Towards a multiple-indicator approach to energy poverty in the European Union

Depósito de investigación de la Universidad de Sevilla https://idus.us.es/ Esta es la versión aceptada del artículo publicado en: This is an accepted manuscript of a paper published in: Energy and Buildings (2019): 15/111/2024 DOI: https://doi.org/10.1016/j.enbuild.2019.03.039 Copyright: © 2019 Elsevier Ltd. All rights reserved. El acceso a la versión publicada del artículo puede requerir la suscripción de la revista. Access to the published version may require subscription. “This is an Accepted Manuscript of an article published by Elsevier in [Energy and Buildings 2019], available at: https://doi.org/10.1016/j.enbuild.2019.03.039 ” page 1 Towards a multiple-indicator approach to Energy Poverty in the European Union: a review Raúl Castaño-Rosa (corresponding author), Department of Building Construction II, University of Seville, Spain, Avenue Reina Mercedes, 4-a 41012, Seville. Email: [email protected]. Jaime Solís-Guzmán, Department of Building Construction II, University of Seville, Spain. Carlos Rubio-Bellido, Department of Building Construction II, University of Seville, Spain. Madelyn Marrero, Department of Building Construction II, University of Seville, Spain. Abstract The term energy poverty (EP), mostly used to capture problems related to an inadequate access to energy suppliers in developing countries, and that of fuel poverty (FP), commonly employed when households suffer from insufficient monetary resources to pay for their basic energy needs, are used interchangeably by many researchers. Energy poverty and/or fuel poverty are considered to be an indication of domestic energy deprivation or energy vulnerability and remain a global problem that has been accentuated in Europe due to the economic crisis. This work has made a comprehensive review of current concepts and indicators of FP across Europe, thereby raising discussion of how the energy vulnerability factors fit FP situations: available infrastructure, energy efficiency, social and economic poverty, and wellbeing and health. In this context, the first weak spot of the FP indicators analysed is that most factors that influence the everyday activities of a household are not considered (comfort, health, and wellbeing), resulting in an incomplete analysis if these indicators are used in an isolated way. This paper provides a multiple-indicator approach to understand an FP situation beyond its traditional definition, thereby establishing a starting point for the reflection of vulnerability in policy decisions and methodological implications. Keywords Energy poverty, Fuel poverty, Vulnerable homes, Energy Vulnerability, Thermal-comfort Highlights • Lack of a clear path in the fight against Fuel Poverty across the European Union. • Fuel Poverty related to the residential sector allows new solutions to be developed. • Current fuel poverty indicators provide an incomplete analysis of vulnerability. 1. Introduction Fuel Poverty (FP), a global problem that has been accentuated in Europe due to the economic crisis, has given rise to government and political party interest, and has caused a major public impact. There are page 2 several official definitions and indicators of FP developed by countries such as the United Kingdom, Ireland, France, and Slovakia [1]. However, no official standard has been established across the European Union that assesses FP in member states, and therefore in certain countries this type of poverty still remains to be recognized as a social problem. Although it is not the main reason, the lack of a specific legislative programme to address the issue can lead to FP rates varying drastically from country to country, or even between regions within a country. For example, based on Eurostat data for 2012, 46.5% of the population of Bulgaria are unable to keep their homes adequately warm during the winter, while in Estonia this figure is officially only 4.2%, in Denmark 2.6%, and in Sweden, 1.4% [2]. According to certain sources, 50 million people in Europe live in FP; other research, however, indicates this figure to be closer to 20% of the total EU population (101 million people) [2]. According to the latter research, the main objective of the UN´s Millennium Development Goals [3] is to eradicate extreme poverty, thus improving living conditions and facilitating the path towards sustainable development. In this context, the European Commission, in a staff working paper, has suggested three criteria to evaluate FP: the inability to keep homes adequately warm; the delay in the payment of utility bills; and the occupation of defective dwellings (those with leaky roofs, damaged walls or foundations, and with adverse conditions such as mould and rot). The FP concept not only indicates the inability to keep a home adequately warm during different periods of the year, or the existence of problems in paying for the energy consumed or the high cost of energy prices, but it has also evolved into a multi-dimensional concept. Fuel poverty is currently defined as that situation which may deprive households not only of heating and/or cooling, but also of hot water, electricity, and other essential household needs [4]. In developing countries, energy consumption levels that satisfy basic necessities may be out of reach [5] due to a complex interplay of power [6], politics and commodification, together with asymmetry regarding the management, generation, distribution and final consumption of energy [7]. This situation, usually defined as Energy Poverty (EP) [4,8], “has been traditionally used to capture problems related to the inadequate access to energy supplies in developing countries, and involves a host of economic, infrastructural, social equity, education and health concerns” (Bouzarovski, Stefan, Petrova, Saska and Tirado-Herrero, 2014, p. 1). The term EP is related to the characteristics of the area where the home is located by identifying those areas with higher levels of poverty, usually located far from the city centre, with inefficient infrastructure and equipment, and a population with high levels of economic deprivation (old town, suburbs, social exclusion zones, rural areas, etc.) resulting in a poor quality of life for their population. In spite of various definitions, EP and FP can be used interchangeably [10], and hence both terms are used with the same meaning throughout this paper. Furthermore, the terms FP and EP have been related to Energy Vulnerability (EV), which is defined as “the propensity of an individual to become incapable of securing a materially and socially needed level of energy service in the home” (Bouzarovski, Stefan, Petrova, Saska and Tirado-Herrero, 2014, p. page 3 10), and this variable is even used in reference to the lack of access to modern energy services in developing countries, mainly in Central and South America, sub-Saharan Africa, and Asia [11,12]. This leads us to focus on the definition of the concept of a vulnerable consumer, which is related to the concepts “resilience” (the ability to recover quickly from illness, misfortune, troubles, or the like) [9,13,14], and “scarcity” (insufficiency or shortness of supply) [15], instead of on FP. This paper makes a comprehensive review of current concepts and indicators of FP across Europe. It also raises a discussion concerning how the energy vulnerability factors fit FP situations (infrastructures available, energy efficiency, social and economic poverty, and wellbeing and health [9]), which leads to the identification of the weaknesses of current FP indicators. The result is a multiple-indicator approach towards the understanding of an FP situation beyond its traditional definition, thereby providing a starting point for the reflection of vulnerability in policy decisions and methodological implications. Those households at risk of FP could therefore be better identified. 2. Fuel Poverty concepts around Europe The concept of FP was introduced by Isherwood and Hancock [16] in 1979 following the energy-price inflation due to the petroleum crisis (1973-1974). However, the definition of FP was given for the first time by Dr Brenda Boardman in 1991 as occurring when households cannot “have adequate energy services for 10 per cent of income” (Boardman, 2010, p. 277). There was much discussion on considering it unnecessary to pay special attention to the issue of FP, since it was related to the poor and to the economic crisis. However, several researchers including Braubach and Ferrand [18], Rosenow et al. [19], and Boardman [20], identify the FP situation as being related not only to those households with low incomes, but that energy efficiency in households can also be an effective policy approach in order to reduce FP in society. An attempt was made by The European fuel Poverty and Energy Efficiency (EPEE) [21] project in 2009, framed within the “Intelligent Energy Europe” programme, to analyse the FP situation in several countries (France, Belgium, United Kingdom, Italy, and Spain). The objective was to identify which of the factors are similar and to determine a wide number of good practices that reduce FP. In addition to this isolated case, there are numerous FP definitions developed in the EU as well as a diversity of indicators for their analysis, although no official standard has yet been established across the EU. Therefore, the diversity of policies regarding FP for these European countries has been analysed. In the case of Bulgaria, neither an official FP definition nor policy strategies have been established; the only policies available are those leading to the eradication of general poverty without any relation to energy supply. Currently, a new plan, which aims to protect vulnerable energy customers, has been signed by the government in order to reach the goal listed by the European Union in the Energy Roadmap Strategy 2050 (Commission, E., 2012), and its latest version stated that a vulnerable situation occurs when “households page 4 that are supplied with electricity and are inhabited by people who, for reasons of age, health or low income, are at risk of social exclusion regarding electricity supply and consumption, and benefit from social protection measures needed to provide them with the necessary supply of electricity” [23]. France set up the “RAPPEL” (Réseau des Acteurs de la Pauvreté et la Précarité Energétique dans le Logement) FP work group in 2007, (Network of Poverty Factors and Energy Precariousness in Housing), which accepted the establishment of a definition of FP based on the official UK definition of FP, although this remains an unofficial definition. With this premise, in Grenelle II law, the concept of energy precariousness and an official FP definition were established in 2009, whereby a person was considered as suffering from fuel poverty “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” [24]. The Department of Communications, Energy and Natural Resources of Ireland states officially in 2016 that “a household that spends more than 10% of its income on energy is considered to be in energy poverty” (“Department of communications, energy & natural resources. A Strategy to Combat Energy Poverty,” 2016). Slovakian Act No. 250/2012 Coll. states that “FP is a status reached when average monthly expenditures of a household on consumption of electricity, gas, heating and hot water production represent a substantial share of the average monthly income of the household” [26]. Two official definitions are used in the United Kingdom depending on the region: “a household is said to be in fuel poverty if it needs to spend more than 10% of its income on fuel to maintain an adequate level of warmth” is the official definition for Northern Ireland, Scotland and Wales [27]; “A household is considered to be fuel poor if they have required fuel costs that are above average (the national median level) and having spent that amount, hence they would be left with a residual income below the official poverty line [60% median income]” is the official definition for England [28]. An official FP definition has not yet been established in Spain; however, four studies are being carried out by the Association of Environmental Sciences (Asociación de Ciencias Ambientales, ACA) [29–32] during 2010-2018 which strive to analyse FP, and therefore to provide a proper definition. Furthermore, the relation between FP and the health of the household was analysed, and an increment of 30% in the occurrence of death during the winter was found to be due to low incomes and poor-quality dwelling, according to the study carried out by Dr. J. D. Healy in 1988-1997 [33]. More specifically, during the winter of 2014, 7,000 deaths out of the total 24,000 deaths in Spain were associated to FP [31]. Another FP study “Fuel Poverty in Spain” [34] was published in 2014 by the research group Economics for Energy of the University of Comillas, which provided a critical analysis of the FP indicators available in Europe and of the current situation in Spain. The study pointed out the need to establish an official definition by considering all factors involved and highlighted that FP cannot be considered an isolated issue. page 5 In this context, the latest work carried out by Castaño-Rosa et al. in Spain [35], which provides a new methodology to identify homes in a vulnerable situation at risk of FP, along with the studies from ACA, may be considered the most comprehensive initiatives to resolve this issue in Spain. 3. Fuel Poverty Indicators: Critical analysis Currently, various indicators are used to analyse and evaluate FP, and therefore this paper provides a revision of the existing FP indicators within the EU and their relationship to the residential sector through a technical perspective. These indicators can be divided into two categories: those based on household income and expenses; and those based on self-reported household conditions. Furthermore, several methodologies and indicators are used in the analysis of those consumers that are more vulnerable to falling into FP, including econometric analysis, the RELS Project, thermal-comfort analysis, and those based on the energy rating of the dwelling (Table 2). 3.1. Indicators based on household income and expenses These indicators analyse household income by using its utility bills and the various day-to-day expenses. 3.1.1. 10% indicator A household is suffering from FP when it cannot “have adequate energy services for 10 per cent of income” (Boardman, 2010, p. 277). This is a simple and versatile indicator, and its results are also easy to understand and communicate, thereby allowing a good policy path to be established. Most critics hold one of two opinions: that either the energy prices exert a major influence on the indicator results by yielding higher values when prices are high, and lower values when prices are low; or that the 10% threshold, which was justified by the socio-economic situation in the United Kingdom at the beginning of the 90’s, is too arbitrary a figure. In recent years, a significant number of households that have high incomes and inefficient dwellings have been labelled as fuel poor by the application of this indicator. This indicator points to the average energy expense of 30% of income of the poorest households in England, and to approximately the double median value of the percentage of energy expense of households. Certain authors [36,37], in the application of the indicator, are critical both of the use of full income without excluding those household expenses and of the non-inclusion of those possible household economic benefits (rent, pension, etc.), and support the use of net income after home expenses in the calculations. In turn, the indicator definition, based on the energy consumption used to maintain the dwelling adequately warm, fails to consider the “heating or eating” effect that occurs when households are unable to afford both food and minimum energy consumption due to their lack of economic resources, and therefore have to choose between paying their energy bills and buying those minimum foods to survive. The impossibility of analysing dwelling page 6 characteristics reveals another weak point of this indicator, since it is known that most households, due to their characteristics, priorities, and customs, use the dwelling systems inadequately (since they remain unaware of how to use the heating system, cooker, lights, etc. properly), leading to a significant increase in energy consumption. This energy-consumption increase can be prevented if good usage information and personal advice are made available to these households. 3.1.2. 2M Indicators Four very similar types of indicators are available: Double the median of household expenditure on energy; Double the mean of household expenditure on energy; Double the median share of household expenditure on energy; and Double the mean share of household expenditure on energy [38]. Only the third indicator out of these four is based on Boardman’s work by following the identification of 5% of the median share of household expenditure on energy in relation to the full income of British households in 1988. These indicators therefore define a household as being in FP when it must pay more than double the median share of its income on energy. The FP threshold is set according to the national median value, which allows this indicator to be recalculated every year, evaluations of possible fluctuations to be computed, and the application of a static indicator to be prevented, thereby leading to its acceptance as the FP threshold in England. Furthermore, from a statistical perspective, using the indicator based on the median instead of the mean is more sensitive to untypical values and habit changes. With this premise, in a hypothetical situation where rich people increase their energy consumption considerably, the use of an FP indicator based on the mean would have a major influence on the threshold and would therefore fail to identify the poorest people whose energy consumption is of a standard level as being fuel poor. Key points of this indicator: households with high incomes are not included as fuel poor; the characteristics of the country are considered; and the threshold definition is related to the distribution of income and expenses, thereby preventing arbitrariness in its calculation. 3.1.3. Minimum Income Standard (MIS) Indicator This was defined by Moore as the minimum income of a household that allows it to “have what it needs in order to have the opportunities and choices necessary to participate in society” [37]. A good example of this indicator is given by “A minimum income standard for Britain” project carried out by Bradshaw et al. [39]. This project features two key points: first, it limits the area of study, since the MIS set is always related to a specific social group; secondly, the participatory character of the project enables people to participate in the establishment of a suitable MIS and in the development of conclusions. The most important part of this indicator involves the definition of the “Minimum Income Standard” (MIS), and two factors are taken into account: first, the definition of “minimum” as set by the Human Rights Convention of the UN as “those page 7 things necessary for physical, mental, spiritual, moral and social well-being of people”; and secondly, the studies carried out by an expert committee in the USA that analysed household incomes in 1980 and defined the Prevailing Family Standard concept as “the standard of living achieved by the typical family”, and was “set at the median expenditure of two-parent families with two children“ [40]. In order to consider the most vulnerable people, various socio-economic parameters of the home were defined: Household size, job status, disability, health, ethnic group, and accessibility to energy. Several types of families with their correspondent MIS (taken as a weekly expense) were then established. In this way, a home is in FP if its income after energy-consumption expense is below the MIS for the corresponding characteristics of the household, and certain authors suggest calculating the MIS after subtraction of home expenses in order to prevent the distortion of these expenses. The “minimumincome.org.uk” initiative in United Kingdom allows the population to calculate their MIS according to several daily-life variables. The determination of the correct definition of the MIS (household income available after home expenses, to defray those minimumenergy needs) is the main disadvantage of this indicator, and results in a complex and arbitrary task. 3.1.4. Low-Income High-Cost (LIHC) Indicator This is the basis of the new FP strategy in England, and identifies an FP situation when “the household income is lower than the monetary poverty threshold and the energy-consumption expense is higher than the threshold” [41]. Two thresholds have to be established for the indicator to be applied: first, Hills uses 60% of the median equivalised disposable income after social transfers and the modelled equivalised energyconsumption expenses; and secondly, the median equivalised energy-consumption expense is used, which is based on all households in the area analysed. The main critics of the LIHC indicator include: Preston et al. [42], who provide a comparative analysis between the LIHC indicator and the 10% indicator; and Moore, who defends the MIS indicator, as follows: - This is an excessively complex and unclear indicator, mainly due to the modelling of the equivalent energy-consumption expenses. - By setting the threshold of energy expenditure to the median without further considerations, the indicator obviates the effect of household energy efficiency, and at the same time makes it difficult to identify those households that can escape from energy poverty through the reduction of their energy expenditure. - The LIHC indicator is not based on the actual costs of the household, but instead on those costs that it would need in order to keep the dwelling adequately warm for the well-being of the occupants. - The most vulnerable people (the elderly, those with chronic illnesses, disabled people, and children) are seldom considered as fuel poor with this indicator [43]. page 8 - This indicator defines a home with low income and poor energy-efficiency and therefore leads to the promotion of energy-efficiency interventions as the main measure to reduce FP; low-income households will consequently continue with problems paying their utility bills and suffering from health problems mainly due to inadequate thermal comfort in their dwellings. The key point of this indicator is based on the possibility of differing between FP and general poverty by defining FP as depending on the household income. 3.1.5. After Fuel Cost Poverty (AFCP) Indicator This indicator is based on the MIS indicator proposed by Moore, the studies carried out by [44], and the applications carried out in Spain and England by [34] and [41], respectively, and it defines a home as being in FP when its income, after home and energy-consumption expenses, falls below the minimum acceptable income (according to the size and type of the household using the modified OECD scale). In short, this indicator attempts to prevent social exclusion of those households with monetary problems by establishing the minimum-income level that can guarantee the well-being of a person. This indicator differs from others, such as the LIHC indicator, in its capacity to identify those households with a lower income in relation to the threshold and low energy-consumption expense, which is not identified as being in FP by the LIHC indicator, but is actually more vulnerable to being fuel poor. 3.1.6. Hidden Energy Poverty (HEP) Indicator This identifies those homes whose low incomes place them in an FP situation since “the absolute energy expenditure is below half the median absolute energy expenditure” [45]. This indicator enables the identification of those households who have to choose between either paying their utility bills or eating (the “heating or eating” effect) due to their low income and consequent low energy-consumption expense. In contrast, this indicator is valid if the absolute monetary expense is used instead of the energy-consumption expense as a share of the income, since high-income households tend to have a higher absolute energyconsumption expense, but a lower energy-consumption expense as a proportion of their incomes (resulting in an FP situation), in contrast to low-income households. In short, those households that have an energyconsumption expense below the standard level can be identified by using the absolute energy expenditure. In conclusion, the results provided by each indicator can differ widely, and therefore lead to the use of as many indicators as possible in an acceptable and appropriate analysis. The challenges of operationalizing each indicator from a methodological perspective can include: - Data availability: the main data resource employed in the application of these indicators is based on the median national data of household income, fuel prices, housing costs, household expenditure, page 15 their day-to-day activities, remain unable to pay their high energy-consumption bills even if monetary grants are provided to carry out energy efficiency interventions in their dwellings. 3.6. The RELS Project The RELS Project [92] provides a good example of those projects which aim to reduce FP by including tenant performance in both the pre- and post-analysis carried out after the energy-efficiency intervention. This project, managed by the Housing Agency of Catalonia, Spain, aims to reduce energy insecurity and to ease the energy-efficiency process in social housing in order to provide Mediterranean households with a beneficial housing model that presents sufficient thermal comfort at a fair cost. The RELS process is defined according to the methodology that the European SQUARE project [93] has used in Austria, Bulgaria, Finland, Holland, Spain, and Sweden, which focuses on ensuring a controlled, optimal, and high-quality energy-efficiency intervention of the interior environment in multi-family buildings. The RELS project consists of: evaluation of the initial state of the buildings under study; identification of the points for improvement; validation of the most efficient measures; implementation of the measures selected; and the monitoring of efficiency measures after the intervention. The measures applied are evaluated by using a six-variable matrix (20% reduction of CO2 emissions; 20% reduction of energy consumption; minimum renewable-energy use of 20%; optimal intervention cost in relation to the project budget; energy-cost reduction in relation to household income; and the guarantee of a reliable analysis in relation to the methodology), which is based on the targets set in the project, and allows a comparative analysis with the initial case to be performed. The initial results of the application of the RELS project in two case studies of the Housing Agency of Catalonia, Spain (residential building blocks with 27 tenants) show the influence of considering household typology together with the socio-cultural and academic-level aspects. Furthermore, the amortization of the cost of energy-efficiency interventions in relation to the health cost of a person in FP living in social housing is analysed (by using data from the UK) in order to provide an approximation of the monetary cost that the Spanish Health Service would have to pay to keep an FP household in a state of perfect health. Therefore, if the cost of energy-efficiency intervention is offset by the health benefits of the household under study, then that intervention is justified. Regarding the thermal-comfort variable, this project uses self-reported surveys on household conditions, and a wide range of results are obtained. The limits of admissible comfort of two homes are analysed during the summer: first, that of an immigrant man from the north of Africa; and second, a family native to Barcelona. The results show that the immigrant man had a lower energy consumption in air-conditioning than the native household, mainly due to his higher limit of admissible comfort. Both analysed studies, [86] and the RELS project, show the need to consider the thermal comfort in the dwelling when the vulnerability page 16 of a household is analysed. Furthermore, the use of an adaptive model of thermal comfort in relation to attaining a reduction in energy consumption shows great potential, and allows a comparative analysis among different study cases to be performed through the application of objective data. Table 2. Summary of analysed indicators (Source: Authors’ own preparation based on the references quoted). Indicator Assessment Reference Based on household income and expenses Energy-consumption expense is higher than 10% of the household income (10%) [17] Energy-consumption expense is higher than twice the national median (2M) [38] The household income is lower than the Minimum Income Standard (MIS) [37] The household income is lower than the monetary poverty threshold and the energy-consumption expense is higher than the threshold (LIHC) [41] The household income after fuel cost is lower than the threshold (after fuel cost in the study area) (AFCP) [44] [34] [41] Absolute energy-consumption expense is lower than the threshold (HEP) [45] Based on self-reported household conditions The inability to keep homes adequately warm during winter European Union Survey on Income and Living Conditions (EU-SILC), 2014 [48] The delay in the payment of utility bills The dwelling has deficiencies, such as a leaky roof, damaged walls or foundations, and conditions such as mould and rot The inability to keep homes adequately cool during summer Survey on perceptions and statements from households (SPS), 2014 [49] Econometric analysis The influence of various demographic, socioeconomic and physical factors in the probability of being fuel poor [50] [51] Thermal-comfort % of hours in which a household is in a thermal- [86] page 17 analysis comfort situation Based on the energy rating of the dwelling The influence of dwelling quality (energy consumption) on fuel poverty. Dwellings with poor quality cause an increase in energy consumption, thereby triggering FP [90] [91] The RELS Project Monitoring and analysis of the tenant, as well as the performance of the tenant, in order to provide an energy-efficiency intervention, and therefore an increase in the indoor-environment quality in the dwelling [92] 4. Discussion This section provides a discussion of current FP measures based on the evidence reviewed and on the author’s experience in arguing the weaknesses of FP indicators to identify those households at risk of FP. The energy vulnerability factors: available infrastructure, energy efficiency, social and economic poverty, and wellbeing and health, have been used and a multiple-indicator approach, which gathers various methods available, is proposed for the definition of vulnerable households at risk of FP (Figure 2). 4.1. Available infrastructure Scarcity in energy supply, mainly due to a variety of problems concerning distribution facilities, leads to the inability of occupants to maintain a minimum level of energy services in the home. None of the indicators analysed here involves an FP situation in relation to the infrastructures available in the area where home is located. The analysis carried out by these indicators is solely based on the monetary data of tenants, energy consumption, and indoor temperature in the dwelling. The benefit for households following an intervention for the energy-efficiency improvement in the existing infrastructures of the area under evaluation is not assessed. Changing the contracted energy supplier, the type of fuel used, and the dwelling systems cannot be evaluated by current FP indicators. Our analysis enables those areas with either an inefficient, bad state of preservation or with a lack of infrastructure (in the form of rural and social exclusion zones) to be identified with higher FP rates, since they cannot change their current energy supplier, type of fuel used, or systems (Figure 2). In this context, it is the econometric analysis, from among those previously analysed, which constitutes the most commonly method used to assess the relationship between FP and this factor. 4.2. Energy efficiency page 18 Dwelling quality is the most complex, and possibly the most influential factor to be assessed in an FP situation. Assessing the energy efficiency of a dwelling in order to improve its quality is always a complex task due to the wide range of factors involved: energy-efficiency rating, thermal isolation, heating/cooling systems, ventilation level, state of maintenance, and age of the dwelling. All the indicators analysed allow a relationship to be established, either in a direct or indirect way, between FP and the energy efficiency of the dwelling since they use data based on the energy consumption for the basic needs of occupants (maintenance of an adequate temperature in the dwelling during winter and summer, etc. (Table 1)). The energy consumption of households can vary greatly depending on the energy efficiency of the dwelling: this may lead to double the energy consumption between dwellings of poor energy efficiency (those with an energy-efficiency rating of “E” or “F”) and high energy efficiency (those with an energyefficiency rating of “C” or above). Therefore, if the initial energy consumption of a household in FP is reduced thanks to an energy-efficiency intervention, then the extent of FP in the home can also be reduced and even solved [59,94]. The assessment of the thermal comfort in the dwelling, in relation to a higher probability of the dwelling being fuel poor, constitutes another methodology that allows energy efficiency of the dwelling and FP to be related [56,60]. If the thermal comfort of the dwelling is inadequate, occupants use air-conditioning or heating systems for longer, thereby leading to an increase in energy consumption, and consequently, an increase in their energy bill. This analysis shows that the adequacy of the level of thermal comfort in the dwelling also plays a major role when evaluating FP and the risk of falling into FP (Figure 2). However, special attention is required since the assessment of thermal comfort does depend on the type of household analysed. 4.3. Social and economic poverty The social impact caused by FP, mainly due to the social exclusion of those households in FP and to their low income, can be considered one of the most complex factors to be evaluated, and may be determined in two ways: first, people in FP tend to limit their relationships with their friends and family either in order to avoid being considered monetarily poor, or since they cannot plan encounters at home due to the poor quality of the dwelling; secondly, young people and children tend to attain low academic results and poor job performance since an inadequate environment prevents them from resting and studying adequately when they are at home [95] (Figure 2). The analysed indicators based on household income and expenses, and self-reported household conditions (Table 1), allow the relationship between FP and monetary poverty to be assessed due to their inclusion of household income as a factor. Only the indicators based on household income and expenses allow a direct analysis of the monetary situation of the household to be analysed, and these show higher FP page 19 rates in those households with low income and in a situation of monetary poverty. The relationship established between household income and energy-consumption expense remains the main drawback of these indicators since they may lead to false-negative results regarding the current household situation, since those households whose energy consumption costs fail to reach the minimum required (depending on each indicator) are excluded from being considered in a FP situation. Regarding this point, only HEP indicators (Table 1) help to identify those households that cannot have a minimum energy-consumption expenditure due to the lack of monetary resources. 4.4. Wellbeing and health It is known that children who live in a dwelling in FP can be affected in terms of their school progress and emotional well-being: it is nearly twice as likely for children living in homes with insufficient heating to suffer from respiratory diseases such as asthma, and it has also been noted that children living in FP homes run a higher risk of being admitted to hospital or of needing primary care facilities [95]. In addition, these households have shown a higher risk of suffering from mental illnesses, such as depression, with a probability greater than 25%, whereas there is only a 5% risk in the rest of the population [96]. Overall, children, pregnant women, and the elderly can be considered as vulnerable, and have a higher possibility of suffering from illness [97]. As well as mental problems, it is known that those households with inadequate temperatures during the winter are more likely to suffer from minor health problems, such as colds and influenza, and from worsening arthritis and rheumatism [98] (Figure 2). With this premise, it is known that the relationships between dwelling quality, FP, and households’ quality of life lead to a higher monetary cost for the national health service. The Affordable Warmth Access Referral Mechanism (AWARM, Manchester) [99] and Affordable Warmth (Bolton) [100] are both good examples of the cost-effectiveness of an energyefficiency intervention, whereby the amount of money the national health service has to pay to keep a person alive in perfect health for one year is compared to the cost of an energy-efficiency intervention. This analysis shows that a relationship between household health and FP cannot be established by the indicators analysed in this paper, and therefore no energy-efficiency intervention of a dwelling to improve the quality of household health can be justified. This factor can only be analysed by combining the available data of FP percentage in the analysed areas and its relation to the health of households in terms of the most commonly used indicators: 10%, LIHC, and MIS (Table 1) [55,101–103]. page 20 Figure 2. Identification of vulnerable homes from a multiple-indicator approach. 5. Conclusions The main purpose of this paper is to provide a revision of current FP indicators as regards their ability to identify those households at risk of FP. To this end, the energy vulnerability factors (available infrastructures, energy efficiency, monetary and social poverty, and well-being and health) have all been employed to analyse the effectiveness of current FP indicators, which can be grouped in accordance with the indicators upon which they are based: income-expenses; self-reported conditions; econometric analysis; thermal comfort; and energy efficiency. The main weakness of all these indicators lies in the impossibility for a single indicator to consider all possible factors that bear an influence on the day-to-day activity of households, such as thermal comfort, health, and wellbeing. Consequently, an incomplete and unreliable analysis is provided if they are used in an isolated way, mainly due to “exclusion” inaccuracies, whereby households that should receive benefits are not recognised by the policy and are therefore left unsupported, and “inclusion” inaccuracies, whereby households that are not fuel poor meet the eligibility criteria and are therefore awarded support. It is page 21 therefore necessary to combine various indicators and to analyse their isolated results together to ascertain whether a current and reliable analysis of both the technical characteristics of the dwelling and the state of the household are achieved. The discussion presented in this paper shows the weaknesses of existing FP indicators in the identification of households at risk of FP and leads to the definition of a multiple-indicator approach that gathers as many factors as possible involved in an FP situation. Furthermore, the lack of suitable definitions and indicators in most EU member states leads this research to provide a starting point for future EU studies, in terms of vulnerable homes, as well as answering the last call from the EU Parliament to evaluate whether sufficient protection has been provided to citizens by means of current measures and indicators in an effort to recognise the source of the problem. Finally, although a wide range of methodologies have been employed to yield a solution to the issue addressed, none to date has provided an objective comparison of approaches. This work leads to the optimal recommendations in action plans and energy policies in terms of prioritising the most vulnerable households and of the protection of vulnerable consumers, and to the reflection of vulnerability in policy decisions and methodological implications. Furthermore, those households in greatest need would be supported, public money would be better deployed, and FP resources would be distributed effectively. Funding We thank the Eaga Charitable Trust [GB1088361] for generously awarding a postgraduate student bursary, which has been essential in carrying out this work. References [1] H. Thomson, C. Snell, C. Liddell, Fuel poverty in the European Union: a concept in need of definition?, People, Place & Policy. 10/1 (2016) 5–24. doi:10.3351/ppp.0010.0001.0002. [2] (EnAct), The ENERGY ACTION project., (n.d.). http://www.coldathome.today/. (accessed January 1, 2016). [3] E. Rehfuess, Fuel for life: household energy and health, Geneva: World Health Organization. (2006) 1– 23. [4] S. Bouzarovski, S. Petrova, A global perspective on domestic energy deprivation: Overcoming the energy poverty-fuel poverty binary, Energy Research and Social Science. 10 (2015) 31–40. doi:10.1016/j.erss.2015.06.007. [5] M. González-Eguino, Energy poverty: An overview, Renewable and Sustainable Energy Reviews. (2015). doi:10.1016/j.rser.2015.03.013. [6] G. Thompson, M. Bazilian, Democratization, energy poverty, and the pursuit of symmetry, Global page 22 Policy. 5 (2014) 127–131. doi:10.1111/1758-5899.12103. [7] E. Brown, J. Cloke, D. Gent, P.H. Johnson, C. Hill, Green growht or ecological commodification: debating the green economy in the Global South, Geografiska Annaler, Serie B: Human Geography. 96 (3) (2014) 245–259. doi:10.1111/geob.12049. [8] D.S. Shonali Pachauri, Energy Use and Energy Access in Relation to Poverty, Economic and Political Weekly. 39 (3) (2004) 271–278. [9] S. Bouzarovski, Stefan, Petrova, Saska and Tirado-Herrero, From Fuel Poverty to Energy Vulnerability: The Importance of Services, Needs and Practices, Science Policy Research Unit Working Paper 2014; 25: 1-28. [10] K. Li, B. Lloyd, X.J. Liang, Y.M. Wei, Energy poor or fuel poor: What are the differences?, Energy Policy. 68 (2014). [11] M. Bazilian, A. Sagar, R. Detchon, K. Yumkella, More heat and light, Energy Policy. 38 (10) (2010) 5409–5412. doi:10.1016/j.enpol.2010.06.007. [12] F. Birol, Energy Economics: A Place for Energy Poverty in the Agenda?, Energy Journal. 28 (3) (2007) 1–6. [13] G. O’Brien, A. Hope, Localism and energy: Negotiating approaches to embedding resilience in energy systems, Energy Policy. 38 (12) (2010) 7550-7558. [14] M. Welsh, Resilience and responsibility: governing uncertainty in a complex world, The Geographical Journal. 180 (1) (2014) 15–26. doi:10.1111/geoj.12012. [15] S. Paugan, The spiral of precariousness: a multidimensional approach to the process of social disqualification in France, Policy Press. (1995) 47–79. [16] R.M. Isherwood, B.C., & Hancock, Household expenditure on fuel: distributional aspects, Economic Adviser's Office, DHSS, London. (1979). [17] B. Boardman, Fixing Fuel Poverty. Challenges and Solutions, Earthscan, London, 2010. [18] M. Braubach, A. Ferrand, Energy efficiency, housing, equity and health, International Journal of Public Health. 58 (3) (2013) 331–332. doi:10.1007/s00038-012-0441-2. [19] J. Rosenow, R. Platt, B. Flanagan, Fuel poverty and energy efficiency obligations - A critical assessment of the supplier obligation in the UK, Energy Policy. 62 (2013) 1194–1203. doi:10.1016/j.enpol.2013.07.103. [20] B. Boardman, Achieving zero. Delivering future-friendly buildings, Oxford, Oxford, 2012. [21] European Commission, European Fuel Poverty and Energy Efficiency (EPEE) Project, 2009. https://ec.europa.eu/energy/intelligent/projects/en/projects/epee (accessed June 11, 2016). [22] E. Commission, Energy Roadmap 2050, communication from the Commission to the European Parliament, the Council, the European Economic and Social Committee and the Committee of the page 23 Regions. (n.d.) 1–20. doi:10.2833/10759. [23] T. Peneva, Mechanism for Protection of Vulnerable Consumers in Bulgaria, 2016. http://fuelpoverty.eu/2016/10/18/mechanism-for-protection-of-vulnerable-consumers-in-bulgaria/. [24] A. De quero, B. Lapostolet, Plan Bâtiment Grenelle Groupe de travail Précarité énergétique Rapport, Ministère de l’écologie, de l’énergie, du développement durable et de la mer. París, 2009. [25] Department of communications, energy & natural resources. A Strategy to Combat Energy Poverty, (2016). http://www.dccae.gov.ie/energy/en-ie/Energy-Efficiency/Pages/Combating-Energy- Poverty.aspx# (accessed October 3, 2016). [26] D. Strakova, Energy Poverty in Slovakia, 2014. doi:http://dx.doi.org/10.2139/ssrn.2546758. [27] Departement of Energy and Climate Change (DECC), Fuel Poverty Methodology Handbook, (2010). [28] Departement of Energy and Climate Change (DECC), Annual Report on Fuel Poverty Statistics, (2013). [29] S. Tirado Herrero, J.L. López Fernández, P. Martín Gracía, Energy poverty in Spain. Employment generation potential derived from housing rehabilitation, 1a, Madrid, 2012. [30] S. Tirado Herrero, L. Jiménez Meneses, J.L. López Fernández, J. Martín Gracía, E. Perero Van Hove, Energy poverty in Spain. Trend analysis, 1a, Madrid, 2014. [31] S. Tirado Herrero, L. Jiménez Meneses, J.L. López Fernández, E. Perero Van Hove, V.M. Irigoyen Hidalgo, P. Savary, Poverty, vulnerability and energy inequality. New approaches to analysis, 1a, Madrid, 2016. [32] S. Tirado Herrero, L. Jiménez Meneses, J.L. López Fernández, E. Perero Van Hove, V. Irigoyen Hidalgo, Energy Poverty in Spain. Towards a system of indicators and a state action strategy, 1a, Madrid., 2018. [33] J.D. Healy, Excess winter mortality in Europe: a cross country analysis identifying key risk factors, Journal of Epidemiology & Community Health. 57 (10) (2003) 784–789. doi:10.1136/jech.57.10.784. [34] J.C. Romero, P. Linares, X. López Otero, X. Labandeira, A. Pérez Alonso, Energy poverty in Spain. Economic analysis and proposals for action, Madrid, 2015. [35] R. Castaño-Rosa, J. Solís-Guzmán, M. Marrero, A novel Index of Vulnerable Homes: Findings from application in Spain, Indoor and Built Environment. (2018) 1420326X18764783. doi:10.1177/1420326X18764783. [36] J. Hills, Fuel Poverty: The Problem and its Measurement – Interim Report of the Fuel Poverty Review, CASE Report 69, 2011. [37] R. Moore, Definitions of fuel poverty: Implications for policy, Energy Policy. 49 (2012) 19–26. doi:10.1016/j.enpol.2012.01.057. [38] R. Schuessler, Energy Poverty Indicators: Conceptual Issues, Centre for European Economic Research (ZEW), Discussion Paper Series. 14 (14) (2014) 37. [39] J. Bradshaw, S. Middleton, A. Davis, N. Oldfield, N. Smith, L. Cusworth, J. Williams, A Minimum page 24 Income Standard for Britain: What people think, Joseph Rowntree Foundation, York, 2008. [40] D.S. Johnson, J.M. Rogers, L. Tan, A Century of Family Budgets in the United States, Monthly Labor Review. 124 (2001) 28–45. [41] J. Hills, Getting the measure of fuel poverty. Final Report of the Fuel Poverty Review. Centre for the Analysis of Social Exclusion, London, 2012. [42] I. Preston, V. White, K. Blacklaws, D. Hirsh, Fuel and poverty: A Rapid Evidence Assessment for the Joseph Rowntree Foundation, Bristol, 2014. [43] L. Middlemiss, A critical analysis of the new politics of fuel poverty in England, Critical Social Policy. 37 (3) (2016) 425–443. [44] P. Heindl, Measuring Fuel Poverty: General Considerations and Application to German Household Data, Finanz Arch. 71 (2) (2015) 178–215. doi:10.1628/001522115X14285723527593. [45] K. Rademaekers, J. Yearwood, A. Ferreira, S. Pye, P. Ian Hamilton, D.G. Agnolucci, J. Karásek, N. Anisimova, Selecting Indicators to Measure Energy Poverty, Rotterdam, 2016. [46] J.D. Healy, J.P. Clinch, Quantifying the severity of fuel poverty, its relationship with poor housing and reasons for non-investment in energy-saving measures in Ireland, Energy Policy. 32 (2) (2004) 207– 220. [47] S. Friel, Housing, Fuel Poverty and Health: A Pan-european Analysis, Health Sociology Review. 30 (2) (2006) 194–195. doi:10.5172/hesr.2007.16.2.195. [48] European Comission, European Union Statistics on Income and Living Conditions (EU-SILC), Eurostat, Brussels, 2014. [49] Spanish National Statistics Institute, Survey on Living Conditions (SLC), Madrid, 2014. [50] B. Legendre, O. Ricci, Measuring fuel poverty in France: which households are the most vulnerable?, Energy Economics. 49 (2014) 620–628. doi:10.1016/j.eneco.2015.01.022. [51] R. Miniaci, C. Scarpa, P. Valbonesi, Fuel Poverty and the Energy Benefits System: The Italian Case, Milano, 2014. [52] S. Bouzarovski, J. Cauvain, Spaces of exception: governing fuel poverty in England’s multiple occupancy housing sector, Space and Polity. (2016). doi:10.1080/13562576.2016.1228194. [53] R. Walker, P. McKenzie, C. Liddell, C. Morris, Area-based targeting of fuel poverty in Northern Ireland: An evidenced-based approach, Applied Geography. 34 (2012) 639–649. doi:http://dx.doi.org/10.1016/j.apgeog.2012.04.002. [54] R. Walker, C. Liddell, P. McKenzie, C. Morris, Evaluating fuel poverty policy in Northern Ireland using a geographic approach, Energy Policy. 63 (2013) 765–774. doi:10.1016/j.enpol.2013.08.047. [55] R. Walker, P. McKenzie, C. Liddell, C. Morris, Estimating fuel poverty at household level: An integrated approach, Energy and Buildings. 80 (2014) 469/479. doi:10.1016/j.enbuild.2014.06.004.