November 18, 2019
Statement from the Child and Family Research Partnership (PDF)
Researchers at the Child and Family Research Partnership at UT Austin’s LBJ School of Public Affairs recently published a paper in Child Maltreatment titled “The Effect of CASA on Child Welfare Permanency Outcomes.” This research uses a more rigorous design and larger sample than prior research on this topic. We find that Texas children in the child welfare system who are appointed a CASA volunteer advocate are less likely than similar children who are not appointed a CASA volunteer to be reunified with their parents or reach a permanent placement outcome. This finding should not be interpreted to claim that CASA is not effective, however, for the following reasons:
- Permanency is only one measured outcome for children who are in the child welfare system. Other outcomes, such as the child’s emotional health, connections to community and culture, and access to supports and services, are not measured and captured in the child welfare data, and therefore we could not examine the impact CASA has on these important outcomes.
- Although our study attempted to control for the fact that CASAs are assigned to more complex child welfare cases (for example the reasons for removal are more severe and there are more and older children on CASA cases), there are likely many factors not measured that contribute to case complexity that were not accounted for in our models.
- Results from the study may be misinterpreted to insinuate a larger difference in permanency outcomes between children with and without a CASA than the results actually find. The results are stated in odds ratios, which is a statistical term commonly used in research. To better understand the difference in permanency outcomes between CASA and No CASA children, it is clearer to share the results in predicted probabilities (which are estimated from the regression models). For example, these findings show that 32.13% of children without a CASA are reunified compared to 28.61% of children with a CASA, for a difference in 3.5 percentage points (in the paper, the difference is stated as 16% lower odds). Although the 3.5 percentage point difference is statistically significant, it may not be considered substantially large.
Predicted Probability of Permanency Outcome by CASA Status
|Permanency Outcome||No CASA||CASA|
|Adopted (if not reunified)||41.45%||44.79%|
|PMC to KIN (if not reunified or adopted)||79.26%||76.39%|
|No Permanency (if not reunified or adopted)||20.74%||23.61%|