Official Statistics

Children in low income families - After Housing Costs Consultation Note

Published 27 March 2025

Proposals for a new statistical series on After Housing Costs counts and proportions of children in low-income families, local area

DWP options and proposals for developing new annual local area Official Statistics on After Housing Costs (AHC) counts and proportions of children in low-income families, alongside the existing Before Housing Costs (BHC) Children in low income CiLIF statistics.

Authors: Russ Bentley and Steve Watkins (DWP) in consultation with Juliet Stone (Centre for Research in Social Policy, Loughborough University)

Email stats.consultation-2018@dwp.gov.uk

Release date: 27 March 2025

1. Executive Summary

The options in this paper to develop After Housing Costs (AHC) measures are presented to illustrate the discovery work that has been carried out. We invite comments on the value of the proposed statistics and on the methodology and scope for any future release. Email stats.consultation-2018@dwp.gov.uk with feedback and queries about the statistics.

AHC statistics for local area child poverty are not currently available as complete source data on social and private sector rents and mortgage payments for individual families is not held on DWP administrative systems. However, given the focus on relative AHC poverty in the Tackling Child Poverty: Developing Our Strategy:

We will be guided by the leading, and internationally recognised, measure of poverty – Relative Poverty After Housing Costs – the proportion of families with below 60% of the median income after housing costs are deducted

we are striving to establish what might be possible and have developed a number of initial proposals in consultation with the Centre for Research in Social Policy that we are seeking comments on.

Centre for Research in Social Policy produce Local Child Poverty Statistics for the End Child Poverty Coalition on a yearly basis. This research provides a figure and percentage of children living in relative poverty, after housing costs, for each Local Authority and Parliamentary Constituency: End-Child-Poverty-Briefing.

For any new Official Statistics, our principle is to build from a granular level i.e. individual and family level data to derive AHC gross income from the established existing statistics on a BHC basis. Estimating families’ housing costs is thus central to the discovery work and the options presented show how this might be done using:

[a] families Housing Benefit and Universal Credit housing element payments

[b] actual Local Authority rental data

Comments are sought on the detailed methodological considerations – and on the scope of the proposed statistics by 31 July 2025. Updates will be communicated via DWP Statistical Work Programme ahead of any proposed publication in March 2026.

2. Introduction

The CiLIF statistics, provide information on the number and proportion of children living in Relative and Absolute low income Before Housing Costs (BHC) by local area across the United Kingdom.

Figures are calibrated to the Households Below Average Income (HBAI) survey UK estimates but provide more granular local area information not available from the HBAI for example, by local authority, parliamentary constituency and ward.

In the absence of DWP Official Statistics on After Housing Costs (AHC), Loughborough University developed a method for adjusting the BHC statistics to estimate the effect of varying housing costs on child poverty in local areas using administrative data on rents for local authorities, combined with household-level data from the Understanding Society longitudinal survey to estimate the relationship between housing costs and the relative risk of being in poverty before and after housing costs. For constituencies, for which local rent data are not available, information on median house prices is also used.

DWP have committed in the Statistical Work Programme to undertake discovery work on the potential to develop and publish Official Statistics on AHC local area child poverty and this paper seeks to summarise what may be possible. Comments are sought on the options and methodological considerations and on the future scope and dissemination of the statistics.

3. Modelling AHC local area statistics

A fundamental principle of the children in low income families(CiLIF) statistics is to utilise the DWP administrative data to build a summary of each child/family equivalised gross income to establish where, across the UK, the poorest children live – calibrated to the HBAI survey UK estimates. This approach offers a level of insight beyond generic modelling at LA level since it is specific to each individual child and to their family composition and estimated gross incomes.

At the heart of this discovery is how best to derive AHC gross Income from BHC gross income at family level: what approach do we use to estimate families’ housing costs ? Option A considers actual Universal Credit / Housing Benefit housing element payments. Option B considers using a generic average LA rental amount adjustment. Both methodologies apply to granular family level data rather than applying adjustments to the existing aggregate published BHC statistics.

CiLIF statistics draw data from the database “RAPID” (Registration and Population Interaction Database) which provides a single coherent view of citizens’ interactions with DWP and HMRC within a tax year for the UK. RAPID provides a basis for analyses of children, the family unit, and gross personal incomes (benefits/tax credits, employment, self-employment, occupational pensions) from which estimates of the number of children in low income families can be derived calibrated to HBAI UK estimates on Absolute and Relative definitions.

RAPID is based on 100% extracts of various DWP benefit systems and is supplemented with 100% data extracts from HMRC systems. RAPID collates information on individual activities (and the income generated from those activities) within each tax year, including benefit, employment and in-work benefit interactions, for example Tax Credits and Housing Benefit. Children have been identified from HMRC Child Benefit scans.

RAPID does not capture information on social or private rents or on mortgage payments and thus historically, AHC metrics have not been derived.

Going forward – for possible publication in March 2026 – two Options are presented on how AHC metrics could be developed:

A. Using Universal Credit / Housing Benefit housing element payments for individual families’ which reflect their family circumstances and composition (and potential imputation for cases where no housing payment is claimed)

B. Using generic Local Authority average rental data for individual families

Option A

Separately from RAPID, data is available on Universal Credit housing element payments for families claiming Universal Credit. This housing element payment is recorded for both the social and private rental sectors.

This source will – going forward – increasingly capture a greater proportion of families in low income as Housing Benefit and Tax Credits are phased out and claimants migrate to Universal Credit. However, similar payment data to also available for families on Housing Benefit.

Indicatively for FYE 2024, coverage is 70% i.e. of all families with children on Universal Credit or Housing Benefit, 70% have recorded housing payments. A proportion are on UC with no housing element payments (for example these may be owner occupiers with a mortgage or conversely claimants living at home or with family or friends with or without housing rental costs).

The Universal Credit housing payments capture for both social rented and private rented sectors, the housing payment entitlements for families with children.

In the social rental sector, housing payments should broadly reflect rent, subject to the spare room subsidy whereby a proportion of the housing award is removed if there are additional spare rooms over and above the entitlement. The bedroom standard is the most widely used occupancy rating used in the UK. The standard assumes that every household should contain a separate bedroom for each of the following (i) adult couple (ii) any remaining adult (aged 21 years or over) (iii) two adolescents (aged 10 to 20 years) of the same sex (iv) one adolescent (aged 10 to 20 years) and one child (aged 9 years or under) of the same sex (vi)two children (aged 9 years or under) regardless of sex (vii) any remaining child (aged 9 years or under).

In the private rented sector, the housing element is based on Local Housing Allowance (LHA) rates or actual rent – whichever is the lower. 100 broad rental area rates – based on valuation office data – are used to determine housing element awards in combination with family composition and circumstances (for example age and disability), income and occupancy.

Note: entitlements are not necessarily the same as actual rental/housing costs. Housing Benefit and UC housing element payments aim to help with housing costs, but they might not always fully reflect actual rents especially in high-cost areas: Housing costs and Universal Credit: What you can get.

Does this shortfall matter for the AHC methodology?

The AHC statistics are not presenting estimates of gross income (£,p) rental costs – rather they are using housing costs data to adjust a families’ gross income and rank – across the UK – where the poorest children live. Thus, the critical factor is how consistent across the UK the housing payment data is. More discovery work is required to better understand this e.g. computation of the housing element shortfall.

How to model families with ‘missing’ housing costs?

Housing element payments are not held for x% of families with children on Universal Credit or Housing Benefit. These may be for example owner occupiers, or people living at home but who may or m ay not pay rent, or those in temporary/supported/shared ownerships.

One option would be to treat these families as having no housing costs. An alternative approach would be to proxy housing costs where these are missing using housing payment for other families in the same local area; potentially in the same street (using Census Output Area (COA) data) or the same ‘town’ (Lower Super Output Area (LSOA)) to seek to mirror as closely as possible local housing costs proxies.

Whilst COA has the advantage of being very granular, there may not be many comparison cases and thus the methodology may not account for family composition. LSOA will capture more cases and thus similar family characteristics (e.g. lone parents, number of children for comparator housing payments) could be used as proxy cases.

The availability of Universal Credit/Housing Benefit data for Northen Ireland data is an issue that requires further discovery under Option A.

Option B

Similar to Option A to the extent that the methodology uses granular administrative data on families’ BHC income and derives an AHC measure by estimating housing costs.

Different to Option A in how these housing costs are estimated. Under option B actual Local Authority rental data is used and applied to families within each LA. Whilst this approach would use similar data to the Loughborough methodology it would be essentially different as it will not be used to transform published BHC statistics, but will instead use family level data at the data derivation stage to estimate gross income AHC.

A number of sources for local area rental statistics require further investigation for use in any methodology:

Private rents by LA (discontinued)

Price Index of Private Rents, UK

Ministry of Housing, Communities and Local Government, Ministry of Housing, Communities & Local Government (2018 to 2021) and Department for Levelling Up, Housing and Communities publish statistics on Live tables on rents, lettings and tenancies together with Unlocking the power of data to better understand private rents

Using a combination of these metrics, a methodology could be developed to derive AHC gross income for families by applying “average” rental costs by LA. These “average” rental costs would need to be defined as a norm e.g. is a 2-bed property a reasonable standard to use ? should median or mean or lower quartile rents be used ? how are outliers handled ?

4. Proposed Scope

DWP welcomes feedback on other breakdowns which users feel are necessary, though some developments may be more easily achieved than others, and some may be hampered by the current availability of data or its quality.

Coverage and content

UK or GB with standard geographical breakdowns by country, region, Local Authority, ward, and small area census based. Relative metrics only – statistics on Absolute metrics would not be developed.

Timing and frequency

Proposed publication in March 2026 alongside existing CiLIF Official Statistics. Data would relate to FYE 2025. Further discovery is required to establish if a back-series for FYE 2023 and FYE 2024 could be provided.

Dissemination

Via the existing DWP StatXplore platform.

5. Next Steps

We are seeking comments by 31 July 2025.

We will also utilise DWP Surveys Expert Group Forum to present and discuss further discovery work throughout the year.

Once we have considered all of the responses to these proposals we will publish a response via the DWP Statistical Work Programme and on CILIF Collection Page.