Constructing Safety Indicators from Child Welfare Events and Trajectories

1
Constructing Safety Indicators from Child Welfare
Events and Trajectories

• Fred Wulczyn, Bridgette Lery
• Center for State Foster Care and Adoption Data
• Chapin Hall Center for Children
• University of Chicago
• June 28, 2007

2
Questions

• At what stage or stages of involvement in the
  child welfare system does disparity originate?
• Do subsequent stages aggravate or mitigate the
  disparity?
• Policy question
• Do policies interact, contributing to
  disproportionality, in the name of other positive
  outcomes?

3
Definitions

• Disproportionality
• One population is out of proportion with respect
  to a reference population
• Disparity
• A lack of equality likelihood of placement,
  likelihood of exit, time to exit, and exit type

4
Trajectories

• Trajectories are strings of events in temporal
  sequence. Captured events in this jurisdiction
  are
• UNSUB
• SUB
• OPEN
• PLACE
• DISCH
• CLOSE

5
Study Details

• Children who first came into contact with the
  child welfare system in 2001 or 2002 in one
  jurisdiction
• Followed each child for two years and captured
  the first four events
• Captured race/ethnicity and age at first contact
• Calculated rates of contact at each level of
  system involvement
• Followed changes in the racial mix of the
  caseload at various junctures including selected
  common trajectories such as
• SUB/OPEN
• SUB/OPEN/PLACE
• SUB/SUB/SUB

6
Basic Data System Involvement by Race and Age
Group
7
Basic Data System Involvement Disparity
8
Caseload Mix at Initial System Contact
9
Distribution of First Contacts by Type and Race
10
Given a Case Opening, What is Likely to Happen
Next?
11
Given a Substantiated Report, What is Likely to
Happen Next?
12
Caseload Mix After Selected Pairs of Events
13
Caseload Mix After Selected Triplets of Events
14
Summary Caseload Mix
15
Babies vs. All Caseload Mix at Initial Contact
16
Babies Distribution of First Events by Type and
Race
17
Babies Given a Substantiated Report, What is
Likely to Happen Next?
18
Summary Caseload Mix for Babies vs. All
19
Limitations

• We have not taken into account the length of time
  between events.
• We have not looked at maltreatment type.
• We only studied one jurisdiction.
• We have not subjected the data to multivariate
  models.
• Case opening is not a very good proxy for whether
  or not services were delivered.

20
Conclusions

• The source of disparity in this jurisdiction is
  primarily at the point of first contact.
• African American children are more likely than
  white children to have a case opened at any
  point.
• Disparity is greatest for infants.
• Could the disparity in case openings suggest an
  effective protective process?

21
Implications

• Given that infancy is a unique developmental
  stage, we need to adjust strategies for dealing
  with that population in developmentally
  appropriate ways.
• We must be careful in evaluating our options for
  addressing disproportionality. Policies and
  practices narrowly aimed to reduce disparities
  may sabotage other, good outcomes.