In depth: Variation in child welfare practice—what should be examined before drawing conclusions?
Updated: Oct 14
Written by Tanyah Cain, special educator with a master’s degree in psychosocial learning difficulties, Cain MindSYNC.
Purpose and position: This article argues for a substantial need for research into lived experience. Research gaps must prompt closer scrutiny of child welfare practice; they must not be used to dismiss children’s and parents’ accounts. Intrusive decisions and substantial resource use demand better knowledge about the decision-making basis, adapted support and actual outcomes. Experiences shared with FvsB help formulate questions to investigate; they are not presented as representative prevalence estimates.
The practical question is where children and families experience gaps between assessment, promised support and the help actually delivered. We need to examine relevant alternative explanations, suitable adaptations and whether intervention changes life for the better.
An in-depth professional discussion in the series on children and child welfare.
Fair services need sound comparisons
Municipal differences may signal a need to examine practice. A difference between numbers does not reveal why it arose. This article considers learning from variation without ranking families or services on a weak basis.
Equitable help does not mean identical interventions. Different needs may require different responses. The question is whether comparable needs receive relevant support and understandable, defensible assessments.
AI-generated illustration. The image does not depict a real event.
A Norwegian study
Vis and colleagues’ 2023 retrospective cross-sectional study examined 883 cases in 16 agencies. Service provision ranged from 21.9 to 60 percent. Much variation was explained by case factors, with small adjusted threshold differences. It studied service provision, not the quality of all care orders. [1]
Relevant comparison matters. The study does not establish that all inequality is explained. Retrospective data depend on records and the variables analyses can address.
Three problems with simple rankings
Needs may differ: services receiving more serious concerns may respond differently. Access to interventions can also differ. Low intervention rates may relate to capacity, other services or recording practices.
The measure may be unclear. Many closed cases could indicate resolved concerns or warrant examination of missed support. Many interventions may reflect availability without proving fit. Numbers need content and outcomes.
Examine several levels
Lauritzen and colleagues describe decisions shaped by case factors, practitioners, organisations and external conditions. This helps formulate questions rather than assign blame through municipal statistics. [2]
In a fictional example, two children have identical absence rates but different learning environments, health and support. A raw comparison misses this. Local procedures should still be examined for unequal treatment of comparable needs.
Proposed quality analysis
I recommend asking across comparable cases: Were needs examined? Was relevant support available? Were decisions explained? What did the child find useful? Describe differences between the cases.
Cross-service review can explore reasoning and alternatives to identify strengths and possible shortcomings. A high or low intervention count should never alone become a target for rewarding staff.
Include children’s outcomes
Evaluation should examine safety, education, relationships and experienced support according to purpose. Many meetings or short processing times are not themselves measures of the child’s situation.
Services handling harder cases may show poorer observed outcomes despite helpful work. Research needs baseline conditions, change and suitable comparisons rather than final outcomes alone.
Experience data as a complement
Organisations can examine experienced differences, separating documented events from participant interpretation. Explain recruitment and perspectives it may miss.
Include positive, mixed and negative experiences. Self-selected accounts may yield important insights without representing differences across all Norwegian services.
Evidence and scope
This is a professional discussion of selected sources, not a systematic literature review. Research descriptions use available abstracts and verified publication pages. Public guidance was checked on 13 October 2026. Qualitative findings illuminate experience without automatically establishing prevalence or causation. Practice proposals and fictional examples are a professional synthesis without established effect sizes. No new FvsB data were analysed.
Comments