Improvements to private actual and imputed rentals in household expenditure in UK National Accounts, Blue Book 2026

Every year, the Office for National Statistics (ONS) updates the sources and methods for the UK National Accounts and publishes the latest estimates (including revisions to past periods) in our annual UK National Accounts, the Blue Book compendiums.

Our article, published on 24 March 2026, described the proposed changes to the UK National Accounts and balance of payments being made in our Blue Book 2026 and Pink Book 2026 releases.

One of the main methodological improvements relates to the way we measure household final consumption expenditure (HHFCE) rental data, which comprise both private actual and imputed rentals.

HHFCE and the expenditure measure of gross domestic product (GDP) include estimates of actual property rentals paid to private landlords and estimates of owner-occupiers' imputed rentals.

Imputed rentals represent the rental income that an owner-occupier foregoes by living in their property, instead of renting it to a tenant. Measurement of imputed rental is important to ensure that the valuation of housing services is calculated on a consistent basis, for comparisons over time and between countries. Different countries and time periods will have a different composition of owner-occupied housing and rentals. For example, two homeowners would still consume the same amount of housing services as if they both rented their homes to each other, and we need to include these services in the National Accounts.

Imputed and private actual rentals for housing are identified separately in the National Accounts and our Consumer trends bulletins; they have Classification of Individual Consumption According to Purpose (COICOP) codes 04.1 and 04.2, respectively. They are an important part of the National Accounts, with a combined GDP weight of around 14% in 2024.

The main HHFCE rental changes are:

• improved rental prices

• improved deflators

• new data for rent-free accommodation

Processing system changes to improve quarterly interpolation of annual prices were also implemented.

Improvements to prices and deflators account for most of the changes to HHFCE rental and are described in our Redevelopment of private rental prices statistics, impact analysis, UK: January 2025 article.

These changes led to annual reductions in HHFCE rental current price levels of between 1% and 3% between 1997 and 2024. Within this time period, changes to year-on-year growth rates were less than 2 percentage points in each year except 2017, where annual growth was revised upwards from 1.0% to 3.1%. Imputed rental chained volume growth rates were revised upwards from negative 0.7% to positive 1.4% in 2017.

This article explores the methodological considerations behind the improvements made to HHFCE rental within the UK National Accounts. It does not consider detailed data impacts. This will be part of a later article to be published in August 2026, which covers all changes implemented in the 2026 Blue Book.

The methods and data sources used in household expenditure were introduced in our Changes to National Accounts: Imputed Rental article for Blue Book 2016, and in our Changes to National Accounts: actual rentals and imputed rentals article for Blue Book 2017. They are aligned to the European System of Accounts 2010 (ESA10).

We estimate the current price actual, and the imputed rental expenditure, by multiplying estimates of dwelling stocks by average rental prices paid in the private market. This is done by tenure, region and property type, and is aggregated up to the levels required for publication. The use of private rental payments to measure imputed rental is part of the rental equivalence approach, described further in our Consumer Prices Index, including owner occupiers' Housing costs (CPIH) compendium.

Deflators are applied to the current price estimates to remove the effect of inflation and obtain the chained volume measure.

Annual housing stocks by property type, region and tenure are obtained from several sources, including the devolved governments and the Ministry of Housing, Communities and Local Government.

Annual rental prices vary according to the characteristics of the property, that is tenure, property type and region, and in the case of actual rentals, whether it is furnished or unfurnished. These are compiled using data from the Valuation Office Agency (VOA) for England and from the devolved governments.

The owner occupiers' housing costs index (CPIH Index 04.2) and the Price Index of Private Rents data are used to deflate imputed and actual rental expenditure, respectively, to give the chained volume measures for these series. These are applied at the UK level and all property type levels.

Annual rental prices are compiled from samples of over 500,000 rental prices per year. Under the current approach, we calculate a simple average of all prices collected in each region, and by property type. We use this alongside dwelling stock estimates to derive current price actual and imputed rental expenditure. However, these average prices present a high degree of volatility because of sample variability and because prices are not adjusted for quality differences. Adjustments have been used to remove some of this volatility from the published household final consumption expenditure (HHFCE) rental data in the past.

In Blue Book 2026, hedonic modelled predicted prices replace simple average prices. These hedonic prices reflect the average expected price for a property of a given type, in each location, rather than the collected price.

Regression model coefficients use these collected prices to predict the rental price for a "typical" property of a specific type, in a particular location, at a point in time. This gives a predicted "average" rental price for a property instead of a potentially extreme "actual" rent price.

The process also includes treatment of outliers. The price-determining characteristics used in the hedonic regression are:

• number of bedrooms

• floor area

• property type (detached, semi-detached, terraced, and flat or maisonette)

• furnished status

• geographic and demographic groupings

• local Authority Districts in England and Wales, and Broad Rental Market Areas in Scotland

• property age bracket

This approach follows an internationally recognised method, which is described in Chapter 5: Hedonic Regression Methods of the Eurostat Handbook on Residential Property Prices Indices (PDF, 8.1MB).

Using hedonic modelled prices results in better consistency with our Price Index of Private Rentals (PIPR) methodology, and with our CPIH index methodology. Improvements to these methodologies are described in our Redevelopment of private rental prices statistics, impact analysis, UK: January 2025 article.

In addition to smoothing out extreme collected prices, hedonic modelling provides prices that are more reflective of the rented stock than simple averages.

Simple averages are based on prices collected over a year and will disproportionately reflect new rental contracts over existing rental contracts. This does not accurately reflect the rented stock, because most rented properties are under existing contracts, where rent prices tend to remain unchanged for a period of time, such as one year. Using hedonic modelling and a 14-month validity period removes volatility from the collected prices, and ensures that rental prices can remain in the sample for up to 14 months. This better reflects the typical length and behaviour of rental prices for the entire rented stock.

Figure 1: Hedonic modelled owner-occupied property prices show smoother trends than simple average prices

In Blue Book 2025, we replaced the household final consumption expenditure (HHFCE) actual rentals deflator, which was produced using aggregated rental prices, with the Price Index of Private Rents (PIPR). We also adopted the improved Consumer Prices Index including owner occupiers’ housing costs (CPIH) Index 04.2 to deflate imputed rental. Both changes were implemented back to 2021, and the revised data are considered to better reflect the inflation seen during the cost-of-living crisis. For Blue Book 2026, data are extended back to 2015 for actual rental, and to 2018 for imputed rental.

PIPR was launched for Great Britain in March 2024, and for the UK in March 2025. It incorporates the hedonic modelling methodology described in Section 3: Improved rental prices section of this methodology.

The same rental data used in the PIPR are also used to construct the CPIH Index 04.2, under the rental equivalence approach to measuring imputed rental.

Figure 2: Rental deflators more accurately reflect price increases since 2018 (imputed rental), and 2015 (actual rental)

Rent-free accommodation covers cases such as:

• people living in accommodation owned by a family member

• nannies who are given free live-in accommodation along with their job

• construction workers who are given rent-free accommodation by their employers

We will replace forecasted data with data on private rented dwellings that are occupied rent free sourced from the English Housing Survey and commissioned by the Ministry of Housing, Communities and Local Government (MHCLG). This change will have minimal effect on published annual levels and growth rates.

Improved prices and the removal of price volatility adjustments led to reduced levels in household final consumption expenditure (HHFCE) rental at current prices, compared with Blue Book 2025 data, especially in the years before 2017. Data before 1997 will also be recalculated in line with these changes.

Figure 3: Household final consumption expenditure (HHFCE) rental levels at current prices have reduced

Changes are also seen in current prices (CP) and chained volume measure (CVM) growth rates, with changes above 1% seen from 2015 onwards. CVM growth rates are further affected by improved deflators.

Imputed rental growth rates for 2017 are notably higher than Blue Book 2025 rates.

Figure 4: Household Final Consumption Expenditure (HHFCE) rental growth rates have been affected by price and deflator changes