By Richard Devine, Social Worker and Author of Messy Social Work: Learning from Frontline Practice with Children and Families and co-host of the Messy Social Work Podcast
After eleven years in Bath and North East Somerset, and five before that in North Somerset, I’m leaving frontline children’s social work for a leadership role in Solihull.
Before I go, I wanted to reflect on some of the lessons the work has taught me. These are not conclusions, certainties, or best practice guidance. Most emerged slowly through relationships, mistakes, conflict, uncertainty, and time.
Looking back, I think many of them amount to learning how to tolerate complexity a little better.
Lesson 1
When I first became a social worker, I think a part of me believed that enough effort, skill, or commitment could meaningfully change the course of people’s lives. Over time, I came to understand the limits of that idea.
My involvement in families’ lives was pretty limited especially when set against difficulties that were deeply entrenched, long standing, and systemic.
It was unrealistic to assume I, alone, through my therapeutic relationship, could undo the effects of poverty, trauma, structural inequality, addiction, violence, or intergenerational harm.
Some parents could not overcome a lifetime of trauma to provide safe care.
Some managed to reduce the most harmful aspects of their behaviour so that removal of their children could be avoided.
Occasionally, some made extraordinary changes: giving up long-term addictions, leaving violent partners, or finally seeking help with their mental health.
And in every case, life continued after I left.
Some parents who made little progress during my involvement later turned things around. Others relapsed. Most continued to face hardship.
My influence was uncertain; often small, sometimes meaningful, never singular.
In some ways, I was unimportant. In other ways, who I was and how I showed up mattered a great deal.
Lesson 2
There were weeks where I finished on a Friday having worked fifty or sixty hours, only to feel further behind than when I started.
Eventually, I realised the problem wasn’t my efficiency, commitment, or resilience. It was the illusion of control.
I couldn’t do everything.
So I had to choose what to prioritise, and, more importantly, what I would allow to fall away.
That meant living with dissatisfaction: unanswered emails, delayed tasks, and the knowledge that someone, somewhere, would probably be unhappy with me.
For a long time, I experienced this as personal failure. Eventually, I understood it differently. The work was not doing everything. The work was choosing what I couldn’t do, and living with the consequences.
That felt more honest than believing that enough hard work could make impossible demands manageable.
Lesson 3
I also realised something else that changed how I practised.
My role was not to make children completely safe. My role was to reduce the likelihood of significant harm.
That distinction mattered.
The question, “How do I make this child safe?” implied certainty, control, and total responsibility.
Reframing the task as, “How do I reduce harm to a level where this child is no longer experiencing significant harm?” created a different kind of practice.
One more honest about limits and what could realistically be achieved.
Lesson 4
At the same time, I came to understand that “the system” was not something abstract or distant. It took shape through ordinary interactions, moment by moment.
Each time I showed gentleness to a child, intellectual rigour in an assessment, compassion towards a parent harming themselves or others, or gratitude to a colleague…
…I was helping define what child protection looked like in practice.
In those moments, I was “the system”. I was social work.
Lesson 5
I learned that parents appreciate me being real. I might swear occasionally, laugh with parents, sometimes even about serious things, and took an interest in them beyond the concerns that led to me being involved.
If they asked how I old I was or how many children I had, I just told them. If they offered me a cup of tea, I accepted it, even when I didn’t really want one. Sometimes receiving is giving.
As well as being real, parents valued respect, often found in small gestures. Can I ask how you like to be called? Is it Chris, Christopher, or something else? How do you spell that? Shall I take my shoes off before I come in?
Lesson 6
I also learned that compassion and accountability were not opposites. I could empathise with behaviour that was self-destructive or harmful to others, including children, while still being clear about what needed to change and what the consequences might be if it didn’t.
In many ways, the main difference between the parents I supported and the children I was trying to protect was simply the passage of time. As Crittenden puts it, ‘today’s children are tomorrow’s parents‘.
Understanding that never removed accountability, but it changed how I saw people.
When I noticed myself becoming judgemental, emotionally closed off, or disengaged, I tried to treat that as an ethical warning signal rather than a personal failing.
I came to think of compassion as something that grows from understanding; so, when it was absent, it was a prompt to deepen my understanding.
Lesson 7
I also learned to tolerate my own mixed feelings. There were parents I admired, parents who frightened me, parents who exhausted me, and others who made me laugh. Sometimes all of those feelings existed at once.
Pretending otherwise would have been dishonest, and in child protection work, potentially dangerous.
Lesson 8
One of the most important relational skills I developed was learning to approach parents from a position of deliberate curiosity and “not knowing.”
I tried to assume that the parent held something important that I did not yet understand.
Tell me, why do you hate social care? Why do you take drugs? What are the benefits for you? How do they help? Can you help me understand…
Consistently, I found that people became less defensive when they felt fully understood.
Even when I strongly disagreed with parents, I made it my job to show I properly understood and was deeply interested in their perspective.
Lesson 9
As my time in B&NES went on, I realised that much of the work could not be reduced to right answers, models, or best practice.
Instead, it was shaped by tensions that never fully resolved.
Less is more, even when it feels uncomfortable
Long child protection plans involving many professionals often reduced my anxiety because they created the feeling that something substantial was happening. It made me feel in control.
Shorter, clearer plans with fewer professionals sometimes felt more exposing, but they often reduced family distress and created greater clarity about what needed to change.
This is why I like Crittenden’s idea of the critical cause of danger and the critical cause of change.
Recording matters, and it’s never enough
Recording provided accountability and protection for families and professionals.
At the same time, not everything important could be recorded. Some of the most meaningful moments in practice left no trace on a system.
Openness is not always a reliable indicator of safety
In practice, we often feel more reassured when parents are open and talk to us. It reduces uncertainty and helps us feel like we understand what’s going on. When a family shares information and engages well, it can give a sense that risks are recognised and manageable.
On the other hand, when parents are more guarded or hard to engage, it can make us feel more anxious. Not having clear information can make the situation feel riskier and harder to assess.
But this difference is really about how certain we feel, not necessarily how safe the child is.
A parent can be open, cooperative and easy to work with, and still be harming their child.
Equally, a parent who is quiet, cautious or difficult to engage may still be providing safe and caring support.
Engagement and agreement are not the same as change
I have worked with some parents who engaged fully with services or accepted professional concerns, yet made very little meaningful change in their behaviour. Others resisted, challenged, or disagreed with professional involvement, but still took steps that reduced risk over time.
Over time, I learned not to place too much weight on engagement, agreement, or expressed insight. While these can feel reassuring, they are not reliable indicators of safety.
Assessments can easily become focused on whether parents are cooperating, attending services, or acknowledging concerns. In practice, I came to see that the more important question is whether anything is actually different for the child.
In some cases, parents did not fully agree with the concerns I held, but were still able to follow safety plans and reduce risk in consistent and observable ways. In others, parents agreed with everything, yet little changed.
I therefore came to focus less on what parents said, or how they engaged with me, and more on whether their behaviour was changing in ways that improved their child’s safety and wellbeing.
Help and harm can coexist
Social work is built on the belief that state intervention can improve the lives of children and families. Sometimes it does. But intervention can also feel intrusive, disempowering, frightening, or harmful.
Good intentions do not guarantee good outcomes.
Some families experience our involvement as helpful. Others experienced it as stressful. Often, both things existed at once.
In many ways, this is also true of parents. Deliberate cruelty towards children is rare. Harm is more often tangled up with trauma, desperation, emotional immaturity, addiction, fear, or survival.
Early intervention helps, except when it doesn’t
There is a common belief in social work that when families reach crisis, it is because intervention came too late. Often, early help is effective and forms an important starting point, helping to understand the nature of a family’s difficulties.
However, for some families, difficulties are longstanding and complex, and cannot be resolved through brief or universal support alone.
It is therefore misleading to assume that earlier intervention would prevent escalation in all cases. The issue is not simply timing, but whether the level of support matches the level of need.
In these situations, the task is not to intervene earlier, but to provide support that is sufficiently sustained and intensive.
Lesson 10
I learned that some deep, long-term problems require equally sustained solutions.
Where harm and need are intense, support must be equally intense, creative, and sustained. There can be a deep resistance to this kind of intensive provision, usually driven by a fear of “creating dependency” or a belief that we cannot afford long-term support.
However, the reality is that some families *already* receive long-term state intervention.
They spend a decade cycling in and out of the system – two years on a child protection plan, a year stepped down to child in need, a crisis, and then straight back into child protection.
By default, the system ends up offering long-term support, except it is delivered in a fractured, short-term, crisis-focused way that doesn’t help.
Furthermore, taking children into foster care is itself an incredibly resource-heavy, intensive form of long-term intervention.
Instead of pouring massive resources into chronic cycles of crisis and family separation, we can choose to offer that same intensity upfront in a constructive, meaningful way – whether through intensive psychological therapy, peer support networks, or fostering entire families together to keep them safe and together.
Lesson 11
I now believe that effective child protection rests mostly on strong fundamentals: thoughtful direct work, clear analysis, solid recording, respectful multi-agency relationships, and the ability to tolerate conflict, uncertainty, and resistance.
Most good social work involved doing basic, ordinary things well, even when they were repetitive or unglamorous.
Lesson 12
After reviewing the research on most effective treatments for traumatised children, American Psychiatrist, Bruce Perry, concludes, ‘what works best is anything that increases the quality and number of relationships in the child’s life‘ (p.80).
Often, this meant that small interventions rather than dramatic ones could make a huge difference, such as referrals to a Family Group Conference, a school breakfast club, and extracurricular activities like Scouts, Brownies, and sports clubs.
I didn’t need to be the hero. Realising that probably made me more effective.
Lesson 13
I have spent 15 years asking parents what they want for their children, their answers were usually remarkably similar: safety, happiness, education and good relationships.
That shifted the question from:
“Why can’t you meet your child’s needs?”
to:
“What psychological, relational, or contextual barriers are getting in the way of you providing your children what you know they need?”
Over time, I started thinking similarly about social workers.
When practitioners are asked what kind of social worker they want to be, their answers are usually aspirational.
The task of leaders and organisations is not simply to demand better practice, but to understand the conditions that support or obstruct it.
Lesson 14
One of the most important moments in my career came when I asked a manager how I was supposed to support a family experiencing an adoption breakdown.
She simply replied:
“I don’t know, Rich.”
She could have hidden behind procedure or offered false certainty, but she didn’t. Her honesty gave me permission to say “I don’t know” too.
Sometimes we really don’t know, and thank fuck for that (I told you!).
I wouldn’t want to carry the expectation of knowing all the time.
Final thought
In the words of Matilda, sometimes you have to be “a little bit naughty.”
In my book, Messy Social Work, I wrote about Mia, and I’ll end here with that memory.
Mia was a small eleven-year-old girl who had recently been removed from her parents’ care and placed into foster care. She had just finished supervised family time with her dad and grandparents. They had left, and we were standing in the car park waiting for her foster carers to arrive.
Suddenly, Mia swung open my car door, scrambled onto the roof, and looked down at me mischievously, waiting for my reaction.
I couldn’t help smiling.
Then she started dancing.
So I leaned into the car, turned the music up loud, and joined in – Mia dancing on the roof, me dancing below her in the middle of the children’s centre car park on a warm sunny afternoon.
That interaction will never appear in a report, assessment, or electronic case file. No inspector will ever know it happened.
But I hope it exists somewhere in Mia’s heart.
I know it exists in mine.
By Richard Devine (25.06.26)
Join a growing community of 1200+ others by signing up for free blogs to be sent directly to your inbox
