Debate on Australian policy and practice about residential care is to be welcomed as it has important implications for the care and wellbeing of vulnerable young people, as well as the allocation of resources to care and protection systems. I therefore welcome the commentary provided by Ainsworth and Holden on the arguments offered in ‘Weighing Up the Evidence and Local Experience of Residential Care’. Whilst these reviewers of my Practice Commentary have accepted many of the points made, I wish to take this opportunity to reinforce my arguments in light of their critique.

To recap the original article, I urged caution in current discussion on expanding residential care – including advocacy for therapeutic models. Importantly, I did not unequivocally state that all children and young people needing care can be placed in community-based and/or foster care settings, or that there is no need at all for residential care. However, I did indicate a number of serious concerns about pressures to increase the number and type of residential units.

First, I argued that residential care can be difficult to manage, draining on the welfare system, and without direct impact on achieving permanency outcomes for children. These are issues not directly addressed by Ainsworth and Holden in their review. Nevertheless, the major purpose in writing the original paper was to examine the theory and practice of therapeutic residential care in the Australian search for ‘permanency’ outcomes for children in out-of-home care. This is particularly important given the capacity of the child protection workforce and Out-of-home care (OOHC) sector, and recent revelations of very poor standards of care (Royal Commission into Institutional Responses to Child Sexual Abuse, 2016). The original paper was written from the perspective of my own agency's 30-year experience. We are concerned by evidence of past poor practice (including well documented sexual abuse and exploitation), use for children under 12 years of age without justification in a permanency plan, and, use of residential care as a crisis short-term measure in a child or young person's life. The paper concluded by suggesting caution in order to avoid the risk of increasing impermanency for children as an indirect or unintended consequence of residential placement where restoration to family cannot be successfully achieved, and no other person is available to provide ongoing care into adulthood/independence.

Second, I argued that only a very small number of young people may benefit from residential care and, consequently, we need to be extremely conservative about its use. The reviewers do not seem to have acknowledged the point I made about some older adolescents requiring residential care until they can move into secure independence – we run one such unit ourselves. I argued that the use of residential care can be appropriate for extremely damaged children who have grown up in foster care and/or have experienced other significant trauma (such as refugee children). Such residential placements can be very stable. However, our position is that most young people need to be in relationships where they feel that they belong and have the stability to move into adulthood – this is a prerequisite for healing within a care environment of therapy and care. Furthermore, it is very important that children under 12 years of age are not in residential units – a previously well-established principle in Australia that is increasingly being breached. Whilst it may occasionally be necessary that young children stay with siblings, residential care can be a very difficult environment for young children because of the influence of older children and difficulty of managing multiple relationships with adults. We fear that residential beds may be increasingly used for younger children only because there are no alternate placements.

Third, I warn that the evidence base for residential care is not yet available and until it is, no expansion should be considered. Ainsworth and Holden contend that I do not present the evidence base and have themselves provided some examples of US research that is seen as ‘promising’. However, my conclusions are based on the consensus deliberations of the International Working Group on Therapeutic Residential Care, which concludes that there is not an adequate evidence base for therapeutic residential care (Whittaker et al., Reference Whittaker, Holmes, DelValle, Ainsworth, Andreasses and Anglin2016). The research evidence offered by Ainsworth and Holden is of course interesting but, as with all work on ‘outcomes’ the length of time over which the assessment of outcomes was made and whose views are considered, needs to be carefully considered. I contend that what is needed for a proper evidence base is medium- and long-term outcome studies that look at the contribution of residential care to finding a permanent solution for young people, and contribute to the young person's development. Short-term changes in behaviour or assessment whilst in a residential unit are not adequate! One thing that Ainsworth and Holden concede, however, is that there is not adequate Australian research on which to base judgments.

Fourth, I contend there are inherent problems with residential care including young people's dislike of them and the problems of shift work. Ainsworth and Holden do not respond to these arguments – neither offering support nor criticism. Yet, young people's experience of residential care is generally negative as has been well documented by CREATE (McDowall, Reference McDowall2013) and, more recently, by the Royal Commission into Institutional Response to Child Sexual Abuse (Moore, McArthur, Roche, Death, & Tilbury, Reference Moore, McArthur, Roche, Death and Tilbury2016). Furthermore, the nature of shift work (which is a core characteristic of residential care) is very disruptive for young people who have had poor history of relationships and are, as a result, likely to have great difficulty forming and maintaining relationships over time. It is a contestable argument as to whether children and young people with disrupted previous relationships are best cared for by shift workers who move on regularly simply because they ‘do not want to be in a family’ or are ‘unable to attach’.

Finally, I would like to point out that the availability of existing residential care beds is not adequate justification for expanding this form of care. Ainsworth and Holden argue that residential care is used widely (for example, in Queensland, NSW and overseas), and therefore there must be a strong basis for this form of care. I believe this to be a weak argument, in part because there will always be pressure to use an available ‘bed’, but more importantly because we need to continually re-evaluate the nature of the care we provide to children and young people, especially in relation to the growing critique of residential practices.

Both the reviewers and I agree that no one wants to see children living in motel rooms, juvenile justice facilities or being homelessness – but further options within Australian policy and practice environments need to be considered in order to produce the best possible permanency outcomes of welfare intervention for vulnerable children and young people (Tregeagle, Moggach, & Cox, Reference Tregeagle, Moggach and Cox2013; Tregeagle, Moggach, Cox, & Voigt, Reference Tregeagle, Moggach, Cox and Voigt2014). I reiterate that the ‘bottom line’ is that residential care facilities need to guarantee permanency, warm and welcoming environments, and be able to meet the developmental needs of the young person.