International Journal of Child, Youth and Family Studies (2022) 13(2-3): 123–142 DOI: 10.18357/ijcyfs132-3202221117 LOCATING STATE ACTORS IN VIOLENCE AGAINST CHILDREN (VAC) NETWORKS IN KENYA: A COMPLEXITY LEADERSHIP LENS Jacqueline Nassimbwa, Doris M. Kakuru, and Malcolm T. Mpamizo Abstract: Kenya has made significant efforts to address violence against children (VAC), but its prevalence remains high. Networking of different actors has shown evidence of benefit in some sectors, but determining its effectiveness in addressing VAC has not received due scholarly attention. We conducted qualitative research, including a desk review, focus group discussions, and interviews. In this article, we apply complexity leadership theory to illuminate the types of networks involved and the influence Kenya’s government actors can exert towards eliminating VAC. We found that these actors operate through structured and unstructured networks. The latter are mainly grassroots responders who work voluntarily. The complexity leadership theory postulates that leadership influence is exercised through key functions, which are reflected in the two types of networks. The political– administrative function in Kenya is shaped by law; we show how it transforms other networks via an adaptive function. An enabling function is executed through enforcing policy, monitoring, and other methods, while a dissemination function involves the translation of ideas into policy, such as the transformation of Childline Kenya, a grassroots organization, into the National Child Helpline. We conclude that government should strengthen child rights networking by building more technical and financial capacity for this role. Keywords: violence against children, child abuse, child rights networking, civil society networks, children, complexity leadership theory Jacqueline Nassimbwa MSc (corresponding author) is a master’s candidate in Child, Youth, Family and Community Studies at the University of Victoria, and a researcher at Nascent Research and Development Organization Uganda, P. O. Box 25382, Kampala, Uganda. Email: jacquelinenassimbwa@uvic.ca Doris M. Kakuru PhD is an associate professor in the School of Child and Youth Care at the University of Victoria, 3800 Finnerty Road, Victoria, BC V8W 2Y2. Email: doriskakuru@uvic.ca Malcolm T. Mpamizo LLM (International Human Rights Law & Policy) is a social justice lawyer and advocate at CivSource Africa, Plot 4310, Balikuddembe Road, Naguru, Kampala, Uganda. Email: mmpamizo@civsourcea.com International Journal of Child, Youth and Family Studies (2022) 13(2-3): 123–142 124 Violence against children (VAC) is a global health and human rights concern affecting over half of children aged 2 to 17 every year (World Health Organization [WHO], 2020), yet the public remains generally unaware of its magnitude and consequences (Mercy et al., 2016). Significant international and national commitments have been made to address VAC, including target 16.2 of the Sustainable Development Goals1 (SDGs), which explicitly aims to “end abuse, exploitation, trafficking, and all forms of violence against and torture of children” (United Nations Department of Economic and Social Affairs, 2015). Countries have adopted international commitments through national regulatory frameworks, programs, and strategies. The WHO’s 2020 global status report on VAC shows that 89% of countries assign the task of reducing VAC to multiple sectors, such as the education, health, justice, and social sectors (p. xi), and states that VAC prevention and response should include targeted interventions at four interrelated levels of risk — individual, interpersonal, community, and society. Horizontal interlinkages are found at the community level between children, parents, caretakers, families, and schools. Addressing VAC also entails vertical involvement of actors at the household, community, district or county, national, and global levels. Kenya has demonstrated commitment to addressing VAC primarily through the ratification of the United Nation’s Convention on the Rights of the Child (CRC; 1989) in 1990 and, in 2001, the African Charter on the Rights and Welfare of the Child (ACRWC; Organization of African Unity, 1990). The latter is the overarching legal guidance for children’s rights in the African region; it promotes an approach to child protection that involves strengthening systems such as laws, policies, regulations, and services that are needed throughout the various social sectors. Government institutions at the national and subnational levels bear the responsibility for laws, policies, and regulations, while civil society actors and community groups deliver services (Organization of African Unity, 1990). The CRC was adopted into the Kenyan constitution of 2010, which includes provisions regarding child protection (Republic of Kenya, 2010). The Children Act, 2001 elaborates on the constitutional provisions for children’s rights and protection and clearly states that it adopts its mandate from the principles of both the CRC and the ACRWC (Republic of Kenya, 2001). The constitution legitimizes the National Children Policy (National Council for Children’s Services [NCCS], 2010), which aims to realize and safeguard the rights and welfare of children. Later policies include the National Social Protection Policy (Ministry of Gender, Children, and Social Development, 2011), the National Policy on Elimination of Child Labour (Ministry of Labour, Social Security and Services, 2016), and the National Plan of Action for Children in Kenya (2015–2022; NCCS, 2015). 1 The SDGs are: (1) no poverty; (2) zero hunger; (3) good health and well-being; (4) quality education; (5) gender equality; (6) clean water and sanitation; (7) affordable and clean energy; (8) decent work and economic growth; (9) industry, innovation, and infrastructure; (10) reduced inequalities; (11) sustainable cities and communities; (12) responsible consumption and production; (13) climate action; (14) life below water; (15) life on land; (16) peace, justice, and strong institutions; and (17) partnerships for the goals. See https://sdgs.un.org/goals. International Journal of Child, Youth and Family Studies (2022) 13(2-3): 123–142 125 Specific to VAC networking, Kenya has a National Council for Children’s Services (NCCS), which oversees the supervision and control of planning, financing, and coordinating child rights and welfare issues. The NCCS is replicated at the subnational level by county area advisory councils (AACs) in 47 counties, 229 subcounties, divisions, and other devolved structures. The NCCS receives reports, and funds and advises the AACs. The NCCS was designed to be multisectoral, reflecting the fact that child protection issues were spread over diverse government ministries. However, the participation of the different ministries was not defined. The lack of statutory guidance on their roles and the consequent deficit of official accountability meant that line government ministries worked in silos, with duplication of roles. The National Child Protection System framework (NCCS, 2011) was developed in 2011 to address this problem by promoting connectivity among stakeholders. This marked the beginning of VAC networking among state actors. In this article, networks are perceived as “social arrangements made up of individuals and representatives of institutions based on establishing and building relationships, sharing tasks and working on mutual or joint activities, enabling new learning and mobilizing alternative action” (Younis, 2017, p. 2). The Kenya National Child Protection System framework acknowledges the complexity of child protection and the need for multidisciplinary and multisectoral actors. The framework facilitates coordination of the various interventions and actors in VAC prevention (NCCS, 2011). The Department of Children Services2 (DCS) is another government institution designed to work through a collaborative approach. One of its seven divisions is focused on child protection; it develops VAC-related programs and works with NCCS to develop policies and guidelines to address VAC. The DCS participates in child protection technical working groups, committees, and alliances. Considering that all these derive guidance from international instruments that promote a system-strengthening approach, one can adduce that networking to address VAC is supported by the Kenyan legal system. The Kenyan government has been recognized both as leading in the care and protection of children at risk in Africa and as unwilling to ensure their care and protection (Chege & Ucembe, 2020; Cooper, 2012). Kenya has a rich legal and institutional framework to address VAC, but the operationalization of laws and policies is limited by various factors attributable to social and economic inequalities (Wangamati et al., 2019). The heavy political investment that has been made is unfortunately matched with poor service-seeking behaviour and low uptake of services (Annor et al., 2022). Therefore, the prevalence of VAC in Kenya is still high, although Annor et al. (2022) pointed out that the 2019 VAC survey shows an improvement in indicators from its predecessor in 2010. For example, the prevalence of any form of lifetime sexual violence among females (13–24 years) dropped from 36.2% in 2010 to 25.2% in 2019, while those who reported any physical violence went from 28.4% to 16.8% over the same period. The VAC survey, published by the Ministry of Labour and Social Protection, Department of Children’s Services (2019), 2 https://www.socialprotection.go.ke/children-services/ International Journal of Child, Youth and Family Studies (2022) 13(2-3): 123–142 126 showed that among children aged 18 to 24 more boys (56.1%) than girls (45.9%) reported having experienced childhood violence (p. 8). However, girls reported a higher risk of sexual violence: 15.6% of girls and 6.4% of boys experienced sexual violence before turning 18, and 62.6% of girls who were sexually abused had experienced multiple incidents (p. 8). Physical violence is the most common type of VAC, experienced by 51.9% of boys and 38.8% of girls (p. 8). Some recent studies on VAC in Kenya seem to point to the importance of strengthening collaborative efforts between various actors in the child protection space (e.g., Undie & Mak’anyengo, 2020). Examples of such studies include some that have examined the topic of sexual violence, including its drivers (Wangamati et al., 2018), its prevalence (Russell et al., 2020), the kinds of perpetrators and the context (Mwangi et al., 2015), and health providers’ post-rape care knowledge, attitudes, and practices (Wangamati et al., 2020). Undie & Mak’anyengo (2020), in their study to understand screening for sexual violence against children in Kenya, collected data from children and parents through collaborating with schools and hospitals, working with parents, schoolteachers, child health care providers, and children. Their results show the importance of building collaborations with all stakeholders in child protection and speak to the extent of networking for VAC prevention and response. The existing literature paints a bleak picture of the magnitude of VAC in Kenya despite the numerous child protection efforts put in place by the Kenyan government and partners. Ayaya et al. (2021) studied VAC among orphans and separated children and adolescents and found that, while they all reported high levels of abuse, street-connected participants had the highest incidence of all kinds of violence. A qualitative study conducted by Embleton et al. (2021) on access to health care for street-connected children also found that they had many needs arising from child neglect. The study recommended multidisciplinary, holistic, and community-based approaches to care. Wangamati et al. (2018) explored community perceptions of drivers of child sexual abuse. They found that participants attributed child sexual abuse to numerous factors, including poverty, girls’ provocative clothing or behaviour (survivor blaming), “developmental stage, peer pressure, huge gender disparities exacerbated by negative social norms and cultural practices, the HIV epidemic and social media platforms that circulate sexualised images” (p. 1394). This is further evidence that a multipronged approach is needed to successfully address VAC in the communities. The importance of forging solid collaborations between various state and non-state child protection actors has been well acknowledged in multiple studies (e.g., Badoe, 2017; Devaney & Byrne, 2015; O’Leary et al., 2015). However, VAC networking has often been perceived as a reserve of not-for-profit actors. The work of Awortwi (2018) recognizes civil society and grassroots networks as having an advantage over state actors in addressing matters that concern grassroots communities, such as social protection. On the other hand, Tshimpaka et al. (2021) viewed civil society networks as well-governed, self-organized entities that function in the same way as the formal arrangements of government technocrats. International Journal of Child, Youth and Family Studies (2022) 13(2-3): 123–142 127 Although most of the existing literature shows that both state and non-state actors play significant roles in addressing VAC, Wessells (2015) found a disconnect between state and non- state child protection efforts. Hence, while there appears to be some connectivity within state actors or non-state actors as shown by Awortwi (2018) and Tshimpaka et al. (2021), the types and functionality of networking for VAC have not received due scholarly attention. In the existing literature, there is a lack of studies analyzing the nature of efforts to address VAC through collaborations or networks among state actors, and between state and non-state actors. This article aims to illuminate the types of networks that have formed, and the role of government actors as the ultimate duty-bearers for the realization of global, regional, and national commitments toward eliminating VAC in Kenya. Theoretical Framework: The Complexity Leadership Theory Networks are complex and diverse, so the actors within are often uncertain of the magnitude or significance of their actions within networking arrangements (Scholten et al., 2015). The traditional leadership theory that assumes hierarchy or bureaucracy cannot take into account this complexity. Anchoring networks in complexity science instead would help us position government actors in this complexity, which changes over time. Coveney (2003) defined complexity science as the “study of the behaviour of large collections of … simple, interacting units … with the potential to evolve with time” (p. 1058). The interconnectedness, interactivity, fluidity, and diversity of a network means it can be defined as a complex adaptive system (CAS), “a complex interplay from which a collective impetus for action and change emerges when heterogeneous agents interact in networks in ways that produce new patterns of behaviour or new modes of operating” (Uhl-Bien et al., 2007, p. 299). In a CAS, leadership denotes not authority or position but an emergent dynamic (Uhl-Bien et al., 2007). In this article, networks for VAC in Kenya are regarded as CASs, and are used as the basic unit of analysis. The subunits in networks are interdependent actors (individuals or institutions) united by a shared goal. Each of these subunits brings unique creativity, learning, and adaptability. Leadership arises dynamically from the interactions of the different subunits (Scholten et al., 2015). The leadership framework in CASs is what has come to be known as the complexity leadership theory (CLT; Uhl-Bien et al., 2007) that anchors this article. CLT underpins our analysis as we explore the types of networks in which the government is an actor and examine how government leadership influences approaches to VAC intervention within the networks. With guidance from Uhl-Bien et al. (2007), we regard CLT as premised on: • The nature of interactions and interdependencies among people, bureaucratic divisions, environments, and organizations. This context is the foundation of a CAS and of its leadership. International Journal of Child, Youth and Family Studies (2022) 13(2-3): 123–142 128 • A distinction between leaders and leadership. The latter emanates from the interconnectivity of subunits leading to adaptive outcomes (adaptive leadership). Leaders are those individuals that influence the dynamics and outcomes. • A distinction between leadership and managerial positions, which are usually bureaucratic (also called administrative leadership). Managers produce orderly results efficiently, while leaders are a force for change (Meijerink & Stiller, 2013). CLT is also premised on three functions of leadership: administrative, enabling, and adaptive (Uhl- Bien et al., 2007). VAC is a global policy issue, and analysis of the leadership in VAC networks demands a framework suitable for examining CASs in political contexts. Thus, we use Meijerink and Stiller’s (2013) redefinition of the functions of leadership: political–administrative, adaptive, enabling, connective, and dissemination (p. 253). The political–administrative function is fulfilled by positional leaders, such as political leaders or managers within networks. It comprises top-down leadership that includes decision-making, and sourcing and allocation of resources; it targets a political context in which policy is made in the CAS (e.g., on climate change, or VAC). Although Meijerink and Stiller (2013) included “adaptive” as a function of leadership, they actually stated that it is not a leadership task; rather, the “adaptive function is an emergent property of the CAS” (p. 252). In our case, the adaptive function concerns the prevention of and response to VAC. The third leadership function is enabling, which entails creating conditions for a CAS, such as a VAC network, to create, disseminate, learn, and adapt. Positional leaders support this function by practising tolerance for diversity and variety, driving networking, or igniting urgent consideration of an issue. Examples of this function include tolerating deviations from policy, finding necessary resources for experiments, and getting buy-in from other government agencies. The fourth function is dissemination, which includes all actions taken to disseminate new approaches generated through the adaptive function of the framework. Political entrepreneurs (drivers of policy change) are essential for this function, and they link the informal networks to formal decision-making arenas. The last function is the connective one, which coordinates all connections horizontally and vertically. Methodology Study Design and Sites This study adopted an exploratory qualitative approach to enable interaction with the participants and generate in-depth perspectives on VAC networking (CivSource Africa, 2021). The study commenced with a desk review (later published by CivSource Africa, 2021), which was conducted to provide insights into the Kenyan VAC response network landscape and identify organizations to possibly engage in the study. Nairobi and Kisumu counties were selected as the study sites through purposive sampling. Specifically, Nairobi county was selected due to having International Journal of Child, Youth and Family Studies (2022) 13(2-3): 123–142 129 been characterized as suffering a high prevalence of VAC and having an informal settlement dynamic (Onyango & Tostensen, 2015), and because it has a mix of national and county government responses to VAC, and multiple organizations working in that space. Nairobi also hosts several network secretariats and donor communities. Kisumu county was selected for having a heavy burden of HIV and orphanhood, a high rate of teenage pregnancy, both rural and urban dynamics (Mkutu et al., 2019), and a concentration of community-based organizations (CBOs) and non-governmental organizations (NGOs) responding to VAC. Study Participants The selection of participants was done through purposive sampling and snowballing. After selecting the study sites, we referred to the CivSource Africa (2021) report to identify child protection organizations operating in these areas. We purposively selected both formal and informal network actors (formal organizations, and informal collaborations between individuals) that work to prevent or respond to VAC in Kisumu and Nairobi. After identifying these existing partnerships, we randomly reached out to and recruited members of the informal collaborations to participate in focus group discussions (FGDs). Earlier participants referred us to peers who were recruited in turn until the desired number of FGD participants had been achieved. We conducted three FGDs with members of informal networks from Nairobi (11 participants), Kisumu Central (8), and Kisumu East (9). From the FGDs, we confirmed the existence of formal VAC networks and invited their leads or representatives to participate in interviews. We conducted key informant interviews with representatives of three national-level, three county-level , and three subcounty- level networks. Data Collection We collected both secondary (from CivSource, 2021) and primary data (using FGDs and in- depth interviews). FGD data were collected using an FGD guide that two research assistants administered during face-to-face sessions. The guide was translated into Swahili (the local language), tested, and revised before being administered to the study participants. The FGD guide aimed to collect data on how VAC networking happens, who is involved, why it is essential, and current challenges and suggestions for the future. The 28 FGD participants were involved in child protection work at the grassroots level. The FGDs were conducted in Swahili at three different locations convenient for the participants, and were audio-recorded. Members of the research team took field notes that were helpful at the data analysis stage. Interview data were collected using an interview guide collaboratively developed by members of the research team. The interview guide was also translated into Swahili, tested, and revised before actual data collection. The interview guide aimed to collect data from network leads about why and how VAC networking happens, who is involved, networking challenges, and tipping points for VAC philanthropic work. Interviews were conducted face-to-face. All interview data were audio-recorded. Members of the research team also took detailed field notes that informed the data analysis process. International Journal of Child, Youth and Family Studies (2022) 13(2-3): 123–142 130 Data Management and Analysis All audio and text files were stored in a password-protected folder accessible only to the research team members. Audio files were transcribed into text and anonymized to conform to the confidentiality requirements. The data were then coded into emerging themes using ATLAS.ti qualitative data analysis software. A thematic analysis was conducted using an iterative approach (Glaser & Strauss, 1967), using data collected from the FGDs to inform the interview phase. The emerging themes were checked with the literature that had been reviewed in the initial stages of the project. Trustworthiness was achieved through both peer (Connelly, 2016) and researcher triangulation (Curtin & Fossey, 2007). Members of the research team held regular debriefing meetings during data collection and analysis to compare notes and iron out any discrepancies in data interpretations. We also diversified the data sources to enhance the quality of our analysis. Ethical Considerations We obtained ethical clearance and a research permit (Ref. #277079) from the relevant institutions before the commencement of the study. Written consent was sought and obtained from all study participants for data collection and audio recording. We removed participants’ identifiers, including names and locations, from all reports to ensure anonymity. Results Types of Network One of the objectives of our study was to understand VAC networking functionality, and we sought this information from both FGDs and interview participants. Our data show that government actors were involved in both structured and non-structured networks aimed at preventing and responding to VAC. Structured Networks Structured networks are those with more or less clearly defined boundaries, including goals and membership. They may or may not have structures at the devolved county (subnational) level. Our research found three types of structured network at the national and devolved government levels: policy, advocacy, and service delivery. Policy networks operate at the national level and include invited membership. Both state and non-state actors unite to work on issues that relate to policy, including research, formulation, dissemination, and implementation, among others. They are mainly donor-funded and may or may not have a defined period for operation. For example, the NCCS was established to be a corporate entity by law; it has representation from different government sectors, NGOs, religious institutions, and the private sector. These members hold a three-year term and can serve only twice (Republic of Kenya, 2001). The NCCS is mandated to exist as long as it is legitimized by law. However, the International Journal of Child, Youth and Family Studies (2022) 13(2-3): 123–142 131 council mainly collaborates with different actors to realize a particular output for a defined period. For example, developing the NCCS framework for the National Child Protection system involved various members, including state and non-state actors, children, and the Kenyan citizenry. This collaboration ended once the framework was developed (NCCS, 2011). Similarly, the VAC survey was conceptualized and conducted by a multidisciplinary and multisectoral partnership between state and non-state actors, dubbed the VACS technical working group, the operations of which halted after the completion of survey-related activities (Ministry of Labour and Social Protection, 2019). There are many other technical working groups that do not have an ordained time scope but are not legally bound to remain in existence. We conceptualized advocacy networks as those whose dominant role or purpose of coming together is to influence a legal or policy action, including formulation, adoption, or implementation. Advocacy networks sensitize communities about children’s rights, but we did not find any government actors in these networks. This is possibly because we operationalize advocacy, in this context, as seeking government action. From interacting with their members, we noted that the networking function in advocacy networks is not expressly stated in their job descriptions. This could reflect a gap in advocacy networks’ understanding of networking, or perhaps it is simply that organizational activities take priority over network activities. Finally, service delivery networks provide VAC services like legal, health, psychosocial support, and shelter, among others. Some are mandated to provide services by law. For example, the NCCS official who participated in our study reported that only one AAC was actively coordinating different actors to address VAC. The participants reported that many AACs at the county and subcounty level were non-functional due to a lack of coordination, supervision, and resources from the NCCS. Instead, sometimes they rely on funding from other networks or organizations to execute their roles. The desk review showed that other networks are self-organized to ensure a systems approach to service delivery for VAC. They include both state and non-state actors with varied expertise in addressing VAC. Most organizations that form these networks rely on donor funding targeted to project activity and the short term. This makes service delivery discontinuous, particularly in VAC response, where multiactor coordination is critical. The situation is further compounded by capacity gaps among court officials, police officers, and others (United Nations Children’s Fund, 2018). Other services like mental health and psychosocial support providers, and health care, are compromised by limited logistical support and supervision, and lack of a reliable supply of drugs and materials, among other challenges (Munene & Okwany, 2016). We also learned from the FGDs that structured network members often have membership in several different networks performing similar functions. This is because being part of a functional network is attractive to donors, and organizations have to compete for resources within the network. This funding dynamic results in dormancy periods, especially for networks with few sources of income. They are revived when actors find financing. International Journal of Child, Youth and Family Studies (2022) 13(2-3): 123–142 132 Unstructured Networks Unstructured networks comprise individuals who have dedicated their time, material resources, and energy to respond to VAC in their communities. Some of them are linked to a particular state or non-state institution, often on a volunteer basis. They may or may not have previously worked in child protection, or had relevant training. They include gender activists, paralegals, opinion leaders, and general volunteers. The term “unstructured” denotes that these individuals collaborate loosely with formal networks and organizations to which they refer VAC cases, and are often not registered members. There are also networks of individuals within defined government structures who volunteer on issues relating to VAC. Community health volunteers are the first level of health service delivery recognized in the Ministry of Health (2020) Community Health Strategy. They are chosen by the community to fulfil roles that include collecting primary health data and identifying referrals for health matters; they are attached to households and to public health facilities to which they make referrals (Aseyo et al., 2018). Our FGD data show that they respond to VAC in teams and that the other networks and organizations designated to receive their referrals are often underfunded. Another category of volunteers is children’s officers, who are recruited by the district or county AACs. They support child protection through identifying, documenting, and reporting cases of VAC. However, like community health volunteers, they do not receive funds for facilitating communications with their communities, for necessary travel, or for accompanying children who require care or protection. FGD interactions generally revealed that unstructured networks are driven by care for their communities, and as such, are more resilient because they develop organically. They are the frontline responders to VAC, supporting both individuals and groups. State officials supervise those within the public system, while those outside are accountable only to themselves. These unstructured networks identify and follow up on VAC cases; structured networks rely on them for this. The submissions we received from unstructured networks showed that they have accumulated knowledge on child protection through training conducted by various organizations. However, given the resource constraints they face, addressing VAC can become secondary, being overridden by activities that earn them income. Some network activities may even be halted altogether, resulting in failure to follow up on cases. Additionally, due to their unstructured nature, these networks — especially those defined in a legal framework — engage both vertically with civil society organizations and the state and horizontally with each other, leading to their resources being stretched too thin. Actors in unstructured networks also expressed being powerless relative to those in the structured networks. They felt that they were being used by the structured networks as evidence of the “boots on the ground” required by many donors. These grassroots actors further reported being victimized by perpetrators of VAC in the communities they serve. International Journal of Child, Youth and Family Studies (2022) 13(2-3): 123–142 133 The Influence of Government Leadership in Networks to Address VAC In the literature we reviewed, we found that the government of Kenya had invested heavily in child protection, but the prevalence of VAC remained high. We therefore set out to analyze the role of government leadership in VAC network functioning. Our findings revealed four categories of government leadership: political–administrative, adaptive, enabling, and dissemination. Political–Administrative Function As the government is the primary duty bearer, its officials bear the responsibility of carrying out the political and administrative function of protecting children. This function is shaped by a legal framework that guides programming for child protection and holds both duty bearers and perpetrators of violence accountable for children’s rights. This legal framework extends to the formation of institutions that oversee policy and child welfare. These include the NCCS and its devolved structures, the AACs and the DCS. The Councils hold a political and administrative function that oversees Kenya’s planning, financing, and coordination of child rights interventions. The NCCS and its partners developed the NCCS framework for the National Child Protection System to support this function. The WHO’s (2020) Global Status Report on Preventing Violence Against Children shows that Kenya has 12 government sectors with functions addressing VAC, of which only eight have a coordinating function. The report further indicates that most sectoral plans are only partially funded. This limited state funding impedes them from effectively delivering on their mandates. The resulting heavy dependence on donors and NGOs leads to failures in policy implementation (Nyamu, n.d). Similarly, devolved structures are expected to enforce national policies and laws at the county level. However, Hyun et al. (2020) showed in their report on gender analysis that, in practice, devolution is applied selectively, creating a gap in vertical networking that compromises the effectiveness of policies addressing VAC. The political–administrative function of government, which should be providing leadership, is thus weakened. Adaptive Function The adaptive function of leadership refers to the dynamics of change within networks — the emergence of new ideas and practices. We illustrate this with an example from the VAC survey process, a collective undertaking by various state and non-state actors led by the Centers for Disease Control and Prevention (CDC) in the United States. The findings of the first Kenya VAC survey in 2010 led to the formation of other teams that contributed to various follow-on processes, including strengthening of language on child protection in the constitution, and implementing the Framework for the National Children Protection System and the Strengthening Child Protection in Kenya Programme Strategy 2011–2014 (CDC, n.d.). Similarly, the VAC survey in 2019 led to the adaptation of networks to produce different outputs necessary for child protection, including the National Prevention and Response Plan for Violence Against Children 2019–2023 (Ministry of Labour and Social Protection, State Department for Social Protection, & Department of Children’s Services, 2019), a child-friendly handbook on child protection (Ministry of Labour and Social Protection, State Department for Social Protection, & Department of Children’s Services, International Journal of Child, Youth and Family Studies (2022) 13(2-3): 123–142 134 2020). A further result was the formation of Childline Kenya, an NGO created by a collaboration of three non-state actors to manage a toll-free helpline. This service was developed so that children could express themselves and receive an appropriate referral (Childline Kenya, n.d). Throughout this survey process, adaptability was evident in the transformations of the networks of VAC survey actors, as shown in our results. Enabling Function Different government actors (positional leaders) enable the development of adaptive networks. The DCS is a crucial actor that influences the adaptive function through a legal mandate. Some DCS functions stipulated in the Children’s Act, 2001 include: supervision of children’s officers and coordinating their work; working with government institutions to embed best practices for children; maintaining records and data management on the degree of access to welfare amenities; safeguarding the welfare of children in foster care; providing care and assistance for children in custody; and intervening on behalf of children who need care and protection. Through its mandate, the DCS is a member and a convener of technical working groups, committees, and alliances. Our engagements with actors in unstructured networks also revealed that government officials at the subnational level can enable effective policy implementation at that level. However, there were reports of allegations that law enforcement personnel had accepted bribes from perpetrators of VAC whose cases had thereafter not been prosecuted. One volunteer mentioned that she had received death threats after reporting a case of VAC by a health worker to the police. Corrupt behaviour of this kind weakens networks at the subnational level. One of the troubling findings of our study was the apparent absence of children’s networks. The NCCS framework for child protection promotes child participation: children were part of the process of developing the framework (NCCS, 2011) and other policy documents such as the National Prevention and Response Plan on VAC (Ministry of Labour and Social Protection, State Department for Social Protection, & Department of Children’s Services, 2019). There is provision for two children (a boy and a girl) in the devolved structure of the AACs, as well as children’s parliaments and assemblies (Nyamu, n.d). The environment is thus conducive to enabling children’s participation. However, there was no mention of these representatives in our interactions, and we did not find any active children’s networks in our literature review. Dissemination Function The dissemination function of leadership is illustrated by new ideas being integrated into formal policies. An example is the Childline toll-free helpline, developed by non-state actors but afterwards merged with the DCS’s Crisis Desk to form the National Child Helpline Service in 2008. This enabled the 11-digit toll-free number used by Childline to change to a nationally recognized short form (116) allocated by the National Communication Authority. To date, the DCS is the custodian of 116, and Childline manages the every-day running of the helpline (Childline Kenya, n.d). International Journal of Child, Youth and Family Studies (2022) 13(2-3): 123–142 135 Discussion The global VAC status report (WHO, 2020) recommended strengthening multisectoral governance and coordination of actors as a critical strategy for the prevention of VAC. This article illustrates that networking is connectivity beyond identity. Some established networks, such as the NCCS, its devolved structures (AACs), and some donor-funded non-government institutions are prevented by financial or capacity constraints from carrying out networking activities. On the other hand, there are individuals who intentionally unite efforts to challenge VAC in their communities using their own resources. The structure providing connectivity between these individuals may have no official name — or any name at all — but the function of networking happens spontaneously through regular communication, referral, and follow-up of cases. The article further shows that government, as the ultimate duty bearer, must participate as a crucial actor in networks that address VAC if they are to realize change. We conceptualize government leadership as operating through a legal framework that legitimizes networking and obligates government actors as positional leaders to actively engage other stakeholders. The NCCS and DCS, as well as their sub-national structures, hold leadership positions that require them to oversee issues on protection and prevention of VAC. Although there is no evidence of formal connections between state and non-state VAC actors, our results do not confirm those of Wessells (2015), who found a deep disconnect between state and non-state actors. We show how policy-level networks can adapt using the example of the VAC surveys of 2010 and 2019, which were carried out through a partnership between state and non-state actors. According to Meijerink and Stiller (2013), the adaptive function of leadership is not a task, but an emergent property of a network. To illustrate, the VAC surveys mentioned above were conducted by one network (the VACS technical working group), and when the results were released, other groups organized themselves into new networks to translate the findings into actionable interventions. In 2019, for example, the VAC survey results motivated the establishment of one network in the form of the team that developed the National Prevention and Response Plan for Violence Against Children 2019–2023, and another network that developed a child-friendly handbook on child protection. The emerging networks included participants from the technical working group, some of whom were government actors. Our findings show that although networking for VAC is not limited to non-state actors, government involvement in networking is more visible at the national level than at the subnational level, impacting formulation or revision of policy. Donor funding seems to be a significant defining factor for the functionality of these policy-level networks, and it is doubtful whether the policy environment to address VAC would be as rich if external funding were reduced or eliminated — a phenomenon we call donor-induced networking. The strength that national networks have in policy formulation is undermined by barely functional sub-national networks (like the AACs under the NCCS) because the latter have limited funding, if any, for policy implementation or supervision. Other scholars (Chege & Ucembe, 2020; Cooper, 2012; Nyamu, International Journal of Child, Youth and Family Studies (2022) 13(2-3): 123–142 136 n.d.) have highlighted the inadequacy of government funding for implementation of laws and policies on child protection in Kenya. Despite this lack of government funding, there have been improvements in addressing VAC since 2010, which Annor et al. (2022) attributed partly to the successful implementation of policy strategies. This suggests that the implementation efforts have mainly been carried out by non-state actors funded from the private sector. Future research could compare the direct contributions to child protection of networks that include state actors with those solely comprising non-state actors. Additionally, while we see government actors as enablers in policy-level networks, they become disablers at the grassroots level when they take bribes from perpetrators of VAC or threaten network members who are advancing child protection. This reflects several problems. First, there is lack of supervision from government actors, as Nyamu (n.d.) highlighted in her critical analysis of the NCCS child protection framework. A report by Hyun et al. (2020) also shows the selective devolution of government structures, and a resulting gap in responsibilities between the national level and government actors at devolved levels. Second, there seems to be a poor understanding among government actors of networking for VAC, and the importance of the unique role played by networks in referral of cases, investigating and adjudicating cases, and apprehending perpetrators. Third, we think it is possible that government actors disregard VAC as a societal nuisance, and accord it low priority. Relatedly, we found that children participated in different policy formulation processes as information providers, but we did not find active children’s networks, especially at the implementation level. Although determining the existence of such networks was not part of our research design, it is notable that there was no mention of them in all of our engagements with participants; this may indicate that active children’s networks indeed do not exist, or perhaps that different stakeholders in VAC, including government actors, regard their leadership as insignificant. Individuals like Greta Thunberg, a young Swedish woman who has been at the forefront of global child activism for climate change since age 15 (Zhanda et al., 2021), and the students who led a pivotal series of demonstrations in 1976 against the use of Afrikaans as an official language in the South African education system (Banda, 2000), are examples of the power of young people’s leadership in networking and their influence on government policy and practice. On the whole, our results agree with Embleton et al. (2021), who recommended a holistic approach to child protection involving the creation and sustenance of collaborations between state and non- state child protection efforts (Badoe, 2017; Devaney & Byrne, 2015; O’Leary et al., 2015). Conclusion Since the government is the primary duty bearer in the prevention of VAC, the many recommendations in the WHO’s (2020) global VAC status report begin with a key government task: strengthening the governance and coordination of multisectoral actors. To carry out this role, the report recommends that governments ensure that there is “an appropriately resourced agency International Journal of Child, Youth and Family Studies (2022) 13(2-3): 123–142 137 … to coordinate multisectoral action to end violence against children” (p. 63). As well, governments must strengthen the roles of various ministries, departments, and agencies, giving them clearly mandated responsibility for delivery of programmes in areas that may not be covered by existing roles, such as “norms and values, parent and caregiver support, safe environments, and income and economic strengthening” (p. 63). Although the Kenyan government has demonstrated political commitment to addressing VAC and achieved partial success, the prevalence of VAC remains unacceptable, possibly due to gaps in implementation and supervision, and limited capacity in networking, among government actors as the main duty bearers. Previous studies have investigated VAC without focusing on connectivity among various actors, and our research aimed to provide this missing perspective. Our study findings show that both government and civil society collaborations are essential for VAC prevention and response. We recommend that the Kenyan government be more intentional in supporting VAC network actors at all levels for accelerated response to, and eventual elimination of, VAC. Better networking among state and non-state actors is needed to eliminate duplication of roles. Additionally, better connections among actors will increase network functionality and reduce donor-induced networking and the associated disadvantages. We recommend that further research be undertaken to explore the enabling environment for children’s networks in both preventing and responding to VAC. Acknowledgements We are grateful to all our participants at the grassroots, subnational, and national levels. We acknowledge Nascent RDO Uganda for hosting the research, and specially recognize other members of the research team (Annah Kamusiime, Auleria Munene, Lydia B. Sandi). We are further grateful to Dr. Lina Digolo for reviewing the first draft of the manuscript, and the reference group (Agnes M. Wasike, Dr. Lina Digolo, Joel K. Kiiya, Jacqueline Asiimwe, Katie Davies, and Krista Riddley) for providing overall quality assurance. Finally, we thank Catherine Mugabo for extensive administrative support, and the Wellspring Philanthropic Fund through CivSource Africa for funding the research. International Journal of Child, Youth and Family Studies (2022) 13(2-3): 123–142 138 References Annor, F. B., Chiang, L. F., Oluoch, P., Mang’oli, V., Mogaka, M., Mwangi, M., Ngunjiri, A., Obare, F., Achia, T., Patel, P., Massetti, G. M., Dahlberg, L. L., Simon, T. R., & Mercy, J. A. (2022). Decline in violence among children and youth in Kenya: A comparison of the 2010 and 2019 Kenya Violence Against Children and Youth Surveys. The Lancet: Global Health, 10(1), e124–e133. doi:10.1016/S2214-109X(21)00457-5 Aseyo, R. E., Mumma, J., Scott, K., Nelima, D., Davis, E., Baker, K. K., Cumming, O., & Dreibelbis, R. (2018). Realities and experiences of community health volunteers as agents for behaviour change: Evidence from an informal urban settlement in Kisumu, Kenya. Human Resources for Health, 16(1), Art. 53. doi:10.1186/s12960-018-0318-4 Awortwi, N. (2018). Social protection is a grassroots reality: Making the case for policy reflections on community‐based social protection actors and services in Africa. Development Policy Review, 36(52), O897-O913. doi:10.1111/dpr.12364 Ayaya, S., DeLong, A., Embleton, L., Ayuku, D., Sang, E., Hogan, J., Kamanda, A., Atwoli, L., Makori, D., Ott, M. A., Ombok, C., & Braitstein, P. (2021). Prevalence, incidence and chronicity of child abuse among orphaned, separated, and street-connected children and adolescents in western Kenya: What is the impact of care environment? Child Abuse & Neglect, 104920. doi:10.1016/j.chiabu.2020.104920 Badoe, E. (2017). A critical review of child abuse and its management in Africa. African Journal of Emergency Medicine, 7(Supp.), S32–S35. doi:10.1016/j.afjem.2017.09.002 Banda, F. (2000). The dilemma of the mother tongue: Prospects for bilingual education in South Africa. Language, Culture and Curriculum, 13(1), 51–66. doi:10.1080/07908310008666589 Centers for Disease Control and Prevention. (n.d.). Working towards violence-free generation in Kenya. https://www.cdc.gov/violenceprevention/childabuseandneglect/vacs/kenya- successstory.html Chege, N., & Ucembe, S. (2020). Kenya’s over-reliance on institutionalization as a child care and child protection model: A root-cause approach. Social Sciences, 9(4), 57. doi:10.3390/socsci9040057 Childline Kenya. (n.d.). Keeping children safe. https://childlinekenya.co.ke CivSource Africa. (2021). Networking to address violence against children in East Africa: An exploratory study [Research report]. https://static1.squarespace.com/static/593ea10db8a79bc4102e122f/t/6152bed7e84d71050c1e8 4e8/1632812817955/vac+research+report.pdf Connelly, L. M. (2016). Trustworthiness in qualitative research. Medsurg Nursing, 25(6), 435. International Journal of Child, Youth and Family Studies (2022) 13(2-3): 123–142 139 Cooper, E. (2012). Following the law, but losing the spirit of child protection in Kenya. Development in Practice, 22(4), 486–497. doi:10.1080/09614524.2012.673554 Coveney, P. V. (2003). Self-organization and complexity: A new age for theory, computation and experiment. Philosophical Transactions of the Royal Society A: Mathematical, Physical and Engineering Sciences, 361(1807), 1057–1079. doi:10.1098/rsta.2003.1191 Curtin, M., & Fossey, E. (2007). Appraising the trustworthiness of qualitative studies: Guidelines for occupational therapists. Australian Occupational Therapy Journal, 54(2), 88–94. doi:10.1111/j.1440-1630.2007.00661.x Devaney, C., & Byrne, P. (2015). The value of family welfare conferencing within the child protection and welfare system. Child Care in Practice, 21(4), 340–356, doi:10.1080/13575279.2015.1027173 Embleton, L., Shah, P., Gayapersad, A., Kiptui, R., Ayuku, D., Wachira, J., Apondi, E., & Braitstein, P. (2021). Recommendations for improving access to healthcare for street- connected children and youth in Kenya: A qualitative study. Children and Youth Services Review, 131, 106302. doi:10.1016/j.childyouth.2021.106302 Glaser, B. G., & Strauss, A. L. (1967). The discovery of grounded theory: Strategies for qualitative results. Aldine. Hyun, M., Okalo, W., & Munene, A. (2020). USAID Kenya gender analysis report. United States Agency for International Development. https://banyanglobal.com/wp- content/uploads/2020/05/USAID-Kenya-Final-Gender-Analysis-Report.pdf Meijerink, S., & Stiller, S. (2013). What kind of leadership do we need for climate adaptation? A framework for analyzing leadership objectives, functions, and tasks in climate change adaptation. Environment and Planning C: Government and Policy, 31(2), 240–256. doi:10.1068/c11129 Mercy, J. A., Hillis, S. D., Butchart, A., Bellis, M. A., Ward, C. L., Fang, X., & Rosenberg, M. L. (2017). Interpersonal violence: Global impact and paths to prevention. In C. N. Mock, R. Nugent, O. Kobusingye, & K. R. Smith (Eds), Injury prevention and environmental health. (3rd ed.; Ch. 5). The International Bank for Reconstruction and Development / The World Bank. https://www.ncbi.nlm.nih.gov/books/NBK525208/ Ministry of Gender, Children, and Social Development. (2011). Kenya national social protection policy. Republic of Kenya. https://socialprotection.or.ke/images/downloads/kenya-national- social-protection-policy.pdf International Journal of Child, Youth and Family Studies (2022) 13(2-3): 123–142 140 Ministry of Health. (2020). Kenya community health strategy 2020–2025. https://www.health.go.ke/wp-content/uploads/2021/01/Kenya-Community-Health-Strategy- Final-Signed-off_2020-25.pdf Ministry of Labour and Social Protection, Department of Children’s Services. (2019). Kenya Violence Against Children survey report 2019. https://bettercarenetwork.org/sites/default/files/2020-10/2019 Violence Against Children Survey %28VACS%29 .pdf Ministry of Labour and Social Protection, State Department for Social Protection, & Department of Children’s Services. (2019). National prevention and response plan on violence against children in Kenya 2019–2023. Republic of Kenya. https://www.socialprotection.go.ke/wp- content/uploads/2020/07/National-Prevention-Response-Plan-on-VAC-WEB-v5-July-3- 2020.pdf Ministry of Labour and Social Protection, State Department for Social Protection, & Department of Children's Services. (2020). Protecting children against violence: Child friendly booklet. Republic of Kenya. https://www.socialprotection.go.ke/wp-content/uploads/2020/07/Child- Friendly-VAC-Bookle29t.pdf Ministry of Labour, Social Security and Services. (2016). National policy on elimination of child labour: Towards a child labour free society. https://repository.kippra.or.ke/handle/123456789/739 Mkutu, K., Mulinge, M., Yogo, C., Kajumbi, O., Owino, E., Shiundu, M., Muriuki, P., & Atieno, E. (2019). Kisumu County: Crime and violence rapid assessment. The World Bank. https://policycommons.net/artifacts/1272041/kisumu-county/1855848/ Munene, A., & Okwany, A. (2016). Interrogating the ‘artificial’ divide between health and education for children aged 0–3 years in urban poor locales in Kenya. South African Journal of Childhood Education, 6(2), a465. doi:10.4102/sajce.v6i2.465 Mwangi, M. W., Kellogg, T. A., Brookmeyer, K., Buluma, R., Chiang, L., Otieno-Nyunya, B., & Chesang, K. (2015). Perpetrators and context of child sexual abuse in Kenya. Child Abuse & Neglect, 44, 46–55. doi:10.1016/j.chiabu.2015.03.011 National Council for Children’s Services. (2010). The national children policy: Kenya 2010. Republic of Kenya. https://repository.kippra.or.ke/handle/123456789/2047 National Council for Children’s Services. (2011). The framework for the national child protection system for Kenya. Republic of Kenya. https://www.socialserviceworkforce.org/resources/framework-national-child-protection- system-kenya International Journal of Child, Youth and Family Studies (2022) 13(2-3): 123–142 141 National Council for Children’s Services. (2015). National plan of action for children in Kenya (2015–2022). Republic of Kenya. https://bettercarenetwork.org/sites/default/files/2020- 10/National-Plan-of-Action-for-Children-in-Kenya-2015-2022.pdf Nyamu, I. K. (n.d.). More protection or more harm? A critical review of the child protection framework in Kenya. https://www.academia.edu/8113428/More_protection_or_more_Harm_A_Critical_Review_of _the_Child_Protection_Framework_in_Kenya O’Leary, P., Hutchinson, A., & Squire, J. (2015). Community-based child protection with Palestinian refugees in South Lebanon: Engendering hope and safety. International Social Work, 58(5), 717–731. doi:10.1177/0020872815584427 Onyango, P., & Tostensen, A. (2015). The situation of youth and children in Kibera [CMI report 2015: 3]. Chr. Michelsen Institute. Organization of African Unity. (1990). African charter on the rights and welfare of the child. https://au.int/sites/default/files/treaties/36804-treaty-0014_- _african_charter_on_the_rights_and_welfare_of_the_child_e.pdf Republic of Kenya. (2001). Children Act: Chapter 141 (Revised Edition 2012 [2010]. National Council for Law Reporting. https://www.ilo.org/dyn/natlex/docs/ELECTRONIC/61290/115495/F1975053 Republic of Kenya. (2010). The Constitution of Kenya. National Council for Law Reporting. http://kenyalaw.org/kl/index.php?id=398 Russell, D., Higgins, D., & Posso, A. (2020). Preventing child sexual abuse: A systematic review of interventions and their efficacy in developing countries. Child Abuse & Neglect, 102, 104395. doi:10.1016/j.chiabu.2020.104395 Scholten, P., Keskitalo, E. C. H., & Meijerink, S. V. (2015). Bottom-up initiatives toward climate change adaptation in cases in the Netherlands and the UK: A complexity leadership perspective. Environment and Planning C: Politics and Space, 33(5), 1024–1038. doi:10.1177/0263774X15605929 Tshimpaka, L. M., Nshimbi, C. C., & Moyo, I. (2021). Regional economic communities and integration in southern Africa: Networks of civil society organizations and alternative regionalism. Springer. doi:10.1007/978-981-15-9388-8 Uhl-Bien, M., Marion, R., & McKelvey, B. (2007). Complexity leadership theory: Shifting leadership from the industrial age to the knowledge era. The Leadership Quarterly, 18(4), 298–318. doi:10.1016/j.leaqua.2007.04.002 International Journal of Child, Youth and Family Studies (2022) 13(2-3): 123–142 142 Undie, C.-C., & Mak’anyengo, M. (2022). If we ask, will they tell? (and then, what?): Screening for sexual violence against children in Kenya. Child Abuse Review, 31(1), 11–26. doi:10.1002/car.2680 United Nations Children’s Fund. (2018). Situation analysis of children and women in Kenya, 2017 [Report]. https://www.unicef.org/kenya/reports/situation-analysis-children-and-women- kenya-2017 United Nations Department of Economic and Social Affairs. (2015). Transforming our world: The 2030 agenda for sustainable development. https://sdgs.un.org/2030agenda Wangamati, C. K., Gele, A. A., & Sundby, J. (2020). Post rape care provision to minors in Kenya: An assessment of health providers’ knowledge, attitudes, and practices. Journal of Interpersonal Violence, 35(5–6), 1415–1441. doi:10.1177/0886260517696863 Wangamati, C. K., Sundby, J., & Prince, R. J. (2018). Communities’ perceptions of factors contributing to child sexual abuse vulnerability in Kenya: A qualitative study. Culture, Health & Sexuality, 20(12), 1394–1408. doi:10.1080/13691058.2018.1438666 Wangamati, C. K., Yegon, G., Sundby, J., & Prince, R. J. (2019). Sexualised violence against children: A review of laws and policies in Kenya. Sexual and Reproductive Health Matters, 27(1), 16–28. doi:10.1080/26410397.2019.1586815 Wessells, M. G. (2015). Bottom-up approaches to strengthening child protection systems: Placing children, families, and communities at the center. Child Abuse & Neglect, 43, 8–21. doi:10.1016/j.chiabu.2015.04.006 World Health Organization. (2020). Global status report on preventing violence against children 2020. https://www.who.int/teams/social-determinants-of-health/violence-prevention/global- status-report-on-violence-against-children-2020 Younis, M. (2017). Evaluating coalitions and networks: Frameworks, needs, and opportunities. Center for Evaluation Innovation. https://efc.issuelab.org/resources/29445/29445.pdf Zhanda, K., Dzvimbo, M. A., & Chitongo, L. (2021). Children climate change activism and protests in Africa: Reflections and lessons from Greta Thunberg. Bulletin of Science, Technology & Society, 41(4), 87–98. doi:10.1177/02704676211049690