Housing

Introduction

This extended post publication signposts readers to the official sources of key housing statistics predominately collated and reported by the Ministry of Housing, Communities and Local Government (MHCLG), with some wider reference to UK-wide statistics that the Office of National Statistics (ONS) reports, as well as to other sources, where appropriate and illuminating, most notably the 2022-23 English Housing Survey (EHS).  

It is a comprehensive recast, extension and update on Making Sense of the Housing Supply Statistics published last year on this website and now includes six sections, each provided with a commentary provided populated with links to relevant tables reporting the source data and to a summary identifying possible policy implications.

Sections One to Four review the four main relevant source housing supply time series published by the MHCLG: the housing supply, net additional dwellings (net new supply), the dwelling stock, the affordable housing, and the often-neglected social lettings series.

These sections focus on the most recently published MHCLG affordable housing (December 2023), social letting statistical series (January 2024), and the dwelling stock estimates series (May 2024).

Section Five covers the MHCLG indicators of new supply series (termed in this post new housebuilding estimates and new indicators of supply and final tenure estimate series for reasons it explains).    

Section Six focuses on house prices, private rentals, and affordability data.

The aim, first, is to provide a user-friendly guide to the availability of such statistics and their most appropriate application and use depending upon the purpose of the enquiry, without being overwhelmed by detail.

Second, to provide structured and evidence-based context for independent policy analysis across the new policy environment that has been ushered-in by the 2024 general election.

Links to the source data mean that readers can always access it to assist user-led interpretation and on-going reference and review.  

They are also provided to interactive dashboards published by the MHCLG within their respective series and to the House of Commons Library that will allow readers to access data (sometimes incomplete) at an individual local authority level with a touch of a mouse

The post includes some new initiatives, including:

1. Table-1A collated by this website from different MHCLG sources provides a time series for the 2011-23 period providing an average per 1000 dwellings new supply figure, by region. This was felt to be important given the wide variance in dwelling supply that exists between local authorities (LAs) often unrelated to local demand or need;  

• To partially relieve the lack of a definitive official data tenure-based net new supply source, at least at a stock rather than flow level, Annex Table Three, itself derived from MHCLG dwelling estimates series table 104, breaks down annual net changes to dwelling stock to sub-tenure, including renting from local authorities;

• Table 3A is designed take account of, and to statistically adjust for, the shift away from Social rent (SR) provision towards government Affordable Rent (AR) to provide more like-for-like decadal supply comparisons;

• Tables 5A and 5B  and accompanying commentary projects 2024-27 new build completion and supply outcomes with reference to the new government’s 1.5m delivery target.     

Some housekeeping points: all data and commentary refer to England unless otherwise stated.

The UK Housing Review provides information at the Great Britain and at an individual country level for Wales and Scotland.

More detailed information for London can be found in the Greater London Authority’s  Housing in London publication.  

The labels (Private) Registered Provider (RP), more commonly known as Housing associations (HAs) are used interchangeably. Section Three provides more definitional information, including the distinction between ‘social rented’ dwellings and dwellings let at Social Rents).

Many table references will include references to the Department of Levelling Up, Housing and Communities when compiled prior to the 2024 general election and the subsequent departmental name change.

Numbers, depending on the nature of the data, have been rounded to nearest thousand or million as appropriate and percentages to the nearest single data when prefixed “about” or to the nearest single figure, percentages normally to one decimal point.

The direction of travel or trend is often more illuminating that a falsely precise figure subject to sampling or other methodological issues, to incomplete and patchy returns subject to later revision. Round figures also sometimes stick in the memory more. 

Decadal averages are employed in the tables as much as possible to illuminate longer term trends and five-year averages for shorter term ones.

Annual and even more so quarterly figures should not be considered in isolation, given their proclivity to funding programme cycles and other contingent factors.                                                                                                   

Comments and observations are welcomed to asocialdemocraticfuture@outlook.com.

1          Housing supply: net additional dwellings (net new supply) series

The net new supply series is presented by the MHCLG as the primary and most comprehensive available new housing supply metric for England. This is for two primary reasons.

First, it includes changes to the stock other than from new build completions, which can be higher or lower than net new supply. The measure thus records the complete flow of net additions or losses that add or subtract from the total dwelling stock.

The large-scale post war slum clearance and redevelopment demolitions that marked the 1961 to the 1980 period meant that net new supply lagged new build completion totals, when, across recent decades, gains to the stock from net conversions and from net changes in use have tended to outpace losses from demolitions, resulting from annual net new supply exceeding new build completion totals.

Second, as the MHCLG technical notes that accompany the series set out, its method of compilation is more comprehensive and accurate.

LAs outside London are expected to record all the changes to the housing stock within their areas over the previous financial year, before inputting them into what is officially termed a standardised housing flows reconciliation (HFR) form for submission no later than the subsequent September.

The Greater London Authority (GLA) collates similar information from the London boroughs.

Each LA, therefore, has up to five months from the end of the financial year to submit their annual HFR return. This compares to the much shorter six-week submission period connected with the building control based quarterly new housebuilding series, described in Section Five.  

By having more time to reconcile diverse potential sources of completed new build activity, including from council tax, building control and other records, as well as site visits, LAs can consequently compile a more complete and comprehensive count of additions to their local stocks.

Each November successive to the preceding financial year, the MHCLG publishes a housing supply (net additional dwellings) statistical release and an updated set of accompanying live tables that together describe changes to the net supply position.

The latest such release published in November 2023 reported net supply change between 1 April 2022 and 31 March 2023.

It also utilised 2021 census data to rebase the total annual net supply figures to the change in the total dwelling stock by enumerated by the 2011 and 2021 censuses.

58,860 dwellings additional to the aggregated net supply figures previously reported by the series, as identified in the 2021 census, were added to the stock.

This was done by making a constant 5,880 to the annual 2012-21 net additions (supply) figures, much smaller than the annual 16,190 census adjustment uplift made across the previous 2002-11 period.

The latest 2021 census dwelling count now provides the baseline for succeeding net supply and dwelling stock estimates (see next section) until the next adjustment is made to both series to reflect the future 2031 census total dwelling count.

The components of total net supply, including new build completions, however, are not census-adjusted and remain based on historic annual net supply LA statistical returns (see Annex Table One).

Figures reported for 2022-23 figures are provisional, subject to revisions to be reported in the next November 2024 release.

Revised data from 27 LAs resulted in an increase of 1,646 net additional dwellings recorded for 2021-22.

Live Table 118 of this net supply series reports total annual net additions, by region, since 2000-1, taking account of successive census decadal adjustments, according to its components since April 2006

Table 120  also breaks down the different annual components of net supply since April 2006, including:

• new build completions;
• conversions from houses to flats;
• net changes to the dwelling stock resulting from gains or losses,
• from 2015-16 onwards, changes in use resulting from permitted development rights have been separately catalogued
• any retrospective census adjustment, when applicable and known.

Demolitions are then subtracted to provide the annual total net new additions (new supply) figure.

The immediate short story conveyed by Annex Table One is that annual net additions (new supply) steadily rose from 146,700 in 2001-02 to 223,500 in 2007-2008.

The impact of the Great or Global Financial Recession (GFC) then drove a collapse to an ultimate new supply floor of 130,600 in 2012-13.

Subsequently annual new supply steadily recovered to reach a peak of 248,600 in 2019-20.

Although it has subsequently dropped back, during the most recent 2018-23 period, annual new supply averaged 236,600 dwellings during that period, the highest five-year average recorded since the series began.

It should, however, be borne in mind that in England, the population grew by almost 3.5 million (6.6%) between the 2011 and 2021 censuses to 56,489,800 people, increasing the pressure on the housing stock.

212,600 new builds were completed in 2022-23, accounting for about 90% of the 234,400 total net supply increase reported for that latest year.

Net change in use attributable to permitted development rights (PDRs, which allow change of planning use without formal planning permission) accounted for 18,900 of the 37,900 total change in use total in 2016-17.

In the latest 2022-23 year such PDRs accounted for 9,500 (7,900 reflecting changes in office to residential use) of the total 2022-23 net change to the residential use 22,800 dwelling total (see columns C and D, Annex Table One).

Demolitions, which are subtracted from the net additions total, have fallen progressively and continuously from 22,300 in 2006-2007 to 5,500 in 2022-23.

Table 1  confirms that about 90% of new supply during the last two 2021-23 years are accounted for by new builds, compared to about 82% in 2016-17, when numbers in net changes in use, including those resulting from the newly liberalised permitted development rights peaked, giving rise to concern about the size and quality of some of the additional residential dwellings provided through that route.

Even though new build completions and total new supply may have reached a peak in 2018-20, the table shows that they still fell well short of the prevailing 300,000 new supply target.

It also demonstrates the proneness of net new supply and its components to fluctuate in a lagged response to wider macro-economic and housing market conditions.

Net supply in 2012-13 (reflecting the collapse of private speculative activity during and in the wake of the 2008-10 global financial recession) was barely half of the recovered 2018-20 level.

MHCLG Tables 122 and 123 reports total net additional supply by LA district and their component flows, respectively, since 2001-2.

MHCLG Table 124 reports annual changes in the flows of communal non-self-contained accommodation by LA district.

These changes are separate to and are not included in the other tables or the net supply figures, which only records self-contained accommodation in accord with the census definition.

The House of Commons library has produced a very useful online housing supply dashboard in which users at a touch of their mouse can access and print net new supply data for individual LAs covering the most recent ten year reported, or as  users require.

Alternatively, the MHCLG’s own interactive dashboard can be consulted.

Although it is less print-friendly, it does report net new additions (supply) broken down into its components at a local LA level, as well as at regional and national level, on a per 1,000 dwellings basis, for individual years.

It does not, however, allow users to derive average per 1,000 dwelling figures over an extended period, in contrast to year-to-year comparisons.

These can mislead due to the frequent lumpiness of new supply at the local level, connected, for instance, to a completion of a large development in a particular year.

Examples taken from two regions, the East Midlands and London, demonstrate that clearly.

The MHCLG net supply interactive dashboard reports that net new supply per 1,000 dwellings across the Nottingham City Council area in 2022-23 was 14.3 (reported here to nearest decimal point): much higher than the 4.9 reported for 2020-21 and nearly six times its 2011-21 decadal average figure of 2.4.

Net new supply per 1,000 dwellings across the East Lindsey council area (second largest by area for England; mainly rural but including market and coastal towns, such as Louth and Skegness) in 2022-23 was six, nearly than five times less than its 2011-21 average of 30.2.

In London, Barking and Dagenham, net new supply per 1,000 dwellings in 2022-23 was 15.9 compared to 4.7 as recently as 2017-18.

Table-1A collated by this website from different MHCLG sources provides a time series for the 2011-23 period providing an average per 1000 dwellings new supply figure, by region.

It reports the 2011-21 decadal average and the most recent half decadal 2018-23 average, as well as individual years 2011-23, all by region.

The data shows both some intertemporal (between individual years) and spatial (between regions) variation, although less than one would expect if new supply performance was more closely aligned to relative demand.

Taking the half decadal 2018-23 average as the benchmark, London reported the highest 10.9 average per 1,000 average and the North East the lowest, 7.1, within an overall England average of 9.6.  

Such region-wide averages mask much larger differences occurring within them.

Whilst the incompleteness of data coverage reported at that local level should be borne in mind, the range of performance that can be displayed between even neighbouring authorities is striking, noting that due to the often lumpiness of new supply at a local level, decadal or half decadal rather than individual year figures should be considered, as was explained above.

The 2011-21 decadal average for East Lindsey, for example, was 30.2, when for North Kesteven it was 10.3. In Lincoln, it was 5.6.  

In London, the 2011-21 decadal average for Richmond (in London) was 3.6 while in neighbouring Hounslow it was 10.4.

The decadal average in Tower Hamlets was 22.7 compared to 13.4 across the river in Southwark.

There is no necessary relationship between such outcomes and local housing need and demand.

Summary

Since April 2006 the net supply series identifies and measures the different components of net new supply, including new build completions.

It consequently provides the primary and most comprehensive and accurate available statistical record of the different components of new supply at both national and LA levels, consistent with the census definition of a self-contained dwelling.

The series, however, is only published annually in November for the preceding financial year; given that and because it does not cover dwelling starts, it is not forward-looking.

Accordingly, to obtain a more timely but imperfect indicator of new build completions for the year preceding November publication of this series and of future activity, the demonstratable inaccurate new housebuilding series must be consulted (see Section Five, new housebuilding and new indicators of supply section for a full discussion).

The net supply series also does not break down the reported net supply total by tenure, a significant omission insofar that gross reported new affordable supply greatly exceeds net new affordable supply – as the affordable housing series sub-section in Section Three will later show and discuss.

However, the next and successive sections will also identify sources that can be used to overcome at least partially that and related shortcomings.   

2          The dwelling stock series

The dwelling stock series reports total stock estimates back to 1801, when the first census took place, and, from 1961 onwards, also its tenure breakdown.

This series is published each year in May to include data for the past but one 1 April to 31 March period: the May 2024 statistical release (see below) and accompanying tables accordingly reports dwelling stock data for 2022-23.

Until April 2000, the annual dwelling stock estimates provided the source of the annual net additions figure. That, hitherto, had not been separately produced as a standalone series.

From 2000-01 onwards, the annual total dwelling stock estimate has been equal to the previous year’s dwelling stock estimate plus the latest annual net new supply of housing figure.

The total dwelling estimate is broken down (disaggregated) into owner-occupied and private rented sector tenures and into ‘rented from’ local authority (LA) housing and housing association (HA) sub-tenures (together, social housing).

This is done using periodic survey data mainly for the owner occupied and privately rented sector split, including the English Housing Survey (EHS), Labour Force Survey (LFS), and the Local Authority Housing Statistics (LAHS)  and the Statistical Data Return (Regulator of Social Housing) returns for the social housing sub-tenure split.

The private housing stock is not split into owner-occupied and private rented sector at the local authority district level, however.

Dwellings are defined in line with the census definition: a self-contained unit of accommodation where all the rooms (including kitchen, bathroom, and toilet) in a household’s accommodation are behind a single door that only that household can use.

A dwelling therefore can consist of one self-contained household space or two or more non-self-contained household spaces at the same address.

Ancillary dwellings (for example, ‘granny annexes’) are included provided they are self-contained, pay separate council tax from the main residence, do not share access with the main residence (for example, through a shared hallway) and that there are no conditional restrictions on occupancy.

Communal establishments providing managed residential accommodation, are not counted. These include university and college student, hospital staff accommodation, hostels/homes, hotels/holiday complexes, defence establishments (not married quarters) and prisons.

But purpose-built (separate) homes (for example, self-contained flats clustered into units with four to six bedrooms for students) are included in the dwelling estimates, with each such self-contained unit counted as a dwelling.

Non-permanent (or ‘temporary’) dwellings are included if they are the occupant’s main residence and when council tax is payable on them as a main residence.

Such dwellings can include caravans, mobile homes, converted railway carriages and houseboats.

Permanent gypsy and traveller pitches are also be counted as dwellings, if they are, or likely to become, the occupants’ main residence.

The series technical notes provide more background information.

Shared ownership dwellings are currently counted as owner occupied within the dwelling stock series.

The English Housing Survey (EHS) estimated back in 2019–2020 that there were around 202,000 households living in shared ownership properties in England (could be closer to 250,000 in 2022-23), representing approximately one per cent of homeowners and less than one per cent of all households.

The dwelling stock estimates series also does not split PRS properties into sub-tenures, including to buy-to-let (BTL) and build to rent (BTR), nor does the EHS for households, although both sub-tenures are important segments of the housing system.

Table 56 of the UK Housing Review reported that in 2022 there were over two million BTL mortgages outstanding in the UK compared to 836,000 in 2006 and less than 30,000 in 1998, suggesting that is now home to around two million households and to around 45% of private renters: a momentous economic and social change that deserves some attention within the official statistics and wider analysis.

The BTR sub tenure comprises homes built for private market rental. Industry sources suggest that there are around 100,000 BTR dwellings, most of which are concentrated in London and other metropolitan centres.  

Dwellings classified as ‘other public’ include dwellings owned by government departments, such as Ministry of Justice and of Defence. These can be vacant awaiting sale or redevelopment.

As Section 1 explained, both the dwelling and net new supply series are subject to a decadal retrospective and a new baseline census adjustment, usually applied in the year following each successive census.

In that light, the technical notes (scheduled revisions) of the May 2022 statistical release reporting for the year and period ending 31 March 2021 advised that a level 5,880 dwelling upward adjustment to the previous reported annual dwelling estimates for the decadal 2011 to 2021 period had been made, reflecting the 58,880 self-contained dwellings enumerated by the 2021 census that were additional to (on top of) what the new supply series had had previously reported for that decadal period, thus mirroring the adjustment made to that series.

In short, the adjustment ensured that the total dwelling stock estimate for 31 March 2021 equated to the 2021 census total dwelling count.

The end March 2023 figures were reported in the May 2024 statistical release.

This reported that there were 25.4m dwellings in England, as of 31 March 2023, an increase of 234,400 dwellings or 0.93% on the previous year (an increase corresponding to that reported in November 2023 net new supply (additions) series, (see Annex Table One).

Other headline results (reported to nearest 100,000; for example, LA stock figures reported as 1.6 million below were reported as 1,571,000, in Table 100) were:

• 16.3m dwellings were owner-occupied dwellings, an increase of 176,000 dwellings on the previous year (64.1% of total dwellings compared to its 69.5% peak in 2001);
• 4.9m dwellings were private rented dwellings (19.4% compared to its recent peak of 20.3% in 2016) an increase of 35,000 dwellings on the previous year;
• 4.1m dwellings (16.3%) were social and affordable rented dwellings (combined Private Registered Providers (RP) and Local Authorities figure), an increase of 25,000 dwellings on the previous year, of which 2.6m. were rented from Private Registered Providers (also known as Housing Associations) and 1.6m. dwellings were rented from Local Authorities.  
• 31,000 dwellings (0.1% of total) were other public sector dwellings, a decrease of 2,000 dwellings on the previous year;
• 699,126 vacant dwellings (2.8% of total) were recorded in England on 2 October 2023, an increase of 22,822 or 3.4% from 676,304 on 3 October 2022;
• 261,474 dwellings were long-term vacant dwellings in England on 2 October 2023 (one per cent of total), an increase of 13,325 or 5.4% from 248,149 on 3 October 2022.
• England had a dwelling density of 1.95 dwellings per hectare as of 31 March 2023

All figures are rounded (as above), are provisional and are subject to revision. They are estimates rather than precisely accurate, as are most official statistics.

The latest English Housing Survey (EHS) provides information on tenure at the household rather than dwelling level, but, to all intents and purposes, mirror the dwelling estimates.

Owner occupation remained the largest tenure group in England (65% of households) in 2022-23, the social rented sector the smallest (16%) including 10% or 2.5 million households rented from housing associations, and 6% or 1.5 million households from local authorities, while the private rented sector has hovered around a similar proportion (19%) since 2013-14.

Unsurprisingly there were regional variations:

• 31% of households in London were private renters, compared to 17% in the rest of England;
•  21% of London households were social renters, compared to 16% in the rest of England, with the 10% proportion renting from a local authority much higher than across England generally (6% as above);
• 49% of London households were owner occupiers compared to the 68% who were across the rest of England. 

MHCLG live Table 100 reports, from April 2009 onwards, the total dwelling estimates and their tenure breakdown down to LA level (grouped according to type of authority, not to region).

Owner occupied and private rented sectors reported at the LA level are subsumed into one private sector category. At the national level LA and PR sub-tenures are separately reported.

MHCLG Table 104 provides a historical annual dwelling stock series, dating back to 1801, with a tenure breakdown first provided in the census year 1961 and then annually from 1969 onwards.

The short story it tells is that in 1961 the LA and the private rented (PR) stock when combined exceeded the then c6.1m dwelling owner-occupied stock.

The LA stock continued to climb to its 5.2m peak in 1980, after which it plummeted to its current trough level of less than 1.6m. compared to about 2.6m RP owned dwellings, when, until 2008 there were more LA-owned than RP dwellings.

The PRS stock slid from 4.4m in 1961 to 1.6m in 1986, before subsequently recovering robustly to its current level of about 4.9m.

The owner-occupied dwelling stock progressively increased in total numbers throughout the period to its current c.16.3m dwellings, notwithstanding that its percentage tenure share fell from its recorded high of nearly 70% (69.5%) in 2001-2002 (EHS figures reported nearly 71%) to 64.5% in 2010-11, where it has broadly remained since.

Annex Table Two derived from MHCLG live Table 104 reports the total stock dwelling estimates, for each year since April 1991 up to 31 March 2023, by tenure and sub tenure.

MHCLG Table 109 provides a summary breakdown for the same post April 1991 period, by tenure and region.

Table 2 provides summary data for dwelling stock, by tenure and sub-tenure, as percentage of total for the census decadal years 2001, 2011, 2021, and then for 2023.

It reports that the owner-occupied tenure share fell from nearly 70% in 2011 to about 64% in 2021, while the percentage of the total dwelling stock taken by the private rented sector (PRS) nearly doubled as a percentage between 31 March 2001 from about 10% about 19% in 2023, with the share taken by LA rented more than halving from about 13% to about six per cent during the same period, with the RP tenure share rising to about 10%.

Since 2001, the combined LA and HA total (labelled social and affordable rented in the table) fell from about 20% to about 16%.

The EHS reports that since 2013-14 outright owners have outnumbered mortgagors 35% of households reported in 2022-23 as owning outright compared to 29% as mortgagors in 2022-23, a product of an aging population leading to larger numbers of people paying off their mortgages and becoming outright owners, as well as probably of an increased unquantified number of cash purchasers.  

Table 2A reproduces Table 3 of the MHCLG 2024 dwelling stock statistical release, which presented and reported an increase in the combined LA and RP tenure stock (labelled as ‘social and affordable rented’) of c25,000 dwellings in 2022-23.

Since April 2009, according to that official table, their combined total has increased on average annually by about 14,000 dwellings (disregarding dwellings categorised as ‘other public sector’).

That table, however, made no split between Social Rented and Affordable Rent’ sub-tenures let up to 80% of market levels – a significant omission insofar that many commentators consider the affordable housing sub-tenure not to be ‘genuinely affordable’.

Annex Table Three, reports the annual change by numbers in the estimated dwelling stock for each year since April 1991 up to 31 March 2023, by tenure and sub tenure, withLA and RP figures split.

It reports a continuing decline in the LA rented stock, which since April 1991 has fallen by more than 2.3m dwellings, compared to an increase of around two million in the HA stock.

The primary driver for that outcome is the impact of Right-to-Buy (RTB), introduced as a new flagship social policy by Mrs Thatcher in 1979 and statutorily enacted in the 1980 Housing Act.

MHCLG Table 678 reports that during the 40-year period from the introduction of the Right-to-Buy (RTB), starting in 1981 and ending 2022, more than two million dwellings were sold to LA and RP tenants exercising their RTB, c93% of which were to sitting LA tenants.

Such annual combined RTB sales peaked at c84,000 in 2003-04, before collapsing to c.3,100 (mainly because of changes in discount and eligibility arrangements) in 2009-10, before rising to a more stable average of 13,200 dwellings across the 2013-22 period. 11,303 dwellings were sold to sitting tenants in 2022-23.

The fluctuating RTB stock losses, summarised in Table 2B, contributed to the results reported inAnnex Table Three.

These tables, when taken together, show that the LA lost nearly 100,000 dwellings annually during the eighties, reducing to 45,000, on average, during the nineties, and to 35,000 in the noughties, and further still to about 10,000 annually during 2010-20 and the most recent 2018-23 half decade.

Although the outflow from RTB has decelerated, since 1991 the LA net stock has still lost a net c2.3m dwellings.

During 2022-23, there were 11,200 RTBs in the LA sector, 4,600 RP sales to sitting tenants, a total of 15,800 sales to social housing (LA+RP) sitting tenants.

Summary and policy implications

The dwelling stock series is a core series calibrated to the past decadal census then to the new supply series until the next census.

It conveys more information than is generally recognised on tenure composition trends, derived from an additional range of sources, including the English Housing Survey.  

Annex Table Three, itself derived from series table 104, in contrast to Table 2A, breaksdown net changes to dwelling stock to sub-tenure, including renting from local authorities, confirming that it lost about a further 5,000 dwellings in 2022-23, despite an uptick in new builds and, by the standards of 1981-2006 at least, a relatively modest level of about 11,000 RTBs, and provides a good indicator, albeit measured at a lagged stock rather than flow level, of net changes in tenure composition.

The new Labour government is expected to trim RTB discounts but given its fiscal rules will forbid increases in net current expenditure that add to debt (in case, say, a revenue contribution to new council provision, financed by RTB proceeds), it will need to navigate a trade-off between future RTB volumes and receipts generated that could be recycled into new investment.  

That 35% of households reported in the 2022-23 EHS as owning outright compared to 29% as mortgagors in 2022-23, clearly has implications for economic and social policy beyond housing, insofar that it can be expected to reduce the effectiveness of interest rate changes on house prices and broader macro-economic outcomes, as well entrench generational inequalities.

More granular and assessable information of sub-tenures, including Build to Rent (BTR), and Buy to Let (BTL) would be helpful.

Quite clearly, however, the most pressing and important policy issue or implication is the growing concentration of poorer, younger people into the PRS, often poorer quality stock in locations offering limited economic opportunities.  

3          Affordable housing series

Introduction and definitions

The affordable housing series reports the gross annual new supply of additional affordable dwellings for rent or sale that is provided for specified eligible households whose needs are not met by the market.

Dwellings classified and recorded as affordable housing, according to the official definition, encompass a range of sub-tenures. These are defined briefly below.

Social Rent (SR) dwelling rents are set according to national guidelines involving the calculation of a ‘formula rent’ for each property based on its relative local market value, on relative local income levels, and on its size.

This is done with reference to a stated aim to ensure that “similar rents are charged for similar properties” that is not always realised in practice.

Dwellings let in accordance with such arrangements are termed Social Rent (with initial block capitals) dwellings in this post.

The MHCLG in its January 2024 MHCLG Social housing-lettings April-2022 to March-2023 (Tenancies) statistical release advised that the national median SR in 2023 was 47% of market rental levels, although that proportion can vary within and between areas and regions.

Another sub-tenure akin to SR, as explained here, is London Affordable Rent, calibrated to 2016 Social Rent levels, save that the annual rental increase allowed for SR lets was back then capped until April 2020 at Consumer Price Index (CPI) less one per cent (CPI-one per cent), but LAR rents could be updated annually by CPI inflation plus one per cent.

Since April 2020 CPI + one per cent been the base ‘social rent’ cap set for all SR and AR tenancies let by LA’s and RPs, as set out in this 2020 government statement, save where subject to an overarching ‘cost of living’ cap.

In this post, data on LAR delivery is subsumed within the SR category.

Affordable Rented (AR) housing was introduced in 2011 by the incoming Coalition government to be let by providers of social housing to eligible households at a rent of no more than 80% of the local market rent (including service charges, where applicable).

According to the above same MHCLG release, it was let at a median 71% of market rent levels in 2023-23.

Intermediate Rent (IR), according to official definitions, should not exceed 80% of the current market rate – although some exceptions to that rule may be included and recorded in the affordable series data.

SR + AR + IR is officially defined as ‘social rented’ or ‘low-cost rental’ housing (no capitals), and thus is wider in sub-tenure scope than the Social Rent (with capitals) sub-tenure, as defined above.

That can prove confusing, especially when SR and AR are not specifically differentiated in the data.

Accordingly, for the purposes of this post, what is officially defined as ‘social rented’, is generally described as Social Rent (SR) and Affordable Rent (AR) and Intermediate Rent (IR) combined, or as is applicable.  

Intermediate affordable housing are homes for sale and rent (not SR or AR) provided at a user cost above Social Rent levels but below market levels, subject to it remaining at an affordable price for future eligible households and/or to the recycling of expended subsidy into the support of alternative affordable housing provision.

Such housing can include shared ownership and other tenure forms where some equity is retained by the provider, as well as other low-cost homes provided either for sale or for intermediate rent.

In 2021-22, First Homes was created as a sub-tenure form intended to eventually account for at least 25% of all affordable housing units delivered through planning obligations but was embraced by LA’s generally with reluctance.

Its future following the 2024 General Election result is uncertain and it could possibly be replaced with an alternative government-sponsored home ownership product.  

The new Communities and Housing Secretary in her July 2024 announcement confirmed that LA’s would not be required to provide a prescribed proportion of affordable home ownership housing and emphasised a refocus on SR provision.

Dwellings purchased under the post-2013 Help-to-Buy programmes are not recorded as affordable housing, on the ground that purchaser access was not subject to an income qualification.

Social landlords of what is officially defined as affordable housing can be a local authority (LA) – letting dwellings generally known as council housing – or be a private registered provider (RP) registered with the Regulator of Social Housing (RSH).

Historically, only non-profit-making organisations, previously known as housing associations, and still sometimes referred to as such given that most private registered providers are HAs, (this post will also use the terms HA and RP interchangeably), could be registered as providers of social housing or as social housing landlords.

Since April 2010 profit-making organisations also have been able to register with the RSH.

RPs can now include:

• organisations providing supported housing and care;
• local authority subsidiary companies;
• community groups seeking to develop new housing;
• commercial developers setting up small subsidiaries to receive Section 106 affordable housing;
• subsidiaries of investment companies and funds;
• entities established by registered provider groups, either new parents for their group structures or new subsidiaries;
• small charities, such as alms houses.

Dwellings provided by non-registered providers should not be counted as affordable housing. However, in practice, they sometimes are (they are, for instance, separately defined in MHCLG Table 1013, see below).

Affordable housing can be newly built, acquired, or result from a net gain secured through conversion or from a change in use.

The affordable housing seriesreport Gross totals that take no account of demolitions (representing a direct dwelling loss) or of sales of existing affordable dwellings overwhelmingly into owner occupation.

The series, therefore, does not report changes in annual net affordable supply: the net flow of stock gains minus losses across each sub-tenure – which can be positive or negative – nor its availability, as indicated by new lets.

Rather, it measures new affordable provision activity (predominantly of new build, but also including acquisitions and net gains or losses from conversions and from changes in use).

Such gross new provision of sub tenures therefore influences, but does not necessarily determine, either their net stock position – which Annex Table Three reported in the preceding section acting as a p[atrial corrective – or their letting supply availability, as Section Four will shortly show.

Affordable housing series data

Affordable starts and completions for the preceding financial year (2022-23) are published in the succeeding late November/early December (2023) alongside updated live tables, which are also subject to scheduled revisions each June/July.

MHCLG Table 1000 reports the most up-to-date summary estimates of affordable housing provided for the period since the series started in 1991-92 (primarily new build completions, but also acquisitions and net additions resulting from conversion or from change in use).

Henceforth affordable housing series terms ‘completions’ and ‘provided’ are used interchangeably, unless otherwise stated

MHCLG Table 1000C additionally breaks down that summary data to sub-tenure by type of scheme and funding. Table 1000S reports also affordable starts on a similar basis but – as across the series in cases where starts are reported – from 2015-16 onwards only.

MHCLG Tables 1008C and 1008S break down the total affordable completion and start information to region and LA district level, while Tables 1006 and 1007 do likewise for different defined sub-tenure types, reporting both starts and completions.

MHCLG Table 1009 breaks the total affordable completion data down, according to whether the provided dwellings were newly built or acquired, by sub-tenure.

MHCLG Tables 1011C and 1011S report from 1991-92 onwards, affordable total completions and starts (from 2015-16), according to type and source of funding (including nil grant S106).

A dropdown box at the top of these tables allows users to access data at region and LA level.

MHCLG Table 1012 reports affordable housing starts and completions, funded by Homes England, combined with data with the GLA from April 2012, both for starts and completions, between 2009-10 to 2023-24.

The latest Homes England  and GLA dedicated affordable supply data for 2023-24 can be accessed via the aforementioned links.

Homes England and GLA are focused on the delivery of dwellings subject to their programme funding support.

The MHCLG affordable housing series has a wider remit to provide a complete picture on affordable housing delivered, irrespective of funding mechanism or its source, using more disparate sources, most notably LA annual housing statistical (LAHS) returns.

LAs are asked to only record affordable housing that has not been reported by Homes England or the GLA.

This should predominately be affordable housing that did not receive grant funding or developer contributions under planning agreements.

However, Homes England has confirmed that some nil grant S106s are indeed included in its data returns, and it is possible that some LAs double count by including them also in their LAHS returns

As their 2023-24 start and completion data are published in the following June, rather than in December, as is the case with the MHCLG series, these  two Homes England and GLA are utilised in a summary table in the scan forwards sub-section that follows.

New MHCLG Table 1013 breaks down affordable provided completions, by type of provider, whether LA, or registered, or non-registered providers, or unknown.

It advised that RPs provided 49,844 of all the 63,605 affordable dwellings recorded for 2022-23, compared to LAs that provided 8,906, with non-registered providers accounting for 2,273. Another 2,582 dwellings were classified as of unknown sub-tenure.

Of the LA provided affordable dwellings, 60% were let at Social Rent or London Affordable Rent (LAR) levels (3,926 SR plus 1,419 LAR: 5,345 in total): the highest figure since the series started in 1991-92, but still only 8.4% of the total affordable supply provided in 2022-23.

Annex Table Four summarises affordable completion data across the entire 1991-92 to 2022-23 series period, taken from MHCLG Tables 1009 and 1011C, broken down to sub-tenure.

The 63,600 new affordable homes provided in England in 2022-23 is the highest reported since 2014-15 and a seven per cent increase on 2021-22.

Gross annual total or all affordable provision across the most recent April 2018-23 period averaged about 58,000 dwellings, representing the highest level recorded since April 2007-11.

It coincided with the end of the 2016-23 Affordable Homes Programme, as did the earlier 2014-25 uptick peak with the end of the 2011-15 Affordable Homes Programme: funding programme approval and delivery profiles result in years where starts and completions are bunched.

Although the latest peak was considerably above the April 1999-2003 low point average of 33,500, it still fell short of the 68,200 annual average dwellings provided during the April 1992-96 entire series peak point period.

Although Table Annex Four appears to report relatively stable total affordable provision totals on a decadal average basis, in practice they could fluctuate quite sharply within each decade in response to programme profile issues, to funding availability, and to wider economic conditions.

Total new build affordable completions increased progressively decade to decade in total and as a share of all affordable completions as acquisition and conversion provision decreased (subject to the compositional issues discussed below).

The table also reports that since 2001, annual affordable new supply has accounted for a stable close to 27% share of total net new housing supply as reported in MHCLG Table 120 (see Annex Table One).  

But as Annex Table Four also shows, that share increased sharply in the wake of economic downturns (when private speculative activity contracts).

This it did across the post-GFC 2010-13 period when it rose above 40%, as affordable dwellings funded by a relatively previous generous affordable programme settlement were completed during a period when private completions were at record low levels as a lagged impact of the GFC.

The share tends to decrease, albeit less sharply, during upswing periods marked by private sector recovery. This it did during 2002-06 and 2015-17 periods to a share of 24% or below.  

Taking account of the compositional mix

The sub tenure composition of provided affordable housing has changed markedly since the mid-nineties.

As Annex Table Four, summarised in Table 3, shows, the predominance of dwellings let at Social Rent (SR) levels started to decline from 1997-98 onwards.

56,900 Social Rent dwellings were provided in 1996-97 – greater than two and half times of the 21,700 SR completions reported in 2004-2005, and four times the 13,900 dwellings (including LAR) provided in 2022-23 – the most since 2012-13.

While the sub tenure still provided 82% of all affordable dwellings in 1999-2000, its share had declined to 57% by 2009-10.

This was  a combined product of falling SR volumes and the rising intermediate sub tenure completions that accounted for the remaining share of the programme, to about 42%.

The share taken by SR then collapsed to 14% in 2014-15, although most affordable dwellings were let at Social Rents until 2012-13.

This was the result of the introduction by the 2010-15 coalition government of the Affordable Rent (AF) sub tenure and the associated diversion of public grant funding in support of AR provision, which that year provided 62% of total affordable dwellings.

Intermediate tenures accounted for the remaining 24%. Their volume had decreased from 24,800 in 2009-10 to 15,800 in 2014-15 before dipping further to 9,300 in 2015-16.

Since 2014-15, shared ownership (part owned, part rented) homes have provided about a third of all affordable housing supply, becoming the predominant affordable home ownership sub-tenure.

Overall, all affordable home ownership sub tenure types since 2018-19 have accounted for the same 35% to 38% proportion of all affordable dwellings that it did twenty years ago.

During the latest 2018-23 period nearly three quarters (74%) of the average annual 36,100 combined social and affordable rent completions total were let at higher Affordable Rental levels, with SR providing an average 9,700 dwellings compared to 26,400 AR completions. Intermediate tenures provided an average 16,400 dwellings annually.

SR volume recently increased in 2022-23 to nearly 14,000 dwellings, enlarging its sub tenure share (including LAR) to 22%. By then AR’s share of affordable completions had fallen slowly but progressively to 38%, totalling 24,300 dwellings.

These compositional changes in the affordable housing programme concerning the volume and tenure share shift away from SR in favour of AF, as was identified above – as well as possibly other potential ones not reported or highlighted here, including changes in the bedroom composition of the provided dwellings and/or their location – provide a salutary example of the care needed to interpret time-series data on a consistent like-for-like basis.

Here, to take account of, and to statistically adjust for the shift away from SR provision towards government Affordable Rent (AR) housing provision, Table 3A  has applied the following weightings: Social Rent = 1; Government Affordable Rent = 0.625; Other Intermediate = 0.5.

When these are applied, the 2011-21 decadal average annual completion figure compared to that of previous 2001-11 period, total affordable completions transpose from around nine per cent higher to around nine per cent lower.

The most recent 2018-23 annual average affordable completion figure is also greatly reduced from about 27% higher than the 2001-11 average to about one per cent higher.  

Tweaks to the weightings applied would vary that outcome.

Table 3A also shows that, disregarding programme and other cyclical variations – and after adjusting for the compositional effects identified above – gross affordable supply when measured on a cross-decadal average basis has proved broadly stable in total volume terms.

Total affordable completions on an average annual decadal basis, weighted as above, reduced from 36,900 during 2001-11 to 33,700 during the following decade. The most recent 2018-23, half decadal average was slightly higher at 37,300.

Such apparent relative long-term stability of affordable housing supply, of course, is a different issue as to whether gross affordable supply in volume and compositional terms was – and is – sufficient and/or well-targeted relative to prevailing and changing socio-economic circumstances and needs.

These can include demographics, wider housing supply and its composition, and its affordability relative to individual household income and other circumstances rather than official definition.

And, as will be shown in the following Section Four, new gross affordable supply crucially also does not necessarily equate with new housing letting opportunities.

The paramountcy of S106 nil grant to new affordable supply

Annex Table Five uses MHCLG Table 1011C to estimate the number and proportion of all completed affordable dwellings using Section 106 (S106) affordable housing obligations to deliver affordable housing without the use of public grant both in total and by sub tenure.

Such affordable dwellings are generally secured through the planning system through developer cross subsidies realised from sales of dwellings sold at market values.

This table identifies a general trend since the nineties – save for some trend interruption in the wake of the Global Financial Crisis (GFC) – for the proportion of affordable housing across tenures provided through S106 to progressively increase

This is to a point where S106 nil grant has become the primary reported funding mechanism of affordable housing.

They exceeded 50% of total affordable completions in 2019-20, and 47% of them in the latest reported 2022-23 year.

The story of how S106 became the primary funding mechanism of affordable housing is recounted in Section 1 of  The new infrastructure levy: going-round the mulberry bush.

In terms of sub tenure composition, nil grant S106, in general, delivers a greater proportion of intermediate tenure properties – about 53% in 2022-23 – than Social Rent (SR), where it delivered about 38%.

As ever, averages can mask variation. In London, only four of the 1,567 dwellings provided through conventional Social Rent (excluding London Affordable Rent) was provided though nil grant S106, although S106 nil grant did account for 39% of the London Affordable Rent dwellings provided.

Scanning forwards

Affordable starts for the April 2018-23 period by sub-tenure (when known) reported in MHCLG Table 1011S, provide an indication of future short-term gross affordable completion levels.

About 71,800 affordable starts were recorded in 2022-23. Annual average starts recorded during the 2018-23 period was 61,000 dwellings.

Unfortunately, it is likely that this figure over the 2023-25 period will fall back to below 50,000 dwellings.

The most recent data on affordable starts and completed funded by Homes England and the GLA, summarised in Table 3B, shows that in total, starts fell to about 31,000  in 2023-24 from 55,000 dwellings in 2022-23. In London they plummeted even more dramatically by about 90% to about 2,300.

The precise reasons for this sudden and apparently calamitous drop are not totally clear or, at least, accepted by competing stakeholders, but programme profile and approval issues, market conditions and costs, and the impact of post Grenfell and other regulatory requirements all seem to have played a part.

Public funded start data reported as funded by Homes England and the GLA can be expected to fall short of what will be reported in the more comprehensive MHCLG Table 1011S in December, for reasons explained earlier.

The 55,100 affordable starts reported by Homes England and the GLA for 2022-23 accounted for about three quarters of the 71,800 total later reported in Table 1011S.

Conducting a rough rule of thumb exercise, replicating that relationship would suggest about 42,000 starts in 2023-24. The December reported figure could, of course, be lower or higher than that, but the omens overall are to the downside.

Such a level in practice would require affordable starts to increase to about 58,000 in 2024-25 to get to an 2023-25 annual average of 50,000 dwellings, and a similar completions level projected forward into 2026-28.

Summary and possible policy implications

This section has reported a feast of data on affordable housing produced by MHCLG. So much, in fact, that it is easy to be overwhelmed by detail, missing the wood for the trees.

It is therefore probably sensible to summarise the main points or contours of the story.

First, the series does not purport to report changes in annual net affordable supply: the net flow of stock gains (predominately now new build provision, but also acquisitions, conversions etc) minus losses mainly from RTB and other sales and to a lessening extent demolitions) across each affordable sub-tenure.

That net supply flow can be either positive or negative (see Annex Table Three).

Nor does it measure affordable housing availability, as indicated by new lets (see Section Four and its accompanying tables).

Second, broadly speaking, contrary to the political heat and noise and spin often put on comparisons between periods generally bookended by changes in government and/or cherry picked to outlier peak or trough years, over successive medium term 1991-21 decadal periods total average affordable housing completions remained relatively stable within a 46,000 and 52,000 range.

Shorter-term gyrations linked to affordable housing programme approval and funding profiles, as well as to wider public expenditure funding choices occurring over comprehensive spending review cycles (see Annex Table Four) also feature.   

Example of the former was 2015-16, and of the latter, the squeeze on the affordable housing programme during the early noughties as result of the decision of New Labour to stick to the previous government’s spending programmes and to prioritise other areas of social spending.

Third, there have been, however, significant compositional changes, most notably in the secular volume and share decline of Social Rent (SR) provision, within the programme.

Average annual SR provision fell progressively from about 41,000 dwellings in the 1991-2000 decadal period, to 27,700 the following decade, to under 12,000 in the 2011-21 period.

Until 2011 that was largely a combined product of declining grant funding support of affordable housing and a shift to intermediate tenures within the programme.

The decision by the 2010-15 Coalition government then to further reduce and shift government funding support to affordable rent (AR), let at closer to 80% rather than  the closer to 50% market rental levels that SR is let at, meant that by mid-decade the proportion of the programme taken by that new sub tenure began to exceed 50%, before starting to drop back from 2019-20 onwards towards 38% (see Table 3).

Most affordable dwellings provided before the mid noughties were let at Social Rents; during the latest 2018-23 period, nearly three quarters (74%) of the average annual 35,800 combined social and affordable rent completions total were let at higher AR levels.

Fourth, even when such compositional changes are taken account of and adjusted for, as this section did (recognising that results could change if different weightings were applied), affordable dwelling provision since 2001 appears stable on a cross decadal comparison basis (see Table 3A).

The shift to AR allowed more output, officially defined as affordable, to be squeezed from a given level of resources. This, however, was at the cost of higher rents, higher housing benefit costs, worse affordability at a household level, and of heightened risk of pushing more families into poverty after housing costs, even though the rationale lent to its introduction included that AF would improve such metrics if it allowed households to move out of temporary accommodation or the PRS.

Shifting back to SR, however, will involve higher levels of grant support, which given the new government’s fiscal stance, may not be forthcoming, in which case such a shift could involve SR taking a continuing rising proportion of a stable or even a reduced total affordable provision total, while most poor households remained in the PRS.

Efficiencies such as moving to longer-term funding settlements and the targeting of funding support to the areas most in need of affordable housing, such as London, could help to weaken but not lift such trade-offs over time.

Since 2014-15, shared ownership (part owned, part rented) homes have provided about a third of all affordable housing supply, becoming the predominant affordable home ownership sub-tenure.

Overall, all affordable home ownership sub tenure types since 2018-19 have accounted for the same 35% to 38% proportion of all affordable dwellings that it did twenty years ago.

Fifth, a general trend since the nineties – save for some trend interruption in the wake of the Global Financial Crisis (GFC) – is for the proportion of affordable housing across tenures provided through S106 affordable housing obligation process to progressively increase from negligible levels in the nineties and noughties to become the primary reported funding mechanism of affordable housing, accounting for 47% of affordable completions in the latest reported 2022-23 year.

This provision route, in a sense, represented a privatisation of public housing investment into affordable housing (as did the stock transfer in the late eighties and nineties) allowing this time a portion of increased land values to be captured for affordable housing purposes.

In the absence of substantial increases in direct public grant support, this now established provision mechanism will need to continue to do much of the heavy lifting to support affordable housing provision, even though recent market conditions and cost pressures could reduce the scope for such a cross-subsidy mechanism, in the short term, if developer expected profit margins are not to suffer.

Longer term, increased delivery through this route in the absence of wider strategic reform of a private speculative housing business model predicated on profit rather than output maximisation within a market with multiple failures, will rely on higher house prices with attendant affordability and access problems.  

Sixth, the overarching message is that since the nineties both Labour and Conservative governments have struggled to maintain affordable supply, especially of SR, due to increased fiscal pressures and a political unwillingness to prioritise housing relative to other spending programmes and/or tax cuts.

This is unlikely to change with the Starmer government. Although its fiscal rules will allow borrowing for investment, its commitment for public debt to reduce as a proportion of gdp (total economic output) by the end of the new parliament could cause problems and even prove self-defeating, given the potential capacity of increased affordable housing to contribute to growth generation.

Since 2001, annual affordable new supply has accounted for about 27% of total annual new net additions (supply) on a decadal average basis (see Annex Table Four).

For the new government’s commitment to deliver 1.5m new homes by 2029 to be achieved that share will need to increase substantially (see section five for a more detailed analysis).

There is simply little or no prospect of private speculative annual completions reaching the 200,000 to 250,000 dwelling level (250,000 + 50,000 affordable = 300,000; 200,000+100,000=300,000; the more realistic private speculative maximum of 180,000, would require an affordable gross provision annual level of 120,000 dwellings).

The view of this website is that such a step change requires a recasting of both public and delivery systems where the provision of a variegated range of affordable housing sub-tenures is mainstreamed within both, but that is another story.

4          Social Housing Lettings

Data and some commentary reported in this section is taken from the MHCLG January 2024 Social housing-lettings April-2022 to March-2023 (Tenancies) statistical release (January 2024 MHCLGsocial lettings release), unless otherwise stated.

Again, definitions can be confusing. The release describes ‘social rented’ housing as that let at sub-market rentals including Social Rent (SR), Affordable Rent (AR), and Intermediate rent (IR) sub-tenures (see definitions set out in Section 3).

The distinction between new lettings and new first lets to tenants not previously social tenants, should be understood.

New lettings are the combined total of first lets of newly built or acquired properties and reletsof existing properties.

13% of new lets in 2022/23 were in properties that were let for the first time as social housing. The majority of these were new builds (11.5%) with the remainder being made up of new leases (0.5%) and conversions, acquisitions and rehabilitations (0.8%).

87% of new lettings were relets of existing social housing (overwhelmingly LA+RP) stock.

Total social sector (LA+RP) new lettings represent new housing opportunities to both their existing tenants and to households offered and accepting a letting new to the sector, including those coming from temporary accommodation (TA) and/or statutorily homeless.

Around a third (32%) of households starting a new social tenancy in 2022-23 were existing social tenants renewing or transferring within the sector – the remaining 68% entered from outside the sector.

New first lets to tenants not previously social tenants, while they can be the first lettings of newly built or acquired properties or be a relet of an existing property, are those made to households not previously social tenants – whether they formerly resided in the PRS or in other tenures, shared with others, or were in temporary accommodation or otherwise homeless or otherwise were not existing social housing tenants.

51% of total new Affordable Rent lettings were first let in 2022-23, compared to 38% of new Intermediate Rent lets, whereas only five per cent of new Social Rent lets were first lets in contrast to being relet.

SR, however, has remained the primary sub-tenure of the existing social housing stock, explaining why, notwithstanding the secular trend shift towards Affordable Rent (AR) since 2011, 83% of total new social housing (SR+LR+IR) lettings (including relets), 209,190 dwellings, were let at Social Rent levels in 2022-23.

AR accounted for the remaining about 17% of the 2022-23 total new lettings total, compared to 13% in 2015-16. In 2012-13 soon after the sub tenure was introduced, it accounted for seven per cent of total new sub-market lettings.

Intermediate Rent (IR) accounted for a negligible one per cent of total 2022-23 lettings.

The 20 largest social housing providers (in terms of total new and relets) provided 27% of total lets. Of these 20, only one was a local authority (Leeds City Council).

The top three largest PRP providers were Riverside Housing Group, Anchor Hanover and Sanctuary Housing Association, providing in total 16,000 new lettings or 6% of all new lettings in 2022/23.

Since 2007/08 the share of total new lettings provided by local authorities has fallen from 40% to their current 29% share.

PRPs in 2022/23 provided 71% of total 2022-23 social housing (LA+RP) lettings; LAs the remaining 29%.

75% of total new lettings were for General Needs (GN) and 25% were Supported Housing (SH) lettings.

SH is housing provided with special design facilities or features targeted at a specific client group requiring support, for example housing designed for older people or those with disabilities.

38% of GN lettings in 2022/23 were one-bedroom properties or bedsits; 41% had two bedrooms, 19% had three bedrooms, and two per cent had four or more bedrooms.

New letting opportunities and churn rates

Total LA new lets including relets, averaged over 400,000 between 1981-2000,  according to MHCLG Live Table 602, sourced from Local Authority Housing Statistics (LAHS) data.

Their churn rate (percentage of properties let as proportion of stock) increased during the nineties to reach a record 12.3% in 1996-97. That year nearly 417,000 LA dwellings were let, higher than around the 379,000 that were let in 1989-90 and the 404,000 in 1981-82.

Social Rent LA lettings thus held up well during most of the last two decades of the twentieth century despite the massive net loss of stock from the RTB that marked that period.

But, as Table 4 (also derived from MHCLG live table 602) charts, total LA new lets had fallen to about 327,000 by 2000-01 and then continued to calamitously collapse to about 80,000 in 2021-22, before experiencing a small uptick to about 87,000 total lets (including mutual exchanges and short-term lets) in latest reported year 2022/23.

In parallel their churn rate also continuously reduced from 2000 onwards, save for flatlining between 2007 and 2014, halving from 11.5% in 2000-01 to 5.5% in 2022-23.

Instructively, Table 4 also records that LA new lettings to households not previously living in a social housing dwelling plummeted to about 55,000 by 2022-23: less than a quarter of their 2000-01 level.

As discussed later, a secular trend related to the recent ballooning numbers of households in local authority provided and largely funded temporary accommodation (TA).

Neither MHCLG Table 602 nor Table 4 report RP lettings, which are not are not tracked by a consistent long term national time series.

However, the Continuous REcording of Lettings in Social Housing in England (CORE) series does provide data, using a different collection methodology to the LAHS.

It is a ‘flow’ measure of all new social housing lettings recording data at case level (i.e. individual lettings), whereas data in LAHS is a ‘stock’ measure of all social housing stock in local authorities. 

CORE also excludes mutual exchange, short-term and some other lets. Other classificatory changes, in addition, mean that comparisons over time should be made with caution.

Although CORE is designed to be a complete census of new social housing lettings provided by local authorities and private registered providers that own social housing stock, the provision of local landlord information is voluntary. Some LAs do not respond, necessitating a weighted imputation process to be undertaken by CORE using LAHS data.

Further detail and explanation can be found in the MHCLG technical notes that accompany the latest 2022-3 social housing lettings statistical release.

These differences at least partly explain why the LA lettings figures and the relet (churn) rate that Table Four report are higher than the CORE levels reported in Tables 4A and 4B, derived from CORE April 2022-23 social housing lettings data (tenancies, summary table), which includes a compendium of data tables reporting a massive range of tenancy information, some of which is disaggregated to LA and RP landlord level.

Table 4A reports trends in the numbers or volume of all lettings by RPs and LAs (including general needs and supported housing let at SR and AR levels) during the 20011-23 period.

In short, total social sector lets fell from 394,500 in 2011-12 to 251,900 in 2022-23.

It also charts a secular trend reduction in the LA relet or churn rate, similar to what Table 4 reported for all LA lettings between 2011-12 and 2022-23 (7.5% falling to 4.6% compared to 8.3% falling to 5.5% in Table 4). RP lettings as proportion of RP stock also fell from 11.2% in 2011-12 to 6.8% in 2022-23.

Table 4B drills down on general needs tenancies, separating SR and AR lets across LA and RP sub tenures, confirming that their combined lets of SR and AR general needs tenancies dropped from 252,000 in 2007-08 to 182,900 in 2022-23 (the 2020-21 figure of 166,500 was likely a Covid-related outlier).

Total RP lettings including for both general and supported housing needs across all tenancy types fell from 267,200 in 2011/12 to 179,300 in 2022-23 across England.

As Figure 7 of the January 2024 MHCLG social lettings statistical release shows graphically, this was despite the RP stock increasing in net numbers whilst new LA lettings have fallen by more than LA stock did.

Reduced churn relet rates and the secular falling letting trend are a linked feature of both sub tenures.

Declining social housing churn or relet rates in isolation, or in combination, could be attributed to:

• a widening affordability gap between social and market rents;
• previous and continuing net stock losses (a RTB cannot be relet)
• an intensifying inability or unwillingness of social tenants to exit the tenure due to their persistent and unchanging disadvantaged labour market and their other socio-economic characteristics;
• changing tenancy age structure profiles;
• longer void turnaround times;
• allocation policies; or
• other unidentified factors.

The MHCLG in its January 2024 social lettings statistical release, pointed out that generally a higher proportion of stock was relet (churned) in northern England than it did in the south: 7.1% of the combined social stock (LA + RP) in the North East region was relet during 2022/23, compared to 2.3% for London.

It went on to suggest a trend connection with regionally varying social rented (SR + AR) levels and local market affordability.

Figure 24 of the same release graphically represented median (mid-point) weekly rents (£) by rent type, between 2007/08 – 2022/23, across England, indicating a median market rental rent of about £200 in 2022/23 compared to about £180 for AR and about £130 for SR.

Such figures, however, will be influenced by the generally higher churn (relet rate) of social sector rental properties in northern and midland areas. Such properties are let at relatively low rents across those areas compared to high-cost areas, such as London that exbibit low churn. This is another example of a compositional effect.

Figure 26 of the same release plotted the average rent for new social lets (lets at sub-market rents, not just SRs) as a proportion of market rent by local authority area, for 2022/23. They ranged between 25% and 80% of market rentals at the extremes.

Although linking regional rental variances to churn rates accords with common sense observation, supported by any passing acquaintance with the London’s prevailing house prices and market rental levels (recognising their variation within the capital) –(see section six).

Yet the data and current evidence is insufficient to delineate and to definitively quantify separately all the contributory causes of both declining churn rates and letting volumes.

Statutory homeless acceptances

Definitions and interpretations of the homelessness legislation is largely taken from the government’s Homelessness Code of Guidance, unless other stated.

Homeless households found to be eligible for assistance, who are unintentionally homeless and fall within a priority need group (responsibility for dependent children, pregnant women, people over retirement age and others deemed vulnerable, as well as those facing an emergency, such as fire and flood) are owed a main statutory homelessness duty under Part Seven of the 1996 Housing Act by the local housing authority that they have a ‘local connection’ with.

The Homelessness Reduction Act (HRA 2017) further focused on preventing individual and family homelessness through the introduction of new prevention and relief duties. All applicants now are entitled to a minimum 56 days of assistance under the prevention and relief duties, prior to a main duty assessment and decision.

A prevention duty is now owed to applicants threatened with homelessness within 56 days, including when a Section 21 eviction notice (ending an assured shorthold tenancy) has been served which expires within that period, including the provision of advice and assistance.

It applies regardless of priority need status, intentionality or local connection, but generally will not cover applicants subject to immigration control (see chapter seven of the guidance for detail).

To discharge the duty, LAs are required to “take reasonable steps to help prevent any eligible person who is threatened with homelessness from becoming homeless… either (by) helping them to stay in their current accommodation or helping them to find a new place to live before they become actually homeless” (para 14).

The prevention duty continues for 56 days unless it is ended by an event, such as accommodation being secured for the person, or by their becoming homeless.

The relief duty applies where the applicant is or becomes or already is homeless, despite prevention stage intervention. It requires responsible LAs “to take reasonable steps” to relieve an applicant’s homelessness, focused on helping the applicant to secure accommodation available and suitable for their occupation, for a period of at least six months.

Where the housing authority has reason to believe a homeless applicant may be eligible for assistance and possesses a priority need, they must be provided with such interim accommodation (para 15).

The relief duty also lasts for 56 days, by which time the applicant should be provided with a decision as to whether they are owed a main homelessness duty or not, although if it is yet to decide on whether the applicant is in priority need and/or intentionally homeless and there is reason to believe that the applicant is in priority need, the interim accommodation duty owed under the relief duty continues until a decision is made on the main housing duty.

If no main home homelessness duty is owed the responsible LA possesses no further homelessness duties for the applicant.

During calendar year 2023, the relief duty ended for 63,650 households of which:

17,860 were secured accommodation in supported hostel or housing, 17,220 in the PRS, 11,860 in RP accommodation, 7,160 secured a council tenancy, while 3,710 stayed with friends and family (Table R2, homelessness statistics, see link below).

A responsible local authority owes an applicant, who is homeless, in priority need, and not intentionally homeless, a main homelessness duty under the 1996 Housing Act, as set out at the beginning of this sub-section.

That definition was not changed by the 2017 HRA. However, such applicants are now only owed a main duty if homelessness was not prevented during the preventative and relief duty stages and where they fall within the priority need categories (set out in more detail below) and are not intentionally homeless (paras 16 and 17).

Certain categories of household have priority need if homeless, such as pregnant women, families with children, and those who are homeless because of they are a victim of domestic abuse or due to an emergency such as a fire or flood.

Other groups may be assessed as having priority need because they are vulnerable because of old age, mental ill health, physical disability, or having been in prison or care or because of becoming homeless due to violence.

When determining whether an applicant in any of the categories set out above is vulnerable as defined under the homelessness legislation, the responsible LA should determine whether, if homeless, the applicant “would be significantly more vulnerable than an ordinary person would be if they became homeless”.

The main duty requires responsible LAs to provide temporary accommodation until the duty is discharged.

Usually this will be through the offer of a settled home (whether accepted or refused by the applicant), which could be suitable secure or introductory tenancy with a local authority, an offer of accommodation through a private registered provider (also known as a housing association) or a suitable tenancy for at least 12 months from a private landlord made by arrangement with the local authority.

The duty can also be discharged by the LA providing advice and assistance that is sufficient to secure an offer of such secure accommodation that is reasonable and suitable given the applicant’s circumstances. 

It will also end for the responsible authority where the applicant turns down a suitable offer of temporary accommodation, abandons or loses it, or otherwise ceases to be eligible.

Alternatively, it may be possible for the applicant to be ‘homeless at home’, where, for example, an unsustainable family or sharing arrangement is extended on the basis that alternative secure and suitable accommodation will be made available in the future.

During calendar year 2023, the main duty was ended for 43,020 households. Three quarters – 31, 530 – accepted a social tenancy (Part Six 1996 Housing Act social housing offer). 2,860 accepted an offer of PRS accommodation (Table MD2, homelessness statistics).

The MHCLG homelessness statistics collection includes a range of data tables on reasons why applicants became homeless, their circumstances, and actions taken under each of the homelessness duties.

Information at a local LA level can also be obtained from MHCLG detailed local authority tables at a quarterly level and for financial year 2022-23, as well as from its data dashboards on homelessness  and rough sleeping, although this website found them tricky to navigate.

Table 4C reports statutory main duty homelessness acceptances from 1998-99 onwards, at an England level, using the MHCLG source collection.

Until 2004-2005, such acceptances (as then defined) exceeded 100,000 but then fell sharply to about 40,000 in 2009-10.

Subsequently acceptances then rose steadily to about 59,000 in 2016-17, before nearly halving to about 31,000 in 2018-19.

That year represents a discontinuity in the data as it marked the introduction of statutory prevention and relief duties enacted by the 2017 HRA.

Table 4C records that main duty acceptances have since returned to around the 60,000 level for 2023 calendar year (which is also the expected estimated figure for 2023-24).

Recourse to temporary accommodation

Annex Table Six records the numbers placed in temporary in accommodation (TA), by LAs further to their homelessness duties under the 1996 Housing Act and the 2017 HRA defined above.

By 1 January 2024, a record 112,600 households were in such accommodation: more than double than the close to 50,000 households placed during calendar years 2010 and 2011.

Indeed, after falling from 101,000 in 2004 (all dates refer to TA position at 31^st December of the year cited) to 48,000 in 2009, numbers in TA have steadily since increased to their current level.

A July 2024 Audit Commission tackling homelessness report (NAO 2024 homelessness report) advised that the DLUHC (now MHCLG) had concluded that the government achieved the reduction between 2004 and 2010 “through a public commitment to halve the number of households, coupled with significant investment and proportionately higher Housing Benefit” (para. 1.23).

The current near 113,000 total encompasses:

• 29,400 nightly paid self-contained private accommodation placements, including self-contained annexes to hotels with very basic cooking and washing facilities (compared to less than 6,000 between 2002 and 2014);
• 26,700 in dwellings leased by the accepting LA and/or a RP (compared with more than 40,000 between 2004 and 2007);
• 33,400 put in temporary accommodation outside their home LA area – more than three times the levels that was recorded between 2003 and 2012;
• 16,000 were in hotel bed and breakfast accommodation, even though in 2003 such accommodation was prohibited for households responsible for dependent children except in an emergency, and then for no longer than six weeks.

Recourse to TA is uneven, and is concentrated, as one would expect, across high housing need/stress areas, most especially Greater London and Greater Manchester within the North West.

A 2022 Smith Institute study reported a worsening situation in the majority of London boroughs and in the City of Manchester, where the number of households per thousand households was found three times the England average (in volume more than the whole of Yorkshire & Humber) during 2021-22.

The problem was most acute it found in some London boroughs: in Newham it was 12 times more; Southwark, Redbridge six times; in Wandsworth and Westminster five times.

More up-to-date information is provided in Figure 7 of the NAO 2024 homelessness report, which whilst confirming that use of temporary accommodation is particularly concentrated in London and parts of the South East and North West, especially City of Manchester, shows that its use is also scattered across the country and is not confined to those areas.

That said, in last quarter 2023, according to the latest available official data, 63,240 of the total 112,600 households placed in temporary accommodation were in London: 56% of the England total.

Such outcomes can also be linked to declining LA and RP letting availability. Directly provided CORE data reports sharp falls in new lettings to households new to the social sector, falling from 240,700 in 2007-08 across England to 147,000 in 2022-23.

Table 4D documents an even more calamitous long-term fall across London. LA first lets to new social tenants (including those accepted as statutorily homeless) in London fell from 40,500 in 1996-97 to 13,700 in 2011-12, before falling a further 33% to a reported 8,500 by 2021-22.

According to that same GLA Housing in London 2023 source, RP lettings to new social housing tenants fell even more sharply (45%) across the 1996-2022 period from 9,300 to 5,000, even though – unlike the LA sector – the RP sector grew in net stock size during that period.

In total social sector (LA+RP) lets to households new to social housing in London fell from 49,800 in 1996-97 to 13,600 in 2021-22.

CORE data also advises that total social sector lettings in the capital to tenants new to the sector halved from 28,800 in 2012-13 to 14,400 in 2022-23.

Also, according to CORE information that Tables 4E and 4F reproduce, 7,500 combined lettings were made by LAs and RPs in London to households coming from TA or who were statutorily homeless in latest reported year 2022-23 (compared to about 73,000 nationally).

That flow of new lettings to households previously in TA or who were statutorily homeless undershot the 12,040 main duty homelessness acceptances recorded by London responsible authorities during 2022-23 (Table MD3, MHCLG homelessness statistics, detailed local authority statistics, 2022-23).

A gap that seems to provide the main reason why the use of TA is so concentrated in the capital.

International comparisons

A recent article by the authoritative Financial Times (FT) data analyst,  John Burn-Murdoch, noted that homelessness among OECD countries was worst in the UK, when measured by the total numbers, who in 2023 or later, were placed in temporary accommodation.

The UK exhibited a staggering rate of 50 per 10,000 households in temporary compared to about 30 in France, 25 in Germany, 20 in the United States (US), and six in Denmark.

It, however, exhibited a rate of less than one for rough sleeping (varies between the UK’s constituent countries and cities) in contrast to nearly five in Germany and eight in the US.

On the face of it, such statistics suggest relative UK institutional effectiveness relative to its international peers in responding to rough sleeping that obvious and stark manifestation of homelessness, amid the patent ineffectiveness of its overall housing system(s) in meeting general housing need.

It should be borne in mind, however, that the metrics involved will vary in definition, in interpretation, and in measurement methodology between countries, as will the social context in which rough sleeping, especially, takes place. 

The MHLCG Ending Rough Sleeping Data Framework, provides quarterly management information on five core indicators underpinning the four aims of the vision for ending rough sleeping: that it is prevented wherever possible, and where it does occur, it is rare, brief, and non-recurring.

In that light, 3,898 people sleeping rough were recorded on a single Autumn 2023 snapshot night, higher for the second year in a row, but lower than the peak of 4,751 previously recorded in 2017 (Para 2.11 of the NAO homelessness report).

Table 4.4 of Housing in London 2023 report advised that outreach teams recorded 1,614 people sleeping rough in London for the first time, in the second quarter of 2023, representing an increase of seven per cent on pre-lockdown figures.

Most new rough sleepers spent one night only sleeping rough, including during the height of lockdown, when emergency rough sleeper accommodation programmes run by the GLA and London boroughs helped to reduce the numbers of rough sleepers, particularly in early and mid-2020.

Summary and policy implications

Resource to temporary accommodation is not only disruptive and unsuitable for settled family life and for future social-economic advancement but is also costly in short term current public expenditure terms.

Figure 5 of the of the 2024 NAO homelessness report showed that in 2022-23 spending by English LAs on the use of temporary accommodation exceeded £1.6bn: more than double in real (inflation-adjusted) terms than it was in 2010-11.

Local authorities pay for temporary accommodation and reclaim the costs as a subsidy from Department of Works and Pensions (DWP). The Temporary Accommodation Subsidy that they then receive is based on Housing Benefit rules and the Local Allowance Rate (LHA) rate, not updated for long periods from January 2011.

As that central subsidy source has been in effect capped on welfare expenditure control grounds, it has not kept up with rising costs at the local TA coalface. The resulting deficits that LAs incur on TA accommodation have consequently swelled, piling on their financial pressures.

These, according to the same NAO report, given the share of their net budgets taken by TA costs, could soon drive some LAs into bankruptcy.  TA now costs London LAs £1bn per year with £332m of that falling on their own budgets.

The NAO report also pointed out that a significant portion of the centrally financed Homelessness Prevention Grant has been diverted to fund the provision of temporary accommodation, particularly in areas of poor affordability, rather than being spent on prevention work.

This is a classic case of Peter robbing Paul resulting from immediate pressures trumping longer term best value and delivery outcomes – a tendency that has become increasingly prevalent across welfare state provision.  

Such are the facts on the ground that the new chancellor, Rachel Reeves, when deciding public housing investment allocations in accord with set fiscal rules (in Labour’s case, borrowing for investment is allowed but its volume must be consistent with a reduced public debt/gdp ratio by 2029), will need to grapple with in her first 30 October 2024 budget and the subsequent 2025 Comprehensive Spending Review (CSR) process.

Targeting and increasing housing investment funds to the LAs with the highest TA bills, mainly in London, presents a strong investing now to save later case, a baton  picked up in a densely researched and argued New Economics Foundation (NEF) March 2024 Buying Back Better report.

After reviewing the GLA Right to Buy-back (RTBB) 2021-23 scheme, which saw around 1,300 homes bought by councils quicker and often more in accord with client requirements than they would have been if built or obtained through protracted developer affordable housing obligations, the report highlighted that its successor scheme, the Council Homes Acquisition Programme (CHAP) – contingent on sufficient central funding support and tweaks to Approved Development Programme (ADP) rules – could allow London councils to acquire 10,000 homes over the next ten years, or 1,000 annually.  

Such an outcome, the report estimates, would over the next two decades reduce their TA costs by £1.5bn, trim central government housing benefit subsidies by £340m, and generate additional indirect savings of £440m, including projected £39m short term HB savings. Prioritising TA acquisitions within such a programme would save £26m; prioritising general needs acquisitions £49m.

Annual government-wide savings from CHAP would, the report projected, outweigh annual costs including loan repayments after 16 years, when annual savings would begin to accumulate, further noting that if the housing benefit subsidy cap is uprated, greater savings to LAs would result (although central government HB expenditure would increase, all other things being equal).   

A 2022 Audit Commission Review of the AHP since 2015 had earlier revealed, using government research, that in London, future housing benefit savings over 30 years would cover the cost of 69 per cent of the grant cost of providing new homes for social rent, and that this would go up to 110 per cent over a 60 year period, leaving aside savings in temporary accommodation and social care costs.

One step back. Grant rates under CHAP are higher than were under the RTBB scheme, providing grants for up to £200,000 for each social rented home; and for £85,000 for each TA home acquired.

This spend now to save later approach is a microcosm of the wider argument that transferring PRS households receiving housing benefit (HB) to the social sector would accumulate savings in central revenue expenditure over time.

Lower SR rents would require less HB support, underpinning arguments for a SR programme approaching 100,000 new homes annually, as, to take one example, is argued here.

Expanding such funding would therefore be a call on constrained public investment resources with the financial return extended into the long term beyond the lifetime of two parliaments.

While the increased availability of local authority-owned accommodation let at SR rents could produce significant long-term savings for taxpayers via reduced housing support costs (whether centrally or locally funded), that is a hope rather than a certainty.

Future HB costs will depend on a range of future unknown as well as expected factors.

Political focus, on the other hand, is geared to the short to medium term, as are Labour’s fiscal rules: not 15 or 25 years, let along 30 or 60 years, unfortunately.

In truth, the prospect of an annual new SR centrally funded programme being scaled up to approach 100,000 dwellings by 2029 (five years ahead), even when supported by a more certain and robust developer contribution policy regime and helped by a reformed Compulsory Purchase Order (CPO) process effective in reducing land acquisition costs closer to existing use value, seems uncertain at best, and unlikely, probably.

This is, even though, as the next section will demonstrate, Labour’s flagship commitment to deliver 1.5m additional homes in this parliament will depend on – amongst other outcomes, such as expeditious and effective delivery of its New Town, urban extensions, and devolution ambitions – an expanded affordable homes programme rising to a sustainable 100,000-150,000 annual level, sooner rather than later.

As the 2022 Audit Commission report cited above highlighted, the existing regional distribution of AHP grant fails to reflect either the variations in need for investment in different parts of the country or the benefits that would flow from altering it.

The data reported in this section has also demonstrated the linkage between declining social sector lettings and rising recourse to TA, concentrated in London and other high need hotspots.

The case for adjusting the AHP as part of the 2025 CSR in favour of London and tilting it towards supporting additional social housing provision across such areas appears overwhelming on both housing need and expenditure efficiency grounds.

Likewise moving to a ten-year rather than a five-year AHP, as proposed in the 2020 Double or Quits  University College of London (UCL)  Bartlett School of Construction and Project Management  report, and subsequently by many  stakeholders, accompanied by other efficiency enhancing programme rule flexibilities, should allow potential supply volumes and efficiency to be increased, as well as help to smooth delivery over the duration of the programme, avoiding programme-end peaks.  

More broadly, greater attention on the lettings data reported and their implications seems justified and necessary. At the end of the day, it is the letting capacity and the flow of annual new lettings of social landlords that determines new housing opportunities for households in housing need.

Given the low churn rate in high cost and high need areas, especially London, the systematic routemap to maximise LA letting capacity through the development of letting chains that a former Southwark Council cabinet lead for housing, set out in another 2022 Smith Institute publication, appears pertinent.

Given RPs dominant sub-tenure and repository of public investment status and position, more transparent and topical information on the volume trend of LA nominations to RPs could also be helpful.  

5        New housebuilding, indicators of new supply, and final tenure estimates

The MHCLG in September 2020 changed the title of its “House building: new build dwellings’ statistical release to “Housing Supply: indicators of new supply”.

As the series live tables were predominately concerned with new build, this was a confusing change, in the view of this website, risking conflation with the “Housing supply: net additional dwellings” series covered in section one.

A recent quarterly published statistical release stated the purpose of the series is: “to provide an indication of the levels of and trends in new housing supply in England” advising that it should “be regarded as a leading indicator of overall housing supply”.

Section Four of the release continues to highlight the building control reported new build starts and completions data (termed in this post, the new housebuilding series), even though, as explained and demonstrated below, it covers 80% or less of house building activity in England and systematically undercounts new build activity.

Section Five of the release does, however, cover other indicators of housing supply, advising that along with the above they are intended (in totality) to “provide a suite of information which give a rounded estimate of the current trends in housing supply and an indication of what the more complete estimate of housing supply, net additional dwellings, will show when it becomes available in November 2024”.

New housebuilding series (MHCLG ‘new build dwelling’ figures based on building control inspection data): the live tables

MHCLG Table 213 covers both new build starts and completions, by tenure of the provider, both on an annual financial year (from 1969-70) and on a quarterly non-seasonally adjusted basis, (from quarter, Q1, 1978 to Q2 2023 at the time of writing, then updated by the quarterly reporting cycle).

It starkly reports that during 2023-24 only 134,780 new build dwellings were started (the lowest since 2013-14), 104,310 of which were reported as built by private enterprise (the lowest since 2012-13).

During the same period, 153,800 dwellings were completed, 115,780 of which were reported as built by private enterprise (both figures the lowest reported since 2015-16).  

MHCLG Table 222 provides similar but quarterly seasonally adjusted figures.

MHCLG Table 244 catalogues the longer-term new housebuilding record, broken down by tenure, on a calendar year basis, from 1946 onwards.

MHCLG Tables 217, 253 and 253a report dwellings completed and started by tenure, broken down to region (quarterly) since 1990-91 (217), LA district (253, by financial year, 252a, quarterly since 1980-81).

MHCLG Table 254 reports completions and starts, by house or flat type, by bedroom size, and by tenure. 

The Office of National Statistics (ONS) in its House building, UK: permanent dwellings started and completed by country statistical series also continues to publish new housebuilding data according to country, as well as at a pan Great Britain and UK levels. 

ONS tables relating to England replicate the live tables published by the MHCLG and are updated to incorporate revisions, unlike the annual statistical release commentaries.

New housebuilding series (MHCLG ‘new build dwelling’ figures based on building control inspection data): development, methodology and undercounting

The new housebuilding series began in 1946, sourced solely from LA building control information, considered the best source to identify when new build dwellings are started (specifically the commencement of construction in laying of foundations) and the timeliest measure of new build completions (as measured by the completion certificate).

However, such sources became increasingly fragmented. National House Building Council (NHBC) data was added from 1985 and then in 2007 independent approved inspectors also became a statutory source of data.

The MHCLG estimates that the building control sourced data source currently provides information on about 80% of house building in England.

In computing the series, MHCLG statisticians make allowances for non-response (reported as high as 78% for independent approved inspectors in Q1 2023). This done through a process of imputation from past returns to take account of late submissions.

The technical notes accompanying the series recognises, however, that late and non-reporting continues to prove a problem.

Indeed, both the MHCLG and ONS live tables under-enumerate recent new build completions, often by a large magnitude.

Annex Table Seven reports that the annual discrepancy (undercount) between the new completion totals reported in Table 213 of this series and the more accurate net supply Table 120 reached an annual average of about 24% across the most recent 2018-23, period, compared to an average of about 17% across the entire 2006-2023 period covered by Table 120.

This ONS new housebuilding series similarly consistently undercounts recent new build activity.

Reporting new housebuilding by tenure more accurately

The MHCLG interactive dashboard published with this series provides important new data on new build completions by final tenure, when occupied.

This is the final tenure usage of the new dwelling, whether for future private ownership, for housing association or for local authority housing use.

The final tenure use of a completed dwelling is not necessarily the same as the tenure of the developer that built it because private developers also build dwellings for housing associations and local authorities, as well as predominately for onward private sale.

Indeed, as was discussed in Section Three and documented in Annex Table Five, a progressively larger proportion of affordable dwellings since the early 1990 have been built by private developers in accordance with S106 planning agreements concerning affordable housing obligations.

Some, perhaps most, of these dwellings are reported and recorded as private enterprise completions within the new housebuilding series tables, even though they are later transferred or sold to RPs for future use within their stock.

Annex Table Eight reproduces tabulated information that can be accessed from the series dashboard’s final tenure estimates tab (see previous link), cataloguing new build completions from 2006, by final tenure.

It utilises the more accurate net supply (total new build completions) and the affordable housing series (LA and RP new build completions) from 2006 onwards – when the new build component in the net additional dwellings estimates was first published – to model and estimate more accurately private enterprise final tenure figures (Section 7of the statistical release describes in more detail).

Prior to 2006, the tenure of dwellings, by provider, is derived from building control sources, meaning that the table reproduces new housebuilding series data.

The table also conveys the cyclical sensitivity of private speculative new build activity to the wider economic environment.

Annual private enterprise completions more than halved from about 164,000 between 2006-8, the high-water mark of an era of low inflation, interest rates, and steady economic and household income growth, to less than 70,000 between 2009-12, as the lagged impact of the GFC played out.

During the 2006-23 period, LAs and RPs accounted for – on an average annual basis – about 26% of all new build completions, by final tenure.

That proportion varied, however, not only with the annual trend of new build affordable completions but also with that of private sector completions: their collapse in 2010-11 largely explains why the social sector share of all new completions rose to about 42%, by final tenure that year, as it took a growing share of a dwindling total (the denominator fell proportionately more than the increase in the numerator).

As way of contrast, social sector (LA+RP) annual new build completions fell to historic lows of between 13,000 to 14,500 dwellings – less than 10% of total new build completions during the 2000-2003 period: a result that reflected the lagged impact of early New Labour housing investment levels having been crimped by its self-imposed commitment to keep to the previous John Major government’s spending requirements, as well to the higher priority that New Labour accorded to other social programmes.

It was only towards the end of its period in office, when, in response to the 2004 Barker Report, social housing investment allocations were quite substantially increased, enabling some recovery in social housing new building activity, which subsequently partially blunted the impact of the GFC on total completions. 

Projecting 2023-24 new building completions and housing new supply

Both the Conservative and Labour parties during the 2024 general election committed to deliver at least 1.5 million new homes during the lifetime of the next parliament – equivalent to 300,000 new homes each calendar year, 2025-29 inclusive.

Putting that vaunted ambition into perspective, about 212,000 new builds were completed each year from April 2021 to March 2023, according to the most accurate MHCLG new additions (supply) series, while net new supply (including gains from net conversions, change of use, less demolitions, additional to the 212,000 new builds) hovered around 234,000 dwellings (see Annex Table One).

Concerning the expected 2023-24 outturn, Section Five of the MHCLG’s latest new indicators of housing supply release as discussed above, provided analysis of indicators other than building control new build data, including Energy Performance Certificates (EPS) lodged for new dwellings, building and council tax data.

Instructively, the release went on to advise that the EPC statistics provide a very close estimate or proxy to net additions, almost to the point of giving an official steer that are the most reliable indicator of future new supply.

At time of writing, however, such EPC data is relevant to the immediate past 2023-24 financial year to be reported in November 2024 – at least until sufficient 2024-25 EPC data is reported.

In that light, Table 5 compares the number of EPC of certificates lodged for new domestic properties (new build, conversions and change of use to domestic) with MHCLG Table 120 net new additions (new supply) across 2018-23 actual net new supply outturns.

It indicates that net supply on an annual average basis was 0.98 below the reported number of EPCs granted across that period (although, this was after a retrospective census adjustment was made for the 2018-21 new supply figures).

Applying that 0.98 coefficient to the reported 232,473 new dwelling EPCs reported lodged during 2023-24, suggestsa 2023-24 new net supply figure of about 227,000 dwellings, continuing its recent downward trend. 

Projecting future 2024-27 new build activity and new supply: implications for Labour’s 1.5m delivery target

MHCLG new housebuilding and other indicator of new supply series data, as customised in Tables 5A and 5B, strongly suggests that the recent downward new supply trend will not be reversed during the early 2024-26 period of the Starmer government.

MHCLG Table 213 data reproduced in Table 5A reports that 481,240 new build dwellings were started between 2021-24, an annual average of 160,400.

To compensate for the systemic tendency for new housebuilding series to undercount, as was laid out above, an adjustment multiplier of 1.24 was applied to the to the Table 213 starts and completion data reported in Tables 5A and 5B to produce adjusted figures.

That adjustment made was consistent with 2018-23 under-enumeration of Table 213 MHCLG Table 120 new build completions of about 24% that Annex Table Seven demonstrated for the 2018-23 period.

The average difference between Table 213 reported completion totals across 2018-23 and the later net new supply annual reported totals reported for that period was 41.5% and an adjustment coefficient of 1.415 was applied on Table 213 start and completion data to produce projected future new net supply figures

On that basis, Table 5B projects a 2024-25 and 2025-26 net new supply outturn (assuming a two year start to completion gestation) of about 244,000 and 191,000 dwellings, respectively (the reported adjusted new build 2022-24 start totals rolled forward two years as completions), suggesting a 2024-26 financial year net new supply total of about 435,000 dwellings; or an annual average of about 220,000 across those two years.

Such a level would generate a cumulative delivery shortfall of about 160,000 dwellings relative to the 300,000 annual target by April 2026.

Such rule of thumb projections are inherently tentative and indicative. Although yesterday starts are tomorrow’s completions, starts data provide a forward but imperfect indicator of future completions: they are subject to cyclical fluctuation related to the external economic and market environment and to funding programme profiles and associated delays.

New build completions, in practice, can also lag for longer than might be expected if a one-to-three-year construction period is assumed. Large scale regeneration projects involving multi-phases invariably straddle many years.

Lags could also possibly relate to incentives to record starts for programme monitoring purposes or result from other unquantified reasons.

In short, there is no way of knowing precisely when reported starts will be converted into a conversion.

Given the subdued nature of the private housing market in 2024 and the expected continuing impact of high (albeit hopefully further softening in line with the July 2024 0.25% base rate cut) interest rates, little reason currently exists that 2024-2026 completion totals will prove much different to the expected and projected weak 2023-24 outturn.

Indeed, as Section Three projected, a downtick in new affordable supply annual starts over the 2023-25 financial year period to an average 50,000 dwellings or less can be expected, suggesting a similar level of affordable completions during 2025-27.

Future affordable supply prospects beyond that are uncertain given its dependence on private cross-subsidy (less is available in a weak private market), the future impact direction of building and labour costs, and RP ability to use reserves to finance new building considering post-Grenfell and environmental regulatory requirements that require increased investments in existing properties.

In such an environment, increased grant funding needs to be made available to secure a future step change in affordable supply. Angela’s Rayner’s 30^th July statement to parliament on that score was vague in terms of whether the forthcoming CSR housing settlement would involve substantial additional resources rather than a refocus of resources towards Social Rent (SR) provision.

Increased ADP resourcing would involve a time lag of least two years before it translated into new build completions, although acquisitions involving programmes such as CHAP (discussed in section four) could be quicker.  

The new Labour government also intends to introduce a revamped mortgage guarantee schemes for first time buyers in place of the retired Help-to-Buy (HtB) and to prioritise implementation of its pledge to ‘build rather than block’ planning reforms and compulsory purchase order (CPO) process reforms.

But planning reform by itself, while necessary, will not result in the current speculative private market-led system delivering future housing volumes on the scale required.

The outlook for completion outturns going into 2027 is not much brighter, therefore, on any realistic reckoning.

 Barratt, example, in July 2024 announced a seven per cut in planned starts for the coming year into 2025 that will then inevitably translate into future 2026 and 2027 completions.  

Research conducted earlier in the year by Savills included a chart that indicated – assuming no increase in government housing support – annual new build completions not breaking through 160,000 throughout the remainder of this decade, meaning that annual net new supply would undershoot 200,000 let alone reach 300,000 (less than 1m rather than 1.5m).

That doomsday scenario is unlikely to come to pass given Labour’s linkage of planning reform to its growth agenda, New Towns programme, and commitment to boost affordable housing supply, but the undoubted delivery shortfall projected above by 2026-27, supported by the work of other commentators, means that for the 1.5m  target to be achieved, supply would need to reach well beyond 300,000 during the 2027-30 period.

Even a 2024-27 (financial year as measured by MHCLG Table 120) cumulative new supply total of 750,000 or 250,000 annual average new supply across that period (on the current evidence, unlikely, as presented above, with undershooting much more likely), would require annual new supply delivery (let alone new builds) across the 2027-29 financial years to exceed 320,000 dwellings annually to achieve the 1.5m homes target (1.5m minus 750,000 = 750,000, which divided by two = 325,000).

This was something achieved last during the 1964-70 Wilson governments, and then not really, as it straddled a period when net supply was lower rather than higher than new build totals because of accompanying slum clearance demolition and redevelopment activity.

Starmer and Labour’s target presupposes that several New Towns and urban extensions, at least in substantial part, are planned, land acquired, prepared, approved, funded and completed by the end of 2029, amid a wider step increase in both private and public activity, measured by completions.

Angeal Rayner, herself, no doubt fortified by internal MHCLG briefings, intimated in her July parliamentary statement that supply would need to rise to 375,000 annually to compensate for early under-delivery.

Such outcomes, on the face of it, seems unrealistic, pure and simple.

Perhaps, a more realistic but still stretching target, requiring early concrete and effective policy action and reform, would have been to achieve 300,000 starts by 2026 calendar year on a future sustained basis.

Even that target would be tricky to measure, given the new housebuilding series (which measures quarterly without much time lag) related tendency to under count.

Indeed, the measurement problem where MHCLG table 213 new build starts and completions can be counted on a quarterly and annual basis but involve systemic undercounting, while the more accurate MHCLG table 120 new supply (net additions) table reports for financial years 18 months will, in any case, make the measurement of Labour’s flagship target, if not a fool’s errand, problematic, heavily exposed to political spin.  

Besides, because of construction lags and the overlapping impact of previous policies and conditions, a total production or supply target based on completions over the lifetime of a parliament didn’t and doesn’t make much sense, other than as a statement of political intent (hopefully) or spun aspiration (more usually).

Summary and policy implications

At a statistical reporting level, this website has previously recommended that MHCLG should advise users that the new housebuilding series should only really be used for two main reasons.

First, to consider longer-term trends related to the tenure breakdown of housing starts and completions, especially prior to April 2006.

Second, although the Housing supply: net additional dwellings series provides the best measure of new net supply and new housebuilding completions, it remains a lagged backward-looking metric that does not cover starts, requiring that recourse to the new housebuilding series as a leading but imperfect indicator of future near-term new build activity is necessary, complemented by other sources of evidence, where available and appropriate.

The new housebuilding series demonstrated tendency to undercount that, however, can be compensated for, as was done in this section, albeit on an imperfect rule of thumb basis, subject to phasing and contingent factors and uncertainties.  

The placing and enforcement by the MHCLG of more effective statistical collection and reporting requirements relating to housing starts and completions, although administratively challenging, would relieve the problem at source.

More importantly, scenarios based on the evidence currently available and considered in this section, point to a conclusion that the new government could possibly struggle to deliver over one million, let alone one and a half million, additional homes during the next 2024-29 parliament.

Even approaching Labour’s lifetime parliament target of 1.5m homes will require many New Towns and/or urban extensions to be identified, planned, authorised, funded, and at least partially completed within five years (something that would not necessarily show up in official statistics at the earliest in November 2030, probably more than a year after the next election).

Such a feat would require sustained political overarching focus and institutional coordination and drive of a nature unprecedented and unexperienced for decades.

It would also require public pump-priming infrastructure investment mid-decade rather than end-decade, enabled by a firm but flexible, rather than ironclad, interpretation of fiscal rules, focused on sustainable growth and best use of public resources over the medium term rather than mechanical and rigid calculations of future debt levels.

A focused, co-ordinated, streamlined and effective public-private partnership approach to housing delivery on lines suggested by the 2017 Letwin report, yet going further, in line with an overarching and primary political commitment to achieve a step change in housing, especially affordable, delivery accounting for a much larger than the current 27% share of total delivery.

These are preconditions but are not guarantors of the achievement or near achievement off the 1.5m target in practice rather than rhetoric.

The Starmer government in its July 2024 proposed planning reforms demonstrated its intent to remove blockages connected to local Nimbyism and other planning-related obstacles.

However, setting local delivery targets consistent with an annual supply level of 375,000 is not the same as securing their actual delivery on the ground. Councils do not currently build dwellings at scale (other than a few homes built for rent within the Housing Revenue Account (HRA)): developers do.  

Achieving such an elevated target, requires not only sufficient planning permissions but their implementation by profit-maximising private developers, whose current business model requires them to rather build and release dwellings for sale in step with that imperative, not government targets.

6          House Prices, Rents, Affordability and Access

This final section examines house price, rental, and affordability data, explaining their differences between the different sources available and their related uncertainties, complexities and limitations.

2022-23 English Household Survey (EHS) data is also utilised and reported to provide relevant information on household tenure and housing affordability circumstances

House prices: the different indices

Getting a handle on house price outcomes and trends is tricky. Several different sources often reported in the media use different data and methodologies covering different points in the home purchasing process, which vary in coverage.

For example, Rightmove use asking prices. Lender sources, such as Nationwide and Halifax, use their own mortgage approvals data. The UK House Price Index (UK HPI) published by the ONS uses mainly Land Registry data produced at the end of the conveyancing process and reports transacted prices.

The ONS explains the various house price index measures and their strengths and limitations in their publication Comparing house price indices in the UK.

This points out that although Rightmove’s index, which is based on advertised asking prices is the timeliest offering a leading indicator of future sale prices, properties may not ultimately sell and complete at their advertised prices, resulting in consequent inaccuracy.

Nationwide and Halifax indices are based on their own mortgage approvals, which means that they also can publish quickly. However not all approvals lead to a transaction; moreover, individual lender approvals are not necessarily representative of all approvals; even more seriously, these indices do not include the 30% to 40% of transactions that are cash purchases. These indices, therefore, may provide a biased estimate of actual sale prices.

The ONS series and the LSL Acadata series are the least timely as they use transaction (registration of the sale in the Land Registry) data at the end of the conveyance process, published usually around two months after the latest period reported.

These series are not, therefore, forward-looking and tend to be of less interest to the media when hungry for a timely and catching headline and for forecasters wishing to project or predict future trends; they are, however,more complete in ultimately capturing the set of actual transacted properties during the period that they report.

The average prices that UK HPI reports are a (weighted) geometric mean and by excluding the inflating impact of a few highly priced properties the UK HPI measure aligns reported mean (average) prices closer to median (midpoint values)

This in contrast to the simple average that the ONS quarterly house price tables, based on a on a sub-sample of the Regulated Mortgage Survey, report.

Figure 2 of the ONS guide cited above, which compared the average price time series for the various indexes, indicates that the UK HPI reported values, are as much as 25% lower than those reported by the LSL Acadata and Rightmove series. The same divergence is discernible between it and the ONS’s own quarterly series.

The UK HPI series is also a mix-adjusted index, which means that it is weighted to reflect annual changes in the mix and other attributes of the properties transacted.

The period with the base period for its price series is updated every five years, when the whole price series is rescaled to align with the new base period. Halifax also update their weights annually, while the Nationwide index updates its weights every two years.

To put it simply, mix adjusted indexes aim to report changes in average prices that take account of the changing composition of its source data of registered transacted dwellings, screening out possible distortions associated with outlier transactions, such that may arise with the sale, say, of highly valued properties.

The UK House Price Index (UKHPI) publishes UK HPI average house price estimates, associated indices and percentage growth rates down to regional and local authority level, according to property type, to buyer status (first time or former occupier), and to funding and property status (new build or existing).

The series thus offers greater granularity than do the other house price indices. Because it is based on all completed sales (rather than advertised or approved), including and cash and mortgage-financed purchases, it better reflects true house price trends, but with a backward-looking publishing lag.

National real house price trend since 1980

Annex Table Nine provides a time series between 1970 and 2024 for mix adjusted house prices across England, according to type of property, derived from that UK HPI series.

This table also offers a conversion of those cash prices into real (inflation-adjusted) terms.

A warning: the conversion of cash into real prices can be sensitive to the choice of deflator. Reported outcomes could differ if another conversion deflator was applied, such as the Retail Price Index, while, as explained above, the reported cash prices can vary substantially depending on the methodology and coverage of the index used.

Accordingly – as the case generally with house price indices – regard should be paid to trend variation rather than precise yearly figures conveyed.

Taking that on board, the main contours of the English house price story since 1980 can be broadly summarised thus.

1. House prices in real terms were stable between 1980 and 1985, increased sharply between 1985 and 1990, before falling in both nominal and real terms between 1990 and 1996 – a period bookended at the start by a sharp recession in the early nineties that led to record levels of repossessions and negative equity; and at the end by the ushering-in of a period of low inflation, higher growth, and rising household incomes – sometimes called the Great Moderation.  
2. By 2007, real prices had increased by a factor of between two and three. Cash or nominal prices during the 2000-2005 period exhibited an average annual percentage inflation increase of about 14% when annual inflation was around two per cent.
3. The came the Great Financial Crash (GFC). House prices fell sharply in cash and real terms in 2008-2009, before recovering slowly in real terms by April 2014.
4. Prices then the rose sharply again during 2015-2018, and 2020-22, before falling back slightly in real terms during the 2022-24 period, responding to rising interest rates and the cost-of-living crisis.  

Given the compilation and methodological issues briefly discussed above, the overall trend direction, as broadly summarised, should be taken as a guide with the average price levels reported for any year its conversion into inflation-adjusted constant price as indicative rather than precisely accurate.

Even more crucially, snapshots of house prices at any time point very much vary between regions and often areas (and sub-areas) within regions.

Country and regional trends

Table 6  reports average prices across the countries of Great Britian and by English administrative region, according to buyer (first time or former occupier) and property status (new build or existing), recorded by the UKHPI for April 2024.

It reports an average house price paid for all property types (taken as a proxy for all buyers) buyers in England of about £298,000 (nearest thousand), £158,000 for the North East, £203,000 in Yorkshire and Humberside, £375,000 in the South East, and £502,000 in London; £229,000 and £208,000 for Scotland and Wales, respectively.

With respect to the average price paid by first time buyers, it reported £250,000 for England, £136,000 for the North East, £175,000 for Yorkshire and Humberside, £299,000 for the South East, and £435,000 for London, and £153,000 and £180,000 for Scotland and Wales, respectively.    

Recent July 2024 data on lower quartile price non mix adjusted data also published by the ONS by indicates that in England that was £190,000, with lower quartile prices (price at the 25^th point of the price distribution) ranging between £104,000 in the North East to £397,000 in London.

Table 6 also reports data on average prices paid for new and existing properties. The average reported new build price in England was £420,000 compared to £293,000 for an existing dwelling.

This new build premium of over a third appears surprising. This reported outcome is likely to relate to the compositional mix of new builds, and the tendency that they tend to be bigger in size and more likely to be detached at least in some regions (not London).

Such considerations underscore the need to treat headline house price data with caution, commensurate to the purpose of the inquiry.

The ONS provides an interactive guide to track average (estimated reported) house prices at the LA level in March 2024 (at time of writing, June 2024), by type of property.

A House of Commons Library dashboard, does so by single parliamentary constituency to neighbourhood level. It reports the non-mix adjusted median price paid for properties during the preceding twelve months on a rolling three monthly time series basis with latest reported at (at June 2024) for September 2023.  

Private rental levels

The primary source is the ONS’s private rental summary statistics series. This reports monthly rental prices for the private rental market in England according to bedroom category, by region and LA administrative area, calculated mainly from the Valuation Office Agency (VOA) data.

The ONS warns that that its source “samples are not statistical” and they should not be used to compare figures over time and between areas; nor do they include housing benefit (HB) claimants: a glaring omission insofar that 24% of PRS tenants according to the 2022-23 EHS were in receipt of HB.

The series reports lower quartile (the datapoint splitting the lowest 25% of rents from the highest 75%), the median (midpoint rent), and upper quartile (the datapoint splitting the highest 25% from lowest 75%), as well as recorded mean rent figures, for non-self-contained ‘room’, one bedroomed, two, three, and four bedroomed dwellings, for England, by region, across the 1 October 2022 to September 2023 period.

It is reproduced in Tables 6A to 6C for ‘rooms’ (bedsit-type accommodation as defined in note 1), one, and for three bedroomed dwellings.

The recorded median monthly rents for one bedroomed properties over that period for England, the North-East, the South East, and for London was £750, £450, £850, and £1,400, respectively.

For family sized three bedroomed dwellings, it was, £925, £600, £1,300, and £1,950, respectively.

These figures should be taken as a rough snapshot guide for the period in question.

They, of course, will also vary with area and neighbourhood; nor are they quality adjusted.  

Affordability and access (owner-occupation)

Another ONS dataset, using house price statistics for small areas (HPSSAs) derived from Land Registry recorded transactions, provides time series information on median house prices, earnings, and the median house price to median gross annual residence-based earnings affordability ratio, all by country and region (tables 1a to 1c),  by county (tables 3a to 3c), and by LA district (tables 5a to 5c).

Table 1c reports that the median house price to median gross annual residence-based earnings affordability ratio rose from 5.11 in England and 6.9 in London in 2002 to 9.06 and 13.62, respectively, in 2021, before subsequently slightly falling back in 2022 and 2023 as house prices dipped in response to increased mortgage rates.

The same ratio calibrated to the July 2024 ONS lower quartile series reporting the ratio of lower quartile house prices to lower quartile residence-based earnings, indicated a ratio in September 2023 of 7.25 in England, ranging from 4.32 in the North East,  to 13.1 in London, and 10.23 in the South East.

Such data does not report the affordability ratio for households who possessed sufficient savings or other resources (increasingly bequests) and income to purchase, nor the actual mortgage repayments of such households relative to their incomes, nor does it record properties sold at less than full market value.

They, accordingly, provide an indicator of the direction of affordability over time, but little about individual circumstances and access to home ownership for households with different circumstances, other than undoubtedly it has become much more expensive in general to purchase, relative to individual earnings.

Another House of Commons library (HOC) datasource using the ONS house price statistics for small areas series (HPSSA), reports house prices in each parliamentary constituency and the house price to earnings ratio.

It compares the median house price in each constituency (three-year 2020-22 average) with the median salary for full-time employees living there in each parliamentary constituency.

Fourteen London constituencies record a median house affordability ratio of more than fifteen. At the other end of the scale, eight constituencies report a median house affordability ratio of less than four, including Burnley, Copeland, Liverpool Walton, and Easington in England, and Blaenau Gwent, Rhondda, Aberavon, and Cynon Valley, in Wales.

As was earlier explained, the UK HPI is weighted to reflect the mix of properties sold in the previous year to make it broadly representative of the mix of properties in the overall dwelling stock.

The HPSSA, unlike the UKHPI, however, is not mix-adjusted, but uses rolling years to better reflect the actual mix of property sold.

As the UK HPI provides a measure of the changing market value of properties in the housing market whereas the HPSSAs measure the price paid for properties sold in each period, the two sets of statistics provide different figures, albeit exhibiting a similar time trend.

Unfortunately, the HPSSA series has been discontinued by the ONS and the final data available is for September 2022.

Another problem with all these measures is that earnings data is an incomplete measure of income, not taking account that many households have multiple earners, nor for the impact of other sources of income, such as benefits. At a local level, small sample sizes may aggravate variances.

Actual affordability ratios based on house price and mortgage advances data,  such as Table 44a of the 2024 UK Housing Finance Review,  reporting mortgage cost-to-income ratios for first-time buyers to the country/region level, indicate that the average ratio in England was 19.5% in 2022, while across its most expensive areas, including London and the South East, it registered between 20 and 21%.

The problem with such series – the converse to the one inherent with overall price to affordability ratios reported in the ONS datasets cited above – is that they only capture households that have been able to access and purchase through a mortgage a home, not those who were unable to access due to deposit requirements and/or unaffordable repayments whether by their own or lender decision.

The housing costs and affordability chapter of the 2022-23 EHS reported that the average (mean) deposit of a first time buyer in 2022-23 was £50,051 (£30,000 median), with 58% of such buyers coming from the top two income quintiles.

This piece of information helps to explain why 47% households in 2021-22 led by someone aged 25-34 were homeowners, less than the 59% recorded in 2003-4.

In 2021-22, 36% of purchasers reported receiving help from family or friends, while 9% used an inheritance as a source of deposit, a significant increase on the 22% of first-time buyers who reported help from family and friends in 2003-04.

89% of mortgagors reported in 2022-23 that they found it “very or fairly easy” to afford their mortgage, compared to 93% of mortgagors in 2021-22, with the proportion finding it “very easy” fell from 48% to 37%.  

However, an increased proportion of respondents reported a rise in some level of difficulty in affording repayments, with 9% of mortgagers finding it “fairly difficult” and 1% finding it “very difficult”.

The ONS has highlighted that monthly repayments on newly issued mortgages increased significantly recently in line with rising interest rates, noting that 1.4 million households in the UK faced the prospect of interest rate increases when they renewed their fixed rate mortgages in 2023.

Its map based affordability calculator indicates that purchasing a semi-detached property in the UK at the average December 2022 price of £286,000 would involve a monthly mortgage repayment of £1,262, assuming a mortgage term of 25 years: a £481 (61%) increase compared with the corresponding monthly repayment estimate in December 2021.

Purchasing the average detached UK property on the same terms in December 2022 would have resulted in a monthly mortgage repayment of £2,041 (up by 60.7% on December 2021).

For terraced houses, it would have been £1,063 (up by 59.6%). For flats and maisonettes, it would have been £1,028 (up by 54.6%).

In December 2021, a loan-to-value rate of 75% and a budget of £1,000 per month would have enabled a purchaser to afford an average semi-detached property in nearly two-thirds of local authorities in Great Britain.

With the same budget just one year later in December 2022, however, the same purchaser would have been able to afford the average semi-detached property in less than a third (30.1%) of areas.

Annex Table Two of the 2022-23 EHS, reported mortgage/rent as a proportion of household income for mortgagors and renters, resident for at least three years in 2022-23. 

That EHS table did not include housing-related costs, such as water and fuel bills, insurance, maintenance costs and council tax, were not included in that calculation. Income was taken to be the gross weekly household income with separate tables including and excluding benefits. Outright owners were excluded.

Two ratios were offered: one based on the household income (i.e. the income of all the members of the household, who are assumed to contribute to the mortgage); and the other based on the Household Reference Person (HRP) and their partner income only.

Between 2012-13 and 2022-23, the proportion of household income that mortgagors on average spent on their mortgage remained similar (19% in 2012-13 and 18.7% in 2022-23, according to the HRP plus partner measure).

Access and affordability: renters

According to the 2022-23 EHS, 59% (2.4 million households) of social renters and 24% (1.1 million households) of private renters received housing support (HB or the housing element of Universal Credit) to help with the payment of their rent.

A seminal June 2023 Institute of Fiscal Studies(IFS) study on renter housing quality and-affordability for lower-income-households reported that low-income households are increasingly private renters, renting relatively low-quality homes.

Younger low-income individuals today are more likely to privately rent than did older generations at the same age.

According to the same study, as at 2023Q1, just five per cent of private rental properties were affordable for housing benefit recipients.

Such properties were also more likely to have an energy rating of D or below, and on average have 19% higher heating and hot water costs. More generally they are more likely to be in low-employment and high-crime areas.

Moreover, most such tenancies are likely to be insecure and to suffer from quality issues. 

A 40% gross earnings rental affordability ratio, or more pertinently a 30% disposable income (including benefits) ratio may well push households further in poverty or into poverty, depending on individual household circumstances.

In that light, Annex Table Two of the 2022-23 EHS reported that for private tenants the average rental affordability was 40% (including housing support) and 46% (excluding housing support) relative to HRP plus partner income and for housing association tenants 30% and 38%, respectively (to nearest percentage points).

Although the 2023 Autumn Statement confirmed that Local Housing Allowance (LHA) will increase in April 2024 after having been frozen for long periods since January 2011, returning its coverage up to 30th percentile rent level, according to bedroom size, applicable to the local market area, making 30% of the market theoretically affordable to those reliant on housing benefit.

However, a recent Savills analysis of over one million Zoopla asking rents in 2023 showed that across Britain only 8.5% of such new listings would have been affordable using the new LHA rates.

The HoC datasource, using the ONS private rental summary statistics series, reports monthly private rents for two bedroomed homes, expressed as a proportion of monthly earnings, by LA area, for 2021-22.

Over 30 LAs, concentrated in London, reported a renter affordability ratio greater than 30.

Tables 6D-F  drills down on the ONS private rental summary statistics series.

It calculates monthly/annual gross incomes consistent with an arbitrary 40% affordability benchmark relative to room, one bedroomed and three bedroomed lower quartile rents, by region.

That an annual gross income according to that yardstick was required in London of £19,500, £35,100 and £49,500, respectively, for the period ending September 2022, suggests that the problem of affordability endemic for low-income households in the capital is both concentrated and endemic.

A more up-to-date and forward looking source is an ONS ‘statistics in development’ dataset on new renter affordability.

This estimates the proportion of actual gross earnings taken by rent paid by individual renters, starting new tenancies during the reported months up to May 2024, excluding guarantors, or individuals with very low incomes who are likely supported by another income source or whom have very high incomes that could skew the results.

The ONS advises that these reported estimates “reflect affordability at snapshots in time and are therefore not precisely comparable over time although the large sample size, covering around 30% to 40% of all rentals in the UK, ensures that the time series is useful”.

In May 2024, the proportion was 28.2% that month compared with 25.2% in May 2021. That estimate was an average, which, as ever, will mask variations, some of which will involve individual renters paying a higher proportion. It will exclude potential renters ‘priced out’.   

Summary and policy implications

This section identified a range of issues relating to the reporting of house prices and their affordability measures. It advised that attention should focus on the broad direction of direction and its regional/area differentiation rather than on short term headlines and precise figures, which, are always averages, if not estimates.

The assessment of affordability is particularly problematic. One common yardstick is that the maximum rent or mortgage payment should not exceed a third of after-tax household income or 40% of pre-tax income, as was employed above.

Some measures include other housing costs, such as utility and heating costs; and others don’t, and sometimes exclude housing benefit (HB) claimants, and 40% or whatever of household income, however, defined, differs in impact with income.

A feature common to both buying with a mortgage and renting affordability metrics is that taken across the board is that they can suggest affordability.

That conclusion is bedevilled, however, by the reality of the ‘priced out’ generation and ballooning levels of homelessness, often linked to insecurity and lack of access to, and unaffordable rents prevailing in the PRS, relative to individual circumstances.

Those able to buy, and to a much lesser extent, to rent in the PRS, are, by definition, self-selecting, insofar that they have managed to surmount the hurdles to entry, whether through their individual household circumstances and/or increasingly through being able to access family financial support.

The clear ensuing implication that follows, of course, as is well known and incessantly advocated, is a greatly expanded supply of affordable housing, especially let at SR levels.  

True, but this website worries that this is unlikely to be achieved anytime soon for a range of reasons touched on throughout this post and is easy to proclaim rather than activate and that in the unbounded meantime more extensive, granular, and concerted measures are needed.

It also rather suggests a bifurcated housing system with SR catering for those unable to purchase and able to qualify for it, with the remaining majority left to access the private speculative market.

In the absence of wider strategic cross tenure reform, the housing double binds that so mark the English housing system will continue to get tighter.

To take just one salient example, any dependence by the new Labour government on the private sector to achieve its delivery target, presupposes rising house prices to deliver required profit incentives and to maintain S106 cross subsidies within a constrained fiscal environment.

But that outcome will itself undermine access and affordability, especially to those wishing to become homeowners but with moderate and less secure incomes, while bestowing continuing greater wealth (and ability to assist offspring with their purchases on established owners, especially in high value/need areas: a self-perpetuating process, entrenching societal inequalities.