Long-term collateral effects of parent programs on child maltreatment proxies: Can administrative data provide useful insights?

Collecting child maltreatment data from participants is expensive and time-consuming, and often suffers from substantial attrition rates. Administrative population data may prove fruitful to overcome these barriers. The aim of this study was twofold: (1) to illustrate how administrative data may be used in evaluating long-term intervention effects; and (2) to examine collateral effects of three preventive early childhood interventions offered to families in the Netherlands (Supportive Parenting, VoorZorg, and Incredible Years). Using population data, four proxies of child maltreatment were assessed to examine collateral intervention effects: incidences of child protection orders, placements of children in residential care, crime victimization of children or their parents, and parental registrations as a crime suspect. The results revealed no significant differences between experimental and control conditions on any of these proxies, with very small effect sizes (ranging from Cramer's V = 0.01 to Cramer's V = 0.10). We conclude that the results do not provide support for collateral effects, but that studying other outcomes may provide this support. We further discuss that small sample sizes and low prevalences challenge studies using administrative data. Notwithstanding these limitations, we conclude that administrative data can strengthen the evidence base for collateral and direct intervention effects.