Racial Disparities in Child Protective Services

1
Racial Disparities in Child Protective Services
Based on January 19, 2008 Presentation
Introducing the Qualitative Story
Qualitative study presented by Margaret
(Griesgraber) Skrypek, Susan J. Wells and Maxie
Rockymore University of Minnesota School of
Social Work Contact swells_at_umn.edu
A Study of the African American Racial
Disparities Committee, the Minnesota Department
of Human Services and four Minnesota counties

• Principal Investigator Susan J. Wells Research
  Assistant Margaret (Griesgraber) Skrypek
• Minnesota DHS Coordinator Maxie Rockymore
• Data Analysis Contributors Ila Kamath, Scotty
  Daniels, Alex Beutel,
• Mary Pfohl, Louis B. Carter, Dinorah
  Martinez-Osorio, and Nancey Riley

2

• Study Background
• and
• Quantitative Findings

3
Analyses Completed

• Quantitative - assigning numerical values to
  variables in order to understand behavior by
  statistical modeling and measurement
• Qualitative - subjective judgment based on
  non-quantifiable information, such as text, notes
  and comments

Quantitative analysis is a powerful tool for
evaluation, but the story is more complete when
it is combined with qualitative analysis.
4
Selection of Cases

• All neglect cases in 4 counties
• Selection requirements
• a single type of maltreatment.
• children not previously placed
• race all African American or all white
• no children identified as Hispanic ethnicity
• ages 0 through 11
• parents born in the USA

5
Total Cases Selected

• Total of 1,095
• Substantiated for neglect
• From four Minnesota counties in 2001
• African American (58) and Caucasian (48)
  children

6
Matched Case Study

• 103 pairs of cases were matched on
• Age group (0-5 and 6-11)
• Reason for referral - type of neglect
• Gender
• County

7
Evolution of Number of Cases or Pairs

• 1,095 original cases
• After case matching, 103 original pairs 206
  cases
• Some cases were eliminated later
• For example
• Child was of more than one race
• Family was immigrant or other culture
• 81 pairs remained 162 cases
• (If pairs of eliminated cases were included -
  180 cases)

8
Descriptive Data for Study
Counties

• Anoka 18 children
• Hennepin 76 children
• Olmsted 20 children
• Ramsey 48 children
• N 162

9
Descriptive Data for Study (cont.)
Type of Neglect

• Abandonment 4
• Educational neglect 28
• Endangerment 52
• Inadequate Supervision 14
• Neglect (food, clothing, shelter) 58
• Prenatal exposure 6
• N162

10
Three Parts of Record Review Questionnaire (for
Paired Cases)

• Assessment 162 cases
• 81 pairs
• Case Management 55 cases
• 13 original pairs
• Reunification - 39 cases
• 7 original pairs

11
Todays Focus

• Very quick review of 162 81 pairs
• Overview of quantitative analysis
• of 180 cases
• Qualitative analysis of 180 cases

12
Qualitative Variables Used for Quantitative Study

• History of maltreatment
• Extensive multi-generational
• Interaction of worker and family
• Worker negative parent uncooperative
• Poverty yes or no
• Substance abuse extensive
• Relative was a resource
• Primary problem contributing to maltreatment
• Family moved, case outcome unclear
• Police arranged informal placement

13
Family Composition at Assessment
Percentage of Pairs with Biological Father in
Household
McNemar level of significance p
.004 significant finding
14
Other Significant Differences at Assessment
N 81 pairs
15
Police Involvement at Assessment
Percentage of Pairs in Which Police Arranged an
Informal Placement
McNemar level of significance p
.039 significant finding
16
Cases Opened for Case Management Services
Percentage of Pairs in which the Case was Opened
for Case Management Services
McNemar level of significance p 1.0 Not
significant
17
Cases Opened for Reunification Services
Percentage of Pairs in which the Case was Opened
for Reunification Services
McNemar level of significance p .690 Not
significant
18
Similarities among Cases

• There were no significant differences in parental
  drug abuse, inadequate housing, felony history,
  contact with the law, fathers disabilities,
  termination of parental rights for older
  siblings, or death of one or both parents.

19
Multivariate Analysis

• That is, was the case referred for Reunification
  Services at any time during the intervention
  process?
• (Does not include cases that were in placement
  initially but went home right away.)

20
New Variables

• Several scales were created to reduce the number
  of variables
• Three variables of particular interest follow

21
Maltreatment Summative Scale (0-3)

• None of the following
• Maltreatment history found at investigation
• Maltreatment extensive (qualitative var.)
• Maltreatment multi-generational (qual.var.)

22
Mom Drug and Law Probs Scale

• No drug problems or problems with the law (0)
• Problems with drugs or problems with the law (1)
• Problems with drugs and problems with the law (2)

23
Parent-Worker Relationship (0-1)

• Worker negative
• And/Or
• Parent uncooperative
• This is not a scale there is only a yes/no answer

24
Multivariate Analysis Results
25

• Variables in the Equation
• B S.E. Sig. Exp(B) 95.0 C.I.for
  EXP(B) Lower Upper
• Q01children -.849 .237 .000 .428 .269 .680
• WrkrJudgPrnt
• UncoopScale 2.011 .424 .000 7.471 3.252 17.167
• Ages1_4 1.507 .684 .028 4.512 1.180 17.243
• AfricanAmerican
• by Ages1_4 -2.327 1.043 .026 .098 .013 .753
• MaltreatHistScale .214 .342 .531 1.239 .634 2.421
  
• AfricanAmerican
• by MaltreatHistScale 1.067 .427 .013 2.906 1.258 6
  .714
• MomDrugLawProbs 1.062 .321 .001 2.892 1.541 5.427
  
• SubsAbuse_Extensive 1.979 .616 .001 7.237 2.164 24
  .208
• Constant -2.418 .684 .000 .089

26
Other Important Variables

• Bio Dad in the household
• Mom cognitive, physical or MH disabilities
• County
• Source of report was health
• Relative is a resource

27
Percent Children Placed when Parent
Uncooperative by Race
African American Uncooperative or Worker
Negative-gt Yes29, No58 Caucasian
Uncooperative or Worker Negative-gt
Yes29, No65
28
Percent Children PlacedMaltreatment History by
Race
Number of Children Maltreatment History Scale
by Race 0 1 2 3 Total African
29 33 21 4 87 American Caucasian 36 31 24 2 93
29
Percent Children PlacedMaltreatment History by
Race
Number of Children Maltreatment History Scale
by Race 0 1 2 or 3 Total
African 29 33 25 87 American Caucasia
n 36 31 26 93
Maltreatment History Scale Indicates the number
of instances of the following history of
maltreatment, maltreatment extensive, and/or
multigenerational maltreatment.
30
Percent Referred for Placement (in Which Referral
for Reunification Services was Made) by Age
Children Identified as Both African American
Caucasian are not included on this graph.
Number of Children - Age by Race
-1 0 1 2 3 4 5 6
7 8 9 10 11 Total
African 1 13 4 6
4 2 11 10 6 7 9 7
7 87 American Caucasian 3 14
5 7 5 5 4 7 7 12
5 11 8 93 Total 4 27
9 13 9 7 15 17 13 19 14
18 15 180

31

• Qualitative Study

32
So, What Can We Know From the Qualitative Study?

• Examples of questions we had
• What was the problem financial difficulties
  about?
• How did age differences work?

33
Introduction to the Qualitative Study

• Use of the same data sources as the quantitative
  study
• Opportunity to clarify findings from the
  quantitative study
• Deeper look at whether and how race may affect
  case decisions and outcomes

34
Qualitative Methods Step 1

• Researchers re-read all 206 original case record
  surveys
• When data was confusing or unclear, researchers
  re-read the case record to fill in missing
  information
• Text from these readings and selected
  quantitative variables were used to create a new
  qualitative database

35
Qualitative Method Step 2

• Data in the database was checked and rechecked
  against the surveys and the case records for
  accuracy
• Data was imported into NVIVO software a
  specialized system for qualitative analysis

36
Qualitative Methods Step 3

• To gain insight from community members with child
  welfare experience, a sample study was completed
  with 36 selected cases
• 3 community members and 2 researchers
  participated in the sample study
• The cases were selected randomly but were
  representative of the larger group of 206 cases
• Using lessons learned from the sample reading, a
  full qualitative analysis was completed on the
  text of 162 cases

37
Age Discovered to be a Defining Variable

• Childs age was important in understanding
  differences among placement decisions and race.
• The following age groups were established
• Less than one year old
• One to four years old
• Five to nine years old
• Ten or eleven years old

38
Highlights

• Themes that emerged across age groups
• Parental competence
• Disinterest
• Inability (might include mental health)
• Crisis
• Resources (educational neglect)
• Domestic Violence
• Drugs/Alcohol
• Use
• Other involvement

Racial patterns within these case categories vary
by age!
39
Highlights (continued)

• Themes that occurred across age groups with
  consistent racial patterns
• Police involvement
• Arrests
• Formal vs. informal placements
• County may also be important

40
Children Less than One Year Old (N30)

• Positive toxicology and pre-natal cases
• Similar outcomes regardless of race
• Omitted from further qualitative analysis

41
Children One to Four Years Old (N30)

• Parental Competence Case Types
• Parental disinterest
• Inability to parent
• Child emotional or physical problems
• Crisis Case Types
• Short-term emergencies
• Domestic violence
• Extensive Substance Abuse Case Types
• Multiple relapses during case management
• Methamphetamine use
• Indicates possible racial pattern

26 of the 30 cases in this age group (86.7) were
in three categories.
42
Patterns in 1 4 Age Group (N30)

• Caucasian families more likely to experience long
  term placement and to undergo TPR/TLC.
• Police were involved in 12 cases
• Racial pattern observed in whether police made
  formal placements or allowed informal placements

43
Children Five to Nine Years Old (N74)
Crisis Parental Competence Case
Types Domestic violence Poor
parenting Alcohol use Educational Neglect Case
Types Resource deficit Mental health of
caregiver Parent Out of Home Case Types Police
involvement Abandonment Parent/Child Mental
Health Case Type Indicates possible racial
pattern
67 of 74 cases in this age group (91) were in
four categories
44
Patterns in 5 9 Age Group (N74)

• African American children more likely to enter
  long term placement for this age range.
• All cases (6) that resulted in a child being
  placed permanently outside the home were TLCs to
  relatives.
• Police were involved in many of these cases, 10
  of which included out of home placements.
• All 10 cases were either drug raids, parental
  substance abuse, or criminal activity.
• 6 cases were arrests - all were African American
  parents (one during case management.)

45
Children Ten or Eleven Years Old (N30)
Parent Drugs/Alcohol Case Types Parental
substance abuse Police drug raids Educational
Neglect Case Types Resource deficit Parent or
child mental health Parental Competence Case
Types Disinterest in parenting Inability to
parent Short-term Crisis Case Types Domestic
violence Resource issues Indicates possible
racial pattern
30 of 30 cases in this age group (100) were in
four categories
46
Patterns in 10-11 Age Group (N30)

• African American and Caucasian children equally
  unlikely to enter long term placement
• Police arranged out of home placements in 3 cases
  in this age range.
• One formal placement for a Caucasian child whose
  homeless mother requested 72 hour hold for child.
  
• Two informal relative placements for children
  whose parents were arrested for drug raids. One
  of these children was Caucasian and one was
  African American.

47
Other factors to consider

• Several cases closed early because the family
  moved, which may affect the data related to case
  outcomes.
• What is the impact of parents perceived level of
  cooperation on workers attitudes?
• More Caucasian families using other mechanisms to
  avoid child welfare involvement i.e. family
  court to transfer custody to a non-custodial
  parent.

48
Implications

• Case types differ by race
• So, to some extent, other differences might be
  expected
• Nevertheless, what can we learn from every day
  practice about potential differences?
• How can we keep children out of the system when
  they dont need to be there?

49
Conclusion

• Thank you for your attention
• Questions?