Transformation of child welfare Institutions in Bandung, West Java: A case of deinstitutionalization in Indonesia

Bandung level (26%) and (iii) medium to large, secular, non-government Child Welfare Institutions that deliver privately funded institutional care and other family support services at the provincial level (24%). While there is evidence of deinstitutionalization activity in Bandung, the large number and relative size of privately operated Child Welfare Institutions suggests a high dependency on institutional care to meet children ’ s general welfare needs. The cluster analysis also indicates limited government funding and the missions of religious organizations as possible constraints on system transformation. Further research is needed to understand the enablers and barriers to transforming the system from predominantly institutional care to a continuum of services in the context of deinstitutionalization in Bandung.


Child welfare Institutions within Indonesia's child welfare system
Families in lower-middle-income countries like Indonesia live with social and economic vulnerabilities that hinder their ability to meet their children's needs.In a context of limited social assistance, these vulnerabilities drive demand for institutional care.Research suggests poverty, lack of access to basic services, discrimination, and exclusion are the primary causes of institutionalization, particularly for children who are not orphans (Boothby et al., 2012;Goldman et al., 2020;Martin & Zulaika, 2016).Circumstances such as parental remarriage, paternal abandonment, and lack of civil and legal documentation are other common reasons for children entering institutional care (Save the Children Indonesia Country Office, 2013).Disability and being born out of wedlock also heighten children's risk of institutionalization (PUS-KAPA & UNICEF, 2014; Sutinah & Aminah, 2018).
Institutions have become instrumental in fulfilling children's basic needs in Indonesia and providing access to formal or informal education (Ministry of Social Affairs & UNICEF, 2015; PUSKAPA & UNICEF, 2014).
Institutions have also been sustained through cultural customs encouraging donations to orphans and impoverished children, and faith-based organizations continue to run institutions to help disadvantaged children and families as an expression of their religious mission (Davidson et al., 2017;McLaren & Qonita, 2019).In 2020, the Ministry of Social Affairs (MoSA) recorded 102,482 Indonesian children aged 0-17 living in 3,575 institutional care facilities (PUSKAPA et al., 2020).Previous research suggests that most of these children lived in non-government child welfare institutions that deliver services at the district/city level (Martin & Sudrajat, 2007).However, MoSA still operates centers providing institutional care services.As outlined in the latest ministerial regulation on the structures of technical units for social rehabilitation under MoSA (No.3/2022), the national government is responsible for operating thirty-one centers across the thirty-four provinces, which provide both community and facility-based services for vulnerable groups that can include children, the elderly, people with disabilities, and victims of disasters and other emergencies.Provincial government manage their own institutional-based services in which each institution is provided for a specific group, for example for children under five years of age.
Children living in institutional care are considered among the most vulnerable groups of children due to the risks of growing up without stable relationships with caring adults, of being exploited and isolated, and of living in poor quality and distressing care environments (Berens & Nelson, 2015;Fluke et al., 2012;Ismayilova et al., 2014).Previous studies have reported that some children raised in Indonesian child welfare institutions do not receive a formal education or access to health services (Martin & Sudrajat, 2007;PUSKAPA & UNICEF, 2014;Sutinah & Aminah, 2018).A recent study of 500 children in child welfare institutions in five districts in East Java found that many children experienced physical abuse in the forms of beatings or being forced to work (29.5 %) and psychological abuse, such as being scolded, threatened, locked away(13.8%), by other children and caregivers in the institutions (Sutinah & Aminah, 2018).A separate study of 625 children in institutions in three Indonesian provinces reported that 17 % of children felt lonely at the time of the survey, and 42 % of children experienced physical violence at least once during their time in an institution (PUSKAPA & UNICEF, 2014).

The process of deinstitutionalization in Indonesia
Following the large number of children orphaned by the Tsunami in 2004, Save the Children supported MoSA to investigate the country's child welfare institutions for orphaned children.The research report indicated many children were relinquished to live in institutions unnecessarily (Martin & Sudrajat, 2007).Following the research, MoSA declared a shift towards deinstitutionalization and the establishment of a continuum of child welfare services that places institutional care as the most restrictive child welfare intervention (Kusumaningrum & Irwanto, 2011;Save the Children Indonesia Country Office, 2013).
Deinstitutionalization is the progressive replacement of an institutional-based care approach with a continuum of child welfare services to prevent family separation, reunite children and families, or provide quality family-based alternative care.Fig. 1 shows a continuum of child welfare services.The continuum of child welfare services comprises family support services, family-based alternative care in the form of kinship care, foster care, and adoption, and lastly, small, temporary, transitional institutional care (Faith to Action Initiative, 2015; Fluke et al., 2012;Goldman et al., 2020).
Within the service continuum (see Fig. 1), family support services may include direct services to birth families such as counseling, family development sessions, social, health, and psychological services within a case management approach, or economic support for families through cash payments under social welfare arrangements (Daly, 2018;Daly et al., 2015).Gatekeeping is an ongoing assessment process determining if alternative care is necessary, the type of alternative care children receive when required and whether reunification is possible.A coordinated gatekeeping process is necessary to prevent family separation, help reunite children with their families, promote family-based alternative care and ensure institution-based care is provided only when necessary (Faith to Action Initiative, 2015).
Indonesia's child welfare system was not formalized until the Government of Indonesia enacted the national Child Protection Law in 2002.Overall, the government and non-government sectors hold essential roles within the system.As previously mentioned, prior to the enactment of deinstitutionalization policies, the system had relied on institutionbased care (Martin & Sudrajat, 2007) and the national government managed institutions for vulnerable groups, including children not attending school and children with disabilities.MoSA also operated shelters for child victims of violence and exploitation.Non-government organizations provided mostly institutional care for children, operating few family or community-based services.Studies on the implementation of the national child welfare system thus far indicate that the continuum of services within the system has focused heavily on tertiary services and programs have been implemented in an uncoordinated and fragmentary manner (Kementerian Sosial RI & UNICEF, 2010).
The National Standard of Care for Child Welfare Institutions (the Standard) was enacted in 2011 to support the paradigm shift towards deinstitutionalization in Indonesia and establish a continuum of child welfare services (Save the Children Indonesia Country Office, 2013).The Standard aimed to transform the operations of all child welfare service providers, including government institutions, non-government child welfare institutions, and child welfare organizations.Following the enactment of the National Standard, all types of child welfare providers were registered as a Child Welfare Institutions (LKSA), 1 whether they provided institutional care or not, to reflect the new role enshrined by the Standard (National Standard of Care for Child Welfare Institutions, 2011).Transformation of the Child Welfare Institutions (LKSA) from institutional facilities into family services centers to prevent institutionalization is a key parameter of the national deinstitutionalization process.
Under the Standard, Child Welfare Institutions operate as family service centers responsible for conducting referrals or providing family support services to vulnerable children and families.Institutional care delivered by the government or by non-government providers is limited to small-scale residential services for children who required therapeutic or specialized services, including children in need of special protection, children with disabilities, victims of violence, exploitation, or maltreatment, and children in conflict with the law.The Standards also provide guidelines for the minimum Standard of care for children in institutions (National Standard of Care for Child Welfare Institutions, 2011).Indonesian child protection services, which respond to 1 The Standard defines LKSA (Lembaga Kesejahteraan Sosial Anak) as "institutions built by the government, local government, or civil society in carry out care for children" (pg.iv).The name Panti Asuhan was changed into LKSA for "any institution or organization that provide care for children" (pg.5).
allegations of physical and sexual abuse against children, are not registered as LKSA.These government services exist as part of the development of the national child welfare system by providing case management, family support services, and referral services.Before the deinstitutionalization process, many of these services were still provided through institutional care.

An investigation of deinstitutionalization in Bandung, West Java, Indonesia
The province of West Java has one of the largest numbers of registered non-government Child Welfare Institutions in Indonesia.This province has 435 Child Welfare Institutions, with 13,046 children receiving residential services from the institutions (PUSKAPA et al., 2020).West Java is also the most populous province in Indonesia and Bandung City is Indonesia's third most populous city (Tarigan et al., 2016).As the capital of West Java, Bandung attracts people from its surrounding districts, including rural areas, who relocate or settle there to find jobs or access education (Argo, 2015;Paramita et al., 2021).
Understanding of the transformation process is critical to navigate the future direction of Indonesia's child welfare system.Whether Child Welfare Institutions (LKSA) in Indonesia have transformed to facilitate deinstitutionalization is currently unknown.A research study was conducted in two sequential phases to understand the transformation of Child Welfare Institutions in Bandung, West Java, specifically to explore the progress towards deinstitutionalization and identify enablers and barriers to facilitate the national deinstitutionalization process.This paper reports the findings from the Phase 1 study which aimed to address the following research question: What local services exist for neglected children in Bandung, West Java?By understanding the capacity of the local service system, the Phase 1 study will provide evidence on the stage of system transformation.

An online survey of child welfare service providers in Bandung, West Java, Indonesia
The online survey explored the transformation of Child Welfare Institutions and government child protection services into family service centers.Survey questions and hypothetical cases were derived from key issues identified in the authors' review of available literature on institutional care in Indonesia and transformation of institutional care in other developing countries.The first part of the survey contained closeended questions regarding the organizations' sector, affiliation, resources, and services.Participants were then asked to respond to three hypothetical scenarios of children at risk of family separation due to different social and economic vulnerabilities to capture rich information about the organizations capacity to deliver family support and gatekeeping services.The survey was developed in English and translated to Bahasa Indonesia.To ensure suitability of the survey to the current work of child welfare institutions in Indonesia, the authors piloted the survey with selected child welfare practitioners in Indonesia.Ethics approval was obtained from the University of Melbourne in June 2020 (HREC ID# 2056371.1).

Population and recruitment
Child welfare providers in scope for the study included Child Welfare Institutions (LKSA) and government child protection services.Seventythree child welfare service providers in Bandung were identified through the MoSA registry of Child Welfare Institutions or by MoSA social workers.Author 1 contacted all 73 service providers to introduce the study and invite them to participate.This initial contact was followed by two reminders.Recruitment and data collection were conducted between July and September 2020.

Data analysis
Data analysis was conducted using a mixed-method approach.Quantitative analysis of the survey data was conducted using IBM SPSS Statistics 27.Two-step clustering (Rezankova, 2009;Tkaczynski, 2017) was used to identify homogeneous subgroups of child welfare service providers based on sector and affiliation, service area, funding source, and scale of activity (size).Qualitative analysis of the open-ended responses was conducted using a thematic analysis approach (Braun & Clarke, 2006) to describe and summarize the types of services within a continuum of services and to further understand the characteristics of service providers in each cluster.Three initial steps in thematic analysis were implemented: familiarization with data, coding (using predetermined codes), and generalization of themes (Braun & Clarke, 2006).Pre-determined codes were developed prior to the coding process.The three pre-determined codes are: family support services, family-based alternative care, and institution-based care.Next, responses were grouped for each of the scenarios and the responses coded based on the pre-determined codes.Lastly, the themes were matched to the groups of LKSA identified through the cluster analysis.

Sample characteristics
Qualtrics recorded 82 survey responses, and after data cleaning, responses from 50 local service providers were included in data analysis.Of the 50 respondents who completed the online survey on behalf of their organization, 33 (67.3 %) had worked in their organization for more than five years.More than half (29, 58 %) held leadership positions such as the organization's director, head, or leader.Most of the respondents were full-time employees (61.2 %).

Characteristics of child welfare service providers
Table 1 presents the sector, operational budget, and source of funding of the child welfare service providers in the sample.Table 2 shows variables that further describe the human resources and capacity of the providers.Table 3 shows variables that describe the scope of service of the providers.Missing data are excluded from the calculations.

Table 1
Sector and funding of child welfare service providers (N = 50).N. Luh Putu Maitra Agastya et al.

Clusters of child welfare service providers
To further understand the resources and operations of the child welfare service providers and to help make sense of findings from the open-ended survey items, providers were grouped into clusters, identified via a two-step clustering process.Three clusters were identified based on 1) sector (non-government or government); 2) affiliation (secular or faith-based); 3) annual funding (less than 9,800 USD, between 9,800 -39,000 USD, more than 39,000 USD); 4) primary funding source (government funding, philanthropic funding, individual donations); and 5) service footprint (district (Bandung), provincial (West Java), national coverage).The three clusters and their associated operational profiles are presented in Table 4.

Scope of services among child welfare service providers
Qualitative responses in the survey were analyzed to understand the scope of services provided by service providers generally and within each cluster.The survey asked participants to respond to three hypothetical cases by identifying whether the child welfare provider could provide a service response.Author 1 coded the open-ended responses to the question: "What type of services can your organization provide in responding to the case?" for each hypothetical case using the predetermined codes with the aim of identifying the types of services available for each hypothetical scenario.The researcher then compared the response patterns identified by this thematic analysis between the clusters of service providers.Cluster 1 (half of all providers) included privately funded, medium to large, faith-based non-government LKSA based in Bandung and funded privately.Cluster 2 comprised government services and small secular Bandung-based LKSA, funded by government and private funds.Cluster 3 included secular based nongovernment LKSA that operated across West Java and were privately funded.

Scenario 1: Child at risk of school dropout
In Scenario 1, respondents were presented with the case of Ana, from a low-income family, who was at risk of dropping out of primary school.Most organizations in each cluster indicated they could provide a service response neglect (96 %, 76.9 %, and 100 %, respectively), as school dropouts are often viewed as a case of.Most Cluster 1 providers were inclined to offer institutional care without family support, although children were still able to communicate with their families: Ana can live in the dormitory for us to provide services, care, and education until she graduates to high school/college level, while her parents still focus on working for their family at home.The family is not given any assistance for children who enter the dormitory.(Cluster 1) Some Cluster 1 providers indicated they would implement gatekeeping activities before providing institutional care; that is, thinking about the most appropriate service depending on the needs of the child.These organizations considered the availability of suitable extended family members and the circumstances of the family that could influence the child's access to education, such as the number of siblings, parent's employment, and distance to school: We'll see how far the school is from her house and calculate how much money is needed for transportation.If the transportation costs are too high and exceed the daily food needs, we look for extended family members close to the school to accommodate the child.But if no relatives can accommodate and care for them, we offer the last option to be taken to an orphanage that solves the education issue.(Cluster 1) Contrary to the inclination to provide institution-based care of providers in Cluster 1, Cluster 2 providers could mostly offer prevention services that would keep Ana with her family through referrals for education, counseling, financial aid, and economic empowerment initiatives for the parents, as well as providing financial assistance for schooling derived from donations: Conduct an initial assessment of Ana's family problems.Provide individual and family counseling.Explore the parents' skills for work   Most Cluster 3 providers were inclined to offer a combination of family support, education services and institutional care depending on the specific circumstances of Ana's family.Some Cluster 3 providers indicated that they would collaborate with government agencies to provide family support, while others would initiate their own services: If economic conditions are the only factor in dropping out, the children are categorized as clients living with families.As a social welfare organization, we foster economic independence for families.
[We provide] family-based care so that children are still in school, the families receive information about parenting, we conduct child and family strengthening meetings in which families interact with the Child Welfare Institution.(Cluster 3)

Scenario 2: Unplanned pregnancy.
Scenario 2 described a young woman (Ika) whose pregnancy was unintended and who was not legally married to the father of her child.Approximately one-quarter (24 %) of the Cluster 1 providers indicated they could offer a service response, compared to 69.2 % of the Cluster 2 providers and 41.7 % of the Cluster 3 providers.However, Cluster 1 providers could not provide family preservation services, although several respondents indicated that they could offer a referral response: "There are cases of pregnancies out of wedlock, their children being handed, from poor families.We directed him to be referred to [baby orphanage] to facilitate the delivery.We don't handle it ourselves, but we provide referrals".(Cluster 1).
Another provider also stated: A person has had a case like the one above.She is pregnant with her child because of an illegal relationship [and is] asking us to take care of her child.We refuse because our institution does not have the facility and human resources (to respond to the case).Still, as our form of empathy, we can give referrals but only to certain institutions in the city of Bandung.We are not ready yet regarding the adoption because they must undergo a lengthy procedure, including the court.
(Cluster 1) One Cluster 1 provider stated that support for children of unmarried parents was not aligned with the organization's mission:" [We] do not accept cases of babies who are abandoned by their parents and adoption.Orphanages do not legalize relationships outside of marriage and are not shelters for those who throw their children away".
One Cluster 1 provider offered long-term institutional care.In this situation, the institution appeared to prioritize the rights of the child to live: "[Providing] care from baby to school age and sending to school until graduation and work.In addition, it acts as a substitute parent for the child for the rest of their life".
Similar to responses to Scenario 1, Cluster 2 providers were able to provide prevention services, combined with referrals to other services, including institutional care if this was the recommendation following assessment: Depending on the recommendations given after the assessment is carried out.Possible services are the facilitation of referral to an orphanage and assistance in the adoption process for the baby, safe houses for clients, or referral and advocacy with the client's family if the recommendation is family reunification.(Cluster 2) Some Cluster 2 providers were able to offer temporary institutional care for the mother and baby as well as other services and support to prevent family separation: We will not advise Ika to leave the baby in the institution's care.We will try so that Ika can continue her education while caring for the baby at the institution.[We] make an agreement and [give] understanding [to] Ika as the baby's parent, and we will not recommend the baby to be adopted.Ika will get information related to institutions and be able to complete her education.Meanwhile, we will provide care for the baby.At the same time, Ika lives outside the LKSA (Child Welfare Institution) and, as soon as possible, hands over care to Ika as her biological mother if Ika is ready.(Cluster 2) Most Cluster 2 providers and Cluster 3 providers (but only one Cluster 1 institution) offered counseling services that included the partner and the extended family to prevent the separation of the baby from their mother.The counseling services have different goals: to prepare the mother and family to become parents or make sound decisions regarding the care of the baby, including options for adoption.One Cluster 3 provider suggested kinship care: "Accompanying the baby's mother and contacting her family to discuss the baby's fate so it won't be adopted.Inviting related parties who can provide a way out".
A Cluster 2 provider also hoped to involve the father: Counseling for decision making.Facilitate Ika and her partner to build readiness to care for the baby and identify parties who can assist her in raising the baby.If the choice is to care for the baby, Ika, and her partner are facilitated [to access] needed and available social support and how to address the gap between the two.(Cluster 2)

Scenario 3: Street-Associated child from a Single-Parent family
Scenario 3 presented a case of a street-associated child (Didi).Just over half (53.8 %) of Cluster 1 providers indicated they could respond, compared to 26.9 % of Cluster 2 providers and 19.2 % of Cluster 3 providers.Overall, the service that was offered by service providers across the clusters was mainly institutional care, incorporating vocational education or schooling, as stated by a provider in Cluster 1: "Didi will be given religious education because it is a religion-based institution and provided a place to live because the institution has an orphanage for orphans and underprivileged children to live, giving school education as well".
Access to education, particularly vocational training, was therefore viewed by both Cluster 1 and Cluster 2 providers as an intervention that can prevent the child from returning to the streets: The prioritized formal services are mental and spiritual assistance to restore age-appropriate behavior.If they are unwilling to return to school, they can go to the vocational training center, while another option is to go to a community-based learning center to get a formal school diploma.(Cluster 2) Some Cluster 2 and Cluster 3 providers indicated that they would offer temporary institutional care for Didi and family support with the goal of family reunification: Didi has the right to obtain the fulfillment of formal education, health, identity, and services without discrimination.On the other hand, Didi's mother must be visited and offered social assistance to enhance her social functioning to care for her biological child.(Cluster 3) Interestingly, family-based alternative care was rarely mentioned as a care option.Only one Cluster 1 provider suggested kinship care as a response to the case of street children, saying: "Find information about the whereabouts of an extended family and whether the client can be cared for by an extended family."(Cluster 1).
One Cluster 2 provider indicated it could utilize its resources to prevent children from living or working on the streets through various prevention services.
[For] handling street children, shelter houses no longer exist.In the community, [we] educate the parents of street children.This is carried out regularly, [including] explaining the risks of trafficking.
[We] coordinate with anti-trafficking organizations to prevent child trafficking.The focus [of services] is social rehabilitation.(Cluster 2) Findings from the cluster analysis and open-ended questions are combined in Fig. 2. below.It shows the type and range of services that each provider Cluster delivers.

Discussion
Cluster analysis identified three groups of child welfare providers responding to the needs of neglected children in the City of Bandung.The largest cluster (Cluster 1; 50 % of providers) were faith-based, nongovernment Child Welfare Institutions (LKSA) for orphaned and neglected children that operated only in Bandung and received funding from non-government sources, including philanthropic organizations and private donations.Cluster 2 (26 %) providers were governmentmanaged and delivered child protection services and non-government, secular LKSA that received government and non-government funding and only operated in Bandung.Cluster 3 (24 %) providers were nongovernment LKSA that received non-government funding but were secular organizations with a regional service footprint.
Descriptive analysis of the providers' organizational characteristics indicated that service providers are primarily community-based organizations with smaller number of children in care, contrary to international experts' perceptions of childcare institutions as large-scale government organizations, such as those in Post-Soviet bloc countries (Ismayilova et al., 2014).The analysis suggested that service providers tended to provide services for school-aged children instead of families and offered single services instead of services along a continuum.This situation conflicts with the objective of the National Standard of Care to transform LKSA into family service centers, highlighting that a continuum of child welfare services to reduce reliance on institution-based care, has not been fully established.
Responses to the hypothetical scenarios of children at risk of family separation showed that each cluster had a unique response, reflecting a particular orientation in service delivery.Cluster 1 providers were far more inclined than Cluster 2 and Cluster 3 providers to offer institutional care.Fewer Cluster 1 providers indicated gatekeeping efforts and provision of family support services as an alternative to institution-based care.Cluster 1 providers generally could not provide a preventive response to an unplanned pregnancy.Cluster 2 and Cluster 3 providers were more able than Cluster 1 providers to provide family support services and family-based alternative care.The services offered in Cluster 2 were mainly family support.While Cluster 3 providers offered institutional care, some provided family support services as well.
These results suggest that the scope of services for neglected children in Bandung is still limited.Some progress should be noted since the "Someone that Matters" research in 2007 where gatekeeping mechanisms were not implemented (Martin & Sudrajat, 2007).However, at least half of the non-government Child Welfare Institutions (LKSA) in the study only provided shelter, food, and education, and did not offer family services.The findings suggest that implementation of the National Standards of Care has only influenced a small number of LKSA to transform into family service centers, or to provide services at the prevention end of the child welfare service continuum.Based on their service profile, it appears that LKSA in Cluster 1 were traditional child welfare institutions that had not transformed or were at the early stages of transformation.These institutions offered some gatekeeping and referral, but their scope of services was primarily institutional care.Cluster 3 providers were further ahead in the transformation process, offering both institutional care and family support services (such as economic support programs).Child Welfare Institutions in Cluster 2 offered services at the prevention end of the service continuum in the form of family support services.
When the findings from the cluster analysis and responses to the opened ended survey items are taken together, it appears that the capacity of LKSA to transform is influenced by government funding arrangements.In Cluster 2, child welfare providers that received a combination of government funding and private donations were more likely to offer family support services when a child was at risk of family separation than providers in Cluster 1 and Cluster 3 that only received funding from private donations.Service providers in Cluster 2 are also government child protection services mandated to provide family support services.This finding is consistent with the experience of deinstitutionalization in other countries, which indicate that service transformation requires significant funding from the government and non-government sectors, including local private donors to provide family support services instead of institution-based care (Goldman et al., 2020;Wilkea et al., 2020).
At the beginning of deinstitutionalization, MoSA provided the same financial and technical support for all Child Welfare Institutions (LKSA) to provide direct services to children and families.However, this research suggests that LKSA may require different types and levels of financial and technical support to enable them to reorient their services towards a continuum of care.Further, it appears that the national government's support for deinstitutionalization has been wound back.The Social Welfare Program for Children (PKSA) that combined cash transfer for children in institutions with case management and direct support from social workers to Child Welfare Institutions ceased in 2015 (Kaybryn et al., 2015;Ministry of Social Affairs & UNICEF, 2015).Since then, the government has not launched any specific programs to support Child Welfare Institutions.
Religious affiliation of child welfare providers is another factor that appears to be affecting deinstitutionalization.Child welfare providers in Cluster 1 and Cluster 3 only receive funding from private donations, yet secular child welfare institutions in Cluster 3 appear to be expanding their scope of services without government funding, while faith-based child welfare providers in Cluster 1 show little signs of transformation.Indeed, the service response of Cluster 1 child welfare providers appeared to be influenced by religious values in some respects, such as not providing services for children conceived out of wedlock, and prioritizing children's upbringing in an environment such as an institution that provides structure and fulfills children's basic needs.Child Welfare Institutions in Cluster 1 may also be hesitant to change their scope of services due to concerns that this would lead to a loss of funding because of the entrenched community practice of giving charity donations for neglected children in institutions combined with the fact that charity is an important tenet in in Islam (Davidson et al., 2017;McLaren & Qonita, 2019).
Unlike South-Sahara countries and some countries in the Asia region, faith-based and secular institutions in Indonesia do not receive or attract much international funding.The ability of secular Child Welfare Institutions in Cluster 3 to attract private funding for family support services may indicate changed perceptions of the importance of family support services on the part of service providers and donors alike.Further investigation is needed to understand factors that influence the mission and values of secular child welfare providers that are more supportive of deinstitutionalization.
Achieving deinstitutionalization in Bandung requires the transformation of many child welfare providers, especially the traditional, faith-based child welfare institutions that make up most child welfare service provision in Bandung.While a relative lack of government funding for family support services may be a key barrier, further information is needed about the barriers and enablers of institutional transformation.This will be explored in phase two of the current study.

Limitations
This study only recruited Child Welfare Institutions registered with the Ministry of Social Affairs and the National Communication Forum of Child Welfare Institutions (LKSA) in 2019-2020 or referred by MoSA social workers in Bandung.As the number of non-government child welfare institutions is unknown, the study may not capture operations of non-regulated institutions.Although the response rate was above 50 %, a few survey responses were insufficient to be included in the analysis due to missing data.The study was conducted during the COVID-19 pandemic when the study site was under restrictions, preventing the researcher from developing initial communications with participants and supporting them in completing the survey through face-to-face interaction.

Conclusion
This study provides insight into the transformation of child welfare services for neglected children in Bandung, West Java, following the national deinstitutionalization process that commenced in 2011.Almost ten years on, the study identified three clusters of providers: (i) traditional non-government Child Welfare Institutions (LKSA) with faithbased affiliations and private funding with local coverage, (ii) government child protection services and transitioning secular nongovernment LKSA with combined government and private funding with services in Bandung, and (iii) transitioning non-government secular LKSA operating with private funding and serving the whole province of West Java.The study found that only a small proportion of LKSA provide family support services and alternative family-based care with gatekeeping activities for neglected or at-risk due to neglect children.
The transformation of the LKSA to centers of family services is a key strategy of Indonesia's deinstitutionalization process.The findings of this study recommend the national government considers the varying capacity and stages of the transformation of LKSA in developing a continuum of child welfare services within the national system and in providing support for LKSA.Within a continuum of services, a limited number LKSA may continue to provide institutional care while more LKSA are supported to provide prevention, family support services, and family-based alternative care.The provision of government financial and technical support should follow the designated role of the LKSA within the continuum.
Although the study suggests the availability of government funding and religious beliefs and values may be significant, other community, organizational and governmental factors may also be influencing the deinstitutionalization process.Phase two of the current study will investigate further the factors enabling and hindering the emergence of a child welfare continuum in Bandung, West Java, Indonesia.

Fig. 2 .
Fig. 2. Clusters of child welfare service providers and scope of services in the survey.

Table 2
Human resources capacity of child welfare service providers (N = 50).

Table 3
Scope of services of the child welfare service providers (N = 50).
* n < 50 due to missing values in some variables and value percentages are presented excluding the missing values.

Table 4
Clusters of child welfare service providers (N = 50).be economically empowered while Ana continues her schooling.To support Ana, coordinate with several other government agencies such as the Office of Social Service, National Zakat Agency, Office of Education, Ana's school, and the family's sub-district.(Cluster 2) and