Center for Advanced Studies in Child Welfare School of Social Work, College of Education and Human Development, University of Minnesota

Center for Advanced Studies in Child Welfare

Research & Evaluation

Foster Care Re-Entry Study

Title IV-E Curriculum Module

Module Elements

Module Summary

Summary

This study was part of the Minnesota Department of Human Services Results Initiative conducted in cooperation with Hennepin County.   Because rates of re-entry into foster care in Minnesota exceeded the national rate, the state and county sought to better understand the discrepancy by investigating reasons for re-entry into care.  By identifying and targeting certain variables, it was hoped that Minnesota's high re-entry rate could be explained and/or reduced.  The purpose of the study was to explore factors that may help predict re-entry or stability after exit from foster care, and provide direction for policy or practice changes that would help provide better stability for children exiting care. 

Relationship to policy/practice

Some of the issues noted during the study in relationship to Hennepin County also have a bearing on practice in child welfare generally.  The first was the importance of assessment with respect to improved family conditions including determining compliance with the case plan.  Next, was an issue that has been commonly noted:  the degree to which substance abuse and poverty play a role in family functioning.  Service planning and policy must begin to develop more useful approaches to dealing with these problems.  The interaction of many family factors was also an issue for assessment requiring the most highly qualified personnel in order to determine the ability of the family to care for the child.  There was a high indication of the need for further research on the effectiveness of intervention with this population.

Background

A review of the literature regarding foster care re-entry revealed factors associated with re-entry that could be organized into four categories: child characteristics, parental characteristics, family descriptors, and prior CPS history and placement characteristics. 

Child Characteristics

Child characteristics associated with placement re-entry included age, health problems, mental health/behavior problems, and African American racial identity.  The review revealed differing opinions about child age.  Several studies found that infants were more likely to re-enter foster care (Barth, 1997; Berrick et al., 1997; Frame et al., 2000).  This may be explained by the pressure to reunify younger children with their biological family (Barth, 1997).  That is, pressure to reunify younger children with biological families may have caused premature reunifications, with the end result being re-entry into the foster care system.  An exception was one study conducted in Wisconsin which found the opposite to be true.  In this study, older children had a higher rate of re-entry into foster care compared to younger children (Courtney, 1996; Courtney et al., 1997).  An additional study found that younger children re-enter the foster care system at a slower rate (Wells et al., 1999).

Research regarding foster care re-entry found that children with health problems were at higher risk for re-entry (Courtney, 1995; Courtney et al., 1997; Davis et al., 1993).  Likewise, children with mental health/behavior problems were at a greater risk of re-entry (Fraser, 1996; Jones, 1998, Lewis 1995, Thomlinson, undated).  Finally, re-entry research revealed that African American children were more likely to re-enter the foster care system compared to other racial groups (Barth, 1997; Berrick et al., 1997; Courtney, 1995; Courtney et al., 1997; Jones, 1998; Jones, 1999; Maluccio et al., 1994; Terling, 1999; Wells et al., 1999). 

Parental Characteristics

The literature review of factors associated with re-entry into foster care also revealed parental characteristics tied to re-entry including substance abuse problems, criminal history, severity of problems, parental competency/care-giving skills, and non-resolution of problems prior to reunification.  For example, children born to parents with substance abuse problems were more susceptible to re-entry (Berrick et al., 1997; Frame et al., 2000; Hess et al., 1992; Jones, 1999; Rzepnicki et al., 1997; Terling, 1999; Wilson, 2000).  Additionally, children of parents who had a criminal history were more likely to re-enter the foster care system (Frame et al., 2000; Jones, 1999; Rzepinicki, 1997; Terling, 1999). 

Parental competency and care-giving skills were also associated with re-entry.  More specifically, children born to parents with low parental competency and minimal care-giving skills were at a higher risk of re-entering the system (Festinger, 1996; Jones, 1999; Maluccio et al., 1994; Maluccio, 1999; Maluccio, 2000; Newman, 1995; Terling, 1999; Thomlinson, undated).  Lastly, non-resolution of parental problems prior to reunification increased the likelihood that a child would return to foster care (Hess et al., 1992). 

Family Descriptors

Family descriptors associated with placement re-entry included social isolation and/or no support system, unmet service needs, housing instability, and poverty.  Not surprisingly, children reunified with caregivers who were socially isolated and/or had no support system were at a higher risk of re-entry (Berrick et al., 1997; Festinger, 1996; Jones, 1999; Maluccio et al., 1994; Maluccio, 1999; Maluccio, 2000; Thomlinson, undated).  In addition, children of parents who were described as having unmet service needs were more likely to re-enter the foster care system (Festinger, 1996; Lewis, 1995).

In numerous studies housing instability was another predictor of recidivism (Berrick et al., 1997; Festinger, 1996; Frame et al., 2000; Jones, 1998; Lewis, 1995; Rzepnicki et al., 1997).  A final family descriptor heavily associated with placement re-entry was poverty.  As would be expected, poverty led to greater likelihood of re-entry (Barth, 1997; Courtney, 1995; Courtney et al., 1997; Hess et al., 1992; Jones; 1998).  In particular, children who were AFDC eligible were more susceptible to re-entering foster care (Barth, 1997; Courtney, 1995; Courtney et al., 1997).  One noteworthy study found that the re-entry rate for African American children from AFDC eligible families was twice as high as for Caucasian children from families who were not AFDC eligible (Courtney, 1995). 

Prior CPS History and Placement Characteristics

The final category of factors discussed in the foster care re-entry literature is prior CPS history and placement characteristics.  These factors included total number of CPS reports on the family, total number of prior placements, exiting from a prior kinship care, placement stability, and last placement in group home.  The higher the number of CPS reports on the family the greater the likelihood of recidivism (Berrick et al., 1997; Frame et al., 2000; Fraser, 1996; Jones, 1999; Terling; 1999).  The total number of prior placements also increased the likelihood of re-entry (Wulzcyn et al., 2000).    Exiting from a prior kinship care placement was not a risk factor for re-entry and often served as a protective factor that prevented re-entry (Barth, 1997; Courtney et al., 1997; Davis et al., 1993; Frame et al., 2000; Wells et al., 1999; Wulzcyn et al., 2000).  One study found that children placed with kin prior to reunification were 80% less likely to re-enter foster care (Frame et al., 2000).  Placement stability also served as a protective factor that prevented re-entry (Courtney, 1995; Courtney et al., 1997).  Finally, children whose last placement was in a group home were more likely to re-enter care (Wells et al., 1999; Wulzcyn et al., 2000).

In sum, prior research outlines four broad categories of characteristics associated with foster care re-entry.  They include child characteristics, parental characteristics, family descriptors, and prior CPS history and placement characteristics.  Although some re-entry patterns emerged across states, it is important to note that there were differences in re-entry characteristics between both studies and states.

Back to top

Project description, results summary

The project, a case record reading study, was the result of a tremendous amount of effort on the part of Hennepin County, the SSIS personnel from Minnesota’s DHS, the Gamble-Skogmo Chair at the University of Minnesota School of Social Work, and the staff who collected and analyzed the data.  It yielded a large amount of information that was combed through over many months and reviewed numerous times by state and county personnel on the Results Initiative Panel to ensure the quality of the findings.   The cases were drawn from a random sample and so should be, at least to some degree, representative of the population of cases from which they were drawn.

Given the care with which this work was conducted, the findings became even more significant in informing practice.  It was the goal of the project team that they be used in further analytic and planning discussions and that they provide a foundation on which to make a number of practice and policy recommendations.

In the present study, case characteristics that increased the odds of re-entry into the foster care system were indicated.  Of the variables known to the worker at the time of initial placement, parent’s incarceration at the time of the initial placement was the strongest predictor of not re-entering foster care.  For placement/service characteristics known at the time of, and during, placement I ending in 2000, the following variables were the best predictors of re-entry in to the foster care system: (1) having had a placement of less than 72 hours, (2) two or fewer living situations in Placement 1, (3) Title IV-E eligible v. eligibility unknown or not eligible, and (4) duration of placement 1 equaled more than 10 days.  For all case characteristics combined the strongest predictors of re-entry were (1) reason for placement 1 is not neglect, (2) parent(s) not incarcerated at the time of the initial placement, and (3) child’s placement is Title IV-E reimbursable. 

The findings of this study must also be considered in the larger context of re-entry rates in other studies and with respect to the federal standards for foster care re-entry within 12 months of exit.  There may be many reasons for seeing higher rates, such as serving different types of cases in the child welfare system.  For example, New York City is likely to have very different caseload characteristics compared to a much more sparsely populated area. 

Examination of the cases in this study suggested that there are major categories of cases that go into foster care.  The first consists of those cases that are less serious and that may occur almost accidentally.  An example might be those cases in which the parent is incarcerated and the child has no immediate place to stay.  The other consists of cases that are quite serious and often contain ongoing chaotic situations that are not resolved by the time the placement ends.  For example, the parent may show good faith in his or her efforts to follow the plan but ultimately fails after the child is returned home.  Children in these cases may go home quickly, reflecting the current preferences in the field for the shortest possible placements as well as the system of case processing.  In the latter, cases were traditionally closed when the child exited care, rather than staying open for “trial home placements” which would not count as re-entries if the child returned to care. These are all hypothetical observations drawn from the quantitative analysis, the qualitative analysis, other studies and conversations with policy makers in other states.

Back to top

Conclusion and recommendations

Conclusion

As previously mentioned, the findings of the project should be considered within the larger context of re-entry rates in other studies and with respect to federal standards for foster care re-entry within one year of leaving placement.  The discrepancy in rates across states may have a number of causes.  Caseloads can vary immensely from one state to another.  For example, Minnesota is likely to have different caseload characteristics than a state like California. 

Examination of the cases in this study suggests that there were two categories of cases that enter the foster care system.  The first category is those cases that are not as serious and may occur accidently (a child enters placement because his/her mother is incarcerated and the child needs an immediate place to stay).  The second category of cases is significantly graver in nature and is characterized by continuous chaotic situations that remain unresolved once the placement ends.   An example of this case category would be when a parent shows good faith in efforts to follow the case plan, but is unsuccessful following reunification.  Premature reunification may be the result of current child welfare preferences for the shortest possible placement. 

Case processing may also have had an effect on this study’s findings.  If a child is returned to care following a “trial home placement” the case is not counted as a re-entry.  Thus if when the child exits care the case stayed open for a “trial home placement”, the higher number of re-entries might have been reduced.  Again, these were hypothetical observations drawn from quantitative analysis, qualitative analysis, other studies and conversations with policy makers in states other than Minnesota. 

Selected Recommendations

Several recommendations could be made based on previous findings and the project’s results.  First, a method for accurately evaluating parental readiness to resume child care-giving responsibilities should be instituted.  Suggested areas of improvement included: practice guidelines, worker training, and supervisor guidelines.  It was recommended that in addition to a safety assessment other assessments should be conducted such as ensuring stable housing, a proper social support system, parents are engaged in ongoing services, a basic level of family financial security, and parents are armored with competent caregiving skills. 

Second, post-reunification services must be available for reunified families.  Monitoring was not enough to prevent re-entry.  This was an issue of both service use and engagement.

It was also recommended that practitioners and policy makers acknowledge that the degree of disorganization in some families predicts a failure on the parents’ part to protect their children.  Careful consideration must be given to what families need to do in order to show that they are able and willing to keep their children safe.

Finally, it would be helpful to determine the degree to which children are placed when there are other alternatives.  For example, when a child is at risk for imminent harm, first responders should be encouraged to seek and find possible family members with whom the child can stay with before the child is taken to a shelter placement.

Back to top

Sample description

Below is a description of the random sample of children selected.  The items of most interest are presented here.  These are purely descriptive characteristics and should not be identified as the final risk factors. 

  • Child and Family Factors: Demographics
    • Age
      • The children were, on average, 5 years old at the time of their initial placement [1], with 23% of the children being 1 year old or younger.
      • When they entered Placement 1 ending in the year 2000, the average age was 6.6 years, with 25% under 2.2 years of age.
      • Of non-reentries, 23% had been infants at the time of their initial placement, compared to 32% of the reentries. 
    • Race/ethnicity [2]
      • Thirty-two percent of the non-reentries as compared with 13% of reentries were American Indian. 
      • Caucasians comprised 41% of the non-reentries and 29% of the reentries. 
      • African Americans comprised 63.6% of the non-reentries and 73.7% of the reentries.
    • Gender
      • Half the non-reentries and 55% of the reentries were male.
      • Child and Family Factors: Maltreatment
      • All of the cases read were either child protection assessment or child protection management cases.
      • For all but two cases, the initial placement was for child protection reasons.
    • Placement Characteristics
      • Children were in Placement 1 (the placement ending in the year 2000) an average of 4.5 months, with a wide range of time in care from 2 days to 24.6 months.   Half the children stayed 1.2 months or less. 
      • Within Placement 1, children experienced 1 to 4 different placement settings, with half experiencing at least two settings, including St. Joseph's shelter. 
      • The children experienced 1 to 10 continuous placements. 
  1. In this study, the first placement in the case record or the SSIS system was referred to as the child’s initial placement.  Because SSIS was initiated in the year 2000, it is not possible to be absolutely certain that there were no other placements, e.g., in another county, before the initial placement.
  2. The percentages of different races in each group may add to more than 100% because more than one race may be chosen for each child or family member.