Social housing sales and demolitions 2023-24: Net supply of social housing for rent

Q: 1. Headline figures

In the year ending 31 March 2024: +25,830 was the estimated net change in the stock of social housing for rent. -650 homes was the estimated net change in social rent stock, driven by sales of local authority stock. This is in contrast to the observed gain of 696 homes in the social rent stock of registered providers over the same period, as published by the Regulator of Social Housing. However, this is not directly comparable for reasons explained later in this bulletin . +25,240 was the estimated net change of affordable rent stock (including London affordable rent). 77% of all losses of social housing for rent were to social rent, increasing to 96% when restricted only to sales.

Question 1

1. Headline figures

Accepted Answer

In the year ending 31 March 2024: +25,830 was the estimated net change in the stock of social housing for rent.

-650 homes was the estimated net change in social rent stock, driven by sales of local authority stock. This is in contrast to the observed gain of 696 homes in the social rent stock of registered providers over the same period, as published by the Regulator of Social Housing. However, this is not directly comparable for reasons explained later in this bulletin.

+25,240 was the estimated net change of affordable rent stock (including London affordable rent).

77% of all losses of social housing for rent were to social rent, increasing to 96% when restricted only to sales.

Question 2

2. Introduction

Accepted Answer

This bulletin presents gains and losses of social housing for rent in England. It is split into two sections.

The first section presents accredited official statistics comparing over time the number of sales and demolitions of social housing for rent with new supply of these homes provided by local authorities and private registered providers. This provides a time series back to 1997-98, and figures down to the local authority level, but it does not represent a complete measure of net supply of social housing for rent because some small categories of losses are not covered (but all categories of new supply are included).

The second section presents official statistics in development ^{[footnote 3]} estimating the net supply of social housing for rent for the two years up to 31 March 2024. These estimates show the full profile of gains and losses by provider and tenure, but they cover a shorter period and fewer providers, rely on newly developed data sources, and are available at England level only. They also show some differences when compared with the more direct measure of the year-on-year change in stock of social housing for rent. The reasons for these differences are discussed later in this bulletin. The methodology in this second section differs from previous years’ releases to make use of more detailed data now available for local authorities. These statistics have been developed to meet demand from users for a net measure of the supply of social housing for rent, including a request from the July 2020 recommendation by the Housing Communities and Local Government select committee.

Question 3

3. Comparing sales and demolitions to new supply

Accepted Answer

This section compares the number of sales and demolitions of social housing for rent with the new supply of these homes provided by local authorities and private registered providers ^{[footnote 4]}. This data is presented as a time series from 1997-98, although it should only be used for comparison of general trends. For official statistics in development which calculate net figures for the supply of social housing for rent since 2022-23 broken down by tenure, please see the next section.

While sales and demolitions are the two main sources of losses to social housing, the data presented in this section does not account for all losses. There are a few reasons for this:

the local authority demolitions data only covers local authorities with a Housing Revenue Account (those that own at least 200 homes). These local authorities own more than 99% of local authority housing for rent in England ^{[footnote 5]}.

the sales and demolitions data only covers private registered providers that own at least 1,000 homes (the Regulator of Social Housing’s definition of a large provider). These providers own 96% of private registered provider housing for rent in England ^{[footnote 6]}.

this data does not cover losses through conversions and change of tenure, which we estimate to be 25% of losses of social housing for rent in 2023-24 ^{[footnote 7]}.

this data does not cover non-registered providers or where the provider was unknown. We expect these to be a relatively small number.

In 2023-24, local authorities delivered 8,295 new units of social housing for rent, while private registered providers delivered 30,149, representing a combined total of 38,444 new social housing for rent ^{[footnote 8]}. Meanwhile, there were 17,589 sales and demolitions of social housing for rent this year.

The chart below compares sales and demolitions with new supply of social housing for rent since 1997-98. Between 1997-98 and 2007-08, the combined sales and demolitions were consistently higher than new supply of social housing for rent. By contrast, in all but 2 of the 16 years from 2008-09 to 2023-24, new supply of social housing for rent has exceeded losses through sales and demolitions.

Figure 4.1 Comparison of sales and demolitions to new supply of social housing for rent, England 1 April 1997 to 31 March 2024

The data for gains in this chart is available in the Affordable Housing Supply statistics open data, while the data for losses is available in the Social Housing Sales and Demolitions open data.

The change in the pattern was due to several factors, including the sharp decrease in sales between 2003-04 and 2008-09, the general decrease in the number of demolitions since 2001-02, and the investments through different affordable homes programmes since 2008. In particular, the peak in completions in 2014-15 can be explained by a surge in delivery at the close of the 2014-15 Affordable Homes Programme.

This is also consistent with trends in stock estimates presented in Live Table 104 which show an increase in the combined local authority and private registered provider stock since 2008. Estimates in this table follow the Census definition of “dwelling” and include an adjustment for bedspaces and non-self-contained dwellings. Therefore they do not exactly match data on the number of units (which include bedspaces as well as dwellings) available for affordable or social housing.

Question 4

4. Net supply of social housing for rent

Accepted Answer

4.1 Background

This section presents official statistics in development ^{[footnote 3]} which estimate the net supply of social housing for rent at various tenures by comparing the inflows and outflows of stock in England. Users should be aware that official statistics in development may have a higher degree of uncertainty than accredited official statistics.

The most direct measure of the change in the stock of social housing for rent can be calculated by comparing private registered provider and local authority housing stock over time ^{[footnote 1]}, as this covers all local authorities and private registered providers. However, this measure lacks information on the constituent gains and losses which drive the change between annual stock figures.

We have therefore collected new data and further developed a methodology to estimate the net change in social housing for rent by looking at the profile of gains and losses at each tenure. This is to meet a recommendation from the July 2020 Housing, Communities and Local Government Select Committee, as well as demand for a net measure from other users.

However, the coverage of these estimates of net supply is slightly lower than the coverage of the figures reported by the Regulator of Social Housing on the change in stock of social housing for rent. The estimates of net supply presented here cover only local authorities with a Housing Revenue Account (those that own at least 200 homes) and private registered providers that own at least 1,000 homes (the Regulator of Social Housing’s definition of a large provider). We estimate these providers own 97% of social housing for rent in England ^{[footnote 2]}. Later in this section we compare and explain the differences between these two measures of the change in stock of social housing for rent.

The estimates of net supply in this section are calculated by combining data on gains and losses of social housing for rent from the Local Authority Housing Statistics and the Statistical Data Return collections.

Social housing for rent is defined in section 69 of the Housing and Regeneration Act 2008, where it is also referred to as low cost rental accommodation. For the purpose of these statistics, it includes homes owned by local authorities and private registered providers at the following tenures: social rent, affordable rent (including London affordable rent), and intermediate rent. For definitions of these tenures, please refer to the housing statistics glossary.

As these are official statistics in development, we are particularly keen to receive feedback via housing.statistics@communities.gov.uk on whether readers find them relevant, valuable, or whether there are any other comments.

In 2023-24, the estimated net increase in social housing for rent was 25,830, compared with a net increase of 21,140 in the previous year. By provider, social housing for rent owned by local authorities saw a net increase of around 240 units in 2023-24, while the rental stock owned by private registered providers increased by 25,590 units.

The graph below presents these estimates of net supply of social housing for rent (for the social rent and affordable rent tenures separately) in England for 2022-23 and 2023-24 by provider ^{[footnote 7]}.

Figure 4.2 Net supply of housing for social rent and affordable rent by provider, 2022-23 to 2023-24

The data used for the graph is available alongside this release.

In 2023-24, there was an estimated net loss of 650 homes for social rent, with just under 4,000 social rent gains for private registered providers offset by more than 4,600 social rent losses for local authorities. Overall, the net loss of social rent in 2023-24 was 84% lower than the net loss of 4,060 in 2022-23.

But there were considerable gains in homes for affordable rent, with a net increase of 25,240 in 2023-24, driven by 20,410 gains in affordable rent homes owned by private registered providers. Meanwhile, local authorities saw a net increase of 4,830 homes for affordable rent in 2023-24. Overall, the net increase of affordable rent in 2023-24 was 6% higher than the net increase of 23,870 in 2022-23.

4.3 Comparisons with other available data

Another way of measuring the change in stock of social housing for rent is to compare the stock held by local authorities and private registered providers from one year to the next, as published by the Regulator of Social Housing ^{[footnote 1]}. This method shows there were an additional 25,450 units of social housing for rent at 31 March 2024 compared with the previous year. This is around 380 units lower than the overall estimate of net supply presented above.

There are several definitional differences between these two measures which explains this:

The figure for private registered provider stock published by the Regulator of Social Housing includes both small and large providers (those that own more or less than 1,000 stock), whereas the data on gains and losses in the stock balance sheet of the Statistical Data Return additional tables, which informs our estimate of net supply, excludes small providers ^{[footnote 9]} ^{[footnote 10]}. Large private registered providers own 96% of social housing for rent in England ^{[footnote 6]}.

All local authority housing stock is covered by the Regulator for Social Housing’s Local Authority Data Return. Whereas data from the Section K of the Local Authority Housing Statistics used to calculate our estimate of net supply is only mandatory for local authorities with a Housing Revenue Account (those that own at least 200 stock). These local authorities own more than 99% of local authority housing for rent in England ^{[footnote 5]}. However, this data was not returned by all these local authorities in full ^{[footnote 11]}.

There may also be differences due to the timing of gains and losses recorded in different data sources, particularly with transfers of stock from private registered providers to local authorities (and vice versa) near the start or the end of the financial year. For instance, we are aware of around 200 transfers from private registered providers to local authorities being recorded as losses in 2022-23, which seem to only appear recorded as gains for local authorities in the following year.

It is possible to account for some of the differences in coverage by restricting the Regulator of Social Housing’s data to local authorities with a Housing Revenue Account and by using stock data for large private registered providers only. Any remaining differences are likely due to the timing issue discussed above, as well as minor data quality issues discussed further in the technical notes accompanying this release. After accounting for differences in coverage, the difference between the two measures are as follows:

Overall, the difference between measures increases slightly from 350 to 620 units, as shown in Table 3 of the technical notes accompanying this release.
For social rent, both measures show a net decrease, of 650 from our estimate of net supply and 1,150 in the stock figures from the Regulator of Social Housing (see Table 4 of the technical notes). This is in contrast with the increase of 696 units in the overall social rent stock when directly comparing the Regulator of Social Housing’s figures across years. However, these figures are not directly comparable because, in addition to the definitional differences discussed above, the Regulator of Social Housing stock figures for social rent also include intermediate rent units in some tables, although other tables indicate these are a relatively low number.
For affordable rent, both measures show a similar net increase. Our estimate shows a net increase of 25,240 affordable rent homes in 2023-24, compared to 25,600 from the Regulator of Social Housing’s data (see Table 5 of the technical notes) after accounting for differences in coverage between the two measures. The Regulator of Social Housing’s unadjusted comparison of stock across years showed an increase of 24,760 units.

This section presents estimates of net supply of social housing for rent broken down by tenure, provider, and the constituent types of losses and gains.

The chart below shows a breakdown of gains and losses of housing for social rent and affordable rent for local authorities and private registered providers combined.

Figure 4.3 Net supply of housing for social rent and affordable rent by type of gain or loss, 2022-23 to 2023-24

The data used for the graph is available alongside this release.

This chart shows that losses to social rent stock consist mainly of sales, with 12,480 social rent sales taking place in 2023-24, making up 61% of social rent losses. By contrast, the 400 sales of affordable rent stock were considerably lower, making up only 9% of all losses to affordable rent.

Gains in affordable rent were mostly driven by new build, with 24,940 new build homes for affordable rent in 2023-24, making up 84% of all affordable rent gains. This is 9% higher when compared with 2022-23. By contrast, 8,610 social rent new builds represented only 43% of all social rent gains in 2023-24.

There were 4,080 affordable rent acquisitions recorded in 2023-24, a 65% increase compared with 2022-23. For social rent, 4,020 acquisitions made up 20% of gains, while other gains (due to conversions, transfers in stock, or changes in tenure) made up 37%.

The chart below shows a breakdown of gains and losses of housing for social rent and affordable rent in the latest year, 2023-24, this time with local authorities and private registered providers shown separately.

Figure 4.4 Net supply of housing for social rent and affordable rent by provider and type of gain or loss, 2023-24

The data used for the graph is available alongside this release.

Most gains to affordable rent were new build, with a smaller number of acquisitions. But gains to social rent stock, especially that owned by private registered providers saw a greater mix of new builds, acquisitions, changes of tenure and conversions.

Most of the 26,780 stock losses were of social rent stock, representing 77% of all losses of social housing for rent. Social rent accounts for 96% of all sales. 40% of all social rent stock losses were owned by local authorities and 60% owned by private registered providers.

An estimated 99% of losses of local authority stock were social rent properties, similar to the 97% of stock held by local authorities that is for social rent in 2023-24. By contrast, just 67% of private registered provider losses were social rent properties, a substantially lower proportion than the 87% of stock held by private registered that was for social rent in 2023-24 ^{[footnote 12]}.

Private registered providers saw 5,600 gains of social rent through changes in tenure during 2023-24, which was partly offset by 3,730 losses of affordable rent homes through changes in tenure over the same period.

The data used for our estimates is available at local authority level for stock owned by local authorities ^{[footnote 13]}. Data for private registered providers is not currently available by geography. We will continue to explore ways to obtain this data in future.

Available in Table 1.4 of the Regulator of Social Housing’s ‘Registered provider social housing stock and rents in England’ collection. ↩ ↩² ↩³
Calculated by dividing the sum of large private registered provider rental stock in Table 3.1c of the Statistical Data Return additional tables and local authority rental stock in the stock information sheet of the Local Authority Data Return for local authorities with a Housing Revenue Account by the total rental stock owned by registered providers (available in Table 1.4 of the Regulator of Social Housing’s ‘Registered provider social housing stock and rents in England’ collection). ↩ ↩²
Official statistics in development are official statistics that are undergoing a development; they may be new or existing statistics, and will be tested with users, in line with the standards of trustworthiness, quality, and value in the Code of Practice for Statistics. ↩ ↩²
The figures for new supply are a subset of the total presented in Live Table 1000, and include not just new build, but also acquisitions of stock that were not previously used as social housing for rent. ↩
Calculated by dividing local authority rental stock in the stock information sheet of the Local Authority Data Return for those local authorities with a Housing Revenue Account by the total rental stock owned by local authorities (available in Table 1.1 of the Local Authority Data Return additional tables). ↩ ↩²
See Table 3.1c of the Statistical Data Return additional tables. ↩ ↩²
For the purposes of this release, London affordable rent is counted as affordable rent. Intermediate rent accounts for a relatively small percentage of stock and is therefore not commented on in this bulletin, but [data on gains and losses of intermediate rental stock is available alongside this release. ↩ ↩²
See Affordable Housing Supply Live Table 1013. ↩
The stock balance sheet only covers private registered providers owning at least 1,000 social housing units who are registered on 31 March each year. This means that stock flow data for smaller providers, or those who have developed within year but subsequently deregistered, will not be captured. Additionally, as providers only report activities which they have undertaken, units transferred into a provider may include development activity carried out by precursor bodies prior to a merger (e.g. Provider A develops 200 units prior to merging with into Provider B and transferring all of its stock. Provider B does not record this as development activity because it did not carry out the work and the 200 units would be recorded under the total units transferred into it). Units developed by subsidiaries or third parties that have ownership moved to a large private registered provider may also sometimes be recorded as something other than new build activity depending on the legal form the change of ownership took. ↩
To calculate our measure of net supply, we also removed from the stock balance sheet 5,810 “other losses” of social housing for rent reported by Sparrow Shared Ownership Limited within the Regulator of Social Housing’s Statistical Data Return. These losses were confirmed as transfers to other private registered providers, and therefore do not represent a net loss of social housing for rent at a sector level. Regulator of Social Housing analysts advised us that, after removing these losses, there are a small number of additional cases (totalling fewer than 500 units each year at a sector level) where transfers to other providers are recorded as “other losses”. For more information on these cases, see page 16 of the briefing note “Private registered provider social housing stock in England - sector characteristics and stock movement”. From 2024-25, it is expected that the data collection will be expanded to allow for such cases to be directly identified. ↩
Eight local authorities returned non-responses to some of the questions in this section, and a further three submitted zero losses or gains in Section K which contradicted responses in other sections. The questions were made mandatory for the first time in the 2022-23 collection and we expect completion rates will improve as the new questions become more established. ↩
Stock figures from Table 1.1 in the Regulator of Social Housing’s Local Authority Data Return additional tables and Table 1.7a in the Regulator of Social Housing’s Statistical Data Return additional tables. ↩
See Section K of the Local Authority Housing Statistics collection. ↩