The Role of Kinship and Siblings in Young Children’s Placement Preferences

Abstract

Although considerable attention has been directed toward the most appropriate placement for children following removal from home due to maltreatment, very little of this attention has focused on children’s stated preferences, particularly when children are young. Specifically, children below 12 years of age are often presumed incompetent to form reasoned judgments about their best interests in placement. This assumption, however, has rarely been tested directly. We surveyed 100 4- to 11-year-olds removed from home because of maltreatment about their placement preferences. Children were less likely to indicate they wanted to return home if they were placed with siblings or with kin, consistent with statutory placement preferences. These results suggest that young children may express more mature preferences than recognized by the law, and that there may be value in asking even relatively young children about with whom they would like to live following their removal from home as a result of maltreatment.

Each year, more than 200,000 children in the United States are removed from parental custody due to substantiated maltreatment (U.S. Department of Health and Human Services [USDHHS], 2017). When this occurs, one of the first and most critical determinations concerns where and with whom the children should live. Particularly in the last few decades, social services agencies and the courts have prioritized placing children with relatives in what is called kinship care (Cuddeback, 2004). When relatives are not available or appropriate, children are placed with foster families or in congregate care facilities. Regardless of the type of placement, out-of-home care is for the most part intended to be temporary, given that family reunification is the primary initial goal for a vast majority of children who have been removed (Child Welfare Information Gateway, 2016b; Wulczyn, 2004). When children cannot safely return to their parents, the temporary placements may become permanent, or long-term placement solutions must be found.

When determining an appropriate placement for children, numerous factors are considered. The most important ones include children’s well-being, needs for safety and stability, and the capacity of caregivers to meet those needs (Child Welfare Information Gateway, 2016a). What does not figure prominently in these factors, however, is children’s own preferences. In fact, in an examination of state laws regarding overarching principles of assessing children’s best interests, including in regard to placement, the Child Welfare Information Gateway (2016a) was only able to identify 12 states that explicitly required children’s preferences be considered when making placement decisions. Even then, the state’s requirements typically contained the caveat that children need to be of sufficient age and maturity to express a “reasonable preference,” with 12 years of age being the most common stated cutoff (Child Welfare Information Gateway, 2016a) and younger children being presumed incompetent to express a reasonable preference.

California’s statute provides an example of a type of ambivalence regarding children’s competence to provide input on their placement. In dependency cases, in which children are removed from their home because of parental maltreatment, the child’s attorney is charged with assessing the child’s preferences. The law requires that “if the child is four years of age or older, [the child’s] counsel shall interview the child to determine the child’s wishes and assess the child’s well-being, and shall advise the court of the child’s wishes.” Given that attorneys are typically charged with the duty to zealously advocate for their client’s wishes, this requirement suggests that California recognizes that some weight should be put on even a young child’s placement preferences, provided via the child’s attorney. However, the next sentence warns that the child’s attorney “shall not advocate for the return of the child if, to the best of his or her knowledge, return of the child conflicts with the protection and safety of the child” (California Welfare & Institutions Code Section 317(e)(2), 2019). The statute thus implicitly assumes that children’s wishes are often contrary to their protection and safety, and provides no guidance with respect to how the child’s attorney is supposed to assess children’s competence to provide a placement preference.

One way that the legal system may assess whether young children are competent to express a preference is to evaluate whether their stated desires track legal recommendations. In California, as in many other states, when children are removed from their parents, social services are required to give “preferential consideration” to relatives (California Welfare & Institutions Code Section 361.3, 2019), and to make “a diligent effort” to place siblings together in the same placement (California Welfare & Institutions Code Section 16002, 2019). Assessing children’s competency by comparing their preferences to legal preferences avoids the difficulty that competency will be underestimated if children are asked to explain the basis for their desires. As an example, Garrison (1991) investigated children’s ability to express reasonable preferences about post-divorce custody. Nine-year-olds appeared as competent as 14-year-olds and 18-year-olds when the reasonableness of their preferences was assessed, but inferior when their stated rationale was assessed. Garrison acknowledged that the younger youth’s difficulties explaining their preferences could be attributed to their more limited cognitive inferential abilities relative to those of the older youth.

In the child maltreatment context, the extent to which children’s placement preferences are consistent with the notion of “preferential consideration” is unknown. To address this question, we asked 4- to 11-year-old children who had been removed from their homes due to substantiated maltreatment about their preferences regarding placement. We assessed whether their expressed preferences varied depending on their age and their current placement, specifically whether they were living with relatives or with siblings. We also asked separately about their preferences in the short term (where they wanted to stay, if they could “stay anywhere”) and in the long term (where they wanted to “live until they were a grown-up”). Consistent with well-accepted guidelines regarding interviewing children (American Professional Society on the Abuse of Children [APSAC], 2012; Lamb, Hershkowitz, Orbach, & Esplin, 2008), we prompted children with an open-ended recall question about “where” they would like to live, rather than with a yes/no or forced-choice question.

Children’s Placement Preferences

Prior work on children’s preferences about their placement provides only limited insight into young children’s ability to express a preference. In a review of extant research on children’s attitudes about out-of-home placement, Fox and Berrick (2007) noted that most research on children’s functioning in and perceptions of foster care has relied on other sources of information, such as caseworkers, caregivers, and administrative materials; very few studies have asked children directly about their experiences, perceptions, and preferences. Among those studies that did question children, most were “retrospective and therefore provide data filtered by subjects’ long-term memory, while only a handful of studies have involved interviews with children while in care” (Fox & Berrick, 2007, p. 24). Furthermore, among studies with potential to provide knowledge about children’s own perceptions of placement with extended relatives or siblings, Fox and Berrick (2007) noted that “[e]ven fewer have specifically examined the experiences of children living in kinship care” (p. 24).

A notable exception is the National Survey of Child and Adolescent Well-Being (NSCAW; USDHHS, 2001, 2005), a nationally representative survey of several thousand children involved in Child Protective Service investigations. Although many of the questions included in the survey were administered only to those aged 11 years and older, children as young as 6 years were asked a number of questions about their out-of-home placement experiences, including two directly about their placement preferences: A yes/no question asking whether children wanted their current placement to be their “permanent home,” and a recall question asking with whom children would live if they “could live anywhere or with anyone.”

The NSCAW questions were based on those initially asked in a smaller study of 100 6- to 14-year-old children in foster care, most of whom had been in placement for over 5 years (Fox, Frasch, & Berrick, 2000, 2008). In the initial study, 77% of children answered “yes” when asked the yes/no question about whether they wanted their current placement to be their permanent home, and “yes” responses were correlated with reporting feeling very safe in their placement, liking living in their placement, and feeling part of the placement’s family. “Yes” responses were also positively correlated with the length of time children had spent in the placement. These results suggest that children were expressing legally reasonable preferences. However, the researchers warned that children’s responses to the yes/no question were often discrepant with their responses to the subsequent “where would you live if you could live with anyone” question, because, to the latter, only 37% named their current placement. No correlates with the “where” question were reported. Furthermore, the study did not assess whether children’s preferences were related to their placement with a relative or sibling. Indeed, the researchers omitted the yes/no question with children “who might have assumed their current placement was permanent (mostly children in kin placements)” (Fox, Frasch, & Berrick, 2000, p. 152). With respect to age, although the age range of the children in the sample included young children, who would be presumed incompetent to express reasonable preferences, no age analyses were reported.

In the subsequent larger national survey, two sets of 6- to 15-year-olds completed the placement questions: 320 children comprising the OYFC sample (One Year in Foster Care, also known as the Long-Term Foster Care or LTFC sample), and 641 children comprising the Child Protective Service (CPS) sample. As the names indicate, children in the OYFC sample had been in out-of-home placement for over a year, whereas children in the CPS sample, in contrast, had been in out-of-home placement for less time, typically 7 months (USDHHS, 2001, pp. 7-16).

In the OYFC sample, 50% of children answered “yes” to the yes/no question about wanting their current placement to be permanent. Although kinship care appeared preferable to children (65% vs. 50%), this difference was not statistically significant, with the only significant difference appearing among the children in congregate or group care, who assented at a substantially lower rate (22%). When asked the “where” would you live if you could live with anyone question, 66% named their biological mother or father as their preferred placement (USDHHS, 2001). Subsequent reports revealed comparable trends (Barth, 2002; Chapman, Wall, & Barth, 2004). In the CPS sample, 38% of children answered “yes” to the yes/no question suggesting that they wanted their current placement to be permanent, with children in kinship care (61%) answering yes more so than children in foster care (27%) (USDHHS, 2005, Table 7-16, pp. 7-17). When asked the “where” question, 81% named their mother or father. The reports did not examine the correlates of the “where” question and did not consider sibling placement.

In both the OYFC and CPS samples, about half of the children were 11 years or older. Thus, the studies combined children who would be legally presumed competent with children presumed incompetent. Age effects across the studies were inconsistent. Younger children in the OYFC sample were more likely to state that they wanted their current placement to be permanent (Chapman et al., 2004), whereas no age differences emerged in the CPS sample (USDHHS, 2005, pp. 7-17). A subsequent study, which combined children’s answers across different questions (including the yes/no question about permanence), and incorporated later waves of questioning among the CPS sample, found that younger children were less likely to be satisfied with their placement than were older children (Merritt, 2011).

We have identified only one study using the NSCAW data that assessed placement with siblings. Hegar and Rosenthal (2009) examined 1,823 observations, which contained a combination of the OYFC and CPS samples and subsequent waves of questioning of the CPS sample (at either 18 or 36 months after the first interview). The authors compared children’s responses to “Do you like living with the people you live with?” and “Do you feel like you’re part of this family?” between those who were and were not placed with siblings. Results were mixed. Sibling placement interacted with the sample (OYFC vs. CPS), such that children living with siblings were nonsignificantly less likely to say they liked their placement in the OYFC sample, and significantly more likely to say that they liked their placement in the CPS sample. Although the interaction was not predicted, the authors speculated that perhaps siblings increase children’s happiness with their placement mainly in the short term. Other measures examined in the study were only administered to children 11 years of age and older, and the authors did not examine, in relation to sibling placement, children’s responses to the yes/no question about wanting their current placement to be permanent or the where question regarding with whom children would like to live.

Two important concerns have been noted by researchers relying on the OYFC and CPS samples in regard to interpreting children’s responses. First, children were interviewed in their current placement, and the reports acknowledge that “children may not have felt comfortable enough in the interview setting to share more negative thoughts about their current placement” (USDHHS, 2005, pp. 7-17; see also USDHHS, 2001, p. 94). This would have led to an exaggeration of children’s preference to remain in their placement. Second, the discrepancy between children’s answer to the yes/no question (in which children tended to answer yes indicating that they liked and wanted to be a part of their current placement) and the “where” question, in which children were much more likely to endorse wanting to return to their parent, led the authors to question the competency of children’s answers desiring placement with parents, arguing that “implicit in that desire may be that they would like to live with another, more idealized version of their parent” (USDHHS, 2005, pp. 7-17). It is difficult, though, to reconcile these two concerns, given that the former would also imply that children should name their current placement to both questions if they felt uncomfortable expressing negative thoughts about that placement. An alternative interpretation for the inconsistency is that the yes/no questions might have led to response biases, a concern often raised when evaluating children’s responses to yes/no questions (e.g., Bruck, Ceci, & Hembrooke, 2002).

In summary, although the NSCAW studies provide some insight into the potential for children below 12 years to express reasonable placement preferences, the studies also highlight the need for further research. First, many of the measures were only administered to children 11 years and older, and the samples that included young children (as young as 6 years) contained a large percentage of older children, with inconsistent age-related changes in reporting being evident. Second, the relation between children’s preferences and either kinship or sibling placement were inconsistent, and hampered by focusing attention on the yes/no question with respect to kinship placement, and the yes/no liking placement question with respect to sibling placement. Third, because children were questioned in their placement, children might have been reluctant to express negative feelings about their placement.

The Present Study

We asked children, aged 4 to 11 years, who had been removed from their homes due to substantiated maltreatment about their placement experiences and preferences. Children were questioned while awaiting court appearances at the Los Angeles County Juvenile Court, Dependency Division. We asked open-ended questions about children’s preferences for short and long-term placement to determine whether children differentiated these placement experiences, and we compared children’s responses across age, placement type, and placement with siblings. We predicted that children would be more likely to express a desire to return home if they were in foster care (not in kin placement) and separated from their siblings. We tested whether children’s preferences were related to age, although given the inconsistent results of prior studies, we did not predict how age would affect children’s preferences.

Method

Participants

Participants included 100 children, aged 4 to 11 years (M = 7.74 years, SD = 2.10) who were under the jurisdiction of the Los Angeles County Juvenile Court due to child maltreatment. Fifty-six percent were male; 38% were African American, 35% were Latino, and 25% were Caucasian. Two children’s ethnicities were unknown. About 30% of the children were new to the dependency system, and present for a detention hearing, having been removed from their parents within the past few days; 70% of the children were post-detention and had been out of home for a longer period of time (M = 2.48 years, SD = 2.75, Median = 1.68, range: 0.03 −11.70). None of the children were present in court for adoption proceedings or for contested adjudication or disposition hearings at which they might have to testify. About 45% of children had experienced two or more placement changes (M = 2.15 placements, SD = 1.83; range: 1-11). A small number of children (n = 9) had been removed from a guardian (e.g., grandparent) who was not a biological parent. For children who had lived with that guardian for 12 months or longer (n = 5), that guardian was considered the offending caregiver; otherwise we considered the biological parent the offending caregiver. (Changing the classification of these nine children did not affect the results.)

Materials and Procedure

The study was approved by the relevant university’s Institutional Review Board, and consent to question children was granted by the Presiding Judge of the Los Angeles County Juvenile Court and children’s dependency attorneys. Interviews were conducted over the course of 1 year. Children were questioned individually in a quiet corner of a care facility while awaiting court hearings for their case. Children who were incapable of communicating in English were excluded. Children’s answers were shared with their attorneys. All but one child agreed to participate in the study.

At the beginning of each interview, an interviewer explained the study and obtained children’s assent. She then asked a standard series of questions about children’s placement and court experiences. Those relevant to the present study are described here. First were demographic questions about children and their current living situation (e.g., name, age, date of birth, where they are living now, whether they are with a sibling). Second were questions about prior placements, such as where and with whom children had lived prior to their removal from parent or other guardian custody. Third, and most important, were two open-ended placement preference questions. One concerned children’s immediate preference: “Where would you like to stay if you could stay anywhere?” and the other concerned their long-term desire: “Where would you like to live until you are a grown-up?”

Court records verified demographic characteristics (e.g., date of birth, gender, race, ethnicity), case characteristics, and placement history. From this information, length of time in out-of-home care and number of prior placement changes were determined.

Results

A majority of children resided in nonrelative foster care (70%), while the remaining 30% resided in kinship care. More than half of the sample (63%) shared their current placement with at least one sibling (all but five children came from families with siblings; these latter five children were excluded from analyses concerning sibling placement). Preliminary analyses (t tests, chi-square analyses, and correlations, as appropriate) revealed that, across placement type and sibling arrangements, children were comparable in gender and ethnicity, χ^2s(l) ≤ .47, ps ≥ .51, as well as age, length of time in years in out-of-home care, and number of placement changes, ts(89-98) ≤ 1.30, ps ≥ .19. These demographic variables were, as well, unrelated to children’s preferences for either short- or long-term placement, rs(93) ranged from −0.15 to 0.07, ps > .05, and are not considered further (results not shown). Finally, type of case (new detention vs. post-detention) was unrelated to children’s placement preferences, χ²(1) ≤ 1.39, ps > .24.

When asked about short-term placement preferences, shown in Table 1, just over half (52%) of the children indicated that they wanted to return home. Children who did not want to return home immediately most often listed their current placement as their preference, followed by another adult (often relative) caregiver, or their nonoffending parent. One child stated a preference for returning to a caregiver with whom the child had lived for 2 years, but from whom she had been removed. A very small portion of children provided an ambiguous response or said, “I don’t know.”

Table 1.

Children’s Placement Preferences (N = 100).

Variable	%
Short-term preference: Where would you like to stay if you could stay anywhere?
Return home	52
Current placement	25
Other adult caregiver	14
Nonoffending parent	3
Ambiguous response	3
Don’t know	2
Prior offending caregiver	1
Long-term preference: Where would you like to live until you are a grown-up?
Return home	57
Current placement	17
Other adult caregiver	15
Nonoffending parent	4
Ambiguous response	2
Don’t know	3
Prior offending caregiver	2

Note. Prior offending caregiver refers to a caregiver with whom the child had resided with, but then from whom the child was removed.

When asked about their long-term placement preferences, a slightly larger percent (57%) indicated that they wanted to return home. The remaining children listed their current placement, another adult caregiver, a nonoffending parent, or a prior offending caregiver with whom the child had lived briefly but then from whom the child had been removed. Five children gave either an ambiguous response or an “I don’t know” response (see Table 1).

Children’s desire to return home did not significantly vary when asked the short-term and long-term questions, McNemar’s test p = .45. One fifth of children (17%) showed differentiation in their desires: 11% reported that they wanted to live in out-of-home care temporarily and return home eventually, and 6% reported that they wanted to return home temporarily, but live in out-of-home of care permanently.

We next tested whether children’s placement with extended relatives or with siblings affected their placement preferences. Because children’s answers to the short-term and long-term questions were similar, we conducted a generalized estimating equation (GEE). Children’s responses to the questions regarding their short- and long-term placement preferences, coded as 0 = any other placement or 1 = home, were included as correlated dependent measures. Given the binary nature of these responses, a binomial distribution and logit link function were specified, along with an unstructured covariance matrix. Independent variables included children’s current placement type (0 = nonkinship care, 1 = kinship care), placement with siblings (0 = no, 1 = yes), question type (0 = the question about children’s short-term preferences, 1= the question about children’s long-term preferences), and the interaction between each of the placement variables (i.e., kinship care and sibling placement) and question type. Robust standard errors were obtained.

Results, shown in Table 2 and in Figure 1, revealed that children living in kinship care were less likely to report wanting to return home in the short term and long term than children in nonkinship care, b = −1.25, SE = 0.49, p < .05 (odds ratio [OR] = .28, 95% confidence interval [CI] [−2.21, −0.30]). Similarly, children living with siblings had a lower likelihood of wanting to return home across these two time frames than children living without siblings, b = −1.17, SE = 0.56, p < .05 (OR = .31, 95% CI [−2.28, −0.07]). Stated another way, children who were separated from relatives or siblings were consistently more likely to want to return home to their maltreating caregivers, both in the short and long term, than those living with some form of kin. Neither question type nor the interactions between question type and placement variables emerged as significant correlates, ps > .05. However, inspection of the figure suggests that children placed with kin or with siblings exhibited some discrimination between their short- and long-term desires, with a stronger preference to return home in the long term.

Table 2.

Predictors of Children’s Short- and Long-Term Placement Preferences.

Model	b	SE	z	95% CI	OR
Kinship placement	−1.25*	0.49	−2.58	[−2.21, −0.30]	0.28
Sibling placement	−1.17*	0.56	−2.08	[−2.28, −0.07]	0.31
Question type	0.04	0.45	0.09	[−0.84, 0.91]	1.04
Kinship Placement × Question Type	0.37	0.40	0.96	[−0.40, 1.15]	1.46
Sibling Placement × Question Type	0.25	0.46	0.55	[−0.64, 1.16]	1.29

Note. The overall model Wald χ² (5) = 14.93, p < .05. For kinship placement, the reference group is nonkinship care. For sibling placement, the reference group is nonsibling placement. The question type variable is coded as 0 = short-term question, 1 = long-term question. The dependent variables are coded as 1= wants to return home; 0 = wants to live at any other placement. Robust standard errors are displayed. CI = confidence interval; OR = odds ratio.

^*p < .05.

Figure 1.

Percentage of children who wanted to return home (0 = no, 1 = yes) as a function of placement type. The four placement types are displayed (kinship care, nonkinship care, sibling placement, and nonsibling placement). Short-term refers to children’s short-term placement preferences and long-term refers to children’s long-term placement preferences. Comparisons between children living in nonkinship versus kinship care and children living in sibling versus nonsibling placements were statistically significant for both short- and long-term preferences, ps < .05.

Discussion

The purpose of the current study was to evaluate young children’s placement preferences, and determine how placement with kin, either caregivers or siblings, influenced those preferences. We focused specifically on children ages 11 years and younger, given that they are typically presumed incompetent in making reasoned judgments about their placement, and given limited research directly testing this assumption. And, unlike in some prior work, we asked children open-ended questions about with whom they wanted to live, in the short term and in the long term, and were thus able to ascertain whether their stated preferences were consistent with legal preferences. We found that children’s preferences were affected by whether they are placed with family, either kin or siblings, consistent with legal preferences, suggesting that children presumed legally incompetent can nevertheless express sensible placement preferences.

Prior work, much of which has relied on data from the NSCAW, has not been able to explicitly address the question of whether young children can provide valuable input into legal assessment of their appropriate placement. First, the NSCAW questions were not designed to examine the legal competency of young children’s preferences about placement. Instead, older and younger children were combined, or measures and questions were omitted with younger children (e.g., USDHHS, 2001, 2005). Second, as the NSCAW researchers emphasized, their methodology had some constraints that limit inferences about children’s placement preferences. First, because children were questioned in their current placement, they may have been reluctant to share negative perceptions about that placement, and instead endorsed wanting their current placement to be permanent. In contrast, we asked children for their opinions before a court hearing at which their attorneys would share their desires with the court charged with making decisions about the children’s future placement. If anything, children concerned about their parents’ reactions should minimize their happiness with their current placement. Second, the NSCAW researchers took the position that inconsistencies between children’s responses to the yes/no question (“Do you want your current placement to be your permanent home?” to which many children assented) and recall question (“Where would you live if you could live anywhere or with anyone?’ to which many children indicated a preference for returning to a biological parent) raised concerns about children’s ability to answer the recall question, and subsequent analyses focused more often on their responses to the yes/no question about their current placement being permanent and other yes/no questions.

In the present study, consistent with growing research suggesting the benefits of wh- questions (i.e., who, what, when, where, how; APSAC, 2012; Lamb et al., 2008) when conducting interviews with children, we asked children open-ended questions about where they wanted to live to allow them to generate responses. Our results clearly showed that 4- to 11-year-old children could express a reasonable preference regarding their desired placement, and, of importance, their responses were consistent with legal recommendations that prioritize placement with relatives and siblings. Our results, as well, are contrary to presumptions of competence that indicate only youth aged 12 years and older can provide input into their placement. Children far younger than 12 years are, in many circumstances, able to answer questions about where they would like to live, and securing their input would be a valuable addition to current approaches to placement determinations.

Unlike in prior work, which has produced inconsistent results regarding children’s preferences to be placed with kin or siblings, our research revealed that the presence of siblings and kin emerged as key influences on children’s placement preferences for both short-term and long-term placement. Children who were separated from extended family or siblings (i.e., in nonkinship care without any siblings present) were consistently more likely to want to return home to their maltreating caregivers, whereas children who were living with either form of kin more often endorsed a preference for remaining in their current placement or indicated a desire to live with another close relative. Hints emerged, as well, suggesting that the latter children may exhibit some discrimination between their short- and long-term desires, with a slightly greater preference for returning home in the long term. Thus, at least some children may recognize that it is in their best interests, at least initially, to live with someone other than the caregiver from whom they were removed, while still maintaining hope in returning home eventually. Future work should explore this possibility in greater detail, ideally with open-ended follow-up prompts to allow children to elaborate on their perceptions and reasons.

Our findings relevant to kinship and sibling placements are important in, first, highlighting the potential maturity of even relatively young children’s responses. That children express clear preferences for maintaining connections with extended relatives and siblings aligns remarkably well with statutory preferences, which prioritize family preservation and thus children’s placement with kin and siblings whenever possible and appropriate (e.g., California Welfare & Institutions Code Section 361.3, 2019; California Welfare & Institutions Code Section 16002, 2019). These findings therefore suggest that young children are able to provide “reasonable preferences,” as recognized by the law, at least when asked directly about with whom they want to live. Second, these findings may in some ways clarify those of the NSCAW, which have not consistently found that children’s preferences are tied to kinship or sibling presence (Barth, 2002; Chapman et al., 2004; USDHHS, 2005). NSCAW has often incorporated varied samples (i.e., CPS and OYFC), has at times omitted placement questions with children in kinship care (“who might have assumed their current placement was permanent”; Fox et al., 2000, 2008, p. 152), and has not, as mentioned, focused on children’s responses to open-ended questions about their placement. Our findings are clear in showing that children’s preferences—for both short and long-term placement—are influenced by their current placement with kin and siblings.

The evident desire to maintain strong family connections with extended family and siblings among maltreated children removed from home is perhaps unsurprising. Family often provides a sense of security and familiarity, and siblings, especially in maltreating homes, often form strong attachments, and even caregiver types of relationships with one another (e.g., Katz & Hamama, 2018). These relationships likely provide important support, including in times of transition and uncertainty, such as that caused by children’s removal from home, and when they experience extended periods of separation from their primary caregivers (Chapman et al., 2004; Hegar & Rosenthal, 2009). Indeed, other work, though not explicitly focused on placement preferences, has recognized the value of kin and sibling relationships for maltreated children’s well-being following removal from home. Across studies, including those involving former foster youth, children often describe siblings and extended family members as vital sources of support, security, and continuity, and routinely report wishing that they had greater contact and connection with these family members while in out-of-home care (Barth, 2002; Chapman et al., 2004; Festinger, 1983). Children placed together in foster care or with kin caregivers also tend to demonstrate better adjustment than those separated from siblings or placed in nonrelative foster care, as evidenced by lower rates of emotional and behavioral problems, better mental health, and better academic performance (Chapman et al., 2004; Hegar, 2005; Hegar & Rosenthal, 2009; Herrick & Piccus, 2005; Keller et al., 2001; Leathers, 2005; Milojevich, Quas, & Adams, 2017; Rubin et al., 2008; Tarren-Sweeney & Hazell, 2005). Thus, for children in out-of-home care, kinship and sibling placements may confer a range of protective effects on their behavior and well-being.

Limitations and Future Directions

While the present study makes a novel contribution toward understanding young children’s placement preferences, limitations should also be noted. For one, the small sample size limited our ability to consider other potentially relevant influences on children’s placement preferences, including their length of time in out-of-home care, number of placement changes, type and severity of maltreatment, and contact with parents, siblings, and other relatives. This will be an important direction for future research, the results of which can advance understanding regarding the nuances of young children’s preferences. Furthermore, our sample size did not allow us to examine whether kin and sibling placement jointly influenced placement preferences. This possibility could be directly tested in future research involving the NSCAW data. Moreover, it would be valuable for such research to isolate children younger than 12 years and examine these interactions specifically in the age range of children generally not considered competent. This would provide additional insight into how the combination of kin and siblings, directly and across age, further shapes children’s preferences.

A final limitation to our work, which is common in other studies of children’s perceptions of their placement, concerns how more complex dynamics of children’s background directly and potentially indirectly affect their placement perceptions. One such background characteristic concerns children’s race or ethnicity, which can affect their preferences in at least two ways. First, children’s likelihood of being placed with a sibling or kin versus in a group or foster home setting varies by race and ethnicity: Compared with White children, Black children are 18% more likely to be placed in a group home rather than a family-based setting (Wiltz, 2018), which then leads to differences in whether the children even have the opportunity to be with siblings or kin. Second, even when Black or other ethnic minority children are placed in foster homes, their race and ethnicity at times does not match those of their caregivers (Coakley & Orme, 2006; Libby et al., 2006). It is unknown whether children’s placement preferences vary systematically in relation to racial matching, even though evidence indicates that children recognize race and ethnicity at a fairly early age in development (Quintana, 1998). By studying larger and more diverse samples, and by collecting data on children’s and caregivers’ race, ethnicity, and other background characteristics, more comprehensive insight into placement preferences in children could be gleaned.

Conclusion

Taken together, our findings indicate that concerns that children cannot provide useful input regarding their placement are largely unwarranted. Children can— and do—at times express desires to live with caregivers other than their maltreating caregivers and further seem to recognize the value of maintaining placements that preserve family connections with extended relatives and siblings. Thus, ascertaining children’s own preferences about their placement may well be an important and helpful practice when making determinations about the kinds of placement that would most appropriately serve children’s best interests and well-being. Furthermore, the finding that children legally presumed to be incompetent are capable of expressing reasonable preferences has potential implications for legal assumptions about children’s decision-making in other contexts, including custody post-divorce and medical decision-making. By allowing children to have some input and be a part of ongoing discussions about their own placement, children’s sense of involvement may well increase, thereby promoting their sense of justice (Fagan & Tyler, 2005) and feelings that they have a voice in what is arguably a complicated and potentially long-lasting legal case. Indeed, evidence from other work on legally involved populations indicates that feelings of being heard are an important and unique predictor of perceptions of legitimacy of the legal system, adherence, and feelings of fairness (Cascardi, Poythress, & Hall, 2000; Casper, Tyler, & Fisher, 1988; Mazerolle, Bennett, Davis, Sargeant, & Manning, 2013), separate from whether the decisions themselves are perceived of as favorable or not. As such, asking children about their placement preferences may confer a range of benefits, both in relation to the disposition of their case and in terms of their feelings and attitudes.

Biography

Kelli L. Dickerson is a doctoral student in psychological science at the University of California, Irvine. Her research focuses broadly on the effects of adversity exposure, including maltreatment, trauma, and legal system involvement, on children and adolescents. She is particularly interested in using empirical research to develop effective tools, programs, and policies to promote positive development in adversity-exposed youth.

Thomas D. Lyon is a developmental psychologist and legal scholar who has studied maltreated children’s competencies as witnesses for over 20 years. His career goal is to maximize children’s productivity as witnesses while minimizing error. He is past president of the American Psychological Association’s Section on Child Maltreatment (Division 37) and a former member of the board of directors of the American Professional Society on the Abuse of Children.

Jodi A. Quas is a professor of psychological science at the University of California, Irvine. Her work focuses on children’s eyewitness capabilities and abuse disclosure, consequences of legal involvement on children, and children’s coping with stress, maltreatment, and trauma. She teaches and conducts trainings across the globe on these topics and has written extensively on them in academic journals, technical reports, and edited volumes.