Resolves to the local authority via postcodes.io.
England only: homelessness law and reporting are devolved, so Scotland, Wales and Northern Ireland are shown as outline only, no data.
SDWH Limited has no relationship with the local authorities shown. All figures are compiled from public sources (gov.uk statistical releases, MHCLG and, for the partial funding column only, local press reporting of individual council announcements). This is general information, not financial or policy advice, and 12 authorities have TA totals suppressed in the source data itself (small-numbers disclosure control), not a gap in this build. Rough sleeping figures are not subject to the same disclosure control and are shown for all 296 authorities.
England-wide summary
Headline totals and the household-type breakdown, summed across all 296 authorities in this snapshot, 31 March 2026.
Rough sleeping and the £442m package
People sleeping rough on a single night, autumn 2025 (official statistics), and the latest monthly management-information estimate (RSDF).
Accommodation by type
Net homelessness spend by type
Net current expenditure by TA type, summed across the matched authorities, 2024-25 outturn. Green bars are a net income position (subsidy and rent recovered exceeded gross cost on that type); blue bars are a net cost.
Don't divide these figures by the household counts above. The spend figures are a flow: money spent across the whole 2024-25 financial year. The household counts are a stock: a headcount on a single date, 31 March 2026, over a year later. Dividing one by the other produces a number that looks like a unit cost but isn't one. For example, nightly paid self-contained accommodation nets to roughly £602m against 49,325 households recorded on the later date, an implied “£12,202 per household” that is not a nightly rate, an annual rate, or any other real unit; it's an artefact of dividing a year of spending by a headcount from a different date entirely. Comparing it against the actual length-of-stay data below (roughly two years average for this type) shows why: over a period that long, the two sources are closer to describing the same population, but they still don't line up cleanly enough to divide safely. Read spend and household figures side by side, not as a ratio.
Length of stay in temporary accommodation
Household-weighted average length of stay by accommodation type, England, over the five quarters this breakdown has been published (31 March 2025 to 31 March 2026). Longer stays are why the spend and household figures above shouldn't be divided into each other.
Sources & vintages
All figures below are drawn from official UK government statistical releases (gov.uk and MHCLG) and reproduced here under the Open Government Licence v3.0. This page reorganises and joins those releases against a common local authority list for information purposes only. It does not create new data, and the licensing and sourcing conditions of each original release are unaffected by their presentation here. For the authoritative figures, follow the links below to the original release.
- Households in TA, by type, and household composition: H-CLIC / "Live tables on homelessness", table TA1, gov.uk, as at 31 March 2026. 12 of 296 authorities have a suppressed or missing TA total in the source data itself ([x]/[c] disclosure control on small numbers), listed as "suppressed" rather than zero.
- People sleeping rough, single night: Rough sleeping snapshot in England, autumn 2025, gov.uk, accredited official statistics, published 26 February 2026. Not subject to small-numbers suppression; all 296 authorities have a figure. England total 4,793, the highest on record, up from 4,667 in autumn 2024 and above the previous 2017 peak of 4,751.
- Rough sleeping, monthly management information: Rough Sleeping Data Framework (RSDF), MHCLG, March 2026 (latest available month; published quarterly in arrears). MHCLG's own term for this: management information, explicitly not official statistics, based on local returns rather than an independently audited count. Shown here as a more current, higher-frequency companion to the annual snapshot above, not a replacement for it.
- Local authority funding from the £442m rough sleeping package (partial): Prime Minister Andy Burnham announced £340m in July 2026 and a further £102m on 18 August 2026, aimed at getting everyone sleeping rough into accommodation by Christmas 2026. No single consolidated local-authority-level table has been published yet; the 11 figures shown in the table below (of 296 authorities) are drawn from named local press reporting of individual council announcements between 18 and 19 August 2026: North Northants Telegraph (Northamptonshire), West Bridgford Wire (Nottinghamshire districts), and Blog Preston (Preston). Treat these as a reliable but incomplete first cut, not an official allocation table; the remaining 285 authorities are marked "not yet published", not zero. This column will be replaced with the official table as soon as MHCLG publishes one.
- Band D council tax, excluding parish precepts: gov.uk Live tables on council tax, 2026-27 rates.
- Local authority homelessness spend (gross and net current expenditure, by TA type): MHCLG Revenue Outturn housing services (RO4), TOTAL HOMELESSNESS lines, 2024-25 outturn (the latest audited year available; outturn data always lags the current year). Barnsley and Sheffield have no matching RO4 2024-25 return under any code variant tried, a genuine gap in that source's own publication (410 of 411 returns included), not a join failure.
- Reading the net spend figures: per MHCLG's own RO4 guidance (Group 80, Homelessness), councils record gross costs against each TA type (furniture, property maintenance, rent guarantees, nomination fees, and Housing Benefit paid to homeless people), then net off two things recorded specifically against that same type: Housing Benefit subsidy received for it, and rent contributions from tenants/licensees. This netting rule applies identically to every TA type, B&B included, not just LA/HA stock. A £0 for a type means the council reported no net RO4 expenditure against that category for 2024-25: a real reported value, not a placeholder ("n/a" is reserved for genuinely unavailable figures: Barnsley/Sheffield's RO4 gap, and the "Other" category, which has no RO4 equivalent). A negative figure means the subsidy and rent recorded against that type exceeded the gross cost recorded against it that year, which can happen for any type. It shows up most easily at small counts: with only a handful of households in a category (as at many smaller authorities), a single case's Housing Benefit subsidy, including backdated amounts relating to an earlier period but received in-year, can outweigh the modest gross cost recorded, without implying anything unusual happened. Because H-CLIC's household counts (31 March 2026) and RO4's spend categories (2024-25, a different year and a slightly different classification) don't share a period or an exact definition, a type with households recorded but £0 net spend against it is not proof that accommodation was free. It more often reflects the mismatch between the two sources.
- Stock vs. flow, and why not to divide spend by households: the household counts on this page (H-CLIC) are a stock, a headcount on one date. The spend figures (RO4) are a flow, a total across a twelve-month year. Dividing a year of spend by a one-day headcount from a later date does not produce a nightly rate, an annual rate, or a meaningful unit cost of any kind, even though the arithmetic is easy to do. This page still shows a "spend per household in TA" map metric and table column because, read as a rough per-authority indicator rather than a literal unit cost, it is useful for comparison; treat it that way, and see the length-of-stay chart in the summary section for why the mismatch matters less for some types (long average stays) than others (short ones).
- Average length of stay in TA: H-CLIC / "Live tables on homelessness", table TA7, gov.uk. Published quarterly only since the quarter ending 31 March 2025; the chart in the summary section above uses all five quarters published so far (31 March 2025 to 31 March 2026). MHCLG publishes local-authority-level median and mean days by type, not a national figure; the England-level lines shown here are this page's own household-weighted average of the published local authority means (weighted by each authority's household count in that type, from the same quarter's TA1 table), not an official MHCLG statistic. Medians cannot be validly combined this way, so only means are aggregated to England level.
- These sources are different vintages: 31 March 2026 TA household counts, autumn 2025 rough sleeping counts, March 2026 RSDF management information, 2026-27 tax rates, 2024-25 spend outturn, and funding figures announced 18-19 August 2026, shown together for local context, not as a single "as at" snapshot. Treat cross-source comparisons (e.g. spend per household in TA) as indicative, not exact.
- Boundaries: Office for National Statistics, 2024 local authority districts. Local government reorganisation is live in parts of England; this snapshot uses the pre-reorganisation 296-authority geography that matches the data pull itself.
- Postcode geocoding: postcodes.io, built on ONS data.
- Basemap tiles: © OpenStreetMap contributors, © CARTO. Also credited in the map's own attribution control (bottom right corner of the map).
All 296 authorities
Sortable by any column. Click a row to locate it on the map.
Produced by SDWH Limited for information only, compiled and rejoined from the public sources listed above. SDWH Limited has no relationship with the local authorities shown, accepts no liability for errors, omissions, or decisions made in reliance on this information, and does not warrant its accuracy or completeness. This is a one-off dated snapshot, not a live feed. For authoritative, current figures, always refer to the original statistical releases linked in the sources panel above.