The rights of the child in voluntary care in Ireland: a call for reform in law, policy and practice

Voluntary care agreements form a significant part of child protection systems in many jurisdictions. From a children’s rights perspective, they enjoy numerous advantages over court-ordered removals of children. However, when loosely regulated, voluntary care agreements can give rise to significant concerns in respect of compliance with international children’s rights law. This paper will present findings from the Voluntary Care in Ireland Study, one of the first in-depth empirical examinations internationally of voluntary care agreements. It will present qualitative data on the system in operation in Ireland that indicates that voluntary care agreements are less adversarial, time-consuming and costly than court proceedings. This frees up resources for early intervention and facilitates a more collaborative relationship between parents and social services, making it more likely that children will remain at home or eventually return home from care. At the same time, the findings suggest that the voluntary care system currently operated in Ireland suffer from numerous flaws, including absence of independent oversight; unlimited duration; potential instability (since parents can withdraw consent at any time); weak mechanisms for child participation; and inferior resource allocation compared to courtordered care placements. The paper examines legislative provisions from a number of comparable jurisdictions and makes recommendations designed to ensure that the voluntary care system in Ireland complies more strongly with principles of international children’s rights law.


Introduction
In Ireland, as in many other countries, there are two main pathways through which children may be placed in the care of a person other than their parents or guardians: through a care order granted by the District Court (which is a non-specialist general court that also deals with minor criminal offences, as well as other issues (O'Mahony, Burns, Parkes, & Shore, 2016)) or through the voluntary care pathway. The voluntary care pathway is a social work-led administrative system which involves a parent consenting to their child being cared for by another party (such as a kinship carer or foster carer). Although parents agree to cede day-today care of the child, they retain all other parental rights and responsibilities, and may cancel the agreement at any time. In Ireland, a voluntary care agreement obliges the Child and Family Agency (which operates under the name 'Tusla') to provide care for the child for as long as the child's welfare requires it. A slight majority of children (55% in 2018, down from a high of 70% in 2014) who enter alternative care do so pursuant to a voluntary agreement (Tusla, Child and Family Agency, 2020). Although a precise breakdown of type of care placement is not available specifically for children in voluntary care, the vast majority of children (an average of 90% between 2016 and 2018) in the Irish care system are placed in foster care, with approximately 20% in relative foster care and 5% in residential care (Tusla, Child and Family Agency, 2020). Foster carers receive a foster care allowance, which applies both to children in care pursuant to care orders and children in voluntary care.
The use of voluntary care agreements is commonplace internationally, featuring in child care legislation in common law jurisdictions including England and Wales, New Zealand and Ontario. In Europe, voluntary care agreements form a prominent feature of the care systems in Germany (Haug & Höynck, 2017), Finland (Pösö & Huhtanen, 2017) and Sweden (Svensson & Höjer, 2017). The effect of voluntary care arrangements may vary across jurisdictions with respect to the obligations that are placed on social services to provide care for the child; for example, in several Australian states, voluntary care agreements do not place children formally in the care of social services, although they are regulated by legislation to support the safety and wellbeing of the children. The common denominator of out-of-home voluntary care systems is that children are placed in the care of parties other than their parents or guardians in the absence of a court order, during which time their parents or guardians retain their rights and responsibilities in respect of the children.
Voluntary care has an important role to play in child protection systems. Numerous studies across jurisdictions have documented the cost, delay and adversarialism that may be involved in courts or courtlike processes (O'Mahony, Burns, et al., 2016;Hollander-Blurnoff & Tyler, 2011;Sheehan & Borowski, 2014;Wood, 2008). By contrast, the available literature in Ireland has suggested that voluntary care is a less adversarial and more partnership-focused intervention (Burns, O'Mahony, Shore, & Parkes, 2017;Burns, Pösö, & Skivenes, 2017). Corbett (2018), for example, lists the positives of voluntary care: it is quick, cost effective, accessible to parents in crisis, can be temporary, provides refuge for a child, can be a family support intervention, and avoids the court environment. These positives are echoed in analysis of section 20 of the Children Act 1989 in England, where the opportunity provided by voluntary care for partnership between family and the state has been highlighted (Dickens, 2016;Masson, 2018). Where parents consent to the agreement, social services can focus on working with them to provide the support they need to care for their child, instead of working against them to prove that the threshold for involuntary removal has been met. (Indeed, voluntary care allows for out-of-home care to be provided in cases where perhaps the strict legal threshold might not be met, but the parents and social services agree that the placement is in the best interests of the child). Voluntary care is of particular benefit in cases where a parent has requested short-term respite and does not propose to contest an application for a care order. It is also beneficial in longer-term scenarios where reunification is more remote, but parents wish to remain involved in their children's lives; the fact that parental rights are retained means that continuous input from the parents on decisionmaking for the child is necessary. This involvement may help to smooth over reunification occurring due to the child leaving the voluntary care placement, or ageing out of care and returning home thereafter.
By eschewing the necessity to make a court application, voluntary care keeps cases within a social work decision-making paradigm where social work values and ethics predominate. Dickens (2016) has observed that this differs substantially from legalistic court-based decision-making, and allows for cases to be handled in a less judgmental way. However, courts or court-like processes also bring certain advantages, in that they tend to have various in-built safeguards designed to protect the rights of the children and parents involved. From a children's rights perspective, these include mechanisms designed to ascertain the views of the child, and the presence of an independent decision-maker such as a judge who can take steps designed to ensure that the child's rights are vindicated by child protection and welfare services. These processes and mechanisms may be more or less effective at protecting children's rights; but at least they exist, and there is a body of empirical research available that assesses their operation in practice (Parkes, Shore, O'Mahony, & Burns, 2015;Burns, O'Mahony, et al., 2017;Burns, Pösö, et al., 2017;Eronen, Korpinen, & Pösö, 2020).
By contrast, laws regulating voluntary care agreements do not always contain the same sorts of safeguards; and voluntary care systems have yet to be the subject of any detailed empirical examination. A recent comprehensive text examining child welfare systems in eight countries (Burns, O'Mahony, et al., 2017;Burns, Pösö, et al., 2017) found that while all eight countries had a voluntary care pathway, none of the eight had research data examining the operation of voluntary care. As such, it is unclear whether, and to what extent, legislation and policy governing voluntary care agreements are effective in ensuring that child protection and welfare services protect children's rights in the absence of independent oversight by a judge or other decision-making body.
This paper presents data from one of the first international studies of its kind: the Voluntary Care in Ireland Study, 2018-2021. Through the medium of a mixed-methods study of professional experiences of voluntary care agreements in Ireland, it addresses the following research questions: (1) What are the strengths and weaknesses of voluntary care agreements in Ireland in terms of compliance with children's rights principles? (2) What elements of legislation governing voluntary care agreements in other jurisdictions might be adapted to enhance the protection of children's rights in the Irish voluntary care system?

Children's rights in voluntary care
The Convention on the Rights of the Child 1989 (CRC) contains multiple provisions setting out the rights that should be enjoyed by children who enter state care. The most relevant for the purposes of this paper include: • Article 3(1): the best interests of children should be a primary consideration in all matters affecting them. • Article 3(2): children have the right to such protection and care as is necessary for their well-being. • Article 12: children have the right to express their views freely in all matters affecting them, and their views should be given due weight in accordance with the age and maturity of the child. • Article 20: children temporarily or permanently deprived of their family environment have the right to special protection and assistance provided by the State. • Article 25: children in alternative care have the right to have their placement reviewed periodically by the competent State authorities.
Cutting across all of these rights is the principle of nondiscrimination under Article 2 CRC, which requires States Parties to ensure the rights set forth in the CRC to each child within their jurisdiction without discrimination of any kind, and to take all appropriate measures to ensure that the child is protected against all forms of discrimination. This principle 'is not a passive obligation, prohibiting all forms of discrimination in the enjoyment of rights under the Convention, but also requires appropriate proactive measures taken by the State to ensure effective equal opportunities for all children to enjoy the rights under the Convention. This may require positive measures aimed at redressing a situation of real inequality' (Committee on the Rights of the Child, 2013). In the present context, this requires that children in voluntary care should not be treated less favourably than children in care pursuant to a care order, and should have the above rights protected to the same degree. This paper will assess whether or not this is the case in Ireland through an examination of law and policy and of empirical data on professional experiences and perspectives. To the extent that the analysis identifies issues and challenges that arise in the Irish context, it is reasonable to question whether similar, and as-yet unidentified, issues may exist in the voluntary care systems of other jurisdictions.
In Ireland, there are numerous important legal provisions for the protection of children's rights in the context of care orders resulting from court proceedings. Article 42A.4.2 • of the Constitution requires that laws be enacted giving effect to the child's right to be heard in the context of court-based child care proceedings. Current provisions in this regard include section 26 of the Child Care Act 1991 (which allows for the appointment of a guardian ad litem to report on the child's views and best interests), and section 25 of the same Act (which allows for a solicitor to be appointed to represent a child). Research indicates that the power to appoint a solicitor is rarely used, while the rate at which guardians ad litem are appointed varies widely from district to district, with some judges appointing them in over 80% of cases, while others appoint them in as few as 20% of cases (Parkes, Shore, O'Mahony, & Burns, 2015). Draft legislation is currently under development designed to ensure that a higher proportion of children in care proceedings benefit from a guardian ad litem by establishing a legal presumption in favour of their appointment (O'Mahony, 2020). This is driven in part by the need to comply with the strengthening of the constitutional obligation to ascertain the views of children in court proceedings on foot of the enactment of Article 42A, which was approved by referendum in 2012.
None of the above provisions extend to voluntary care agreements, which are regulated by a single section (section 4) of the Child Care Act 1991: (1) Where it appears to the Child and Family Agency that a child requires care or protection that he is unlikely to receive unless he is taken into its care, it shall be the duty of the Agency to take him into its care under this section.
(2) … nothing in this section shall authorise the Child and Family Agency to take a child into its care against the wishes of a parent having custody of him or of any person acting in loco parentis or to maintain him in its care under this section if that parent or any such person wishes to resume care of him. (3) Where the Child and Family Agency has taken a child into its care under this section, it shall be the duty of the Agency-(a) subject to the provisions of this section, to maintain the child in its care so long as his welfare appears to the Agency to require it and while he remains a child, and (b) to have regard to the wishes of a parent having custody of him or of any person acting in loco parentis in the provision of such care.
A number of other, more general legislative provisions that apply to all children in care (whether pursuant to a court order or a voluntary care agreement) are relevant to the analysis of children's rights in voluntary care. These include section 3(2) of the Child Care Act 1991, which provides that in the performance of its function of promoting the welfare of children who are not receiving adequate care and protection, the Child and Family Agency shall: (b) having regard to the rights and duties of parents, whether under the Constitution or otherwise-(i) regard the welfare of the child as the first and paramount consideration, and (ii) in so far as is practicable, give due consideration, having regard to his age and understanding, to the wishes of the child; and (c) have regard to the principle that it is generally in the best interests of a child to be brought up in his own family.
Further provision to this effect is made in the Child and Family Agency Act 2013, which provides in section 9 that in the performance of all of its functions, the Agency (which operates under the name 'Tusla', and will be referred to as such for the remainder of the paper) shall have regard to the best interests of the child in all matters and shall ensure that consideration is given to the views of children. Thus, there is a clear legal obligation to give effect to both the best interests principle and the child's right to be heard in the context of voluntary care agreements. However, while voluntary care is captured by this general obligation, it is important to note that the obligation to 'give consideration' to the wishes of the child is somewhat weaker than the obligation stated in the context of court proceedings in Article 42A.4 of the Constitution, which requires that courts ascertain the views of children capable of forming them and to give them due regard in accordance with their age and maturity. Moreover, while the Child Care Act 1991 provides specific mechanisms in sections 25 and 26 through which the views of children can be ascertained during court proceedings, the Act is silent on what mechanism is to be used to achieve this aim in the context of voluntary care agreements. Literature has shown that the views of children are often not ascertained in court-based proceedings, notwithstanding the presence of specific statutory mechanisms (see Parkes, Shore, et al., 2015). The danger therefore arises that the absence of such mechanisms will make this tendency more pronounced in voluntary care agreements. Assessing whether this is the case in practice was one key focus of the Voluntary Care in Ireland Study.
As previously mentioned, the other main difference between care orders and voluntary care agreements is that the former are made by a judge, who may scrutinise the care plan devised for the child at the time of making the order (and, in some cases, may even schedule court-based reviews of the case at intervals thereafter (see Coulter, 2014)). By contrast, voluntary care agreements are concluded between the parent (s) and Tusla with no independent oversight at the point of entry to care. Thus, there is no safeguard at the outset to ensure that the placement and care plan are adequate to meet the child's needs.
As regards periodic review, the review process for all children in care (whether pursuant to a voluntary care agreement or a care order) is the same on paper. Regulations made pursuant to section 42 of the Child Care Act 1991 require that care placements be reviewed within two months of the initial placement; at least once every six months in the first two years of the placement; and thereafter not less than once in each calendar year. In conducting the review, the Agency shall have regard to any views or information furnished by the child, the parents of the child, the foster parents and any other person whom the Agency has consulted in relation to the review, as well as reports of visits to the child, school reports, and any other information which the Agency considers relevant. Reviews shall consider (among other things) whether all reasonable measures are being taken to promote the welfare of the child and whether the care being provided for the child continues to be suitable to the child's needs.
On their face, these regulations appear sufficient to discharge the obligation imposed by Article 25 CRC. However, these reviews are internal Tusla exercises; and while there may be input from external medical or other experts, or by members of the child's family or community, the outcome of the review is entirely at the discretion of the Agency. This contrasts with at least some care order cases where proactive judges schedule court-based reviews and may make directions as to the care of the child. The process also lacks provision for independent scrutiny by a Tusla social worker not involved with the child's case (like, for example, Independent Reviewing Officers in England and Wales (Beckett, Dickens, Schofield, Philip, & Young, 2016)). Limited research is available on the operation of reviews in Ireland, but one case study found that 24% of the sampled reviews were not conducted within the statutory guidelines (O'Meara, 2018).
The importance of effective reviews to prevent drift in voluntary care agreements has been acknowledged in previous research (Dickens, 2016). Concerns have been raised about the effectiveness of periodic reviews in voluntary care placements in Ireland, particularly in longterm voluntary care placements, which may be less visible within the system than care order cases involving more complex needs or more serious abuse or neglect. Geoffrey Shannon, the former Special Rapporteur on Child Protection, expressed concern that: [i]n view of the stretched resources of Tusla and the fact that urgent cases will demand the immediate attention of social workers with significant caseloads, the concern is that those children in voluntary care agreements may not always be visible. The voluntary nature of their placement and the consent of their parents may mean that their needs fail to be prioritised (Shannon, 2018, p. 153).
It was reported in 2020 that the Health Information and Quality Authority (HIQAa statutory body whose remit includes inspection of foster care services) had sent correspondence to Tusla which stated that inspectors had found 'no evidence … that the appropriateness of the child remaining in the voluntary care of Tusla had been reviewed and considered' (Power, 2020). The effectiveness or otherwise of reviews was a point of particular concern to both Shannon and HIQA in light of the fact that voluntary care agreements in Ireland have a potentially unlimited duration. HIQA has documented cases where voluntary care persisted for up to 10 years, and has raised concerns with Tusla that children were being 'subjected to voluntary consent for significant periods of time' with no efforts made to formalise the arrangement by securing a legal care order for a child, 'despite clear indications that they would not be reunified' with their parents (Power, 2020). The Child Care Law Reporting Project documented a case in which a voluntary care agreement had persisted for eight years; the District Court judge noted that agreements of this length could 'lead to complications' (Child Care Law Reporting Project, 2018). The potentially unlimited duration of voluntary care agreements in Ireland is unusual by comparative standards; for example, lengthy voluntary care placements are discouraged in Norway (Skivenes & Søvig, 2017) and Finland (Pösö & Huhtanen, 2017), and-as will be seen in the discussion section below-multiple jurisdictions place legislative limits on the duration and potential renewal of voluntary care agreements.
Having set out the potential strengths and weakness of voluntary care agreements as operated in Ireland in safeguarding and protecting children's rights, this paper will proceed to examine the reality of such agreements in practice. This will be achieved through the presentation of quantitative and qualitative data collected from legal and social work professionals with experience in voluntary care.

Methodology
The Voluntary Care in Ireland Study employed a mixed-methods approach consisting of an online national survey of social workers that collected quantitative data, followed by an in-depth exploration of themes identified in the survey and literature review through qualitative focus groups with social workers and individual interviews with solicitors. (Some reference is made to the quantitative data from the survey, but the focus in this article will be on the qualitative data, which built on the survey in more depth and provided richer data.) Permissions to undertake the qualitative phases of the study were secured from the Chief Social Worker of Tusla; the Principal Social Workers of each child protection and welfare social work team; and the managing solicitor of each participating Legal Aid Board Law Centre. Solicitors in private practice sought permission to participate from a senior partner. Additionally, each individual participant provided their own informed consent. The Tusla sample was limited to social workers and managers as they are the staff involved in voluntary care agreements. Ethical approval was secured from the social research ethics committee of the researchers' institution and from Tusla's research ethics committee.
Due to the very small literature base examining voluntary care, and the even smaller amount of Irish-specific literature, an online consultation was conducted with existing networks of front-line child protection and welfare social work staff, legal practitioners, academics and civil society organisations to assist with the development of the survey questionnaire. Responses to the consultation (n = 29) were analysed and combined with our analysis of the literature to finalise the questionnaire. The questionnaire was piloted, revised and circulated to social work practitioners and managers (max. possible sample n = 1,300-1,400) on behalf of the researchers by Tusla through its internal email list. The online questionnaire was live between the end of January and the end of March 2019. 243 responses (c.18% participation rate; 85% female, 13% male and 2% other) were received. 87% of the participants were aged between 26 and 55 years: 29% aged between 26 and 35, 36% aged between 36 and 45 and 22% aged between 46 and 55. 50% of respondents were front-line social workers, 11.5% were senior social work practitioners, 26% were social work team leaders, 11% were principal social workers and 1.5% were other senior managers. 20 solicitors (11 female and 9 male) took part in semi-structured interviews in seven counties across Ireland. 10 worked for the Legal Aid Board representing parents, seven worked in private law firms representing the Child and Family Agency and three worked in private practice with experience of representing both parents and guardians ad litem. For all counties, except one, we interviewed solicitors who represented parents and also solicitors who represented the state. This selection strategy, in conjunction with at least one focus group with social workers and social work managers (six focus groups with social work teams were conducted across seven counties, with 26 participants) ensured that we included a very large sample in each county. For these counties, the solicitors interviewed represented between 80 and 100% of all of the eligible legal practitioners who were active in this field of practice. Interviews and focus groups were based on an interview guide that asked open questions seeking participants' views on the strengths and weaknesses of voluntary care; parental consent, and the support available to parents; the welfare and best interests of children in voluntary care; and child participation. In addition to these open questions, participants were asked for their views on a number of specific reforms that had been proposed in previous reports or which were based on comparative experience, including placing a time limit on voluntary care placements; requiring a notice period before consent to voluntary care could be revoked; and requiring the assent of mature children to voluntary care placements. The transcripts were coded and analysed in nVivo qualitative analysis software using thematic analysis.
The main limitation of the study was that the views of children and young people with experience of voluntary care have not yet been included, nor have the views of parents with experience of placing their children in voluntary care. Extensive efforts were made to recruit participants from these groups through various channels, but it proved impossible to recruit a sufficient number to have a usable sample. The impact of COVID19 and the inability of the research team to meet with participants in person exacerbated this difficulty, as children and parents proved even more reluctant to conduct interviews via video conference than they were to conduct them in person. We hope to re-visit this phase of the research in future once face-to-face interviews are possible again. A further limitation was the limited nature of some of the quantitative statistics collected and published by Tusla; this point will be returned to in the discussion section below.

Strengths of voluntary care in safeguarding children's rights
Participants in the Voluntary Care in Ireland Study shared the view proposed in the extant literature that voluntary care may allow resources that would otherwise be spent on expensive court proceedings to be channelled into early intervention that might prevent children coming into care in the first place: … the costs of the court process is quite expensive. Whereas … if some of the money spent in the litigation could be put into the early intervention family support at a community level to prevent children coming into care and to support parents, obviously it would be well worth it. (Solicitor 14 (Parents/GALs/Children), County B) This is in line with the requirement set down in both the CRC and the European Convention on Human Rights (ECHR) that children be cared for by their own parents wherever possible, with alternative care being a last resort. The data from our study also suggests that professionals feel that voluntary care is conducive to a good working relationship between the parents and the social workers, which is in turn good for the child. Previous research (O'Mahony, Burns, et al., 2016;O'Mahony, Burns, Parkes, & Shore, 2016;O'Mahony, Parkes, Shore, & Burns, 2016, Burns, O'Mahony, et al., 2017Burns, Pösö, et al., 2017;Burns, O'Mahony, Shore, Parkes, 2018;O'Mahony, 2018) has found that the adversarial nature of District Court proceedings makes it difficult for social workers to effectively work with parents thereafter. As voluntary care agreements are agreed rather than imposed, they avoid the highly adversarial and stressful dynamic commonly seen in court, as there is no need to demonstrate parental neglect or abuse to justify a care order: … it can create a good working relationship between parents and social workers … when something goes into court and you have to hammer the parents it can put a strain on working relationships. (Solicitor 2 (Parents), County A) Social workers reported that voluntary care supports a collaborative relationship with parents which benefits the child in care. This is illustrated by this quote from a focus group with a social work team: I think that in order to preserve relationships with parents Section 4 [of the Child Care Act 1991, which governs voluntary care] is much easier to do that. You can work with parents. You can be on their level. Once you go into court and people get lawyers it just turns into a very different process. And, you know, you could have lawyers, barristers-[if] there's two parents, you could have two solicitors, two barristers, a guardian ad litem with their legal representation. It just becomes so adversarial, so complex, and the relationship that you've worked hard to build up with parents can just go very quickly in that process. (Social Worker Focus Group 1, County B) Through consenting to voluntary care, parents acknowledge that there are difficulties to be addressed. In contrast to engaging in a long and stressful court process, professionals view voluntary care as allowing parents and social workers to begin focusing on addressing and resolving these difficulties. This, they suggest, makes family reunification more likely in the long-term and reduces traumatisation of the child: So, for me, it is about reducing trauma … Once we have taken care orders and assumed parental responsibility, we then set those children up for a pathway in care that may not serve them very well; whereas maybe if we hadn't alienated the parents in the first place, we might be able to get them back home easier. (Social Worker Focus Group 6, County C) … a lot of the time the child is completely conflicted when they see their parents reacting to the social workers or, you know, shouting and roaring at them … Whereas if you see a positive working relationship between your parent and the social worker, well, then there's no need for you to worry and you can just get on with being a child and knowing that things are, you know, progressing, because in those cases it is usually a scenario where there's going to be a piece of work done which will hopefully allow reunification. (Solicitor 5 (Tusla), County B) Participants also expressed the view that potential reunification is made easier by the fact that in voluntary care agreements, parents can retain important decision-making functions over issues relating to their children's education or medical treatment, as a voluntary care agreement in Ireland does not transfer these decision making powers to Tusla: In this regard, it should be noted that ECHR case law (as approved by the Irish courts) specifically stipulates that any placement in care should in principle be viewed as a temporary arrangement, with the ultimate goal being family reunification. 1 As such, if voluntary care agreements are more conducive to reunification than involuntary removals, this is an important reason for using them in suitable cases.

Lack of independent oversight
As stated earlier, Article 25 CRC requires that children in state care receive periodic review of their placement. It was noted earlier that some District Court judges engage in the practice of court-based review (Coulter, 2014). Where this occurs, it was highlighted by some study participants as an advantage of care orders over voluntary care agreements: … sometimes when the child's in voluntary [care] things just go on, and nobody's putting any pressure on the social workers to do anything and sometimes things aren't being done properly. But once it gets into the court system then you find that the judge is looking at this, that and the other and the solicitors are looking at this, that and the other and making sure that things are done for the child or for the parent … and there's more of a defined plan, I suppose. (Solicitor 10 (Legal Aid Board), County B) It was also noted earlier that concerns have been expressed about the effectiveness of internal Child and Family Agency child-in-care reviews in the context of voluntary care agreements, particularly given that the Irish system does not limit the duration of such agreements and they can in some cases persist for many years. Participants in our study shared the concern that voluntary care agreements suffer from comparatively ineffective periodic review. There was a broad consensus among solicitors in our study that internal Child and Family Agency child-in-care reviews are insufficient in voluntary care due to a lack of independent oversight: Now, there are some cases where … there has been a kind of drift from a parent's point of view or a change of staff with the social worker, where there isn't the same management happening of a case, more through effluxion of time than anything else. And in those situations the regulations, the requirements that reviews take place in a certain period of time, they can be missed, there's no doubt about it. I've been aware of situations where-missed by months as opposed to by years, you know. (Solicitor 15 (Tusla), County B) Often these arrangements can be quite sort of benign in that the child may go to a relative carer who then facilitates contact. So it's a low enough workload, if it works well, for the social worker. And we all know-I'll go back in and the first cases on my desk that'll get the attention are the ones that people are making noise and demands and writing solicitor's letters. So I do think there should be a built-in mechanism, you know. I know the care planning derives from the statutory reviews but I just wonder sometimes in terms of the degree of care planning are they getting everything that a child in [courtordered] care might get. (Solicitor 1 (Tusla), County A) As already stated, a lack of independent oversight was linked in the data to resource provision, but there were also concerns around 'drift' in placements which continue for lengthy periods: … when a child's in voluntary care there's no one reviewing whether that child-in-care review took place. Whereas at least if it's in a care order the Child and Family Agency have to answer to the court if a child-in-care review did or did not take place, or more often than not there'll be a solicitor in the case who can advise the parents that this 1 See Johansen v Norway (17383/90, 7 August 1996)  Some social workers defended the efficacy of the review process: … you would get a total oversight of all aspects of the child's care with that. And like we've an obligation for them to be held and for all the children regardless of their care order to have a care plan. So, I mean, and they are subject to the same statutory guidelines as [care orders] … (Social Workers Focus Group 3, County D) However, others acknowledged the weaknesses in the system: … the review process here is very overloaded, very under-resourced … maybe it is the children in care or the children subject to voluntary care orders should be given priority because that's the only place they're going to be reviewed. (Social Workers Focus Group 2, County A) Thus, as currently operated, the review system for voluntary care agreements in Ireland raises concerns regarding the quality of compliance with the right to periodic review under Article 25 CRC, as well as a potential issue of differential treatment contrary to Article 2 CRC as between children in voluntary care and children subject to care orders.

Unlimited duration
The concerns expressed by study participants around the effectiveness of reviews of voluntary care agreements were strongly connected to the potentially unlimited duration of the placements. Participants indicated that the involvement of a court at the outset puts care placements on a better footing due to the involvement of a judge and legal representation for the parents (and, as will be seen below, the involvement of a guardian ad litem), and the fact that these parties may intervene subsequently focuses minds: … where there are court proceedings, there's a legal team involved who will be doing things like, you know, [in] a one-year care order I would automatically looking for a six-month review. The client has access to me during the course of the one-year care order, so if issues arise in relation to the placement or the access or whatever might arise during the course of the care order they have a mechanism to bring the matter before the court … And a court order I think probably focuses the mind maybe a little bit more than a voluntary arrangement. (Solicitor 13 (Legal Aid Board), County A) Voluntary care agreements running to several years in duration were thus viewed as being more susceptible to drift than care orders: I think a big issue for us was we would have brought children into voluntary care, we'd have no time limit on it … We have probably got some children in long-term voluntary care that would not meet the threshold for an interim care order or for legal care orders in the court and we have had to come back and review them and parents just haven't challenged us in the court … that's a fault in the system really for allowing voluntary care orders to go on and on and on.

(Social Workers Focus Group 3, County D)
A further finding of the study, which is reported in a separate paper (O'Mahony, Brennan, & Burns, 2020), was that the absence of time limits was problematic from the perspective of parental rights due to question marks over the extent to which parental consent to voluntary care agreements is always freely given and fully informed.
In 2018, the Special Rapporteur on Child Protection stated that 'it is not ideal for any child to be in [voluntary] care for an indefinite period of time, without any certainty or stability' and recommended that a maximum duration of twelve months should apply to voluntary care placements in Ireland (Shannon, 2018, p. 153). This recommendation was put to participants in the study, and it generated a mixed response. The majority of participants favoured a limit, with solicitors being particularly supportive; but different views were expressed about how long the maximum duration should be: ….for certainty for that child, for planning and everything that is involved, I think it would be important because … it would put pressure on the Agency to make decisions at an early stage regarding the planning for the children. (Solicitor (Tusla) 18 Counties E & F) There's also a problem, which the courts are recognising, of voluntary care agreements going on for maybe ten, fourteen years or whatever … I think, yeah, 12 months, because that would give a defined period for the parent to address their issues and if they can't address them within that time, well, then you can take further steps. (Solicitor 10 (Legal Aid Board), County B) I think it should be time-limited. I think there's an argument for it. And I would say no more than three months … I think 12 months is far too long. (Solicitor 3 (GALs and Parents), County B) A sizeable minority of participants disagreed with the concept of a maximum duration on the basis that 12 months is not long enough, or that transitioning to a care order might be damaging in cases that are working well. Social workers were more likely to express this view: I don't believe that 12 months is long enough. I think if people are badly affected by alcohol or drug misuse their brain doesn't work like other people and it takes a long time. I think the minimum would be two years for me. (Social Workers Focus Group 3, County D) I do think in the longer term things can drift on … That said, my slight concern … is that if we then-we've had agreement and mum's never tried to pull the child out, says that she can't cope. But in going to court … would that then make mum think that she's never going to see the child? (Social Worker Focus Group 5, Counties E & F) …if mum and dad are consenting till the child is 18 why would you need a care order? It might upset the applecart if there was a care order obtained. (Solicitor 16 (Tusla), County D) It was noted above that Ireland's approach of allowing for potentially open-ended voluntary care agreements is out of line with jurisdictions such as Finland and Norway. Given the impact this has in exacerbating concerns around the effectiveness of reviews, there is a strong case for adopting time limits, with the possibility of renewing the agreement if an effective review takes place first. This is the approach taken in numerous jurisdictions, and this will be examined further in the discussion section below.

Potential instability of placements
The potentially open-ended nature of voluntary care agreements also contributes to a separate concern that was flagged in the pre-existing literature: namely, the risk to child welfare that is posed by the fact that parents in Ireland can cancel voluntary care agreements at any time and demand the immediate return of their child. The Child and Family Agency could make an application to the court for an emergency care order if it is concerned that the child should not return home; but depending on the timing, it may not always be possible to secure an order in time to prevent the transfer of care.
Articles 3 and 20 CRC require that children receive such care and protection as is necessary for their well-being, including alternative care where this is necessary, with the best interests of the child being a primary consideration. Similarly, the ECHR case law has consistently held that state authorities have a positive obligation to take steps to protect children from risks of ill-treatment of which the authorities are or ought to be aware, including removing the children from the family home if necessary. 2 As such, international human rights law clearly requires states to take measures to ensure that children are not allowed to remain in or be returned to the care of their parents in circumstances where their parents are not capable of adequately caring for them. Voluntary care agreements pose a risk in this regard, and social workers were particularly concerned about the potentially negative impact on the child of this situation: I think especially if you have a relative placement as well, because they'd often ring the relatives and say, well, I'm taking it back, I don't want it. And if that happens over the weekend we're kind of caught on what response just they can give, I suppose, if the out-ofhours aren't aware of the situation, do you know what I mean… That can be scary, yeah … the parent could actually come in and say, 'I want my child back today, give me the form, I want to sign the form,' and that just puts you in a very difficult position. (Social Worker Focus Group 4, County B) … you can have a parent that is, like, 'I've gone to a counsellor and now my counsellor says-'. And they'll come rocking into the office and demanding this, that and the other when, you know, you're like, 'whoa, where's this all come from?' And that can be scary for a kid, you know. Because I know a parent that used to do that often, you know, and then they'd cool off and then they might disappear and you wouldn't see them for four or five weeks and then they'd come back again. (Social Worker Focus Group 1, County B) As a response to concerns around return of a child to an unsafe environment, a 72 h notice period was proposed by the authors to the study participants as a potential safeguard for revocation of parental consent. Participants provided mixed views in response to this proposal. The majority of participants supported it as a buffer to protect the child: The obvious difficulty with Section 4 is a parent at 5 O'Clock on a Friday can ring up and say, 'Right, I revoke my consent,' and the CFA have no other option but to return the child. Good luck getting a section 13 [emergency care order] at that stage, you know. So that's the real kind of drawback … The solution is if the parent has had legal advice and has signed a voluntary arrangement, that you have a minimum period, you know, for revocation. One of the social work teams in this county reported having received legal advice that such a clause would be viable, but that a court later said the parent could have the child back any time: I got legal advice about whether that was possible and the legal advice was that yes, of course you just write that into the consent form so that when the person is giving consent that's part of it. But then there were different legal views about it and the courts took a different view about it and I was told in the end just the parent can take their child back. (Social Workers Focus Group 1, County B) However, this arose in a context where the notice period incorporated into the agreement did not have any basis in statute, and would seem to conflict with the terms of section 4(2) of the Child Care Act 1991. The position would be quite different if the notice period were to have a statutory basis. Overall, the possibility of a child being returned to an unsafe home environment appears incompatible with the obligations set down in the CRC and ECHR relating to the child's best interests and ensuring that the child receives alternative care where his or her wellbeing requires this. Making provision in law for the incorporation of 72 hour notice periods into voluntary care agreements is a proportionate measure that balances the rights of both children and parents, and we will return to this issue below.

Weak mechanisms for ascertaining children's view
It was noted above that the statutory provisions governing the ascertainment of the views of children in voluntary care placements are weaker than those governing child care court proceedings. Our study participants confirmed that in practice, children in voluntary care fare worse in this regard than children in care pursuant to court orders, which appears contrary to both Articles 2 and 12 CRC. The majority of lawyers ascribed this tendency to the absence of a guardian ad litem: … possibly the voice of the child isn't really heard. So there's no-like in the court system you have your guardian ad litem with an independent person looking at that. So that's another thing that could be done. (Solicitor 17 (Legal Aid Board), Counties E & F) I think that is one of the perhaps downsides of voluntary consent is the independent voice of the child is often missing from that sort of roundtable discussion we spoke about, you know. I think if you have a child who has been repeatedly placed on sort of voluntary consent arrangements then it should give rise to who is speaking independently on behalf of the child in those situations. (Solicitor 1 (Tusla), County A) It is evident from the above that solicitors place a high value on the independence of the guardian ad litem and their ability to advocate on behalf of the child. Children in voluntary care must rely on social workers to perform this role; and unlike a guardian ad litem, social workers are not independent as they are employed by Tusla. In our focus groups, social workers described how they employ a range of methods to obtain the views and wishes of children: You know, we are also a voice of the child as well, too … There's a hundred different ways … in Signs of Safety you can do your Three Houses. You know, we talk to kids all the time. We bring kids out. We do individual work. Social care workers. Observations. Talk with other professionals. (Social Worker Focus Group 3, County D) However, many agreed the absence of a guardian ad litem is a disadvantage for the child in voluntary care in terms of representation for the child: … they can't be appointed a GAL because they're not subject to court proceedings. And like GALs are constantly advocating for different assessments for children … that's a large proportion of their work is going to the court and asking for particular assessments for children. The lack of a representative for the child such as a guardian ad litem was also said to be to the detriment of periodic reviews in voluntary care: …the other perhaps vulnerability for children who are on rolling kind of voluntary consent agreements is that they are somewhat not as represented in that discussion as they should be … It doesn't even have to be a guardian ad litem, but in an ideal world some sort of body or some sort of oversight of, you know, how long these arrangements have been continuing in respect of a child. (Solicitor 1 (Tusla), County A) In a number of comparable jurisdictions, including England and Wales, New Zealand, South Australia, Western Australia, Ontario and Finland, the law stipulates that children may or must consent to voluntary care agreements once they reach a certain age (ranging from 12 to 16). This possibility was put to our study participants. 11 out of 20 solicitors and the majority of social workers did not favour putting such a weighty responsibility on the child, noting that the child may not be well placed to make such a decision, due to their age, level of understanding or the trauma they have suffered; and that the child may have conflicted emotions or loyalties in relation to the parent. Some participants expressed the fear that a requirement for child consent would push a lot of voluntary care cases into the courts. However, the vast majority of participants still favoured facilitating the child to express their views: We have to be careful … in an effort to redress the voiceless child that we then burden them with matters that are, you know, very, very difficult even for the adults to navigate. (Solicitor 1 (Tusla), County A) I think that is a dangerous road to go down. I am absolutely in favour of the voice of the child, but I think that careful consideration has to be given to the weight that's going to be given to the child's views … What I actually think is more important is how the scenario's explained to children, in a child-friendly and appropriate way, depending on the children's own ages and the capacity to understand. (Solicitor 13 (Legal Aid Board), County A) Putting all the responsibility on a child, which is really a very questionable and inappropriate thing to be doing. And it definitely could be significantly abusive … (Solicitor 4 (Parents), County A) It's too much to put on them, you know. (Social Worker Focus Group 4, County B) A sizeable minority of solicitors (9 out of 20) and a small minority of social workers were in favour of older children being required to consent to voluntary care agreements. Reasons given included the fact that 16year-olds can consent to medical treatment; that it would be in line with the constitutional amendment on children; and that there is a risk that teenagers who do not want to be in a voluntary placement might 'vote with their feet' and abscond. There was some variation on the age at which respondents felt it would be appropriate to consent, with replies ranging from 12 to 16. On the question of age, they said it would depend on the particular child. Participants were strongly in favour of the child's views being sought more generally: I think I'd definitely feel happy with a 15-year-old being asked. I'm just thinking of a case where we have-there is a teenage child and the CFA took an ICO [interim care order] application but it was adjourned on the basis the mother gave undertakings [that] the child go to school. And there was sort of neglect issues and stuff like that. But that could have been probably circumvented if that type of process was available, because if they had the consent of the child, that would have probably added more. They would have consulted with the child … so I can that that would be beneficial. I still think 12 is quite young. (Solicitor 17 (Legal Aid Board), Counties E & F) I suppose it makes sense. I mean, physically you can't actually get a child in your car. I've had a few where I have had the difficult ones and they've said, okay, we'll all agree, and you've finally got a placement. You've got a placement for this difficult 15-year-old … -'I'm not going'-and then it kind of calms down …. I'd never force taking a child in. (Social Worker Focus Group 5, Counties E & F) The emphasis placed on child participation and evolving capacities in the CRC suggests that there is a strong case for Ireland to follow the approach of multiple other jurisdictions by stipulating that the assent of older children to voluntary care agreements must be obtained in addition to the consent of the parent(s). We will return to this point below.

Inferior resource allocation
As noted above, Article 2 CRC requires that children in voluntary care should receive the same level of care as children placed in care pursuant to a court order. In this light it is noteworthy that the broad consensus from participants in our study was that children on care orders get access to better resources than children in voluntary care, with 71% of social workers who participated in our national survey either agreeing or strongly agreeing that this is the case. The reason most highlighted in this regard was the independent oversight provided by the courts: It is a two-tier system. There's the voluntary care and then there's the care orders. The kids on the care orders get the funding. They get the oversight there. (Social Work Focus Group 4, County B) The lack of a guardian ad litem in voluntary care was identified by solicitors as a disadvantage for children, as a guardian ad litem will often make recommendations for access to resources and services: … is the child in voluntary care getting as much without a GAL putting in recommendations in a report to get A, B, C, D and E? I wonder. I doubt it, you know. I have my doubts about that. (Solicitor 9 (Tusla), County C) …the child who has gone down the 'court' route for want of a better word and the guardian has been appointed then they could end up with more resources and be better looked after because they have somebody fighting their corner … in voluntary care there is probably nobody. (Solicitor 20 (Tusla), County G) Other participants highlighted the fact that Section 47 applications (for directions from the Court regarding the welfare of a child in care) are not made in voluntary care cases, since parents usually do not have legal advice and again, there is no guardian ad litem who could make one: … realistically under Section 4 [i.e. voluntary care cases] who's going to be bringing Section 47 applications? There's no guardian. The parents are unlikely. The child's not going to bring the application themselves. (Solicitor 11 (Legal Aid Board), County B) Another factor that was suggested was that voluntary cases may be seen as low maintenance cases due to their consensual nature which often involves relative placements where the onus is on family members to address children's issues: … even when you are on a statutory care order there is a battle to try to get these services for children in care and I just wonder then are children in voluntary consent that little bit quieter; their problems maybe go either managed by family or unnoticed by family. (Solicitor 1 (Tusla), County A) Participants stressed that this tendency continues even after the child leaves care: And aftercare planning as well. All the kids that are coming through voluntary they're left last for aftercare services. They're bottom of the list for everything. (Social Workers Focus Group 4, County B) Thus, it seems quite clear from the above findings that voluntary care agreements raise significant concerns vis-à-vis discriminatory treatment contrary to Article 2 CRC. These concerns are linked to the absence of independent oversight and inadequate mechanisms for ascertaining the views of the child, as discussed above. Thus, as will be explored further below, reforms which address these issues could also help to mitigate the risk of discriminatory resourcing of care placements.

Comparative analysis and recommendations
The limited extant literature prior to the Voluntary Care in Ireland Study suggested that there were a number of areas worthy of further research in voluntary care, with one of those being the rights of the child and to what extent these were being protected (Corbett, 2018;Hill, 2017;Shannon, 2018). The data from this study agrees with previous suggestions that voluntary care can bring certain advantages from a children's rights perspective; but at the same time, the views and experiences of social work and legal professionals substantiate concerns around safeguarding children's rights in the loosely regulated Irish system and confirm the need for reform in Ireland. The sparse legal framework provided by section 4 of the Child Care Act 1991 gives rise to numerous risks to the rights of children and parents involved in voluntary care agreements, including the absence of independent oversight; the potentially unlimited duration of placements; potential instability; weak mechanisms for ascertaining the views of the child; and inferior resource allocation by comparison to children who are subject to care orders. Taken together, these issues raise concerns that the voluntary care system in Ireland is either non-compliant or merely partially compliant with the obligations set down in Articles 2, 3, 12, 20 and 25 of the CRC. (The proposals for reform that follow are specific to the Irish system; other jurisdictions with different systems may have different strengths and weaknesses that suggest different reforms. The findings of this study highlight the importance of conducting detailed empirical research to inform any such process.) Section 4 of the Child Care Act 1991 lacks several of the safeguards seen in various comparable jurisdictions. These include limits on the duration of voluntary care agreements; limits on whether, or how many times, an agreement can be renewed; provision for legal advice for parents; provision for independent oversight or review; provision for notice periods before an agreement can be cancelled, and provision for child participation in the consent process. Some of these safeguards interlock with each other, and strong provision for one safeguard would allow for less strong provision to be made for another. For example, concerns around lengthy voluntary agreements would lessen if there was independent oversight of the operation of the agreement to prevent drift. Conversely, the absence of such oversight would strengthen the case for voluntary care agreements to be limited to a maximum duration, with an application for a care order being required after a set period.
Based on the findings of the Voluntary Care in Ireland Study, we argue that section 4 of the Child Care Act 1991 should be completely overhauled to provide for safeguards that draw on some of the best practice seen internationally (while adapting them as necessary to the Irish system). Reform of section 4 should be accompanied by a number of policy and practice measures designed to ensure that the legislative framework has the desired effect. The following specific reforms are recommended:

Limitations on duration and renewal
Ireland is not the only jurisdiction to allow for open-ended voluntary care agreements; England and Wales is another example. Nonetheless, statutory limits on the duration and renewal of voluntary care agreements are common in comparable jurisdictions, although they vary considerably in their form. Western Australia limits voluntary care agreements to a maximum of three months per agreement, with extension possible once the agreement operates for a total of not more than six months (Children and Community Services Act 2004, s 75(9)). Victoria makes provision for short-term agreements which have a maximum duration of 6 months and may be renewed up to a maximum total period of 12 months; while long-term agreements have a maximum duration of two years, and may renewed more than once with the approval of the Secretary to the Department of Human Services of the Government of Victoria (Children, Youth and Families Act 2005, ss 135-150). In Ontario, voluntary care agreements are known as 'temporary care agreements' and have a maximum duration of 6 months. They can be extended up to 12 months; but if new voluntary agreements are made subsequently, the outer limit of how long a child can be in care under such an agreement is 12 months for children under six years of age, and 24 months for older children (Child, Youth and Family Services Act 2017, ss 75(5) and 75 (6)). In New Zealand, the law distinguishes between short term placements (which have a maximum duration of 28 days and may only be renewed once) and extended placements (which have a maximum duration of 6 months for children under 7 years of age or 12 months for older children, but may be renewed more than once if approved by a family group conference) (Children and Young People's Wellbeing Act 1989, ss 139 and 140). Open-ended voluntary care agreements of unlimited duration are not permitted in any of the jurisdictions mentioned; and the common thread is that once the outer limit of duration and renewal has been reached, it becomes necessary to obtain a court order if the child is to remain in out-of-home care.
Limits on duration and renewal may lead to the diversion of a proportion of voluntary care cases into the court system, although the size of this proportion will depend on the form that such limits take. Cases which are suitable for voluntary care should be allowed to remain in that system and to enjoy the comparative benefits of voluntary care over court proceedings discussed above. If risks to children's rights in voluntary care can be addressed through the provision of safeguards within the framework for voluntary care, this is preferable to diverting cases into the more costly and adversarial court system. The findings presented above in relation to the advantages of voluntary care agreements in Ireland suggests that an absolute upper limit on the duration for which a child may remain in voluntary care is neither necessary nor desirable-provided that appropriate safeguards are in place to mitigate the risks associated with a lack of independent oversight. Drawing on the comparative examples set out above, it is recommended that such safeguards should include the stipulation in the Child Care Act of a maximum duration for individual voluntary care agreements, and of a requirement of independent review before an agreement may be renewed.

Reviews
The data from this study confirms that the absence of independent oversight of reviews may put children in voluntary care in a disadvantaged position compared to children on care orders. Legal professionals in particular expressed the need for some provision for independent oversight to be made in voluntary care, and social work teams spoke about onerous workloads which placed child care court proceedings at the top of task lists while voluntary care cases received less attention. The need for additional scrutiny of the cases of children in voluntary care, to include child and family support and ongoing efforts towards reunification, has been a finding of previous research in England and Wales in relation to Section 20 of the Children Act 1989 (Lynch & Boddy, 2017;Masson, 2018). Enhanced provision is also needed for the voice of the child and their views and wishes in voluntary care; in the absence of an independently appointed guardian ad litem, there is no standardised model for representation of children's views in voluntary care in Ireland. (This difficulty is not limited to voluntary care; the limited research available indicates that child participation in reviews is weak and in need of strengthening for all types of care (O'Meara, 2018)).
Our recommendation is that a formal review should take always place before the expiry of an agreement, regardless of whether the duration of the agreement was for the maximum permitted or for a shorter period. The review would assess the feasibility of the child returning home, or (if this is not feasible) whether the child should remain in voluntary care or whether transition to a care order is appropriate (depending on the level of co-operation evident between the parents and Tusla). The care plan for the child should also be thoroughly reviewed. In order to address concerns related to children's rights highlighted above, this process should provide for independent oversight so as to ensure adequate provision is made for voluntary care placements and to avoid drift; and should include a mechanism designed to ensure effective child participation in reviews. Children in voluntary care should have access to an advocate who would, inter alia, participate in reviews. Children who are too young to work with an advocate should have access to a guardian ad litem for the review process to ensure that independent representation is provided for the child's views and best interests.
Reviews should be chaired by an independent person who has not been employed by Tusla recently, and who has at least five years' experience of child care (e.g. retired judges, solicitors, guardians ad litem, independent social workers). This chairperson should have the power to make a binding recommendation as to whether the agreement should be renewed or not. In the event of non-renewal, Tusla would then apply professional judgment as to whether an application for a care order is necessary. This recommendation is aimed at ensuring that the chairperson enjoys institutional independence (by not being answerable to Tusla management); perceived independence (so that parents and children do not perceive them as being a Tusla insider); and effective independence (so that they are actually capable of influencing events) (Dickens, Schofield, Beckett, Philip, & Young, 2015).

Notice period
As explored above, the fact that parents can revoke their consent to voluntary care at any time and demand the immediate return of their child creates a risk that in some cases, children may be returned to an unsafe home environment, even if only for a short period of time. This appears contrary to Articles 3 and 20 CRC. It is recommended that the Child Care Act 1991 should be amended to provide for a 72 hour notice period before a voluntary care agreement can be cancelled by a parent in order to mitigate the risk that a child may be returned to an unsafe home environment before there is an opportunity to apply for and obtain an emergency care order. An example of a statutory requirement for a notice period can be seen in New Zealand, where the governing statute provides that every voluntary care agreement shall 'specify the manner in which it may be terminated and, unless so specified, shall provide that the agreement may be terminated by either party on giving 7 days notice [sic] in writing' (Children and Young People's Wellbeing Act 1989, s 146(1)(c)). Flexibility could be provided by stipulating that Tusla may waive this period if it is in the best interests of the child to do so.

Assent of mature children
In order to comply with Article 41 of the Constitution (which protects the rights of the family, and has been interpreted as giving parents a strongly-protected right to care for and make decisions for their children), Ireland could not allow children to sign themselves into care over the objection of their parents. However, in cases involving mature children, there are strong reasons of both a principled and practical nature to require that parental consent be supplemented by the assent of the children being placed in care. In such a model, a voluntary care agreement could not be concluded without the agreement of both children and parents; children would not be able to conclude a voluntary care agreement without their consent of their parent(s), but the same would be true in reverse. If a child did not assent, and Tusla remained of the opinion that the child's welfare required placement in care, an application for a care order would be necessary in the same way as it would be if the parent(s) did not consent to voluntary care.
Such a measure-versions of which are in place in multiple comparable jurisdictions-would be in keeping with the principle recognised in the CRC that children have evolving capacities and should have a greater say in decisions affecting them as they mature. It would also mitigate the practical concern expressed by participants in the study that teenagers who object to a voluntary care placement may 'vote with their feet' and abscond. Children and young people who have been asked for and given their assent to the placement have an element of ownership over the decision that makes this outcome less likely. Conversely, where a child refuses to assent, it seems ill-advised to proceed with the placement without first going through court proceedings in which the child would have a full opportunity to participate.
This raises the question of the age at which assent of the child should become mandatory. Different jurisdictions take different approaches to this issue. In England and Wales, it is not mandatory to obtain the assent of children to voluntary care. However, accommodation may be provided to children from the age of 16 (Children Act 1989, s 20(5)); parental objection to such accommodation will have no effect where a child aged 16 or over agrees to being accommodated (Children Act 1989, s 20 (11)). In Ontario (Child, Youth and Family Services Act 2017, s 22 (1)), children can consent to voluntary care from the age of 16. In South Australia, a child of 16 or older can initiate negotiations for a voluntary care agreement and must consent to that agreement (Children and Young People (Safety) Act 2017, s 96.). Other jurisdictions set the age much younger; in New Zealand (Children and Young People's Wellbeing Act 1989, s 144.) and Finland (Pösö & Huhtanen, 2018) children over 12 must consent. Since the proposal being made here is for a requirement of assent that is in addition to parental consent rather than standalone consent given by the child (and is thus more modest than some of the comparative examples given above), the lower age of 12 would be preferable. But at a minimum, reform of section 4 should require the assent of the child from the age of 16 so as to bring the law on this issue into line with international best practice and with Irish law on other issues (including, for example, the fact that 16 year-olds can consent to medical treatment).

Conclusion
A comprehensive review of the Child Care Act 1991 is currently underway in Ireland. This provides an opportunity for enhancing safeguards for the protection of children's rights in voluntary care. However, in order to reach its potential, more detailed statistics should be compiled and published by Tusla so as to inform the review of the Act (the recent adoption by Tusla of a new information management system should assist in this task). This should include issues such as placement types; duration of stays in voluntary care; frequency of reviews; reason for terminating voluntary care agreements; and outcomes when agreements are terminated (i.e. family reunification or transition to care order). Future comparative research is also recommended; the dearth of literature to date on voluntary care is such that many jurisdictions may have as-yet unidentified difficulties in their systems, and there is significant potential for law, policy and practice in each jurisdiction to be informed by legal frameworks for voluntary care agreements seen elsewhere. In the meantime, jurisdictions in which no empirical data on voluntary care is yet available may have much to learn from Ireland, which demonstrates the risks that attach to a light-touch regulatory approach.

Declaration of Competing Interest
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.