Exploratory analysis of child protection mediation permanency placement outcomes

Children and Youth Services Review 50 (2015) 20–27

Contents lists available at ScienceDirect

Children and Youth Services Review journal homepage: www.elsevier.com/locate/childyouth

Exploratory analysis of child protection mediation permanency placement outcomes Donna M. Aguiniga a,⁎, Elissa E. Madden b, Alicia Hawley b a b

University of Alaska Anchorage, United States University of Texas at Arlington, United States

a r t i c l e

i n f o

Article history: Received 11 August 2014 Received in revised form 22 December 2014 Accepted 23 December 2014 Available online 2 January 2015 Keywords: Child protection mediation Permanency Foster care Child welfare

a b s t r a c t Using a merged data set constructed from empirical data obtained from Child Protective Services (CPS) and a nonprofit agency tasked with conducting an evaluation of a child protection mediation pilot project in the state in which this study was conducted, this study (N = 311) explores how various child and family factors in child protection mediation cases affect placement outcomes for children in care. Results of the multinomial logistic regression (MLR) analyses found that children for whom parent/caregiver substance abuse and/or mental illness was an issue were less likely to be reunified with their parents than remain in care. Those children who experienced a higher number of placements and those who were male were less likely to be placed with relatives. Children whose parents experienced mental health issues were less likely to be placed with a relative. With regard to the permanency outcome of adoption, the findings showed that as the age of the child at time of removal increased, children were less likely to be adopted than remain in care. Additionally, African American children in the sample were less likely to be adopted. © 2015 Elsevier Ltd. All rights reserved.

1. Introduction and background During the last two decades, child protection mediation has increasingly been used in child abuse and neglect cases to resolve disputes and expedite the permanency process for children in foster care. While the presence of permanency cannot guarantee a positive childhood experience, research on children and attachment suggests the need for increased efforts to achieve permanency for children in foster care (Jones-Harden, 2004; Testa, 2004). Drawing on the work of Hardin (1992), Jenson and Fraser (2006) outline several reasons why permanency is important for children. First, it is widely accepted that a child's attachment with his or her caregiver is believed to be the basis for other attachments and relationships. Children who have consistent and responsive caregivers are more likely to develop the foundation that is necessary for them to learn how to form healthy attachments with others. Second, children need caregivers who are committed and willing to tend to their needs. This is more easily navigated when the relationship between the caregiver and the child is permanent and not subject to ongoing supervision and regulation by external agencies. The absence of a committed and loving caregiver can undermine a child's self-worth and their subsequent ability to trust others. Third, children desire and require a certain amount of predictability in their lives. Children who lack a permanent placement are often painfully aware of the precarious and tentative nature of their situation. A permanent placement for a ⁎ Corresponding author. E-mail address: [email protected] (D.M. Aguiniga).

http://dx.doi.org/10.1016/j.childyouth.2014.12.021 0190-7409/© 2015 Elsevier Ltd. All rights reserved.

child creates a stable and predictable environment for them to be able to grow and mature as they encounter new and different situations. On a practical level, independent families are also generally more capable of making timelier and more fully-informed decisions for a child than child welfare agency professionals or the courts. Furthermore, Pecora, Whittaker, Maluccio, and Barth (2000) note that an additional benefit of permanency is that permanent placements result in more consistent decision-making for the child, as opposed to the fragmented, drawn-out decision making process that is often utilized for children in foster care. Because of the importance of permanency, it is critical that continued research and attention be given to determine the effectiveness of service interventions, such as child protection mediation, in helping facilitate positive permanency outcomes for children. Therefore, this study seeks to explore how different child and family factors in child protection mediation cases affect permanency outcomes for children in care. 2. Literature review 2.1. Child protection mediation Mediation is a form of conflict resolution that involves the facilitation of communication between two or more interacting and conflicting parties by a third party who does not possess the authority to impose a particular outcome (Wall, Stark, & Standifer, 2001). Originally seen as a means to assist with managing court dockets, mediation was first used in child abuse and neglect cases in the early 1980s (National Resource

D.M. Aguiniga et al. / Children and Youth Services Review 50 (2015) 20–27

Center for Foster Care and Permanency Planning, 1998; Thoennes, 2009). Using a neutral third party intermediary, child protection mediation brings parties together to explore issues related to the child or children in foster care and propose mutually agreeable solutions to help them achieve a permanent home. Once an agreement is made, it is presented to the court for acceptance, rejection, or modification (Stack, 2003). The purpose of child protection mediation is not to impart blame on the alleged perpetrator of the abuse or neglect (Barsky & Trocmé, 1998), but rather to facilitate problem-solving among the parties and, ultimately, find a safe and stable environment for the child or children (Dobbin, Gatowski, & Litchfield, 2001; Eaton, Whalen, & Anderson, 2007; Edwards, 2009). The process is intended to provide families with an alternative to the protracted adversarial court process that critics argue is deficit-focused and polarizing (Crush, 2005; Vidmar, 1992). In contrast to the conventional adversarial process, mediation appears to offer greater opportunities for families and child welfare agencies to participate in the decision-making process and in determining how the case will be resolved. Furthermore, child protection mediation has been shown to result in faster resolution of child abuse and neglect cases when compared to cases that go to trial (Koh, 2004). Advocates of mediation suggest that increased family engagement, improved communication, comprehensive service plans, and parental compliance contribute to earlier resolutions of cases (Colman & Ruppel, 2007). While not conclusive, some evaluations of the child protection mediation process suggest that mediated cases have shown to progress to permanency more quickly and with less involvement of the court (Anderson & Whalen, 2004). There is also evidence that mediation reduces costs by lessening the time that parties spend preparing for court hearings (Bernstein, 1998; Giovannucci, 2007; Nasworthy & Tarver, 2005). While these evaluations were not necessarily designed to demonstrate the effectiveness of mediation in improving permanency outcomes for children in foster care, they can help elucidate the benefits of mediation and how mediation may help facilitate permanency. 2.2. Permanency Permanency is defined as “a legal, permanent family living arrangement, that is, reunification with the birth family, living with relatives, guardianship, or adoption” (U.S. Department of Health and Human Services [USDHHS], 2005, p. 2). The Adoption and Safe Families Act of 1997 (ASFA), the legislation that currently guides permanency planning efforts, outlines five possible permanency options for children in foster care (in order of preference): 1) family reunification; 2) adoption; 3) legal guardianship; 4) permanent placement with a “fit and willing” relative; and 5) another planned permanent living arrangement. It is widely acknowledged that permanency, and the emotional attachments that can accompany it, are of paramount importance to a child's social and emotional development (Jenson & Fraser, 2006; Jones-Harden, 2004). Research has found that positive experiences with supportive adults may contribute directly to behavior and relationships later in life (Appleyard, Egeland, & Sroufe, 2007). Some developmental studies have concluded that children who do not have at least one strong attachment to a caring adult are at increased risk of experiencing challenges later in life with regulating their own emotions, interacting with others, and coping with stressful situations (Carlson & Sroufe, 1995; Murphy, Bandy, Schmitz, & Moore, 2013; Scales & Leffert, 1999). Furthermore, research has shown that children who lack long-term attachments to appropriate and nurturing caregivers are at greater risk of experiencing difficulty in achieving and maintaining self-sufficiency and independence as young adults (Avery & Freundlich, 2008; Children's Permanency Cooperative, 2000; Masten & Coatsworth, 1998; Plunkett & Osmond, 2004). 2.2.1. Barriers to permanency Agencies and courts work to provide children in the foster care system with safe, permanent homes in the shortest amount of time possible (Olson, 2003). However, overcrowded dockets and inadequate

21

child welfare and court resources often serve as barriers to the establishment of timely permanent placements for children in foster care (U.S. General Accounting Office [USGAO], 2002). Furthermore, the adversarial nature of child protection cases can also result in lengthy delays and appeals that prevent the timely resolution of cases (Olson, 2003; U.S. Department of Health and Human Services, 2006). In recent years, studies have identified a number of procedural and systemic issues that negatively impact permanency outcomes for children in care (e.g., Macomber, Scarcella, Zielewski, & Geen, 2004; U.S. General Accounting Office, 2002; Wilson, Katz, & Geen, 2005). Court issues such as poor communication between courts and child welfare agencies, inadequate numbers of judges and attorneys, insufficient training for judges and attorneys regarding child protection issues, and delays in scheduling termination of parental rights hearings are some of the issues that have been found to serve as obstacles to permanency (U.S. General Accounting Office, 2002). Additional agency-related barriers that impede permanency include caseload size and caseworker turnover, which can negatively impact staff responsiveness and communication, as well as other factors such as the complexity of the foster and adoption process, difficulties with recruiting resource families for “special needs” children,1 and the lack of available supportive services for families involved in the child welfare system (Flower, McDonald, & Sumski, 2005; Murphy, Van Zyl, Collins-Camargo, & Sullivan, 2012; U.S. General Accounting Office, 2002; Weigensberg, 2010). Research has shown that some child characteristics may also influence and, in some cases, serve as barriers to permanency for children in foster care (Kemp & Bodonyi, 2000; Weigensberg, 2010). Child characteristics studied typically relate to demographic characteristics of the child such as age, gender, and race/ethnicity (e.g., Becker, Jordan, & Larsen, 2007; Kemp & Bodonyi, 2000; Weigensberg, 2010). However, there is evidence that the presence of mental and physical disabilities can also negatively impact permanency outcomes for children (Akin, 2011; Connel, Katz, Saunders, & Tebes, 2005). Additional case characteristics identified in the literature that may influence permanency outcomes include the child's prior removal history, reasons for the child's removal, and the child's initial placement setting (Connel et al., 2005; Koh & Testa, 2008; McDonald, Poertner, & Jennings, 2007; Weigensberg, 2010).

2.3. Research on child protection mediation and permanency Those who support the use of child protection mediation generally believe that the mediation process has a positive influence on permanency outcomes. However, few studies have tested this assertion empirically. To date, much of the research on child protection mediation has focused on short-term measures that may contribute to permanency, such as settlement rates, parental compliance, and cost savings (Supreme Court of Virginia, 2002). Only a limited number of studies have examined the impact of child protection mediation on permanency outcomes (e.g., Anderson & Whalen, 2004; Gatowski, Dobbin, Litchfield, & Oetjen, 2005; Madden & Aguiniga, 2013; Thoennes, 2001). Examination of the available literature on child protection mediation reveals a number of methodological inadequacies that limit the overall generalizability of past studies' findings and what conclusions can be generated. First, most evaluations of child protection mediation have been limited by small sample sizes, incomplete data, and the lack of a comparison group. Second, time constraints appear to have prevented most evaluators from following cases over a lengthy period of time to examine long-term outcomes. Third, few evaluations have been 1 Within the context of the foster care system, the term “special needs” refers to children who possess characteristics that may present challenges to locating adoptive families for them, including older age, membership in a sibling group, the presence of developmental disabilities, emotional or behavioral problems, and/or membership in a ethnic/racial minority group (McRoy, Lynch, Chanmugam, Madden, & Ayers-Lopez, 2009; Rosenthal, 1993).

22

D.M. Aguiniga et al. / Children and Youth Services Review 50 (2015) 20–27

conducted on stable, well-established mediation programs that have a high enough volume of cases needed to support more rigorous research designs (Thoennes, 2009). While the limitations above apply to both process evaluations and outcome studies on child protection mediation, an additional criticism of studies that have specifically attempted to examine permanency outcomes is the lack of consensus across studies regarding how permanency should be defined, and therefore, operationalized (Anderson & Whalen, 2004). Madden and Aguiniga (2013) compared 315 mediated with 315 matched (propensity score analysis), non-mediated child welfare cases. Children were classified as having achieved permanency (i.e., yes/no regarding whether they were reunified, adopted, placed with family, or achieved another form of permanent placement) or failed to achieve permanency. Results indicated that while there was not a statistically significant difference in the rate of permanency between the two groups, there were several factors that were associated with the likelihood of a child achieving a permanent placement. Older children, Hispanic children, parent/caregiver substance abuse, parent/caregiver mental health concerns, parents who were incarcerated, and inadequate housing were all characteristics that contributed to decreased odds of children achieving permanency. 2.4. Current study This study attempts to address gaps in prior research by exploring how child and family factors in child protection mediation cases2 affect permanency outcomes for children in care. Given the highly adversarial nature of the child welfare system, mediation provides families with a much needed forum to discuss concerns and participate in the decisionmaking process. Because states tend to prefer and prioritize legal, family-based permanency options (i.e., family reunification, formal guardianship by a relative, and adoption) (Freundlich, Avery, Munson, & Gerstenzang, 2006; U.S. Department of Health and Human Services, 2001), the findings of this study can provide child welfare agency and court personnel with insight on the types of cases that are most likely to result in more desirable permanency outcomes as well as which children are most likely to benefit from child protection mediation. Using a merged data set constructed from empirical data obtained from Child Protective Services (CPS) and a nonprofit agency tasked with conducting an evaluation of a child protection mediation pilot project in the state in which this study was conducted, this study explores, among mediated child protection cases, what child and family factors may lead to notable differences between cases regarding permanency outcomes. 3. Methods 3.1. Data sources This study utilized secondary mediation participant survey data obtained from a nonprofit agency tasked with conducting an evaluation of a child protection mediation pilot project in the state in which this study was conducted in between September 1999 and September 2005.3 The pilot project was designed to promote the use of mediation as a means of resolving child protection cases in a more expeditious, less adversarial, and less costly manner than traditional litigation. An evaluation of the pilot project was conducted by a nonprofit agency between the fall of 1999 and spring of 2005. Child protection mediators involved in the project were asked to complete 40 h of basic mediation training, as well as an additional 2 It should be noted that the terms “case” and “child” are used somewhat interchangeably; however, for the purposes of this analysis, each “case” refers to a single, unduplicated child. 3 Although the pilot project evaluation extended from 1999 to 2005, administrative data for children removed prior to FY 2002 was inconsistent with CPS data compiled after FY 2002. To maintain consistency between data sets, only data for children removed between FY 2002 and FY 2005 was included in the study.

24 h of mediation training specific to working with families, and 24 h of training in Child Protective Services Mediation. While figures specific to this analysis are not available, findings from the original mediation pilot project evaluation reports (i.e., Nasworthy & Tarver, 2002; Nasworthy & Tarver, 2005) suggest that the vast majority of mediators who participated in the pilot program from which data for this study was utilized, reported that they were female and attorneys. Information from the pilot project evaluation also suggests that most mediation sessions involved more than one issue (i.e., type of abuse or safety concern) and were completed in less than 6 h. While mediation can be used at any stage of a child protection case, following 2002, mediation was predominately used in the later stages of child protection cases, typically after 90 days. Secondary data from the evaluation pilot project was utilized to complete this study. Additional information and context about the mediation pilot project, including mediator and participant postmediation survey responses, can be found at Nasworthy and Tarver (2002) and Nasworthy and Tarver (2005). In addition to the child protection mediation pilot project evaluation data, this study utilized case-level administrative data from Child Protective Services (CPS) for all children removed between 2002 and 2005 in the 43 counties that participated in the Child Protective Services Mediation pilot project. Data for the children spanned from January 2002 (the earliest point of entry) to March 2009 (when the data were obtained). This seven year time frame provided the opportunity to examine the permanency outcomes of children whose families participated in child protection mediation (between 2002 and 2005) who exited foster care during that period, as well as the status of children who had not yet achieved permanency as of March 2009. Data from the two data sets were merged to create the data evaluated in this study. After randomly selecting one child from each sibling group contained in the data set,4 the combined data set was found to include a total of 317 children whose families participated in child protection mediation. 3.2. Study sample Of the 317 cases in which mediation was utilized, three were excluded due to missing data. Additionally, three other cases were excluded because they were determined to be extreme outliers and not representative of the general population and thus, their inclusion would potentially bias the study results. The final sample utilized in this study includes 311 children whose families participated in mediation. 3.3. Dependent variable The dependent variable in this study was final permanency outcome. This variable was used to indicate the placement status for children at the time that they achieved permanency or at the end of an 18 month evaluation period. Eighteen months was used as the time frame to determine if a child had achieved a permanent placement as this length of time parallels the guidelines for the state in which this study was conducted that children must have a permanent placement within 12 or, with extenuating circumstances, 18 months. Based on their permanency status, children were classified in one of the four categories: 1) family reunification, 2) adoption; 3) relative placement; or 4) no permanent placement. 3.4. Independent variables The 18 independent variables examined in this study included child, parent/caregiver, and case variables discussed in the literature that have been shown to potentially influence or delay the permanency process 4 One random sibling was selected from each sibling group in the merged CPS/ mediation pilot data set in order to avoid the potentially confounding effect that may have resulted from the inclusion of siblings in the analyses. In addition, children without siblings (“only children”) are also represented in the final study sample of 315 children.

D.M. Aguiniga et al. / Children and Youth Services Review 50 (2015) 20–27

23

variables examined, parent/caregiver death (no/yes) is used to indicate whether one or more of the child's parents or caregivers were deceased at the time of the child's removal.

for children in foster care (e.g., Connel et al., 2005; Kemp & Bodonyi, 2000; McDonald et al., 2007; Weigensberg, 2010). It should be noted that other potentially influential variables were considered for this study (e.g., child's disability status, child's mental health); however, it was determined that the data for those variables were insufficient to include in the analyses. Child's age at removal was coded as a metric variable that was computed from the child's birth date to the date of the child's removal. Child's gender (dichotomous; male/female) was determined by what was recorded in the CPS administrative data. Child's race/ethnicity was recoded into three separate dichotomous variables with White/ Anglo used as a reference category: African American (no/yes); Hispanic (no/yes); and Other (no/yes). Number of siblings is a metric-level variable that represents the number of siblings removed from the home and placed in the foster care system. Placement stability is a metriclevel variable that represents the overall stability of the child's residence through a count of the number of placements they lived in during the 18 month study timeframe. Parent/caregiver substance abuse is a composite variable (dichotomous; no/yes) derived from mediation pilot project survey data and CPS administrative data that indicates whether substance abuse was identified as an issue in the case by one or more parents/caregivers at the time of the child's removal or at some point during the CPS case. Parent/caregiver mental health concerns is a composite variable (dichotomous; no/yes) derived from mediation pilot project survey data and CPS administrative data that refers to the presence of a diagnosed mental illness. Types of maltreatment the child experienced were recoded into five dichotomous (no/yes) variables: physical abuse, sexual abuse, abandonment, neglect, and other. Geographic setting (dichotomous, urban/rural) was determined by the location of the court which had jurisdiction over the case. Inadequate housing is a dichotomous variable (no/ yes) used to indicate whether or not the child's family provided the child with safe and adequate shelter necessary to sustain the life or health of the child. Parent/caregiver incarceration (no/yes) indicates whether one or more of the child's parents or caregivers were incarcerated at the time of the child's removal from the home. Similar to other parent/caregiver

3.5. Data analysis SPSS version 22.0 was used to complete a univariate analysis of sample characteristics and perform the multinomial logistic regression (MLR) analyses. MLR was selected because of its ability to predict group membership with an unordered categorical dependent variable having more than two categories (Moutinho & Hutcheson, 2011). Discriminant analysis, another method frequently used when predicting group membership, was considered and deemed inappropriate to use because the data could not be successfully transformed to meet analysis requirements. MLR was used to predict the probability of a case achieving a specific permanency outcome type. Three MLR models were performed with the outcome “No permanent placement” (i.e., “remaining in care”) used as the dependent variable's reference category. 4. Results 4.1. Sample characteristics Of the 311 mediated cases included in the analysis, 33.1% (n = 103) failed to achieve permanency within 18 months. The largest group of children, 107 (34.4%), was reunified with their parents/caregivers, and the smallest group (n = 23; 7.4%) was adopted. Seventy-eight children (25.1%) were permanently placed with a relative. As described in Table 1, children in the sample were an average of 3.96 years old at the time of removal (SD = 4.17). Children had, on average, at least one sibling (M = 1.52, SD = .82) and, during the study's timeframe, resided in three placements (M = 3.07, SD = 1.49). Slightly more than half of the children (55.3%) were female and children of color (African American, 30.9%; Hispanic, 23.3%; or “Other” 2.6%). Parent/ caregiver substance use was indicated in 53.4% of the cases while

Table 1 Characteristics of mediated cases. Sample

Child's gender Female Male Child's race/ethnicity African American Hispanic White Other Parent substance abuse Parent mental health concerns Type of maltreatment Physical abuse Sexual abuse Abandonment Neglect Other Geographic setting Urban Rural Inadequate housing Parent incarceration Parent death

Permanency not achieved

Family reunification

Relative placement

Adoption

N = 311

%

n = 103

%

n = 107

%

n = 78

%

n = 23

%

172 139

55.3 44.7

52 51

50.5 49.5

66 41

61.7 38.3

40 38

51.3 48.7

14 9

60.9 39.1

96 72 135 8 166 127

30.9 23.2 43.3 2.6 53.4 40.8

32 27 39 5 61 56

31.1 26.2 37.9 4.9 59.2 54.4

41 21 43 2 61 39

38.3 19.6 40.2 1.9 57.0 36.4

21 19 37 1 30 21

26.9 24.4 47.4 1.3 38.5 26.9

2 5 16 0 14 11

8.7 21.7 69.6 0.0 60.9 47.8

151 30 22 201 13

48.6 9.6 7.1 64.6 4.2

49 12 9 72 6

47.6 11.7 8.7 69.9 5.8

52 9 8 67 6

48.6 8.4 7.5 62.6 5.6

37 2 4 46 1

47.4 2.6 5.1 59.0 1.3

13 1 1 16 0

56.5 4.3 4.3 69.6 0.0

273 38 53 20 3

87.8 12.2 17.0 6.4 1.0

89 14 22 5 1

86.4 13.6 21.4 4.9 b1.0

95 12 14 10 2

88.8 11.2 13.1 9.3 1.9

69 9 11 5 0

88.5 11.5 14.1 6.4 0.0

20 3 6 0 0

87.0 13.4 26.1 0.0 0.0

Permanency not achieved

Family reunification

Relative placement

Adoption

SD

M

M

M

M

Sample M Child age (at removal) Placement stability (# of placements) Number of siblings

3.96 3.07 1.52

4.17 1.49 .82

4.32 3.31 1.6

SD 4.29 1.70 .94

3.58 2.65 1.50

SD 3.85 1.25 .77

4.9 3.31 1.54

SD 4.53 1.41 .80

.97 3.13 1.17

SD 1.72 1.49 .081

24

D.M. Aguiniga et al. / Children and Youth Services Review 50 (2015) 20–27

parent/caregiver mental health concerns were reported in 40.8% of the cases. Most of the children's cases (87.8%) originated in urban counties. While multiple reasons were often indicated for a given child's removal, neglect (64.6%) and physical abuse (48.6%) were the most common reasons cited. Inadequate housing was a factor in 17% of the cases. With regard to parent or caregiver incarceration, 6.4% of the children had one or more parents or caregivers who were incarcerated at the time of the child's removal from the home. Additionally, parent or caregiver death was indicated in 1% of the cases. 4.2. Multiple logistic regression Three iterations of the MLR were performed; after the first iteration, four variables were removed due to an insufficient number of cases: parent/caregiver incarceration, parent/caregiver death, type of maltreatment: other, and child's race/ethnicity: other. A second model, while statistically significant, had several extraneous (i.e., non-significant) variables. A final, more parsimonious model evaluated the significant variables from the second model: parent/caregiver substance abuse, parent/caregiver mental health concerns, age at removal, placement stability, and African American race/ethnicity, plus one approaching significance (gender, p = .064). While there was some hesitation in including gender as an independent variable given that it only approached significance in the second model, its inclusion in the final model was based on the review of the literature and the exploratory state of this research. The final model, with the more parsimonious set of variables, was statistically significant (χ2 = 85.767, df = 18, p b .001), indicating that the model differentiated between the groups. MLR models are determined to be useful when they represent at least 25% improvement over by chance classification accuracy (Bayaga, 2010). The by chance classification accuracy without a model was 29.65%. The MLR model needed to obtain a minimum classification accuracy of 37.06% to be retained; the final model obtained 48.6% classification accuracy and, thus, is a useful model for further analysis. The results of the final multinomial logistic regression (MLR) model are presented in Table 2. The model compares permanency outcomes of children whose families participated in mediation, using the outcome of “No permanent placement” (i.e., “remaining in care”), as the reference category. 4.3. Family reunification With regard to the permanency outcome of family reunification, the findings indicate that children whose parents/caregivers experienced Table 2 Final MLR model predicting permanency outcomes. Variable Family reunification Age at removal Number of placements African American Parent substance abuse Parent mental health concerns Male Relative placement Age at removal Number of placements African American Parent substance abuse Mental health concerns Male Adoption Age at removal Number of placements African American Parent substance abuse Parent mental health concerns Male ***p b .001; **p b .01; *p b .05.

Exp(B)

p

1.007 .986 .687 .326 .254 1.087

.854 .892 .298 .001*** .000*** .795

.958 .677 1.750 .856 .432 .540

.239 .001*** .085 .616 .005** .042*

.610 1.040 .200 .617 .627 .652

.002** .821 .043* .352 .346 .393

Model parameters LR χ2 (df) = 85.767 (18); p = .000

issues with substance abuse were about 67% less likely (OR = .326, Wald = 11.27, p = .001) to reunify with their families than remain in care. Cases with noted parent/caregiver mental health concerns were about 75% less likely (OR = .254, Wald = 16.25, p = .000) to result in reunification. 4.4. Relative placement Children who experienced more placements were approximately one-third (OR = .677, Wald = 12.043, p = .001) less likely to be permanently placed with a relative than remain in care at the end of the 18 month observation period. Children whose parent or caregiver reported mental health issues were about half as likely (OR = .432, Wald = 7.742, p = .005) to result in placement with relatives rather than remain in foster care (i.e., no permanent outcome). 4.5. Adoption For the permanency outcome of adoption, the analysis showed that as the age of the child at time of removal increased, children were 39% less likely to be adopted (OR = .610, Wald = 9.182, p = .002) than remain in care. Additionally, African American children were 80% less likely (OR = .200, Wald = 4.098, p = .043) to be adopted than remain in care at the end of the observation period. 5. Discussion Child protection mediation is a promising intervention used with child maltreatment cases that shows great potential in helping increase parental compliance and satisfaction with child welfare services, as well as in lowering court costs (Anderson & Whalen, 2004; Colman & Ruppel, 2007). Using secondary data from a pilot mediation project, the focus of this research was on increasing understanding of the impact of child and family characteristics in mediated cases with regard to specific permanency outcomes. 5.1. National context The permanency outcomes of children at the end of the study's 18month evaluation time period should be examined within the larger context. The findings indicate that one-third of the children in the study (n = 103) did not acquire a permanent placement within 18 months, and only 34.4% of the children were reunified with their parents. In 2009, the same year as which data for this study was last collected, 51% of children nationwide who exited foster care were reunified with their parent or caretaker (U.S. Department of Health and Human Services, 2010). It is unclear why there is such a disparate experience with this sample. It may be there were extenuating factors or variables not accounted for in this study that made the mediated cases particularly challenging. In addition, a review of the study's child and family characteristics found some differences when compared to national averages. For example, nationally, the average age of children in foster care in 2009 was 9.6 years, more than five years older than the mean age of this sample (U.S. Department of Health and Human Services, 2010). The authors could not find evidence of national level data specific to child welfare mediation and its relationship to children's age and placement outcomes. However, research has shown that older children may be more likely to be reunified with the parents, whereas younger children—infants in particular, are often viewed as more vulnerable and may be less likely to be reunified (Akin, 2011; Wulczyn, 2004). It should be noted that 35% of the children in this sample were 1 year old or less at the time of their removal. Although age was not a significant factor for family reunification, it is possible that the children's young ages may have had a nominal negative impact on the likelihood of reunification in favor of other outcomes. This supposition is supported by a post-hoc analysis, F(2, 205) = 9.021, p = .000, that showed that

D.M. Aguiniga et al. / Children and Youth Services Review 50 (2015) 20–27

children who were reunified with their family (M = 3.58, SD = 3.85) or were placed with kin (M = 4.9, SD = 4.53), were significantly older than children placed for adoption (M = .97, SD = 1.72). A second noted difference in the distribution of foster children in care in 2009 was by gender. Whereas this mediation sample was 55.3% female and 44.7% male, these numbers were flipped at the national level, with 47% female and 53% male (U.S. Department of Health and Human Services, 2010). It may be that while gender was not significant for reunification in this sample, the 8.3% difference in gender distribution at the national level could impact reunification rates.

5.2. Family reunification The findings indicate that in cases that involved parent/caregiver substance, children were about two-thirds less likely to be reunified with their parents than remain in care. Furthermore, cases that involved mental health concerns were almost 75% less likely to achieve reunification. Choi, Huang, and Ryan (2012) found that mothers receiving substance abuse treatment who also had a mental health concern were less likely to be reunified with their children. Examination of the data revealed that many of the parents/caregivers in this sample (n = 56) experienced both of these conditions. These findings provide important insight for the use of mediation in child protection cases, as they suggest that mediation may not be the most effective tool for families struggling with these issues who are seeking family reunification as a permanency outcome for their children. Additional research is needed to clarify if there were specific factors that, in general, reduced the overall likelihood of reunification, or if cases with this constellation of factors may be better served by a different intervention than mediation. Furthermore, if it were to be found that additional children were reunified after the 18 month evaluation period of this study, questions still arise about the efficacy of mediation in reducing lengthy stays in out-of-home care.

5.3. Relative placement Children who experienced a higher number of placements during the study period were less likely to be permanently placed with relatives. It may be that when children experience multiple placements, their connections to extended family members suffer. This finding is consistent with prior research suggesting that multiple placements reduce the likelihood that children will be permanently placed with relatives (McDonald et al., 2007). It is not unreasonable to assume that moving between placements can make it more challenging to develop or sustain bonds with extended family. Without these bonds, relatives may be less likely to open their homes to the children for permanent placement. Conversely, the lack of a family network might result in a higher number of placements for children. Children who have strong familial ties are likely to be placed with relatives, whether formally or informally, and there is evidence that suggests that relative placements can result in fewer initial placement disruptions (Koh & Testa, 2005) and stable, long-term commitments to raise children (Testa, 2005). The findings also suggest that cases involving parent/caregiver mental health concerns were about one-third less likely to be permanently placed with relatives than remain in care. It is not uncommon for persons with mental illness to have strained or estranged relationships with family (Nicholson, Sweeney, & Geller, 1998; Saunders, 2003). Relatives could be less likely to want to take in a child when the placement would likely require continued contact with a parent with mental illness. It may also be that, due to poor relationships, relatives were unaware that children had been placed in foster care. Efforts by states to identify, locate, and educate family members about mental illness may help reduce this placement disparity for children in care.

25

5.4. Adoption Older children in the sample were less likely to be adopted. This finding is consistent with other studies (c.f. Maza, 2009; Swope, 2012), and highlights the challenges of finding adoptive homes for older children. Several well-documented explanations exist for why older children are less likely to be adopted. First, studies on barriers to adoption from the foster care system have consistently shown that courts and child welfare agencies are reluctant to pursue termination of parental rights, a necessary precursor to adoption, in cases when an adoptive home has not yet been identified for the youth (Macomber et al., 2004). In recognition of this issue, state and federal agencies have dramatically increased efforts in recent years to promote the adoption of older youth and to identify and recruit adoptive families for this population. Additionally, there is evidence that older youth are often reluctant to pursue adoption out of fear that they will lose connections with their birth parents, siblings, and other important family members (Macomber et al., 2004; North American Council on Adoptable Children, 2009). While additional barriers to the adoption of older youth have been identified in the literature, the aforementioned barriers are likely to have a direct impact on placement outcomes of children in this sample, as they directly relate to parties who are typically involved in the mediation process. Race/ethnicity was found to be a significant factor in placement outcome for African American children. The findings of this study indicate that African American children whose families participated in mediation were 80% less likely to be adopted than remain in care at the end of the study period. African American children represent 30.9% of the study sample, but only 8.7% of the children in the study who were adopted. While the evidence of racial disparity in the child welfare system is somewhat mixed, a number of studies have shown that African American children have historically received differential treatment in care such that, in many jurisdictions, they are less likely to be reunified with their parents (Lu et al., 2004; U.S. Department of Health and Human Services, 2011; Wildfire, Barth, & Green, 2007) or adopted (Akin, 2011; McDonald et al., 2007). While increased attention and resources have been dedicated by child welfare agencies in recent years to ameliorate these disparities and increase the overall number of adoptions from care (Ledesma, Pettaway, McRoy, Madden, & Cody, 2011; U.S. Department of Health and Human Services, 2011), the findings of this study suggest that more work is left to be done with regard to African American children. For example, 33.5% of the children who exited care in the state in which this study was conducted in 2009 (the same year in which the data for this study was last collected) were placed for adoption (Texas Department of Family and Protective Services, 2010). Examination of the findings suggests that adoption is a relatively unlikely outcome for children in this mediation sample, with African American children being the least likely to experience the permanency outcome of adoption. When examining adoption outcomes for the overall sample of 311 mediated cases, only 7.4% of the children were placed in adoption; only 2.1% of African American children achieved permanency via adoption. Given that it takes time to recruit adoptive families for children, it is possible that the 18 month time period of this study was insufficient to accurately assess this permanency outcome, as the average wait time for adoption in 2009 from removal to consummation was 38 months (U.S. Department of Health and Human Services, 2010). Overall, there is insufficient data to determine if mediation inhibited adoption or if it actually helped facilitate adoptions at a quicker pace. However, the disparity in adoption rates for African American children remains a concern that needs further exploration. 5.5. Limitations of the study While this study addresses a gap in the literature, the findings of this study should be viewed with some caution. The first limitation pertains

26

D.M. Aguiniga et al. / Children and Youth Services Review 50 (2015) 20–27

to the scope of the study. Given that the sample only included mediated cases, the authors were unable to compare permanency outcomes of the children in this sample with unmediated cases processed through the traditional court process. Second, the pool of cases, from which the 311 mediated used in this analysis were selected, was limited to children whose cases originated in 43 counties in the state in which this study was conducted. Therefore, the findings are not generalizable to children in the state foster care system or nationally. Third, the selection of variables evaluated in the study was based on secondary data, and thus, limited the range of available variables that could be examined. Child protection cases tend to be complex and are influenced by a number of child, family, agency, and systemic factors. This study was not able to address or control for agency/organizational or court related characteristics, such as staff turnover, caseload size, or other factors such as the families' prior history with Child Protective Services. Rather, the variables examined in the study were largely confined to child and family characteristics. Lastly, while these cases were referred to mediation, limited information is known about the mediation referral process, such as the specific stage of the process that cases were referred for mediation or what criteria or factors were used to refer families to mediation. Additionally, for this particular sample, no information was available about how the individual mediation sessions were facilitated or other variables that may have impacted the outcome of the process (e.g., location of mediation, whether joint sessions or “caucusing” was utilized). It is probable that these additional factors may have had an impact on permanency outcomes as they can set the tone for the mediation session and whether parties believe the process is truly unbiased.

6. Conclusion Child protection mediation is a welcome alternative to the often adversarial nature of court proceedings in Child Protective Services. However, there is still insufficient data known about its impact on permanency and other case outcomes. The value of this research lies in its potential to help clarify the types of cases with which mediation may be most beneficial. Ultimately, a greater understanding of the potential factors that influence mediation's outcomes will allow for more targeted referrals and for a more critical analysis of the mediation process and changes that may allow for the facilitation of better permanency outcomes for children in foster care.

References Akin, B. (2011). Predictors of foster care exits to permanency: A competing risks analysis of reunification, guardianship, and adoption. Children and Youth Services Review, 33(6), 999–1011. http://dx.doi.org/10.1016/j.childyouth.2011.01.008. Anderson, G.R., & Whalen, P. (2004). Permanency planning mediation pilot program: Evaluation final report. Lansing, MI: Michigan State Court Administrative Office. Appleyard, K., Egeland, B., & Sroufe, A. (2007). Direct social support for young high risk children: Relations with behavioral and emotional outcomes across time. Journal of Abnormal Child Psychology, 35(3), 443–457. http://dx.doi.org/10.1007/s10802-0079102-y. Avery, R.J., & Freundlich, M. (2008). You're all grown up now: Termination of foster care support at age 18. Journal of Adolescence, 32(2), 247–257. http://dx.doi.org/10.1016/j. adolescence.2008.03.009. Barsky, A., & Trocmé, N. (1998). The essential aspects of mediation in child protection cases. Children and Youth Services Review, 20(7), 629–656. Bayaga, A. (2010). Multiple logistic regression: Usage and application in risk analysis. Quantitative Methods Inquiries, 5(2), 288–297. Becker, M., Jordan, N., & Larsen, R. (2007). Predictors of successful permanency planning and length of stay in foster care: The role of race, diagnosis and place of residence. Children and Youth Services Review, 29(8), 1102–1113. http://dx.doi.org/10.1016/j. childyouth.2007.04.009. Bernstein, M.M. (1998). Child protection mediation: Its time has arrived. Canadian Family Law Quarterly, 16(1), 73–119. Carlson, E., & Sroufe, L.A. (1995). The contribution of attachment theory to developmental psychopathology. In D. Cicchetti, & D. Cohen (Eds.), Developmental processes and psychopathology: Theoretical perspectives and methodological approaches, Vol. 1. (pp. 581–617). New York: Cambridge University Press. Children's Permanency Cooperative (2000). Fast forward to permanency: Final report (October 1997–September 2000). Houston, TX: Harris County.

Choi, S., Huang, H., & Ryan, J.P. (2012). Substance abuse treatment completion in child welfare: Does substance abuse treatment completion matter in the decision to reunify families? Children and Youth Services Review, 34(9), 1639–1645. http://dx.doi.org/ 10.1016/j.childyouth.2012.04.022. Colman, R., & Ruppel, J. (2007). Child permanency mediation pilot project: Multi-site process and outcome evaluation study. Rensselaer, NY: New York State Office of Children and Family Services. Connel, C.M., Katz, K.H., Saunders, L., & Tebes, J.K. (2005). Leaving foster care: The influence of child and case characteristics on foster care exit rates. Children and Youth Services Review, 28(7), 780–798. http://dx.doi.org/10.1016/j.childyouth.2005.08.007. Crush, L. (2005). Child protection mediation: A model for Ontario. Ontario, Canada: Queen's University. Dobbin, S.A., Gatowski, S.I., & Litchfield, M. (2001). The Essex County Child Welfare Mediation Program: Evaluation results and recommendations. Technical Assistance Bulletin, Vol. V, No. 4, Reno, NV: NCJFCJ. Eaton, M., Whalen, P., & Anderson, G.R. (2007). Permanency planning mediation pilot program: The Michigan experience. Michigan Child Welfare Law Journal, 10(3), 2–10. Edwards, L. (2009). Child protection mediation: A 25-year perspective. Family Court Review, 47(1), 69–80. http://dx.doi.org/10.1111/j.1744-1617.2009.00240.x. Flower, C., McDonald, J., & Sumski, M. (2005). Review of turnover in Milwaukee County private agency child welfare ongoing case management staff. Retrieved from http:// legis.wisconsin.gov/lc/committees/study/2008/SFAM08/files/turnoverstudy.pdf Freundlich, M., Avery, R.J., Munson, S., & Gerstenzang, S. (2006). The meaning of permanency in child welfare: Multiple stakeholder perspectives. Children and Youth Services Review, 28(7), 741–760. http://dx.doi.org/10.1016/j.childyouth.2005.08.008. Gatowski, S.I., Dobbin, S.A., Litchfield, M., & Oetjen, J. (2005). Mediation in child protection cases: An evaluation of the Washington, D.C. family court child protection mediation program. Reno, NV: NCJFCJ. Giovannucci, M.T. (2007). Future trends in state courts: Dependency mediation. Retrieved from http://www.nmma.info/images/ADRMedTrends2007.pdf Hardin, M. (1992). Establishing a core of services for families subject to state intervention. Washington, D.C.: American Bar Association. Jenson, J.M., & Fraser, M.W. (2006). Social policy for children & families: A risk and resilience perspective. Thousand Oaks, CA: Sage. Jones-Harden, B. (2004). Safety and stability for foster children: A developmental perspective. The Future of Children, 14(1), 31–48. Kemp, S.P., & Bodonyi, J. (2000). Infants who stay in foster care: Child characteristics and permanency outcomes of legally free children first placed as infants. Child and Family Social Work, 5(2), 95–106. http://dx.doi.org/10.1046/j.1365-2206.2000.00156.x. Koh, H.J. (2004). “Yet I shall temper so justice with mercy”: Procedural justice in mediation and litigation. Law and Psychology Review, 28, 169–176. Koh, H. J., & Testa, M. F. (2005). Propensity score matching of children in kinship and nonkinship foster care: Do permanency outcomes still differ? Social Work Research, 32(2), 105–116. Koh, E., & Testa, M.F. (2008). Propensity score matching of children in kinship and nonkinship foster care: Do permanency outcomes still differ? Social Work Research, 3(2), 105–116. Ledesma, K., Pettaway, S., McRoy, R., Madden, E.E., & Cody, P. (2011). Engaging African American communities and organizations to support foster care and adoption for children in the child welfare system. The Roundtable, 25(1), 1–12. Lu, Y.E., Landsverk, J., Ellis-MacLeod, E., Newton, R., Ganger, W., & Johnson, I. (2004). Race, ethnicity and case outcomes in child protective services. Children and Youth Services Review, 26(5), 447–461. http://dx.doi.org/10.1016/j.childyouth.2004.02.002. Macomber, J.E., Scarcella, C.A., Zielewski, E.H., & Geen, R. (2004). Foster care adoption in the United States: A state by state analysis of barriers & promising approaches. Retrieved from http://www.urban.org/UploadedPDF/411108%5FFosterCareAdoption.pdf Madden, E.E., & Aguiniga, D.M. (2013). An evaluation of permanency outcomes of child protection mediation. Journal of Public Child Welfare, 7(1), 98–121. http://dx.doi.org/10. 1080/15548732.2012.740442. Masten, A.S., & Coatsworth, J.D. (1998). The development of competence in favorable and unfavorable environments: Lessons from research on successful children. American Psychologist, 53(2), 205–220. http://dx.doi.org/10.1037/0003-066X.53.2.205. Maza, P. (2009). A new look at the role of ASFA and children's ages in adoption. The Roundtable, 23(1), 1–3. McDonald, T.P., Poertner, J., & Jennings, M.A. (2007). Permanency for children in foster care: A competing risks analysis. Journal of Social Service Research, 33(4), 45–56. http://dx.doi.org/10.1300/J079v33n04_04. McRoy, R.G., Lynch, C.J., Chanmugam, A., Madden, E.E., & Ayers-Lopez, S. (2009). Children from care can be adopted. In G. Wrobel, & E. Neil (Eds.), International advances in adoption research for practice (pp. 97–118). London: John Wiley & Sons. Moutinho, L., & Hutcheson, G. (2011). The SAGE dictionary of quantitative management research. London: SAGE Publications, Ltd. Murphy, D., Bandy, T., Schmitz, H., & Moore, K. (2013). Caring adults: Important for positive child well-being (publication #2013-54). Retrieved from http://www. childtrends.org/wp-content/uploads/2013/12/2013-54CaringAdults.pdf Murphy, A., Van Zyl, R., Collins-Camargo, C., & Sullivan, D. (2012). Assessing systemic barriers to permanency achievement for children in out-of-home care: Development of the Child Permanency Barriers Scale. Child Welfare, 91(5), 37–71. Nasworthy, C., & Tarver, T. (2002). Report on the implementation of the Children's Justice Act Mediation Pilot Projects. Austin, TX: The University of Texas at Austin. Nasworthy, C., & Tarver, T. (2005). Report on the implementation of the Children's Justice Act Mediation Pilot Projects 2003–2005. Austin, TX: The University of Texas at Austin. National Resource Center for Foster Care and Permanency Planning (1998). Tools for permanency: Tool #3: Child welfare mediation. New York: Hunter College School of Social Work.

D.M. Aguiniga et al. / Children and Youth Services Review 50 (2015) 20–27 Nicholson, J., Sweeney, E.M., & Geller, J.L. (1998). Mothers with mental illness: 2. The competing demands of parenting and living with mental illness. Psychiatric Services, 49(5), 643–649. North American Council on Adoptable Children (2009). It's time to make older child adoption a reality because every child and youth deserves a family. Retrieved from http:// www.nacac.org/adoptalk/MakeOlderChildAdoptionReality.pdf. Olson, K.B. (2003). Lessons learned from a child protection mediation program: If at first you succeed and then you don't. Family Court Review, 41(4), 480–496. http://dx.doi. org/10.1177/15312445030414008. Pecora, P.J., Whittaker, J.K., Maluccio, A.N., & Barth, R.P. (2000). The child welfare challenge: Policy, practice, and research (2nd ed.). New York: Aldine de Gruyter. Plunkett, R., & Osmond, O. (2004). Permanency planning: Choosing between long-term foster care and adoption. Ontario Association of Children's Aid Societies (OACAS) Journal, 48(1), 7–14. Rosenthal, J.A. (1993). Outcomes of adoption of children with special needs. The Future of Children, 3(1), 77–88. Saunders, J. (2003). Families living with severe mental illness: A literature review. Issues in Mental Health Nursing, 24, 175–198. http://dx.doi.org/10.1080/01612840390160711. Scales, P., & Leffert, N. (1999). Developmental assets: A synthesis of the scientific research on adolescent development. Minneapolis: Search Institute. Stack, K. (2003). Information packet: Child welfare mediation. Retrieved from http:// www.hunter.cuny.edu/socwork/nrcfcpp/downloads/child-welfare-mediation.pdf Supreme Court of Virginia (2002). Child dependency mediation. Retrieved from http:// cdm266901.cdmhost.com/cdm4/item_viewer.php?CISOROOT=/ p266901coll6&CISOPTR=146&CISOBOX=1&REC=2 Swope, H. (2012). The challenge of older youth adoptions from foster care. The Roundtable, 25(2), 1–2. Testa, M.F. (2004). When children cannot go home: Adoption and guardianship. Future of Children, 14(1), 115–129. Testa, M.F. (2005). The quality of permanence — Lasting or binding? Subsidized guardianship and kinship foster care as alternatives to adoption. Virginia Journal of Social Policy and the Law, 12, 499–534. Texas Department of Family and Protective Services (2010). 2009 Annual report and data book (Retrieved from http://www.dfps.state.tx.us/documents/about/data_books_ and_annual_reports/2009/2009databook.pdf). Thoennes, N. (2001). Permanent custody mediation. Denver, CO: Lucas County Court of Common Pleas Juvenile Division.

27

Thoennes, N. (2009). What we know now: Findings from dependency mediation research. Family Court Review, 47(1), 21–37. http://dx.doi.org/10.1111/j.1744-1617. 2009.00237.x. U.S. Department of Health and Human Services (2001). Assessing the context of permanency and reunification in the foster care system. Washington, D.C.: Assistant Secretary for Planning and Evaluation. U.S. Department of Health and Human Services (2005). A report to congress on adoption and other permanency outcomes for children in foster care: Focus on older children. Washington, D.C.: Children's Bureau. U.S. Department of Health and Human Services (2006). Working with the courts in child protection. Retrieved from http://www.childwelfare.gov/pubs/usermanuals/courts/ index.cfm U.S. Department of Health and Human Services (2010). The AFCARS report preliminary FY 2009 estimates as of July 2010 (17). Washington, D.C.: Children's Bureau. U.S. Department of Health and Human Services (2011). Addressing racial disproportionality in child welfare. Washington, D.C.: Children's Bureau. U.S. General Accounting Office (2002). Foster care: Recent legislation helps states focus on finding permanent homes for children, but long-standing barriers remain (GAO-02585). Washington, D.C.: Author. Vidmar, N.J. (1992). Procedural justice and alternative dispute resolution. Psychological Science, 3, 224. http://dx.doi.org/10.1111/j.1467-9280.1992.tb00033.x. Wall, J.A., Stark, J., & Standifer, R.L. (2001). Mediation: A current review and theory development. Journal of Conflict Resolution, 45(3), 370–391. http://dx.doi.org/10.1177/ 0022002701045003006. Weigensberg, E.C. (2010, January 17). Child welfare agency performance: How are child, agency, and county factors related to achieving timely permanency outcomes for children in foster care? Paper presented at Society for Social Work Research 14th Annual Conference. San Francisco, CA: SSWR. Wildfire, J., Barth, R. P., & Green, R. L. (2007). In R. Haskins, F. Wulczyn, & M. B. Webb (Eds.), Reunification of children from foster care at 18-months: Findings from the National Survey of Child and Adolescent Well-Being. Child protection: Using research to improve policy and practice. (pp. 155–170). Washington, DC: Brookings. Wilson, J.B., Katz, J., & Geen, R. (2005). Listening to parents: Overcoming barriers to the adoption of children from foster care. Cambridge, Mass: Harvard University, John F. Kennedy School of Government. Wulczyn, F. (2004). Family reunification. The Future of Children, 14(1), 95–113.