Housing supply statistics in England have retained their official accreditation, but the UK statistics regulator has ordered clearer labelling of quicker housing delivery estimates based on Energy Performance Certificates (EPCs). The Office for Statistics Regulation published its review on 29 October 2026, setting a requirement for the next release expected in November.
The review covers three established datasets: quarterly indicators of new housing supply, annual net additional dwellings and estimates of dwelling stock. The regulator concluded that these continue to satisfy its professional standards. Its concern is narrower, but potentially significant: the newer EPC-derived estimates need to be identified more precisely.
That matters to councils, planning practitioners, housebuilders and anyone tracking the government’s commitment to 1.5 million homes. I would keep one distinction clear: early estimates help fill the gap before final annual figures arrive, but a certificate lodged for a property is not interchangeable with a completed net addition. Different measures can tell different stories about delivery.
What the regulator actually concluded
The Office for Statistics Regulation compliance review did not withdraw accreditation from the three main housing datasets. It found that they still met the standards of trustworthiness, quality and value under the Code of Practice for Statistics. The review was the first of these series since 2018.
However, the regulator said the status and governance of the EPC-based estimates of net additional dwellings were not clear enough in published material. Those estimates are now prominent in housing supply statistics and are widely used because they give an earlier indication of delivery than the annual net additions count. That makes a clear explanation of their standing particularly important.
The report imposed one formal requirement. The Ministry of Housing, Communities and Local Government (MHCLG) must clearly label their status
, explaining whether the new estimates are official statistics, official statistics in development or another form of analysis. The regulator expects that distinction in the next relevant release, currently scheduled for November 2026.
How the quicker housing estimate is calculated
Three commonly cited measures answer different questions. Building control returns describe starts and completions of new construction, the annual net additions series includes conversions and changes of use as well as new builds, and dwelling stock estimates measure how many homes exist. Housing supply statistics become misleading when those definitions disappear from the headline.
The annual net additions figure is particularly comprehensive, but it takes months to assemble. To offer a more timely picture, MHCLG combines the latest annual returns with weekly and quarterly EPC lodgements for new homes and makes an allowance for estimated demolitions. The calculation uses historical ratios, with different adjustments for English regions.
The Department explains that method in its published housing supply indicators. EPCs can provide a useful indication of newly available properties, but the method remains an estimate of net additions, not a direct count of keys handed to owners or homes occupied.
Our earlier report on housebuilding starts and Building Safety Regulator reporting shows another difficulty with comparisons. Changes in the authorities and reporting systems supplying data can move a headline series independently of what has happened on building sites. Context is as important as the number.
Why housing supply statistics need better caveats
The regulator identified a gap in published evaluation of the EPC-based estimates. It wants MHCLG to explain more clearly how the early figures are revised, why revisions occur, whether they tend to move upwards or downwards and whether the pattern differs between regions. A national adjustment will not necessarily describe every local area equally well.
For example, certificates may be lodged in one reporting period while construction milestones fall in another. Subsequent data can change a provisional trajectory when the complete annual statistics become available. The regulator did not calculate or announce a particular national overcount in this review, so it would be wrong to claim that it has found the government’s delivery totals to be false.
Annual dwelling stock estimates bring their own revisions, including changes following Census reconciliation. The regulator said important explanations of uncertainty are too often kept in technical notes rather than brought forward in the main publications. Housing supply statistics are more useful when the reader can see the caveats without hunting through several documents.
What it means for local planning and housing targets
England’s national planning framework uses local dwelling stock estimates as a baseline in its calculation of local housing need. Our guide to local housing need and the standard method explains how that baseline feeds into policy. This connection makes accurate descriptions of the underlying datasets especially important for planning decisions and public debate.
There is an important limit to the finding. The regulator has not said that the standard method itself is unlawful or incorrectly calculated, and its review does not amend a Local Plan, change an adopted housing requirement or overturn an appeal decision. It is a request for better transparency about the statistical measures and their appropriate uses.
A five-year housing land supply assessment is also a different exercise. It relies on evidence of deliverable sites, the applicable planning policies and the relevant assessment period. A quarterly EPC trend may inform the wider picture, but it cannot alone establish whether a council has demonstrated a sufficient supply of deliverable housing land.
What the department must do next
Alongside the single formal requirement, the regulator made four recommendations. They cover a stronger explanation of the methodology and uncertainty, evaluation of revisions, better guidance for readers and broader engagement with those who use the statistics. Those recommendations are not the same as the mandatory status-labelling requirement, and should not be presented as four additional statutory deadlines.
The Office for Statistics Regulation expects MHCLG to correct the labelling in the next publication of EPC-based estimates, presently due in November. It also expects a progress update on the broader recommendations by March 2027. The regulator has not announced any immediate cancellation of accreditation for the established housing series.
The next useful test will be straightforward. Does the release clearly say what kind of estimate it is, how its assumptions differ from the annual outturn and how much uncertainty remains? If not, even a technically sound calculation may be given more weight in public argument than the evidence warrants.
How to read the next housing figures
I would approach housing supply statistics with three questions. Is the number a start, a completion, a net addition, a dwelling stock total or an EPC lodgement? Is it a final annual outturn or a provisional estimate? And have revisions or reporting changes made a simple year-on-year comparison unreliable?
Territory matters too. This regulatory review concerns England’s housing supply and dwelling stock statistics, while EPC releases can cover England and Wales together. Scottish, Welsh and Northern Irish statistics are produced through their own arrangements. Putting figures from different nations or different indicators alongside one another without explaining those differences is an easy way to mislead readers.
The significance of the review is not that the established housing statistics have failed. It is that the regulator has accepted the value of quicker estimates while insisting on clearer descriptions of what they do, and do not, measure. Better-labelled housing supply statistics should support decisions about actual delivery rather than add confusion to them.
0 Comments