Fuel Poverty Measurement in Europe:
A rapid review of existing knowledge and
approaches conducted for eaga Charitable Trust
Harriet Thomson
October 2013
1. INTRODUCTION
Using academic, policy and ‘grey’ literatures this review will provide a concise outline of the range
of approaches that have been used to measure fuel poverty across Europe, and will debate the
suitability and transferability of these approaches to different contexts. A key focus is on how
previous surveys have attempted to capture different elements of fuel poverty, such as household
difficulties in affording adequate energy services, the use of coping strategies, and the
measurement of income and fuel expenditure. The review will focus exclusively on the
measurement of fuel poverty levels at the national and pan-European scale, rather than discussing
the identification of households at the local level for targeting purposes. The review will
commence with a brief discussion of the definitional issues, with a particular focus on
terminological disparities at the European level.
2. DEFINING FUEL POVERTY
Conceptualising fuel poverty is a necessary first step prior to devising measurement indicators.
Indeed as Boardman remarks, “who is fuel poor depends on the definition; but the definition
depends on who you want to focus on and this involves political judgement” (Boardman, 2010: 21).
However, as will become apparent over the following discussion, only a minority of countries have
adopted formal definitions of fuel poverty, whilst in some other member states the governments
remain unwilling to acknowledge the existence of fuel poverty issues in their country (Bouzarovski
et al., 2012), thus hindering a rigorous assessment of the problem.
2.1 Terminology
In the United Kingdom, it is commonplace to refer to the phenomenon of being unable to afford
adequate energy services as fuel poverty, with Liddell et al. (2012) holding the view that this is the
most commonly accepted term throughout the industrialised world (Liddell et al., 2012). However,
at the European and global scale there is an inconsistent use of terminology (Thomson and Snell,
2013), with the term energy poverty sometimes used interchangeably with fuel poverty, whilst at
other times it is used to conceptualise a more extreme set of circumstances.
Page 2 of 47
Overall, three key terminological standpoints can be identified in the literature: firstly the terms can
be understood to mean the same thing (Boardman, 2010), and indeed, the terms have been used
interchangeably in a number of key EU policy documents (for example, European Parliament, 2010;
European Commission, 2010a; 2010b; European Economic and Social Committee, 2011). Secondly,
they can be considered as distinct terms, with energy poverty referring to the lack of access to
modern energy services in developing countries (as used by Bazilian et al. (2010); Birol (2007) and
Sagar (2005)), and fuel poverty referring to “a problem of affordability rather than access, which is
present in some of the world's most developed countries” (Househam and Musatescu, 2012: 2).
Lastly, the two terms can be considered as related concepts, with the distinction being the fuel
types covered by each term, for example, as the European Commission states “The energy sources
covered by the term fuel poverty…are broader than those considered in the energy poverty
references in the internal energy market legislation (electricity and gas)” (European Commission,
2010a: 10).
2.2 Member State level definitions of fuel poverty
As stated earlier, only a minority of countries have adopted formal definitions of fuel poverty; for
example, across the twenty-eight member states of the European Union, only France, the Republic
of Ireland and the United Kingdom have official definitions of fuel poverty (see Table 1). In addition,
Romania is in the early stages of creating a national definition of fuel poverty through the United
Nations Development Programme project ‘Improving Energy Efficiency in Low-Income Households
and Communities in Romania’ (see Househam and Musatescu, 2012). Elsewhere, fuel poverty does
not rank high on the political agenda, as confirmed by the decision makers and experts that
Bouzarovski et al. interviewed in Brussels who stated that some member states are unwilling to
accept fuel poverty exists in their country (Bouzarovski et al., 2012: 78).
Country
Definition of a fuel poor person/household
France
[précarité énergétique] “If he/she encounters particular difficulties in
his/her accommodation in terms of energy supply related to the
satisfaction of elementary needs, this being due to the inadequacy of
financial resources or housing conditions” (translation of Plan
Page 3 of 47
Bâtiment Grenelle, 2009: 16).
Republic of
“A household is considered to be energy poor if it is unable to attain
Ireland
an acceptable standard of warmth and energy services in the home at
an affordable cost…a household is defined as being unable to afford
its energy needs if it is required to spend at a level greater than twice
the national average (median) share (currently 10%) of disposable
income spent on energy services to achieve an acceptable standard of
warmth” (Department of Communications, Energy and Natural
Resources, 2011: 12)
United Kingdom
Previous definition: “One that cannot afford to keep adequately warm
at reasonable cost. The most widely accepted definition of a fuel poor
household is one which needs to spend more than 10% of its income
on all fuel use and to heat its home to an adequate standard of
warmth.” (Department of Trade and Industry, 2001: 6).
New definition: a household is fuel poor if:
• “Their income is below the poverty line (taking into account energy
costs); and
• Their energy costs are higher than is typical for their household
type” (Department of Energy and Climate Change, 2013: 11)
Table 1: Summary of existing national definitions of fuel poverty
Of the three countries that have official definitions, the United Kingdom is the most advanced in
defining and measuring fuel poverty. This is principally as a result of Brenda Boardman’s seminal
monograph in 1991, which provided the foundations for the ’10 per cent’ definition, and more
recently, as a consequence of the Government-commissioned review of fuel poverty by Professor
John Hills (Hills, 2012), which led to a revision of the UK definition. By comparison, France and
Ireland were much later adopters of a fuel poverty definition, for example as Dubois (2012)
outlines, the formal definition of “energy precariousness” (précarité énergétique) in France
emerged in 2007 as a result of the ‘Grenelle de l'environnement’ (see Plan Bâtiment Grenelle,
2009), which were multi-party roundtable discussions about various aspects of the environment.
The Republic of Ireland followed a slightly different trajectory, with a Private Members Bill in 2008
entitled the ‘Fuel Poverty & Energy Conservation Bill’, which called on the Government to develop a
Page 4 of 47
national fuel poverty strategy. However this was defeated in a vote (Public Health Policy Centre,
2008). Despite this set back, an Inter-Departmental/ Agency Group on Affordable Energy was
formed in 2008, which led to the establishment of Ireland’s first strategy on affordable energy and
energy poverty in 2011 (Department of Communications, Energy and Natural Resources, 2011a),
which outlined how to define and measure energy poverty, as well as what policy would be
implemented to alleviate energy poverty.
2.3 Pan-EU definition of fuel poverty
At the European level arguably less progress has been made to address fuel poverty; policy
responses have been piecemeal and limited (Thomson and Snell, 2013) and no pan-European
definition of fuel poverty or energy poverty exists. This is despite entry of the terms fuel poverty
and energy poverty into European policy literature more than a decade ago in 2001 and 2002
respectively (see European Coal and Steel Community Consultative Committee, 2001; European
Commission, 2002), and subsequent high profile discussions on fuel poverty issues during the
preparatory stages of the 2009 European Directives concerning the internal markets in electricity
and natural gas (2009/72/EC and 2009/73/EC respectively). Indeed, one of the core legislative
institutions of the EU, the European Parliament, regularly petitioned the European Commission
during 2008 to define energy poverty (European Parliament 2008a; 2008b; 2008c) , and proposed
several amendments to the draft 2009 Directives, including the following which introduced a
description of energy poverty and mandated member states to recognise it:
“38. “energy poverty” means the situation where the members of a household cannot afford to
heat their home to an acceptable standard based on the levels recommended by the World
Health Organisation” (European Parliament, 2008c: 9)
“(d) paragraph 5 shall be replaced by the following:
‘5. Member States shall take appropriate measures to protect final customers, and shall, in
particular, ensure that there are adequate safeguards to protect vulnerable customers, including
prohibiting the disconnection of pensioners and disabled people in winter. In this context,
Member States shall recognise energy poverty as defined in Article 2(38) and shall provide
definitions of vulnerable customers” (European Parliament, 2008c: 10)
Page 5 of 47
However, the European Commission has remained opposed to establishing a pan-European
definition and to offering basic descriptions of fuel poverty/energy poverty and vulnerable
customers, stating:
“Given the diverse situations of energy consumers in different parts of the EU, the Commission
does not consider it appropriate at this stage to propose a European definition of energy poverty
or of vulnerable customers” (European Commission, 2010a: 12)
Thus in the main pieces of legislation currently governing fuel poverty at the member state level, no
definition, basic description or guidance is provided on energy poverty or vulnerable customers.
The consequences of this are far reaching; firstly, it means that fuel poverty is not recognised as a
policy problem, as evidenced by the lack of national definitions of fuel poverty/energy poverty in
the majority of member states, and subsequently the lack of intermediate targets to alleviate fuel
poverty. Secondly, it reduces the likelihood that the overlap between fuel poverty policy and other
policy domains, such as climate change, will be sufficiently recognised by policymakers, in spite of
research that states the importance of recognising the relationship between fuel poverty and
climate change goals (Snell and Thomson, 2013; Ürge-Vorsatz and Tirado- Herrero, 2012). Lastly, it
leads to a fragmented approach to addressing fuel poverty, with the European Economic and Social
Committee commenting that, “Not all Member States are addressing this problem and those that
are, act on their own, without seeking synergies with others, which makes it harder to identify,
assess and deal with energy poverty at the European level” (European Economic and Social
Committee, 2011: 4). Furthermore, Bouzarovski et al. argue that the lack of a common approach at
the European level has “hampered the adoption of unified monitoring and evaluation
methodologies” (Bouzarovski et al., 2012: 78).
3. MEASURING FUEL POVERTY
Moving on from the issues of defining fuel poverty, this next section will consider what approaches
have been used to date in measuring fuel poverty across various European countries. This review
focusses on the measurement of fuel poverty levels at the national and pan-European scale for
monitoring purposes, rather than small-area identification of fuel poor households for targeting
Page 6 of 47
purposes. The following three approaches to measuring fuel poverty will be discussed:
temperature, expenditure, and consensual.
3.1 Temperature approach
The temperature approach, whereby internal room temperatures are collected to determine if
households are attaining ‘adequate’ levels of warmth, is rarely used to measure fuel poverty at the
national level, and has never been employed at the European level. As Healy explains, whilst the
temperature approach seems simple in theory it is problematic for a number of reasons, “chiefly
because of the inadequacy and unreliability of data on household temperatures, while intermittent
occupancy may also distort results using this approach” (Healy, 2004: 35). Nevertheless, the
temperature approach does merit some discussion as it can be useful as a secondary approach used
in conjunction with other indicators.
A central issue when using a temperature approach is deciding on what temperature thresholds to
use. Boardman states that the temperatures needed to provide adequate warmth are “21°C in the
living room and 18°C elsewhere, when these rooms are occupied” (Boardman, 2010: 174). This
temperature range is used in the English, Welsh and Northern Irish fuel poverty models, and
originates from standards established by the World Health Organization (1987). By comparison,
Scotland uses a slightly higher threshold of 23°C in the living room for disabled, infirm and elderly
households (Boardman, 2010), reflecting the additional heating requirements that these
households may require. However, the 18°C to 21°C range is not accepted by all as the correct
range, with some level of debate as to what constitutes a comfortable internal temperature. As
Healy states “many physiological, psychological and environmental variables play a part in humans’
perception of thermal comfort” (Healy, 2004: 131), whilst Critchley et al. (2007) found that some
householders preferred temperatures lower than the prescribed limits due to various reasons other
than financial constraints.
Page 7 of 47
Large scale empirical temperature data is scarce at the national level, particularly as the English
House Condition Survey stopped taking spot living room temperatures in 1996 (Boardman, 2010);
since then only a limited number of studies have been conducted. For example, Oreszczyn et al.
(2006) conducted a comprehensive study of internal living room and bedroom temperatures in
1,604 English houses, taking half-hourly readings for two to four weeks across two winters
(Oreszczyn et al., 2006: 246). These measurements were collected from low-income households
that were receiving energy efficiency improvements to their property through the Warm Front
scheme. The data produced in this study is more reliable than the EHCS temperature data as it
collected around one thousand data points per dwelling (Oreszczyn et al., 2006) compared with the
EHCS’s single data points.
Healy and Clinch (2002b) have also conducted research into internal room temperatures via their
national household survey of fuel poverty and thermal comfort in the Republic of Ireland. In total,
1,500 households were recruited by random probability-based sampling, and were questioned
about their ability to heat their home adequately and had their living-room temperature
measurements taken. Healy and Clinch found that 29.4% of fuel poor households had a living-room
temperature of 18°C or less, compared with just 8.8% of other households (Healy, 2004: 134).
However, they also discovered that a mismatch exists between households classified as fuel poor
and those that are enduring thermal discomfort, stating “not all those who are defined in the study
as fuel-poor can be classified as enduring thermal discomfort” (Healy, 2004: 135). Healy argues that
this highlights the problems associated with using living room temperatures as an indicator of
thermal comfort, particularly as social desirability bias may cause households to heat the living
room to a higher level than normal in anticipation of the interview (Healy, 2004: 134). Furthermore,
in the context of Central and Eastern European countries where many dwellings are served by
district heating systems that do not allow individuals to control their heat consumption, it is argued
by Tirado Herrero and Ürge-Vorsatz (2012) that indoor temperatures are not a good indicator of
fuel poverty as the internal temperatures are “typically adequate, or in cases even too high” (Tirado
Herrero and Ürge-Vorsatz, 2012).
Page 8 of 47
3.2 Expenditure approach
The most commonly used fuel poverty measurement approach is the expenditure approach, which
explores the ratio of household income to fuel expenditure. Broadly speaking, under an
expenditure definition, a household is considered to be fuel poor if they spend more than X% of
their income on fuel (Healy, 2004). Within this approach, there are two main considerations to be
made at the outset: firstly, whether to use an absolute or relative expenditure threshold, and
secondly, whether to use actual or required fuel expenditure data, although this second point is
likely to be driven by data availability. Following on from these decisions, there are a number of
other elements that require thought, namely issues around measuring household income, and
calculating household energy requirements.
Expenditure thresholds
As with the measurement of income poverty, whether to use a relative or absolute fuel poverty
threshold is a contested topic. Absolute measures of fuel poverty increase in line with rising fuel
prices, and make the eradication of fuel poverty a possibility (Boardman, 2012), whereas fuel costs
under a relative threshold are “determined relative to the median cost to income ratio for all
households” (2012: 21). Moore states that a relative threshold is problematic in practice for the
measurement of fuel poverty as, unlike incomes, fuel prices do not remain static, “reflected in the
increasing median % of income required for fuel by all households, but not by the number in relative
fuel poverty” (2012: 21). On the other hand, whilst acknowledging that a varying proportion of
income will make policy more complex due to the potential annual changes, Boardman states that
it does “accurately reflect the level of a relative hardship faced by the fuel poor” (Boardman, 2010:
231). Indeed, the basis for the UK’s 10 per cent absolute threshold was that it represented twicemedian expenditure (Boardman, 2010), and as will be outlined shortly, the Republic of Ireland has
set three thresholds for fuel poverty that relate to twice, three times and four times median
expenditure (see Department of Communications, Energy and Natural Resources, 2011b).
The choice of median rather than mean expenditure is an important distinction to draw attention
to. As highlighted by Fahmy (2011) and Moore (2012) fuel expenditure is asymmetrically
Page 9 of 47
distributed, thus the use of the mean can be misleading as it gives weight to ‘atypically’ high values.
Liddell et al. argue that the smoothing out of extreme values is one of the reasons why medians
“are internationally favoured for representing distributions related to income and expenditure”
(Liddell et al., 2012: 27). They further argue that conceptualising fuel poverty in median terms has
been useful for comparing prevalence across countries “since it absorbs real variations in the
amounts which residents of very diverse countries customarily pay for heat, power and light”
(Liddell et al., 2012: 27).
In terms of choosing between actual or required fuel expenditure, actual fuel expenditure is easier
to calculate than required fuel expenditure as the latter “requires a detailed knowledge of the
energy efficiency of the housing stock” (Moore, 2012: 3), however, it is regarded as a poor
indication of fuel poverty (Moore, 2012; Liddell et al., 2012), particularly as both the 1991 and 1996
English House Condition Surveys showed that low income households often spent significantly less
on fuel than would be required to maintain a warm home (Moore, 2012). Indeed, a comparison of
actual fuel expenditure from the UK Living Costs and Food Survey and modelled required
expenditure showed that households in the lowest income decile group only spent around 62 per
cent of the required spend to maintain adequate warmth (Department of Energy and Climate
Change, 2011: 69). Furthermore, the use of actual fuel expenditure may be problematic in several
member states where meter readings may be as infrequent as every two years (Darby, 2012).
Required fuel expenditure, as is currently modelled for UK fuel poverty statistics, is considered to
be more meaningful (Moore, 2012), but, as Moore highlights, “The UK is almost unique in having a
series of large national house condition surveys that enable such fuel costs to be accurately
determined and compared directly with corresponding household incomes” (Moore, 2012: 3),
limiting the application of a required expenditure approach in other European countries.
Household income
In both required and actual fuel expenditure models, an accurate assessment of household income
is needed. However, as Boardman notes in relation to the UK “this part of the definition has
remained controversial, particularly in relation to housing costs” (Boardman, 2010: 24). Overall,
Page 10 of 47
there are three main contentions: firstly, whether to use a before housing costs or after housing
costs measure, secondly, and related to the first point, is what benefits should be included as
disposable income, and lastly, whether income should be equivalised. Equivalising incomes,
whereby incomes are adjusted to “reflect the fact that larger households need a higher income
than smaller households to achieve a comparable standard of living” (Moore, 2012: 20), is a
debated topic. Hills states that equivalisation seeks to “set al.l households sizes and types on the
same basis compared to a standard” (Hills, 2012: 35), but in effect this makes it appear that “small
households have a higher income and that large families have less” (Boardman, 2010: 32), which
may mean that the measure favours large families. Boardman argues that equivalent income may
not be entirely appropriate for fuel poverty analysis (Boardman, 2010: 32), and in addition, cannot
be used on the doorstep to check household eligibility for free energy efficiency improvements
(Boardman, 2010: 32).
Including disability related benefits, such as the UK’s Disability Living Allowance (DLA), in disposable
income calculations is controversial as these benefits exist specifically to “help with the extra costs
caused by a disability” (Department for Work and Pensions, 2013). Some argue that their inclusion
exaggerates disabled people’s incomes and artificially pushes some above the poverty threshold
(Bevan Foundation, 2009; Parckar 2008). Indeed, in his review of UK fuel poverty, Hills argues that
“classifying DLA as general income for measuring fuel poverty implicitly assumes that its recipients
are better off than those who do not receive it” (Hills, 2012: 92). Hills further states “removing DLA
from the income calculation would be appropriate, reflecting more general arguments about the
way in which its inclusion leads to understatement of the proportion of disabled people who have
low incomes” (2012: 92). Analysis of English Housing Survey data from 2010/2011 by Thomson et
al. (2013), demonstrates that the level and depth of fuel poverty does indeed increase significantly
when DLA is excluded from income calculations.
The remaining point of contention is whether to measure income before housing costs or after
housing costs. Within England, two income definitions relating to housing costs are used in fuel
poverty statistics. These are:
Page 11 of 47

Basic income: includes all income, but excludes income related directly to housing (i.e.
after housing costs)

Full income: the basic income plus all benefits relating to housing, including housing
benefit, income support for mortgage interest and council tax benefit (i.e. before housing
costs)
(Boardman, 2010: 29)
As Moore states, the case for excluding housing costs from income in the definition of fuel poverty
appears self-evident as “Households cannot spend their housing costs on fuel, any more than they
can so spend the national and local taxes which are specifically excluded from income” (Moore,
2012: 20). Furthermore, the use of a full income (before housing costs) measure can lead to an
anomalous situation whereby an increase in rent can lift a low-income household in receipt of
housing benefit out of fuel poverty. Boardman explains this is because their housing benefit will
increase to match the rise in rent, and thus, their total income will rise (Boardman, 2010: 29). Hills
recommended that the British government measure income for fuel poverty purposes after housing
costs (Hills, 2012), however, the government favours a full income measure as the basic income
risks including too many richer households (Boardman, 2010: 29). Ultimately, the decision about
housing costs is “a political decision primarily about who should be helped most” (Boardman, 2010:
30).
In the context of European fuel poverty measurement, it becomes apparent that not all data
sources will be sophisticated enough to disaggregate income in the ways suggested above. For
example, in his analysis of ECHP data, Healy (2004) analysed the household’s main income source
as a proxy for income amount, stating that they are a good indication of income level (Healy, 2004:
56). In the ECHP, six categories of income source were available: wages (employee); selfemployed/farming; pensions; unemployed; other social transfers; and private (Healy, 2004). Healy
argued that a dependence on unemployment assistance or other social transfers “implies that such
Page 12 of 47
households live on a modest level of household income. Such income may preclude households
from heating their home adequately” (Healy, 2004: 56).
Household energy requirements
In the United Kingdom, modelled required energy consumption takes into account the energy
required for space heating, water heating, lights and appliances, and cooking (Department of
Energy and Climate Change, 2010). However, this is a complex model that needs to account for the
dwelling characteristics, as well as the household’s specific energy requirements. As the British
Department of Energy and Climate Change outline, the amount of energy required to heat a
property will depend on elements such as “insulation levels, heating systems, the geographical
location of the dwelling and construction type” (2010: 27). By comparison, household energy
demand depends on “the number of people within the household and the lifestyle and habits of
these individuals” (ibid), in addition to the hours of heating and the proportion of the house to be
heated (Boardman, 2010).
With regard to space heating and required internal temperatures, as discussed earlier in the
temperature approach section, 21°C in the living room and 18°C elsewhere is the normative
temperature range, and originates from standards established by the World Health Organization
(1987). This range is used in the English, Welsh and Northern Irish fuel poverty models. Scotland is
the exception in the UK as it uses a slightly higher threshold of 23°C in the living room for disabled,
infirm and elderly households (Boardman, 2010), reflecting the additional heating requirements
that these households may require. Indeed, a review by Snell and Bevan (2013) highlighted the
additional energy needs that some disabled households may require, including “a need for higher
indoor temperatures for longer periods, the use of energy intensive equipment, increased laundry
requirements” (Snell and Bevan, 2013: 11).
Occupancy is another determinant of household energy needs, and is a two-fold element. Firstly,
the hours of occupancy during the day will determine how many hours of heating a household is
Page 13 of 47
likely to require. Table 2 displays the standard format for asking a household how long they require
the heating to be on during winter, and is used in the English Housing Survey, Northern Ireland
House Condition Survey, and the Living in Wales Survey.
Generally speaking, during winter when heating needs are greatest, when would you or someone
else in your household have your heating on to stay warm?
All day/all the time
Weekday evenings
Weekday morning
Weekend daytimes
Weekday lunchtime
Weekend evenings
Weekday afternoon
Don’t know
Table 2: Standard question for winter heating requirements
Secondly, in broader terms, the UK government assess whether a property is under-occupied, that
is, if there are both surplus bedrooms and floor area. If a dwelling is under-occupied, then it is
“assumed that some of the rooms in the dwelling are not heated and a “half-house” heating regime
is applied” (Department of Energy and Climate Change, 2010: 32). However, Moore (2012) has
voiced concerns that the half-house heating regime “may be insufficient to prevent condensation
and mould growth in unheated rooms” (Moore, 2012: 20).
Household needs will also vary depending on the energy efficiency of the dwelling, particularly in
terms of the fabric of the dwelling, the efficiency of its energy systems, and the types of fuel
available to the household. The approach used in the United Kingdom relies on detailed
information to be collected about all aspects of the dwelling, for instance, to calculate the heated
volume and heat loss areas, information is collected on:

Internal & external wall areas

Roof area

Room specific floor areas

Habitable floor area and footprint area

Perimeter of building

Ceiling height

Window areas
Page 14 of 47

Number of floors and rooms in a dwelling
Department of Energy and Climate Change, 2010: 28
As may be apparent from the UK-centric nature of this section, knowledge and expertise in
modelling required fuel expenditure is centred in the United Kingdom; no other European country
collects such detailed data nor conducts in depth modelling. Nevertheless, emerging literature from
other countries shows that in developing a pan-European approach to measuring fuel poverty,
country specific contexts will need to be acknowledged. For example, Tirado Herrero and ÜrgeVorsatz (2012) demonstrate that particular attention needs to be paid to the characteristics of
district heating systems as many do not allow individual dwellings to regulate the temperature or
timings. Such systems are widespread in Central and Eastern European countries, leading to Tirado
Herrero and Ürge-Vorsatz (2012) identifying a new variant of fuel poverty, whereas they are
scarcely used in the United Kingdom.
Previous assessments of fuel poverty using the expenditure approach
There have been various studies that have used an expenditure approach to quantify fuel poverty
in European countries. To date, expenditure-based assessments of fuel poverty have been made in
Austria, Belgium, France, Hungary, the Republic of Ireland and the United Kingdom. Starting with
the United Kingdom, Isherwood and Hancock (1978) are credited with being among the first to
define the issue of fuel poverty (Osbaldeston, 1984). Their analysis of the 1977 Family Expenditure
Survey resulted in high fuel expenditure being defined as “those spending more than twice the
median (i.e. 12%) on fuel, light and power” (Isherwood and Hancock (1979) cited in Osbaldeston,
(1984: 368)). However, it was not until 1991 that the issue of fuel poverty measurement was
formalised in the UK, with the publication of Brenda Boardman’s monograph (1991), which
provided the foundations for the present day definition and measurement of fuel poverty.
Boardman found that the poor spent twice as much on fuel, as a proportion of income, than the
rest of the population (Boardman, 1991), and determined that households unable to achieve an
adequate level of energy services for ten per cent of income are fuel poor, which represented
around 6.6 million households (Boardman, 1991: 207).
Page 15 of 47
Reflecting the advanced nature of fuel poverty research in the UK, the Department of Energy and
Climate Change have been officially monitoring fuel poverty rates for a number of years, using a
complex algorithm to calculate the level of expenditure on energy required in order to achieve an
adequate standard of warmth and use of lighting and electrical items. The modelling is based on
detailed house condition survey data, and takes into account a number of factors such as the size
and energy efficiency of the property, and the number of occupants (see Department of Energy and
Climate Change, 2010 for further details). Table 3 outlines the trends in official fuel poverty rates in
the UK from 1996 to 2010. As can be seen, fuel poverty rates were highest in 1996, with 6.5 million
households classified as being fuel poor, but between 2001 and 2005, the incidence of fuel poverty
fell dramatically to around 2.5 million households.
Fuel
poverty
(millions of
1996 1998 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010
HHs)
All
Vulnerable
6.5
4.75
2.5
2.25
2
2
2.5
3.5
4
4.5
5.5
4.75
5
3.5
2
1.75
1.5
1.5
2
2.75
3.25
3.75
4.5
4
Table 3: Rates of fuel poverty in the UK from 1996 to 2010. Taken from Department of Energy and Climate Change (2012: 10)
The validity of the substantial fall in fuel poverty rates has been questioned by Professor Hills, who
was appointed by the Government in 2011 to review fuel poverty in the UK (see Hills (2012)). The
terms of reference called for Hills to review whether fuel poverty is a distinct problem or just part
of general poverty, and if it is distinct, how it should be measured, and the implications of
measurement for policy (Hills, 2012). In his interim report, Hills reported that fuel poverty was a
distinct problem, but criticised the dramatic ‘V’ shape that the current indicator produces, and
asked “Did the underlying problem of fuel poverty really improve by nearly four-fifths in just seven
years—suggesting that it was well on the way to being solved with little further action needed?”
(Hills, 2011: 13).
Page 16 of 47
Professor Hills has proposed two new fuel poverty indicators for the UK: firstly, the Low Income
High Costs (LIHC) indicator, which measures the extent of fuel poverty, and defines a fuel poor
household as one that experiences a combination of high energy costs and low household income.
Under this definition, high energy costs are defined as required fuel costs above the median,
adjusted for household composition, whilst low household income is defined as income below the
sixty per cent median poverty line, adjusted for household size and composition. Fuel poverty rates
remain moderately static from 1996 to 2009 when applying the LIHC indicator, with around 2.7
million households defined as having both low income and facing higher-than-median fuel costs in
2009. The second indicator Hills proposes is the fuel poverty gap, which measures the depth of fuel
poverty by determining the amounts by which the assessed energy needs of fuel poor households
exceed the threshold for reasonable costs (Hills, 2012).
By comparison, in Austria no official or published analysis of fuel poverty exists, however, Kalliauer
and Moser (2011) did produce a short report on the issue of energy poverty, which summarises a
number of issues, including rising gas and electricity prices and disconnections. Using national
statistics, Kalliauer and Moser demonstrate that average household energy expenditure has
increased significantly for households in the lowest income quartile between 2004/5 and 2009/10,
whilst energy expenditure has decreased marginally for households in the highest income quartile.
However, the authors misinterpret the United Kingdom’s ten per fuel poverty threshold, stating
that “at least every tenth household spends more than ten percent of the household budget for
energy. That is that boundary that is supposed commonly [used] as a definition for energy poverty”
(translation of Kalliauer and Moser, 2011: 2).
In Belgium, Huybrechs et al. (2012) have conducted a comprehensive investigation into fuel
poverty. Using actual expenditure data from the 1999 and 2009 Belgian Household Budget Survey,
Huybrechs et al. examined the average share of expenditure on energy, by income decile, and
found that energy expenditure represented a larger proportion of expenditure for households in
the poorest income deciles, compared with households in the richest income deciles. Huybrechs et
al. also found that energy expenditure has increased for households across all of the income
deciles, but particularly for households in the bottom three deciles. However, the authors misapply
Page 17 of 47
the United Kingdom’s ten per cent definition, and assert that potentially all households in the first
income decile could be in fuel poverty as they spend more than ten per cent of income on fuel
(Huybrechs et al., 2012: 70). In addition, the authors use mean rather than median figures, which as
discussed earlier, may distort the results as the mean can be affected by extreme values.
Agence nationale de l’habitat (2009) have carried out a detailed analysis of fuel poverty in France,
using 2006 data from the National Housing Survey, with a representative sample of 37,000
households (Agence nationale de l’habitat, 2009). They found that the national average fuel
expenditure represented 5.5 per cent of income, and subsequently chose a 10 per cent threshold to
determine fuel poverty. However, the use of mean fuel expenditure figures is problematic and so
the reliability of the 5.5 per cent figure and 10 per cent threshold is questionable. Nevertheless, the
research by Agence nationale de l’habitat reveals that whilst the majority of households across all
income quartiles spend less than ten per cent of their income on fuel, there is a disparity across the
income quartiles for households spending more than fifteen per cent. Around eighteen per cent of
households in the lowest income group spend more than fifteen per cent, compared to just 0.3 per
cent of households in the highest income group (Agence nationale de l’habitat, 2009).
Within Hungary, there have been two key studies that have used expenditure data. Firstly, Tirado
Herrero and Ürge-Vorsatz (2010) used data from a Hungarian survey on financial and living
conditions and found that the mean energy expense of Hungarian households was around 9.7 per
cent of net income between 2000 and 2007. However, as the authors used macro survey data,
which is data that has been derived from micro data by aggregating household variable results into
averages and frequencies (see United Nations Statistical Commission and Economic Commission for
Europe, 2000), they were unable to provide an estimate of the actual number of households
affected, furthermore the authors used a mean figure of energy expense, rather than median.
More recently, researchers from ENERGIAKLUB have investigated the characteristics of the fuel
poor in Hungary (Fellegi and Fülöp, 2012), using data collected for a separate research project,
entitled NegaJoule2020. Fellegi and Fülöp applied three different measures of fuel poverty, as
Page 18 of 47
outlined in Table 4. Firstly the authors used the British ten per cent level, without transferring the
underlying methodology, and subsequently found that 80 per cent households would be classified
as fuel poor, as the authors state: “this would obviously be a nonsensical approach and would make
it impossible to treat the problem” (Fellegi and Fülöp, 2012: 1). It is unclear why a twenty per cent
measure was applied, but this reduced fuel poverty levels to around forty per cent. The authors
finally calculated the fuel poverty threshold using a twice median (actual) spend definition,
resulting in a figure of 34 per cent; under this definition, around 8 to 10 per cent of the Hungarian
population are estimated to be in fuel poverty.
Fuel poverty rates (% and
households)
Fuel poverty measure applied
>10% actual
>20% actual
>twice median actual
expenditure
expenditure
expenditure (34%)
Percentage of population
80%
37 -40%
8-10%
Households
3,040,000
1,400,000 –1,500,000 300,000-380,000
Table 4: Fellegi and Fülöp’s (2012) fuel poverty results for Hungary
However, the reliability of the data is an issue in this research due to the sampling method
employed; as outlined in a methodology report (ENERGIAKLUB, 2011), data was collected using a
two-step, layered, quota-based sampling. In the first step, the sample quota was designed to be
representative on the basis of housing types and region, using data from the Central Statistical
Office (ENERGIAKLUB, 2011: 1). However, the second step is not as objective as “The interviewers
had to select the households for the completion of the questionnaire by random walking. This
means that it was the interviewer who could decide on the addresses and households visited within
the specified area – by observing the quotas specified” (ENERGIAKLUB, 2011: 2). Nonprobability
sampling methods, such as quota-based sampling, have been criticised for introducing selection
bias and unreliability, with no way to measure the precision of the sample (see Bryman, 2008). In
addition to analysing actual expenditure data, the authors also calculated the theoretic energy
demand required to heat a house to 20°C and provide hot water in the given household (Fellegi and
Fülöp, 2012: 1), however, it is unclear what criteria was used to calculate energy demand, nor what
the rationale for excluding electricity for appliances, lighting and cooking.
Page 19 of 47
Moving on to the Republic of Ireland, fuel poverty has been estimated there using actual
expenditure data from the Irish Household Budget Survey. Due to an absence of data on housing
conditions, the Irish Government has developed a so called ‘preliminary measure’, which defines a
household as fuel poor if annually it spends more than ten per cent of its disposable income on
energy (see Department of Communications, Energy and Natural Resources, 2011b). This threshold
currently represents twice-median fuel expenditure in Ireland, and is regarded as the least severe
form of fuel poverty in Ireland. Households that spend “more than three times the national median
ratio of household energy expenditure to household disposable income (currently 15%)”
(Department of Communications, Energy and Natural Resources, 2011b: 14) are classified as
experiencing severe energy poverty, whilst households spending more than four times the national
median ratio, currently 20%, are classified as suffering from an extreme form of fuel poverty
(Department of Communications, Energy and Natural Resources, 2011b: 14). However, this
preliminary measure is due to be replaced with a ‘comprehensive measure of energy poverty’ that
combines a housing conditions survey with a modelling framework (Department of
Communications, Energy and Natural Resources, 2011b).
Strengths and limitations of the expenditure approach
The expenditure approach is the most widely used method for measuring fuel poverty across
Europe, in part due to the objective and quantifiable nature of the approach. To reiterate the
comments made by Moore (2012), collecting detailed data on housing enables “fuel costs to be
accurately determined and compared directly with corresponding household incomes” (Moore,
2012: 3). However, whilst the benefits of this approach are self-evident, there are numerous
criticisms that can be levied on the approach.
Firstly, the (mis)interpretation of the UK’s expenditure method in other countries has shown that
the underlying methodology is complex and not easily transferred. By using a ten per cent actual
expenditure threshold that is not grounded in the specific context of the country under study,
researchers risk producing meaningless results. Indeed, the confusing nature of the expenditure
approach has been highlighted by Healy and Clinch who state “it can be misleading, as several
Page 20 of 47
formulae now exist for calculating fuel poverty, some with housing costs included in net household
income, other calculations exclude housing costs from the denominator of the formula, while other
calculations analyse gross household income as opposed to net” (Healy and Clinch, 2002b: 5).
Furthermore, Healy and Clinch point out that expenditure based estimates of fuel poverty are far
higher than consensual based estimates, and state this has “led some commentators to wonder
whether the two approaches are measuring the same type of fuel poverty, i.e. persistent versus
intermittent fuel poverty” (Healy and Clinch, 2002b: 5).
Harrington et al. (2005) condemn the methodology for calculating required fuel expenditure,
stating: “a formula-based fixed model of acceptable heating, perhaps driven by the ‘tyranny of
numbers’, may give a misleading picture of household need” (2005: 266). This point corresponds
with that made by Bouzarovski et al., who assert that the delimitation of the causes of fuel poverty
to ‘low income, inadequate building quality and high energy prices’ ignores the importance of
energy needs and socio-demographic circumstances at the household scale (Bouzarovski et al.,
2012: 78). Healy and Clinch also further criticise the method for its inability to “capture the
deprivation and social-exclusion elements of fuel poverty” (Healy and Clinch, 2002a: 9).
The most pertinent critique of the expenditure approach, however, is that it is not easily applied at
the European scale. As stated previously, the UK is almost unique in its production of a series of
large national housing condition surveys, which allow required fuel costs to be accurately
determined (Moore, 2012). Without the replication of this model in other member states, a
required fuel expenditure approach cannot be applied on a European basis. The absence of
standardised data concerning fuel expenditure (Healy and Clinch, 2002a) also excludes the
application of an actual fuel expenditure approach, necessitating the use of alternative methods.
3.3 Consensual approach
Given the criticisms and difficulties associated with the expenditure approach, particularly in a
European context, some academics (most notably Healy and Clinch, 2002a; Healy, 2004; Thomson
Page 21 of 47
and Snell, 2013) have proposed the use of consensual indicators to quantify fuel poverty. This
method is grounded in the consensual poverty approach pioneered by Gordon et al. and is based
on the inability “to afford items that the majority of the general public considered to be basic
necessities of life” (Gordon et al. 2000: 7). Typically, this has involved asking households whether
they can afford to heat their home, pay their utility bills on time, and live in a damp and rot free
home.
The consensual approach has tended to be used to measure pan-European rather than national fuel
poverty, with Whyley and Callender (1997) first using consensual indicators from the European
Community Household Panel (ECHP) to quantify fuel poverty across the United Kingdom, Ireland,
the Netherlands and Germany. Healy and Clinch (2002a) later expanded on this work by measuring
fuel poverty longitudinally from 1994-1997 across fourteen EU countries. Both sets of researchers
used six indicators from the ECHP, as displayed in Table 5 below. Three of the indicators were
subjective and required the respondent to make a value judgement, and the remaining three were
objective indicators, and required the respondent to describe factual characteristics of the dwelling.
Subjective indicators
Objective indicators
Households unable to heat home adequately
Presence of damp walls and/or floors
Households unable to pay utility bills
Lacking central heating
Households lacking adequate heating facilities
Rotten window frames
Table 5: Consensual indicators used by Whyley and Callender (1997) and Healy and Clinch (2002a)
However, in 2001 the ECHP survey was stopped and replaced with the EU Statistics on Income and
Living Conditions (EU-SILC). The EU-SILC dataset aims to be a “reference source for comparative
statistics on income distribution and social exclusion at European level” (Eurostat, 2010: 10), and as
Clemenceau and Museux (2007) comment, is a significant improvement on its predecessor, which
suffered from issues of reliability, varied response rates and incomplete geographical coverage
(Clemenceau and Museux, 2007). However, several variable changes occurred during the transition
from ECHP to EU-SILC which has made it more difficult to model EU fuel poverty, including the loss
Page 22 of 47
of a variable asking if a dwelling has central heating, and the merger of two separate housing
condition variables (presence of damp walls and/or floors, rotten window frames) to form a single
variable in EU-SILC.
To date, there have been two key studies that have used EU-SILC micro data (EPEE (2009) and
Thomson and Snell (2013)). The first study, the European Fuel Poverty and Energy Efficiency
project (EPEE), was co-financed by the European Commission, and analysed fuel poverty in the
United Kingdom, Spain, Italy, Belgium and France, using EU-SILC data from 2005. Later, Thomson
and Snell (2013) quantified fuel poverty across twenty-five EU member states. Both studies
employed just three proxy indicators to measure fuel poverty, compared with the six available to
earlier researchers; these variables are displayed below in Table 6.
Subjective indicators
Objective indicators
Household unable to pay to keep home adequately
Leaking roof, damp walls/floors/foundation
warm
and/or rot in window frames or floor
Household unable to pay utility bills on time
Table 6: Consensual indicators used by EPEE (2009) and Thomson and Snell (2013)
EU SILC data has also been used on a national scale to explore fuel poverty in Belgium (see
Huybrechs et al., 2012) and in Hungary (see Tirado Herrero and Ürge-Vorsatz, 2010). However,
these latter two studies used macro EU SILC data, which is data that has been derived from micro
data by aggregating household variable results into averages and frequencies (see United Nations
Statistical Commission and Economic Commission for Europe, 2000). This limits analysis to basic
descriptive statistics and prevents the researchers from estimating the actual number of
households affected. Nevertheless, both studies do begin to address the gap in knowledge
concerning the existence of fuel poverty in European countries.
Page 23 of 47
In addition to the research that has used ECHP and EU SILC data, there have been four studies that
have collected primary data and reported on fuel poverty levels using subjective indicators (see
Healy and Clinch, 2002b; Waddams Price et al., 2012; Gordon et al., 2013; Agence nationale de
l’habitat, 2009). As highlighted in the temperature approach section, Healy and Clinch (2002b)
conducted a national household survey of fuel poverty and thermal comfort in the Republic of
Ireland in 2001, and collected data from 1,500 households. Their key fuel poverty indicator asked
households how often they were unable to heat their home, and provided the following four-point
response scale:
Yes, without
any problems
Usually with
occasional
difficulty
Usually not
No, not at all
By ascertaining how often households were unable to heat their home, Healy and Clinch were able
to distinguish long-term (chronic) sufferers from short-term (intermittent) sufferers (Healy and
Clinch, 2002b). Using this indicator they found that 12.7% of Irish households suffered occasional
fuel poverty, whilst a further 4.7% suffered persistent fuel poverty, amounting to approximately
227,000 households in total (Healy and Clinch, 2002b: 9). Interestingly, these results are much
higher than the estimates for Ireland produced by Healy and Clinch (2002a) using ECHP data. They
speculate that a large portion of the intermittently fuel-poor are not declaring problems of fuel
poverty in the ECHP data and that this may be due to the binary-response variable format used in
the ECHP (Healy and Clinch, 2002b: 9).
In the study by Waddams Price et al., they were interested in low income households and collected
data from 3,417 households using the questionnaire outlined in Cooke et al. (2001). The key
questions they used to measure feelings of fuel poverty were:
In general, do you feel that you are able to heat your home adequately?
Do you feel that you can afford enough fuel for all your water heating and cooking needs?
Page 24 of 47
As well as collecting subjective information about feeling fuel poor, Waddams Price et al. also
defined respondents as expenditure fuel poor if they actually spent more than ten per cent of their
income on fuel. Their main findings were that the two measures gave very different results, stating
“Many households who spend more than 10% of their income on energy do not feel fuel poor, and
not everyone who feels fuel poor spends more than 10% of their income on fuel” (Waddams Price
et al., 2012: 37). However, it is important to note that Waddams et al. used actual fuel expenditure,
and the sample was intentionally skewed to “represent low income households typical of those
who use prepayment metres to pay for their household fuel” (Waddams Price et al., 2012: 33),
therefore the results are not representative for other household types.
By comparison, the Poverty and Social Exclusion project (Gordon et al., 2013) is a much more
comprehensive study. Although not strictly focussed on fuel poverty, the PSE living standards
survey contained a block of questions on fuel poverty (see Appendix 1, or Dermott et al., 2013) and
the study focusses heavily on various aspects associated with fuel poverty. The PSE study involved
three steps: an attitudes survey to determine what things are considered necessities; a living
standards survey, which distinguished between people who lacked necessities by choice and people
who could not afford the necessities; and finally a deprivation count to identify how many people
cannot afford groups of necessities (Gordon et al., 2013). In terms of opinions about necessities,
the PSE study found that 96% of adults considered heating to warm living areas of the home a
necessity, and 94% considered a damp-free home a necessity (Gordon et al., 2013: 5). However, the
study found that 9% of UK households (2.3 million) cannot afford to heat the living areas of their
home and 10% of households (2.7 million) live in a damp home (Gordon et al., 2013: 8).
The final empirical study of subjective fuel poverty is that by Agence nationale de l’habitat (2009).
As outlined in the expenditure approach section, this study was conducted in France in 2006, and
collected both expenditure and consensual data from a representative sample of 37,000
households. The main consensual indicators asked respondents if they felt cold in their house, and
what caused this. Table 7 below presents the results for these two questions; as can be seen, the
most common reason for feeling cold across all the income quartiles was poor insulation, followed
by insufficient heating.
Page 25 of 47
Quartile 1
Quartile
Quartile
Quartile 4
(lowest)
2
3
(highest)
Total
Feeling cold in the house (for
at least 24 hours)
Yes
19.8%
15.1%
11.0%
9.9%
14.0%
No
80.2%
84.9%
89.0%
90.1%
86.0%
Total
100%
100%
100%
100%
100%
1,303,000
998,000
723,000
653,000
3,677,000
Insufficient heating
37.2%
33.4%
25.9%
28.9%
32.5%
System failure
13.2%
16.4%
22.0%
29.0%
18.6%
Financial reasons
26.4%
21.9%
17.0%
11.4%
20.6%
Poor insulation
44.5%
43.5%
40.1%
32.7%
41.3%
Other reasons
11.2%
12.8%
18.9%
17.7%
14.3%
Number of households that
felt cold
Reasons for feeling cold
Table 7: Number of households feeling cold in France. Taken and translated from Agence nationale de l’habitat (2009: 3)
A commonality between the comparative and national consensual fuel poverty studies outlined
above is the focus on the following three aspects of fuel poverty: attaining adequate warmth;
paying utility bills on time; and living in a dwelling that is free from leaks, damp and rot. In Table 8,
these three indicators are discussed in more detail.
Indicator
Reasons for use
Criticisms/comments
Adequate warmth


As Healy and Clinch state,
it ‘‘encompasses the
At the European level, it has
been framed as a
Page 26 of 47
standard definition of fuel-
dichotomous variable (see
poor households’’ (Healy
Appendix 1), with
and Clinch, 2002a: 12)
respondents only able to
answer ‘yes’ or ‘no’, which
fails to capture variation in
experiences, and fails to
determine why people are
unable to heat their home.

Healy and Clinch’s
household survey in
Ireland, the PSE study and
the French housing survey
are notable exceptions,
however.

Ürge-Vorsatz and Tirado
Herrero (2012) and Healy
(2004) suggest that the
indicator should be made
broader to encompass
cooling relating difficulties
in summer

However, the2007 ad-hoc
housing module of EU SILC
has been the only survey to
ask if respondents are able
to keep comfortably cool
during summer time,
consequently there is
limited European evidence
around summer time
cooling difficulties.
Page 27 of 47
Arrears on utility bills

Experiencing financial
In the ECHP and EU SILC
difficulties with utility bills
variables, gas, electricity
may indicate that a
and water are grouped
household is struggling to
together, which may
afford adequate services
overestimate the problem
(Thomson and Snell, 2013).



There are anomalies with
Healy and Clinch (2002a)
the EU SILC data as it shows
suggest that people unable
UK households as not
to keep up to date on
experiencing any arrears
utility bills may suffer from
(Thomson and Snell, 2013)
disconnection of supply

Some surveys have only
mentioned gas and
electricity, despite many
households across Europe
relying on additional fuels
such as oil, firewood or
coal, as highlighted by
Bouzarovski et al. (2012)
and Tirado Herrero and
Ürge-Vorsatz (2010)
Leaking roof, damp

Provides “an objective

Many national household
walls/floors/foundation,
measure of the condition
surveys, such as the English
or rot in window frames
of the dwelling’’ (Eurostat,
Household Survey ask
or floor
2010: 175).
respondents about the
The presence of leaks,
severity, sometimes using
damp or rot may indicate a
photographs.

property is being

However, the ECHP and EU
continuously unheated or
SILC use a dichotomous
ineffectively heated (Healy
variable and do not ask
and Clinch, 2002a).
about severity, nor
distinguish between a
Page 28 of 47
leaking roof, damp walls, or
rotten windows.
Table 8: Summary of the three main consensual indicators
Strengths and limitations of the consensual approach
The consensual approach to measuring fuel poverty has numerous strengths. Firstly, it can be less
complex to collect consensual data than expenditure data, particularly actual modelled expenditure
data, thus it may be suitable as an interim measure of fuel poverty in countries that lack a
comprehensive house condition survey. Secondly, at the European level there are no standardised
micro data concerning household fuel expenditure or house conditions (Thomson and Snell, 2013;
Healy and Clinch, 2002a), and so by employing consensual indicators from the EU SILC survey
researchers have been able to circumvent data issues and quantify EU fuel poverty. A third
strength, and arguably the most important, is that a consensual approach to fuel poverty has the
potential to “capture the wider elements of fuel poverty, such as social exclusion and material
deprivation” (Healy and Clinch, 2002a: 10). By using subjective indicators that require value
judgements, researchers can gauge an individual’s experience of fuel poverty and their perceived
burden. However, this potential has yet to be fully exploited at the European level as the ECHP and
EU SILC surveys have framed the key fuel poverty indicators in dichotomous terms, and have not
followed up on the reasons why people feel unable to heat their home.
Conversely, the subjective indicators used in the consensual approach have been criticised for their
error of exclusion, whereby households may not identify themselves as fuel poor even though they
may be characterised as fuel poor under other measures (see Dubois, 2012). Furthermore, the
degree to which subjective measures overlap with expenditure measures is a concern. For example,
using English House Condition Survey data from 2005, Palmer et al. (2008) found that very little
overlap exists between fuel poverty using a subjective measure and fuel poverty using the UK’s ten
per cent expenditure threshold. Indeed, just 6% of households in fuel poverty by the standard
expenditure definition said that their living rooms were not warm in winter because of the cost it
took to do so (Palmer et al., 2008: 16). Additionally, Palmer et al. found that a third of households
that declared they were unable to keep their living rooms warm in winter had average or abovePage 29 of 47
average incomes, which they suggest means the subjective indicator is “picking up something other
than income (e.g. attitudes to expenditure)” (Palmer et al., 2008: 16).
McKay is also critical of consensual deprivation indicators, stating they “assume that there is a
broad consensus on what goods/services families should be able to afford, and that an inability to
afford those items can measure deprivation” (2004: 201). Consequently, if the underlying
assumptions are incorrect, a person may appear poor due to their consumption preferences rather
than lacking resources (McKay, 2004). Indeed, with regard to air conditioning, Sampson et al. found
that some participants actively opted not to use air conditioning for reasons other than financial,
for example “they disliked colder temperatures, sounds from the air conditioner, or associated
physiological reactions they experience when using air conditioners, including aggravated
respiratory problems or arthritis” (Sampson et al., 2013: 479). This highlights the importance of
gathering public opinion on what items are necessary, and reveals a weakness of previous panEuropean consensual fuel poverty work – the indicators have not been tested with the general
public prior to analysis, thus consensus is assumed to exist across twenty-eight diverse countries.
No research exists that assesses the quality and validity of consensual fuel poverty indicators in a
pan-European context, however, a basic analysis of Eurobarometer data from 2010 by the author
shows that while respondents from the majority of European countries concur that adequate
warmth and access to gas, electricity and water are essential, it is not unanimous across the EU.
Indeed, as shown in Table 9, just 34.2 per cent and 44.5 per cent of respondents in Slovakia and
Bulgaria consider access to gas, electricity and water a necessity, and only 22.7 per cent and 32.9
per cent of respondents in Malta and Portugal respectively, consider adequate warmth a necessity.
Country
% who think it’s necessary to
% who think it’s necessary to have
keep home adequately warm
access to gas, electricity, tap water
Austria
70.0%
68.2%
Belgium
65.6%
76.2%
Page 30 of 47
Bulgaria
74.7%
44.5%
Cyprus
52.4%
65.1%
Czech Republic
65.9%
70.2%
Denmark
64.5%
77.1%
Estonia
61.7%
60.0%
Finland
71.6%
73.0%
France
60.4%
79.8%
Germany East
72.9%
69.9%
Germany West
75.5%
70.6%
Great Britain
63.4%
72.6%
Greece
60.5%
50.0%
Hungary
71.9%
66.6%
Ireland
82.8%
71.0%
Italy
54.5%
59.1%
Latvia
53.7%
64.2%
Lithuania
57.2%
61.8%
Luxembourg
53.8%
66.7%
Malta
22.7%
72.7%
Netherlands
61.5%
88.2%
Northern Ireland
75.5%
68.1%
Poland
55.5%
62.2%
Page 31 of 47
Portugal
32.9%
56.9%
Romania
51.0%
59.6%
Slovakia
66.4%
34.2%
Slovenia
68.4%
80.7%
Spain
45.5%
64.1%
Sweden
71.1%
62.1%
Table 9: Public opinion about what is necessary to afford to attain a minimum adequate standard of living. Author analysis of
Eurobarometer 74.1 (2010)
4. Conclusion
It is evident from this short review that the analysis of fuel poverty across Europe is lacking, at both
the pan-European and national level. Indeed, at the national level, attempts to measure fuel
poverty have been made in just six member states – representing a little over twenty per cent of EU
members. An additional concern is the quality of previous research, for example, all five panEuropean analyses of fuel poverty have used data from before 2008, which means it is likely the
data will have been collected before households experienced the worst increases in gas and
electricity prices, as well as decreasing incomes overall due to the global financial recession. By
comparison, national level analyses have often incorrectly transferred the United Kingdom’s 10 per
cent methodology by using actual spend data rather than modelling required spend, and applying
mean rather than median figures. Furthermore, policy frameworks to address fuel poverty, and
formal definitions of fuel poverty, are in place in just a minority of EU member states, whilst at the
European level, policy provision is limited and piecemeal.
Nevertheless, numerous lessons and best practice can be taken from the existing knowledge on fuel
poverty measurement. Starting with the temperature approach, this review has established the
following:
Page 32 of 47

The temperature approach is rarely used at the national level and has never been used at the
European level.

There are many difficulties associated with collecting accurate internal temperatures. In
particular, intermittent occupancy may distort results, and social desirability bias may cause
householders to heat rooms to a higher level than normal.

The normative temperature range used to determine adequate warmth is 21°C in the living
room and 18°C elsewhere.
In terms of the consensual approach to fuel poverty, the following key points emerge:

It is important to distinguish between people who lack items out of personal preference, and
people who lack necessities because they cannot afford them.

A weakness of existing consensual fuel poverty studies is that they have not tested the
indicators with the general public prior to analysis, thus in relation to the EU, consensus is
assumed to exist across numerous diverse countries.

More accurate data may be collected when using a multi-item scale instead of a binaryresponse variable as the former can capture households that are intermittently fuel-poor as
well those that are persistently fuel poor.
Lastly, in relation to the expenditure approach to fuel poverty measurement, this review found:

Modelling required fuel expenditure is the most rigorous approach to measuring fuel poverty,
but requires complex data.

Actual fuel expenditure data is more readily available at the national level, however, it is a poor
indicator of fuel poverty as fuel poor households often spend less than required.

The use of median values is more appropriate than mean values as it can smooth out the
extreme values inherent in fuel expenditure distributions.

The calculation of energy needs should take into account the additional energy needs that
specific household types, such as disabled households, may require. This may include the need
for higher indoor temperatures for longer periods.
Page 33 of 47

Replication of the required fuel expenditure approach beyond the UK would necessitate the
establishment of national house condition surveys across the majority of the EU member states.

An actual fuel expenditure approach is also not possible at the European scale due to the lack of
micro data on fuel expenditure.
Page 34 of 47
Appendix 1 – Selection of survey questions currently and previously employed to
measure fuel poverty
Pan-European
European Community Household Panel





Can the household afford keeping its home adequately warm?
Has the household been unable to pay scheduled utility bills, such as electricity, water, gas during the
past 12 months?
Does the dwelling have hot running water?
Does the dwelling have heating or electric storage heaters?
Does the accommodation have lack of adequate heating facilities?
EU Statistics on Income and Living Conditions







Can your household afford to keep its home adequately warm?
In the last twelve months, has the household been in arrears, i.e. has been unable to pay on time due to
financial difficulties for utility bills (heating, electricity, gas, water, etc.) for the main dwelling?
Do you have any of the following problems with your dwelling / accommodation?
- a leaking roof
- damp walls/floors/foundation
- rot in window frames or floor
Dwelling equipped with heating facilities? (2007 housing module only)
Dwelling comfortably warm during winter time? (2007 housing module only)
Dwelling equipped with air conditioning facilities? (2007 housing module only)
Dwelling comfortably cool during summer time? (2007 housing module only)
National
English Housing Survey




Generally speaking, during winter when heating needs are greatest, at which of these times are
you or someone else in your household regularly at home? 8 possible answers
During the cold winter weather, can you normally keep comfortably warm in your living room?
Yes, No.
How easy or difficult is it for you to meet your heating/fuel costs? 5 possible answers
Have you (or your landlord/freeholder if applicable) had any of the following work done to your
heating or insulation in the last 12 months? Work to central heating, Work to storage heating,
Work to other fires/heaters, Work to insulation, None of these
Poverty and Social Exclusion UK 2012 living standards questionnaire (Dermott et al., 2013)




Looking at this card, do you have any of these problems with your accommodation? 12 possible
answers.
Did your household cut back on fuel use at home in any of these ways last winter, because you
could not afford the costs? 7 possible answers.
Describe the overall level of warmth in your home last winter? 6 possible answers
You said that you had problems with [Accommodation Problem] and that your home was
warmer/colder [Comfort] than you would have liked. Did this affect you or other members of
your household in any of the following ways? 6 possible answers.
Page 35 of 47

Sometimes people are not able to pay every bill when it falls due. Have you (or your household)
been in arrears on any of the things on this card during the last 12 months, due to a lack of
money? 13 categories including ‘Electricity, gas, fuel bills’.
England –Cooke et al. (2001) household survey on affording gas and electricity










In the winter, do you usually keep the heating on when everyone in the home is asleep, or do
you turn it down or do you turn it off altogether?
And again in winter, do you keep the heating on at all other times when there is someone in the
home?
Are there any rooms in your home that are hardly ever heated?
In general, do you feel that you are able to heat your home adequately?
Is this because your home is difficult to keep warm because of its condition (for example
draughts, damp, exposed location etc.), or because you find it difficult to afford the fuel?
Do you or anyone in your household have any special heating needs, for example because of
health, disability, or age - young or old, or for any other reason?
Do you feel that you can afford enough fuel for all your water heating and cooking needs?
Has your household’s electricity/gas ever been disconnected because of unpaid electricity bills?
If the cost of electricity/gas went DOWN, would you use MORE electricity/gas or use the same
electricity/gas and use the savings for something else?
If the cost of electricity/gas went UP, would you use LESS electricity/gas or use the same
electricity/gas?
French National Housing Survey


Feeling cold in the house (for at least 24 hours).
Reasons for feeling cold. 5 possible responses, including insufficient heating, and poor insulation.
Ireland – Healy and Clinch (2002b) household survey

How often are households unable to adequately heat their home? Four-point response variable.
Living in Wales Survey




What is the main way of heating your rooms in winter? 11 possible answers.
Can you tell me how effective you think the heating is in your home? Very effective, Fairly effective, Not
very effective, Not at all effective, Don’t know.
In order to help establish how much your heating is used, during winter when heating needs are
greatest, at which of these times are you or someone else in your household regularly at home? 9
possible answers including All day/all the time, and Highly variable.
Which of these methods do you use to pay for your electricity/mains gas? 11 possible answers.
New Zealand Warmer Winters Campaign Survey




Were you cold in your house last winter? Never/Sometimes/Often/Always
What’s your average monthly energy bill during winter?
Do you sometimes struggle to pay it?
Have you had your power disconnected because you couldn’t pay a power or gas bill within the last two
years?
Page 36 of 47
Northern Ireland House Condition Survey


How satisfied are you with each of the following aspects of your heating system? The type of heating;
The cost of running your system; The amount of heat that you can get; The control over the level of
heat; The ease of use of the system.
Generally speaking, during winter when heating needs are greatest, when would you or someone else in
your household have your heating on to stay warm? 8 possible responses, ranging from All day/all the
time, to Weekend evenings.
Scottish House Condition Survey



Are there any rooms which are not heated on winter weekdays/weekends, for example, if the radiator is
switched off?
During the winter months, do you generally find that your heating keeps you warm enough at home, or
not? 4 possible answers.
To what extent do you monitor your use of energy in your property? 5 possible answers.
Page 37 of 47
Bibliography
Agence nationale de l'habitat (2009) Analyse de la précarité énergétique à partir des résultats de
l’Enquête Logement 2006 de l’Insee. [online] http://www.precarite-energie.org/IMG/pdf/StatsAnah-ENL.pdf
Bazilian, B., Sagar, A., Detchon, R., and Yumkella, K. (2010) More heat and light. Energy Policy, 38:
5409-5412
Bevan Foundation (2009) Paying the Price of Being Poor. [Online]
http://www.bevanfoundation.org/publications/paying-the-price-of-being-poor
Birol, F., (2007) Energy economics: a place for energy poverty in the agenda? The
Energy Journal, 28: 1–6.
Boardman, B. (2012) Fuel poverty synthesis: Lessons learnt, actions needed. Energy Policy, 49: 143148.
Boardman, B. (2010) Fixing Fuel Poverty: Challenges and Solutions. Earthscan, London.
Bouzarovski, S., et al.. (2012) Energy poverty policies in the EU: a critical perspective. Energy Policy
49: 76–82.
Bryman, A. (2008) Social Research Methods. 3rd ed. Oxford University Press.
Page 38 of 47
Clemenceau, A., and Museux, J-M. (2007) EU-SILC (community statistics on income and living
conditions: general presentation of the instrument). In: Comparative EU statistics on Income and
Living Conditions: Issues and Challenges. Proceedings of the EU-SILC Conference, Helsinki, 6–8
November 2006. Office for Official Publications of the European Communities, Luxembourg.
Cooke, D., Ferrari, A., Giulietti, M., Sharratt, D and Waddams Price, C. (2001) Affording
Gas and Electricity: Self Disconnection and Rationing by Prepayment and Low Income
Credit Consumers and Company attitudes to Social Action. Electricity Association
Critchley, R., Gilbertson, J., Grimsley, M., Green, G. and Warm Front Study Group (2007) Living in
cold homes after heating improvements: Evidence from Warm-Front, England's Home Energy
Efficiency Scheme. Applied Energy, 84: 147-158
Darby, S. (2012) Metering: EU policy and implications for fuel poor households. Energy Policy, 49:
98-106
Department for Work and Pensions (2013) Disability Living Allowance [Online]
http://www.dwp.gov.uk/healthcare-professional/benefits-and-services/disability-living-allowance/
Department of Communications, Energy and Natural Resources (2011a) Warmer Homes: A Strategy
for Affordable Energy in Ireland. Dublin: Department of Communications, Energy and Natural
Resources.
Page 39 of 47
Department of Communications, Energy and Natural Resources (2011b) Warmer Homes: A Strategy
for Affordable Energy in Ireland – Technical Annex. Dublin: Department of Communications, Energy
and Natural Resources.
Department of Energy and Climate Change (2013) Fuel Poverty: a Framework for Future Action.
London: HMSO.
Department of Energy and Climate Change (2012) Annual report on fuel poverty statistics 2012.
London, HMSO [online] http://www.decc.gov.uk/assets/decc/11/stats/fuel-poverty/5270-annualreport-fuel-poverty-stats-2012.pdf
Department of Energy and Climate Change (2011) Annual report on fuel poverty statistics 2011.
London, HMSO [online]
https://www.gov.uk/government/uploads/system/uploads/attachment_data/file/48138/2181annual-report-fuel-poverty-stats-2011.pdf
Department of Energy and Climate Change (2010) Fuel Poverty Methodology Handbook. London,
HMSO [online]
https://www.gov.uk/government/uploads/system/uploads/attachment_data/file/66018/614-fuelpoverty-methodology-handbook.pdf
Department of Trade and Industry (2001) UK Fuel Poverty Strategy. London: HMSO.
Dermott, E., et al. (2013) Poverty and Social Exclusion in the UK: Final Questionnaire. PSE UK.
Page 40 of 47
Dubois, U. (2012) From targeting to implementation: The role of identification of fuel poor
households. Energy Policy, 49: 107-115
ENERGIAKLUB (2011) METHODOLOGY OF THE STATISTICAL SAMPLING. [Online]
http://www.negajoule.hu/sites/default/files/methodology.pdf
EPEE, (2009) Tackling Fuel Poverty in Europe: Recommendations Guide for Policy Makers.
http://www.fuel-poverty.com/files/WP5_D15_EN.pdf
European Anti-Poverty Network (2010) EAPN Working Paper on Energy Poverty. European AntiPoverty Network, Brussels.
European Coal and Steel Community Consultative Committee (2001) Opinion of the ECSC
consultative Committee on the European climate change programme and emissions trading. Official
Journal of the European Communities, C 170/8
European Commission (2010a) Commission Staff Working Paper: An Energy Policy for Consumers.
European Commission, Brussels.
European Commission (2010b) Communication from the commission to the European Parliament,
the Council, the European economic and social committee and the committee of the regions. The
European Platform against Poverty and Social Exclusion: A European Framework for social and
territorial cohesion. COM. 16.12.2010. Luxembourg, Office for the Official Publications of the
European Communities.
Page 41 of 47
European Commission (2002) Communication from the Commission to the Council and the
European Parliament: Energy cooperation with the developing countries. COM(2002) 408 final,
Brussels.
European Economic and Social Committee (2011) Opinion of the European Economic and Social
Committee on ‘Energy poverty in the context of liberalisation and the economic crisis’ (exploratory
opinion). Official Journal of the European Union, C 44/53
European Parliament (2010) Resolution of 15 December 2010 on Revision of the Energy Efficiency
Action Plan (2010/2107(INI)). Official Journal of the European Union, C 169 E/66
European Parliament (2008a) Legislative resolution of 9 July 2008 on the proposal for a directive of
the European Parliament and of the Council amending Directive 2003/55/EC concerning common
rules for the internal market in natural gas. Official Journal of the European Union, C 294 E/142
European Parliament (2008b) Resolution of 19 June 2008 on Towards a European Charter on the
Rights of Energy Consumers. Official Journal of the European Union, C 286 E/24
European Parliament (2008c) Legislative resolution of 18 June 2008 on the proposal for a directive
of the European Parliament and of the Council amending Directive 2003/54/EC concerning common
rules for the internal market in electricity. Official Journal of the European Union, C 286 E/106
Eurostat (2010) Description of Target Variables: Cross-sectional and Longitudinal. EU-SILC 065/08.
European Commission, Eurostat.
Page 42 of 47
Fahmy, E. (2011) The definition and measurement of fuel poverty: A Briefing Paper to inform
Consumer Focus’ submission to the Hills fuel poverty review. Consumer Focus, London.
Fellegi, D., and Fülöp, O. (2012) Poverty or Fuel Poverty? Defining fuel poverty in Europe and
Hungary. [online]
http://energiaklub.hu/sites/default/files/energiaklub_poverty_or_fuel_poverty_1.pdf
Gordon, D., Adelman, L., Ashworth, K., Bradshaw, J., Levitas, R., Middleton, S., Pantazis, C., Patsios,
D., Payne, S., Townsend, P., Williams, J. (2000) Poverty and social exclusion in Britain. Joseph
Rowntree Foundation, York.
Gordon, D., Mack, J., Lansley, S., Main, G., Nandy, S., Patsios, D., Pomati, M. (2013) The
Impoverishment of the UK - PSE UK first results: Living Standards. PSE UK.
Harrington, B.E, et al.. (2005) Keeping warm and staying well: findings from the qualitative arm of
the Warm Homes Project. Health and Social Care in the Community, 13: 259-267
Healy, J.D. (2004) Housing, Fuel Poverty and Health: A Pan-European Analysis. Ashgate: Aldershot
Healy, J. D., and Clinch, P (2002a) Fuel poverty in Europe: A cross-country analysis using a new
composite measure. Environmental Studies Research Series, University College Dublin.
Healy, J.D. and Clinch, J.P. (2002b) Fuel poverty, thermal comfort and occupancy: results of a
national household-survey in Ireland. Applied Energy, 73: 329-343
Page 43 of 47
Hills, J. (2012) Getting the measure of fuel poverty: Final Report of the Fuel Poverty Review. CASE
Report 72, London
Hills, J. (2011) Fuel Poverty: The problem and its measurement - Interim Report of the Fuel Poverty
Review. CASE Report 69, London
Househam, I., and Musatescu, V. (2012) Improving Energy Efficiency in Low-Income Households and
Communities in Romania: Fuel Poverty Draft assessment report.
http://www.undp.ro/libraries/projects/EE/Assesment%20Report%20on%20Fuel%20Poverty%20%20DRAFT.pdf United Nations Development Programme, Romania
Huybrechs, F., Meyer, S., and Vranken, J. (2012) La Précarité Energétique en Belgique [online]
http://dev.ulb.ac.be/ceese/CEESE/fr/projet.php?menu=1&categorie=3&projet=124
Isherwood, B.C., and Hancock, R.M. (1979) Household Expenditure on Fuel: Distributional Aspects.
Economic Adviser's Office, DHSS, London
Kalliauer, J., and Moser, J. (2011) Energiearmut: In immer mehr Haushalten fehlt das Geld für Strom
und Heizung! [online]
http://www.arbeiterkammer.com/bilder/d154/PKU_Energiearmut_Juli2011.pdf
Liddell, C., Morris, C., McKenzie, S.J.P. and Rae, G. (2012) Measuring and monitoring fuel poverty in
the UK: National and regional perspectives. Energy Policy, 49: 27-32
Page 44 of 47
McKay, S. (2004) Poverty or Preference: What Do ‘Consensual Deprivation Indicators’ Really
Measure? Fiscal Studies, 25: 201-223
Milne, G., and Boardman, B. (2000) Making cold homes warmer: the effect of energy efficiency
improvements in low-income homes. Energy Policy, 28: 411-424
Moore, R. (2012) Definitions of fuel poverty: Implications for policy. Energy Policy, 49: 19-26
Oreszczyn, T., Hong, S.H., Ridley, I., Wilkinson, P., and The Warm Front Study Group (2006)
Determinants of winter indoor temperatures in low income households in England. Energy and
Buildings, 38: 245–252
Osbaldeston, J. (1984) Fuel poverty in UK cities. Cities, 1 (4): 366-373
Palmer, G., MacInnes, T., and Kenway, P. (2008) Cold and Poor: An analysis of the link between fuel
poverty and low income. New Policy Institute, London
Parckar, G. (2008) Disability poverty in the UK. Leonard Cheshire Disability
Plan Bâtiment Grenelle (2009) Groupe de travail Précarité énergétique Rapport. http://www.planbatiment.legrenelle-environnement.fr/index.php/actions-du-plan/rapports.
Public Health Policy Centre (2008) Annual Update on Fuel Poverty and Health. Dublin: Institute of
Public Health in Ireland.
Page 45 of 47
Sagar, A.D. (2005) Alleviating energy poverty for the world’s poor. Energy Policy, 33: 1367–1372.
Sampson, N.R., Gronlund, C.J., Buxton, M.A., Catalano, L., White-Newsome, J.L., Conlon, K.C.,
O’Neill, M.S., McCormick, S., and Parker, E.A. (2013) Staying cool in a changing climate: Reaching
vulnerable populations during heat events. Global Environmental Change, 23: 475-484
Snell, C., and Bevan, M. (2013) Fuel poverty and disability: a review of existing knowledge conducted
for eaga charitable trust. The University of York.
Snell, C., and Thomson, H. (2013). Reconciling fuel poverty and climate change policy under the
Coalition government: Green deal or no deal? In: G. Ramia and K. Farnsworth (eds.) Social Policy
Review 25: Analysis and debate in social policy, 2013, The Policy Press.
Thomson, H. and Snell, C. (2013). Quantifying the prevalence of fuel poverty across the European
Union. Energy Policy, 52: 563-572.
Thomson, H., Snell, C., and Bevan, M. (2013) Fuel Poverty and Disability: a statistical analysis of the
English Housing Survey. University of York.
Tirado Herrero, S. and Ürge-Vorsatz, D. (2010) Fuel Poverty in Hungary: A first assessment. Central
European University, Hungary
Tirado Herrero, S. and Ürge-Vorsatz, D. (2012) Trapped in the heat: A post-communist type of fuel
poverty. Energy Policy, 49: 60-68
Page 46 of 47
United Nations Statistical Commission and Economic Commission for Europe (2000) "Terminology
on Statistical Metadata", Conference of European Statisticians Statistical Standards and Studies,
No. 53, Geneva.
Ürge-Vorsatz, D. and Tirado Herrero, S. (2012) Building synergies between climate change
mitigation and energy poverty alleviation. Energy Policy, 49: 83-90.
Whyley, C., and Callender, C. (1997) Fuel poverty in Europe: evidence from the European Household
Panel Survey. National Energy Action, Newcastle upon Tyne.
World Health Organization (1987) Health impact of low indoor temperatures. WHO Regional Office
for Europe, Copenhagen.
Page 47 of 47