Using Linked Administrative Data to Examine the Educational Outcomes of Children in Care in Manitoba
26/08/2016 | 09:25 - 09:45     Auditorium

Marni Brownell
Manitoba Centre for Health Policy, University of Manitoba

Presentation Type: Plenary

Themes: Applied projects, Data and linkage quality and Delivering and measuring impact

Session: Plenary F

Authors:

Marni Brownell, Mariette Chartier, Wendy Au and Leonard MacWilliam


Objective:

Manitoba has one of the highest rates of children in out-of-home care in the world. Our aims were to: 1) describe characteristics of children in care; 2) determine differences in educational outcomes for children in care, children receiving in-home protection services and children not involved with child welfare; and, 3) identify factors contributing to educational success for children in care.

Approach:

Population-level administrative data from the Population Health Research Data Repository at the Manitoba Centre for Health Policy, were used. Several data sources were linked together at an individual level: child welfare services (e.g., in care, in-home services; hospital birth records (e.g., birth weight, gestational age); health status (e.g., from hospitalizations, physician visits); family risk factors (e.g., maternal prenatal smoking and/or drinking, maternal education); family characteristics (e.g., number of children in family, receipt of income assistance); education assessments at kindergarten, grades 3, 7, 8, 9 and 12. Children in care from 1998 onward were included. Number of children analyzed differed depending on the educational outcome examined: e.g., analyses at kindergarten include 1,893 children in care, 4,229 children receiving in-home services, and 32,280 children not involved with child welfare services; for high school completion these numbers were 11,234, 40,671, and 162,265 respectively. Multivariate regressions identified characteristics associated with educational outcomes for children in care.

Results:

Over one-third of children in care first entered care as infants, and of those, close to half were apprehended at birth. 5.5% of children in Manitoba spent some time in care before their 15th birthday, with an almost tenfold difference for Indigenous (16.6%) and non-Indigenous (1.7%) children. Across the range of educational outcomes, children in care did more poorly than children receiving in-home services, who in turn did more poorly than children with no child welfare involvement. For example, only 33.4% of children in care completed high school compared to 66.8% of children receiving in-home services and 89.3% of children with no child welfare involvement. Adjusting covariates reduced differences but all remained statistically significant. Factors associated with positive outcomes for children in care differed depending on the outcome and included measures like poverty (at Kindergarten, p<0.05; at high school, p<0.0001) and total number of care placements (at Kindergarten, p<0.05; at grade 9, p<0.01).

Conclusion:

Children in care are at risk of experiencing poor educational outcomes, from school entry onwards. Identifying factors associated with educational success for children in care can inform policy and program development.


Conference Proceedings Published By

International Journal of Population Data Science