Kent Housing Conditions: Government Data Models One in Five Homes as Non-Decent in Four Council Areas

Less than two years after Kent’s own housing officers said the evidence base on local housing conditions was “scarce”, a new government model has put numbers against the problem. Dover, Folkestone & Hythe, Thanet and Tunbridge Wells all sit above the South East’s statistical band for non-decent homes. These aren’t local survey totals and they aren’t a league table. But they’re a warning that now needs testing at the front door.

In November 2024, housing officers working across Kent and Medway admitted something rather awkward. Their evidence base on local housing conditions was “scarce”. They hadn’t found good evidence showing how many homes contained the most serious Category 1 hazards, how many had actually been found after inspection, or even which hazards were turning up most often. The Shepway Vox Team reported that gap the following month. It was a sizeable blind spot in an area where 145,566 households — 19.2% of the total — were classed as living in the private rented sector, including rent-free households, at the 2021 Census.

Whitehall has now tried to illuminate part of it.

On 25 June, the Ministry of Housing, Communities and Local Government (MHCLG) published its 2024 local authority housing stock condition model. It estimates, for every local housing authority in England, the proportion of occupied homes likely to be non-decent, to contain a Category 1 hazard or to have a damp problem. The government labels the series “Official Statistics in Development”: the statistics are official, but the methods and outputs are still being developed and evaluated.

The headline for Kent is difficult to shrug off. The English Housing Survey estimates that 16% of homes across the South East were non-decent in 2024. Because surveys contain uncertainty, MHCLG gives that regional figure a 95% confidence interval of 15% to 18%. Put simply, 15% to 18% is the band within which the regional estimate would be expected to fall 95 times out of 100 if the survey were repeatedly run in the same way.

Against that band, four Kent and Medway council areas are modelled above it: Dover at 21.3%, Folkestone & Hythe at 20.9%, Thanet at 20.5% and Tunbridge Wells at 19.8%. Dartford, at 13.9%, and Tonbridge & Malling, at 14.0%, are below it. Ashford, Canterbury, Gravesham, Maidstone, Medway, Sevenoaks and Swale sit within it. MHCLG itself singles out “the South coast of Kent” as an area where modelled non-decency exceeds the South East’s upper bound.

That’s roughly one home in five modelled as non-decent in each of the four areas above the regional band. But there’s an important wrinkle in the geography. Three — Dover, Folkestone & Hythe and Thanet — fit the government’s coastal-Kent description neatly.

Tunbridge Wells doesn’t.

Any easy story that this is simply a seaside-deprivation problem therefore falls apart before it’s got going.

Here’s the complete Kent and Medway picture. We’ve kept it alphabetical deliberately, because MHCLG says these figures shouldn’t be used to rank individual councils against one another.

The first percentage column comes from MHCLG’s accompanying South East figure file. The last comes from the department’s revised 2024/25 Local Authority Housing Statistics return and means something much narrower: private rented dwellings which, following an inspection, were found to have one or more Category 1 hazards. Section F was first published in January and the return was revised on 25 June. Kent County Council isn’t missing by accident: this particular dataset works at local housing authority level, meaning Kent’s 12 district and borough councils plus unitary Medway.

What “non-decent” actually means

A non-decent home isn’t simply one with mould on a wall, nor does the label automatically mean a landlord has broken the law. Under the existing Decent Homes Standard, a home is decent only if it meets four tests: it passes the statutory minimum safety standard, is in a reasonable state of repair, has reasonably modern facilities and services, and provides a reasonable degree of thermal comfort. A Category 1 hazard — the most serious HHSRS safety category — causes a home to fail the first of those tests.

That distinction matters especially in private renting. The existing Decent Homes Standard is regulatory in the social rented sector, but it isn’t currently the regulatory standard for private rented or owner-occupied homes. Private rented housing is presently policed on condition mainly through the Housing Health and Safety Rating System, or HHSRS. Nationally, 22% of private rented homes were non-decent in 2024, compared with 15% of owner-occupied homes and 10% of social rented homes. Category 1 hazards were found nationally in 10% of private rented dwellings, 9% of owner-occupied dwellings and 5% of social rented dwellings.

So Folkestone & Hythe’s 20.9% does not mean 20.9% of its homes contain a Category 1 hazard. A home can fail the Decent Homes Standard because of serious disrepair, outdated facilities or inadequate thermal comfort without having a Category 1 hazard. Conversely, when a council actually identifies a Category 1 hazard under the HHSRS, it has a legal duty to take enforcement action. Those are related measures, not interchangeable ones.

There’s another update worth making to our own earlier coverage. In December 2024 we correctly described the HHSRS as containing 29 hazards. That changed on 23 June this year. The revised system now groups them into 21 hazards, with updated scoring and descriptions after statistically similar hazards were amalgamated. Councils must act when a hazard reaches Category 1 and may act on Category 2 hazards. Since 22 June, councils have also been able to impose a financial penalty of up to £7,000 where the legal conditions are met for a landlord’s failure to remove a Category 1 hazard.

That makes the quality of local intelligence rather more than an academic concern.

A warning map, not 13,000 surveyors with clipboards

Here’s the catch with the new MHCLG figures: nobody inspected every home in Dover, Folkestone, Margate or Tunbridge Wells to produce them.

The model was built from the English Housing Survey, including physical inspections carried out between April 2023 and March 2025, then combined with local information including 2021 Census data, 2025 Experian dwelling data and Energy Performance Certificate records lodged up to March 2025. It uses characteristics associated nationally with poor housing to estimate what the local picture is likely to look like. The local figures cover occupied homes, and MHCLG says the resulting numbers are indicative and mustn’t be treated as surveyed totals.

There’s even a technical wrinkle buried beneath that. The local authority estimates are household-weighted, broadly meaning households in occupied properties are being counted, whereas the regional figures used for context are dwelling-weighted, meaning properties are counted. MHCLG nevertheless explicitly uses the South East’s regional confidence interval as its framework for deciding whether a council area looks above, below or similar to the regional picture. That’s another reason not to pretend these are simple raw percentages collected in identical ways.

The strongest predictors help explain why Kent’s pattern looks the way it does. Older housing — particularly pre-1919 stock — pushes modelled non-decency and serious hazards upwards. Low levels of central heating matter. A larger private rented sector is associated with higher non-decency and damp. For damp, older homes, private renting, disrepair and uninsulated cavity walls are among the strongest drivers. Newer, post-1980 stock and a larger social rented sector tend to pull predicted non-decency down.

That fits part of the evidence Kent already had. The Kent Housing Group’s work found those 145,566 Kent and Medway households in private renting in 2021, a rise of 14.2% since 2011. Its housing officers subsequently said poor conditions were likely to affect a large number of properties, particularly in the private rented sector, while identifying local housing-condition information as a key data gap. What MHCLG has now added is a consistent model applied across every authority, rather than thirteen councils measuring the problem thirteen different ways — or, in some cases, barely measuring it at all.

But the model can’t tell us which street to knock on. MHCLG expressly warns that small, atypical pockets of poor housing can be missed. It says councils should use these figures alongside local intelligence, not as a replacement for a proper local stock-condition survey. Nor should anyone use the release to claim one council’s enforcement team is better than another’s, or to say conditions have definitely improved or deteriorated since an earlier model.

And this is crucial: MHCLG says its confidence-band approach “does not represent true statistical testing”. Outside that approach it specifically discourages direct comparisons between local authorities, whether in the same year or across different years.

It would therefore be wrong to write “Dover has Kent’s worst housing” merely because 21.3% is fractionally higher than Folkestone & Hythe’s 20.9%. The model simply isn’t precise enough for that contest.

What it does support is a stronger and more defensible conclusion. MHCLG has identified characteristics in Dover, Folkestone & Hythe, Thanet and Tunbridge Wells suggesting their housing quality is markedly worse than the South East pattern. What that patterns does show is landlords are averse to spending money to prevent issues and only spend reactively in most cases.

Then there’s the awkward bit: what councils actually find

Now put the model beside the councils’ own 2024/25 returns and things get interesting.

Folkestone & Hythe reported 197 private rented dwellings found after inspection with at least one Category 1 hazard. Thanet reported 92, Medway 56 and Tunbridge Wells 30. Dover, despite being in MHCLG’s “above” group for modelled non-decency, reported 19. Swale reported two, Sevenoaks three and Maidstone zero. These are the figures in MHCLG’s revised June return, not estimates produced by us.

The temptation is to call that a contradiction.

It isn’t.

The 197 and the zero measure discovery, not prevalence. A council can only put a home into that particular figure after an inspection has found one or more Category 1 hazards. A high number might reflect poorer stock, energetic inspection work, a targeted programme, more tenants coming forward, better recording — or some mixture of them. A low figure could reflect healthier housing. It could equally reflect fewer inspections, less proactive enforcement, fewer complaints reaching the authority or gaps in its information. Without knowing how many homes each authority actually inspected, the number on its own can’t tell us which explanation is right. MHCLG itself warns that local-authority administrative information is gathered for particular purposes and requires careful interpretation.

Kent’s housing professionals had already spotted this problem before the new model arrived. Their November 2024 note said LAHS data had shown some authorities with no Category 1 hazards and concluded it wasn’t a complete picture. The final Kent and Medway housing evidence base was built precisely because local knowledge needed strengthening. The new model doesn’t magically cure that problem.

It makes the gap harder to ignore.

For Folkestone & Hythe and Thanet, two different datasets now point in broadly the same uncomfortable direction: both have above-band modelled non-decency and comparatively large numbers of inspected private rented homes where serious hazards were actually found.

Dover and Tunbridge Wells pose a different question. If the characteristics of their housing stock point to unusually high non-decency, what does their inspection programme find when officers get through the door?

Then look the other way. Maidstone reported no private rented dwellings found with a Category 1 hazard in that return. Swale reported two and Sevenoaks three. Are those pleasing numbers evidence of healthier housing, or are too few properties being inspected for serious problems to be uncovered?

The published figures can’t settle that. That’s precisely why the councils now need to.

There’s a public-health reason not to leave this sitting in a spreadsheet. The HHSRS is explicitly a health-and-safety system, judging the risk housing defects pose to occupants and visitors, particularly vulnerable people. In the national 2024 survey, the most common Category 1 hazards included falls on stairs and excess cold, alongside falls on the level, falls between levels and damp. Significant damp for these purposes means penetrating damp, rising damp or serious condensation — not a bit of harmless bathroom mist.

Kent now needs the numbers behind the numbers

The next useful dataset shouldn’t come from another model. It should come from the councils.

Every Kent housing authority ought to be able to say when it last carried out a statistically robust private-sector stock-condition survey; how many private rented homes it inspected in 2024/25 and 2025/26; how many inspections were proactive rather than triggered by a complaint; which of the 21 current hazards are being found most often; where those homes are concentrated; and how long it takes from a complaint or inspection to the danger being removed.

It should also be possible to see how many cases are settled informally, how many lead to improvement notices, prohibition orders or emergency action, and how many end with civil penalties or prosecution. Without that information, residents get a numerator — 197, 92, 19, zero — but no meaningful denominator.

Staffing matters too. A council can have every power Parliament can devise, but an enforcement regime is only as real as the people available to use it. The sensible follow-up is therefore the number of qualified officers, full-time-equivalent posts, vacancies and inspection caseloads at each of the 13 housing authorities, placed beside the number of properties actually inspected.

That would tell us considerably more than a league table ever could.

The timing is hardly academic. The new private rented tenancy system under the Renters’ Rights Act has been in force since 1 May 2026. The revamped 21-hazard HHSRS followed in June, together with the additional power to issue penalties of up to £7,000 for qualifying failures involving Category 1 hazards. And from 2035, the new Decent Homes Standard is due to become regulatory across both social and private rented housing. The direction of travel is unmistakable: councils are going to need better intelligence about the condition of rented homes, not less.

The government’s model is therefore neither a conviction nor an excuse.

It doesn’t prove that one particular home is unsafe. It doesn’t prove that one council is failing. And it certainly doesn’t give anyone licence to turn uncertain estimates into a cheap league table.

What it does is point.

It points towards south-coast Kent.

It points, unexpectedly, towards Tunbridge Wells as well.

It points back to the same evidence gap Kent’s own housing officers acknowledged in 2024.

And it asks a question that can no longer be answered with “we don’t know”.

A model can’t knock on a front door. Councils can.

Have a story that needs telling, or information you think may interest us? Please get in touch. All correspondence will be treated in the strictest confidence. TheShepwayVoxTeam@proton.me

The Shepway Vox Team

Dissent is NOT a Crime