Kent & Medway Homelessness by District (2022/23–2024/25): Section 21 Surges, While Thousands of Homes Sit Empty

Between 2022/23 and 2024/25, every council in Kent and Medway has been feeding detailed case-by-case homelessness data to Whitehall through the national H-CLIC system – the Homelessness Case Level Information Collection. These returns, made under Part 7 of the Housing Act 1996 as amended by the Homelessness Reduction Act 2017, are the statutory backbone of England’s homelessness statistics.

Using those datasets for 2022/23, 2023/24 and 2024/25, this article looks at what is happening in Kent and Medway when people present as homeless or on the brink of it – and sets those figures against the number of long-term empty homes across the county. It also explains, in plain language, what “statutory homelessness” actually means, and what the law expects councils to do.

What the data covers – and where it comes from

Every time someone goes to a local housing authority saying they are homeless or about to lose their home, the council must carry out a homelessness assessment. That process and the resulting duties are set out in:

• Housing Act 1996, Part 7 – the main homelessness law.

• Homelessness Act 2002 – which broadened prevention duties.

• Homelessness Reduction Act 2017 – which introduced new prevention and relief duties for anyone who is actually homeless or at risk within 56 days.

Councils must “have regard to” the Homelessness Code of Guidance for Local Authorities (18_Nov_2025) issued by central government. This Code, first re-issued in 2018 and updated several times since, tells councils how they are expected to interpret the law and how they should collect and report data.

Under this framework, councils upload case-level records to DLUHC via H-CLIC. The technical specification for H-CLIC defines every field – from how to record whether someone is threatened with homelessness, to the outcome of their case – and is backed up by detailed data quality and protection guidance.The spreadsheets you provided are DLUHC’s Table A1: “Number of households by initial assessment of homelessness circumstances and needs.” They show, for each authority:

• how many households were assessed;

• how many were judged threatened with homelessness within 56 days;

• how many were already homeless; and

• how many the council accepted a legal duty to help.

These are statutory homelessness figures – they are not rough sleeping head-counts, and they do not include everyone sofa-surfing or staying with friends.

Headline picture: thousands of households in crisis every year

Across Kent and Medway combined, the figures from 2022/23 to 2024/25 show a system under steady, unrelenting strain.

Using the official data for all 12 Kent district councils plus Medway:

• In 2022/23, at least 8,409 households were owed a prevention or relief duty.

• In 2023/24, that rose to 8,798.

• In 2024/25, the total dipped slightly to 8,549, but still remained above 2022/23 levels.

These totals combine two groups:

• households threatened with homelessness within 56 days (the prevention duty); and

• households already homeless and owed a relief duty.

The first chart shows how that breaks down.

• The lower part of each bar shows households threatened with homelessness.

• The upper part shows households already homeless.

Over the three years, the threatened-with-homelessness numbers are remarkably flat – around 4,500 households a year. The pressure is constant. The homeless part of the bar is what moves: households already without a home rose from about 3,886 in 2022/23 to 4,283 in 2023/24 and remained above 4,000 in 2024/25.

This matches the Kent County Council bulletin for 2023/24, which reported 11,627 assessments, with 3,468 households threatened with homelessness and 3,438 already homeless in the Kent County Council area alone (excluding Medway).

Which districts are most affected?

The picture is not even across the county. The second chart looks at 2024/25 and asks: How many households were owed a duty, per 1,000 households in the area? That lets us compare districts of very different sizes.

To do this, the DLUHC tables provide:

• the number of households owed a duty; and

• an estimate of the total number of households in each district.

Dividing one by the other gives a rate per 1,000 households – in other words, “out of every thousand households in this area, how many needed formal homelessness help this year?”

In 2024/25:

• Medway has the highest rate, at just under 19.4 households per 1,000.

• Ashford follows at around 14.3 per 1,000.

• Maidstone and Gravesham both sit above 13 per 1,000.

• Dartford and Thanet are a little lower but still in double figures, just under 13 per 1,000.

• At the other end of the scale, Sevenoaks and Tunbridge Wells have rates below 8 per 1,000.

Two caveats are important here:

1. For Canterbury, several key figures are suppressed or not available in the national release, so its rate cannot be reliably calculated from this table.

2. For Tonbridge & Malling, the 2024/25 “total owed duty” field is missing in the published data, so again the rate is not calculated.

Where the data exist, however, the pattern is clear: homelessness is a normal, regular occurrence, not an edge case, with several districts seeing more than one in every 75 households in crisis with the council in a single year.

Section 21 – the private-rented “tripwire” you can see in the homelessness data

One of the most revealing lines in your MHCLG/DLUHC Table A1 dataset is “of which: due to service of valid Section 21 Notice”. This doesn’t measure rough sleeping. It doesn’t even measure physical eviction. It measures something earlier — the moment a household has been served a valid Section 21 notice that is due to expire within 56 days, which (in law) automatically counts them as “threatened with homelessness”. 

What is Section 21 (in plain English)?
Section 21 of the Housing Act 1988 is the legal route that (historically) allowed many private landlords to end an assured shorthold tenancy without needing to prove the tenant did anything wrong — often described as “no-fault” eviction.

What does the Table A1 “Section 21” figure actually count?
It counts households who approached their council and were assessed as threatened with homelessness, specifically because a valid Section 21 notice had been served and would expire within 56 days. It is, in effect, a visible gauge of private rented sector instability feeding into statutory homelessness — but it is not the same as “evictions carried out”. Some notices are withdrawn, some tenants leave earlier, and some never approach the council at all. 

What the Kent & Medway numbers say (and why they matter)

Across Kent and Medway, the Section 21-linked figures in your dataset rise sharply over three years:

• 2022/23: 1,195 households recorded as threatened due to Section 21

• 2023/24: 1,673

• 2024/25: 1,787

That’s a jump of +592 households from 2022/23 to 2024/25 (around +50%). Even more striking is what it does to the mix of homelessness pressures:

• In 2022/23, Section 21 cases were about 26% of all “threatened with homelessness (prevention duty)” cases.

• By 2024/25, they are about 40% of that threatened-with-homelessness group — roughly two in five.

Put simply: a growing share of homelessness prevention work is being driven by people losing private rented homes through Section 21 notices.

Where the pressure is highest (district-by-district)

Your 2024/25 data show wide variation across Kent and Medway when Section 21 cases are expressed as a rate per 1,000 households (so that large and small districts can be compared fairly).

The highest Section 21 pressure in 2024/25 is in:

• Medway: 519 cases (about 4.53 per 1,000 households)

• Thanet: 219 (about 3.26 per 1,000)

• Gravesham: 128 (about 2.97 per 1,000)

• Folkestone & Hythe: 161 (about 2.93 per 1,000)

• Maidstone: 179 (about 2.39 per 1,000)

• Ashford: 137 (about 2.35 per 1,000)

At the lower end (where data are present), rates fall to around 1.09 per 1,000 in Sevenoaks and around 1.52 per 1,000 in Dover.

Two important data notes for transparency: in the 2024/25 table as supplied, Canterbury and Tonbridge & Malling have missing values in the Section 21 field, so they cannot be ranked reliably in the Section 21 chart.

Why Section 21 shows up as homelessness — even before anyone is “evicted”

The reason Section 21 matters in homelessness statistics is built into the legal definition of “threatened with homelessness”. The Homelessness Code of Guidance explains that a person is threatened with homelessness if they are likely to become homeless within 56 days — and it explicitly includes cases where a valid Section 21 notice has been served and will expire within 56 days. 

That means the homelessness system is effectively being asked to act as the shock absorber for the private rented sector: once a notice arrives, councils have a duty to step in and try to prevent homelessness or relieve it — often under severe time pressure.

The policy horizon: Section 21 is being abolished — but not yet

This is not just a historical story. In November 2025 the government published guidance explaining that the Renters’ Rights Act will abolish Section 21 evictions, moving landlords onto defined “grounds” (via Section 8-style routes) instead. 

However, the widely-used government “Evicting tenants” guidance notes the rules are changing on 1 May 2026, which means Section 21 pressures can remain highly relevant through 2024/25 and 2025/26, with the data likely to shift materially as implementation bites.

Changes over time: who is going up, who is going down?

Looking at the period from 2022/23 to 2024/25, and taking the number of households owed a duty in each district:

• Medway shows the largest absolute increase, from 1,793 households owed a duty in 2022/23 to 2,220 in 2024/25 – an extra 427 households a year.

• Ashford rises from 613 to 832, up 219.

• Folkestone & Hythe rises from 524 to 572.

Some districts have seen slight falls:

• Swale drops from 761 to 578, down 183, although the 2023/24 total was very low, suggesting local data or service changes.

• Gravesham falls from 658 to 585.

• Maidstone, Sevenoaks and Tunbridge Wells show modest reductions.

These shifts do not necessarily mean fewer people in housing need. They may reflect changes in who qualifies for help, how cases are recorded under H-CLIC, or how early people seek advice.

Empty homes: could they house everyone who is homeless?

The obvious question, once the homelessness numbers are put on the table, is: what about empty homes?

Kent County Council’s latest Housing Stock in Kent publication uses DLUHC’s Council Taxbase vacant dwellings tables to show how many homes have been empty long-term – defined as vacant for six months or more.

Table 12 of that report shows that in 2024:

• Kent’s district councils together had 7,317 long-term vacant dwellings.

• Medway had 1,135.

• Kent & Medway combined therefore had 8,452 long-term empty homes.

Compare that with the homelessness data:

• In 2024/25, Kent and Medway together owed a prevention or relief duty to 8,549 households – slightly more than the total number of long-term vacant homes in the area.

It is a powerful visual: one year’s flow of homelessness roughly equals the stock of long-term empty homes. In very crude terms, if every long-term empty home were magically brought into use, you could – on paper – house almost every household the councils had a statutory duty to help that year.

But reality is more complex:

• Some empty homes are in very poor condition, or awaiting demolition or redevelopment.

• Many are in the wrong place or the wrong size for the households who need help.

• Others are tied up in probate, legal disputes or speculative investment.

• Crucially, the homelessness figures are annual flows – new households each year – while the empty homes are a snapshot stock. Even if you filled all existing empties once, new homelessness would still arise the next year.

Nevertheless, the comparison highlights a policy choice: leaving thousands of homes empty while thousands of households rely on statutory homelessness duties is not inevitable, it is the result of how powers and incentives are used.

Councils already have tools to act, including:

• Council Tax premiums on long-term empty homes, which were strengthened again from April 2024.

• Empty dwelling management orders and compulsory purchase powers in extreme cases.

• Planning and regeneration powers to bring redundant stock back into use.

The Kent housing stock report itself notes that Ashford, for example, saw a 112.5% increase in long-term vacant dwellings in a single year, while Tonbridge & Malling saw a sharp fall. That suggests local policies and enforcement do make a difference.

What do these duties actually mean for a household?

For a lay reader, the alphabet soup of “prevention duty”, “relief duty” and “main duty” can be confusing. In simple terms:

• Initial assessment / statutory homelessness assessment
  This is the interview and investigation the council must carry out when someone applies as homeless. It looks at whether they are homeless or threatened with homelessness within 56 days, their local connection and their support needs. This is what Table A1 is counting.

• Threatened with homelessness (Prevention duty)
  The council has accepted that the household is likely to become homeless within 56 days. It must take “reasonable steps” to help them keep their current home or find somewhere else before they become homeless. This is the prevention duty created by the Homelessness Reduction Act.

• Homeless (Relief duty)
  The council accepts that the household is actually homeless. It must take reasonable steps to help them secure accommodation for at least six months. This is the relief duty. The household may be placed in temporary accommodation while this is happening.

• Main housing duty
  If the relief duty does not end successfully and the household is in a “priority need” group (for example, families with children, pregnant women, some care leavers, and people made homeless by domestic abuse) and is not intentionally homeless, the council owes a longer-term main duty to secure suitable accommodation. The detailed rules and examples are set out in the Homelessness Code of Guidance.

• Not homeless, not threatened / not eligible / withdrew
  Some assessments end with the council deciding that the applicant is not homeless, is not eligible (for example because of immigration status), or has withdrawn their application. These outcomes are also captured in the H-CLIC data and appear in the same table, but they are not the focus of this article.

All of these terms, and the way councils must apply them, are laid out in the Code of Guidance and the H-CLIC technical notes.

Beyond the spreadsheet: what this means on the ground

Numbers can be dry; what they represent is not.

On any given working day in Kent or Medway:

• Dozens of people are having those first, anxious conversations with housing officers.

• Others are being told they are “threatened with homelessness” and must work with the council on a personalised housing plan.

• Some will be placed into temporary accommodation – often hotels, nightly-paid rooms, or out-of-area placements – while a more permanent solution is sought.

The Kent County Council bulletin reports that, across the Kent districts alone (excluding Medway), an average of 2,091 households were in temporary accommodation at any one time over 2022/23–2023/24.

The legislation and guidance are clear that councils should not simply process people through a system:

• They must provide free, accessible information and advice on homelessness to everyone in their area, not just those already in crisis.

• Children’s services and housing authorities have joint responsibilities where 16- and 17-year-olds are at risk, and the “duty to refer” expects a wide range of public bodies to flag people at risk of homelessness early.

• Yet the statutory data show year after year of high volumes. That raises questions not only about the availability of affordable housing, but also about:

• the stability of the private rented sector;

• the use of Section 21 “no fault” evictions (which DLUHC tracks separately in the same tables);

• the adequacy of Local Housing Allowance and other benefits; and

• the way councils manage their own stock and nominations to housing associations.

Could things be different?

Taken together, the homelessness data and the empty homes figures suggest three broad conclusions.

First, homelessness in Kent and Medway is structural, not accidental. The fact that about 8,500 households a year need a statutory homelessness duty is not a one-off shock; it is how the system is functioning, year after year.

Second, the near-parity between long-term empty homes and annual homelessness duties is too close to ignore. No one can seriously claim that every empty home could or should be used for homeless households. But neither can anyone reasonably pretend that long-term empties are irrelevant to the homelessness crisis.

Third, the legal framework gives councils both duties and tools, but not always the resources to match. The Code of Guidance emphasises early prevention, personalised support and partnership working. H-CLIC provides detailed, comparable data that can be used to scrutinise local performance. Yet the persistence of high numbers suggests that policy levers – from planning, to enforcement on empty homes, to welfare levels – are being pulled only part-way.

As new updates to the homelessness Code of Guidance are published and H-CLIC continues to be refined, the data for Kent and Medway will keep coming. The question for local and national politicians is whether they are prepared to treat those numbers not just as a statistical bulletin, but as a mandate to act: on empty homes, on genuinely affordable housing, and on the precarious conditions which keep feeding people into the homelessness system in the first place.

The Shepway Vox Team

Discernibly Different Dissent