In recent years, Norwegian child welfare services have been subject to heightened legal and political scrutiny, particularly following a series of rulings by the European Court of Human Rights (ECHR) between 2018 and 2023 concerning violations of the right to family life. These rulings primarily addressed restrictive contact regimes and insufficient efforts to support family reunification (Tellesbø, Meland, & Jullum, 2024). In response, legislative amendments were introduced in the new Child Welfare Act (2021), alongside national competence strategies aimed at strengthening professional practice and managerial accountability. The revised legal framework places increased emphasis on documentation, follow-up, and systematic evaluation of child welfare interventions, thereby intensifying demands on local child welfare managers to monitor practice and demonstrate compliance with human rights standards (NOU 2023:7; Prop. 133 L (2020–2021)). At the same time, nationwide inspections have identified persistent weaknesses in investigation practices and quality assurance, which supervisory authorities have explicitly linked to shortcomings in management and leadership (The Norwegian Board of Health Supervision, 2022). Together, these developments underscore the growing importance of administrative data as a governance and management tool in Norwegian child welfare services.

Taken together, these legal, regulatory, and supervisory developments have significantly reshaped the conditions for leadership in Norwegian child welfare services. Child welfare managers are increasingly expected to translate human rights obligations, statutory requirements, and national reform objectives into everyday practice, while simultaneously demonstrating quality, accountability, and compliance. Administrative data have thus become a central tool in managerial work, serving both operational and strategic purposes.

Norwegian child welfare services operate within a decentralized governance structure in which municipalities are responsible for the provision, organization, and delivery of services, while national authorities primarily the Directorate for Children, Youth and Family Affairs (Bufdir) provide institutional care and foster care arrangements. The Ministry define legal frameworks, guidelines, and reporting requirements. Municipalities have considerable discretion in how services are organized, including staffing, use of inter-municipal cooperation, and prioritization of resources. At the same time, they are subject to extensive national regulation, supervision, and mandatory reporting systems. This creates a hybrid governance model characterized by strong central regulation combined with local autonomy in implementation. Such decentralization is likely to shape how administrative data are used, as municipalities differ substantially in size, analytical capacity, and access to resources.

The use of management data as a tool in quality development in the public sector is a well-known phenomenon (Klausen, 2020). The public sector has been influenced by marketisation, which has introduced market mechanisms and concepts, such as an increased emphasis on cost-effectiveness, enhanced accountability, competition, and customer focus (Shanks et al., 2015). Consequently, public services are increasingly described as being “managerialized” (Noordegraaf, 2016). The increased focus on the use of management data can be understood within this context.

The challenges this development have brought to social work are well-documented (Hyslop, 2018). For instance, the Munro Review of Child Protection in England (2010, 2011) emphasized how a managerial focus on standardized assessments, productivity measurements, documentation, procedural adherence, and risk aversion weakens decision-making quality. Research from Scandinavia highlights that child welfare managers encounter significant challenges and dilemmas in their managerial roles. These challenges are particularly tied to an increased focus on control, quantitative performance targets, administrative tasks, and budgetary responsibilities (Tham & Strömberg, 2020; Shanks et al., 2015). Furthermore, the alignment of the sector with marketisation principles has introduced a cost-driven approach to handling child welfare cases (Ebsen, 2018).

As a part of the child welfare reform in Norway, national authorities have emphasized the use of management data in the effort to develop a knowledge-based child welfare service (Djupvik et al. 2019). The primary aim of the reform is to ensure that more children and adolescents receive the appropriate assistance at the right time. This necessitates the importance of robust data to assess whether this goal is being achieved (Pedersen et al. 2022, pp. 26-28). Empirical studies show that if performance management systems become overly complex and burden organizations with numerous performance indicators, targets, and benchmarks, this can increase costs related to data collection and analysis, diverting time, attention, and resources away from improvement activities (Raudla et al. 2023). Performance goals as a management tool bring the conflicting relationship between efficiency, quality and accountability of managerialism clearly to light (Reiter & Klenk, 2018). Many CWMs spend most of their time on daily tasks such as checking, reporting, and accounting (Olsvik & Solstad, 2024).

Managerialism and marketisation have led to changes in the conditions for child welfare leadership as CWMs have been given more responsibility for budgets, departments, and staff (Shanks, Lundström, & Wiklund, 2015; Olsvik & Saus, 2022, p. 464). This implies that leadership must be exercised in new ways and adapted to new situations. Leadership skills are increasingly in demand, and expectations towards managers are more explicit. In Norway, expectations and requirements for leadership in the CWS have been specified in a professional recommendation from the state level (The Norwegian Directorate for Children, Youth and Family Affairs, 2017). It contains a clear expectation that leadership should be professionalized (Olsvik & Saus, 2022, p. 465). There is a need to improve access to knowledge, and a greater focus should be placed on research into implementing knowledge-based measures (NOU 2023: 79).

To address complex problems and challenges in social work, many scholars (Shlonsky and Mildon, 2017; Gotvassli & Moe, 2019) emphasize the importance of maintaining a delicate balance between interprofessional collaboration and managing the relationships with families, both in terms of closeness and distance. This approach requires a comprehensive strategy that integrates prevention, investigation, and treatment within the leadership framework of CWS (Sanders, Jackson, and Thomas, 1996).

The concept of knowledge-based child welfare aims to integrate evidence-based methods with experiential knowledge. A knowledge-based approach encompasses not only empirical evidence but also the experiences of professionals, children, and families, and accounts for contextual factors (Moe & Gotvassli, 2023, pp. 226–227). A practice-based perspective on knowledge management (KM) rests on the understanding that knowledge is created in and through practice, shared within overlapping and interwoven practices, and impeded when these practices fail to interact with one another (Newell, 2015). In the context of child welfare leadership, this perspective applies to both internal personnel management and the leadership of interdisciplinary collaboration. Consequently, knowledge management is also essential in multidisciplinary collaboration, where interaction and coordination are highly emphasized.

KM in CWS is a broad concept that involves leveraging knowledge assets to benefit children, families, and the organization (Alavi & Leidner, 2001; Hislop et al., 2018, p. 50). According to CWS, it also highlights the importance of a process and dialogue with children and families, and of «knowledge as a process», a perspective from Alavi and Leidner (2001; Hislop et al., 2018, p. 50).

These include promoting more efficient processes within different parts of the organization to increase innovation and thereby providing a strategic advantage. KM is, therefore, more than just managing knowledge, implying that it is primarily a technological challenge that can be addressed by applying knowledge management systems. It also includes organizational factors such as culture, structure, human resource management (HRM), and leadership (Heisig, 2009). Consequently, KM is considered the management of knowledge processes and knowledge work rather than merely the management of knowledge itself (Newell, 2015; Hislop et al., 2018, p. 50).

Recent scholarship has expanded the discourse on KM within the public sector, moving beyond early foundational work to more nuanced conceptual and empirical analyses. For instance, Laihonen et. al. (2024) emphasize the social processes underlying knowledge formation in public organizations, suggesting alternative theoretical lenses for public sector KM. Knowledge formation is perceived as an iterative process in which policymakers play a key role in interpreting societal issues and in creating, using, and evaluating information to solve wicked problems and give meaning to public policy. Thus, a key aspect of knowledge management that enhances the technical process of information production is the collective interpretation process, in which information is constructed and justified, and where the core mission of public administration is defined (Laihonen et al., 2024, p. 227). Adequate administrative data and an information base are essential for making informed decisions in CWS. However, the mental models that guide selection and interpretation are necessary for developing KM and quality in public administration (Laihonen et al., 2024, p.229). There seems to be a necessary connection between leadership practice, as reflected in priorities, decision-making capacity, and collaboration in CWS.

Kassa and Ning (2023) conducted a systematic review on the role of Knowledge Management (KM) in the public sector, identifying three key themes: KM for organizational improvement, citizen satisfaction, and collaborative innovation management. Stella Hill (2025) offers a practical framework for understanding how KM is implemented across different public administrations. Other recent work has examined the sustainability of KM as an innovation in public reforms (Vargas, 2025) and tested empirical models linking KM with organisational outcomes such as learning and intellectual capital (Huerta-Chávez & Figueroa-Ochoa, 2023).

Limited knowledge exists regarding how administrative data is utilized by local child welfare authorities in Norway. This study explores how child welfare managers (CWM) across eight different municipalities in Norway utilize data to enhance service delivery and improve outcomes. By examining the types of data used, their employment, and the perceived strengths and weaknesses of these practices, this research provides insights into the current state of data-driven practices in CWS in Norway across diverse settings.

To investigate the current state of data-driven practices in CWS in Norway across diverse settings, we used a qualitative interview study, specifically utilizing semi-structured interviews. This approach was selected because it enables researchers to develop insights into nuanced accounts of the variation in how child welfare managers across various municipalities in Norway use data for leadership and planning.

The use of semi-structured interviews facilitates the identification of key themes related to data use, such as strategic planning, quality improvement, and compliance. This aligns with the research objective of understanding the impact of data-driven practices on child welfare services. Managers at various municipal levels were asked about their thoughts and experiences regarding leadership in CWS and the use of administrative data for planning and managing services.

The sample was recruited by sending requests via email, along with an information letter describing the project, to the chosen municipal directors. The request to participate was sent to twelve municipalities, of which eight agreed to participate in the study.

A total of sixteen leaders from eight municipalities were interviewed. Seven of whom were the chief municipal executive, one was the department head of health and social services, and eight were CWS managers. Leaders from mid and top-level management within the municipality were included to obtain the perspectives from leaders with different positions within the organisation.

The participating municipalities were selected through purposive sampling to capture variation in organizational and managerial conditions relevant to the use of administrative data in child welfare services. Selection criteria included (i) municipal size, (ii) organizational model for child welfare services (self-sustained services, inter-municipal host arrangements, and partial inter-municipal cooperation), and (iii) geographical diversity. The aim was not to achieve statistical representativeness, but to include municipalities operating under different structural and resource conditions that may shape managerial data practices.

In this study, municipalities were categorized pragmatically based on organizational scale rather than national distribution. Municipalities serving populations below 50,000 were classified as smaller, reflecting differences in organizational capacity and access to analytical resources. This categorization does not reflect the national distribution of municipality sizes in Norway, where the median population is substantially lower, but was used analytically to distinguish between varying administrative capacities relevant to the study.

The largest metropolitan municipalities were not included, as their child welfare services typically operate under substantially different conditions in terms of organizational complexity, with several agencies within the municipality, different levels of specialization, and access to analytical resources. Including these municipalities would likely have introduced dynamics requiring a separate analytical focus and was therefore considered beyond the scope of the present study.

Inter-municipal cooperation in Norwegian child welfare services is not mandated by law but is encouraged as a strategy to address capacity constraints, particularly in smaller municipalities. Such arrangements allow municipalities to pool resources, share specialized competencies, and achieve economies of scale. In host-municipality models, one municipality assumes responsibility for delivering services on behalf of others, which may influence both managerial autonomy and access to administrative data. These cooperative structures may enhance analytical capacity but can also introduce coordination challenges that shape how data are used in practice.

There were three self-sustained CWS organisations, meaning that they would do screening of referrals, investigations and service delivery only for the population within the municipality. Two CWS’, labelled ‘some local cooperation’, would have cooperation with other services for specific tasks, such as screening of emergency referrals or delivery of more specialised services. Three agencies were inter -municipal hosts, meaning that they would do the entire CWS process, i.e. screening, investigation and service delivery for one or more other municipalities in addition to their own. Geographical diversity was also taken into consideration.

The interview guide is designed to explore the use of management information in CWS in Norway. It was created to elicit the informant’s views and experiences related to three main themes:

Which types of information are seen as important for planning and management of the CWS, by CWS managers and municipal directors, respectively?

The interviews were conducted by two to three researchers online between December 2023 and May 2024. All the informants were interviewed individually, and the interview lasted approximately 45 minutes to an hour. An audio recording of the conversation was made with the web form “Dictaphone mobile app” and was transcribed into text. All the authors participated in conducting and transcribing the interviews.

We employed the approach to thematic analysis recommended by Braun and Clarke (2006), which provides a flexible method for identifying patterns and themes in qualitative data analysis. While the interview guide provided an initial deductive structure by defining broad thematic areas, the analytical process was primarily inductive. Codes were developed from the data, using the informants’ own language as far as possible, and were refined through collaborative discussion within the research team.

The analysis was conducted both individually and collaboratively within the research team. In the first phase, all members of the research team familiarised themselves with the data. One of the authors wrote a memo that served as the basis for discussion within the entire group. This formed the foundation for the second phase, during which the research team collectively discussed potential codes. We emphasised that the coding should reflect the language used by the informants. Once we had a draft of the codes, we revisited the interviews to ensure that the codes accurately reflected the data material. The development of themes occurred during the third and fourth phases, with the research team engaging in multiple rounds of discussion before finalising the themes in the fifth phase. Initial codes and sub themes were developed concerning each of the main topics of the interview guide, i.e. (i) types of data used (seven types identified), (ii) how that data is used (four sub-topics) and (iii) perceived strengths and challenged associated with data usage (Table 2).

The themes were reviewed and refined to ensure they accurately represented the data. This involved checking whether the themes were effective concerning the entire dataset, resulting in a thematic map of the analysis. The themes formed a basis for identifying differences and similarities across informants. A summary of the four main themes was created for each informant. Furthermore, tables were created comparing which subthemes were mentioned by each informant and comparing themes between participants from small and large municipalities. Cross-municipal comparisons were conducted through thematic summaries and simple comparison matrices, enabling systematic examination of similarities and differences across municipalities with varying size and organizational models.

The analytical focus of this study is on identifying cross-municipal patterns in managerial use of administrative data rather than on reconstructing individual narratives or meaning-making processes. Given the small number of informants in senior leadership positions and the need to ensure strict anonymity across municipalities, extensive use of illustrative quotations was considered of limited analytical value. Achieving full anonymity would require substantial contextual stripping of quotations, which would remove much of the organizational and managerial context necessary for meaningful interpretation. Instead, the analysis emphasizes systematic cross-case comparison, thematic synthesis, and structured summaries to ensure analytical depth and transparency. This approach is consistent with the study’s purposive sampling strategy and comparative design, where the primary contribution lies in elucidating patterns across organizational contexts rather than in presenting individual accounts.

This study adheres to ethical guidelines for qualitative research and, following Norwegian practices, was reported to SIKT, the Norwegian Agency for Shared Services in Education and Research, which confirmed that the processing of personal data complies with relevant regulations. Informed consent was obtained through an information letter and by providing participants the opportunity to ask questions before the interviews commenced. To ensure anonymity, all municipalities and informants have been pseudonymized.

The interviews reveal a diverse utilization of data types to manage and enhance child welfare services. The CWS employ administrative client data from electronic systems of record-keeping, key performance indicators, quality indicators, internal control data, financial data, user surveys, and qualitative data from client evaluation. An overview of the key characteristics of data types used is shown in Table 3.

The administrative data described above originate from multiple sources within the Norwegian child welfare system. A substantial proportion of data such as client-level information and internal reporting data, and financial data is generated and maintained by municipal child welfare services through electronic systems of recordkeeping. In addition, municipalities are subject to mandatory reporting requirements to national authorities, including Statistics Norway, which compiles and aggregates data across municipalities. The case level mandated reporting consists of events such as reports, investigations and servives provided for each end every child in contact with CWS. Each event has a start and end date and is linked to the childs national indentity number. These national systems produce standardized indicators that are subsequently used for benchmarking and performance assessment. Performance evaluations are performed by supervisory bodies at the state level. Thus, administrative data used by managers are shaped both by local data production and nationally defined reporting frameworks.

Taken together, the table illustrates that administrative client data and performance indicators form the core informational infrastructure for managerial work, while user-centred and qualitative data sources are used more selectively.

Administrative data, such as extracts from electronic systems of recordkeeping (ESR) and the agency aggregate of those data, accessible through Statistics Norway, are commonly used across agencies to track and manage cases and ensure compliance with statutory requirements. This type of data supports the operational management of services and aids in the transparency and accountability of the agencies. Statistical data at the agency level, including key figures like the number of reports, investigations, and interventions, is used for operational and strategic decisions, helping managers adjust services to meet community needs better.

User feedback, although not systematically collected by all agencies, plays a role in several agencies where it is used to gauge the effectiveness and reception of services by parents, predominantly. This feedback is then reviewed in team meetings, and both statistical and narrative forms may be used. However, the underutilization of this feedback in several agencies indicates a potential area for improvement. Overall, user feedback appears as a complementary but underutilized data source, with considerable variation across municipalities.

Data usage practices varied significantly among the municipalities. Each CWS employs data uniquely, tailoring its use to meet specific administrative, operational, and strategic needs.

Common uses included strategic planning, quality improvement, resource allocation, and compliance with legal standards. Two of the largest municipalities (agencies 4 and 8) utilized data primarily for strategic planning and quality improvement, while smaller agencies (1 and 6) emphasized compliance and operational monitoring.

Operational Management: Data is used for day-to-day management across the agencies. This includes tracking and managing child welfare cases, monitoring service quality, and ensuring compliance with statutory requirements. Administrative data, such as extracts from ESR, supports the operational management of services, enhancing transparency and accountability within the agencies. One agency also has key personnel with specialized analytical skills that help prepare and analyse data. The manager of this agency states: «These key figures are both used for analysis and to see how we can further develop our services about other services in the municipality». This use of data was common across municipalities, though it was particularly emphasized in smaller agencies where compliance and case flow dominated managerial priorities.

Strategic Decision-Making: CWS use data to inform decisions at leadership meetings, helping adjust services to meet community needs better and align CWS with broader municipal health and social goals. For instance, the data in CWS 8 is used to inform decisions at leadership meetings and helps adjust services. In CWS 2, data from comparisons with other municipalities, made available through the Norwegian Statistical Agency, helps assess service performance and identify areas needing attention. Strategic use of administrative data was reported primarily in larger municipalities with greater analytical capacity and access to comparative data.

Quality Improvement: Several agencies highlight the use of data for quality improvement, where it aids in identifying areas needing attention and evaluating the effectiveness of interventions using feedback from families. Agencies 3 and 4 use feedback data from families for evaluating the effectiveness of services, identifying areas for improvement, and planning developmental projects.

Reporting: Data is also used extensively for reporting purposes to various stakeholders, including municipal management. Graphical representations of data are produced to track changes, which are then discussed at various management levels within CWS.

Taken together, these findings suggest a clear distinction between municipalities where administrative data is primarily used for operational compliance and monitoring, and those where data is more actively integrated into strategic planning and quality improvement work.

The potential strengths identified include enhanced monitoring capabilities, user-centred approaches, systematic improvements, and strategic alignment with broader community goals.

Enhanced Monitoring and Compliance: A common strength mentioned across several agencies is the use of data to monitor service quality and ensure compliance with statutory requirements. For instance, CWS 1 utilizes key performance indicators and quality indicators to conduct detailed and regular reviews, which help in assessing compliance and the effectiveness of interventions. This systematic monitoring enables agencies to track standards of service and identify areas that require attention or improvement.

User-Centred Approach: Several agencies emphasize the importance of incorporating feedback directly from service users, which fosters a user-centred approach to quality improvement. For example, CWS 1 conducts user surveys after each case to gather feedback, which is then reviewed during team meetings to adjust practices and address identified issues. This direct feedback from service users provides valuable insights into the effectiveness and reception of services, ensuring that the services are responsive to the needs and experiences of those they serve.

Systematic Quality Improvement: The use of data for systematic quality improvement is mentioned as a notable strength in agencies like 3, where participation and feedback from families help reduce the need for placing children in foster homes. Similarly, CWS 4 emphasizes the ability to generate detailed reports helping to understand and improve service delivery, particularly through new software system implementations.

Strategic Planning and Alignment: Data-driven strategic planning is a priority in CWS agencies (CWS 8), where comprehensive data collection and analysis support strategic decisions that align child welfare services with broader municipal health and social goals. This aligns with the objectives of recent welfare reforms in Norway that emphasized better integration of services and coordination of efforts across sectors. Such strategic use of data is thus seen as a contribution to enhancing the overall impact of child welfare initiatives.

Network Engagement and Learning: Engagement in the networks for larger municipalities, as mentioned by CWS 4, provides a platform for learning and sharing best practices with other large municipalities. This network engagement is seen as a significant strength as it allows agencies to benchmark against others. Smaller municipalities (CWS 1,2 & 3) have developed local networks based on their own initiatives. These are to a lesser degree motivated by and used for comparison of administrative data, focusing instead on learning from each other’s experiences.

Despite these strengths, challenges related to data management efficiency, system capabilities, and the need for enhanced data analysis skills are noted as areas for further improvement. Addressing these challenges will be important for maximizing the benefits of data utilization in CWS.

While all municipalities emphasized monitoring and compliance as key strengths of data use, larger municipalities more frequently described the use of administrative data for strategic planning, system development, and cross-sector coordination. Smaller municipalities, in contrast, highlighted the role of data in maintaining oversight and managing statutory requirements within limited resource frames.

Despite the benefits, several challenges were identified, including challenges in data quality and integration, resource constraints, and the complexity of data analysis. For example, several agencies reported difficulties with new IT systems and the integration of data sources, which hindered the utilization of data. Additionally, the complexity and resource-intensive nature of data collection and analysis were common concerns across several municipalities. Other challenges revolve around the complexity of data management, reliance on manual processes, system inefficiencies, underutilization of user feedback, and the need for enhanced data analysis skills.

Complexity and Volume of Data: A significant challenge highlighted by several agencies, such as CWS 1, is the high volume and complexity of data, which can be difficult to manage effectively. This complexity often leads to challenges in extracting meaningful insights, especially when coupled with tight reporting deadlines. Managing this complexity requires robust data management systems and processes, which some agencies currently lack.

Reliance on Manual Processes: The reliance on manual processes for data extraction and analysis is another notable weakness. For instance, CWS 1 mentions the use of manual processes for analysing certain types of data, which can introduce inefficiencies and errors. This reliance not only slows down the process but also increases the workload on staff, potentially leading to delays and reduced accuracy in data handling.

System Inefficiencies and Transition Challenges: Several agencies face challenges related to inefficiencies in their current data systems or issues arising from transitions to new systems. CWS 5 discusses significant challenges with the transition to a new data system, which has been problematic in terms of data migration and accuracy. Similarly, CWS 8 mentions disruptions in data accessibility and management due to recent transitions to new data systems. These system inefficiencies can hinder the ability to monitor key metrics effectively and make informed decisions.

Underutilization of User Feedback: Although some agencies like CWS 1 effectively utilize user feedback to improve services, others do not systematically collect or integrate this feedback into their quality improvement processes. For instance, CWS 8 acknowledges that feedback from service users is not systematically collected, which could hinder the ability to understand service impact and areas for improvement fully. CWS 2 solicits only verbal feedback during meetings but aims to create a standardized template for user surveys following the child welfare assessments. This underutilization of user feedback is by some seen as a missed opportunity to engage service users and tailor services to better meet their needs.

Need for Enhanced Data Analysis Skills: A recurring theme across the interviews is the need to have access to enhanced data analysis skills. Several agencies recognize that they do not have the skills to analyse complex data sets (CWS8), which may affect the depth of data-driven insights. This skills gap can limit the agency’s ability to leverage data fully for strategic decision-making and continuous improvement. Taken together, these challenges point to a structural mismatch between increasing expectations for data-driven governance and limited analytical capacity at the local level, particularly in smaller municipalities.

The interviews with managers from CWS in both small and large municipalities reveal distinct differences in terms of data usage, strengths, and weaknesses, mainly influenced by the scale of operations and resource availability.

In small municipalities, data usage tends to focus heavily on operational management and compliance with statutory requirements. These agencies often rely on administrative data and key performance indicators to monitor service quality and ensure regulatory adherence. For instance, one agency utilize administrative data for detailed service quality reviews. However, strategic decision-making and sophisticated data analytics appear less emphasized, possibly due to limited staff expertise and technological resources. In contrast, several larger municipalities often employ data in more diverse ways. One CWS uses comprehensive data collection for strategic planning and aligning CWS with community health and social goals. Small municipalities often demonstrate strengths in focused areas, such as enhanced monitoring and compliance, as seen in one agency’s detailed and regular reviews that utilize key performance indicators. These agencies may also exhibit a strong user-centred approach, leveraging direct feedback from service users to tailor services effectively. The collection and use of user feedback aim to balance and address some of the shortcomings in aspects of CWS services that administrative data typically does not cover. This can be seen as an attempt by CWS managers to move beyond managerialism (Noordegraaf, 2016) and create learning organisations for the improvement of services, a call sounded by Munro almost 15 years ago (Munro, 2010, 2011). On the other hand, larger municipalities benefit from broader resource bases and typically exhibit strengths in systematic quality improvement and strategic planning. Thus, they can engage in network learning and implement best practices from other large municipalities, which can lead to innovative service delivery and continuous improvement.

The weaknesses observed in small municipalities often stem from resource constraints. They may struggle with the high volume and complexity of data, which can be overwhelming for the limited staff available. Additionally, administrative data may be less valuable for statistical analysis and cross-municipal comparison in smaller municipalities where there may not be enough cases to allow for direct comparison with others. Larger municipalities, although generally better equipped, still face challenges, such as the difficulties associated with transitioning to new data systems and the need for enhanced data analysis skills among staff. They may, for instance, face challenges with system inefficiencies and underutilization of user feedback, which can hinder their ability to fully leverage data for service improvement.

In summary, while both small and large municipalities utilize data to enhance child welfare services, the scale of operations and resource availability significantly influence their data usage practices, strengths, and weaknesses. Addressing these tailored challenges is necessary for optimizing the effectiveness of child welfare services in different municipal contexts.

The use of administrative data for quality improvement within services presents both significant possibilities and challenges. On the positive side, administrative data can provide insights that help in strategic planning and resource allocation. However, the challenges are equally notable, particularly concerning data integration and the relevance of the available data. As noted by several informants, particularly the CWS managers, the available administrative data tends to focus on counting what can be easily quantified, typically deadlines and formal requirements in case processing. Quality in terms of short- and long-term outcomes for children and families is, however, a more complex phenomenon that routinely collected data cannot fully understand. We can understand this as a conflict between professional assessments and “managerialised” measurement tools, which come at the expense of professional quality. It shows the difficulty of creating bridge-building, coherence, and consistency using only quantitative management data (Klausen, 2020). This may be the reason why some CWS also collect information about their operations directly from users, both through routine feedback evaluations and by more qualitative measures. This is believed to help complement, at least some of the “blind spots” regarding service quality that are perceived to exist within administrative datasets.

The capacity and capabilities that are needed from CWS are not only dependent upon the amounts and types of referrals they receive during operation but are connected to how services for children and families more broadly are organized and the capacity within the health and educational sectors. An informant pointed out the difficulty in obtaining a holistic view from the data available, stating the need for a “system that could aggregate data across different sectors,” which suggests a gap in current capabilities to synthesize and interpret complex data sets effectively. This calls for greater integration of different sectors and a more interdisciplinary approach that can help overcome the strong fragmentation of the public sector created by managerialism and marketization (Reiter & Klenk, 2019). This is in line with the experience from projects abroad that have attempted to assist local authorities in making better use of administrative data (Trocme et al., 2019; Davey et al., 2022). Although there are many possibilities for linking administrative data from different sources to individuals, through personal identification numbers, in the microdata service provided by the Norwegian statistical agency, this does not seem to be widely used by local authorities. Probably because significant skills in data analysis are required, a point we will revisit. Moreover, issues of data timeliness and accuracy can hinder the ability to make informed decisions promptly. These examples illustrate the dual nature of administrative data’s role in service improvement. While offering a path toward more informed and strategic management, it also demands robust systems and processes to handle the complexities involved effectively.

The competence requirements for better utilization of administrative data vary between the agency and municipal levels, each demanding specific skills and knowledge. At the CWS agency level, there is a need for specialized data analysis skills to interpret and apply data insights directly to service improvements. This requires a deep understanding of what the data represents and what it does not say anything about. To effectively utilize such data, it is necessary to be familiar with both the Child Welfare Act and the organization of child welfare services. Additionally, technical proficiency in data handling and analysis is required, as it is crucial for translating complex datasets into actionable strategies. It is probably a rare find to have both areas covered by one single person. Routines are therefore necessary for professionals and administrative personnel to work together within the agency to maximise the potential of administrative datasets.

On the municipal level, the competence requirements extend beyond data analysis to include strategic data integration across various departments and services. This requires municipal employees not only to understand data but also to facilitate its integration, providing a comprehensive view that supports broad strategic decision-making. These examples highlight the distinct, yet complementary competencies needed at different governance levels to enhance the effectiveness of administrative data in public service management.

As previously discussed, the Norwegian Board of Health Supervision (2022, p. 24-30) concluded that professional failure is a symptom of inadequate quality assurance by management. The criticism pertains to various aspects of leadership. A national committee was tasked with assessing measures to strengthen the rule of law throughout the child protection system. The committee holds the view that all informational materials, statistics, and overviews related to the child protection system should be consolidated in a centralized child protection portal. Furthermore, regular reporting should be introduced regarding the status of the child protection sector about human rights obligations, particularly concerning the United Nations Convention on the Rights of the Child and the European Convention on Human Rights (Summary, NOU 2023:7).

Both Human Resource Management and the monitoring of the quality of CWS are highlighted as areas of concern (NOU 2023:7, p. 3). There is a lack of statistics and information on key aspects of child welfare operations, which adversely affects the quality of support provided to children and parents (NOU 2023:7, p. 13). The committee proposes the establishment of research and analysis centers, as well as a data infrastructure for the child protection sector, through the collection, compilation, linkage, dissemination, and accessibility of data across the full scope of CWS (NOU 2023:7, p. 104).

A child welfare system characterized by knowledge management includes a focus on culture, structure, and Human Resource Management (Hislop et. al., 2018), as well as an emphasis on evidence-based practice. While measuring this can be complex, it is essential to strive for robust governance data that contributes to illuminating and enhancing quality. As part of knowledge management and the professionalization of leadership, there is also a need for knowledge-based measures in relation to understanding and analyzing management data. One MD emphasized the importance of CWS in determining what key performance indicators are important and ensuring that the data being monitored is meaningful. This illustrates that, in addition to strong analytical skills, professional knowledge management in child welfare is also required, emphasizing various aspects of the work. This concept encompasses a broad approach that involves leveraging knowledge assets for the organization’s benefit (Alavi & Leidner, 2001; Hislop, 2018). To address complex problems in Norwegian Child Welfare, it is crucial to have interprofessional collaborations, utilize various types of power, and strike a balance between professional proximity and distance to families (Shlonsky & Mildon, 2017; Gotvassli & Moe, 2019). Knowledge management, as reflected in administrative data, also encompasses perspectives on organizational factors such as culture, structure, and leadership (Heisig, 2009; Hislop et al., 2018). It is also important to consider the management of knowledge processes in the CWS. At the same time, instances where administrative data is used for reflection, coordination, and service improvement resonate with theories of learning organizations, suggesting that data can support organisational learning when analytical capacity and leadership resources are available.

Based on the findings, several recommendations can be made to enhance the effectiveness of data utilization in child welfare services. Municipalities should focus on integrating different data systems to provide a more comprehensive view of services.

Furthermore, developing training programs for child welfare staff on data analysis and usage could significantly enhance the effectiveness of data-driven practices. By equipping agencies with staff that has necessary skills to interpret and apply data insights, municipalities can foster a more analytical and evidence-based approach to service delivery. Finally, standardizing data collection and analysis practices, while allowing for customization to meet local needs, may help balance the need for consistency and flexibility. This could involve the definition of specific metrics and indicators to be consistently tracked across all municipalities. The core elements of the framework could include identifying key metrics essential for evaluating child welfare services, such as the types of referrals and re-referrals, types of interventions, duration of services, compliance with statutory requirements, and outcomes for children and families. Indicators should be chosen based on their relevance to assessing service quality and effectiveness. Recognizing the diversity in municipal contexts, the framework should be flexible enough to accommodate local variations. Municipalities could add additional metrics that reflect local priorities or challenges while still adhering to the standardized core elements.

The scientific strengths of this study lie in its qualitative content analysis approach, which provides a detailed, context-rich exploration of how administrative data is utilized across various municipalities, offering insights into the practical application of data in child welfare services. The inclusion of multiple municipalities enhances the generalizability of the findings within Norway, allowing for a nuanced understanding of regional variations and practices. However, several limitations should be acknowledged. Firstly, the depth of qualitative analysis, while comprehensive, is inherently limited by the sample size and the scope of the interviews conducted. With 16 interviews across eight municipalities, the findings may not capture the full diversity of practices and experiences present in all Norwegian municipalities particularly because we did not include any of the three major metropolitan area cities. This limitation suggests that the themes identified may not be exhaustive or fully representative of all possible perspectives within the child welfare sector. These largest cities may have other capabilities and needs than what is discussed here, potentially affecting the generalizability of the study’s conclusions. Despite these limitations, the study offers a foundational understanding of administrative data use in child welfare services in Norway and suggests areas for further exploration and improvement.