Impact assessment - proposed change to collection of non-self-contained units in LADR (accessible version)
Updated 17 February 2022
Applies to England
© Crown copyright 2022
This publication is licensed under the terms of the Open Government Licence v3.0 except where otherwise stated. To view this licence, visit nationalarchives.gov.uk/doc/open-government-licence/version/3 or write to the Information Policy Team, The National Archives, Kew, London TW9 4DU, or email: psi@nationalarchives.gov.uk.
Where we have identified any third party copyright information you will need to obtain permission from the copyright holders concerned.
This publication is available at https://www.gov.uk/government/publications/local-authority-data-return-proposal-to-cease-dwelling-equivalent-count/impact-assessment-proposed-change-to-collection-of-non-self-contained-units-in-ladr
1. Introduction
The Regulator of Social Housing are proposing to end the collection of non-self-contained units as clustered dwelling equivalents in the Local Authority Data Return collection. From 1 April 2022 the RSH will collect data on all non-self-contained units by counting these as bed spaces.
This document provides supplementary information on the background and impact of the proposal, along with details around the complexity of the current methodology, to allow data users to provide informed feedback and views on this proposal.
Background
When the LADR collection was created in 2019, it took questions from the existing Local Authority Housing Statistic and replicated these in order to maintain the timeseries for data publication. Since then, we have sought to ensure and improve the quality of the data returned to us through the LADR. An area which we are aware causes significant confusion, and which has been reported as having resulted in errors in previous years’ submissions, is the collection of the dwelling equivalent count of non-self-contained units.
Our concerns over the continued capture of the dwelling equivalent figures are outlined in the rest of this document, falling into the following areas:
Methodology concerns
The methodology employed to calculate the dwelling equivalent figure creates a value which could be representative of between 1 and 6 bed spaces for each dwelling equivalent reported.
More details are provided in section 2.
Data quality concerns
During our LADR cleaning process we identified a number of data errors within the dwelling equivalent count. These stem from the complexities in the methodology for calculating dwelling equivalents and have led providers to report inaccurate data over a number of years (including within previous collections (such as LAHS)). The main type of error has been the reporting of individual bed spaces instead of employing the methodology to create the dwelling equivalent value.
More details are provided in section 3.
Burden concerns
As providers report bed spaces in the rent data sections of LADR already, the proposed change is expected to reduce the burden on providers, as it removes the need for them to calculate the dwelling equivalent count and to respond to queries resulting from the cleaning of these data.
More details are provided in section 4.
Impact
There are just 4,537 non-self-contained units (when recorded as bed spaces) owned by 64 (out of 213) local authority registered providers (LARPs). This is equivalent to 0.3% of the total LARP owned social stock. The proposed change impacts on only these units, increasing the count from 1,690 dwelling equivalent figures to 4,537 bed spaces. This impact at a sector level represents a value increase of just 0.2%.
Given the issues with historic data quality (see section 3), the analysis which provides an estimated multiplier for estimating dwelling equivalent figures in the future (see section 5) and the limited use of the data and its comparability (see section 6) we feel that the dwelling equivalent count does not provide significant insight into the number of non-self-contained units and leads to issues with burden and future data quality.
Proposal
From 1 April 2022 we intend to cease the dwelling equivalent collection and collect non-self-contained units as bed spaces only across the LADR. This will improve the accuracy of the data (providing actual counts of the number of bed spaces owned by local authorities), while ensuring comparability with our private registered provider statistics.
This will mean no dwelling equivalent data will be published after this publication.
If this change or the loss of continuation of the dwelling equivalent timeseries poses significant issues with your analysis then please provide details of the concerns to us as soon as possible and by 31 January 2022 at the latest.
You can contact us by email at enquiries@rsh.gov.uk, or by calling 0300 1245 225.
You can also write to us at:
Regulator of Social Housing
Level 2
7-8 Wellington Place
Leeds
LS1 4AP
2. Current clustering methodology
The dwelling equivalent count is based on a clustering of non-self-contained units to provide an estimate of the number of properties in which bed spaces are present. It is not an accurate count of these properties, as it employs a maximum threshold for the number of bed spaces clustered together (which is different depending on whether the bed space is in a hostel or in a house of multiple occupation). The calculation is complicated and results in a figure where between 1 and 6 bed spaces (in hostels and Houses of Multiple Occupation) could combine to a single dwelling equivalent (i.e. 100 dwelling equivalents could represent anywhere between 100 and 600 bed spaces).
Further breakdown
The dwelling equivalent of multi-occupied dwellings has two components – a figure derived from bed spaces in hostels and a figure derived from ‘clusters’ in HMOs. These equivalents should be calculated as follows:
-
Hostels – for hostels the dwelling equivalent is derived from the number of groups of three bed spaces in the hostel. Divide the total number of bed spaces in each hostel at 31 March by three, with any balance counting as one dwelling.
-
HMO – In HMOs with shared facilities, the dwelling equivalent is derived from the number of ‘clusters’ in the dwelling. A ‘cluster’ is a group of rooms in a HMO serving as separate accommodation for two or more persons but sharing common kitchen, bathroom and lavatory; where such a dwelling accommodates six of fewer persons, this counts as one cluster; where it accommodates more than six persons, the number of clusters is calculated by dividing the number of persons by six, with any balance counting as one cluster. The figure derived as the number of clusters is the dwellings equivalent for the HMOs.
Impact of clustering
Employing the methodology outlined above in order to report dwelling equivalents means that every single reported dwelling equivalent could represent between 1 and 6 bed spaces.
We feel that the ambiguity in this count, with the potential for 1,000 dwelling equivalents to represent between 1,000 and 6,000 units has limited use for users.
Worked example
The local authority has 5 HMOs. Two have ten bedrooms, one has eight bedrooms, one has six bedrooms and one has four bedrooms (with each bedroom let to a different tenant).
The local authority also has 5 hostels. One with 20 bed spaces, one with 16 bed spaces, one with 12 bed spaces, one with 7 bed spaces and one with two bed spaces.
HMO Cluster element calculation:
| Number of units per HMO | Maximum number of units per cluster | Number of clusters (equivalent to dwelling equivalent | Total HMO Clusters (equivalent to dwelling equivalent - rounded to full clusters) | |||
|---|---|---|---|---|---|---|
| 10 | / | 6 | = | 1.67 | 2 | |
| 10 | / | 6 | = | 1.67 | 2 | |
| 8 | / | 6 | = | 1.33 | 2 | |
| 6 | / | 6 | = | 1.00 | 1 | |
| 4 | / | 6 | = | 0.67 | 1 | |
| Total | 8 |
Hostel element calculation:
| Number of bed spaces per hostel | Maximum number of bed spaces | Number of dwelling equivalents | Total hostel dwelling equivalents (rounded to full DE) | |||
|---|---|---|---|---|---|---|
| 20 | / | 3 | = | 6.67 | 7 | |
| 16 | / | 3 | = | 5.33 | 6 | |
| 12 | / | 3 | = | 4.00 | 4 | |
| 7 | / | 3 | = | 2.33 | 3 | |
| 2 | / | 3 | = | 0.67 | 1 | |
| Total | 21 |
Total dwelling equivalent units is the combined total of HMO clusters and Hostel dwelling equivalents.
In this example 8 + 21 = 29
In this example, the dwelling equivalent count of 29 could represent anywhere between 29 and 174 individual bed spaces, but actually represents 95 bed spaces.
3. Data quality
We believe that it is the dwelling equivalent count which causes the most frequent significant data errors made by providers and that this appears to have been a historical issue that predates the LADR return – see examples in Annex B.
When compared with the 2019 LADR pilot data, and dwelling equivalent reporting in the 2019 LAHS, 11 out of the 87 participating LARPs (13%) appear to have made errors when calculating dwelling equivalents (for example reporting the same number of dwelling equivalents and bed spaces). In all cases investigated this was reported as an error in calculating dwelling equivalents and not an error in counting bed spaces.
In the 2020 LADR, where the return rate was 100%, we have identified 27 out of 197 LARPs (14%) where data errors appear to have been made when calculating dwelling equivalents (as above).
In the 2021 LADR collection process, additional time was spent by analysts at RSH identifying these types of errors. Within the overall LADR cleaning process, queries on dwelling equivalent calculations were raised with 29 out of 213 providers (14%) (or 41% of the 64 LARPs reporting non-self-contained units).
Most of these queries were resolved and the final suspected error rate was reduced to 3 out of 213 LARPs (1%). However, pursuing the resolution of these errors this came at a significant resource cost to RSH. The three LARPs which have suspected errors all confirmed they were confident in their reporting. They reported that their calculations were consistent with the guidance for calculating dwelling equivalent clusters (found in the LAHS guidance documents issued each year). RSH have attempted to clarify the LADR guidance and provided some examples, but it is possible that differences in interpretation may remain with some LARPs.
As a result, we accepted and published data from all LARPs in 2021, even where errors in the dwelling equivalent calculation are suspected. Users should note that this level of error was equivalent to 81 non-self-contained bed spaces, with the methodology used in the calculation likely to be the cause.
4. Impact of proposal on Local Authority Registered Providers
Based on the information reported in 2021 LADR, just 64 LARPs out 213 will be affected by the proposed change. This represents 30% of the overall LARP population.
| Region | Total number of LARPs | Number of LARPs reporting non-self-contained stock | Percentage of LARPs reporting non-self-contained stock |
|---|---|---|---|
| East Midlands | 28 | 3 | 10.7% |
| East of England | 31 | 15 | 48.4% |
| London | 32 | 13 | 40.6% |
| North East | 10 | 2 | 20.0% |
| North West | 19 | 7 | 36.8% |
| South East | 41 | 11 | 26.8% |
| South West | 20 | 5 | 25.0% |
| West Midlands | 17 | 2 | 11.8% |
| Yorkshire and the Humber | 15 | 6 | 40.0% |
| England | 213 | 64 | 30.0% |
The majority of affected LARPs are located in the London, East of England and South East regions. These regions contain 61% of the LARPs who own non-self-contained units (39 of the total 64).
Data provider burden
As providers report bed spaces in the rent data sections of LADR already, the proposed change is expected to reduce the burden on providers, as it removes the need for them to calculate the dwelling equivalent clusters. This was something the Central and Local Information Partnership (CLIP-H) group has previously asked us to consider, during the original move of questions form LAHS to LADR.
In addition, significant time was spent by analysts at RSH identifying likely errors, when reviewing bed spaces versus dwelling equivalent counts in the 2021 return and led to providers being contacted further to review these data.
Within the overall LADR cleaning process in 2021, queries on dwelling equivalent calculations were raised with 29 providers (14% of total population).
5. Impact of proposal on RSH published data and national statistics
During the data cleaning period for the 2021 LADR dataset, emphasis was placed on ensuring the reporting of non-self-contained units was accurate. Additional checks were undertaken to allow RSH analysis to highlight when dwelling equivalent clusters may have been calculated incorrectly. Any suspected errors were raised with providers and corrected where possible prior to sign off. This additional work has allowed us to estimate the impact of removing the requirement to calculate dwelling equivalent clusters on the published figures with a high degree of certainty.
Impact on RSH publication
The data presented in the 2021 publication focuses on bed space data for non-self-contained units, although dwelling equivalent data are provided in our additional tables. The tables below show the impact of this change on our 2021 data.
Difference between dwelling equivalent count and bed spaces count (2021)
| Region | General needs (social rent) | Supported housing (social rent) | Affordable rent general needs | Affordable rent supported housing | Total |
|---|---|---|---|---|---|
| East Midlands | +3 | +42 | 0 | +7 | +52 |
| East of England | +147 | +160 | +18 | 0 | +325 |
| London | +1,047 | +465 | 0 | 0 | +1,512 |
| North East | 0 | +15 | 0 | 0 | +15 |
| North West | +2 | +247 | 0 | +2 | +251 |
| South East | +69 | +37 | +9 | + 79 | +194 |
| South West [footnote 1] | +12 | +18 | +12 | +19 | +60 |
| West Midlands | +250 | 0 | 0 | +8 | +258 |
| Yorkshire and the Humber | +53 | +8 | +110 | +9 | +180 |
| England | +1,583 | +992 | +149 | +124 | +2,847 |
The move to the bed space count adds a value of 2,847 to the overall social stock owned. The table below shows the impact, by region, of this change.
Percentage difference in total stock including dwelling equivalent and including bed spaces (2021)
| Region | General needs (social rent) | Supported housing (social rent) | Affordable rent general needs | Affordable rent supported housing | Low cost home ownership | Total |
|---|---|---|---|---|---|---|
| East Midlands | 0.0% | 0.2% | 0.0% | 0.9% | 0.0% | 0.0% |
| East of England | 0.1% | 1.0% | 0.6% | 0.0% | 0.0% | 0.2% |
| London | 0.3% | 3.5% | 0.0% | 0.0% | 0.0% | 0.4% |
| North East | 0.0% | 0.2% | 0.0% | 0.0% | 0.0% | 0.0% |
| North West | 0.0% | 3.9% | 0.0% | 0.6% | 0.0% | 0.3% |
| South East | 0.0% | 0.3% | 0.3% | 12.5% | 0.0% | 0.1% |
| South West | 0.0% | 0.2% | 0.9% | 7.9% | 0.0% | 0.1% |
| West Midlands | 0.1% | 0.2% | 3.1% | 5.5% | 0.0% | 0.1% |
| Yorkshire and the Humber | 0.0% | 0.2% | 3.1% | 5.5% | 0.0% | 0.1% |
| England | 0.1% | 1.0% | 0.6% | 3.3% | 0.0% | 0.2% |
As non-self-contained units make up just 0.3% of the total owned social stock, the impact of the change on total social stock is very small at a regional level.
The reported value of all units across England is just 2,847 (0.2%) greater when using a count of bed spaces than when using a count of dwelling equivalents within the total stock count. The impact is greater on supported housing than on general needs (1.0% compared to 0.1%). This is due to there being proportionally more supported housing provided as non-self-contained units (1.6% compared to 0.2% of general needs).
Users should note that Affordable Rent supported housing shows some relatively large percentage increases in some regions but that this is due to very small numbers of these units reported overall. For example, the South East shows a 12.5% increase but represents an increase in reported value of just 79 (from a total of 633 (when including dwelling equivalents) to 712 - when including bed spaces).
We feel that this shift is minimal and that the move to only bed space level data will continue to provide high quality, comparable data in future years whilst reducing the burden on our data suppliers (LARPs).
Impact on national statistics
The RSH has already moved away from presenting dwelling equivalent data in the local authority registered provider stock and rents statistics, focusing on the more granular bed space data collected in the LADR return.
However, we recognise that historic timeseries information is based on the dwelling equivalent count due to bed space information only being introduced in 2020.
We have recognised the impact of this and detailed it within the published statistics in 2021.
Difference between dwelling equivalent and bed spaces when viewed as component of total social stock (2021)
| Region | Total social stock –with non-self-contained units as dwelling equivalents | Total social stock –with non-self-contained units as bed spaces | Difference | Percentage difference |
|---|---|---|---|---|
| East Midlands | 173,111 | 173,163 | +52 | +0.0% |
| East of England | 155,404 | 155,729 | +325 | +0.2% |
| London | 388,645 | 390,157 | +1,512 | +0.4% |
| North East | 88,884 | 88,899 | +15 | +0.0% |
| North West | 81,905 | 82,156 | +251 | +0.3% |
| South East | 170,569 | 170,763 | +194 | +0.1% |
| South West | 93,819 | 93,880 | +60 | +0.1% |
| West Midlands | 195,835 | 196, 093 | +258 | +0.1% |
| Yorkshire and the Humber | 225,350 | 225\530 | +180 | +0.1% |
| England | 1,573,518 | 1,576,370 | +2,847 [footnote 2] | +0.2% |
We feel that there is little disruption to any timeseries where historical dwelling equivalent figures have been used in previous years due to the small impact this shift to bed space count has on overall stock numbers. This is further illustrated in Table 1.1a of the LARP 2021 Additional Tables document (available as part of the 2021 statistical publication).
A table showing total social stock counts and percentage difference for all individual LARPs reporting non-self-contained stock can be found in Annex A.
Potential future estimation of dwelling equivalent figures
By comparing dwelling equivalent and bed space data from the 2021 LADR collection (where significant time was spent reviewing and working with providers to confirm the dwelling equivalent data) we were able to estimate the average number of bed spaces which are clustered together to make a dwelling equivalent.
This multiplier is different for each region, with between 2.2 and 3.1 bed spaces forming a dwelling equivalent in each region. The overall average across England is 2.7 bed spaces to each dwelling equivalent.
Dwelling equivalent to bed space multiplier (2021 data)
| Region | Total dwellings equivalent to HMOs/Hostels | Total bed spaces | Difference | Multiplier |
|---|---|---|---|---|
| East Midlands | 39 | 91 | +52 | 2.3 |
| East of England | 184 | 509 | +325 | 2.8 |
| London | 845 | 2,357 | +1,512 | 2.8 |
| North East | 7 | 22 | +15 | 3.1 |
| North West | 184 | 435 | +251 | 2.4 |
| South East | 98 | 292 | +194 | 3.0 |
| South West | 47 | 107 | +60 | 2.3 |
| West Midlands | 134 | 392 | +258 | 2.9 |
| Yorkshire and the Humber | 152 | 332 | +180 | 2.2 |
| England | 1,690 | 4,537 | +2,847 | 2.7 |
We can take data from the LADR 2020 return to demonstrate using the above multiplier to calculate an estimate of the dwelling equivalent count [footnote 3].
| Total dwellings equivalent to reported in LADR 2020 | Total bed spaces reported in LADR 2020 |
|---|---|
| 1,507 | 4,202 |
If the multiplier of 2.7 is used to estimate the number of dwellings equivalent then, based on the number of bed spaces reported in 2020, we obtain an estimated figure that is only 49 units different from the true figure captured in the return in 2020.
In this example 4,202 ÷ 2.7 = 1,556 (rounded to nearest unit)
Based on the likely historical inaccuracy of this method of counting non-self-contained units (as outlined in section 3), we feel that the accuracy of this dataset means that if a dwelling equivalent figure was needed by any user in future years, this could be estimated using the multipliers above. If further accuracy is needed, it is recommended that the bed space count is used, as this is comparable with our private registered provider dataset.
A table showing individual unit counts and dwelling equivalent-to-bed space multipliers for all LARPs reporting non-self-contained stock can be found in Annex A.
6. Impact on other publications
To our current knowledge no external publications rely directly on the LADR dataset. However, we are seeking feedback from any users of this data which will improve our knowledge and understanding of how the LADR data, and specifically this dwelling equivalent count, is used.
We are aware that the figures from the LADR are used by the Department for Levelling Up, Housing and Communities to supplement headline figures presented in some live tables.
DLUHC publish several dwelling stock live tables that use LADR to supplement headline figures. We have identified those below where DLUHC use figures from the Local Authority Housing Statistics for overall stock totals, and LADR data for further breakdown by stock category (e.g. social or Affordable Rent) or by unit size in the event that this is requested. These data are not published in a structured way (outside of the LARP statistics published by us) but are used on an ad hoc basis.
The live tables where we believe LADR data are used as supplementary information are:
- Table 100: number of dwellings by tenure and district, England
- Table 104: by tenure, England (historical series) (and Chart 105)
- Table 109: by tenure and region, from 1991
- Table 116: local authority dwelling stock, by district, England, from 1994
Live tables on dwelling stock are available online from Live tables on dwelling stock (including vacants) - GOV.UK
These tables use the Local Authority Housing Statistics data as their source. We are already aware of a small difference in coverage of stock between the LAHS and LADR collections. This means that even when using current data from the LADR where dwelling equivalents are present a degree of caution should be applied to comparing the two data sets.
For example, the stock totals featured in the above live tables are drawn from the LAHS (field a1a). This question captures the total stock of local authority dwellings within the authority’s geographical area, including those owned by other local authorities. The LADR survey asks the LARP to only report units that they own (both inside and outside the authority’s geographical area), and which meet the definition of social housing (as defined in sections 68 to 71 of the Housing and Regeneration Act 2008).
With a minimal impact on overall stock counts, the potential to estimate dwelling equivalent values (see section 5), the issues with data quality (see section 4) and the limitations in the comparability of the LADR data to LAHS information (above) we feel that the use of the dwelling equivalent figures to supplement the DLUHC figures could be replaced with the bed space information without substantial impact on users.
7. Annex A
Dwelling equivalent and bed space data for all LARPs holding non-self-contained stock – by individual provider (2021)
| LARP name | LARP code | Total dwellings equivalent | Total bed spaces | Multiplier | Total social stock –with non-self-contained units as dwelling equivalents | Total social stock –with non-self-contained units as bed spaces | Difference | Percentage difference |
|---|---|---|---|---|---|---|---|---|
| London Borough of Lewisham | 00AZ | 320 | 928 | 2.9 | 14,082 | 14,690 | +608 | +4.3% |
| London Borough of Lambeth | 00AY | 211 | 590 | 2.8 | 23,952 | 24,331 | +379 | +1.6% |
| Birmingham City Council | 00CN | 129 | 379 | 2.9 | 59,941 | 60,191 | +250 | +0.4% |
| London Borough of Camden Council | 00AG | 82 | 193 | 2.4 | 22,971 | 23,082 | +111 | +0.5% |
| Manchester City Council | 00BN | 77 | 213 | 2.8 | 15,611 | 15,747 | +136 | +0.9% |
| Blackpool Borough Council | 00EY | 62 | 62 | 1.0 | 4,732 | 4,732 | +0 | +0.0% |
| London Borough of Hammersmith and Fulham | 00AN | 61 | 180 | 3.0 | 12,092 | 12,211 | +119 | +1.0% |
| Southend on Sea Borough Council | 00KF | 61 | 123 | 2.0 | 5,981 | 6,043 | +62 | +1.0% |
| London Borough of Hackney | 00AM | 59 | 81 | 1.4 | 21,760 | 21,782 | +22 | +0.1% |
| City of York Council | 00FF | 56 | 168 | 3.0 | 7,583 | 7,695 | +112 | +1.5% |
| London Borough of Wandsworth | 00BJ | 56 | 231 | 4.1 | 17,124 | 17,299 | +175 | +1.0% |
| Epping Forest District Council | 22UH | 44 | 131 | 3.0 | 6,431 | 6,518 | +87 | +1.4% |
| Harrogate Borough Council | 36UD | 43 | 50 | 1.2 | 3,887 | 3,894 | +7 | +0.2% |
| London Borough of Havering Council | 00AR | 38 | 95 | 2.5 | 9,426 | 9,483 | +57 | +0.6% |
| Kingston upon Hull City Council | 00FA | 32 | 75 | 2.3 | 23,462 | 23,505 | +43 | +0.2% |
| Stockport Metropolitan Borough Council | 00BS | 30 | 118 | 3.9 | 11,061 | 11,149 | +88 | +0.8% |
| Leicester City Council | 00FN | 28 | 64 | 2.3 | 19,940 | 19,976 | +36 | +0.2% |
| Portsmouth City Council | 00MR | 23 | 85 | 3.7 | 14,651 | 14,713 | +62 | +0.4% |
| Canterbury City Council | 29UC | 22 | 49 | 2.2 | 5,079 | 5,106 | +27 | +0.5% |
| Dacorum Borough Council | 26UC | 21 | 32 | 1.5 | 10,228 | 10,239 | +11 | +0.1% |
| Dorset Council | 5099 | 19 | 31 | 1.6 | 40 | 52 | +12 | +30.0% |
| New Forest District Council | 24UJ | 19 | 46 | 2.4 | 5,147 | 5,174 | +27 | +0.5% |
| Bristol City Council | 00HB | 15 | 33 | 2.2 | 26,907 | 26,925 | +18 | +0.1% |
| Stevenage Borough Council | 26UH | 13 | 44 | 3.4 | 7,965 | 7,996 | +31 | +0.4% |
| Derby City Council | 00FK | 10 | 26 | 2.6 | 12,724 | 12,740 | +16 | +0.1% |
| Selby District Council | 36UH | 10 | 15 | 1.5 | 3,027 | 3,032 | +5 | +0.2% |
| Teignbridge District Council | 18UH | 10 | 29 | 2.9 | 10 | 29 | +19 | +190.0% |
| Cambridge City Council | 12UB | 9 | 24 | 2.7 | 7,181 | 7,196 | +15 | +0.2% |
| Central Bedfordshire Council | 00KC | 9 | 26 | 2.9 | 5,293 | 5,310 | +17 | +0.3% |
| Cherwell District Council | 38UB | 8 | 33 | 4.1 | 238 | 263 | +25 | +10.5% |
| Harlow District Council | 22UJ | 7 | 31 | 4.4 | 9,160 | 9,184 | +24 | +0.3% |
| London Borough of Greenwich | 00AL | 7 | 19 | 2.7 | 20,411 | 20,423 | +12 | +0.1% |
| South Lakeland District Council | 16UG | 7 | 21 | 3.0 | 7 | 21 | +14 | +200.0% |
| Wokingham Borough Council | 00MF | 7 | 19 | 2.7 | 2,651 | 2,663 | +12 | +0.5% |
| North Tyneside Council | 00CK | 6 | 18 | 3.0 | 14,467 | 14,479 | +12 | +0.1% |
| Ryedale District Council | 36UF | 6 | 19 | 3.2 | 6 | 19 | +13 | +216.7% |
| Wigan Council | 00BW | 6 | 13 | 2.2 | 21,574 | 21,581 | +7 | +0.0% |
| Barnsley Metropolitan Borough Council | 00CC | 5 | 5 | 1.0 | 18,259 | 18,259 | +0 | +0.0% |
| London Borough of Hillingdon | 00AS | 5 | 15 | 3.0 | 10,199 | 10,209 | +10 | +0.1% |
| Shropshire Council | 00GG | 5 | 13 | 2.6 | 4,066 | 4,074 | +8 | +0.2% |
| Tandridge District Council | 43UK | 5 | 15 | 3.0 | 2,580 | 2,590 | +10 | +0.4% |
| West Suffolk Council | 5068 | 5 | 23 | 4.6 | 66 | 84 | +18 | +27.3% |
| Arun District Council | 45UC | 4 | 13 | 3.3 | 3,400 | 3,409 | +9 | +0.3% |
| Babergh District Council | 42UB | 4 | 24 | 6.0 | 3,433 | 3,453 | +20 | +0.6% |
| Reigate and Banstead Borough Council | 5127 | 4 | 11 | 2.8 | 11 | 18 | +7 | +63.6% |
| Thurrock Council | 00KG | 4 | 8 | 2.0 | 9,834 | 9,838 | +4 | +0.0% |
| London Borough of Enfield | 00AK | 3 | 5 | 1.7 | 10,504 | 10,506 | +2 | +0.0% |
| Castle Point Borough Council | 22UE | 2 | 17 | 8.5 | 1,512 | 1,527 | +15 | +1.0% |
| East Devon District Council | 18UB | 2 | 14 | 7.0 | 4,181 | 4,193 | +12 | +0.3% |
| Guildford Borough Council | 43UD | 2 | 5 | 2.5 | 5,256 | 5,259 | +3 | +0.1% |
| Mid Suffolk District Council | 42UE | 2 | 12 | 6.0 | 3,281 | 3,291 | +10 | +0.3% |
| Runnymede Borough Council | 43UG | 2 | 8 | 4.0 | 2,891 | 2,897 | +6 | +0.2% |
| Woking Borough Council | 43UM | 2 | 8 | 4.0 | 3,396 | 3,402 | +6 | +0.2% |
| Ashfield District Council | 37UB | 1 | 1 | 1.0 | 6,635 | 6,635 | +0 | +0.0% |
| Brentwood Borough Council | 22UD | 1 | 6 | 6.0 | 2,449 | 2,454 | +5 | +0.2% |
| East Suffolk Council | 5070 | 1 | 6 | 6.0 | 4,454 | 4,459 | +5 | +0.1% |
| Gateshead Metropolitan Borough Council | 00CH | 1 | 4 | 4.0 | 18,942 | 18,945 | +3 | +0.0% |
| London Borough of Merton | 00BA | 1 | 14 | 14.0 | 1 | 14 | +13 | +1300.0% |
| London Borough of Tower Hamlets | 00BG | 1 | 2 | 2.0 | 11,633 | 11,634 | +1 | +0.0% |
| Norwich City Council | 33UG | 1 | 2 | 2.0 | 14,552 | 14,553 | +1 | +0.0% |
| Ribble Valley Borough Council | 30UL | 1 | 3 | 3.0 | 7 | 9 | +2 | +28.6% |
| Rossendale Borough Council | 30UM | 1 | 5 | 5.0 | 11 | 15 | +4 | +36.4% |
| Royal Borough of Kensington and Chelsea | 00AW | 1 | 4 | 4.0 | 6,705 | 6,708 | +3 | +0.0% |
8. Annex B
Case studies
We have presented a small number of case studies taken from actual local authority data. These examples illustrate the difficulty that organisations have had calculating the dwelling equivalent metric.
To avoid spotlighting individual authorities directly, we have not included organisation names or direct figures in these examples. However, these are based on real data, which has been taken from the LADR 2019 Pilot, 2020 and 2021 datasets. These have been compared with the 2019 LAHS dataset, as this was the last LAHS return to capture a breakdown of stock totals per unit size.
Case A
The LARP reported the same figure for dwellings equivalent and bed spaces, under social rent (general needs) in the LADR 2019 pilot. This 1:1 ratio of dwelling equivalents to bed spaces indicates an error has been made when calculating one of these figures (as multiple bed spaces should be grouped together in a dwelling equivalent figure).
The same figures were reported as the dwelling equivalent total under social rent in the 2019 LAHS dataset – this indicating that the issue is likely to be present in the LAHS data as well. This issue was repeated in the LADR 2020 return, where again the same figure was reported for dwellings equivalent and bed spaces.
In 2021, RSH analysts dedicated additional resources to checking these figures and were able to discuss this directly with the LARP. It was established that the LARP had not been calculating the dwelling equivalent figure due to being unaware that this was required. They had been counting non-self-contained units as individual bed spaces across both returns and had been doing this historically in the LAHS return.
After discussion with the LARP, RSH analysts confirmed the correct clustered figure for dwellings equivalent and the 2021 LADR data was corrected. Once the accurate figure was obtained, this gave a ratio of 2.67 bed spaces to 1 dwelling equivalent for this LARP (this in line with the overall sector average).
Case B
Similar to example given above, this LARP had reported the same figure for dwellings equivalent and bed spaces under Affordable Rent in the LADR 2019 Pilot, 2020 and 2021 datasets.
This was queried by RSH in 2021 and corrected, resulting in a ratio of 4.1 bed spaces to 1 dwelling equivalent cluster. It was established that the LARP had been uncertain in the dwelling equivalent clustering methodology so had been reporting individual bed spaces instead.
When we looked at the historical LAHS returns for this LARP, the unit breakdown questions were blank, and thus the dwelling equivalent figures had not been reported historically. This was confirmed through our conversations with this LARP in 2021, where they confirmed they had not been implementing the dwelling equivalent methodology and therefore, suggests a possibility that the stock totals reported in LAHS may contain issues with non-self-contained units counted as bed spaces rather than as dwelling equivalents.
Case C
In this case, a LARP reported matching figures for dwellings equivalent and bed spaces under Affordable Rent in the LADR 2019 Pilot, 2020 and 2021 datasets.
As with the previous cases, this was queried by RSH in 2021 and corrected resulting in a ratio of 2.4 bed spaces to 1 dwelling equivalent cluster.
Looking at the 2019 LAHS data for this LARP, they had reported an even higher figure for non-self-contained units here. This indicates that they had not only reported these units as individual bed spaces instead of dwelling equivalents in the LAHS but had also included additional units not reported in the LADR – most likely because of the broader scope of what is required to be reported in LAHS compared with LADR. This suggests that there may be some issues with the historical reporting of non-self-contained units in the LAHS, and highlights the issues of comparability between the two surveys.
Case D
A LARP had been reporting the same figure for dwellings equivalent and bed spaces under Affordable Rent in the LADR 2019 Pilot, 2020 and 2021 datasets in the same manner as the above cases.
When contacted by RSH in 2021, they confirmed a history of reporting their non-self-contained units as bed spaces only in the LAHS and LADR returns. However, on further investigation, due to a misinterpretation of which units should be clustered within the dwelling equivalent figures, the LARP confirmed they had been reporting these units under the wrong size category. These units were self-contained bedsits rather than non-self-contained bed spaces.
-
South West total excludes one unit of LCHO dwelling equivalent from calculation. Thus total impact is one unit less than the sum of the impact for rental units. ↩
-
Total unit difference figures shown figure are based on unit differences between dwelling equivalents and bed spaces for each region. Comparison of totals for South East, South West and England will not equal figures shown in this column due to 4 unit variance in self-contained figures in South East, and 1 unit variance in LCHO in South West. ↩
-
In the example below we have removed data from the 27 LARPs who we strongly suspect made errors in their dwelling equivalent count in 2020 ↩