF worker put it, ‘the research fell into fertile earth’. Also, UNICEF understood the importance of bringing in both the MSWGCA and the CP Com since the ministry would be more likely to move in new directions when it had the support of its main national partners. These understandings, together with David Lamin’s skilful door opening and relationship building, enabled the researchers to concentrate their energies on the people and agencies who were at the centre of power on issues pertaining to child protection.
On an ongoing basis, UNICEF also served as key adviser and intermediary in managing power relationships. Understanding how perceptions of exclusion or privileging particular agencies can derail a collaborative process, Lamin helped to navigate the inter-agency environment and to insure that the inter-agency process and workshops were respectful of diverse partners. For example, there was little, if any, privileging of particular agencies or of expatriates in the inter-agency workshops, which included and were respectful of different agencies and Sierra Leonean workers. The fact that different agencies had a seat at the table and an equal voice created an environment of mutual respect and trust, both of which are necessary for reducing the competitiveness and perceptions of privileging that can be harmful. Further, the emphasis on Sierra Leonean voices and views was critical in reducing perceptions that the community-driven approach was somehow an external imposition.
Since UNICEF was a key broker of relationships, an important question is how power relations with UNICEF were managed. In short, they were managed through a consultative, trustful process that involved significant leadership by David Lamin. As both a research team member and a key UNICEF staff member, Lamin was in a position to balance the interests of the research with the longer-term, multifaceted UNICEF agenda. Since he saw the importance of the research for Sierra Leonean children, the government and UNICEF, he was not disposed towards backgrounding the research or allowing its interests to be eclipsed by UNICEF’s wider agenda. Having seen the equanimity and good judgement Lamin exercised, Wessells regularly turned to him for advice, which Wessells then followed.
When issues arose that could have strained the relationship with UNICEF, Wessells acceded to Lamin’s judgement on how to address the issue. For example, when the ethnographic findings were about to be fed back to the CP Com and the MSWCGA, Wessells became concerned that the ministry might react defensively or see the presentation as disrespectful since it showed the failure of the CWCs. Having greater knowledge of the minister and having had prior discussions with him of the findings of the child protection systems mapping, Lamin advised presenting the data in a straightforward manner. Wessells followed this advice with a positive outcome, as noted above. Later in the research, when Wessells had learned that there were discussions under way about a new Child and Family Welfare policy, he asked Lamin before a meeting with the minister whether and how to advocate on behalf of a policy change that provided greater support for community action. Lamin’s counsel was that UNICEF was already promoting such a change and that Wessells did not need to actively lobby but should only mention what the research was finding. In Wessells’ view, there was no need to negotiate this issue further since Lamin was clearly in the know and there was excellent convergence between the UNICEF agenda and the research agenda.
Needless to say, it does not always happen that relations between researchers and a large agency with its own interests, such as UNICEF, go so smoothly. Perhaps the main implication, though, is that researchers who want to have a significant impact on policy should identify and cultivate a positive relationship with a well-positioned person who can serve as both a power broker and a trusted adviser.
Traditionally, much research is done by a single agency that designs a study and its methodology, collects and analyses the data, and then presents the results afterwards to policymakers and practitioners with a request for changes in policy and practice. Although this approach has value, it is limited by low levels of inter-agency collaboration and collective ownership. In many cases, neither practitioners nor policy leaders will see the relevance of the research or view it as sufficiently important to change their own practice. This approach may also leave policy leaders and practitioners wondering whether the research group had bothered to learn about their priorities and strategies or why no or little effort was made to engage deeply with them in advance. This approach may also leave policy leaders and practitioners feeling that they have been disrespected, with the result that the research is left sitting on the shelf.
This case study highlights a very different approach that features collaboration and collective ownership at all phases of the research, leading to more positive impact. Indeed, this case study underscores that the social processes around the research may influence its impact as much as the technical merits and the findings of the research. However, it would be misguided to suggest that such a collaborative approach is a ‘silver bullet’ that will boost impact in all situations. A collaborative approach may not be possible if the research agency or an NGO or UN agency that commissioned the research wants to pursue its own agenda or claim the glory for itself. Also, circumstances could lead governments not to take a collaborative approach. Still, a truly collaborative approach with high levels of collective ownership may yield higher impact and build the coordination that is critical to the success of efforts to strengthen child protection systems. Perhaps the time has come to give increased attention to moving beyond the priorities of one’s own agency and working in a more collaboration manner. Ultimately, a collaborative approach can help to achieve what ought to be an overarching global priority: realising child rights even in challenging circumstances.
African Child Policy Forum; African Network for the Prevention and Protection against Child Abuse and Neglect; Environnement et Développement du Tiers-monde… World Vision International (2013) Strengthening Child Protection Systems in Sub-Saharan Africa: A Call to Action, Dakar: African Child Policy Forum
Behnam, N. (2011) ‘Awkward Engagement: Friction, Translation and Human Rights Education in Post-Conflict Sierra Leone’, unpublished doctoral dissertation, University of Pennsylvania
Boothby, N.; Strang, A. and Wessells, M. eds. (2006) A World Turned Upside Down: Social Ecological Approaches to Children in War Zones, Bloomfield CT: Kumarian Press
Bronfenbrenner, U. (1979) The Ecology of Human Development, Cambridge: Harvard University Press
Child Frontiers (2010) Mapping and Assessing Child Protection Systems in West and Central Africa: A Five-Country Analysis Paper, Bangkok: Child Frontiers
Child Protection Working Group (2012) Minimum Standards for Child Protection in Humanitarian Action, Geneva: Child Protection Working Group
Davis, R.; McCaffery, J. and Conticini, A. (2012) Strengthening Child Protection Systems in Sub-Saharan Africa: A Working Paper, Dakar: Inter-agency Group on Child Protection Systems in sub-Saharan Africa
Fernando, C. and Ferrari, M. (eds) (2013) Handbook of Resilience in Children of War, New York: Springer
Freymond, N. and Cameron, G. (eds) (2006) Towards Positive Systems of Child and Family Welfare: International Comparisons of Child Protection, Family Service, and Community Caring Systems, Toronto: University of Toronto Press
Kostelny, K.; Lamin, D.; Manyeh, M.; Ondoro, K.; Stark, L.; Lilley, S. and Wessells, M. (2016) ‘Worse than the War’: An Ethnographic Study of the Impact of the Ebola Crisis on Life, Sex, Teenage Pregnancy, and a Community-Driven Intervention in Rural Sierra Leone, London: Save the Children
Krueger, A.; Thompstone, G. and Crispin, V. (2013) ‘Learning from Child Protection Systems Mapping and Analysis in West Africa: Research and Policy Implications’, Global Policy 5.1: 47–55
McKay, S.; Veale, A.; Worthen, M. and Wessells, M. (2011) ‘Building Meaningful Participation in Reintegration among War-Affected Young Mothers in Liberia, Sierra Leone and Northern Uganda’, Intervention: International Journal of Mental Health, Psychosocial Work and Counselling in Areas of Armed Conflict 9.2: 108–24
Stark, L.; Macfarlane, M.; King, D.; Lamin, D.; Lilley, S. and Wessells, M. (2014) A Community-Driven Approach to Reducing Teenage Pregnancy in Sierra Leone: Midline Evaluation Brief, London: Save the Children
Stark, L.; Muldoon, K.; Lilley, S.; King, D.; Hotanga, P.; Lamin, D. and Wessells, M. (2013) Change that Counts: Baseline Report for the Evaluation of an Action Research Intervention to Strengthen Community Based Child Protection Mechanisms in Sierra Leone, London: Interagency Learning Initiative on Strengthening Community-Based Child Protection Mechanisms and Child Protection Systems
Stark, L.; Wessells, M.; King, D.; Lamin, D. and Lilley, S. (2012) A Grounded Approach to the Definition of Population-based, Child Protection and Well-being Outcome Areas, London: Interagency Learning Initiative on Community-Based Child Protection Mechanisms and Child Protection Systems
UNICEF; UNHCR; Save the Children and World Vision (2013) A Better Way to Protect ALL Children: The Theory and Practice of Child Protection Systems, Conference Report, New Delhi: UNICEF
Wessells, M.G. (2015) ‘Bottom-up Approaches to Strengthening Child Protection Systems: Placing Children, Families, and Communities at the Center’, Child Abuse & Neglect: The International Journal 43: 8–21
Wessells, M.G. (2014) The Case for Population Based Tracking of Outcomes for Children: Toward a Public Health Approach in Child Protection System Strengthening, New York: Columbia Group for Children in Adversity
Wessells, M.G. (2012) ‘Psychosocial Well-being and the Integration of War-affected Children: Toward a Community Resilience Approach’, in I. Derluyn, C. Mels, S. Parmentier and W. Vandenhole (eds), Re-member: Rehabilitation, Reintegration and Reconciliation of War-affected Children, Cambridge: Intersentia
Wessells, M.G. (2011) An Ethnographic Study of Community-Based Child Protection Mechanisms and their Linkage with the National Child Protection System of Sierra Leone, New York: Columbia Group for Children in Adversity
Wessells, M.G. (2009) What are We Learning about Protecting Children in the Community? An Inter-agency Review of the Evidence on Community-based Child Protection Mechanisms in Humanitarian and Development Settings, London: Save the Children
Wessells, M.G.; Lamin, D.; King, D.; Kostelny, K.; Stark, L. and Lilley, S. (2015) ‘The Limits of Top-down Approaches to Managing Diversity: Lessons from the Case of Child Protection and Child Rights in Sierra Leone’, Peace & Conflict: Journal of Peace Psychology 21.4: 574–88
Wessells, M.G; Lamin, D.; King, D.; Kostelny, K.; Stark, L. and Lilley, S. (2012) ‘The Disconnect between Community-based Child Protection Mechanisms and the Formal Child Protection System in Rural Sierra Leone: Challenges to Building an Effective National Child Protection System’, Vulnerable Children and Youth Studies 7.31: 211–27
Wessells, M.G; Manyeh, M. and Lamin, D. (2014) Preliminary Results of the Community-Driven Intervention to Reduce Teenage Pregnancy in Two Districts of Sierra Leone: Findings From Participatory Review Workshops With Children and Adults, London: Save the Children
Wulczyn, F.; Daro, D.; Fluke, J.; Feldman, S.; Glodek, C. and Lifanda, K. (2010) Adapting a Systems Approach to Child Protection: Key Concepts and Considerations, New York: UNICEF
*The authors wish to acknowledge: the Oak Foundation and an anonymous donor for their generous funding of the intervention and the pre-intervention phases of the research; Plan International and World Vision for their support for field testing and validating the survey on outcomes for children; UNICEF/Sierra Leone for co-funding the baseline measurement and intervention and for supporting all phases of the research; and the ESRC-DFID Poverty Alleviation Fund for its generous funding for the mid-point and endline data collections and analyses. We thank the Government of Sierra Leone and the national Child Protection Committee for their support and openness to change. We thank Save the Children for its financial support and its leadership of the global Reference Group. Most of all, the authors thank the many Sierra Leonean people and communities who did inspiring work to protect their children.
 For both ethical and practical reasons, the action research is currently being extended to the former comparison communities with support from the Oak Foundation.