Do Multi-Agency Safeguarding Hubs Improve Outcomes for Children? New Research Raises Important Questions
Multi-agency working has become a cornerstone of safeguarding practice. The principle makes sense. When professionals from social care, police, health, education and other agencies bring together information, expertise and professional judgement, we should be better placed to understand risk and make effective decisions for children.
Multi-Agency Safeguarding Hubs (MASHs) have become a significant part of that approach across England, but an important new study published in The British Journal of Social Work asks a question that safeguarding leaders should be willing to ask of any established safeguarding system:
Is it actually improving outcomes?
The research, Do multi-agency safeguarding hubs improve child protection outcomes? Evidence from national administrative data, by Andy Bilson, Matthew A. Jay and Simon Haworth, examines what happened to child protection activity following the introduction of MASH arrangements across England, and its findings deserve attention.
What did the researchers examine?
Rather than focusing primarily on whether professionals felt communication or information sharing had improved, the researchers looked at national outcome data.
Using Freedom of Information responses alongside national administrative data, they examined MASH implementation across English local authorities and considered trends including:
Section 47 child protection investigations
Child protection plans
Serious incident notifications
Short and longer-term patterns following the introduction of MASH arrangements
This distinction matters. Safeguarding systems are often evaluated through process measures: whether meetings happen, information is shared, referrals are timely, procedures are followed and professionals understand their responsibilities.
Those things are important, but they are not the same as demonstrating that a system is achieving better outcomes.
A substantial increase in investigations
One of the most striking findings concerns the growth in Section 47 investigations. Nationally, the number of child protection investigations increased from 89,300 in 2009/10 to 224,520 in 2023/24 – an increase of more than 150%. Over the same period, referrals increased by just 3%, while child protection plans increased by around 40%. The proportion of investigations resulting in a child protection plan consequently fell from approximately 50% in 2010 to 28% in 2024.
The researchers then considered whether the introduction of MASH arrangements was associated with these patterns. In 9 of the 12 years examined, local authorities implementing a MASH experienced greater increases in Section 47 investigations than authorities without one. Over the longer term, the difference became particularly noticeable. By 2023, the average rate of Section 47 investigations in authorities with a MASH was 194.1 per 10,000 children, compared with 108.5 in authorities without one. Yet the difference in child protection plans was considerably smaller.
The research therefore identifies an association between MASH implementation and increased investigative activity, but not a proportionate increase in children subsequently becoming subject to child protection plans.
What about serious harm?
Perhaps the most important question is whether increased investigation has been accompanied by a reduction in serious harm. The study found little evidence that it has. When serious incident notification rates were compared between authorities with and without established MASH arrangements, the rates were very similar. The researchers therefore found no evidence within the data analysed that the presence of a MASH was associated with lower rates of notifiable serious incidents.
That does not mean that multi-agency working is ineffective, nor does it mean that individual MASHs are not making excellent decisions or protecting children, but it does challenge an assumption that safeguarding organisations should always be prepared to challenge:
More activity does not automatically mean more effective safeguarding.
We need to be careful about what the research tells us
There is an important caveat. This is an observational study and the researchers themselves make clear that the findings demonstrate association rather than causation. Local safeguarding systems are complex. Differences between areas may also be influenced by local policy, resources, demographics, deprivation, inspection activity, thresholds, workforce capacity and many other factors. There are also significant variations in how MASH models operate.
So the conclusion should not simply be:
“MASHs don't work.”
That would be an oversimplification of an important piece of research. Instead, the study gives us an opportunity to ask much better questions about how we evaluate safeguarding systems.
From safeguarding activity to safeguarding impact
This is perhaps the most important lesson for safeguarding leaders. We are very good at counting safeguarding activity.
How many referrals were made?
How many concerns were recorded?
How many Section 47 investigations took place?
How quickly were referrals processed?
How many people completed safeguarding training?
How many audits were completed?
How many cases were discussed?
These figures tell us something about the functioning of a system, but they don't necessarily tell us whether people are safer because of it.
A safeguarding system can be extremely busy while still failing to achieve the outcomes it was designed to deliver. That distinction should matter to every organisation, not only children's social care.
More information does not automatically mean better decisions
MASHs were designed partly around the principle that bringing information together enables professionals to see a fuller picture of risk.
That principle remains important, but information sharing itself is not the outcome. The question is what professionals do with the information.
Effective multi-agency safeguarding still requires:
skilled professional judgement
appropriate challenge
curiosity
proportionate decision-making
understanding of context
recognition of cumulative harm
clear thresholds
appropriate escalation
meaningful involvement of children and families
access to effective help and intervention
Simply creating a multi-agency structure cannot guarantee those things.
There is also a workforce question
The study raises another important issue: capacity. Every investigation requires professional time.
Where safeguarding systems increasingly generate investigations which do not subsequently lead to formal child protection intervention, we should be prepared to examine what that means for practitioners. High volumes of investigative activity may leave less time for relationship-based practice, early intervention, direct work with families and sustained support. It may also contribute to workforce pressure within systems already experiencing significant recruitment, retention and workload challenges.
The safeguarding question therefore becomes broader than:
“Are we identifying enough risk?”
We also need to ask:
“Are we directing our finite safeguarding capacity towards the people who need it most?”
What should safeguarding leaders take from this?
The research presents an important challenge for boards, safeguarding partnerships, senior leaders and those responsible for assurance.
We should not assume that because a safeguarding model is established, widely adopted or professionally supported, it is necessarily achieving its intended outcomes.
Instead, leaders should be asking:
What difference is our safeguarding system actually making?
That means looking beyond compliance and activity data.
It means understanding patterns and trends.
It means exploring the relationship between referrals, investigations, interventions and outcomes.
It means considering unintended consequences.
It means listening to practitioners and those experiencing safeguarding systems.
And crucially, it means being willing to challenge established practice when the evidence tells us something unexpected.
Safeguarding assurance must go beyond compliance
At RLB Safeguarding, we frequently talk about the importance of moving safeguarding assurance beyond compliance. A policy can exist and still be ineffective.
Training can have been delivered and still not change practice, a referral pathway can be clear and still produce poor decisions. A multi-agency structure can be functioning exactly as designed and still require us to ask whether it is producing the outcomes we intended.
Good safeguarding governance therefore asks three different questions:
Are we doing what we said we would do?
Are we doing it well?
Is it making a difference?
The third question is often the hardest but is also arguably the most important.
Final thoughts
The findings of this research should not lead us to reject multi-agency safeguarding. If anything, they should encourage us to strengthen it, but strengthening safeguarding does not necessarily mean creating more processes, collecting more information or generating more investigations.
Sometimes it means examining whether the system itself is helping professionals distinguish more effectively between risk that requires investigation, vulnerability that requires support and circumstances where families need earlier, practical help.
Safeguarding should never become a measure of how much activity a system can generate.
The real measure must always be whether our arrangements are helping us recognise harm earlier, respond proportionately and ultimately improve safety and outcomes for children and families.