Downloadable Files


Conflicting interpretations of the Convention on the Rights of the Child (CRC) and Convention on the Rights of Persons with Disabilities (CRPD) send mixed messages on the safety and legitimacy of residential care, resulting in the replacement of large institutions with smaller ones often called ‘residential care’ or ‘group homes.’ The CRPD requires governments to create protections and supports to allow all people with disabilities to live in the community. CRPD Committee General Comment No. 5 says that, for children, living in the community means growing up in a family – not in a large or small facility. This article demonstrates how the family inclusion mandate of General Comment No. 5 is rooted in the ‘human rights model of disability,’ fundamental to equal protection under the CRPD for all children with actual or perceived disabilities. The article proposes solutions to ensure full implementation of both the CRC and CRPD.

Keywords: Human rights, disability, CRPD, CRC, residential care, group home



The Convention on the Rights of the Child 1989 (CRC) was groundbreaking as it was the first international convention to recognize the rights of children with disabilities, but disability rights and children’s rights experts have raised concerns about weaknesses in CRC protections that reflect paternalistic and outdated assumptions about disability (Kilkelly, 2002, pp. 119-120; French and Kayes, 2008, p. 13; OHCHR European Regional Office, 2011 , pp. 10-11; Bantekas, 2018, p. 201; Rosenthal, 2019, pp. 71, 113-114). The UN Convention on the Rights of Persons with Disabilities (CRPD) was adopted in 2006 to ensure equal protection for people with disabilities – and to fill gaps in existing protections under existing international human rights conventions (French and Kayes, 2008, p. 20; Kanter, 2015, p. 5). Improving protections for children with disabilities[1] was one of the areas targeted by the CRPD’s drafters (see UNGA, 2004, p. 35). (See also Skarstad and Stein, 2018, p. 5 for discussion.)

One of the most important contributions of the CRPD is that it protects the right of people with disabilities to full inclusion in society. In relation to the rights of children with disabilities, this right to full inclusion has been interpreted as prohibiting the placement of children outside the family in large or small residential facilities (UN Committee on the Rights of Persons with Disabilities (CRPD Committee), 2017, para. 37; Rosenthal 2019, pp. 65-137). These new human rights standards require broad changes in domestic policies, professional practices, and international guidelines which have long been built on the assumption that some residential treatment is necessary as part of any child protection system. While grounded in specific CRPD rights, the full measure of these new protections can only be understood through a broader respect for what has been called the ‘human rights model’ of disability that underlies the entire convention (Degener, 2015, pp. 31-50).

The CRPD’s approach is shaped by an ‘evolving’ understanding of disability as a social construct. The CRPD describes disability as the product of the ‘interaction between persons with impairments and attitudinal and environmental barriers that hinders their full and effective participation in society on an equal basis with others’ (CRPD, preamble (e)). The new framework shifts from an individual/medical model to a human rights model (Degener, 2017, pp. 31-49; Rosenthal, 2019, pp. 76-77). Under the ‘medical model’, a person’s ‘deficits’ or medical problems are located within the individual. Where a medical cure for the individual is incomplete, the remaining impairments have been seen as a legitimate ground for restricting or denying rights (CRPD Committee, 2018a, para. 8). Under the human rights model, society must become fully inclusive of all people with actual – or perceived – impairments (CRPD Committee, 2018a, para. 7-11). 

The medical model is often associated with the perception of impairment as a personal tragedy to be pitied, protected, or fixed on an individual level, so it is referred to here as the ‘medical/tragedy’ model of disability. The human rights model, in contrast, emphasizes social action to create legal, policy, or attitudinal changes to make the community fully accessible to all. The human rights model includes a participatory element, recognizing the right of people with disabilities to make choices about their own lives (CRPD, art. 12) and to contribute to ‘the development and implementation of legislation and policies’ to implement rights recognized under the CRPD, art. 4(3)). The CRPD has been described as ‘a paradigm shift’ in the protection of rights for people with disabilities (MacKay, 2007, p. 328; French and Kayess, 2008, p. 20).

This article focuses on a contemporary international controversy about the legitimacy of any placement of children in residential care. In 2018, the CRPD Committee and the UN Committee on the Rights of the Child (CRC Committee) established a Joint Working Group (JWG) to discuss ‘new ways and means of reinforcing the coherence of both Committees’ jurisprudence on children with disabilities’ (CRPD Committee, 2018c, para. 22). While agreeing on many points, the JWG has struggled to reach an agreed understanding about residential care. The adoption of common guidelines would be an invaluable step forward – as long as they reflect the full measure of protection promised by both conventions.


Conflicting interpretations of the CRC and CRPD

Both the CRC and CRPD recognize the importance of the family and provide protections for family life (CRPD preamble para. x and article 23, CRC preamble and articles 8, 9, 10, 16, 20, and 22; see discussion in Rosenthal, 2019, pp. 68-69). But CRC art. 20 permits placement outside a family if the child has no family of their own or if they face abuse and must be removed from the immediate family. The child who is ‘temporarily or permanently deprived of his or her family environment’ is entitled to ‘special protection’ and placement in ‘alternative care’ (CRC, art. 20(1-2)). Alternative care may include ‘inter alia, foster placement, kafala of Islamic law, adoption or if necessary, placement in suitable institutions for the care of children’ (CRC, art. 20(3)). When the CRC was drafted it was widely perceived that ‘some children will always need institutional care’ (Parliamentary Assembly of the Council of Europe, 2005 cited in Cantwell and Holzscheiter, 2008, pp.).

During the thirty years since the CRC was drafted, as scientific evidence has revealed the dangers of institutional placement for the physical and emotional development of the child, there have been a series of efforts to limit placement from both children’s rights and disability rights authorities. The CRC Committee adopted General Comment 9 (2007 para. 47) which made it clear that institutional placement should be understood as a ‘measure of last resort.’ Further, the UN Guidelines for the Alternative Care of Children (UN Guidelines)(UN General Assembly (UNGA), 2010, para. 21-23, 29) – a non-binding but highly regarded set of professional standards intended to help governments improve care consistent with the CRC and evolving professional standards (Cantwell et al., 2012, p. 20) – also stressed this point. These influential Guidelines call for the ‘progressive elimination’ of institutions (UNGA, 2010, para. 23). They take a strong stand on protecting family life, calling on governments to create programs to address the ‘root causes’ of child abandonment (UNGA , 2010, para. 32).

Nevertheless, despite taking this strong stand against large institutions, the Guidelines preserve smaller institutions by calling on governments to maintain a system of what they call ‘residential care’ facilities, to be professionally staffed and regulated (UNGA, 2010, para. 23). They define ‘residential care’ as including ‘long-term residential facilities’ (UNGA, 2010, para. 23). By encouraging the preservation of residential facilities the Guidelines effectively allow for continued placement in smaller institutions where children may grow up without a family.

Decades after the adoption of the CRC, the United Nations embraced a stronger stance against institutions when it adopted the CRPD. Article 19 of this convention establishes a right to community inclusion of ‘all persons with disabilities.’ The protection is universal in its commitment to integration as it applies to all people with disabilities – no matter the type or severity of their impairments (CRPD Committee, 2017, para. 20-21). Article 19 does not contain any test or limitation on who qualifies for the right to community integration, and the right does not depend on what kind of services the government can or cannot provide. General Comment 5 states that the ‘level of support’ needed by a person may not be ‘invoked to deny or limit the right to independent living’ for persons with disabilities, nor may the cost of care or support in the community be used to justify any limitation of the right of the individual to be part of the community (CRPD Committee, 2017, para. 20, 21).

Article 23 of the CRPD provides further guidance as to what to do when care within the immediate family is not possible. Article 23 (5) states that, ‘where the immediate family is unable to care for a child with disabilities [States Parties shall] undertake every effort to provide alternative care within the wider family and failing that within the community in a family setting.’ No alternative is countenanced. If kinship care in the ‘wider family’ is unavailable, the CRPD still requires placement in a ‘family setting’ (CRPD, 2007, art. 23(5)).

General Comment 5 clarifies that ‘[f]or children, the core of the right to be included in the community entails a right to grow up in a family’ (CRPD Committee, 2017, para. 37). The CRPD Committee (2017, para. 16(c)) explains that:

Large or small group homes are especially dangerous for children, for whom there is no substitute for the need to grow up with a family. ‘Family-like’ institutions are still institutions and are no substitute for care by a family.

In her 2019 report on Persons with Disabilities Deprived of Liberty, former UN Special Rapporteur Catalina Devandas points out that General Comment 5 has implications for the definition of an ‘institution.’ Since all children have a right to grow up in a family, ‘[a]ny placement of children in a residential setting outside a family must be considered placement in an institution’ (UN Human Rights Council (HRC), 2019, para. 19).

CRPD General Comment 5 does not define ‘family.’ The CRC’s General Comment 7 explains that ‘family refers to a variety of arrangements that can provide for young children’s care, nurturance and development, including the nuclear family, the extended family, and other traditional and modern community-based arrangements’ (CRC Committee, 2005, para. 15). To be consistent with CRPD General Comment 5, ‘modern community-based arrangements’ should not be understood to include family-like residential homes. Based on what scientific research tells us about the developmental needs of children, a more nuanced understanding of family is needed, i.e. a living arrangement in which a child can form stable emotional connections with a committed adult caregiver (Dozier, et al, 2014, pp. 220, 223). Stable family relations may include a child’s biological family, placement in the ‘wider’ family among relatives (‘kinship care’), placement in a stable foster family, or adoption (Dozier, et al, 2014, pp.223). This understanding of a family is consistent with the CRC Committee’s view that ‘in practice family patterns are variable and changing in many regions, as is the availability of informal networks of support for parents, with an overall trend towards greater diversity in family size, parental roles and arrangements for bringing up children’ (CRC Committee, 2005, para. 18).

Based on the new language of the CRPD, the former Special Rapporteur on Disability, Catalina Devandas, has observed that the UN Guidelines are, on this point, no longer ‘in line’ with the requirements of the CRPD (UN Web TV, 2019). She stated that the UN Guidelines ‘should be reviewed under the higher standards upheld by the [CRPD]’ (HRC, 2019, para. 51). The Special Rapporteur also called into question placement in ‘suitable institutions’ under CRC article 20:

The notion of ‘suitable institutions’ under article 20 of the Convention on the Rights of the Child and the Guidelines for the Alternative Care of Children should be reviewed under the higher standards upheld by the Convention on the Rights of Persons with Disabilities. As article 41 of the [CRC] Convention recognizes, its implementation should not affect any provisions of international law that are more conducive to the realization of the rights of the child (HRC, 2019, para. 51).

Devandas’ reference to CRC article 41 highlights an important opportunity for bringing about harmonization between the conventions. Under article 41 of the CRC, any provision of international law ‘more conducive to the realization of the rights of the child’ than those of the CRC should be respected as a higher standard.

What is ‘more conducive’ to the protection of rights could be understood in many ways. The CRC Committee may look narrowly to protect the ‘best interest of the child’ (CRC, art. 3), a principle that underlies the convention drawing on scientific understanding of the dangers of group care or its alternative. It could also be argued that determining what is ‘more conducive’ to the protection of rights must be understood broadly in terms of general human rights obligations. 

Practices that allow for the institutionalization of children for their own ‘protection,’ even under limited circumstances, may be in tension with principles of non-discrimination and full participation that are core to the principles of the CRPD. Under its general principles, the CRPD art. 3 guarantees ‘full and effective participation and inclusion in society’ as well as protection from discrimination. The CRPD Committee (2018a, para. 46) has stated that governments ‘are under an immediate obligation to eliminate discrimination against individuals or groups of persons with disabilities and to guarantee their equal right to living independently and participation in the community.’ The CRPD Committee (2018a, para.3) points out that ‘[m]any national laws and policies perpetuate the exclusion and isolation’ of persons with disabilities based on ‘multiple and intersectional discrimination or discrimination by association.’ These laws and policies may not appear on the surface to be discriminatory or even directed at individuals with disabilities; they are often said to be ‘for the protection and care of persons with disabilities or in their best interest’ (CRPD Committee, 2018a, para. 3).

The justification of residential care placement as ‘necessary’ to help or protect a child with of an impairment – a medical or psychiatric diagnosis – represents a classic use of the medical/tragedy model to restrict or deny rights in a way that is now inconsistent with the CRPD (CRPD Committee, 2018a, para. 8). To the extent that placement of a child outside their family is viewed as ‘necessary’ (CRC, art. 20) because of society’s failure to develop appropriate support in the community, the CRPD creates an obligation to fill the gap in providing support for placement in an alternative family setting. Where services and support are lacking to enforce a fundamental right, the CRPD requires that such programs be created (art. 19, 23). The law can never mandate that families love their children, but it can require that children be placed in some kind of family structure where attachments can be formed and their needs met.

The placement of children in institutions or residential care as a ‘last resort,’ as contemplated by CRC General Comment 9, is another example of an approach based on the medical/tragedy model that effectively discriminates against children with disabilities. In practice, if there is a lack of support to help children live in families, institutionalization can always be justified as a last resort. If the choice is between an institution or being neglected or even abandoned, then of course residential care will appear the better option and in the ‘best interest’ of the child concerned. But the CRPD demands a different way of responding to the problem. Rather than asking the child to adapt to existing social services, new and better services should be provided.

While protection against discrimination entails immediate obligations, article 4 (2) of the CRPD also recognizes that the implementation of some rights requires the creation of programs or social services that can be implemented progressively. Where new services are needed, the CRPD requires State Parties to begin immediately to plan and to act to bring about the ‘full realization’ of rights (Degener, 2017, pp.-38; Rosenthal, 2019, pp. 109-112).


The global stakes for children

The stakes in this debate could not be greater. As Disability Rights International President Laurie Ahern (2013) wrote:

An estimated 8 to 10 million infants and children live in orphanages around the world, and aid agencies, churches and governments provide hundreds of millions of dollars in the hope that they can help vulnerable children find sanctuary in these institutions. This hope is badly misplaced. Orphanages are not safe places for children.

There is now extensive international funding for both large and small institutions for children. A recent study of charitable giving finds that Christian donors from the United States alone account for $3 billion in annual contributions to residential care for children (Barna, 2021, p.10). Some international development agencies, such as the United Kingdom’s UKAID (2019), have adopted policies against funding any form of residential care. Other donors, however, support residential care, including group homes, as a replacement for large institutions (Rosenthal, 2019, pp. 116-117, 125-126). Disability Rights International (DRI) (2019) and the European Network on Independent Living (2020) have, for example, documented extensive EU funding of group homes.

Without a common understanding of what is an ‘institution’ or whether smaller facilities are safe or legitimate, it is impossible to know whether reforms are successful – or even moving in the right direction. A recent survey of leading international children’s organizations, commissioned by the Elevate Children’s Funders Group, found confusion and conflict around the best way to respond to institutions – and whether smaller residential facilities are institutions at all (Shawar and Shiffman, 2020, p. 1). The impact of these disagreements, according to many leaders of the field who were surveyed, is to undermine political support and funding for reform.

These differences in view are striking, given the enormous and growing evidence about the dangers of both large and small residential institutions. A comprehensive review by The Lancet Commission on the Institutionalization and Deinstitutionalization of Children found ‘compelling evidence that institutional care is associated with negative developmental outcomes’ (Van IJzendoorn et. al., 2020, p. 14). The Lancet Commission included in its definition of institutions ‘orphanages, children’s institutions, group homes, infant homes, children’s villages, and similar residential settings for children.’ As long as children are cared for by professionals or ‘shift workers,’ rather than emotionally-committed family members, even small group care will reproduce the dangers of larger institutions. Forming a healthy attachment to an adult family member cannot be replaced by congregate care, no matter how clean, well-funded, or professionally staffed (Dozier, et al, 2014, p. 220; Lockwood, Friedman and Christian, 2015, pp. 306-315).

While much attention has focused on young children – who are especially vulnerable – recent evidence shows that institutional care also threatens the health and development of older children and adolescents. An international Consensus Statement among child welfare researchers concluded that ‘[g]roup settings should not be used as living arrangements because of their inherently detrimental effects on the health development of children, regardless of age’ (Dozier, et al, 2014, p. 219, emphasis added).

Children with disabilities in institutions are at an especially ‘heightened risk of violence, abuse, and exploitation’ (UNGA, 2019a, para. 34). The UN Special Rapporteur on Torture has found that all children placed in institutions are at an increased risk of torture (HRC, 2015, pp. 4, 15).

Such concerns have been recognized at the UN level. In celebration of the 30th anniversary of the CRC, the UN Secretary General submitted a report to the General Assembly calling for States to ‘end the institutionalization of children’ (UNGA, 2019b, para. 66, 77). The Secretary General’s report raised concerns about the global pattern of ‘moving [children] from one institution to another, usually to smaller facilities, including group homes’ (UNGA, 2019b, para. 55). The report highlighted large-scale implications of group placement, noting that ‘globally, the use of residential placements surpasses that of family-based placements’ (UNGA, 2019b, para. 45, emphasis added).

Underlying the public support for institutional placement is the misperception that these facilities are places for needy children who have no place else to go (Ahern, 2013). In fact, studies in different parts of the world show that 80-90% of children placed in congregate care have at least one living parent (Van IJzendoorn, et al., 2020, p. 4). Children are usually placed in institutions due to poverty or disability (Faith to Action Initiative, 2014, p. 6), including lack of access to health care and education (Van IJzendoorn, et al., 2020, p. 4).

Experience shows that increased support for children to live with families would go a long way toward reducing or eliminating institutions for children (Lemay, 2009, pp. 181-194). If a child does not have parents or cannot live with them, alternatives can be made available to ensure that a child can live and grow up with a family – including kinship care, substitute family care, or foster care (Rosenau, 2000, p. 4; Kanter, 2008). It was once believed that some children with disabilities would have to be placed in residential care (OHCHR European Regional Office, 2011, pp. 10-11). Experience now challenges that assumption: with appropriate support, children with ‘every kind and severity of impairment [are] currently living successfully with a family’ (Rosenau, 2000, p. 7; see also Berens and Peacock, 2015, pp. 3-12; Casey, Irby and Withers, 2017, pp. 828-837).

The challenge of creating family support is real. Developing countries may lack trained professionals to help families respond to the needs of children with specific impairment (Rosenau, 2000, p. 7; DeLashmutt, 2015, pp. 8-9; Rosenthal, 2019, pp. 85-86, 136-137). In countries without a history of foster care, there may be cultural barriers to finding families willing to take in children (Cantwell, 2015, p. 261). Building group homes does not make it easier to deliver professional support. On the contrary, best practices for inclusion depend on what is referred to as ‘natural supports.’ Natural supports are no more or less than family, friends and existing social networks that exist in any society (Rosenau, 2000, p. 7; DeLashmutt, 2015, pp. 8-9). ).

Support services for families can be used to stop new admissions into institutional care and create a moratorium on new placements. But closing down institutions will be harder where children in long-term care have lost contact with families. Despite these challenges, effective reform is possible. Professor Joan Kaufman, co-author of the International Consensus Statement by child welfare researchers finding congregate care ‘inherently detrimental’ for children (Dozier, et. al., 2014), finds that there is evidence-based practice to demonstrate that short-term reform is possible. She says that a target date for the completion of those reforms is necessary to ensure activities are focused and the outcome is realized. An approach she suggests, which could be accomplished in ten years, is as follows:

With will and commitment, proper resourcing, crucial international and national partnerships, and proper data to monitor progress, the practice of institutionalization of children could be eradicated by the end of 2030. Existing residential and group care settings can be transformed into community centers offering assessment, case management, physical therapy, mental health, and other needed services, or transformed into family treatment centers where parents can receive substance abuse treatment or other necessary services and support while staying with children. Institutional care is not just bad for children’s development; group care is more expensive than foster care. It is time to make children’s right to family a reality (Kaufman, 2020, p. 653).

Kaufman’s target date does not mean rights should be delayed for any child who can be immediately accommodated in a family. New resources should primarily focus on helping children remain with or return to families. Significant investments in residential care would take scarce resources away from the ultimate goal of full family inclusion and delay structural reforms.


Lessons from disability experience

The dangers of large and small institutions have been documented worldwide by human rights and disability groups, such as Disability Rights International (DRI). DRI findings in Bulgaria, for example, provide useful insights into the ongoing problem. This country received extensive international funding and guidance in planning reforms leading to deinstitutionalization (DRI, 2019, p. xi). The EU has invested more than 260 million Euros in Bulgaria's reforms and the country has been held up as a model by UNICEF and the international reform movement (UNICEF, 2015, pp. 30-35). Bulgaria’s Deputy Prime Minister reported a 90% rate of deinstitutionalization (Zaharieva, E., 2019). Yet DRI’s report shows that a large proportion of children were simply moved from large facilities to smaller ones called ‘small’ group homes (DRI, 2019, p. 23).

DRI experts found that Bulgaria’s group homes, usually made up of 14 beds and staffed by professionals, ‘are neither small nor are they family homes’ (DRI, 2019, p. i). These facilities:

separate children with disabilities from society and contribute to their continued isolation – leading to a lifetime of segregation for a new generation of children with disabilities. As this report shows, placement in Bulgaria’s group homes exposes children to emotional neglect, inappropriate and potentially damaging models of behavior, and in some cases, violence, bullying and other forms of abuse that are common in institutions (DRI, 2019, p. 1).

DRI noted that the ‘closure of Bulgaria’s old orphanages is an enormous accomplishment that saved the lives of many children’ (DRI, 2019, p. xi). But reforms replaced them with ‘newer, smaller buildings that are still operating as institutions’ (DRI, 2019, p. i). In Bulgaria, investment in group homes turned out not to be a stepping stone to family reintegration. Funds were spent on group homes at the expense of family support, inclusive education, and transition to independent living. As the report is titled, placement in group homes is a ‘Dead End for Children’ likely to last a lifetime (DRI, 2019).

The Bulgaria experience demonstrates the limitations of international standards intended to limit placement in residential care. Indeed, Bulgarian law limits residential care to the last resort (when possibilities for placement in the home are ‘exhausted’). DRI interviewed judges who were authorized to make such determinations for each residential care placement in Bulgaria (DRI, 2019, pp. 28-33). These judges explained to DRI investigators that, when no alternatives exist to support a child to live with their family or foster family, placement in a group home or institution is always the placement of last resort (DRI, 2019, p. 28-33).

The UN Guidelines state that ‘residential care should be limited to cases where such a setting is specifically appropriate, necessary and constructive for the individual child concerned and in his/her best interest’ (UNGA, 2010, para. 21). As the Bulgaria experience demonstrates, the promise of these standards is empty without support to families to keep children with disabilities. Usually there was only one place for a child to go: a group home (DRI, 2019, pp. 28-33). If the only alternative for the child is to be returned to the old, crumbling orphanage or placed in inadequate foster care with no disability support, group homes will always appear to be in the ‘best interest’ of the child.


The danger of conflicting international standards

The UN General Assembly marked the 30th anniversary of the CRC in 2019 by adopting a resolution on children (UNGA, 2019c). A coalition of international child care organizations (UN Disability and Children’s Rights Group on Behalf of Children Without Parental Care, referred to here as the Coalition on Children), supported by UNICEF, sought to use this resolution to bring attention to the millions of children living without parental care. An international coalition of disability and children’s rights groups (referred to here as the Disability Coalition) supported this effort but sought to avoid endorsement of group placement until conflicts between international standards were resolved by the CRPD-CRC working group (Coalition on Children without Parental Care, 2019, p. 3).

The Disability Coalition had significant influence on the ‘Key Recommendations’ adopted by the Coalition on Children. Recommendation #1 recognizes the right to family life (Coalition on Children without Parental Care, 2019, p. 1). The Key Recommendations also state that the UN Guidelines should be implemented ‘in conformity with’ the CRPD (Coalition on Children without Parental Care, 2019, pp. 2, 5). Taken seriously, this would mean that many of the valuable recommendations of the UN Guidelines on supporting children in the community would be recognized – but would end the endorsement of residential care.

Unfortunately, the Disability Coalition was outnumbered on a working group created by the Coalition on Children, and the Key Recommendations included a limited endorsement for group care, stating:

… in specific cases it may be necessary to provide quality, temporary, specialized, care in a small group setting organized around the rights and needs of the child in a setting as close as possible to the family, and for the shortest possible period of time. The objective of such placement should be to contribute actively to the child’s reintegration with his/her family or, where this is not possible or in the best interest of the child, to secure his/her safe, stable, and nurturing care in an alternative family setting or supported independent living as young people transition to adulthood. (Coalition on Children without Parental Care, 2019, p. 3)

The Key Recommendations represent an advance over the UN Guidelines by insisting group care be used only for short-term placement. Disability groups felt that the recommendations were not in conformity with the CRPD. Experience shows that, when no alternatives exist to support children in the family, placement for the ‘shortest possible’ time may last a lifetime. Until social service reform catches up to the rights of the child, all group care can therefore be justified as being for the ‘shortest possible’ time.

There is a larger concern about the language of the Key Recommendations. The acceptance of the ‘necessity’ of group care in ‘specific cases’ suggests that the existence of a residential care system is a permanent and inevitable part of any social service system. The Key Recommendations do not describe residential care services as a temporary arrangement that can be entirely phased out through progressive realization. Thus, the Key Recommendations effectively undermine the full realization of rights through progressive implementation.

In any society, the need for emergency placement will arise – such as the identification of abuse in the family – where a child will be immediately removed from their family. If extended kinship care is not available, emergency foster care has proven effective. The California Evidence-Based Clearinghouse (CEBC) (2006-2021) has ratings of a range of evidence-based foster programs. Some countries have not yet created such programs, especially for children with disabilities. Such countries must be recognized as falling short of their international obligations. This is part of the political pressure that international law can create to foster progressive implementation and the full realization of rights.

There will also be times of national emergency or conflict, as well as other situations, where there are large numbers of internally-displaced or migrant populations impossible to serve through established social service systems. Article 11 of the CRPD (2006) prohibits any limitation of rights based on disability even in situations of emergency. But human rights law allows for the limitation (or ‘derogation’) of some rights on the condition that emergencies be publicly proclaimed and ‘strictly limited to the exigencies of the situation’ (Motz, 2018, p. 323). That said, there is nothing in the Key Recommendations to limit such use of residential care to such emergencies.

The Disability Coalition was not able to persuade UNICEF not to endorse the Key Recommendations, but their arguments proved to be influential on the UN General Assembly, which in its Resolution of 18 December 2019 did not endorse residential care placement (UNGA, 2019c, para. 10-35). The resolution focused on ‘[c]hildren without parental care’ and called on governments to ‘intensify their efforts’ to bring about full community inclusion (UNGA, 2019c, para. 21-35). In addition to citing the CRC, however, the resolution approvingly cites both the CRPD and the UN Guidelines (UNGA, 2019c, para. 34(l), 35(b)). By citing conflicting standards, the General Assembly allows for a mixed message about the legitimacy of residential care’s ongoing role.

The danger of these mixed messages can be seen in the confusion between two papers published in Lancet. The Lancet Commission on the Institutionalization and Deinstitutionalization of Children published two separate reports: the first (referred to here as the Scientific Analysis) being a review of the research literature on the impact of institutional placement on children (Van IJzendoorn et al., 2020); the second (the Policy Analysis) made recommendations for reform and practice (Goldman, et al, 2020a). The Scientific Analysis concluded that ‘every effort should be made to minimize children’s exposure to institutional care,’ and called for ‘care alternatives that are family-based’ (Van IJzendoorn, et al., 2020, p. 703, 704). 

The Lancet Policy Analysis, however, backtracked on the scientific findings regarding the dangers of group care. The Scientific Analysis advised against group home placement, while the Policy Analysis states that ‘[group] care might be necessary in very few situations’ (Goldman, et al, 2020a, p. 607). The Policy Analysis justifies group care based on the Key Recommendations. It refers to the Key Recommendations as prepared ‘for the UN General Assembly’ (Goldman et al., 2020a, p. 608, emphasis added). The Policy Analysis never mentions that the General Assembly declined to endorse the Key Recommendation on group care. The authors point out that the Key Recommendations were endorsed by hundreds of international children’s organizations and UNICEF, but they fail to note that they were opposed by the world’s largest networks of disability groups (Goldman, et al, 2020a, p. 607).

The Lancet published a letter from this author and allies pointing out this discrepancy (Allen et al., 2020). To their credit, the authors of The Lancet Commission do not defend the practice of residential care placement. Indeed, they ‘share the concerns related to the provision of small group care worldwide, including the possibility of structural neglect, isolation from the community and the risk that these conditions become permanent’ (Goldman, Van IJzendoorn and Sonuga-Barke, 2020b). The only rationale they offer is that ‘[t]o our knowledge, no child protection system to date has been able to secure a safe and nurturing family for every child and at all times’ (Goldman, Van IJzendoorn and Sonuga-Barke, 2020b, p. 4).

The Lancet authors may be correct in stating that no country has yet fully eliminated all forms of institutions or residential care. It has, however, been demonstrated that such care can be eliminated for children with any type of disability (Rosenau, 2000, p. 7; Kaufman, 2020, p. 652). The rationale offered by The Lancet’s policy analysis is based on problematic assumptions about the nature of human rights: that human rights law cannot oblige governments to achieve outcomes beyond what they now provide. In fact, there are many areas of international law that do just this. Protection against discrimination is accepted as a fundamental part of international law even though there is no nation in the world without actual discrimination taking place. The right to the highest attainable standard of health is widely accepted even though ‘health’ itself cannot be legislated. This misunderstanding may be at the heart of the difference between adherents and opponents of the ‘necessity’ argument for group home placement.


Conflicting views of progressive realization

A survey of leading international children’s organizations, commissioned by the Elevate Children Funder’s Group, revealed that residential care placement is widely viewed as a form of progressive realization (Shawar and Shiffman, 2020). Identifying residential care as a practice that can be eliminated through progressive realization in effect ensures that residential care will continue to be used, undermining the full implementation of the right of children to live and grow up within a family.

Leading disability experts and the CRPD Committee come to the opposite conclusion. New investments in residential care, even if presented as transitional, are seen as taking funds away from family-based care and therefore as undermining rather than promoting full realization of rights. Under international law, retrogressive actions contrary to full realization of rights are strictly limited (UN ESCR Committee, General Comment No. 3, 1991, para. 9). In General Comment No. 5, the CRPD Committee stated that States are ‘prohibited from taking retrogressive measures with respect to the minimum core obligations’ (CRPD Committee, 2017, para. 45), and the right to grow up in a family list listed as ‘the core of the right to be included in the community….’ (CRPD Committee, 2017, para. 37). According to Gerard Quinn, writing before he became the UN Special Rapporteur on Disability:

One thing is abundantly clear: the creation of (and the expenditure of funding on) new long term care residential institutions (large, or small, or smaller) is not permissible. Investing in any institution is presumptively discriminatory. As the UN Committee puts it – while the program to deinstitutionalize is subject to ‘progressive achievement’ the actual end goal of deinstitutionalisation (‘replacement’ in the words of the Committee) is non-negotiable … [T]ransitioning away from long-term care residential institutions requires foresight and planning. The planning must assume closure – and not way stations toward closure (Quinn, et al., 2018, p. 14-15).

This analysis is also important in case economic arguments are put forth for the necessity of residential care. In fact, in the long run, full integration into a family is more cost-effective than residential care (Mansel, Knapp and Beadle-Brown, 2007, p. 6; Kanter, 2015, p. 102-3; CRPD Committee, 2017, para. 5). In Bulgaria, for example, placement in group homes is more than twice the cost of placement in foster care (Van IJzendoorn, et. al., 2020, p. 5).

The CRPD Committee has taken a stand against investments in institutions, whether they are labelled as such or not, asserting:

States parties should ensure that public or private funds are not spent on maintaining, renovating, establishing, building or creating any form of institution or institutionalization. Furthermore, States parties must ensure that private institutions are not established under the guise of community living (CRPD Committee, 2017, para. 51).

These principles also apply to the ways that States Parties provide funding and assistance through international aid schemes (CRPD, 2006, article 32).

The principle of progressive realization recognizes that certain changes can only happen over time. Yet this legal principle also entails immediate obligations. Decisions on funding and financing services are exactly the kind of decision-making that must be subject to immediate implementation under international law. When governments or donors make the decision to invest in residential care, they discriminate against children by consigning them to a life deprived of family.


Implications of the human rights model of disability: protection of all children

Many residential programs throughout the world are set aside for children with disabilities or children with specific psychiatric labels. This practice creates ghettos of segregation for children with disabilities and represents a clear form of discrimination.

Protection against discrimination opens up a broader application for the protection of the CRPD to children placed in residential care whose impairments may not have been recognised or children wrongfully throught to have impairments (CRPD, 2018, para. 20). The anti-discrimination mandate of the CRPD is not limited to children with visible or actual impairments. The CRPD’s preamble states that ‘disability results from the interaction between persons with impairments and attitudinal and environmental barriers that hinders their full and effective participation in society on an equal basis with others.’ The CRPD Committee has stated that ‘[d]ignity, integrity and equality of the person have been denied to those with actual or perceived impairments’ (CRPD Committee, 2018a, para. 7, emphasis added). Discrimination ‘on the basis of disability’ can take place ‘against persons who have a disability at present, who have a disability in the past, who have a disposition to disability that lies in the future, who are presumed to have a disability, as well as those who are associated with a person with a disability’ (CRPD Committee, 2018a, para. 20).

Perhaps the strongest argument for the application of CRPD protections to children in residential care is that any child could become impaired by the trauma of family separation, or acquire a psychological or developmental impairment as a result of the emotional neglect inherent in institutional placement. The act of labelling any such child as unable to benefit from family-life carries a stigma that could stick with that child, hindering their full participation in society. Denial of the opportunity to grow up with a family for any child based on actual or perceived impairments should be understood, therefore, as a form of per se disability discrimination (Rosenthal, 2019, p. 108).

Failure to enforce CRPD articles 19 and 23 for parents and society at large also results in the denial of rights for a wide range of children. If a child without any impairment is placed in a group home because their parents have a disability and are thought to be unable to care for them, this could also constitute disability discrimination. CRPD General Comment No. 6 recognizes discrimination by association as prohibited by the convention (CRPD, 2018, para. 20). Studies in various countries have estimated that as high as 60% of children are taken from parents with intellectual disabilities (Lawson, 2007, p.568).

Children without a diagnosis or identified impairment are often placed in institutions or residential care because their home life is considered inadequate. Except in cases where abuse is clearly identified, child protection authorities may be discriminating against parents with disabilities who would otherwise be able to take care of their children with adequate support. Neglect is much more common than abuse, and it can usually be remedied with appropriate social supports (Anne E. Casey Foundation, 2013, p. 17). If parents are truly unable to keep a child – after they have been offered the support and accommodations required by CRPD article 19 – then CRPD art. 23(5) still grants that child a right to placement in the community in a family setting. Thus, children without identified impairments are clearly stakeholders in the effective implementation of art. 19 and 23.

The paradigm shift brought about by the CRPD expands rights for all people, benefiting society as a whole and protecting other groups also excluded by disabling barriers (Kanter, 2015, p. 49). Catalina Devandas, the former UN Special Rapporteur on Disability, observed that:

The Convention also has the potential to override traditional understanding of care and assistance for other groups, such as older persons and children. The Convention restores the importance of the ‘human being’ in the human rights discourse by emphasizing the individual and social aspects of human existence. These innovations can and should be incorporated into the implementation of all existing human rights instruments (UN Human Rights Council, 2016, para. 41).

Thus, the human rights model of disability created by the CRPD can help governments innovate new solutions to help a broad range of children placed in residential care with actual or perceived impairments -- or at risk of such placement.


Toward resolution: more inclusive policy-making

Effective solutions to protect children will require an inclusive approach to policy-making. Special Rapporteur Gerard Quinn warned that the CRPD should not be seen as a ‘magic bullet’ for top-down solutions administered by governments (Quinn, 2009, p 218). He suggested that the CRPD be seen ‘as a powerful tool for enabling its revolutionary insights to percolate into the political process’ (Quinn, 2009, pp. 218, 256).

When the CRC and CRPD are both implemented, the right to participation recognized in each convention will create synergies. The CRC (art. 12) provides a mandate to listen to children and requires their views to be heard. The CRPD (art. 4(3)) requires that in the ‘development and implementation of legislation and policies … on issues relating to persons with disabilities,’ governments must ‘consult with and actively involve persons with disabilities, including children with disabilities, through their representative organizations.’ Article 4(3) of the CRPD was written to avoid token representation of individuals with disabilities by requiring engagement to take place with representatives of disability organizations.

Through vigorous advocacy under the CRPD, the disability community has shifted support away from the long-term use of residential care for children, as envisioned in the UN Guidelines. The Key Recommendations, supported by UNICEF and many international children’s organizations, attempt to restrict residential care to temporary transitional use. Disability groups fear that this restriction is not strong enough and will lead to the entrenched use of residential care. But the limiting principle is valuable. And in the two years since the adoption of the Key Recommendations, more than a hundred disability rights organizations, joined by some children’s groups that once supported the Key Recommendations, have adopted a ‘Call to Action’ to ‘Protect the Right to Family Life and Prevent Institutionalization for All Children’ (Autism Europe, Disability Rights International, European Network for Independent Living, et. al., 2021). The Call to Action asks governments to harmonize international standards consistent with General Comment No. 5 and ‘[u]rgently undertake every effort to support family-based care instead of large or small institutions; available resources should be used to support families and create more family-based placements not for building or expanding any new large or small institutions.’ The Call to Action also calls for the international community to review and update the UN Guidelines, to ‘reflect current scientific knowledge about the dangers of non-family-based congregate care and any kind of institutional or residential care placement.’

These are signs that the voices of the disability community may at last be heard and may thus have greater influence moving forward. After a long and hard debate, it is encouraging that the CRC Committee in its preparation for its September 2021 Day of General Discussion – which could lead to improved standards on alternative care – stated in its guidelines for submissions that it is committed to ‘[l]earn from adults who have been placed in alternative care about their experiences both within and upon leaving care, to build understanding of good and bad practices, and draw from their recommendations for reforming and strengthening child care and protection systems, including how these can inform comprehensive deinstitutionalization processes’ (CRC Committee, 2021, p. 8).

It is vital that the disability and child rights communities do come together. On this point, it is encouraging that a recent White Paper, developed in partnership with the GHR Foundation, a major funder of international child care reform and a member of the Elevate Children Funder’s Group, takes just this approach, warning against the ‘silo’ mentality that divides the two communities, stating that:

When you build a system for children with disabilities, you build a system that works for all children. When you compartmentalize from the first planning stage, you bring the stigmatization of disability into your work practice (Burchell, 2021, p. 2).

The success of any reform must bring governments around the world to full realization of rights under both the CRC and the CRPD. This outcome will depend on the full inclusion of children with disabilities and on the voices and full participation of organizations of people with disabilities.



There is no essential conflict between the CRC and the CRPD. Both conventions protect the best interest of the child and both respect the role of families. CRC article 20 allows the placement of a child in an institution (large or small) but never requires it. As institutional or congregate care comes to be understood as a threat to the health, development and equal protection of the rights of the child, there are proven models of family-based care that can be used instead. By recognizing that ‘the core of the right to be included in the community entails a right to grow up in a family,’ CRPD General Comment 5 interprets the CRPD as requiring governments to create the protection and supports necessary to allow family placement of all children.

Through the CRPD’s human rights model of disability, it is evident the language of General Comment 5 is rooted in fundamental principles of the convention. The CRPD protects against discrimination on the basis of actual or perceived disabilities that hinder full and effective participation in society. Residential care creates impairment and results in disability discrimination by limiting the child’s enjoyment of family life like all other children.

Allowing the placement of the child in an institution, even as a ‘last resort’, is a classic medical/tragedy model limitation on the child’s rights based on actual or perceived deficits. Governments can do better, and human rights law requires them to do so.

Recognizing group care as a ‘necessary’ form of transition to community-living for some children, even when used for the ‘shortest time possible,’ suffers the same medical/tragedy model limitation. Experience shows that what is called temporary becomes permanent when community support systems are inadequate. Establishing group care as a ‘necessary’ part of a social service system guarantees that it will not be progressively eliminated and undermines the full realization of rights. Investment in residential care is retrogressive – delaying rather than bringing about full realization of rights.

As they work toward harmonization, the CRC and CRPD Committees must not gloss over the conflicts in existing international standards, including CRPD General Comment No. 5 and the UN Guidelines. These instruments send contradictory messages about steps necessary to bring about full implementation of rights. The right to family life and the protection from residential placement, as described in General Comment No. 5, are rooted in the core principles of the CRPD. While the UN Guidelines are still widely respected, there is now a growing – if not explicitly acknowledged – recognition that they must be updated.

Clear definitions of such terms as ‘family’ and ‘institution’ must be established. As former Special Rapporteur Catalina Devandas pointed out, the definition of an ‘institution’ for a child must include any placement that is not family-based. Once this definition is broadly and clearly established, only small changes have to be made to the UN Guidelines to bring them into accordance with the CRPD. The UN Guidelines call for every country to come up with a plan, with target dates and financing, for the progressive elimination of institutions. The obligation to progressively eliminate institutions must include the obligation to plan for progressive elimination of residential care and group homes.

Arguably, the human rights model of disability provides a framework to require that protection against discrimination in the CRPD does not just prohibit the ‘over-representation’ of children with disabilities in institutions or residential care. It also allows ‘disability’ to be understood as a social construct. Thus, protections against disability discrimination are applicable to all children deemed unable to live in the community or enjoy the benefits of family life. These protections allow any placement in residential care to be seen as a discriminatory denial of the opportunity to live with a family.

Until societies become inclusive for children and adults with disabilities, no carefully crafted standard for regulating or prohibiting institutional or residential care will be fully effective. Broad alliances between child care and disability rights groups can create synergies to build political will and funding for the necessary global reforms. The transformations promised by the CRPD will only take effect when barriers are removed between children with and without disabilities by a universal guarantee that no child may be deprived of family life.


[1] This article uses people first language to refer to the person before the label or diagnosis and to emphasize universality of disability protections for all children.