The formalized framework for decision-making in child protection care orders: A cross-country analysis

Abstract

Care orders within the child protection system are some of the most invasive interventions a state can make. This article examines the discretionary space governments set out for child protection workers when they prepare care orders. We analyse the formalized framework for these decisions in England, Finland, Norway and the United States. We focus on knowledge, timelines, how children and parents are involved and accountability. We find that Norway and Finland have highly de-regulated systems with wide discretionary space, whereas England and the United States are highly regulated systems with narrow discretionary space. The United States differentiates itself with relatively little parent and/or child involvement in decision-making. England and Finland do not have defined deadlines for terminating the process, and Norway has few directives on what information to collect. Such differences will influence the quality of decisions as well as the principles of the rule of law.

Full text

Journal of European Social Policy
2015, Vol. 25(4) 366 –378
© The Author(s) 2015
Reprints and permissions:
sagepub.co.uk/journalsPermissions.nav
DOI: 10.1177/0958928715594540
esp.sagepub.com
Journal Of
European
Social Policy
The formalized framework for
decision-making in child protection
care orders: A cross-country analysis
Jill D Berrick
UC Berkeley School of Social Welfare, USA
Sue Peckover
Sheffield Hallam University, UK
Tarja Pösö
University of Tampere, Finland
Marit Skivenes
University of Bergen, Norway
Abstract
Care orders within the child protection system are some of the most invasive interventions a state can
make. This article examines the discretionary space governments set out for child protection workers
when they prepare care orders. We analyse the formalized framework for these decisions in England,
Finland, Norway and the United States. We focus on knowledge, timelines, how children and parents
are involved and accountability. We find that Norway and Finland have highly de-regulated systems
with wide discretionary space, whereas England and the United States are highly regulated systems with
narrow discretionary space. The United States differentiates itself with relatively little parent and/or child
involvement in decision-making. England and Finland do not have defined deadlines for terminating the
process, and Norway has few directives on what information to collect. Such differences will influence the
quality of decisions as well as the principles of the rule of law.
Keywords
Child protection, comparative country analysis, decision-making, discretion, involuntary care orders
Corresponding author:
Marit Skivenes, Department of Administration and Organization
Theory, University of Bergen, Christies Gate 17, 5007, Bergen,
Norway.
Email: marit.skivenes@uib.no
594540ESP0010.1177/0958928715594540Journal of European Social PolicyBerrick et al.
research-article2015
Article
Introduction
One of the most invasive and consequential decisions
a state can make is to involuntarily remove a child
from his or her parents’ care. As such, these decisions

Berrick et al. 367
must be of the highest quality possible. Despite the
gravity of these determinations, there are notable dif-
ferences across child protection systems, and within
and between actors in the system, allowing for con-
siderable discretion and thus the potential for inequi-
ties for children, parents and families. Discretion in
state interventions raises questions about the princi-
ples of the rule of law, democratic control and legiti-
macy (Dunn et al., 2007; Elster, 1989; Molander
et al., 2012; Piper, 2000). In modern democratic
states, it is usually the court that makes decisions
about care orders, but it is the child protection sys-
tem1 and front-line child protection workers that have
responsibility for the day-to-day interactions with
children and families – and to suggest and carry
through preparations for a care order. The aim of this
article is to examine the discretionary space (com-
pare Molander et al., 2012) governments give their
front-line child protection workers to conduct care
order proceedings.
Within the child protection system, we focus on
the agency level and the decisions where the State
assumes parental responsibility when parents are
unable or unwilling to perform their parental obliga-
tions. By care order, we refer to the processes and
activities associated with recommendations to the
court to pursue an involuntary removal. Discretion
in care order decisions is necessary as there are indi-
vidual and situational factors that inevitably must be
considered in each case in order to make a sound
decision. However, discretion can also be misused
(Brodkin, 2012; Maynard-Moody and Musheno,
2012). The research literature tells us little about the
quality of decision-making in child protection sys-
tems, but some have raised questions about decision
makers’ comprehensive review of evidence and
arguments and the potential for biased information
gathering (e.g. Munro, 1999, 2008). Other signals
from the field suggest system challenges such as
qualified staff and turnover, which may leave deci-
sion makers excessively dependent upon alternative
sources of information and insufficiently prepared to
engage in sound decision-making (O’Sullivan,
2011).
There are many factors and elements that influ-
ence and can distort a decision-making process; our
focus in this article is to examine the formal
framework for involuntary care order decisions in
four child protection systems – England, Finland,
Norway and the United States. A formal framework
is here understood as the formalized rules and proce-
dures that govern and inform organizational activi-
ties (Hatch, 2013; Weber, 2004). This formalization
can consist of the written documents, legislation,
instructions, guidelines and procedures that are made
by legislators and administrative managers to organ-
ize, streamline and make uniform an approach to a
social problem or issue. How strongly do govern-
ments steer and inform workers? What aspects in a
decision-making process are deemed important and
less important? Do these frameworks lay the ground-
work for high-quality and sound care order deci-
sions? In this article, we use four dimensions as
quality standards of a decision process: evidence and
information, involvement of child and parents, time
and accountability. We elaborate on this below.
Finally, we ask if and how the formal frameworks
relate to types of child protection systems and wel-
fare state models.
The article is organized into four parts, including
an outline of in- and cross-country research on the
formalized framework for care order decision-mak-
ing, followed by a theoretical presentation of the
decision-making dimensions that we highlight here.
Thereafter, we present the formal framework for
care order decisions based on four decision-making
dimensions, followed by a discussion of possible
strengths and weaknesses of the four decision-
making systems.
Decision-making in child protection
Decision-making in the child welfare system is com-
plex and made with a high degree of uncertainty (cf.
Munro, 1999, 2008; O’Sullivan, 2011). In child pro-
tection practice, decisions involve – at a minimum –
multifaceted normative issues, many types of research
knowledge, conflicting legal rights, the unique needs
and interests of involved children and their carers, and
prioritization of scarce resources, typically weighted
against each other, reasoned and finalized. Furthermore,
the context and setting for decision-making, such as
the institutional, organizational and legal framework,
matter. The decision-making model and political and

368 Journal of European Social Policy 25(4)
practice level cultures and normative platforms are all
important as well. For example, these four countries
draw the border between private and public responsi-
bility for children at risk differently (Gilbert et al.,
2011), and governments have unique approaches to
delegation of power and authority (Connolly, 2007). In
short, there are any number of factors that influence
when and how decisions are made in child protection
services, and authorized professional discretion is
included (Brodkin, 2012). There are few studies on
how the formalized framework sets the arena for deci-
sion-making on care orders. Cross-country studies on
decision-making in child protection in general and on
care orders in particular are scarce. Research on deci-
sion-making in child protection shows that there are
issues in the quality of decision-making in England,
Finland, Norway and the United States, and there
appear to be variations in how similar cases are han-
dled even within the same agency. Bolton and Lennings
(2010), based on their review of existing research,
state that ‘research consistently indicates that profes-
sional decision-making in child protection is subject to
bias … and varies significantly even between expert
clinicians’ (p. 1300).
One part of the problem for the State in regulating
decision-making concerns the distinction between
discretionary space and discretionary reasoning
(Molander et al., 2012). Discretionary space is about
the types of tools professionals are given to handle
and make decisions about a case, and discretionary
reasoning is about the justification of decisions. In his
reflections on the challenges of handling normative
issues such as the ‘best interests’ principle, as a stand-
ard for decision-making, Freeman (2007) points out
that ‘conclusions should be supported by reasoned
argument and that bias or worse prejudice should be
eliminated’ (p. 28; compare also Dunn et al., 2007).
Following this line of thought as an expression of dis-
cretionary reasoning, we refer to a regulative idea that
decision-making should be a process of deliberation
that rests on adequate information regarding the con-
tents of the case and the parties’ situations, that pos-
sible choices of action and their consequences must
be explored and that possible results should be ranked
in relation to overall goals (Munro, 2008; O’Sullivan,
2011). An important consideration in theories of
argumentation, as understood by Alexy (1989) and
Habermas (1996), is that legitimate discretionary rea-
soning comes through rational discourse in which all
parties involved participate and all relevant arguments
are presented for open and free discussion. Such a
rational discourse builds on the premise that all per-
sons concerned can participate, that they can freely
put forward their viewpoints and arguments, that all
relevant information is included and that there is a
review of the process.
Thus, when we examine the discretionary space
for front-line workers, we focus on the following
four dimensions (cf. Eriksen and Skivenes, 1997;
Eriksen and Weigård, 2004). First, that all relevant
information, evidence and expert knowledge is
included in the process. Second, that there is time to
process the information that is brought forward from
all parties and sources and that the timeline allows
for considered deliberation; and, at the same time,
that it is sensitive to the developmental needs of chil-
dren. Third, that children and parents are involved in
the process to the extent that they are heard, that
their perspectives and interests are included and con-
sidered and that they are given adequate information
in order to make informed choices about their cir-
cumstances and options. Fourth, that there are
accountability mechanisms in place to improve and
monitor decision-making processes and outcomes.
The data material for this analysis includes the
written documents, legislation, instructions, guide-
lines and procedures that are made by governments
(i.e. legislators and administrative managers) to
organize, streamline, inform and make uniform care
order proceedings at the agency level. The data are
different in each country as the state instructs the child
protection agencies in different ways. We pay atten-
tion especially to national legislation and national
guidelines, and since the US child welfare system is
not a single system but a collection of 50 state-
systems with significant variability both across and
within states, information from two local jurisdictions
within the state of California is included here.
Care order decisions in different
child protection systems
A general overview of the four countries’ child pro-
tection systems, their principles and orientations

Berrick et al. 369
regarding protection of children at risk of harm or
abuse or at risk of harming themselves or others is
found in Gilbert et al. (2011). The four selected
countries represent distinctive welfare state models
and child protection systems. Simplified, we will
categorize Norway and Finland as representing one
end of the Esping-Andersen (1990) typology, and
England and the United States as another, with the
former as two social democratic welfare states and
the latter as two liberal welfare states. The Norwegian
and the Finnish child protection systems are catego-
rized as ‘family service systems’, with a relatively
low threshold for providing a large range of family
services and, in general, characterized as having a
greater orientation to the child and the child’s per-
spective (Pösö, 2011; Skivenes, 2011). In contrast,
the United States and England could be categorized
as ‘child protection systems’, with a relatively high
threshold for intervention in the family and a more
limited mix of services offered to the family (Berrick,
2011; Parton and Berridge, 2011). All four countries
set norms for designated individuals to report accord-
ing to, when there is reason to believe a child is at
risk of harm or neglect. The overarching principles
for the child protection systems in these four coun-
tries differ slightly. Although they are all established
to protect the child’s interest, Norway, Finland and
England subscribe to the ‘child’s best interest’ prin-
ciple, although England has chosen a slightly differ-
ent formulation, stating that the ‘child’s welfare is
paramount’. The United States distinguishes itself
with an overarching principle focused on the child’s
‘safety and risk of harm’. The interaction between
the child protection agency and the courts varies in
care order cases. In the two risk-oriented systems,
England and the United States, the court makes
several decisions in care order proceedings, and
as such there is a regular interaction between the
court and the child welfare agency. In the family
service-oriented systems, Norway and Finland, there
is no interaction between court and the agency before
the one decision-making point that occurs when the
agency sends a care order application to court. In this
regard, the court makes a decision, and then it is
implemented by the agency. Thus, the preparatory
work by the agencies in care order cases differs in
these countries.
Following each country’s child protection system
orientation, Norway and Finland, as family service
systems, have approaches where in-home services
dominate. About 69 percent of the children in the
child protection system in Norway live with their
biological parents and receive in-home services, and
a total of 83 percent of all services provided are with
parental consent (Norwegian National Statistics
(NNS), 2011). The emphasis on in-home services is
also evident in Finland where about 7 percent of all
children between the ages of 0 and 17 receive in-
home services and about 1 percent of children were
in care in 2011 (Lastensuojelu, 2011). In the United
States, federal estimates suggest that each year,
about two-thirds of child maltreatment victims may
receive in-home services compared to about one-
third of children removed to foster care (United
States Department of Health and Human Services
(US DHHS), 2012); however, the intensity and dura-
tion of services are substantially less than what
might be encountered in a Nordic context. The pic-
ture in England is similar to the United States,
although direct parallels are difficult to describe pre-
cisely using data. Children with Child Protection
Plans usually live at home while assessment and
planning work is taking place; this will involve a
time-limited multi-agency plan, the outcome of
which will depend upon the individual case.
The position of involuntary care orders in these
four systems varies considerably. Norway and
England have around 70–75 percent of out-of-home
placements defined as involuntary care orders,
whereas in Finland 21–25 percent of the care orders
are involuntary. Estimates in the United States are
hard to come by. Data from one large state, however,
indicate that approximately 89 percent of out-
of-home placements are involuntary (Needell et al.,
2013). The term ‘involuntary’ care order is not
shared by all the countries, and there are variations
in which types of out-of-home placements are
defined as care orders. For example, emergency
placements are not treated as care orders due to their
temporary nature in Finland, but are so in the United
States. These differences in the degree to which one
country relies on involuntary care versus another’s
reliance on voluntary care speak to fundamental dif-
ferences in the role of the family vis-à-vis the State

370 Journal of European Social Policy 25(4)
and the different child protection system orienta-
tions. These differences also point to the underlying
frameworks for decision-making including one
state’s reliance on a litigious system based upon
findings of fact and verification of legal allegations
compared to other state’s orientation towards a best
interest consideration and a search for supportive
services.
Similar among these four countries is the role of
the court in authorizing involuntary care orders and
the child protection agency’s role in recommending
and carrying through the care order proceedings. At
the agency level in Norway, agency managers have
the formal authority to initiate a care order proceed-
ing. In Finland, the decision to propose a care order
to the administrative court is made by two child pro-
tection workers and their team manager. In England,
these recommendations are made and approved by
agency managers, in conjunction with the local
authority legal department. In the United States
(California), child protection workers make recom-
mendations approved by their supervisor.
With this short and necessarily limited backdrop
of the child protection systems, the following sub-
sections outline four dimensions of the broad for-
malized framework for decision-making regarding
care orders and point out areas of similarity and dif-
ference between countries.
Inclusion of information, evidence and
expert knowledge
In all four countries, evidence must be brought to
bear to show that children’s circumstances qualify
for a care order. Courts are involved to ensure that
appropriate procedures have been followed, parents
and children have experienced due process and eligi-
bility for care has been met. Each of the four coun-
tries meets these standards differently, each relying
on unique aspects of evidence and justification.
The Norwegian Child Welfare Act (CWA Article
4-12) sets three criteria for a care order: first, that
there is serious neglect or harm or failure to provide
for the child; second, that in-home services cannot,
or will not, help; and third, that a care order is in the
best interest of the child. Thus, legal expertise is
required. Information gathering is closely connected
with the persons and agencies that are involved in
case proceedings. Parents and the child are impor-
tant providers of information, as are other profes-
sionals who have been or are in contact with the
child and the family. The CWA specifies that all
information should, if possible, be collected in col-
laboration with parents and that parents should know
what information is collected. The agency has a right
to investigate, to visit the family’s home and to bring
the child to an examination if necessary (e.g. at a
hospital). The agency may also order an external
expert report or assessment (Child Welfare Act of
1992, Norway, Article 4-3).
In Finland, when care order decisions are pre-
pared, the focus is on demonstrating that the three
criteria for the care order are met. First, the child’s
health or development is seriously endangered by
lack of care or other circumstances in which they are
being brought up, or the child seriously endangers his
or her health and development by abuse of intoxi-
cants, by committing an illegal act other than a minor
offence or by any other comparable behaviour.
Second, in-home services have been determined
insufficient or irrelevant. Third, the care order and
related substitute care serves the principle of the
child’s best interest. All three criteria play an impor-
tant part, and therefore, as in Norway, evidence
should be brought to bear on all three (Child Welfare
Act 417/2007; Lastensuojelun käsikirja, 2013). In
order to demonstrate the level of endangerment to the
child’s health and development, social workers col-
lect information made available throughout the child
welfare process. The opinions of the child and the
parents as well as the opinion of the people who are
close to the child should be considered as forms of
evidence. The social workers also assess the outcome
of each in-home service provided and whether other
helpful in-home services could still be available.
They are obliged by the Finnish Child Welfare Act to
ask for expert opinions of those professionals who
are involved in the child’s or parents’ lives. This can
mean, for instance, teachers, kindergarten teachers,
doctors who know the child, or people who have
worked with the parents in substance abuse or mental
health treatment. In addition, each care order has to
be evaluated by a multi-professional team – experts
in the issues of children’s health and welfare. Every

Berrick et al. 371
social welfare agency is obliged to provide legal
expertise to support social workers in care order
decision-making.
In California, care orders may commence if any
of several highly detailed conditions are met. These
conditions, codified in law under Welfare and
Institutions Code 300, include harm or substantial
risk of harm relating to physical abuse, sexual abuse,
neglect and so on (all defined with high specificity in
the law). Evidence-based tools are used extensively
in combination with practice wisdom to determine
children’s risk and safety. Most counties use
Structured Decision-making (SDM) tools to help
guide and then check practice decisions (see www.
nccdglobal.org). Two tools in particular, the Safety
Assessment Tool and the Risk Assessment Tool,
must be completed as a case moves through the sys-
tem. Each tool includes a list of items that have been
empirically tested to assess their relationship to
safety and/or future maltreatment. Each item has an
associated glossary that details the exact parameters
of the item in order to reduce ambiguity of meaning
and inequitable decision-making between social
workers. Child protection workers are required to
file with the court a detailed report to justify an
involuntary care order. Among the several issues
included in the report is a federal requirement to
indicate that reasonable efforts were made to avoid
an out-of-home placement (Adoption Assistance and
Child Welfare Act (AACWA), 1980)
Like the United States, an English care order must
demonstrate that the child or young person is suffer-
ing significant harm or is likely to suffer significant
harm. Harm is defined as ill-treatment (including sex-
ual abuse and non-physical forms of ill-treatment) or
the impairment of health (physical or mental) or
development (physical, intellectual, emotional, social
or behavioural), including impairment suffered from
seeing or hearing the ill-treatment of another (Children
Act, 1989: 31–9, expanded by the Adoption and
Children Act, 2002). ‘Significant’ is not defined in the
1989 Children Act, although the court is advised to
compare the health and development of the child
‘with that which could be reasonably expected of a
similar child’. The guiding principles of the 1989
Children Act (p. 1) state both that ‘the child’s welfare
is paramount when making any decisions about a
child’s upbringing’ and that ‘the court shall not make
an order unless it considers that doing so would be
better for the child than making no order at all’. While
the above outlines the threshold criteria for seeking a
care order, the 1989 Children Act and statutory guid-
ance (HM Government, 2013) provide the framework
for inquiries, assessments and decision-making in all
cases where there are child protection concerns. Much
of this involves multi-disciplinary involvement, par-
ticularly in decision-making forums such as case con-
ferences. The local authority will only be able to seek
a care order when child protection workers are able to
demonstrate that the above threshold criteria are met.
Timelines for decision-making
Child protection workers are key decision makers in
each of the four countries, but the time available for
critical decision-making is starkly different. The
variability relates to the period of time available to
assess the family’s circumstances and the timeliness
for court review.
Norwegian legislation suggests that an investiga-
tion should begin as soon as the situation requires,
according to the law, and the agency has 3 months
(6 months under special conditions) to conduct the
investigation. In contrast, Finland does not impose
time restrictions for care orders but it does for emer-
gency placements. There is no point in a linear pro-
cess where a care order preparation might ‘begin’.
Instead, the process is designed to serve the best
interest of the child, and the timing of the process
should acknowledge the child’s interests. England
also does not set a deadline to finalize the prepara-
tions for a care order. In California, the timing for
decision-making is determined by law. The child
welfare system response is triggered by a child mal-
treatment referral. The referral is immediately
assessed by a child protection worker at the ‘hotline’
and an initial determination is made relating to the
possibility of imminent danger. In cases of extreme
concern (i.e. ‘imminent danger’), a child protection
worker must investigate the case within 2 hours. In
less severe cases, workers have up to 10 calendar
days to investigate the circumstances of the referral.
The investigation/assessment process can take up to
30 days. Once a decision is made to take temporary

372 Journal of European Social Policy 25(4)
custody of a child, the child is removed by the child
protection worker (sometimes in collaboration with
the police), followed by a presentation of evidence
to court within 48 hours in order to sustain custody,
and further evidence must be presented to court
within 30 days to detain a child longer and/or to
impose a care plan for services. These strict time-
lines were developed, in part, to convey to all actors
in the system a sense of urgency, given the serious
nature of the proceedings, and to ensure that chil-
dren and their families are not lost to a slow-moving
bureaucracy. Time frames utilized in court offer par-
ents clear guidelines to allow them to contest deci-
sions if they believe children have been removed
inappropriately.
Involvement of children and parents
We focus in this article on care order decisions that
are imposed on families and are, as such, involuntary
although parents and children may sometimes agree
with the decisions made about their fate. Even in cir-
cumstances where families object to the actions of
the state, parents and their children can have a voice
in selecting an appropriate course of state interven-
tion. Yet, we see across the four countries studied
that each country’s formalized structure accounts for
child and parent ‘voice’ quite differently.
The formalized framework for the Norwegian
child welfare system prescribes that both parents and
children should be included in decisions about care
orders. The legislation governing the system requires
that parents are informed, heard, have a lawyer or
support person, and are given the opportunity to
comment on information and assessments presented.
In particular, parents are allowed to present their
views on the case and related circumstances, and
they are provided free legal aid to assist them in their
interactions with the child protection agency. The
legislative framework gives children strong stand-
ing, not only in care order cases but also in all inter-
actions with the child welfare system:
A child who has reached the age of 7, and younger
children who are capable of forming their own
opinions, shall receive information and be given an
opportunity to state his or her opinion before a decision
is made in a case affecting him or her. Importance shall
be attached to the opinion of the child in accordance
with his or her age and maturity. (CWA article 6-3, first
section)
Older children, that is, children who are 15 years or
older, are considered a party in the case (similar to par-
ents), and younger children are invited to participate in
their own care planning. It follows from the legislation
that the impact and degree of inclusion of children are
related to considerations of their age, ability to form an
opinion, maturity and understanding.
The Finnish Child Welfare Act (417/2007),
implemented in 2008, goes even further. Every child
entering the child welfare system – regardless of age
– is entitled to participate: the children’s right to
obtain information in a child welfare case affecting
them, and the opportunity for them to present a view
on the case, must be safeguarded for the child in a
manner keeping with their age and level of develop-
ment. When assessing the need for child welfare, a
decision concerning a child or young person or the
provision of child welfare must pay special attention
to the views and wishes of the child or young person
(Child Welfare Act 417/2007). Ascertaining the
child’s view might sometimes endanger the child’s
health or development or it may be manifestly
unnecessary, in which case the law allows the princi-
ple to be disregarded. A child who is 12 years of age
or older is included in the formal administrative pro-
cess (‘hearing’) at the agency, and his or her opinion
is given the same weight as his or her custodians. If
a child disagrees with the proposal for a care order,
the care order will be treated as involuntary. Parents
and other custodians are also involved in the Finnish
decision-making process. The social work-led pro-
cess involves all the adults who are close to the child
and who provide for the child’s care and upbringing,
and their views and opinions are heard. This can
mean biological parents and step parents, biological
grandparents and relatives as well as the relatives of
the step parent. Yet, when it is time to carry out the
formal hearing for the care order decision, only the
opinion of the legal custodian has legal status. The
child is entitled to have a ‘guardian ad litem’ if there
is concern that his or her opinion would not other-
wise be well presented. The parents and the child are

Berrick et al. 373
both entitled to have legal aid, which is free of charge
for those with financial need.
In the United States (California), child welfare
practice is guided by a series of government manuals
(California Department of Social Services, 2012).
Within these policies, local jurisdictions have latitude
to fashion their practice to meet regional needs. With
regard to the inclusion of family, written parental con-
sent must be obtained prior to a voluntary care order;
however, when an involuntary care order is required,
parents need not necessarily be included directly in
decision-making. Instead, parents must be provided
with a written notice of their right to apply for judicial
review within 24 hours of their child’s placement.
Parents and children are each provided with legal rep-
resentation in court. Children’s right to be heard in
agency decision-making is not noted in the formal
policies and procedures manuals. In practice, we see
wide variations in the inclusion of parents and chil-
dren in decision-making, usually through team deci-
sion-making (TDM) meetings, although these are not
always offered, nor are they uniformly organized.
In England, the principle of working in partner-
ship with parents is central to the Children Act, 1989
and subsequent guidance; thus, parents are
involved in child protection processes, meetings,
assessments and, in some jurisdictions, Family
Group Conferencing, prior to a formal decision to
seek a care order. At this point, the local authority is
required to hold a legal planning meeting and for-
mally inform parents about this intention. Referred
to as the ‘Letter before Proceedings’, this provides
parents with access to legal support and triggers a
formal meeting (Ministry of Justice, 2008). While
partnership with children and young people is a core
principle of the legal framework, this has been
strengthened in recent guidance. A key principle
underpinning effective safeguarding arrangements
in every local area is ‘a child-centred approach: for
services to be effective they should be based on a
clear understanding of the needs and views of chil-
dren’ (HM Government, 2013: para. 8). It later
states,
Social workers, their managers and other professionals
should always consider the plan from the child’s
perspective. A desire to think the best of adults and to
hope they can overcome their difficulties should not
trump the need to rescue children from chaotic,
neglectful and abusive homes. (HM Government,
2013: 22)
Securing accountability
In each of the four countries, there are layers upon
layers of accountability mechanisms that are institu-
tionalized in the child protection system to guard
against capricious decision-making by child protec-
tion workers or judges. We narrow our focus here to
accountability mechanisms that seek to give input to
or review/control over care order proceedings at the
agency level.
The care order proceedings in Norway have built
in several discussion points that provide an opportu-
nity for examining and interpreting information.
Child protection workers and team leaders work
together. Thereafter, a management group including
team managers and agency managers reviews the
case. The written material is available for parents
and their lawyer. If external experts are used, the
Expert Commission on Children reviews the report.
There are several inspectorate or overview organiza-
tions that are in place to control and check on child
welfare agencies. Some of these can receive com-
plaints from service users and others about the child
welfare system. The National Audit Office is author-
ized to investigate many areas of the public sector
and has done so in the area of the child welfare sys-
tem several times over the last 10–15 years. Finally,
care order proceedings enjoy multiple layers of over-
sight as municipality lawyers and the courts regu-
larly review procedural compliance.
The Finnish Child Welfare Act emphasizes the
importance of thorough documentation, and there-
fore, documentation may be seen as a key measure of
accountability. Child protection workers are required
to prepare written documents for the care order deci-
sion. These documents make the care order applica-
tion, which is sent to the administrative court. All
documents are provided to the clients as well. When
preparing the care order application, the relevance of
the care order is tested and evaluated. The Act requires
two social workers (child protection workers) to be
involved in care order preparation; the agency-based

374 Journal of European Social Policy 25(4)
application is then approved by the child welfare
manager. Parents and children of a certain age are
entitled to appeal the decisions made by the social
welfare agency. Appeals and complaints are com-
monly addressed first to the person in charge of child
welfare services. They may also contact the local
social ombudsman for advice. The municipalities in
Finland are obliged to keep a child welfare register of
all their child welfare clients and provide statistical
information based on these to the National Institute of
Health and Welfare which provides a national annual
report on child welfare. The National Supervisory
Authority for Welfare and Health and Regional State
Administrative Agencies monitor child welfare in
general, but they do not have special oversight respon-
sibilities relating to care order proceedings.
The federal government and the state of California
have adopted a detailed scheme to account for the
outcomes of the child welfare system through a lay-
ered data collection strategy. All states are required
to make annual reports to the federal government on
several outcome measures including the number of
children who are re-reported to the system for fur-
ther maltreatment following case closure, the num-
ber of children entering care, the number of
placements children experience while in care and so
on. In California, these data requirements are sup-
plemented with additional data demands, and each
county must report their data regularly and devise
‘System Improvement Plans’ (SIPs) to address sys-
tem weaknesses that may be revealed. Data for every
county are displayed publicly on an ongoing basis so
that public officials, the press and citizens in the
community can review system performance over
time, in comparison with other counties or in com-
parison with the state as a whole (see http://cssr.
berkeley.edu/ucb_childwelfare/).
At the individual level, lawyers for all parties to
the case provide another layer of accountability and
allow for parents to contest judicial decisions. Of
course, judges serve a final function in accounting
for child protection worker practice, although court
proceedings are not typically open to the public for
observation.
Within local authorities in England, managerial
oversight of cases will be evident. This will involve
social work decisions being overseen by managers
within the organization; the local authority legal
department also plays an important role. At a local
strategic level, the Local Safeguarding Children
Board is responsible for ensuring safeguarding
arrangements and also undertakes Serious Case
Reviews (SCRs) following a child’s death or very
serious incident. SCRs are a particular feature of the
English child protection system, and these reports
are now publicly available and often subject to con-
siderable political and media scrutiny. Some of the
most serious high profile cases have led to govern-
ment inquiries and have informed national and local
policy reforms. There are also a number of mecha-
nisms which scrutinize the performance of local
authorities and other agencies concerned with child
protection, most notably inspection regimes, which,
despite claims to reduce bureaucracy, continue
robustly under the current government. Such inspec-
tions are regular, can be unannounced and can have
consequences for the local authority and organiza-
tions involved, for example, job losses, managerial
changes or public opprobrium.
Discussion
A crude summary of the four countries’ care order pro-
ceedings suggests that these governments have quite
different approaches regarding child protection work-
ers’ use of professional discretion. Finland is on one
end of the spectrum, with a highly de-regulated system
in which legislation is general and national guidelines
are few, and by and large the proceedings and deci-
sions are made by child protection workers in collabo-
ration with the child and the family. The Norwegian
system is quite similar to the Finnish, but less de-regu-
lated as it has strict timelines, hierarchical decision-
making and comparatively less involvement of the
child and parents. Still, it is a system that gives child
protection workers much leeway in what to do in care
order proceedings. The California system is on the
other end of the spectrum, highly regulated, with strict
timelines and detailed decision-making tools in place.
Parents and children may be collaborators in the pro-
cess, but they are principally involved as informants
whose material can be used as evidence. The English
system is even more regulated than the California sys-
tem, with strict guidelines and procedures to be

Berrick et al. 375
followed. However, the guidelines require the strong
involvement of parents and children, and there is no
timeline for the care order process. Thus, the space for
discretion and professional judgment for child protec-
tion workers across these four countries is quite differ-
ent. US and English workers have little space for
discretion – the use of judgment within clear restric-
tions – whereas Norwegian and Finnish workers have
much space for discretion – discretion not bound by an
authority and as such protected from overrule
(Dworkin, 1967). In Figure 1, we have illustrated the
systems and relative discretionary features.
The discretionary space differs with the risk-oriented
systems on the one end, and family service-oriented
systems on the other end, indicating that there will be
differences in what aspects of a decision-making pro-
cess will be deemed important and less important. When
we examine how care order proceedings are prescribed
in law and regulations, following the four dimensions
we have identified as decisive for the quality of a deci-
sion process, we find similarities and differences: first,
all countries follow the principle of legality by setting
the agenda with legislative criteria for interventions in
the family and demanding that evidence for harm or
neglect of the child must be provided. In all countries
except the United States, it is also a criteria that a care
order should be in the best interest (or well-being) of the
child. It is notable that these states – with the exception
of the United States – have subscribed to the United
Nations (UN) Convention on the Rights of the Child. In
Norway and Finland, reflecting their service-oriented
systems and the connection between the social
democratic welfare state and the child protection system
(Pösö et al., 2013), in-home services must first be
attempted and shown not to remedy the situation prior
to a care order. A similar principle is also established in
federal law in the United States, although it is generally
recognized that the bar for demonstrating such efforts is
low, in part due to the significant risk of harm that stands
as the threshold for intervention. Criteria used to justify
a care order clearly shapes the evidence and informa-
tion-gathering process. The required evidence across
these four systems is somewhat similar, as there is a
requirement that legal expertise be involved to confirm
that eligibility criteria have been met. Furthermore, pro-
fessionals and lay persons who know the child are que-
ried. Norway stands out as lacking instruction on which
professionals to consult, and instead bases its approach
on general administrative principles.
Second, with regard to the involvement of chil-
dren and parents, we see a gradual convergence
across states. It is a generally held liberal principle
to protect individual freedom, and only to protect
others should the State restrict or interfere with an
individual’s liberty (Heywood, 2007). States have
interpreted the freedom doctrine differently, as is
evident in the welfare state models and child protec-
tion orientations of Norway and Finland on the one
side, and England and the United States on the other.
The question is how the child protection system
views the child and the parents, and what roles they
play in the care order proceedings. In England,
Norway and more so in Finland, parents are involved
in decision-making and informed about care order
proceedings. This speaks to the State’s view of par-
ents as service users, and, perhaps, as owners of the
information gathered. It also gives an indication of
these systems’ views on the principles of individual
autonomy and the right to privacy. The contrast is
the United States where parents are offered legal
representation from the state, but where they may
not be directly involved in decision-making regard-
ing care order proceedings. Government direction
about how or when to involve children in care order
proceedings is quite similar in Finland, Norway and
England. Children of all ages should be engaged,
although the degree of their involvement may
depend on an assessment of their age, maturity and
ability to form an opinion. Subtle differences emerge
Lile US
England
Norway
Finland
DISCRETIO-
NARY
SPACE
Much
De-regulate
dR
egulated
SYSTEM
Figure 1. Relative discretion features in four different
child welfare systems.

376 Journal of European Social Policy 25(4)
between states on these issues, however, from chil-
dren’s participation based on their ability to form an
opinion (e.g. Norway) to children’s age and matu-
rity as measures of competence (e.g. Finland and
England). In fact, very young children have the
capacity to form an opinion, although it is more dif-
ficult to assess competency (Archard and Skivenes,
2009). These understated differences likely result in
large distributions between countries in the average
age of participating children. Otherwise, there is
only the Finnish system that entitles the child to a
guardian or a spokesperson at the agency level, or, if
necessary, a lawyer can be appointed. The risk-ori-
ented child protection system in California shows
that child protection workers primarily consult with
children during the information collection phase of
the case. When cases go to court for care proceed-
ings, legal counsel is appointed. In this regard, their
voice is represented in court, but they are not actors
themselves, unless invited to participate by the
judge. The US system does not require that children
are involved in agency decision-making, and as
such stands in sharp relief from its comparison
states.
Third, the amount of time child welfare workers
are allowed to prepare a care order can have an
impact on the opportunity to collect information,
to speak with the child and parent, to reflect on and
discuss the information and evidence that are gath-
ered and to seek a second opinion. In this regard,
more time may be better, as in Finland and England,
where staff work without strict deadlines in the
care order process. But time can be problematic
from the standpoint of the child who may be living
in an unsafe situation and for parents who may be
unsure about the eventual resolution of the case.
Timelines for care order proceedings differ sub-
stantially from none in Finland and England to
only a few weeks (e.g. United States). These dif-
ferences may be related to the threshold for inter-
vention and the United States being a risk-oriented
child protection system with a defined high thresh-
old for intervention into the family and with a low
provision of supportive services – a combination
that may result in high-risk situations for the child
who may need more responsive protective systems
to maintain safety.
The established internal and external accountabil-
ity mechanisms are many in all four systems. Child
protection staff working cooperatively together and
manager oversight are typical in all four countries.
Parents’, lawyers’ and children’s access to docu-
ments and discussion points is evident within the
family service–oriented child protection systems.
The oversight that is built into decision-making in
the risk-oriented systems of England and the United
States relies heavily on assessment tools and
guidelines.
Although an analysis of formal procedures cannot
describe how the actual care order proceedings are
conducted at the agency level, they tell us something
about how governments weigh various factors, and
what remains for front-line workers to decide. The
overarching question is how these frameworks lay
the groundwork for high-quality care order decisions
or the discretionary justifications of decisions. Both
the English and the US decision-making tools set
some requirements on what factors to consider. Such
an approach may standardize the criteria used for
decision-making. The system in California, guided
by an evidence-based decision-making tool, is
focused on a justification for each decision, with
standards in place to explain the worker’s assess-
ment. The risk-oriented system that revolves around
a high threshold for intervention may allow for these
strict, evidence-based regulations. When the princi-
ple of best interest or child well-being is at stake, and
there are lower thresholds for intervention, the dis-
cretionary space is somewhat relaxed, as seen, in
part, in the complex English framework and in full in
the Norwegian and Finnish systems.
Concluding remarks
States appear to instruct social work decision-making
more subtly in family-service systems than they do in
child protection systems. The weaker and stronger
steering mechanisms may be appropriate in the sys-
tems in which each is embedded. Some have pointed
out the prospects of governing front-line workers as
dim (Marinetto, 2011). This point remains to be
examined in the child protection area, but we expect
that front-line workers in these four countries will
handle care order proceedings differently because of

Berrick et al. 377
the system differences identified herein. Thus, we
expect child protection workers to be influenced by
the ‘accountability to professionals and citizen con-
sumers of services’ (Marinetto, 2011), but also
accountable to the legislative and the system pre-
scriptions within which they work (cf. Maynard-
Moody and Musheno, 2012).
Our analysis of the formal frameworks guiding
child protection in four countries illuminates stark
differences in the principles undergirding a similar
intervention. What is clear is that these systems have
different aims for how to conduct high-quality care
order decisions, depending on which standards for
quality are applied. For example, if legitimacy is
based on the involvement of those concerned, that is,
the child and parents, we see that these four systems
fulfil this standard, albeit to varying degrees, indicat-
ing different perceptions of autonomy and self-deter-
mination for individuals. The analysis also suggests
that when studying child welfare decision-making,
one should acknowledge the system in which the
decisions are made (Duffy and Collins, 2010).
Funding
The article is part of a larger comparative research project
funded by the Norwegian Research Council, examining
decision-making in child welfare systems in four countries.
Note
1. We use the term child protection system to charac-
terize the systems that have responsibility for chil-
dren at risk of harm or neglect from their caregivers
or who may be at risk of harm to themselves or oth-
ers. In some countries, these may be referred to as
child welfare systems. We use the term child protec-
tion workers as the label for the front-line staff who
interact with children and families and who prepare
care orders. These workers can also be labelled social
workers or child welfare workers. Due to space limi-
tations, we are unable to follow up on the importance
of professionals or their education and training.
Detailed overviews of each country’s systems are
outlined in Gilbert et al., 2011.
References
Adoption and Children Act. (2002) England. The National
Archives, available at http://www.legislation.gov.uk/
ukpga/2002/38/contents.
Adoption Assistance and Child Welfare Act (AACWA).
(1980) HR 3434. 96th Congress USA. Public Law
96–272. Summary, available at https://www.congress.
gov/bill/96th-congress/house-bill/3434.
Alexy, R. (1989) A Theory of Legal Argumentation.
Oxford: Clarendon.
Archard, D. and Skivenes, M. (2009) ‘Balancing a Child’s
Best Interests and a Child’s Views’, The International
Journal of Children’s Rights 17(1): 1–21.
Berrick, J. (2011) ‘Trends and Issues in the US Child
Welfare System’, in N. Gilbert, N. Parton and
M. Skivenes (eds) Child Protection Systems:
International Trends and Orientations, pp. 17–36.
New York: Oxford University Press.
Bolton, A. and Lennings, C. (2010) ‘Clinical Opinions
of Structured Risk Assessments for Forensic Child
Protection: The Development of a Clinically Relevant
Device’, Children and Youth Services Review 32(10):
1300–10.
Brodkin, E. (2012) ‘Reflections on Street-Level
Bureaucracy: Past, Present, and Future’, Public
Administration Review 72(6): 940–9.
California Department of Social Services. (2012) Child
Welfare Services Manual: Policies and Procedures.
Sacramento, CA: Health & Human Services Agency.
Child Welfare Act of 1992, Norway. Child Welfare Act of
17 July 1992, Number 100 (Barneverntjenesteloven).
Child Welfare Act 417/2007. Finland, available at http://
finlex.fi/en/laki/kaannokset/2007/20070417.
Children Act. (1989) England, available at http://www.
legislation.gov.uk/ukpga/1989/41/contents.
Connolly, M. (2007) ‘Practice Frameworks: Conceptual
Maps to Guide Interventions in Child Welfare’,
British Journal of Social Work 37: 825–37.
Duffy, J. and Collins, M.E. (2010) ‘Macro Impacts on
Case Worker Decision-making in Child Welfare: A
Cross-national Comparison’, European Journal of
Social Work 13(1): 35–54.
Dunn, M., Clare, I., Holland, A. and Gunn, M. (2007)
‘Constructing and Reconstructing ‘Best Interests’: An
Interpretative Examination of Substitute Decision-
Making under the Mental Capacity Act 2005’, Journal
of Social Welfare and Family Law 29: 117–33.
Dworkin, R. (1967) The Model of Rules. Scholarship
Series, Faculty Scholarship, 1 January. New Haven,
CT: Yale Law School.
Elster, J. (1989) Solomonic Judgements: Studies in
the Limitation of Rationality. Cambridge: Press
Syndicate.
Eriksen, E.O. and Skivenes, M. (1997) ‘The Problem with
Legitimacy in Child Protection (Legitimasjonsproblemer
i barnevernet)’, Bergen, LOS-notat 9705 (in Norwegian).

378 Journal of European Social Policy 25(4)
Eriksen, E.O. and Weigård, J. (2004) Understanding
Habermas: Communicative Action and Deliberative
Democracy. London; New York: Continuum.
Esping-Andersen, G. (1990) The Three Worlds of Welfare
Capitalism. Princeton, NJ: Princeton University
Press.
Freeman, M. (2007) ‘Article 3: The Best Interests of the
Child’, in A. Alen, J. Vande Lanotte, E. Verhellen,
F. Ang, E. Berghmans and M. Verheyde (eds) A
Commentary on the United Nations Convention on the
Rights of the Child, p. 28. Leiden: Nijhoff Publishers.
Gilbert, N., Parton, N. and Skivenes, M. (eds) (2011)
Child Protection Systems: International Trends and
Orientations. New York: Oxford University Press.
Habermas, J. (1996) Between Facts and Norms.
Cambridge, MA: The MIT Press.
Hatch, M. (2013) Organization Theory. New York:
Oxford University Press.
Heywood, A. (2007) Political Ideologies. New York:
Palgrave Macmillan.
HM Government. (2013) Working Together to Safeguard
Children: A Guide to Inter-Agency Working to
Safeguard and Promote the Welfare of Children.
London: Department for Children Schools and Families.
Lastensuojelu. (2011) Child Welfare 2011. Statistical
Report 16. Helsinki: Terveyden ja hyvinvoinnin lai-
tos, Suomen virallinen tilasto, Sosiaaliturva 2012.
Lastensuojelun käsikirja. (2013) The Handbook of Child
Welfare. Helsinki: Terveyden ja hyvinvoinnin laitos,
available at www.sosiaaliportti.fi/fi-FI/lastensuoje-
lunkasikirja/.
Marinetto, M. (2011) ‘A Lipskian Analysis of Child
Protection Failures from Victoria Climbié to “Baby P”:
A Street-level Re-Evaluation of Joined-up Governance’,
Public Administration 89(3): 1164–81.
Maynard-Moody, S. and Musheno, M. (2012) ‘Social Equities
and Inequities in Practice: Street-Level Workers as
Agents and Pragmatists’, Public Administration Review
72: 16–23.
Ministry of Justice. (2008) ‘The Public Law Outline: Guide
to Case Management in Public Law Proceedings’,
Ministry of Justice, available at www.justice.gov.uk.
Molander, A., Grimen, H. and Eriksen, E.O. (2012)
‘Professional Discretion and Accountability in the
Welfare State’, Journal of Applied Philosophy 29(3):
214–30.
Munro, E. (1999) ‘Common Errors of Reasoning in Child
Protection Work’, Child Abuse and Neglect 23(8):
745–58.
Munro, E. (2008) Effective Child Protection. Los Angeles,
CA; London: Sage.
Needell, B., Webster, D., Armijo, M., Lee, S., Dawson, W.,
Magruder, J., et al. (2013) California Child Welfare
Indicators Project (CCWIP) reports. University of
California at Berkeley California Child Welfare
Indicators Project, available at http://cssr.berkeley.
edu/ucb_childwelfare.
Norwegian National Statistics (NNS). (2011) Barnevern
2010, available at http://www.ssb.no/sosiale-forhold-
og-kriminalitet/statistikker/barneverng/aar/2011-06-
27.
O’Sullivan, T. (2011) Decision-Making in Social Work.
Basingstoke: Palgrave Macmillan.
Parton, N. and Berridge, D. (2011) ‘Child Protection in
England’, in N. Gilbert, N. Parton and M. Skivenes
(eds) Child Protection Systems: International Trends
and Orientations, pp. 60–88. New York: Oxford
University Press.
Piper, C. (2000) ‘Assumptions about Children’s Best
Interests’, Journal of Social Welfare and Family Law
22: 261–76.
Pösö, T. (2011) ‘Combatting Child Abuse in Finland: From
Family to Child-centered Orientation’, in N. Gilbert,
N. Parton and M. Skivenes (eds) Child Protection
Systems: International Trends and Orientations, pp.
112–30. New York: Oxford University Press.
Pösö, T., Skivenes, M. and Hestbæk, A.-D. (2013) ‘Child
Protection Systems Within the Danish, Finnish and
Norwegian Welfare States – Time for a Child Centric
Approach?’, European Journal of Social Work 17(4):
475–490.
Skivenes, M. (2011) ‘Norway: Toward a Child Centric
Perspective’, in N. Gilbert, N. Parton and M. Skivenes
(eds) Child Protection Systems: International Trends
and Orientations, pp. 154–82. New York: Oxford
University Press.
United States Department of Health and Human Services
(US DHHS) (2012) Child Maltreatment: 2011, avail-
able at www.acf.hhs.gov/sites/default/files/cb/cm11.
pdf.
Weber, M. (2004) Power and Bureaucracy [Makt og
byråkrati]. Oslo: Gyldendal (in Norwegian).