The Effects of Separation from Parents on Children

Abstract

More than a million children are separated from their parents by government authorities. We review the research on the effects of separation, including separation for parental maltreatment, adoption, illegal immigration, parental incarceration, and Indigenous status. The effects were universally negative, did not differ by reason for separation, and included neurological change from psychological trauma, precocious sexual maturity, physical and sexual abuse, neglect, academic delay, poor peer relations, psychosomatic symptoms, psychiatric disorder, and behavior problems such as depression, anxiety, delinquency, self-harm, substance abuse, and inappropriate sexual behavior. There were indications that separation led to intergenerational cycles of family separation. The research indicated that children are most vulnerable from 9 months to 9 years of age. We suggest that the negative effects of separation be considered explicitly when courts make placement decisions. Professionals who can reduce children’s suffering from separation are legislators, policy makers, social workers, mental health professionals, attorneys, and judges.

1. Introduction

Every year thousands of children are separated from their parents, mostly by government authorities. The outcomes include higher rates of family break-up, intergenerational cycles of family dysfunction and, for individuals, interpersonal dysfunction, psychiatric disorder, and high rates of incarceration. In this review of published research, we ask how many children are separated from their parents, the reasons for separation, and the effects of separation. Possibly the most striking finding of our review is that separation has largely been overlooked as a serious threat to children’s well-being. On the contrary is it widely used to protect children. We conclude with recommendations to reduce the need for separation and its negative consequences when separation is unavoidable. We seek to find a way to use professionals’ skills and society’s resources to bring greater safety and happiness, even joy, to children’s lives and those of their parents.

We note that, from professionals’ perspectives, separation that is accompanied by placement with a safe caregiver is not deemed threatening, with children’s distress being attributed to prior maltreatment rather than to separation. From children’s perspectives, however, loss of their parent is perceived as the absence of protection and, therefore, the maximum threat. Children describe being taken by strangers to live with strangers as kidnapping [1]. This difference in perspectives has led, we believe, to too little attention to the detrimental effects on children of separation from their parents.

We reviewed the literature for five reasons for separation: foster care, adoption, undocumented immigration, parental incarceration, and Indigenous status. The findings were that separation was harmful in and of itself regardless of the reason for separation. We identify younger child age and number of separations as having negative effects. In the concluding discussion, we focus on the emotional experience of separation from children’s words and drawings and suggest three actions that can be taken to better protect children from the harmful effects of separation from their parents. These actions include (1) renaming child protection as ‘family protection’, (2) providing in-home services, including creating a new para-professional role for community members to join families for long-term support, (3) consider of the negative effects of separation when courts make placement decisions, and (4) diverting funds from foster care, adjudication, and administration to family support.

1.1 Children’s symptoms of distress are consistent across all types of separation from parents

The negative effects of separation on children’s development are substantial and this has been known for a long time [2, 3, 4, 5, 6]. Notably, these outcomes are the same that result from child maltreatment, parental mental illness, traumatizing events, etc. That is, children display distress in several familiar ways that do not link stressors and signs of distress to specific eliciting conditions. The outcomes include poor mental and behavioral health as indicated by internalizing symptoms (e.g., depression, anxiety, withdrawal), externalizing behaviors (e.g., physical aggression, relational aggression, defiance, theft, and vandalism), and social and cognitive difficulties resulting in poor self-control, difficult peer relations, and poor school performance [7]. Separation from parents is also significantly related to symptoms of posttraumatic stress disorder, bi-polar disorder and schizophrenia [8] – even after accounting for prior level of trauma exposure [9, 10]. Separation from fathers is related to later violent offending [11].

1.2 How many children are separated from their parents?

In 2018 an estimated 263,000 U.S. children were separated from their parents and placed in foster care [12], with similar placement numbers in 2019 and 2020 [13]. About 10% of separated children were returned home within 30 days [14], suggesting that these separations were unnecessary. Canada does not report national rates of foster care, but the U.S. rate of .69% of the population was compared to an estimated .92% for Canada [15]. Among western nations, the United Kingdom has the highest percentage of children in care, with four concerning processes: (1) the number of families being investigated is rising, (2) the number of children in care is rising, (3) the number of children being returned to parents is dropping, and (4) the number of adoptions is dropping [16].

In sum, more than a million children are at risk of these detrimental outcomes of separation in western, English-speaking nations alone, with minority children and children from non-western countries being at higher risk of separation and its detrimental effects. Any condition that puts millions of children at developmental risk should receive attention, especially when endorsed and implemented by public authorities.

2. Method

This chapter summarizes our comprehensive review of the international literature on the effects of separation on children. The review references 242 citations (192 individual papers, books, or studies, 10 meta-analyses, 25 systematic or narrative reviews, and 15 policy documents). The full review can be downloaded from: The-effect-of-Separation-from-Parents-on-Children-by-Crittenden-Spieker.pdf (familyrelationsinstitute.org).

3. Findings for five groups of separated children

For five types of separation, we (a) summarize our findings, (b) make recommendations for improving children’s well-being, and (c) state some unanswered questions that need research. At the end of the chapter, we make four over-arching recommendations. We note that many of our recommendations are consistent with those made by others.

3.1 Foster Care

Foster care is based on the belief that foster parents will provide safety and improve the development of fostered children in ways that their parents cannot. This belief is not supported by research. Every review and study indicated that fostered children developed less favorably than comparable maltreated children who remained with their parents. The negative effects were greater for children who had changed home more than once; respite care, additional foster placements, reunification, and adoption counted as additional changes of home. The most harmful ages at which to separate children from their caregivers are 9 months to 9 years; this is the age period when the most child-parent separations occur. Although the effect has not been studied, parents also are separated when children are put in foster care and this might adversely affect their parenting during contact or after reunification. Although some children are injured or killed by their parents, others are harmed by foster parents, including higher rates of sexual abuse than with biological parents [17]. Kinship care both reduces these problems and has lower rates of permanent placement [18, 19]. Foster children whose own parents had been fostered were less likely to achieve permanency than foster children whose parents had never been in foster care [20]. For Black children, the disproportionate use of foster care, following the historic family breakup of slavery, was one strand of systemic government policy that weakened Black family structure across generations [21]. These findings suggest that foster care may have intergenerational impact, and that separation in childhood might affect adults’ ability to form stable, nurturing relationships in adulthood.

Overall children in foster care were in poorer mental and behavioral health compared to children in every other family type and to children in low-income families, both prior to and after placement in care [22, 23]. The negative effects include psychological trauma, neurological change from psychological trauma, precocious sexual maturity, academic delay, poor peer relations, psychosomatic symptoms, psychiatric disorder, depression, anxiety, delinquency, self-harm, substance abuse, inappropriate sexual behavior, and all types of maltreatment while in care.

Rather than questioning the advisability of separation per se, professionals have focused on procedures to identify the few children at very serious risk, rather than on preventing separation and its negative effects for the majority. In fact, separation has not been identified by authorities as a threat to children’s well-being and is not considered in court decision-making processes [24, 25, 26]. We think it should be.

3.1.1 Recommendations for foster Care

Child placement is a complex, systemic condition that is necessarily distorted if simplified and reduced to dichotomous choices. Many of the problems seem tied to: (1) misunderstanding of children’s attachment and psychological trauma from separation, (2) a dichotomous either/or approach to solutions, (3) the adversarial court process, and (4) a focus on individual children rather than family relationships. Our recommendations attempt to re-envision these complexities in ways that can help children and can be implemented under real world conditions. Thus, instead of trying to decide which family can best raise a child, the goal could be finding a way for the all people who love a child to contribute to his or her well-being.

To fund these ideas, we recommend that funds be shifted from foster care to training a more diverse child welfare workforce, including men, so that families are more often be evaluated by professionals with whom they can connect on the basis of shared identity.

3.1.2 Training in attachment and psychological trauma

Child protection personnel (including attorneys, guardians ad litem, and family court judges) should have training on attachment, including learning when attachment develops, its enduring quality, its hierarchy when there are changes of protective figure (i.e., biological, foster, kinship, and adoptive parents), its protective function in each child’s specific circumstances, and individual differences in attachment. Child protection personnel should know that some “problem” behaviors, particularly inhibition, avoidance, and defiance, are indicators of being attached [27].

Assessment of potential foster and adoptive parents should include exploration of their own experience with separation and loss and its effects on them as adults. The results should influence the support that foster parents are given. Particular attention should be given to differentiating the effects of separation from children’s presumed trauma at seeing their parents during contact/visitation.

The long-term harmful emotional and developmental costs to children of separation from any parent figure (biological, foster, kinship, or adoptive) should be added explicitly to the factors to be considered when courts make placement decisions.

3.1.3 Increasing ‘both/and’ solutions

Many problems in foster care and adoption are tied to a dichotomous ‘either/or’ model in which children are placed exclusively in one home or the other. This puts the potential parents at odds with each other and forces the children into the dangerous middle ground of needing to appease opposing parents. Our recommendations are meant to encourage a fluid range of shared parenting, for the benefit of both children and parents.

To prevent separation and foster placement, child needs that are neglected should be provided by service personnel (e.g., housekeepers, money managers, child care, educational stimulation, etc.) when the parents cannot manage to meet the need; this should be done even if the parents do not seem deserving – because the children are deserving and prevention of traumatic separation is a higher priority than teaching the parents skills. A greater variety of parenting services, flexibly offered by someone whom the parents respect and trust, are needed to fill the gap between group instruction such as parent education (for the worried well) and individual psychotherapy (for parents with personal traumas). Individualized home visiting services are particularly effective [28]. Having such services or creating bespoke services could prevent separation of children from their parents (e.g., [29, 30]) and the ensuing separation trauma for children and their parents.

Foster parents who seek to support vulnerable families should be preferred over those who seek to raise other parents’ children. Foster parents’ need to replace losses that occurred in their own childhoods should be assessed. If reunification is probable, foster parents who want to foster a family should be preferred.

Children should never be placed solely for assessment or foster parent respite; doing so adds new trauma on top of existing trauma.

The conflict around contact could be reduced by having biological and foster parents share some contact visits and having adult visits in which the parents share ideas about themselves and the children. It is especially important that foster and adoptive parents hear the history of troubled parents from the parents because compassion is founded on shared information. The model could be that of an extended family with troubled family members. Both kinship care and open adoption (see next section) provide tested models for biological and foster parents to work together to improve outcomes for children.

Most important of all, family strengths should be identified and enumerated explicitly. This is especially true for minority children whose families are sometimes evaluated negatively by majority professionals. Strengths are the base upon which to build a family-specific support plan.

3.1.4 Prioritizing non-adversarial approaches

Teamwork needs to replace adversarial approaches. One example is the UK’s use of jointly instructed assessors [31]. Such shared information and consensus planning could reduce reliance on biased adversarial processes.

Child protection records should indicate the source of information about children’s behavior so that video-recorded behavior, professionally observed behavior, and behavior reported by foster parents can be evaluated separately. This can reduce discrepancies between apparently conflicting observations by identifying biases and personal interest.

Children’s transitions between households should be made gradually and, whenever possible, without termination of relationships. For example, when placement is needed, but the immediate situation is not in eminent crisis, the biological and foster families should visit each other, have overnights, then extended overnights until the children are living with the foster parents and visiting their biological parents. Children should leave some of their toys, blankets, and clothing in the foster home. Both sets of parents should be seen by the children in both homes. The same process should be undertaken in reverse for reunification. This process should continue for as long as the child maintains an emotional connection to the less frequently seen parents. The families, in other words, become like members of an extended family. The outcome should be that the child is loved by more people and loses access to no one.

3.1.5 Emphasizing relationships and interpersonal systems

The entire family should be assessed; too often non-problematic individuals are over-looked even when they could clarify the nature of the problem and contribute to successful solutions. Biological parents should receive counseling around their own trauma of being separated from their children. Foster parents might need this as well if reunification is planned or if they have had significant separations in their own history.

Children’s perspectives on placement and placement changes should be sought. Professionals and parents should be prepared for children to express intense, conflicting, and changing perspectives. Children should be helped to articulate their feelings, including especially mixed feelings, in readily understood ways, such as being enacted, spoken by toy figures, or illustrated with drawings. Seeking and actively valuing children’s perspectives does not imply that these are given priority over the broader perspective of adults. But children’s perspectives should be acknowledged and respected and children should be given reasons for decisions in terms that they can understand at their age.

In reunification, biological and foster parents should be helped to predict problems (such as children’s avoidance and rejection). Engaging the foster parents in both discussion and travel between homes can help to stabilize this inherently disruptive process.

When children have been separated from family members or culture (including racial, language, and national groups), effort should be made to include the actual lost parent, siblings, and culture/racial group; if that is not possible, their importance should be discussed, and substitutes should be sought. Additional parent figures who share the child’s culture should be sought and supported financially. Baby boomers, who have raised their own children, can become parent/grandparent figures and single younger adults without children of their own can become ‘aunties’ and ‘uncles’. The focus would be on expanding families for children, and not separating them from any family.

3.1.6 Supporting families with long-term para-professionals

Many parents do not benefit from educational services. After enough services have been tried and the parents have failed to make changes, foster placement is often sought, even if the risk of permanent injury or death is low. When everything has been tried and nothing has worked, parents and professionals can feel defeated. When foster care is chosen as the solution, the children pay the price for what their parents and the professionals could not do.

Instead, we suggest that some parents be deemed ‘supportable’; they and their children should be supported without expectation of parental change [32]. Non-educational services should be offered, again without expectation of change. A particularly useful service that has become less common as child protection has become more professionalized is volunteer family work. Especially when these para-professionals live near the child’s family, they can become like acquired family who remain long after the allocated service time is reached or the case is closed. Because such family workers do not have 9–5 hours and can come when needed for indefinite periods of time, they can both meet family members’ needs and also reduce professionals’ frustration at having to leave families they have come to care for. Grandparent-aged para-professionals would be ideal and there are many in the retiring generation of baby-boomers.

Several foster care issues require further research. How should individual differences among foster children be considered to customize planning for each family (rather than applying the ‘evidence base’ to everyone)? How should foster parents’ sensitivity and nurturance to fostered children, the accuracy of their reports on children, and their own history of personal loss and trauma be used to inform decision-making? Similar research on the experiences and biases of professionals might be warranted.

3.2 Adoption

Adoption seeks to place children in permanent homes that will foster their development when their biological parents cannot do so. Again, the issue is whether adoption achieves that goal.

3.2.1 Outcomes when children are adopted

About half of fostered children are either adopted or in permanent guardianship with kin although percentages vary across localities [33]. About 3% of adoptions fail, with others having significant problems [34, 35]; the reasons include child behavior problems, mental health issues, and cognitive disability [33]. Factors promoting stable adoption and fewer child developmental problems are young age at adoption, fewer prior placements, and less severe early maltreatment [33, 36, 37]. Children over 4 years of age at adoption were at much greater risk of failed adoption and adoptions failed most frequently at puberty [38]. Adoptive parents’ reports of child distress about moving from the foster to adoptive homes was related to poorer outcomes [38, 39]. Adoptions by foster parents were more successful than adoptions that required separation from foster parents.

Inappropriate sexual behavior and self-harm were associated with multiple prior separations [38]. Sexual behavior probably functioned to initiate attachment quickly, but was misunderstood by adults and could lead to rejection and further separations [40]. Adoptive parents and their children understood adoption breakdown differently. Parents thought the children had problems based on prior experiences in other homes [41, 42, 43] whereas children said that they had not been listened to, did not want to be adopted, and wanted more contact with their biological parents. The functioning of adoptive parents is inadequately documented, but a quarter to a third might have problems that contributed to adoption distress [38]. Adoption breakdown at puberty and the finding that about a third of adoptive parents were confused about their reasons for adoption [44] suggests that some adoptive parents had not resolved their desire for biological progeny. Caregivers’ psychological distress from prior loss of children by death, separation, or lack of fertility destabilized some adoptions [38]. Expanding Selwyn and her colleagues’ recommendation, we recommend policies that encourage both children and parents to be in contact for as long as they wished. Open adoptions is a promising possibility [45, 46, 47].

In general, adoption was most beneficial when children were placed in early infancy and not moved thereafter. Put another way, the fewer separations children experienced, the more successful their adoptions were. A similar finding might be true for adoptive parents (regarding their prior losses and separation). The research supports two recommendations, both of which promote having children and caregivers experience fewer separations:

1. Foster parents should be given priority when a child becomes available for adoption. Subsidies should be used when the foster parents love the child but cannot afford to adopt him or her.

2. Adoptive parents should encourage children to maintain prior relationships for as long as they are important to the child.

Further research is needed on adoptive parents’ history of trauma and possible mixed motivations for adopting. The goal is to select caregivers with greater awareness of their vulnerabilities so as to address these in helpful and preventive ways.

3.3 Children separated from their undocumented parents

Over 100 countries detain migrating children, including refugees and families seeking asylum [48]. This has been strongly criticized by a wide range of pediatric, medical, and psychological professional associations (e.g., [49, 50, 51, 52, 53]) because the effects on children are very negative. These effects include physical, behavioral, mental health, and academic problems immediately upon or soon after detention [54, 55, 56, 57, 58, 59, 60]. Some symptoms, such as posttraumatic stress, remained long after reunification with parents [58]. Short separations were as distressing to children as long separations, thus, separation per se was dangerous to children’s well-being [61]. Particularly harmful separations involved separation while the child was sleeping or at school, separation of siblings, and separation without explanation or when the child was too young to understand explanations [62]. This is consistent with the finding for foster care that the harm was greatest when the children were between 9 months and 9 years of age. The most important recommendation is not to separate children from their parents. Laws that unintentionally increase child-parent separation should be modified to reduce separations.

3.4 Parental incarceration

Incarceration of parents almost always results in child-parent separation, with most children being under child protection. Parental incarceration, especially maternal incarceration, produced profound negative effects including poor physical, behavioral, and mental health [63], but these effects cannot be disentangled from pre-existing family conditions. An even more concerning problem is systemic intergenerational processes that increase the probability of both child maltreatment and parent incarceration in the next generation [64]. Black, Hispanic, and Indigenous children suffered greater effects than White children.

There are too few studies and too many correlated negative conditions to draw clear conclusions about causal pathways, but that might not matter once an intergenerational cycle of broken relationships begins. Social and political policies often reinforce the intergenerational process, making it difficult for individuals to break the cycle. This suggests that the causal process operates beyond the individual level, even beyond the family level, and within the sociopolitical ecology of culture. As with separation from undocumented parents, solutions must be developed and executed by local, regional, and national legislators.

3.5 Separation of children from indigenous parents

Indigenous children have been systematically separated from their parents in many cultures, including American Indian tribes, First Nations bands, Australian Indigenous people, Nordic Sami and Kven populations, and Uyghurs in China. In most cases, separation from parents was accompanied by institutionalization, that is, no replacement caregiver at all. This combination produced very deleterious outcomes: widespread sexual abuse of the children [65], high rates of behavioral and mental health problems [66], poor physical and mental health [67], physical and emotional maltreatment in care, and often many of these, that is, poly-victimization [68]. Even after reunification, the children were more likely to miss school without permission, be treated unfairly at school, have poor physical and mental health, and be less likely to be living in a home owned by a family member. Moreover, the detrimental effects were passed to the next generation of children, indicating multi-generational effects of transferring trauma, family instability, violence, and poverty into at least the third generation [69]. Put another way, enforced separation of Indigenous children from their parents stands out as combining all the harmful aspects associated with separation, including enforced separation from loved parents, separation from both parents, lower socioeconomic status, relatively high levels of institutional child maltreatment, frequent incarceration of fathers, absence of any alternative attachment figure in residential schools, separation from non-parental kin, denigration of children’s ethnicity, refusal to permit children to speak their language, maltreatment in the care setting, and sometimes even loss of personal identity and family origins. Family separation created conditions of trauma, with substantial long-term effects [70]. The experience of the descendants of enslaved people in the United States suggests the possible extent of these effects.

Despite forced separation of children to institutional facilities having been discontinued for several decades, child protection authorities continue to place Indigenous children in foster care at higher rates than other children. In part this is because the long-term effects of separation include higher rates of alcohol misuse, familial violence, incarceration, and child neglect. Without substantial and multi-generation support, these costs of separation will continue within Indigenous communities. The goal should be to provide Indigenous children the advantages of both their Indigenous culture and those of the modern dominant culture.

4. Conditions affecting outcomes from separation

Separation of children from their parents harms children. This is universally true. The precise effects vary by age, culture, and socioeconomic status, but there are no conditions under which separation does not cause distress. Three conditions affect the extent of harm to children.

4.1 Special status

In most cases minorities and boys are affected more than children in the dominant culture and girls, in terms of both the proportion of children removed from families and the severity of negative effects. This ‘minority and male gender’ effect contrasts with the white, middle class, and female status of most child welfare professionals in western, high-income countries. This dichotomy almost certainly reduces the effectiveness of services.

We recommend that professional staff, at both entry and policy-making levels, reflect the ethnic, cultural, and gender status of their clients.

4.2 Age at separation

Child age at the time of separation is a major factor in children’s response to separation from their parents.

When the first placement occurs before 6–9 months of age, the effects tend to be transient [71]. If the infant is reunified or permanently placed with the new caregiver by 8–9 months of age, the effects tend to be minimal. Nevertheless, that does not prevent children having concerns at later ages when they are told about the separation. Separations occurring between 3 and 5 years of age usually result in persistent and enduring loss of security in new relationships [72]. The behavioral signs include increased need for attention, clinginess, temper tantrums, defiance, appetite changes, nightmares/sleep problems, and sadness [72]. Possibly most concerning, brain development is affected by stress; for example, early childhood stress was associated with faster maturation of the prefrontal cortex and amygdala during adolescence, that is, roughly a decade later [73, 74]. This is important because brain pathways activated early in life lay the foundation for later development [75, 76]. Although separation from parents in mid-childhood (when language can be used to explain the situation to children) has fewer negative effects than earlier separation [11, 77], it is not without negative effects, including various forms of acting out, somatic distress, and internalizing behavior. Puberty changes neurological functioning in ways that (a) prioritize short-term high risk/high reward functioning over thoughtful, rational functioning, (b) generate sexual desire and (c) increase strength, especially in males. These changes give maladaptation greater implications for the adolescents’ well-being and survival, and for the well-being and safety of others. Put another way, risk in adolescence can have serious negative outcomes, including physical harm and death, for both the adolescent and for others. On the other hand, older adolescents begin to look to their future beyond childhood; when opportunities are available, they can create ways to change a developmental pathway. Thus, both the risks and the opportunities for self-protection are greater in adolescence than in childhood.

A central problem caused by children’s immaturity is that they are unable to communicate precisely about their experiences and, thus, unable to correct their own or adults’ misunderstandings. This is particularly true when children use inhibitory strategies with little or no obvious display of their distress.

4.3 Number of changes

The negative effects of separation increase as the number of separations increases [11, 78, 79, 80]; this is true even when the next placement is ‘better’, for example, in adoption. Notably, desire for comfort by emotionally deprived children is sometimes displayed as inappropriate sexualized behavior, that is, ‘indiscriminate’ attachment [81].

Changes of caregiver are consistently harmful to children. When the change is intended to stop maltreatment, it is not clear whether the cost of separation offsets the cost of maltreatment; this, of course, varies in each case, but the cost to children of experiencing another separation is often overlooked in the effort to prevent further maltreatment. Although young children can attach to a new caregiver [82], this ability diminishes with each separation and subsequent new caregiver [83]. Once impermanence becomes an individual’s primary model, an intergenerational cycle of broken relationships might be initiated.

The effects of separation vary from one child to another and vary somewhat by age, gender, frequency of separation and reason for separation, but they consistently outweigh the effects of other influences on child well-being [84].

5. Limitations

The limitations of this review reflect the limitations of the research. Ethical standards preclude the randomized designs necessary to isolate causal factors or compare equivalent groups. Most separations occur in the context of other adversities which put children at risk prior to separation, making it hard to isolate the effects of separation [84]. Nevertheless, two studies suggest the added effect of separation. Carr-Hopkins and her colleagues [1] found that the removal itself was experienced by 6–12-year-old school-aged children as more traumatizing than the maltreatment that had preceded it. Crittenden observed that maltreated children denied placement in protective daycare remained with their parents longer than children given the ‘advantage’ of daytime separation [85]; in this study, daytime separation was presumed to function as a catalyst, speeding a process of relationship rupture that ultimately occurred in both groups. A common limitation is a lack of exploration of within-group differences. Further, most research on separation relies on professionals’ evaluations or self-report by caregivers. However, self-reports may be distorted [86]. Pre-post separation designs would be helpful, but we found few. Despite these limitations, our conclusions on the harmful effects of separation on children’s development are unambiguous.

Possibly the greatest limitation to our review is the near absence of children’s perceptions, understandings, and feelings.

6. Finding the voices of children

The conclusions are obvious. Separation from people you love, whenever and however it happens, is painful and leaves an indelible scar on one’s mind and development. It has accompanying risks that can be measured. But none is as basic and universal as the separation itself.

We have tried to stay focused on empirical findings, but these almost entirely reflect adult perspectives. Our central point is that separation is suffused with children’s suffering. For example, 34 years after being separated from her parents, Sandy White Hawk said she “remembers the day in 1954 when she was taken away by missionaries from the Rosebud Sioux Reservation in South Dakota. Standing in a red truck beside the stern woman who would become her adopted mother, the toddler gazed up at a pale white arm so different from any arm that had hugged her before. The 18-month-old wouldn’t see her American Indian family for the next 34 years.” [87, 88]. The pain White Hawk suffered remains palpable three decades later and is made vivid to all of us in her description of the moment of separation. That it was carried out by presumably well-intended missionaries does not reduce the pain, nor render the policy it reflected justifiable.

Although many maltreated children are protected from parental maltreatment by placement in foster homes, all suffer greatly from separation. Children are not usually asked if they want to enter foster care, but occasionally their voices can be heard and, like White Hawk’s, they scream of pain and the harm done by separation. For example, one boy described his anger at being separated from his foster parents and placed with adoptive parents as “like a volcano when it erupts” ([38], p. 234).

As a part of our court reports of fostered children, we have regularly asked fostered children to “draw your family”; universally they draw their biological family. Heptinstall and her co-authors [89] noted the importance of their biological families to fostered children, even when they did not know their family members. This was true in the drawing by a 13-year-old ‘Restless Nellie’ [90] who had not seen her biological family since she was 3 years old and had lived in her adoptive family for only 4 years. In her colorful drawing everyone is smiling, and the effect is happy (Figure 1).

Figure 1.

Restless Nellie’s family drawing.

‘Me’ is placed in order within the intact biological family – that includes stepsiblings whom she had not met, and a child born after she was placed in foster care. The parents are identified by their first names, not ‘mom’ and ‘dad’. No one is touching anyone else, suggesting lack of connection. Their arms and legs are shortened and lack hands and feet, suggesting lack of agency. There is no sign of any foster or adoptive family. In sum, at 13 years of age, the girl delusionally idealized her biological family, no longer acknowledging the abuse and neglect that led to her being removed from her parents’ care. Although no pain is visible, the false portrayal of unrealistic happiness hints at ‘unspeakable’ pain.

Below is an evocative exception to the drawing of biological families, drawn by a 9-year-old boy who had been in multiple kin and stranger placements over the previous 2 years (Figure 2).

Figure 2.

The loneliness of being a foster child in multiple placements.

The absence of any family, the smallness of the boy on the empty expanse of paper, the absence of arms (that could indicate agency or connection to others), the motionless pose, and the use of black (when bright colors were available), plus the words ‘Do not know’ as his answer to ‘What are you doing?’ suggest the depth of his feeling of being abandoned by adults. Indeed, what is he doing in this home with strangers?

7. Recommendations: a broad call to action

Despite reviewing hundreds of documents about children being separated from their parents by governmental authorities and by parents themselves, we did not find even one study that identified ‘separation’ as a contributor to children’s distress. Instead, the focus was on parents’ behavior, children’s needs, national policy, and evidence of children’s pathology. We did find evidence that child deaths led to more stringent rules that protected professionals by both increasing the number of families investigated and reducing the threshold at which foster placement was initiated.

The signs of children’s distress that can be seen, heard, and felt are so many and often so extreme that it is hard to understand how separation has not become a major focus of research and clinical practice. Nevertheless, children’s perspectives are rarely mentioned. We offer four primary recommendations to reduce children’s suffering from separation from their caregivers while also creating the possibility of safety, happiness, and possibly even joy for distressed children and their families:

1. Change the term ‘Child Protection’ to ‘Family Protection’ to emphasize the importance of family well-being to child safety and well-being.

2. Offer vulnerable families practical services to meet children’s needs without requiring that parents change to deserve the services and without separating children from their parents. When additional adults are needed, add them to the family rather than taking children out of the family [91, 92]. Using para-professional adults in the neighborhood as ‘protective kin’ can help to meet this need.

3. Consider the harmful effects of separation when making decisions that could separate children from their parents.

4. Shift resources from foster care, court litigation, and administration to direct services to families and neighborhoods.

These four solutions change the ‘either/or’ solutions used now to ‘both/and’ solutions in which children would be protected by having more loving adults, more resources, and safer neighborhoods. Given the number of adults living alone and baby boomer parents whose children have left home, we think there are community adults who could become protective kin to troubled families – if we learn to look for them in new ways and to support them with reallocated funds reallocated. Put simply, shifting our human and monetary resources from disputed separations managed by professionals to enriching troubled families could lighten their burden, engage adults who seek meaningful family connections, and bring satisfaction to the professionals who sought to use their working lives to help families.

Our recommendations are very low tech. They eschew programs in favor of helping relationships. They shift money from fostering children to fostering families. They minimize conflict by focusing everyone on collaborative effort to strengthen families. We have argued that separation hurts children, but it also hurts biological parents [93, 94] and child welfare professionals [95]. As Bowlby said long ago, we believe that “the better approach is ….to seek to rehabilitate the home and family, no matter how costly in time and effort such an attempt may be” ([96], p. 5).

Our recommendations require reversing a half century of protecting children from their parents to a unified and shared effort to protect families. Design and implementation of these changes should include the contributions of parents and advocacy groups of parents with child protection experience [21]. After half a century of increasing numbers of separations, enlarging bureaucratic systems, and escalating costs, maybe a new approach is worth trying?

We hope that home-grown programs, tied to the needs and people in specific communities and containing feedback to inform program improvement, can improve outcomes to children and families and increase job satisfaction for professionals. Participating in something new generates commitment and hope. Hope can be contagious. We hope that a more compassionate approach to preserving families, particularly the most fragile families, will change the futures of children, their families, and their descendants in future generations.