Tag Archive for: Deprivation of Liberty

Acknowledgement

This article was prompted by a thoughtful post by Adelle Banks, a family and safeguarding lawyer, asking whether children lawyers would benefit from greater understanding of the assessment frameworks used by social workers, Cafcass and youth justice professionals.

Banks’ central proposition was not that lawyers should perform other professionals’ roles. It was that better cross-disciplinary literacy may allow lawyers to identify gaps earlier and ask better questions before those gaps derail a final hearing.

The analysis and conclusions below are JSH Law’s own.

The key point

A lawyer does not need to know how to conduct a social-work assessment.

But a lawyer working in children proceedings should understand enough about assessment to ask:

What information was gathered? What analytical framework was used? What need was identified? What competing explanation was considered? What intervention follows from that need? And is that intervention actually deliverable?

That is not professional trespass. It is the difference between accepting a professional conclusion and properly interrogating the reasoning behind it.

Five things to understand first

1. Behaviour is evidence — but it is not necessarily an explanation.
Going missing, offending, aggression, drug use or refusing boundaries may tell us something important. They do not automatically tell us why the behaviour is occurring.

2. Assessment should connect behaviour to need and context.
Current safeguarding frameworks look at the child’s developmental needs, parenting capacity and the wider family and environmental context.

3. A recommendation is only as useful as the analysis beneath it.
“Therapeutic support is required” is not a complete plan unless the need, intervention, availability, timescale and intended outcome are understood.

4. Lawyers must not turn professional literacy into amateur diagnosis.
Knowing enough to challenge an unexplained conclusion is very different from deciding what diagnosis or treatment a child requires.

5. The question should be asked early.
Discovering at the final hearing that the proposed placement, therapy or support package does not actually exist is usually too late.

Why behaviour is not the same as need

A child may present with behaviour which is extraordinarily difficult to manage.

They may:

• go missing;
• self-harm;
• use drugs or alcohol;
• become aggressive;
• damage property;
• reject placements;
• refuse education;
• become involved in offending;
• reject adult boundaries; or
• repeatedly place themselves in dangerous situations.

Those behaviours may generate immediate safeguarding concerns.

They may require immediate protective action.

But description is not analysis.

A good assessment asks what may sit underneath the behaviour.

Possibilities might include:

• trauma;
• exploitation;
• bereavement;
• neurodivergence;
• speech and language difficulties;
• mental-health difficulties;
• domestic abuse;
• attachment disruption;
• educational exclusion;
• placement instability;
• poverty;
• discrimination;
• family breakdown; or
• a combination of several factors.

The behaviour is real.

But the intervention which is appropriate depends upon what is driving it.

Children may need both.

The danger is when the first question replaces the second.

What children’s assessment frameworks actually do

The old visual triangle associated with the Framework for the Assessment of Children in Need and their Families will be familiar to many practitioners.

The underlying model is not simply a historical curiosity.

Working Together to Safeguard Children 2026 continues to use the same three-domain conceptual structure when describing comprehensive assessment in England.

The purpose of assessment is not merely to record information.

Current statutory guidance says children’s social-care assessment should:

• gather important information about the child and family;
• analyse needs and the nature and level of risk or harm;
• determine whether statutory intervention may be required;
• identify support capable of addressing the child’s needs; and
• identify support available within the wider family network.

That word analyse matters.

A report containing forty pages of history is not necessarily a forty-page assessment.

An assessment should transform information into reasoned conclusions.

The three domains lawyers should understand

Working Together 2026 identifies three broad domains:

For lawyers, the importance is not memorising the triangle.

It is recognising when a professional analysis has become disproportionately focused on one corner of it.

For example

A report may describe at length a teenager’s aggressive behaviour.

But where is the analysis of:

• their mental health?
• their educational history?
• possible exploitation?
• neurodevelopmental needs?
• trauma?
• placement instability?
• relationship disruption?
• previous interventions and whether they worked?

A lawyer does not answer those questions professionally.

A lawyer can ask why they have not been answered.

Cafcass has its own assessment architecture too

Cafcass’s current Child Impact Assessment Framework (CIAF) performs another function.

It brings together practice guidance and resources used by Family Court Advisers and Children’s Guardians in public and private law proceedings.

Cafcass says the framework is intended to support practitioners to:

• keep children and families at the centre of assessment;
• understand what makes the individual child unique;
• work through respectful and professionally curious relationships;
• understand what matters to the child;
• understand strengths, harm and future risk; and
• make clear and reasoned decisions.

That creates an obvious set of questions when reading a Cafcass report.

Has the report merely described the parents’ competing accounts?

Or has it analysed what those accounts mean for this child?

Has it identified the source of the information?

Has it distinguished allegation from finding?

Has it explored an alternative explanation?

Has it explained the route from evidence to recommendation?

Has it identified what support would actually be needed for the recommendation to work?

Those are legal-evidential questions informed by an understanding of social-work methodology.

They do not require the lawyer to become the social worker.

Youth justice makes the distinction particularly clear

The Youth Justice Board’s assessment guidance is remarkably explicit.

Good assessment gives weight not only to harm and public protection but also to:

• strengths;
• vulnerabilities;
• unmet needs;
• mental and emotional wellbeing;
• substance misuse;
• education;
• family and environmental context;
• speech, language and learning needs;
• trauma;
• exploitation; and
• positive networks and protective factors.

The guidance expressly says assessment must go beyond merely describing the facts.

That matters enormously when the presenting fact is:

“This child keeps offending.”

The legal system still needs to deal with the offending.

But if the child is being criminally exploited, traumatised or struggling with unmet communication needs, simply escalating punishment may fail to address the mechanism producing the behaviour.

This is precisely where cross-disciplinary literacy becomes valuable.

Deprivation of liberty shows what happens when behaviour and need become dangerously easy to confuse

Banks’ post focused particularly upon teenagers and deprivation-of-liberty cases.

There is good reason for that.

Official Family Court statistics show that between April and June 2026 there were 420 High Court deprivation-of-liberty applications, an increase of 18% on the equivalent quarter in 2025.

Almost all concerned teenagers:

• 66% involved children aged 13 to 15; and
• 26% involved young people aged 16 to 18.

Earlier Nuffield Family Justice Observatory research illustrates the complexity behind those applications.

Among the cases examined:

• 69.2% involved behaviour presenting a risk to others;
• 59.1% involved mental-health or emotional difficulties;
• 55.3% involved placement breakdown;
• 52.4% involved self-harm or suicidal ideation; and
• 46.6% involved absconding.

But those categories did not exist separately.

In almost all cases there was more than one risk factor, and nearly two-thirds involved at least four.

The children had experienced high levels of trauma, adversity and instability.

That changes how we should read a sentence such as:

“The child repeatedly absconds and is difficult to manage.”

The behaviour remains relevant.

But it may be the visible endpoint of a much more complicated needs profile.

Sometimes the “complexity” is not located entirely inside the child

This is one of the strongest conclusions in the Department for Education’s 2025 research into looked-after children in complex situations who were at risk of or subject to deprivation of liberty.

The report describes children experiencing multiple overlapping difficulties across:

• mental health;
• trauma;
• neurodevelopment;
• education;
• exploitation;
• social care;
• placements; and
• family circumstances.

But it also identifies a system problem.

Children may interact with health, social care, education and justice services which are themselves fragmented.

Repeated short-term intervention does not necessarily amount to sustained support.

A child’s presentation may deteriorate while different agencies each deal with one piece.

Then, at crisis point, the behaviour is described as increasingly complex.

There is a difficult question here:

Is this a child whose needs have suddenly become extraordinarily complex — or a child whose overlapping needs have been repeatedly encountered but insufficiently integrated?

Sometimes the answer will be both.

But lawyers should be alert to the distinction.

Does a professional report contain conclusions you cannot trace back to the evidence?

JSH Law can provide defined-scope support for litigants in person who need help analysing a Cafcass report, Child Impact Report or wider evidential record.

Support can include:

• report analysis;
• evidence-source mapping;
• chronologies;
• identifying factual gaps and inconsistencies;
• separating allegation, professional opinion and judicial finding;
• Cafcass response documents;
• position statements;
• witness-statement preparation support;
• questions for hearings;
• court-bundle preparation support; and
• hearing preparation.

The aim is not to substitute a lay opinion for professional assessment. It is to make the reasoning capable of being understood and tested.

Book a consultation

So what should a children lawyer actually know?

Not how to practise social work.

Not how to diagnose autism.

Not how to undertake a psychological formulation.

Not how to decide which therapy a child requires.

But enough to recognise the anatomy of competent professional reasoning.

That is professional literacy.

The questions I would ask of any professional assessment

A report should not become immune from scrutiny merely because its author is professionally qualified.

Equally, disagreement does not mean the professional is wrong.

The useful middle ground is disciplined questioning.

A good assessment can still lead to a bad plan

This distinction deserves more attention.

Assessment answers:

What is happening and what does this child need?

Planning answers:

What are we actually going to do about it?

A care plan can therefore fail even where the assessment is broadly sound.

For example

An assessment identifies significant trauma and recommends trauma-informed therapeutic input.

The care plan says:

“Therapeutic support will be provided.”

That sounds reassuring.

But the proper questions are:

• By whom?
• Has a referral been accepted?
• What intervention?
• When will it begin?
• How frequently?
• For how long?
• What happens if the child cannot engage?
• Who is funding it?
• What support exists in the meantime?

If nobody can answer those questions, the court may be looking at an aspiration rather than a plan.

“The child needs therapy” and “this child will receive appropriate therapy” are not the same proposition.

Why this matters to the court

Family judges do not commission abstract professional analysis for intellectual interest.

Assessments inform decisions.

Those decisions may determine:

• where a child lives;
• whether contact takes place;
• whether contact is supervised;
• whether a child enters care;
• whether a placement remains viable;
• whether liberty is restricted;
• whether reunification is realistic;
• or whether state intervention can safely reduce.

That means the bridge between assessment and recommendation matters.

The court should be able to follow:

Evidence → Analysis → Identified Need → Intervention → Expected Outcome → Welfare Recommendation

If one link is missing, the recommendation may still be right.

But it becomes harder to test.

The JSH Law Six-Question Check works here too

The JSH Law Six-Question Check can be adapted to professional assessment.

The final question may be the one most frequently overlooked.

What can a litigant in person take from this?

You do not need to become an expert in social work theory to read a professional report intelligently.

If a report recommends a major change affecting your child, try to identify:

1. What behaviour or concern is described?
2. What evidence supports it?
3. What underlying need is identified?
4. How was that need assessed?
5. What alternative explanations were considered?
6. What does the child say?
7. What support is recommended?
8. Why is that support expected to help?
9. Is it actually available?
10. What happens if it is not?
11. How does all of that lead to the recommendation made to the court?

Do not respond to every professional opinion with:

“They’re biased.”

Even where you strongly disagree, it is usually more useful to identify the precise reasoning problem.

For example:

That is considerably more useful than simply saying the professional is wrong.

Could AI help lawyers interrogate professional assessments?

Yes — within limits.

AI could help a lawyer or litigant:

• extract every stated need from a report;
• identify the evidence cited for each conclusion;
• map recommendations against assessed needs;
• compare earlier and later assessments;
• identify contradictions;
• find services promised in a care plan;
• create a chronology of professional involvement;
• identify unanswered questions; and
• separate factual assertion from professional opinion.

That is potentially very powerful.

But AI cannot safely turn itself into the missing professional.

If a report contains insufficient evidence to establish why a child behaves in a particular way, an AI model should not quietly fill the gap with:

“This is probably trauma.”

Or:

“This appears to be autism.”

Or:

“The child is displaying attachment disorder.”

That is exactly the professional trespass we should avoid.

Where does professional curiosity become professional trespass?

The dividing line is relatively straightforward.

It is appropriate to ask:

“What is the evidential basis for this formulation?”

It is not appropriate for a lawyer without relevant expertise to declare:

“I have diagnosed the child differently.”

It is appropriate to ask:

“What intervention is clinically or therapeutically indicated?”

It is not appropriate to prescribe that intervention oneself.

It is appropriate to identify that no speech-and-language assessment has taken place despite repeated communication concerns.

It is not appropriate to conduct one.

It is appropriate to challenge whether a care plan actually delivers what the assessment says the child needs.

It is not appropriate to assume professional responsibility for designing the child’s care.

That is the distinction Banks captured so well.

Curiosity without trespass.

Cross-disciplinary literacy may also reduce delay

There is a procedural consequence to all of this.

A final hearing is an expensive place to discover that:

• the therapeutic intervention has never been commissioned;
• the placement cannot provide what the assessment assumed;
• the relevant specialist assessment was never completed;
• the child’s communication needs were not understood;
• the guardian cannot support the care plan as drafted; or
• the intervention will not be available for another six months.

The court may then need further evidence.

The hearing may adjourn.

The child waits.

The family waits.

Costs increase.

And the system describes the resulting months as “delay”.

This connects directly with the wider point I made recently about the human impact of court delay.

Better professional literacy will not remove every adjournment.

But identifying obvious analytical and delivery gaps earlier may prevent some of them.

Behaviour is data. Need requires analysis.

The legal system has to respond to behaviour.

A child running repeatedly into danger cannot simply be left because professionals are still exploring why.

A child presenting a serious risk to others may require immediate controls.

Safeguarding sometimes demands action before the whole picture is known.

But emergency containment should not become the final explanation.

Because there is a profound difference between:

“We have stopped the child doing this.”

and:

“We understand what this child needs.”

The law tells us what decisions must be made. Other disciplines can help explain the child those decisions are being made about.

Family lawyers do not need to become social workers, psychologists, clinicians or therapists.

But we should understand those disciplines well enough to recognise when the chain from behaviour to need to intervention does not make sense.

That is not crossing a professional boundary.

It is making interdisciplinary decision-making work properly.

And when the consequences may include removing a child from home, restricting their liberty or determining their relationship with a parent, asking better questions is not an optional intellectual exercise.

It is part of making the court’s decision safer.

Need help analysing the evidence behind a professional recommendation?

JSH Law provides defined-scope support for litigants in person in children proceedings.

Depending on scope, support can include:

• Cafcass and Child Impact Report analysis;
• chronologies;
• evidence-source mapping;
• identifying factual and analytical gaps;
• responses to professional reports;
• witness-statement preparation support;
• position statements;
• safeguarding evidence organisation;
• court-bundle preparation support; and
• hearing preparation and McKenzie Friend support where appropriate.

The purpose is not to replace the professional assessment. It is to understand it well enough to test whether the evidence, reasoning and proposed plan actually connect.

Book a consultation

Related JSH Law analysis

• Child Impact Reports and Child Focused Courts
• Delay Is Not Administrative: The Human Impact of Court Delay
• From Incidents to Trajectories: Coercive-Control Evidence
• The JSH Law Six-Question Check

Research and official sources

• Adelle Banks — Professional curiosity, children’s lawyers and assessment frameworks
• Department for Education — Working Together to Safeguard Children 2026
• Cafcass — Child Impact Assessment Framework
• Youth Justice Board — How to Assess Children in the Youth Justice System
• Nuffield Family Justice Observatory — Children Deprived of Their Liberty
• Department for Education — Improving Outcomes for Looked-After Children in Complex Situations
• Ministry of Justice — Family Court Statistics Quarterly, April to June 2026