Official Statistics

Leasehold dwellings, 2019 to 2020: technical notes

Published 8 July 2021

Applies to England

Official Statistics status

The statistics in this release were originally labelled as ‘Experimental Official Statistics’ signifying that they were undergoing development before a final decision was made on their eventual designation. That initial evaluation period has now ended and the statistics have since been given the status of unqualified ‘Official Statistics’. This designation means that the statistics have been produced in accordance with the governance and operational arrangements set out in statistical legislation, and in compliance with the standards of best practice set out in the Code of Practice for Statistics.

Data collection and methodology

The Leasehold Dwellings, 2019-20 Official Statistical Release presents estimates of the number of leasehold dwellings in England in 2019-20 by tenure, dwelling type and region.

An estimate of the number of leasehold dwellings was first published by the Ministry of Housing Communities and Local Government (MHCLG) in 2014. The statistics were released as Experimental Statistics to enable analysts at MHCLG to improve on the methodology for producing the estimates. The Good Practice Team of the Office for National Statistics reviewed the statistics in 2015-16 and judged the methodology to be “the best approach given the available data sources”. Since then, a number of adjustments have been made to improve the accuracy of the estimates and, in 2020, MHCLG’s Head of Profession for Statistics recommended that the estimates released in July 2021 are to be published as Official Statistics^{[footnote 1]}.

The methodology used to produce the leasehold estimates is broadly the same as the methodology used to produce the first estimate in 2014. The methodology involves calculating the proportion of dwellings with leasehold titles by tenure and dwelling type by matching the data from HM Land Registry Title Descriptor Dataset to the data of the English Housing Survey (EHS) using the Unique Property Reference Number (UPRN), and applying those proportions to the total number of dwellings in England (which is derived using MHCLG’s Dwelling Stock Estimates (table 4) adjusted to account for dwelling type using Valuation Office Agency’s (VOA) Council Tax Stock of Properties).

While the methodology has remained broadly the same, a number of adjustments have been made in order to improve the accuracy of the estimate. They included: improvements to the calibration; widening the coverage of the statistics; improvements to address matching; imputation of missing data; and the introduction of regional calibration. These adjustments are explained below and summarised in Table 1.

The adjustments improved the coverage, accuracy and geographical precision of the estimate, but have retained the fundamental methodology. Therefore, the changes do not constitute a break in the time series, and estimates are comparable from the 2014-15 release onwards for the private sector, and from the 2015-16 release onwards for all tenures. It is therefore possible to compare the current estimate with previous estimates.

Improved calibration

To ensure the most accurate possible estimate, the leasehold proportions obtained by matching the English Housing Survey with HM Land Registry are applied (calibrated/weighted) to the total number of dwellings in England, using MHCLG’s Dwelling Stock Estimates and adjusted to account for dwelling type using VOA’s Council Tax Stock of Properties. This method was first used to produce the 2014-15 estimate and has been retained in this release. The classification of dwelling type in English Housing Survey data has been adjusted to match the definitions used by the VOA. In particular, bungalows are now counted separately (they were previously classified as either detached or non-detached houses). From 2018-19 the weighting strategy was updated to also control for regional variations, also using the VOA’s Council Tax Stock of Properties.

Prior to establishing the Experimental Statistics series in 2014-15, MHCLG produced an estimate of the number of leasehold dwellings, based on 2012-13 data. In that release, the leasehold proportions were calibrated to the 2011 Census to obtain the total number of properties. This was due to the proximity of the 2012-13 estimate to the 2011 Census. It is expected that the results of the 2021 Census will be used to produce future leasehold estimates.

Widened coverage

Since 2015-16, the estimates have included the small number of leasehold dwellings in the social rented sector. Prior to that, only leasehold dwellings in the private sector were included.

Also since 2015-16, the leasehold estimates have been based on two years of English Housing Survey data to increase the sample size, and to improve the accuracy of the estimate.

Improvements to address matching

The accuracy of the leasehold estimate relies on a high match rate between the English Housing Survey and HM Land Registry data. Given the relative completeness of the English Housing Survey address information, most cases are easily matched in AddressBase and have a UPRN. However, a small number of cases do not. From the release covering 2016-17, a new address matching algorithm was developed which enabled “fuzzy matching” between the English Housing Survey and AddressBase, allowing English Housing Survey addresses without a precise match in AddressBase to be matched to a UPRN. These are typically dwellings where the address has been recorded in a different format, or where the address has changed, for example from a street number to a named house. In 2019-20, this process matched a UPRN to 99% of the issued sample of the English Housing Survey (27,990 cases). Of the cases matched to a UPRN 95% could be matched to HM Land Registry data (26,585 cases), giving an overall rate of 94% of EHS cases matched to a HM Land Registry data.

Imputation of missing data

A small proportion of EHS cases cannot be matched to HM Land Registry data. These “missing” or “unmatched” cases either cannot be matched to a UPRN, or do not have a record in HM Land Registry title data^{[footnote 2]}. To produce the 2017-18 release, MHCLG worked with the Methods Advisory Service at the ONS to implement an imputation strategy for the 836 cases that could not be matched to HM Land Registry data.

Imputation was not undertaken to produce the 2018-19 and 2019-20 releases because revisions to the calibration strategy to include regional estimates of the dwelling stock (see below) performs the same role, thus negating the need for imputation.

Regional calibration

In 2018-19 the release included regional estimates of the number of leasehold dwellings for the first time. Regional estimates are also included in this release (2019-20) and will be included in all future releases.

Table 1: Overview of leasehold methodology

Official Statistics

Survey year	Leasehold data	Calibration	Changes
2019-20	HM Land Registry’s Title descriptor dataset	MHCLG’s Dwelling Stock Estimates distributed by dwelling type using VOA’s Council Tax: Stock of Properties (England and Wales)	No changes to methodology

Experimental Statistics series

Survey year	Leasehold data	Calibration	Changes
2018-19	HM Land Registry’s Title descriptor dataset	MHCLG’s Dwelling Stock Estimates distributed by dwelling type using VOA’s Council Tax: Stock of Properties (England and Wales)	Regional data added - Imputation removed - Regional calibration and estimates added
2017-18	HM Land Registry’s Title descriptor dataset	MHCLG’s Dwelling Stock Estimates distributed by dwelling type using VOA’s Council Tax: Stock of Properties (England and Wales)	Imputation of missing title data - Imputation applied to cases without a UPRN or a Land Registry title record.
2016-17	HM Land Registry’s Title descriptor dataset	MHCLG’s Dwelling Stock Estimates distributed by dwelling type using VOA’s Council Tax: Stock of Properties (England and Wales)	Improvements to address matching - “Fuzzy matching” used to improve the match rate of EHS address data to its UPRN.
2015-16	HM Land Registry’s Title descriptor dataset	MHCLG’s Dwelling Stock Estimates distributed by dwelling type using VOA’s Council Tax: Stock of Properties (England and Wales)	Wider coverage - Estimate of the number of leasehold dwellings in the social rented sector added. - Estimate based on two years of EHS to increase sample size and improve precision.
2014-15	HM Land Registry’s Title descriptor dataset	MHCLG’s Dwelling Stock Estimates distributed by dwelling type using VOA’s Council Tax: Stock of Properties (England and Wales)	Calibration - Calibrated to Dwelling Stock Estimates

Before Experimental Statistics series was established

Survey year	Leasehold data	Calibration	Changes
2012-13	HM Land Registry’s Title descriptor dataset	Census 2011 distributed by dwelling type using the English Housing Survey

Data quality

Things to note on this release

Throughout this report we refer to the survey year (2019-20) rather than the publication year when describing the methodology and data quality.

Assessment of data quality

In 2015, the UK Statistics Authority (UKSA) published a regulatory standard for the quality assurance of administrative data. To assess the quality of the data provided for this release, MHCLG has followed that standard.

The standard is supported with an Administrative Data Quality Assurance Toolkit which provides useful guidance on the practices that can be adopted to assure the quality of the data they utilise.

The Leasehold Dwellings 2019-20 Statistical Release is produced by MHCLG and based on data collected by MHCLG (English Housing Survey, Dwelling Stock Estimates), VOA (Council Tax: Stock of Properties (England and Wales), and HM Land Registry (National Polygon Service: Title Descriptor datasets).

An assessment of the level of risk based on the Quality Assurance Toolkit is as follows:

Risk/Profile Matrix Statistical Series	Administrative Source	Data Quality Concern	Public Interest	Matrix Classification
Estimating the number of leasehold dwellings in England	MHCLG	Low	Medium	Medium risk [A1]

The publication of the annual estimate of the number of leasehold dwellings typically generates a medium level of public interest and moderate economic and/or political sensitivity. Some specialist leasehold interest groups report on the findings and there is sometimes press coverage.

The data quality concern is low because of the robustness of the data on which the estimate is based, and the high match rate achieved between the English Housing Survey and the HM Land Registry data. A full outline of the statistical production process and quality assurance carried out is provided in the flow chart below.

The data are then further quality assured in detail by statisticians responsible for this publication, who perform further detailed validation and checks, spotting and correcting any errors. These checks include comparisons with data provided, published or historical data.

Overall, the annual estimate of the number of leasehold dwellings has been assessed as A1: Medium Risk. This is mainly driven by the medium level of public interest in the statistics.

A full outline of the statistical production process and quality assurance carried out is provided in the flow chart in Figure 1 below.

Further details are also provided against each of the four areas outlined in the Quality Assurance of Administrative Data (QAAD) Toolkit.

1. Operational context and administrative data collection

No new data are collected for this statistical release. Instead, it draws on information from the statistical sources listed below:

• English Housing Survey (National Statistics produced by MHCLG)
• Dwelling Stock Estimates (National Statistics produced by MHCLG)
• Council Tax: Stock of Properties (England and Wales) (Official Statistics produced by VOA)
• National Polygon Service: Title Descriptor dataset (HM Land Registry)

2. Communication with data supply partners

Data suppliers work collaboratively to produce this release to ensure that there is a common understanding of what information is being supplied and why. There is regular contact between the analysts responsible for the leasehold estimate and those responsible for the English Housing Survey and the Dwelling Stock Estimates, who all work in the Housing and Planning Analysis Division at MHCLG.

VOA’s Council Tax: Stock of Properties data are publicly available and downloaded from Gov.uk.

There is a Memorandum of Understanding in place between MHCLG and HM Land Registry which sets out the specific data sharing arrangements for the sharing by MHCLG of data relating to the properties included in the EHS sample, and HM Land Registry of data relating to the tenure (leasehold or freehold) of properties.

3. QA principles, standards and checks by data suppliers

The data sources for these statistics are National (English Housing Survey and Dwelling Stock Estimates) or Official Statistics (Council Tax: Stock of Properties (England and Wales) and HM Land Registry data) collected and published by government departments in line with the Code of Practice for Statistics. More information on the quality of the data sources used to produce the annual estimate of leasehold dwellings can be found at the links for each data source, provided above.

At MHCLG’s request, the Office for National Statistics (ONS) Good Practice Team reviewed the leasehold methodology in 2015-16. They judged it to be “the best approach given the available data sources”.

4. Producers’ QA investigations and documentation

Whilst providers are expected to carry out their own checks before the data is submitted for this publication, further quality assurance is carried once the data is received by the responsible statisticians for this publication. The data are compiled together and compared to the raw data, published figures, historical time series and policy information that may have impacted the figures.

The data, report and tables are quality assured independently by another statistician in the production team. Further final checks are performed to the final end product. These checks use a clear checklist approach to ensure the figures are consistent across the release and live tables, with each check being systematically signed off when it has been completed.

Imputation

A small proportion of EHS cases cannot be matched to HM Land Registry data. These “missing” or “unmatched” cases either cannot be matched to a UPRN, or do not have a record in HM Land Registry title data^{[footnote 3]}. To produce the 2017-18 release, MHCLG worked with the Methods Advisory Service at the ONS to implement an imputation strategy for the 836 cases that could not be matched to HM Land Registry data.

Imputation was not undertaken to produce the 2018-19 and 2019-20 releases because revisions to the calibration strategy to include regional estimates of the dwelling stock performs the same role, thus negating the need for imputation.

Definitions

Dwelling

A unit of accommodation which may comprise one or more household spaces (a household space is the accommodation used or available for use by an individual household). A dwelling may be classified as shared or unshared. A dwelling is shared if:

• the household spaces it contains are ‘part of a converted or shared house’, or
• not all of the rooms (including kitchen, bathroom and toilet, if any) are behind a door that only that household can use, and
• there is at least one other such household space at the same address with which it can be combined to form the shared dwelling.

Dwellings that do not meet these conditions are unshared dwellings.

Leasehold

A long leasehold is a form of property ownership normally used for flats that is simply a long tenancy, providing the right to occupation and use for a long period – the ‘term’ of the lease. This can be a period of over 21 years and the lease can be bought and sold during this term. The term is fixed at the beginning and decreases year by year, until the property returns to the landlord. Houses can also be leasehold. A person who buys a leasehold property on a lease is called a leaseholder.

Freehold

The freehold interest in land is a title in property that can be held in England and Wales. In practice a residential freehold interest applies to the outright ownership of land or property for an unlimited period and applies to the majority of houses.

Share of freehold

Where the freehold of the building is (a) either owned jointly by a number (up to four) of the flat owners in their personal names, or (b) where a company is the owner of the freehold and each of the leaseholders hold a share or membership in that company.

The data most commonly show detached houses to have a single freehold title. For flats, the record most frequently showed either a single freehold title or a single leasehold title. However, some homes were found to have multiple titles of ownership recorded. In these cases addresses with one or more leasehold titles registered were categorised as leasehold. Conversely, cases were designated as freehold only if all titles found in the data were freehold titles.

These definitions have not changed since the statistical series was first established.

The definitions used in this publication can be found in the ‘Housing statistics and English Housing survey’ glossary published by MHCLG.

Revisions policy

This policy has been developed in accordance with the UK Statistics Authority’s Code of Practice for Statistics and the Ministry of Housing, Communities and Local Government Revisions Policy.

It covers two types of revisions:

Non-scheduled revisions

Where a substantial error has occurred as a result of the compilation, imputation or dissemination process, the statistical release, live tables and other accompanying releases will be updated with a correction notice as soon as is practical.

Scheduled revisions

Changes to the component data sources used in this release will be incorporated in the next scheduled release of data.

Revisions in this release

This release does not contain any revisions.

Other information

Comparison with other data sources

English Housing Survey

The English Housing Survey alone underestimates the number the leasehold dwellings. There are two main reasons for this:

• The English Housing Survey cannot provide an estimate of the number of leasehold properties in the social or private rented sectors, because respondents are tenants who will not know if their home is leasehold; and -Owner occupiers may not understand the leasehold questions and misreport their leasehold status.

HM Land Registry Price Paid Data

The HM Land Registry Price Paid Data records whether properties bought and sold (and registered with HM Land Registry) are leasehold or freehold. This is a record of transactions not properties and cannot be used to estimate the prevalence of leasehold properties across the whole stock.

HM Land Registry Title Data

In addition to using HM Land Registry Title Data to produce the estimate of the number of leasehold dwellings, MHCLG have also conducted robustness checks against these high level data.

HM Land Registry keeps a register of leasehold titles in England and Wales^{[footnote 4]}. In 2019, there were a total of 5.3 million registered leases, including non-residential leases such as commercial premises or leases of roof spaces for solar panels, and multiple leases on some dwellings, such as subsidiary leases for flats within a leased block. It is therefore not possible to calculate a definitive number of leasehold dwellings from this dataset. HM Land Registry estimate that just over 160,000 of these leases were in Wales^{[footnote 5]}, and that 14%^{[footnote 6]} of all leases were likely to be non-residential. Excluding these Welsh and non-residential leases gives a total of approximately 4.5 million residential leases in England, including multiple leases on the same dwelling. This figure is a plausible upper limit on the number of leasehold dwellings in England, and is within the confidence intervals reported in this release (4.4 to 4.9 million leasehold dwellings). It therefore provides support for the robustness of this estimate.

Related statistics

Help to Buy (Equity Loan scheme) and Help to Buy: NewBuy statistics The Ministry of Housing, Communities and Local Government publishes quarterly Official Statistics on the number of home purchases and the value of equity loans under the Government’s Help to Buy: Equity Loan scheme, as well as the number of purchases under the government’s Help to Buy: NewBuy scheme (formerly known only as ‘NewBuy’). The latest statistical release contains data on the number of homes purchased through the Help to Buy: Equity Loan scheme that are freehold and the number that are leasehold. Further information, including a breakdown of homes purchased through the scheme by tenure (freehold/leasehold), property type and local authority, is available in the live tables accompanying the release (Release Tables 9, 9a and 9b). The Help to Buy (Equity Loan scheme) and Help to Buy: NewBuy statistics can be found at https://www.gov.uk/government/collections/help-to-buy-equity-loan-and-newbuy-statistics.

The Office for National Statistics (ONS) produces leasehold statistics on England and Wales in a publication entitled ‘Leasehold and freehold residential property transactions in England and Wales’. Unlike the statistics in this release, the ONS publication covers only transactions not the stock of leasehold dwellings.

The Ministry of Housing, Communities and Local Government is part of a cross-government working group, including devolved administrations and the Office of National Statistics, working to improve the trustworthiness quality and value of housing and planning statistics across the UK. This working group has published action plan to make the planned improvements on house building statistics clear and transparent to users.

Details of this work and how you can provide feedback are available via the Government Statistical Service website

Devolved administration statistics

There are no comparable statistics available for Scotland, Wales or Northern Ireland.

User engagement

Users are encouraged to provide feedback on how these statistics are used and how well they meet user needs. Comments on any issues relating to this statistical release are welcomed and encouraged. Responses should be addressed to the contact given in the first page of the release, Stephen Pottinger via email - [email protected].

The department’s engagement strategy to meet the needs of statistics users is published here: https://www.gov.uk/government/publications/engagement-strategy-to-meet-the-needs-of-statistics-users.

1. All the Leasehold Experimental Statistics Releases can be found at: https://www.gov.uk/government/collections/leasehold-dwellings. ↩

2. Dwellings which have not been bought, given or inherited, received in exchange for other property or land, or been mortgaged since 1990 may not be recorded with HM Land Registry. ↩

3. Dwellings which have not been bought, given or inherited, received in exchange for other property or land, or been mortgaged since 1990 may not be recorded with HM Land Registry. ↩

4. Register Stocks by Tenure (HM Land Registry internal report) based on data from the National Polygon Service datasets. ↩

5. HM Land Registry internal calculations ↩

6. Estimated based on unpublished internal regulatory impact assessment for HM Land Registry and https://www.ons.gov.uk/businessindustryandtrade/business/activitysizeandlocation/bulletins/ukbusinessactivitysizeandlocation/2018. ↩