Key points

  • Households are considered by the Government to be in ‘fuel poverty’ if they would have to spend more than 10% of their household income on fuel to keep their home in a ‘satisfactory’ condition.  It is thus a measure which compares income with what the fuel costs ‘should be’ rather than what they actually are.  Whether a household is in fuel poverty or not is determined by the interaction of a number of factors, but the three obvious ones are:
    • The cost of energy.
    • The energy efficiency of the property (and therefore, the energy required to heat and power the home)
    • Household income.
  • All the points below relate to England only.
  • 4.0 million households in England were classified as being in fuel poverty in 2009 (18% of all households).  This is three time the number of households that were in fuel poverty at the low point in 2003, and there have been increases in each year since 2003.  It is, however, still lower than the number in the mid-1990s.
  • Fuel poverty is most common among those live in private rented accommodation: averaging from 2007 to 2009, 20% of households in private rented accommodation were in fuel poverty compared to around 15% in other tenures.
  • Despite their much lower average incomes, those in social rented accommodation are only a bit more likely to be in fuel poverty than owner-occupiers.  This is partly because very little social housing is energy inefficient (see the indicator on energy efficiency) and partly because social housing tends to be small, both factors meaning that relatively little fuel is required to keep the home warm.
  • Analysis of the relationship between fuel poverty and household income depends in part on whether the household incomes are analysed before or after deducting housing costs. 1  A further complication is that the relationship changes when levels of fuel poverty change substantially, as they have done in the last few years.  In this context, the focus of the points below is on the the qualitative patterns rather than the quantitative proportions.  Furthermore, in keeping with the approach used throughout this website, all the quantitative points relate to households grouped according to their incomes after, rather than before, deducting housing costs.
  • The risk of fuel poverty rises sharply as household income falls and, for example, very few households with above-average incomes are in fuel poverty: averaging across 2007 to 2009, around two-fifths of households in the poorest fifth after deducting housing costs were in fuel poverty.  Even so, a majority of households in the poorest fifth were not in fuel poverty and, furthermore, there were a substantial number of households who are not in the poorest fifth but who are nevertheless in fuel poverty (around half of the total number in fuel poverty).  Clearly, therefore, there are factors other than household income which affect whether a household is in fuel poverty or not.
  • One such factor is how energy efficient the home is.  For example, households not in the poorest fifth but in very energy inefficient homes are actually more likely to be in fuel poverty than households in the poorest fifth but in homes with above-average energy efficiency.  More specifically, averaging across 2007 to 2009, around two-fifths of households not in the poorest fifth after deducting housing costs but living in homes with a SAP rating (see the indicator on energy efficiency for a definition) of less than 30 were in fuel poverty compared with around a third of households in the poorest fifth but living in homes with a SAP rating of 50 or above.  One result of this is that households who are both in the poorest fifth and in very energy inefficient homes are at a very high risk of fuel poverty (85%, averaging across 2007 to 2009).
  • A second such factor is the composition of the household.  Single-person households – working-age singles as well as single pensioners – are more likely to be in fuel poverty than either couples or larger families.  Overall, averaging across 2007 to 2009, around 35% of single pensioners and 25% of working-age singles were in fuel poverty compared to around 20% of lone parents (the next highest group), 15% of pensioner couples, and 5% of working-age couples.  Because of their relatively high risk, half of all the households in fuel poverty in England are single-person households even though only a quarter of all households are single-person households.
  • Among those in low income, single-person households are also more likely to be in fuel poverty than either couples or larger families and this applies particularly to working-age singles (rates for pensioner singles are a bit lower).  The analysis in Cold and poor: an analysis of the link between fuel poverty and low income suggests that the reason for the high risk of fuel poverty among single-person households, both overall and among those in low income, is that, whereas their estimated fuel costs tend to be a bit lower than those for other household types, their household incomes tends to be a lot lower.  In other words, fuel costs tend to be a bigger burden, relative to incomes, for single-person households than for larger households.
  • Finally, households in rural areas are more likely to be in fuel poverty than those in urban areas (averaging across 2007 to 2009, 20% in rural compared with 15% in urban), with similar proportional differences those in low income (for the poorest fifth, averaging across 2007 to 2009, 40% in rural compared with 40% in urban)
  • The analysis above suggests that two of the major groups of concern from a fuel poverty perspective relate to those in low income who are either single-person households of working age and or who live in rural areas.  This is notable because these two groups have not been the focus of government’s more general anti-poverty strategy, which has tended to focus on children, older people and deprived urban areas.
  • Within England, fuel poverty is most prevalent in West Midlands and the North East.

Why this indicator was originally chosen

People who are in fuel poverty are either paying a high proportion of their income for the essential purposes of keeping their homes warm, cooking, etc or are not using the amounts of fuel that are required to keep their home in a satisfactory living condition.

Definitions and data sources

The first graph shows the number of households deemed to be in ‘fuel poverty’, with the data shown separately by tenure.

The second graph shows the proportion of households in fuel poverty in each tenure.

Households are considered to be in ‘fuel poverty’ if they would have to spend more than 10% of their household income on fuel to keep their home in a ‘satisfactory’ condition, where, for example, a ‘satisfactory’ heating regime is considered to be one where the main living area is at 21 degrees centigrade with 18 degrees centigrade in the other occupied rooms.  It is thus a measure which compares income with what the fuel costs should be rather than what they actually are.  Household income is disposable household income before deducting housing costs, with Housing Benefit and Income Support for Mortgage Interest both counted as income.  The fuel costs included comprise that used for space heating, water heating, lighting, cooking and household appliances.

The third graph to sixth graphs show how the proportion of households that are in fuel poverty varies by the income of household.  In the third graph, the basic statistics by household income quintile are presented.  In the fourth graph, the data is broken down by the energy efficiency of the home (for the definition of energy efficiency, see the indicator on energy efficiency).  In the fifth graph, the data is broken down by household type (single person, couple with children, etc).  In the sixth graph, the data is broken down by type of area (rural, urban, etc), using the government’s 2004 classification system for small areas whereby household living in settlements of more than 10,000 are classified as ‘urban’ and those who are not are classified as ‘rural’.  To improve their statistical reliability, the data is the average for the latest three years.

The allocation of households to the poorest fifth uses ‘equivalised household income’ after deducting housing costs, which means that the household incomes have been adjusted to put them on a like-for-like basis given the size and composition of the households.  This means that the results are somewhat different than those in some other publications which either use unadjusted household incomes or incomes before deducting housing costs. 2

The seventh graph shows how the proportion of households in fuel poverty varies by region. 3

The data source for all the graphs is the stock dataset from the English Housing Survey (EHS) and relates to England.  The 1996 data in the first graph has been amended by the government from their original estimate of 4.3 million to take account of DTI gas and electricity bill data and is taken from the 3rd annual progress report of the UK fuel poverty strategy.

Overall adequacy of the indicator: medium.  EHS is a well-established, regular government survey, designed to be nationally representative, but the calculation of required fuel costs is both complex and obscure.

External links

Relevant 2007 Public Service Agreements

Overall aim:  Increase long-term housing supply and affordability

Lead department

Department for Communities and Local Government

Official national targets

Increase the number of net additional homes provided per annum to 240,000 by 2016.

Increase the number of gross affordable homes provided per annum to 70,000 by 2010-11 including 45,000 social homes.

Halve the number of households in temporary accommodation to 50,500 households by 2010.

By March 2011, 80% of local planning authorities to have adopted the necessary Development Plan Documents, in accordance with their agreed Local Development Scheme.

Other indicators of progress

Trends in affordability.

Efficiency rating of new homes.

Previous 2004 targets

By 2010, bring all social housing into decent condition with most of this improvement taking place in deprived areas, and for vulnerable households in the private sector, including families with children, increase the proportion who live in homes that are in decent condition.

Eliminate fuel poverty in vulnerable households in England by 2010 in line with the Government’s Fuel Poverty Strategy objective Joint with the department for Trade and Industry.

The numbers

Graph 1

Year Millions
Private rented Social rented Owner occupied Total
1996 not available not available not available 5.1M
2001 0.3M 0.3M 1.2M 1.7M
2003 0.2M 0.2M 0.8M 1.2M
2004 0.2M 0.2M 0.8M 1.2M
20050.2M 0.2M 1.1M 1.5M
2006 0.4M 0.4M 1.7M 2.4M
20070.5M 0.5M 1.9M 2.8M
20080.6M 0.6M 2.1M 3.3M
20090.7M 0.8M 2.5M 4.0M

Graph 2

Private renters Social renters Owner occupiers
20% 17% 15%

Graph 3

Poorest fifth 41%
2nd 25%
3rd 12%
4th 4%
Richest fifth 0%

Graph 4

SAP ratingIn the poorest fifthNot in the poorest fifth All households
SAP rating under 3085% 39% 47%
SAP rating 30-4065% 18% 24%
SAP rating 40-5053% 10% 17%
SAP rating 50-6040% 7% 13%
SAP rating 60-7025% 3% 7%
SAP rating over 7012% 1% 4%
All households41% 10% 16%

Graph 5

GroupIn the poorest fifthNot in the poorest fifth All households
Single people under 6062% 12% 24%
Single people aged 60 or over50% 33% 34%
Couples aged 60 or over51% 10% 15%
Couples under 60 with no dependent children43% 2% 6%
Lone parents31% 6% 18%
Couples with dependent children25% 2% 7%
All households41% 10% 16%

Graph 6

Type of area In the poorest fifth All households
Rural52% 20%
Urban39% 15%

Graph 7

East 13%
East Midlands 18%
London 11%
North East 21%
North West 19%
South East 10%
South West 15%
West Midlands 22%
Yorkshire and The Humber 18%
1. Prior to the latest 2006 data, only before deducting housing cost incomes could be used (because the relevant survey did not collect data on housing costs) but, from 2006, either before or after deducting housing cost incomes can be used. 
2. It could be argued that, because the definition of fuel poverty uses household incomes before deducting housing costs, then the allocation of household to income quintiles should also use the same before deducting housing cost incomes.  Our response is that a) when allocating households to income groups, we use after deducting housing cost incomes throughout this website (for the reasons set out on the page on choices of low-income thresholds) and b) when analysing any particular subjects (such as fuel poverty), we always stick to the official Government definitions.  So, from this perspective, the issue is more for the Government (i.e. should they change the definition of fuel poverty so that it uses household incomes after deducting housing costs?) than for us.  Note that such a change would not be the same as using the ‘basic income definition’ of fuel poverty as this definition does not deducting rent, mortgage interest, etc from the household incomes.  Finally, it is worth noting that, whilst the specific proportions in the graphs would sometimes be materially different if the income groupings were based on before deducting housing cost incomes, their overall patterns would generally be similar to those presented in the graphs using after deducting housing cost incomes. 
3. Note that equivalent data is not available at a sub-regional level because the sample size of the relevant survey (the English Housing Survey) is totally insufficient to derive sub-regional estimates.  Any sub-regional estimates published by academics are the output of computer models and are NOT based on actual data.