What the homelessness statistics don't show us about violence, abuse and prevention
Every quarter, the Department for Levelling Up, Housing and Communities (DLUHC) releases statistics on statutory homelessness in England and on 22nd July 2025 the latest release was published. On the surface, they provide valuable insights into trends around who is approaching local authorities for help, and why. But dig a little deeper, and the gaps become glaring.
These data releases continue to fall short in capturing the true scale, nature and impact of violence and abuse as a driver of homelessness, particularly for those impacted by serious organised crime, exploitation, and coercive forms of control. As our recent work on reviews into deaths and serious harm shows, those gaps in the numbers are matched by critical gaps in systems and responses on the ground.
A prevention duty that comes too late, if at all
The data tells us that domestic abuse is consistently one of the most common reasons for relief duty being owed, i.e. they are already homeless and in crisis. But what about the prevention duty, where a household is still in their accommodation but at risk of becoming homeless within 56 days?
Here, domestic abuse rarely appears among the top recorded reasons for prevention duty, suggesting that many survivors are not identified or supported until they are already in crisis.
This isn't a sign of success in early intervention. On the contrary, it suggests a significant failure to identify risk early, make timely referrals, and offer safe, planned alternatives to crisis-led flight. In our own reviews, we've consistently seen a lack of Duty to Refer activity from partner agencies, missed opportunities where frontline professionals in health, criminal justice, or community safety failed to trigger housing support before crisis hit. Many didn't even know it was a requirement.
The result? Survivors often reach housing teams only after they've already fled, when the situation has become so dangerous they've had no choice but to leave, sometimes with children, and often with no belongings or safety planning in place. Beyond the human cost, this crisis-led approach is far more expensive than early intervention, yet our current system consistently chooses the most costly option.
Under-recording, misidentification and blind spots
Even where violence, abuse or exploitation are present, they often go unrecorded or miscategorised. Unless officers are trained to ask the right questions with compassion and curiosity, they may simply log the reason as "relationship breakdown (non-violent)" or "rent arrears" — without ever exploring whether coercive control, financial abuse, or other forms of harm are underlying causes.
Without this nuance, data becomes distorted and the opportunity for early intervention is lost. These recording gaps particularly impact women, LGBTQ+ individuals, and people with multiple disadvantages, who face compounded barriers to accessing safe housing and may find it harder to disclose abuse.
We also need far greater clarity and consistency in how we record non-domestic forms of violence, including:
Serious organised crime
County lines exploitation
‘Cuckooing’ and home takeovers
Hate crime and targeted violence
Gang exit and retaliatory threats
These are often linked to community safety responses (e.g. closure orders, eviction notices, enforcement), yet housing responses lag far behind. We've seen cases where a person's home was shut down due to "anti-social behaviour" caused by exploitation, yet there were no safety plans, no safeguarding, and no housing response, with the individual subsequently found intentionally homeless and left in danger.
Scrap the vulnerability test, it's not fit for purpose
For those not in a priority group, access to housing support still hinges on a test of "vulnerability", essentially, whether they are more vulnerable than the average person if made homeless.
This approach is not only outdated - it's inhumane.
Let's be clear: anyone experiencing violence or abuse is at risk of serious harm if left without a safe place to live. The idea that we should be assessing whether someone might "cope better" with homelessness, as if survival is a meaningful measure of resilience, is barbaric.
People forced to flee violence aren't choosing between safe and unsafe, they're choosing between degrees of risk. Between staying in a home where they are targeted, or sleeping rough. Between continuing abuse or further exploitation. Between trauma or re-traumatisation.
If you need evidence, look at CHAIN data from London, which shows that a significant proportion of rough sleepers report past experiences of domestic, sexual, or institutional abuse, with particularly high rates among women and LGBTQ+ rough sleepers. We need to stop pretending this is a "vulnerability test" and start treating it like the safeguarding and human rights issue it is.
Missed connections: homelessness, crime, and violence prevention
The Duty to Refer was introduced to ensure public services play an active role in preventing homelessness. But without quality assurance, accountability or proper data linkage, it's becoming a tick-box at best and completely invisible at worst.
We believe this duty should be mapped against local crime data, safeguarding alerts and known cohorts of individuals experiencing violence and abuse. That way, we can see where duty to refer isn't being used, and where local partnerships need to do more. Homelessness data should be triangulated with police, Violence Reduction Unit (VRU), and health data, not kept siloed.
While some Violence Reduction Units are beginning to recognise housing as a core partner, this remains the exception rather than the rule. We want to see VRUs systematically embed housing as a collocated partner in their work, not just an afterthought. Housing must be part of the violence prevention ecosystem. That means placing homelessness prevention alongside other public health measures, with shared responsibility, shared data, and shared accountability.
A blueprint for change
Addressing these systemic failures requires concrete action:
Immediate reforms: Mandatory trauma-informed training for housing officers, regular audits of Duty to Refer compliance, and standardised risk assessment tools that can identify coercive control and exploitation.
System integration: Data-sharing agreements between housing, police, and health services, with homelessness prevention embedded in all violence prevention strategies.
Practice transformation: Recognition that significant variations exist across local authorities, with the best integrated approaches serving as models for national rollout.
Better data, better duties, and a more humane response
If we want to truly understand the relationship between violence, abuse and homelessness, we need data that reflects lived realities, not administrative categories. We need to train frontline professionals to ask the right questions. We need to scrap outdated vulnerability tests that ignore trauma. And we need to strengthen the duty to refer into a system-wide duty to protect.
Because right now, far too many people are falling through the cracks and the cost is measured in harm, trauma and sometimes lives.