Unfortunately, your browser is too old to work on this website. Please upgrade your browser
Skip to main content
Analysis

Rethinking fuel poverty in Great Britain
Where are risks greatest?

Published 16 November 2026
Time to read clock icon About 12 mins
Authors

Key points

  • Fuel poverty – being unable to adequately heat your home – is strongly associated with poor health and inequalities in the UK. The cold, damp living conditions and financial stress that come with fuel poverty are linked to a wide range of adverse physical and mental health outcomes. Fuel poverty is also associated with higher numbers of excess winter deaths, increased injury risk for older people, and respiratory illnesses and barriers to development among children.
  • Definitions of fuel poverty are debated and inconsistent across the country. In Scotland and Wales, fuel-poor status depends primarily on the proportion of income a household spends on energy costs. In England, a household must also live in an energy-inefficient property to qualify as fuel poor. These definitions may underestimate how many households in Britain are unable to adequately heat their homes.
  • We used linked housing, socioeconomic and smart meter data to map fuel poverty risk at the subregional level throughout Great Britain. In this analysis, we explore where in Britain is at greatest risk of fuel poverty, how health outcomes differ in fuel-poor areas and how we can improve our identification of fuel poverty risk to better target support. 
  • Areas at the highest risk of fuel poverty have 18% higher rates of hospital admissions with respiratory diagnoses compared with areas at the lowest risk, as well as higher rates of emergency care use, cardiovascular-related hospital admissions and poor mental health.
  • The risk of fuel poverty is substantially more widespread across Wales and Scotland than in England. In England, fuel poverty risk is highest in the North West, the West Midlands and Yorkshire and the Humber. 
  • Areas at higher risk of fuel poverty tend to have younger populations than areas at lower risk, and higher proportions of residents with Asian or Asian British and black or black British backgrounds. While people in areas at high risk of fuel poverty tend to have lower incomes than those in areas less at risk, fuel poverty risk is not limited to only the lowest income areas. 
  • We propose reforming the Low Income, Low Energy Efficiency indicator used to measure fuel poverty in England to better identify fuel-poor households and meaningfully track progress towards alleviating fuel poverty. Linked data, including from smart meters, should also be used to deliver on the promises of current government policy by dynamically targeting support.
 

Introduction

Fuel poverty, where a household is unable to adequately heat their home, is strongly linked to worse health and inequalities. Unstable energy prices are compounding wider cost-of-living pressures, forcing many of the most vulnerable households to make health-damaging financial trade-offs, most starkly between heating and eating. Being unable to adequately heat your home can lead to colder, damper living conditions, as well as severe financial distress.

All of this has a critical impact on health. Living in an inadequately heated home is associated with adverse respiratory, cardiovascular and mental health outcomes, as well as childhood development issues and increased injury risk for older people. Illness linked to cold homes is estimated to cost the NHS over £2.5bn annually. 

In response, the government has made addressing fuel poverty a key priority. The 2026 Warm Homes Plan sets out a strategy to alleviate fuel poverty by upgrading the energy efficiency of millions of homes and providing direct support with energy bills to households in need. The plan comes in the wake of a broader set of policies aimed at improving the UK’s housing stock and tackling ill health caused by poor quality housing. These include Awaab’s law and the Renters’ Rights Act, as well as updates to the Decent Homes Standard and its extension to the private rental sector. The recent budget has promised financial support for households in council tax bands A–D, as well as £100m for local authorities to spend on additional targeted support.  

The success of these efforts will depend on how effectively assistance can be delivered and new regulations enforced by resource-strapped local authorities. It also depends on identifying the most vulnerable households and effectively prioritising support. We propose that a data-informed understanding of fuel poverty is essential for effectively executing policy and meaningfully measuring progress, and that linked data can play a key role in targeting assistance. 

To support this, we used linked housing, socioeconomic and smart meter data to map fuel poverty risk across Great Britain. In this analysis, we explore:

  • how fuel poverty is currently measured
  • how health outcomes in areas at high risk of fuel poverty differ from other areas 
  • where in Great Britain is at the highest risk of fuel poverty 
  • the demographic, socioeconomic and housing characteristics of areas at high risk of fuel poverty.

We also provide recommendations for governments and local authorities to better measure fuel poverty and target support. 

 

Measuring fuel poverty

Current measures and their limitations

Measures of fuel poverty differ across the UK. In England, fuel poverty is currently measured using the Low Income, Low Energy Efficiency (LILEE) indicator adopted in 2019. Under this measure, a household is considered fuel poor if living in a property with an Energy Performance Certificate (EPC) rating of D or below with a disposable income below the relative poverty line following heating costs. 

However, this definition likely underestimates the full extent of fuel poverty across the country. By restricting fuel-poor status to only the most energy-inefficient properties, a household living in a property with an EPC of C or above cannot be considered fuel poor, regardless of income. Similarly, a household can only qualify as fuel poor under the LILEE indicator if heating costs push its income below 60% of the national median that year, effectively capping the proportion of households that can be considered in fuel poverty. 

Survey data suggest that households have significant difficulties heating their homes far beyond these parameters, with damaging effects on health and quality of life. Governments in Scotland, Wales and Northern Ireland have more expansive definitions of fuel poverty not limited by a property’s energy efficiency rating and using less restrictive income requirements.

Consistently measuring fuel poverty on a granular level – such as at the neighbourhood or household level – poses additional challenges. Fuel poverty may not be constant for many households but rather an intermittent reaction to changes in income, energy prices or energy needs. Existing subregional estimates of fuel poverty, such as those published by the Department for Energy Security and Net Zero, are typically modelled based on annual survey data collections gathered from a small sample of the population. This limits how well this data can be used for responsive and precisely targeted support. 

Measuring fuel poverty risk

Rather than use a binary definition of fuel poverty, in this analysis we aimed to understand the levels of fuel poverty risk affecting local areas across the country. To do this, we measured fuel poverty risk using the uZero dataset produced by UrbanTide. This data maps and quantifies fuel poverty risk for local areas across Great Britain using linked housing, socioeconomic and smart meter data. See Box 1 for greater detail on the data and methods used.

Data on fuel poverty risk were derived from the uZero dataset produced by UrbanTide. uZero is a data insights platform that quantifies fuel poverty risk at the Output Area and household levels across Great Britain using smart meter data provided by the Data Communications Company (DCC). These data are clustered using a machine learning algorithm to identify energy usage patterns indicative of fuel poverty among users of smart prepayment meters. Households that use smart prepayment meters have been observed to be disproportionately exposed to fuel poverty, and the energy usage patterns of those households can provide a bellwether for fuel poverty risk in a local area. 

Fuel poverty risk was then computed for Output Areas on a 1–100 scale based on a combination of:

  • the percentage of smart prepayment meter devices in an Output Area identified as fuel poor from clustering (both gas and electricity)
  • the percentage of all households in the Output Area on Universal Credit (from the Department of Work and Pensions)
  • the mean energy efficiency gap (difference between current and potential EPC) for all households in the Output Area.

While socioeconomic and energy efficiency variables still contribute to the measure, the introduction of smart meter data allows for a granular and timely detection of fuel poverty risk. 

We aggregated Output Area-level measures from uZero to Lower layer Super Output Area (LSOA) level measures by computing the mean fuel poverty risk score of Output Areas within each LSOA (for England and Wales) or data zone (equivalent geographic level in Scotland) in Great Britain. For the purposes of this analysis, we transformed the 1–100 uZero fuel poverty risk scale into deciles and quintiles marking the distribution of fuel poverty scores across LSOAs.

Average household incomes by postcode were also drawn from uZero and aggregated as the median household income for postcodes in an LSOA. Income was measured as gross income from all sources before any other costs. 

Data on lower-level geography populations broken down by age were drawn from the Office for National Statistics’ mid-2024 population estimates for LSOAs in England and Wales and from National Records of Scotland for data zones in Scotland. Data on LSOA populations broken down by ethnicity in England and Wales were drawn from the 2021 census. 

Data on the local prevalence of asthma, chronic obstructive pulmonary disease and poor mental health were derived from GP practice-level Quality and Outcomes Framework indicators for the financial year 2025/26. Data on prescriptions for inhalers and antidepressant and anti-anxiety medications were drawn from the OpenPrescribing database. In both cases, GP practice-level metrics were profiled to LSOAs based on the distribution of GP practice populations according to NHS England’s Patients Registered at a GP Practice dataset. These data were only available for England. 

We obtained data on secondary care events (inpatient admissions with respiratory, mental health or cardiovascular diagnoses) split by patient age band and LSOA of residence for the financial year 2025/26 from NHS England’s Hospital Episode Statistics – Admitted Patient Care dataset. Data on emergency care attendances split by patient age band and LSOA of residence for 2025 (the most recent available 12 months of data) were drawn from the Emergency Care Data Set. Age adjustment of secondary care metrics was performed through direct standardisation, using the mid-2024 England population split into four age bands as the reference population. Differences in rates of health events between high and low fuel poverty risk areas within the same income quintiles were computed using Mantel-Haenszel risk ratios. These data were also only available for England. 

For more information on data sources, methods and limitations, please see our Technical Appendix.

 

How health outcomes differ in areas at high risk of fuel poverty

Areas at the highest risk of fuel poverty tend to have higher rates of adverse respiratory, cardiovascular and mental health outcomes than areas at lower risk of fuel poverty. After accounting for differences in age distribution and income, Lower layer Super Output Areas (LSOAs) in England in the top quintile of fuel poverty risk had 21% more A&E attendances, 18% more respiratory-related inpatient admissions, and 6% more cardiovascular-related inpatient admissions per person than areas in the lowest quintile of fuel poverty risk (Figure 1). These areas also had a 24% higher prevalence of poor mental health observed in primary care data than areas at lower levels of fuel poverty risk. However, areas with the highest levels of fuel poverty risk were not found to have greater rates of asthma or chronic obstructive pulmonary disease, or of prescriptions for inhalers or mental health drugs after accounting for income differences.

Figure 1

These differences cannot be solely attributed to fuel poverty, given that we would expect fuel-poor status to correlate with a wide range of other facets of deprivation, such as greater exposure to environmental hazards. Ill health may also be a cause of fuel poverty in itself, due to its effect on employment and incomes. They nonetheless highlight the complex mix of challenges faced in areas at high risk of fuel poverty and the associated differences in health outcomes. However, even at similar income levels, areas with higher fuel poverty risk tend to have higher rates of respiratory and cardiovascular hospitalisations (Figure 2) and poor mental health than comparable lower risk areas. This suggests that fuel poverty risk may be associated with additional health challenges that warrant targeted support. 

Figure 2

 

Where in Great Britain is at the highest risk of fuel poverty

As our new map of fuel poverty shows, fuel poverty risk is not equally distributed across Great Britain. Close to one-third of LSOAs in Wales are in the highest quintile (top 20%) of fuel poverty risk, compared with around one-quarter of data zones (equivalent to LSOAs) in Scotland, and around one-fifth of LSOAs in England. 

Figure 3

Regionally, areas in the North West of England, Eastern and Southern Scotland, and North and South East Wales had the highest rates of fuel poverty risk, with over 30% of LSOAs in each area in the highest risk quintile (Figure 3). The West Midlands (29%) and Yorkshire and the Humber (27%) also had large proportions of LSOAs at the highest levels of fuel poverty risk, while the South West of England had the lowest proportion of high risk areas (7% in the highest quintile). 

While most high risk areas are urban, rural areas are over-represented among LSOAs with the highest fuel poverty risk. This is particularly acute in areas of Scotland classified as remote rural by the Scottish government, close to 60% of which rank in the top quintile of fuel poverty risk in Great Britain. This is in line with findings from the Scottish government regarding fuel poverty in rural areas and is likely linked to older housing stock and low gas mains coverage in the Highlands and island areas of the country.

 

Characteristics of areas at high risk of fuel poverty

Populations in areas at higher risk of fuel poverty tend to be younger than those in lower risk areas. Around 24% of the population in areas in the highest quintile of fuel poverty risk were aged 17 years or younger, compared with 19% in areas with the lowest fuel poverty risk. By comparison, 14% of the population in the highest risk areas were aged 65 years or older, compared with 21% in the areas at the lowest risk of fuel poverty.  

Areas at higher risk of fuel poverty tend to have greater proportions of residents with Asian or Asian British and black or black British backgrounds, based on the 2021 census. In England and Wales, people from Asian or Asian British backgrounds made up 18% of the population in areas at highest risk of fuel poverty, compared with 5% in the lowest risk areas. The corresponding figures for people from black or black British backgrounds were 8% and 2%. This reflects findings from our previous work, which similarly indicated that people with black or black British backgrounds are disproportionately exposed to fuel poverty, poor quality housing and damp and mould.

Figure 4

Fuel poverty risk is not limited to areas with the lowest incomes (Figure 4). While the median incomes of areas at the greatest risk of fuel poverty are typically lower than those in areas at less risk of fuel poverty, nearly 20% of areas at the highest risk of fuel poverty have median household incomes greater than the national median income. This demonstrates that areas at high risk of fuel poverty are socioeconomically diverse, in which areas which may not otherwise be at lower incomes can still be impacted by energy prices and household energy inefficiencies.    

Fuel poverty risk is also not only limited to the most energy-inefficient housing. More than one-third of areas at the highest risk of fuel poverty had median household EPC ratings of C or above, a markedly greater proportion than any other quintile. This is likely linked to the fact that social housing tends to be more energy efficient than any other housing tenancy type in the UK. But social housing residents are also disproportionately likely to be living in poverty and impacted by rises in the cost of living. This also demonstrates how EPC bands are a blunt instrument for comparing fuel poverty risk – a large majority of areas in all fuel poverty risk quintiles have median EPCs of D, limiting detectable variation between them.

 

Implications and recommendations

Failing to address fuel poverty will widen health inequalities

These findings highlight the stark inequalities behind the headline figures on fuel poverty. Areas at high fuel poverty risk tend to have younger populations, and people with Asian or Asian British and black or black British backgrounds appear to be disproportionately at risk. Vast regional disparities are also evident, with fuel poverty risk markedly more widespread in the North West of England, in Yorkshire and the Humber, and across Scotland and Wales than in the southern regions of England.

Areas at higher risk of fuel poverty tend to have higher rates of respiratory and cardiovascular hospitalisations and poor mental health than those at lower levels of risk, in line with the wealth of evidence on the links between fuel poverty and poor health. The younger age profile of areas at high risk of fuel poverty is a particularly strong warning signal, given the well-documented links between underheated homes, childhood respiratory illness and barriers to development.

Linked data can help target support 

The assistance and upgrades offered under the Warm Homes Plan must be targeted towards those in the greatest need. Reliable, up-to-date data is necessary to identify the most vulnerable households and deliver cross-sector action on recent policy promises. 

As our analysis demonstrates, smart meter data can be used to identify patterns of energy usage indicative of fuel poverty, providing a promising means to detect risk and to directly and dynamically administer support. While we drew on smart prepayment meter data, the DCC’s data covering over 30 million smart meters nationwide have plentiful further uses. These include wider detection of fuel poverty as well as humidity and damp. Unit rates for energy prices can also be directly adjusted for struggling households via smart meters, as recently proposed by the Resolution Foundation. 

However, smart meter data cannot be relied on alone. While around three-quarters of properties in Great Britain use smart meters, around 16 million do not. Some housing types that are less likely to have smart meters, such as older properties in remote rural areas, may also be particularly susceptible to fuel poverty. Further, while the DCC can provide anonymised smart meter data for Output Areas, government and local authorities currently do not have unrestricted access to household consumption data, which are subject to specific permissions and arrangements.

As an earlier NDL briefing showed, linked housing and health data also have a key role to play in identifying inequalities, targeting assistance and evaluating housing improvements. For example, NDL Cheshire and Merseyside developed a Need for Retrofit Index for the Liverpool City Region, identifying areas with high rates of health vulnerabilities associated with poor quality housing. This allows the local authority to pinpoint overlooked areas and directly intervene where support is most needed.

Retrofitting must be carried out with care and intention to avoid unintended health consequences. To produce a healthy home, retrofitting must equip a household to withstand both the cold and the heat. Without adequate ventilation, for instance, indoor air quality may suffer and induce further health issues, as well as contribute to household overheating against a backdrop of increasing summer temperatures. 

Definitions of fuel poverty must be reformed

Our analysis shows that it is not only the lowest income areas at risk of fuel poverty. Even more strikingly, more than one-third of areas in the highest quintile of fuel poverty risk have local median EPC ratings of C or above. This indicates that substantial proportions of households at risk of fuel poverty may not fall under the LILEE definition used in England. 

The Warm Homes Plan and Fuel Poverty Strategy for England pledge to lift 1 million households out of fuel poverty, primarily by upgrading homes to EPC band C and introducing new minimum energy efficiency requirements for landlords. They also commit to maintaining the LILEE measure of fuel poverty to track progress through 2030. However, our findings throw into question whether this measure can meaningfully capture progress. Under the current definition of fuel poverty, upgrading the energy efficiency of homes will reduce measured rates of fuel poverty in England irrespective of rises in energy costs or other squeezes on household budgets. The income stricture of the LILEE measure – requiring a household’s income to be below 60% of the national median following energy costs – will similarly make tracking progress challenging, given that it caps the number of households in fuel poverty based on the income distribution.

The Warm Homes Plan also proposes direct support for households in fuel poverty. But targeting assistance based on the LILEE metric risks overlooking many households most affected by cost-of-living pressures. Wider evidence shows that social housing residents are the most likely of all tenant groups to live in housing with EPCs of C or above, yet are disproportionately likely to be living in poverty.

To meaningfully capture progress and effectively target support, our understanding of fuel poverty must not be strictly limited by EPC brackets or capped by income distribution thresholds. While our analysis aimed to capture the full spectrum of fuel poverty risk across Great Britain, binary definitions may still be helpful for macro-level target setting, such as for fuel poverty reduction as part of the Warm Homes Plan. With this in mind, a UK-wide definition should be adopted to effectively coordinate policy and feed into energy pricing regulations, which are centrally managed for all of Great Britain. Closer alignment with the definitions used by the Welsh and Scottish governments offers a promising start. Smart meter data can also play a role in defining and measuring fuel poverty, allowing dynamic responses to within-year change rather than annual statistics.

 

Conclusion

Without substantial, targeted financial support and real improvements to housing, fuel poverty will continue to compound other disadvantages to widen the gap in health between the most deprived and most affluent areas. More general cost-of-living support must also play a role in alleviating fuel poverty by decreasing the financial trade-offs necessary for households. 

We propose refining the measurement of fuel poverty and leveraging linked housing and health data, including from smart meters. But this improved intelligence will only matter if it enables earlier action on improving homes and energy affordability and providing support with energy costs, along with evaluating that action’s effect on population health and inequalities.

The current government has pledged to provide ‘breathing space’ for households struggling with their finances. This is only a first step – any long-term solution must be accompanied by living spaces that no longer threaten our health. 

Technical Appendix: Rethinking fuel poverty in Great Britain

technical_appendix_-_rethinking_fuel_poverty_in_great_britain_-_where_are_risks_the_greatest.pdf
(1.45 MB)

We would like to thank UrbanTide for providing the data used to measure fuel poverty risk in this analysis, and Dr. Olivia Swann for reviewing and helping to improve this work. We would also like to thank Daniel Law, Alison Ismail, David Finch and Samantha Field for their contributions and comments on earlier drafts, as well as Chamut Kifetew, Tanjina Islam and Zoe Ruziczka for managing the NDL programme at the Health Foundation. This work uses data provided by patients and service users and collected by health and social care services as part of their care and support.

We would further like to thank Lisa Hornung for designing and producing the interactive map of fuel poverty released in tandem with this analysis, and for creating the data visualisations used throughout the text. We would also like to thank Jessica Clinkett, Nadia Kalam, Peter Stilwell, Alex Boyle and Emily Clements for refining the messages and leading on the communications approach of this piece, and Sarah Jost for copyediting the text.

Further reading

Related content