Diagnosing Energy Poverty in Portugal through the Lens of a Social Survey

Abstract

Energy poverty (EP) is a crucial socio-economic problem in current society, as it deprives people of a basic standard of living and quality of life. In general, EP is linked to various factors, but it is primarily associated with high energy costs, low household income, and inefficient building structures. Due to the recent energy crisis in Europe, the importance of these factors has escalated. Bearing in mind the above remarks, the REVERTER EU-funded LIFE project will develop and test nine roadmaps in four European countries (Brezovo—Bulgaria; Athens Urban area—Greece; Riga—Latvia; and Coimbra—Portugal) to alleviate EP by addressing the poor energy efficiency of dwellings. To directly observe real-life scenarios in designated pilot regions, questionnaire surveys were conducted, involving approximately 300 households in each pilot area. This paper presents and evaluates the outcomes of the survey conducted in the central region of Portugal. The findings allow for a deeper comprehension of the factors that impact energy poverty in Portugal, spanning socio-economic aspects, housing characteristics, mechanical systems, energy expenses, and behaviors, as well as the awareness of available support initiatives. Drawing from the survey’s insights, novel strategies are suggested to alleviate energy poverty, with a primary emphasis on crafting tailored, efficient policies that address the genuine requirements of households and individuals.

1. Introduction

1.1. Motivation

Energy poverty (EP) is a crucial socio-economic problem in current society, as it deprives people of a basic standard of living and quality of life. Over the past decade, the European Union (EU) has been committed to tackling energy poverty and protecting vulnerable consumers, but the problem is far from solved. According to the latest available data for EP across Europe, in 2022, 9.3% of people in the EU could not afford to keep their homes adequately warm, up from 6.9% in 2021 [1]. Overall, there are several factors associated with EP, such as housing conditions (e.g., dwelling type and type of heating systems used); household characteristics (e.g., size of household); building characteristics (e.g., building size and energy efficiency); and demographic characteristics (e.g., employment status, education, nationality, and gender) [2]. However, EP is mainly connected with three causes: high energy costs, low household incomes, and energy-inefficient buildings [3]. Due to the recent energy crisis in Europe, the importance of these factors has escalated.

Energy poverty is defined in the 2023 Social Climate Fund regulation [4] and the revised Energy Efficiency Directive [5] as “a household’s lack of access to essential energy services that provide basic levels and decent standards of living and health, including adequate heating, hot water, cooling, lighting, and energy to power appliances, in the relevant national context, existing social policy and other relevant policies, caused by a combination of factors, including but not limited to non-affordability, insufficient disposable income, high energy expenditure and poor energy efficiency of homes”.

In the Clean Energy for All Europeans package, adopted in 2019, the EU increased its efforts and made energy poverty a key concept. The reduction in and mitigation of energy poverty have also been increasingly targeted via energy efficiency, decarbonization, and clean energy policies to support a just energy transition for all [6]. As part of their obligation to assess energy poverty in their National Energy and Climate Plans (NECPs) [7], several EU countries have integrated targeted measures in their national strategies and are developing their own definitions, measurements, and monitoring methods as well as solutions to tackle energy poverty.

If EP was already a major concern being tackled by different policies, 2022 brought an unprecedented challenge for European citizens regarding access to energy and energy bills. With the energy crisis, energy poverty became a major news item while worsening significantly [8]. To address this rising challenge, more policy developments emerged at different levels (at the European level, but also at national and local levels). Energy poverty (and poverty in a broader sense) is a serious reality that needs local, flexible, and “tailor-made” public policies, adapted to the local context and involving various players and stakeholders.

According to the most recent statistics published by Eurostat, in 2022, Portugal recorded the fourth highest rate in the European Union (17.5%) of people who were unable to heat their homes properly [1]. Until recently, this problem has been mostly neglected in national decisions and by policymakers. A recent nationwide study found that households may consider it normal and acceptable to feel both cold and hot at home, either in winter or in summer [9]. This can hinder the social recognition of the EP problem and the need to tackle its negative consequences on the well-being and health of the population. Fortunately, after more than two years under public consultation, a long-term strategy for mitigating energy poverty was finally published in January 2024. This Long-Term Strategy to Combat Energy Poverty 2030–2050 (ELPPE) adopts the official definition of energy poverty provided in the European Directive (EU) 2023/1791: “the lack of access by a household to essential energy services, where such services provide basic and dignified standards of living and health, including adequate heating, hot water, cooling and lighting and the energy needed for household appliances, taking into account the national context, social policy and other relevant national policies, caused by a combination of factors, including at least lack of affordability, insufficient disposable income, high energy expenditure and poor energy efficiency of dwellings”. It also provides a strategic framework for tackling EP in Portugal.

This paper presents the findings of a survey conducted in the central region of Portugal to assess the state of energy poverty. The main objective of the study was to diagnose the energy poverty levels in the central region of Portugal and identify the socio-economic groups affected by energy poverty, using the data collected in a social survey.

1.2. Related Work

Energy poverty typically stems from a combination of factors, including high energy costs, limited income, and homes that are inefficient in terms of energy use. This last issue can be affected by factors such as the age, condition, and construction materials of the building envelope, as well as the energy efficiency of appliances. Additionally, factors like one’s residential status (whether a person owns or rents their home) and the type of heating/cooling system in place also play a role in determining the ability to make energy-related improvements. Individuals with lower incomes often reside in dwellings with subpar insulation and frequently rely on second-hand or outdated appliances that are not energy-efficient. Moreover, they often have to manage their electricity and gas expenses through pre-payment systems, which can result in them incurring higher unit costs compared to those using monthly billing systems [10]. The key drivers of energy poverty are frequently considered to be factors like access/availability, affordability/income inequality/the capacity for investment, the energy efficiency of the envelope and appliances, the needs and expectations of families, practices/awareness related to knowledge about good household use, the finance available for support schemes, and low-cost passive measures [11].

Several works have assessed EP globally. Al Kez et al. [12] compared different methods for measuring EP, considering spatial factors, household preferences, home standards, and cultural differences among countries. Ntaintasis et al. [13] comparatively evaluated objective, subjective, and composite indicators used for measuring and analyzing EP, and Che et al. [14] performed a comprehensive assessment of global energy poverty with an integrated approach. Other works have been focused on analyzing EP at the EU level. Kashour and Jaber [15] evaluated EP for the EU, and Szamrej-Baran [16] assessed the effects of the COVID-19 pandemic on EP in EU countries. Bouzarovski and Tirado Herrero [17] analyzed the spatial and temporal trends in the national-scale patterns of energy poverty. The findings suggest that the traditional economic division between the core and periphery extends to energy poverty, with Southern and Eastern European EU Member States exhibiting notably higher rates. Bouzarovski et al. [18] analyzed the existing policy efforts to address EP at the governance level of the EU and its constituent Member States, providing an agenda for future research and policy that emphasizes key scientific and decision-making hurdles within both the European and global landscapes.

According to several indicators generally used to assess energy poverty, Portugal has been highlighted as being among the countries that are most vulnerable to energy poverty in the European Union [19]. However, there are no official data about the share of Portuguese living in energy poverty. Depending on the indicators used, it is estimated that the proportion of Portuguese households living in energy poverty ranges from 15 to 24% of the total [20,21]. Likewise, regarding excess winter mortality, which is considered an indicator of energy poverty due to the negative impacts on the health of living in inadequately heated environments, Portugal presents one of the highest rates in Europe. Although it has a mild climate, surprisingly, in Portugal there is still a large increase in the number of people dying in periods of cold weather because they cannot afford to heat their homes [22] or do not understand the danger of the improper use of heating devices. According to a national newspaper, between 1 November 2018 and 25 January 2019, 19 Portuguese died from cold weather-related accidents: they lost their lives in fires caused by bad electrical connections or heating appliances, from problems with poorly insulated fireplaces, or due to carbon monoxide inhalation [23]. Regarding cooling, around 36% of people living in dwellings are not able to keep their house comfortably cool during the summer [24].

Combining a low household income and high electricity costs with a relatively mild climate (mild winters), it soon becomes clear that the rationale when it comes to managing the household budget will be to leave air conditioning (heating and cooling) as the last priority, and people will use low-efficiency equipment (individual heaters and open fireplaces). Portugal is among the European countries where poor-quality housing construction and an inability to maintain the thermal comfort of the house all year round prevail. A recent national study [9] found that households may consider it normal and acceptable to feel both cold and hot at home, either in winter or in summer. This can hinder the social recognition of the energy poverty problem and the need to tackle its negative consequences on the well-being and health of the population. However, until recently, this problem has been mostly neglected by national decisions and policymakers.

Nevertheless, driven by European legislation, Portugal has defined the condition of an economically vulnerable consumer as being the beneficiary of extraordinary social support, whose percentage discount is applicable to the invoice for electricity and natural gas. Although these billing support measures (such as social tariffs and energy price reductions) mitigate the financial burden of the most vulnerable families, they have little effect on improving energy comfort. The multidimensional nature of the concept of energy poverty makes it possible to understand that other factors require urgent action since the number of energy-poor Portuguese families is very high and energy poverty seriously affects living conditions, health, and comfort.

In Portugal, several studies have assessed the impacts of EP and its drivers. Oliveira Panão [25] calculated expenditure-based indicators, using raw data from national statistics to evaluate the impact of energy poverty in Portugal. Several studies also characterized the impact of building characteristics on EP. Gouveia et al. [19] developed a high-resolution spatial-scale composite index, focusing on space heating and cooling, to map energy-poor regions and identify hotspots for local action. The proposed index combines socio-economic indicators of the population with a consideration of the building’s characteristics and energy performance. Matos et al. [26] analyzed the EP evolution in Portugal in the EU context, considering the impact of thermal building regulations and energy efficiency policies. Palma et al. [27] analyzed the EP gap for Portuguese residential dwelling stock regarding thermal comfort attainment at a high-resolution geographical scale. In such a context, the emerging one-stop-shop models designed to streamline the deployment of building renovations were critically reviewed by Sequeira and Gouveia [28].

Some studies have implemented surveys to characterize EP in Portugal. Gouveia et al. [29] combined the data collected from smart meters’ registries with socio-economic data, collected from door-to-door surveys, to comprehend the scope of and factors influencing energy usage among two distinct consumer demographics, namely, fuel poverty and fuel obesity groups. Horta et al. [9] conducted interviews with 100 households in 10 hotspots across the country to perform a detailed quantitative analysis combining the use of an energy poverty vulnerability index and mapping. Castro and Gouveia [30] implemented a survey to analyze students’ perceptions of EP and explore their vulnerability to EP, considering their background and the city they live in, using Montevideo in Uruguay, Lisbon in Portugal, and Padua in Italy as the case studies. Valente and Gouveia [31] assessed upper secondary school students’ perceptions of energy poverty at home and thermal comfort inside classrooms.

Despite the previous evaluations, an updated assessment of energy poverty in Portugal is needed in order to consider the impact of the recent increase in energy prices, and the impact of rising inflation generally, on energy poverty. It is also important to evaluate a larger number of aspects associated with EP, considering not only the characteristics of the buildings but also the socio-economic situation.

1.3. Contribution

The REVERTER project aims to alleviate energy poverty (EP) by developing nine roadmaps to improve the energy efficiency of dwellings. These are tailored to building stock characteristics, vulnerable households, and climate conditions. These roadmaps will prioritize the worst-performing homes, address split-incentive dilemmas, and tackle market, information, and behavioral failures through “one-stop shops” (OSS) that enroll vulnerable households in subsidized energy efficiency programs. To test these roadmaps, pilot networks will be established in Brezovo (Bulgaria), the Athens Urban area (Greece), Riga (Latvia), and Coimbra (Portugal), covering diverse climate regions and socio-economic conditions. Engagement with local, national, and EU stakeholders and experts will ensure these roadmaps contribute to future EP reduction policies.

To gain direct insight into real field situations in the selected pilot areas, questionnaire surveys were implemented, involving about 300 households in each pilot, to help prepare the categorization of multiple aspects of EP. It has been largely advocated that researchers combine both objective and subjective measures to identify EP [32], but these measures alone do not fully capture the adverse effects of EP. This paper presents and assesses the results of a survey implemented in Portugal (focused on the central region), which was based on a questionnaire with 25 questions, considering four complementary dimensions: (a) general information about the local context, (b) details about the building and systems, (c) information about household habits, and (d) demographics.

The main aim of the study was to diagnose the energy poverty levels in the central region of Portugal and identify the socio-economic groups affected by energy poverty. The main research questions were as follows: What is the impact of building characteristics and mechanical systems on comfort levels and energy bills? What are the reasons why households do not apply for support programs and how much are they willing to pay for efficiency improvements? Is there a correlation between income, energy bills, and self-declared comfort? Considering different options in terms of indicators, what are the levels of energy poverty?

The results contribute to understanding the factors influencing energy poverty in the central region of Portugal, touching on topics such as socio-economic issues, housing characteristics, mechanical systems, energy costs and habits, and knowledge about existing support programs. Based on the findings of the survey, new strategies are proposed to mitigate energy poverty. These are mainly focused on tailor-made effective policies addressing the real needs of households and people in order to ensure equitable access to basic energy services and promote social and economic inclusion for vulnerable populations. The results of this study will be used to map households’ energy needs and promote energy-saving measures and support for households to reduce their energy cost burden in the short and the long term.

1.4. Paper Organization

The remainder of this paper is structured as follows: Section 2 presents the overall methodology of the REVERTER Project, as well as the methodology used for the survey. Section 3 presents and assesses the results obtained from the social survey. Section 4 diagnoses and discusses the energy poverty situation. Finally, Section 5 presents the conclusions of this work.

2. Materials and Methods

The implemented survey was carried out as part of the REVERTER project, being therefore the methodology for the survey a part of the overall methodology of the REVERTER Project.

2.1. The REVERTER Project

The REVERTER project [33], submitted under the Call LIFE Clean Energy Transition (LIFE-2021-CET) and, more specifically, addressing the topic LIFE-2021-CET-ENERPOV “Addressing building-related interventions for vulnerable districts”, aims to effectively alleviate EP through the deep renovation of houses according to the “worst first” principle. Within this overall context, REVERTER has eight specific objectives:

• Create nine roadmaps.
• Facilitate the renovation of more than 800 houses during the implementation of the project and within five years after its end.
• Create a positive impact on more than 3000 vulnerable people.
• Demonstrate the effectiveness and replicability of the proposed solutions among 20,000 energy-vulnerable households.
• Trigger the investment of about EUR 8 million in sustainable energy and primary energy savings/renewable energy to generate 12.3 GWh/year during the implementation of the project and within 5 years after its end.
• Use 10 existing and new “tailor-cut” complementary indicators to measure energy poverty.
• Create a knowledge database and analyze deep renovation measures by employing economic, environmental (via LCA), technical, and social criteria, and by using cost–benefit and multicriteria approaches.
• Identify viable financial schemes, support best practices, and shape future policies aimed at alleviating EP through energy retrofits by creating/adapting 15 pieces of legislation, policies, or strategies during the implementation of the project and within 5 years after its end.

Building on the achievements of previous and ongoing projects and initiatives aimed at maximizing the effectiveness of the roadmaps, REVERTER is based on five distinct but interdependent and mutually supportive pillars:

• The identification of energy-vulnerable households;
• The analysis of building stock;
• The assessment of deep renovation measures;
• The establishment of “one-stop-shops” (OSS);
• Capacity building using energy ambassadors, awareness, and training campaigns.

2.2. Social Survey

To gain direct insights into real field situations in the selected pilot areas, questionnaire surveys were implemented, involving at least 300 households in each pilot. The questionnaires were focused mainly on energy issues; house and heating system characteristics; and market, administrative, behavioral, and informational barriers. A common set of questions was developed for the social surveys, but some questions were also adapted to the local contexts.

The questionnaires were self-administered (delivered online or in paper-and-pen formats, either in person or through mail) or researcher-administered (i.e., face-to-face interviews that take place by phone, in person, or online). Regardless of the survey administration method, particular attention was given to ethical and data protection considerations. Appropriate information sheets and consent forms were prepared for that purpose and the ethical guidelines were strictly followed.

The results of the questionnaire surveys were used to perform the following tasks:

• Test the EP identification methodology;
• Develop a set of tailor-made materials that will be used in the one-stop shops to support community capacity-building programs;
• Recruit households willing to participate in REVERTER’s field activities, e.g., via home visits;
• Facilitate the development of roadmaps;
• Identify best practices in terms of delivering energy advice during local engagement initiatives/campaigns for low-income households;
• Recognize key incentives for energy efficiency retrofitting for each category of property tenure type, which is necessary for the design of the one-stop shops.

In Portugal, the first concern was to ensure the accomplishment of GDPR rules [34]. GDPR stands for General Data Protection Regulation. It is a European Union law on information privacy, which went into effect on 25 May 2018. This legal framework sets guidelines for the collection, processing, storage, and transfer of personal information from individuals who live in and outside of the European Union. The selected platform for the survey was LimeSurvey 5.5 [35], an open-source survey platform with all the data stored in the University server, ensuring complete control over data storage and compliance. With proper configuration and security measures, LimeSurvey can ensure GDPR compliance.

The social survey aimed to study the impact of housing on the quality of life of households in Portugal. The questionnaire was appropriately designed to collect all the necessary data required to assess energy poverty, taking into consideration diverse contexts and cultural habits. It included 25 questions, mostly multiple-choice questions—where the respondent was asked to choose the best-fitting answers—and was organized into four complementary dimensions: (a) general information about the local context, (b) general information about the building and systems, (c) general information about the household and habits, and (d) demographics. These were arranged as followed:

(a)

Local context (national/regional/neighborhood)—this section was designed to identify the locally available support mechanisms, users’ awareness of their existence, and possible improvements to them. From this knowledge, using the strategies in the other pilot cities, existing mechanisms can be improved and inspire the development of new ones.

(b)

Information on building envelopes and comfort systems—the information collected in this section helps to discover more about the building and the way it is used, helping to design roadmaps.

(c)

Household-related questions, which aim to validate indicators used to quantify energy poverty and existing European statistics with evidence-based information.

(d)

Individual household demographics, including the postal code, are used in order to allow matching between the answers with local climate conditions.

The objective of the questions was to collect the needed data in order to diagnose the energy poverty levels in the central region of Portugal and identify the socio-economic groups affected by energy poverty. In such a context, the questions allowed us to provide the needed data to answer the research questions, namely:

• The impact of the building characteristics and mechanical systems on the comfort levels and energy bills;
• The reasons to not apply for support programs and the willingness to pay for efficiency improvements;
• The correlation between income, energy bills, and self-declared comfort;
• The levels of energy poverty when considering different options in terms of indicators.

Among other factors, the results of the questionnaire surveys, based on the main approaches being used to measure energy poverty—such as using expenditure data, using the self-assessment of housing conditions (so-called consensual), and comparing the level of energy services to a specific benchmark (direct measurement)—help to test the EP identification methodology advocated by the REVERTER project, which attempts to consider the multidimensional aspects of energy poverty.

To assess energy vulnerability in Portugal, several expenditure-based and consensual indicators were calculated from the survey, as follows:

• An inability to keep the home adequately warm;
• An inability to keep the home adequately cool;
• Cutbacks or restrictions in essential products and services;
• ‘Ten-Percent-Rule’—when the energy expenditure is more than 10% of income [36];
• Arrears in energy bills.

The international community now views poverty as multi-dimensional, encompassing non-monetary factors like health, education, medical care, and living standards, with multi-dimensional poverty measurements playing a crucial role in evaluation [37]. Based on the literature survey carried out at the beginning of this project, REVERTER proposed a new indicator, namely the “simplified LILEE” [12], which is a simplified version of the UK’s Low-Income Low Energy Efficiency (LILEE) indicator. Based on the simplified LILEE, a household is considered energy-poor if its income is below 60% of the median income and if it resides in a low-energy-class home. The simplified LILEE indicator is simple and effective and, most importantly, it allows for the identification of energy-poor households residing in the least energy-efficient homes. In addition to its simplicity and effectiveness, the proposed indicator offers policymakers the ability to analyze various policy scenarios.

Self-completed questionnaires are widely used within energy poverty-related research for collecting information. These are crucial data that have a positive impact on aligning public policies with the high energy costs faced by households. However, since an important target is social housing managed by the municipality, in addition to the online survey, the questionnaire was also filled in through personal interviews in the Coimbra region, covering vulnerable people living in social housing.

Besides the link needed to enter the survey, a brief description of the project and an explanation of the importance of this social survey were provided. For those contacted by email, a PDF version of the survey was provided when respondents preferred to see the overall survey before starting to complete it. The survey was conducted from May 2023 to August 2023. To avoid the risk of non-fulfillment, the project team controlled the collection of replies frequently, downloading the xls file every week with the replies to check the completion rate, the level of quality of the answers, and the total number of responses, thus enabling us to change and or adapt the strategy for dissemination (for example, the social workers of the municipality visited several households to fill in the questionnaires; personal emails were sent to friends of friends; reminders were issued when needed, etc.).

The online questionnaire was distributed through different means, including emailing lists of municipality staff and university staff, social networks, and word of mouth. It is not possible to know exactly how many people received the questionnaire, but it was surely more than 1000. In total, 462 people followed the survey link, and 330 people provided valid answers (33%), with 260 replies fully completed and 70 not fully completed, but with most answers considered valid. The criterion used for the validation of the questionnaires for inclusion in this study was completeness. Then, if outliers were found in critical questions, they were also excluded. Some of the not fully completed answers were excluded due to the low level of information being considered, with a total of 299 answers. The high number of invalid questionnaires seems to be related to people entering the link through social networks. They started to fill in the questionnaire but gave up already on the second question which was related to the building envelope. Among the ones that filled in the full questionnaire, there were some comments about the questionnaire being too long and not simple to understand. The distribution of the questionnaire was random, but the answers were geographically identified. The survey was focused on the central region of Portugal, which has about 1 million residential buildings, and the 299 answers ensured a confidence level of 95% for a margin of error of 5.65%. The results are representative of the central region of Portugal, but may not be representative of the entire country.

3. Survey Results

3.1. Socio-Economic and Demographic Characteristics

A total of 299 answers were valid enough to be included in the analysis of the socio-economic and demographic characteristics, which is presented in

Table 1. Socio-economic and demographic characteristics.

Category	Attributes	N	%
Gender	Female	97	32.4%
	Male	136	45.5%
	Prefer not to answer	66	22.1%
Household members	1	35	11.7%
	2	96	32.1%
	3	85	28.4%
	4	47	15.7%
	5	15	5.0%
	6	4	1.3%
	No response	17	5.7%
Household with children	1	38	12.7%
	2 or more	47	15.7%
Household with students	1	82	27.4%
	2	122	40.8%
	3	26	8.7%
	More than 3	38	12.7%
Household with pensioners	1	40	13.4%
	2	17	5.7%
	More than 2	1	0.3%
Full-time-employed household members	1	87	29.1%
	2	139	46.5%
	3	8	2.7%
	More than 3	1	0.3%
Unemployed	1	25	8.4%
	2	1	0.3%
	3	0	0%
	More than 3	0	0%
Education level	Graduated (technical area)	270	35.9%
	Graduated (humanities)	121	16.1%
	Secondary school	100	13.3%
	Secondary vocational	23	3.1%
	Elementary school	58	7.7%
	Primary school	84	11.2%
	Kindergarten	90	12.0%
	Prefer not to answer	7	0.9%

. The survey attracted a larger number of males, comprising 45.5% against 32.4% females. Regarding the education level, 52% of the households had a university degree, mainly in technical areas, which represented 35.9% of the households. In total, 146 respondents had a university degree in a technical area: 92 males and 54 women. It can be assumed that these numbers provide some evidence about the impact of educational background on engaging in energy-related issues and the relationship of education to citizens’ empowerment and engagement.

Concerning the number of members per household, most households include two persons (32.1%), 28.4% are formed by three persons, 15.7% have four persons, 11.7% have single families, 5% have five persons, and only 1.3% have six persons. In the total sample, 6% of respondents did not provide this information. Regarding the employment status of the members of the household, the rate of unemployment was low. About half (46.5%) of the households had two members working full-time regularly, and 29% and 2.7% had one and three members, respectively, also working full-time. Only 8.4% of households had one unemployed member.

3.2. Income and Expenditure

Table 2. Net monthly income.

Net Monthly Income	N	%
Up to 300 €	11	3.7%
301–600 €	14	4.7%
601–900 €	41	13.7%
901–1200 €	22	7.4%
1201–1500 €	35	11.7%
1501–2000 €	35	11.7%
2001–2500 €	45	15.1%
>2500 €	59	19.7%
Prefer not to answer	37	12.4%

presents the results regarding net monthly income and

Table 3. Monthly household expenditure.

Monthly Expenditure	N	%
up to 500 €	37	12.4%
500–1000 €	89	29.8%
1001–2000 €	95	31.8%
>2001 €	47	15.7%
Prefer not to answer	31	10.4%

shows the monthly expenditure by household. It should be noted that even the respondents with a net monthly income larger than EUR 2500 can still be considered when studying energy poverty. The definition of energy poverty in the revised Energy Efficiency Directive [5] is based on the lack of access to essential energy services and not on income, since energy poverty is broadly recognized as a multidimensional problem. Even though energy poverty is, of course, a matter of economic poverty, in countries like Portugal, with relatively mild weather all over the country, many people who can afford to live more comfortably, invest in energy renovations, improve their indoor comfort, etc., are still energy-poor because of habits, illiteracy (energy and others), cultural beliefs, a lack of awareness of the impacts of inadequate indoor air quality, etc. Traditionally, people used to be at home and feel cold and heat, and just put another blanket in bed in winter, or open the window in summer. This was never seen as a problem, and the practice still prevails, particularly among the elderly.

Table 4. Expenditure to net income.

Expenditure to Net Income	N	%
Less than 30%	6	2.0%
30–50%	56	18.8%
50–70%	54	18.1%
70–85%	31	10.4%
85–100%	45	15.1%
Over 100%	40	13.4%
No response	66	22.1%

shows that a significant part of the sample (57% of the households) spent more than 50% of their income on the monthly expenditure.

Figure 1 shows the relatively low level of economic empowerment seen among the sample. Considering that the share of graduated people in the sample is significant, and as these people are considered to be part of the average class in Portugal, it is not inappropriate to state that most people are struggling generally with their expenses. As can be seen, the share of the population for which the ratio of expenses/income is above 1 is significant. Over 30% of respondents spent more than 85% of their income, and for 17% of the households, the available income was not even enough to cover the monthly expenses. Assuming a natural embarrassment in admitting this fact, it can be deduced that this percentage is most probably higher and the real situation is even more dramatic.

When asked how they would describe their current income in terms of expenses, the households were asked to select one option that best matched their situation. The results are presented in Figure 2. As can be seen, the percentage of respondents who admit living comfortably on their current income is not significant. The lowest-income classes represent the group with the higher number with no replies to this question. This is not surprising, since many authors already identified and addressed similar findings and explain that this is related to a feeling of embarrassment, with respondents preferring not to answer [38].

3.3. Housing Characteristics

Regarding the housing characteristics, as can be seen in Figure 3a, most houses were built between 1960 and 2010 (75%), are mainly built on concrete structures, and have poor insulation. The second largest share of houses was built after 2010 (16%), complying with newer regulations, and after the first thermal building code was introduced in Portugal. As presented in Figure 3b, the most common type of accommodation is an apartment.

When looking at the average floor areas (

Table 5. Average floor areas (in m²).

Construction Year	Max	Min	Average
Built before 1960 (thick walls, high inertia) or of vernacular type	300	51	135.9
Built between 1960–2010 (concrete structure and low insulation)	600	30	135.5
Built after 2010, with newer regulations	300	85	153.3

) per period of construction, it is not possible to provide any conclusions. However, looking at the minimum, maximum, and average in each category, it is possible to observe a slight tendency for houses in Portugal to increase in area. As presented in Figure 4, most buildings have two or three rooms.

To ask about the characteristics that influence their need/possibility to intervene, the households were faced with a set of questions evaluating aspects such as the tenure of the ownership status of the property. The respondents were presented with several options that could influence their need and or possibility to intervene. Only 26% of respondents owned their house without financial obligations (Figure 5).

It can be seen in Figure 6 that the houses built between 1960 and 2010, before the first thermal building code was adopted in Portugal, use a larger amount of energy. Furthermore, it is also evident that older houses, built with thick walls and high inertia, also have better performances. Of course, this should be taken with caution, because the number of houses in these two categories is also smaller. Nevertheless, these analyses clearly show a direct relationship between the energy bill and the envelope of the building.

3.4. Mechanical Systems

Figure 7 presents the self-declaration of the perceived comfort for the main options of heating systems. In such context, the option “Extremely hot and cold” means simultaneously “Extremely hot in summer” and “Extremely cold in winter”. It is possible to see that radiant floor and central systems are the solutions that provide the best comfort. It is also interesting to see that those households indicating fewer comfort issues at home have solar photovoltaic (PV) and solar thermal systems installed (Figure 8).

Regarding cooling (Figure 9), about 41% of households have a cooling system, mostly local systems; about 24% of households have one piece of equipment, mostly for one room; central systems are only available in 7% of the households; and 10% use a portable system to meet their cooling needs.

The average ownership rate of air conditioning in Portugal is estimated to be around 27% according to a recent survey [39]. Such numbers are significantly lower than the presented in the results of this survey, achieving a penetration of about 41%. This can be explained by the relatively high percentage of educated people with university degrees, and it is logical to assume that this segment of society has more significant economic power purchase. However, it should be noted that, in a country where the average salary is equivalent to the minimum salary of other EU countries [40] and the energy price is higher, and as more than 70% of the building park has low Energy Performance, the sample is adequate for studying energy poverty. Indeed, if the social policies in place in Portugal are considered, the middle class faces higher problems paying the bills than very poor people, who face support and incentives from the government, since middle-class purchasing power has been decreasing [41].

Market studies also indicate a large difference in the penetration of this equipment according to the related region. For those surveyed in the south, the figures are 73% higher than the national average. There are also significant differences between social classes, with the figures falling from 29.6% among the highest-income group to 17.7% among the lowest-income category.

3.5. Energy Costs and Habits

When asked about heating and cooling habits, households had the possibility to choose the best-fitting answers from the different groups in terms of the number of rooms being heated and cooled; the duration of these processes; and the set point temperatures, both in winter and summertime (Figure 10).

One important limitation of this type of survey is the “denial of reality bias” that many researchers point out [38]. Energy-poor people might deny seeing themselves as being in an uncomfortable situation and, therefore, choose not declare it. In this survey, only 8 respondents admitted to not cooling the house because they cannot afford the bill, whereas, regarding heating, 46 admitted that they do not heat because they cannot afford to pay the bill. Nevertheless, 159 indicated an indoor temperature lower than 18 °C in winter, which is below the recommended indoor temperature in winter. This is a sign of hidden energy poverty and stresses the relevance of using composite indicators to evaluate EP [42]. The survey confirmed that being energy-poor does not only occur in winter since the inability to cool a household during the summer months is also clear. The majority (59%) of respondents did not have any air conditioning, and when air conditioning was available, they indicated a low utilization rate (about 30% only cool the rooms being used and 9% only cool the main rooms).

3.6. Energy Efficiency Interventions and Awareness

The survey in Portugal started with the local context in order to identify national use levels and cultural habits. The first question aimed to identify the available support mechanisms, both at the local and national level, and characterize the user’s awareness of existing support programs and initiatives (Figure 11). Respondents were given three potential measures, namely, the social energy tariff, support to improve energy efficiency, and one-stop shops, and they could also provide free text if they were aware of other interesting initiatives. Regarding social energy tariffs, about 10% are covered by a social tariff and 50% are not eligible. The remaining do not know about it and did not apply. Since the social tariff is attributed automatically, it can be assumed that 40% of respondents are not aware of it (and, also, are not eligible).

Looking at the national or local programs to support energy efficiency, only 9.6% applied and received support, and 55% did not apply, including 23% who were not aware of such programs and 9% who were not eligible. Awareness about one-stop shops seems to be missing and or not popular since 50% of respondents replied that they do not know and about 42% did not apply. Other support programs identified by a few respondents include renewable energy communities, incentives for solar PV and solar thermal systems, and some kind of local support given by the municipality for building renovations.

Those who did not apply for support programs and initiatives had the chance to indicate the 3 main reasons why from a list of 10 options and were asked to rank the three reasons according to the order of importance (Figure 12). The most voted reasons were excessive bureaucracy, uncertainty about what to do in the house, and the complexity of the process, followed by the low share of the subsidy provided and uncertainty about its real impact on energy costs.

When comparing the results obtained by crossing the support mechanisms and incentives to energy efficiency with each tenure type, Figure 13 characterizes the situation within this sample, namely:

The social tariff (Figure 13a) is mainly attributed to households living in social housing, which is logical because the criterion to access social housing is associated with income level, vulnerability, etc.

Those applying to support schemes to improve energy efficiency (Figure 13b) are usually the households that own property and do not have bank loans, rather than those most in need. This can be an indication of a shortage of in terms of budget, and also education level.

When comparing programs and supports by tenure and level of education (Figure 14), it is no surprise that those with a technological degree are keener on energy issues and therefore are applying more frequently for energy efficiency support measures. Those with low levels of education typically have a lower income and can receive subsidies from the state, like the social energy tariff, and are not so engaged with applying for support because of illiteracy, but also because these are usually complicated processes and the energy tariff is a disincentive to implementing energy efficiency measures.

In the potential scenario of carrying out improvements in their homes, households were asked how much money they could give on a monthly basis to repay the costs of the work. Figure 15 shows that very few households are willing to pay a reasonable amount per month for renovations. There were 26 completed surveys with no replies (8.7%), and 66 households (22%) who would not pay anything as they indicated zero euros, which was not even included in the options provided. Overall, 18 households indicated other amounts and provide interesting replies:

• Depends on the upfront costs;
• The house is rented and there is no keenness to spend money on retrofits;
• A retired person indicated that the amount received per month does not allow other expenses than the basics;
• Others indicated the amounts of EUR 500, 1000, and 5000 at once, but that it depended on the payback (5 years’ payback time was indicated by one respondent).

4. Diagnosis of Energy Poverty

This section diagnoses and discusses the energy poverty levels, first by comparing the energy bills, income, and comfort levels. Then, several energy poverty indicators are calculated, namely, the 10% rule, the arrears in energy bills, and the multidimensional indicators ‘Low Income/Low Energy Efficiency’ and ‘Low Income/High energy Cost’.

4.1. Energy Bills, Income and Comfort

Taking into account that the survey was conducted in the summertime and therefore there should be some bias in the replies when they were asked about the comfort they feel, self-declared indicators need to be taken with caution. As can be seen in Figure 16, the correlation between the energy bill and income is not relevant in this sample, since there is no direct correlation between both factors. Of course, the sample may not guarantee robust conclusions, but this exercise provides a rough estimate of the share of the monthly energy bill in winter concerning the available income per month and gives an indication of the effort that low-income people undertake to pay their energy bills. It should be highlighted that the results are representative of the central region of Portugal, but may not be of the entire country.

In general, the survey results do not show a direct relation between household income and energy bills. However, there is an influence of the climate region on the overall energy expenses. Guarda is the coldest city in the sample, located in the inner part of Portugal, and is a “rural” city that is quite small and has a low level of industrialization. The average salaries are among the lowest, and the energy expenditure per person and the percentage of energy expenditure in the income are significantly higher. Those people already suffer from living in a peripheral region with harder climate conditions. It is therefore socially just if incentives include some bonification for those living in remote areas.

Analyzing Figure 17, it is possible to conclude that energy poverty is clearly related to economic poverty and geographical location. Even if care should be taken as this sample is not robust in terms of representativeness, it is clear that, in winter, no matter whether the benchmark used is the monthly cost per area or monthly cost per person, those living in the inner parts of Portugal require higher levels of effort to pay the energy bills.

The self-declared comfort (Figure 18) assessment was compared using the average income, expenses, and energy bill per month within the same region, Coimbra. It is not possible to establish any correlation between the level of income and the self-declared comfort. Those in the highest income range also represent the highest percentage saying the house is extremely hot and cold for this climate. Paradoxically, lower-income households, up to 900 € per month, are among those who self-declared that they were most comfortable, except for one week per season, and adequately prepared for the climate. A possible explanation for this phenomenon, based on the interviews carried out in social houses, is the embarrassment and the fear of losing their home (in case of complaints). These are mostly elderly people living with very low pensions, who are used to living with low temperatures inside.

4.2. Energy Poverty Indicators

Figure 19 presents the distribution of energy bills (as a percentage of income) for the lowest-income group of households in the sample (Figure 19a) and for all respondents (Figure 19b).

Looking at the overall sample, 88% of respondents’ winter energy bills represent less than 10% of their income, and 12% of surveyed households spend more than 10% of their net income on energy services. However, looking only at the lower-income households (with a net income lower than 800 € per month), 29% of surveyed households spend more than 10% of their net income on energy services. Among these, more than 95% indicate they struggled to pay the bills. Overall, 70% of respondents’ winter energy bills represent less than 10% of their income, and 30% represent more than 10%.

In total, 276 households replied to the question about arrears in energy bills (Figure 20). Some 235 households (85%) declared they had never been in arrears; 9 declared that they had already been in arrears for electricity; and 5 had already been in arrears for gas. Overall, 20 households indicate they are rarely in arrears and that, if this happens, it is because they forgot the deadline.

Although energy expenses represent a considerable share of household income, when looking at the households with debts to public utilities, the survey indicates that 8.4% of households have failed to pay the energy bills in the last 12 months. The official statistics indicate that 4.5% of the population in Portugal has debts to public utilities, where energy is included, compared with the EU average of 6.2% [43]. Considering the results of the survey, it seems the trend has been increasing over the last year in Portugal.

However, respondents declaring they are not in arrears means little about the real energy poverty situation (Figure 21). This indicator is irrelevant because having no debts to public utilities does not give any indication of the effort people make to pay the bills. It also does not give any indication about the comfort levels they are living in. Most people make an effort to pay the electricity bill, because they fear being disconnected, but live with high restrictions on the consumption of goods, with a strong impact on comfort and health. It is important to remember that this sample’s representativeness had some bias, as the vast majority of replies were from the Coimbra region, where the standard of living is higher than in the majority areas of Portugal.

When the arrears are analyzed together with the number of children within the households (any age), it seems that there is no positive correlation (Figure 22); on the contrary, this seems to be logical, since a family always tries to protect its children from vulnerabilities. It is however possible to infer that single parents face higher constraints when paying energy bills. Looking closer at the group of ‘single and one child’ and the monthly income, it is possible to understand that those parents living with the minimum salary (between 601–900 € per month) are the ones more often in arrears.

Figure 23 shows that the share of households that indicate overall arrears in energy bills and living in a house that is extremely hot in summer and cold in winter, disaggregated by income range, is aligned with the results of previous analyses pointing to the same group of people facing more constraints: those whose income is in the range 601–900 € per month. When households were asked if they felt that they had been forced to restrict other essential needs, they were offered a list of options to select as many as matched their actual situation. Surprisingly, as Figure 24 shows, only 10% of households overall admitted the need to reduce expenses. Among those, the most voted for options were reducing transport (37.6%); reducing medical treatments, including medicine intake and consultations (31.3%); reducing the number of heating hours (28.8%); and thermostat regulation to reduce heating (32.3%).

Multidimensional indicators were also evaluated. The quantile-based ‘Low Income/Low Energy Efficiency’ (LILEE) indicator was based on six categories of equivalized income and on three classes for home energy efficiency, depending on the type of windows and windows condensation, considering the following:

• Low efficiency—single-glass window and window condensation;
• Medium efficiency—double-glass windows and window condensation;
• High efficiency—double-glass windows, thermal cut, and no condensation in windows.

The considered levels of income were AS FOLLOWS:

• Low income—≤1100 €/month;
• Medium income—>1100 €/month and ≤1700 €/month;
• High income—>1700 €/month.

The results of the LILEE indicator are presented in

Table 6. LILEE indicator.

Home Energy Classe	Income Ranges
	≤800 €	≤1100 €	≤1400 €	≤1700 €	≤2200 €	≤3000 €
Low	5.0%	4.3%	5.7%	2.1%	3.6%	2.9%
Medium	5.0%	2.9%	1.4%	5.0%	3.6%	7.1%
High	8.6%	5.0%	10.0%	9.3%	9.3%	22.9%

. According to the LILEE quantile-based indicator, the percentage of energy-vulnerable households (i.e., low-income households living in very low energy efficiency homes) is 9.3% (dark-shaded area in Table 6). This percentage increases to 17.2% if households at medium risk due to low income and living in medium-efficiency homes are considered (light-shaded area in Table 6).

The quantile-based ‘Low Income/High energy Cost’ (LIHC) indicator was based on the equivalized net income and equivalized winter energy costs of the household (the only value available for most households). We considered there to be a high energy cost when the energy bill exceeded 5% of the income. The results are presented in

Table 7. LIHC indicator.

Energy Bill	Income Ranges
	≤800 €	≤1100 €	≤1400 €	≤1700 €	≤2200 €	≤3000 €
20–30 €	2.3%	1.4%	1.8%	1.8%	0.0%	0.5%
40–50 €	3.2%	3.2%	3.6%	2.3%	2.3%	1.4%
50–60 €	2.7%	1.4%	3.2%	2.3%	2.7%	3.6%
60–80 €	3.2%	2.7%	0.5%	1.8%	5.0%	5.0%
80–100 €	1.4%	1.8%	2.3%	2.7%	3.2%	5.0%
100–150 €	2.7%	0.0%	1.8%	2.7%	5.0%	7.2%
>150 €	0.9%	0.0%	0.5%	0.5%	0.9%	4.1%

, being the objective in terms of calculating the quantiles of equivalized income and equivalized energy costs for the ranges represented. The percentage of households with higher risk due to low income and high energy costs is 15.4% (dar-shaded area in Table 7). If those households with medium income risk but high energy cost are taken also into account (light-shaded area in Table 7), the total percentage of energy-vulnerable households based on the LIHC quantile-based indicator is 23.2%.

5. Conclusions

The REVERTER project is focused on creating roadmaps with which to enhance energy efficiency in residential buildings. To gain insight into real-world conditions, surveys were conducted in each pilot area, involving approximately 300 households, to assess various factors related to energy poverty. This paper evaluates the survey results from Portugal, providing insights into socio-economic factors, housing characteristics, mechanical systems, energy costs and habits, and the awareness of support programs. The main objective was to diagnose the energy poverty levels in the central region of Portugal and identify the socio-economic groups most impacted by it.

Regarding the impact of the building characteristics and mechanical systems on comfort levels and energy bills, the social survey results align with the existing statistics and literature on building stock. Most respondents live in houses built between 1960 and 2010, with concrete structures and low insulation levels. Most buildings have double-glazed windows, but they are mainly without frames with thermal cuts. Therefore, most buildings have a low level of efficiency, leading to high energy bills and low comfort levels. Heating is primarily supplied via electricity and the most common heating systems are local and portable systems. The buildings with air conditioning typically only have one unit for a single room. The low efficiency of the mechanical system has a strong impact on comfort levels, since the options identified by the households as providing the best comfort levels (radiant floor heating and central systems) are present in a low number of households.

Comparing self-declared comfort among different socio-economic households, including average income, expenses, and energy bills, reveals no correlation between income level and comfort. Surprisingly, the highest-income households report the highest discomfort levels, and lower-income households report the highest comfort levels. One possible explanation for this phenomenon is that lower-income households may feel embarrassed to admit their discomfort, coupled with cultural habits in Portugal, where many people consider it normal to endure some level of thermal discomfort and often prioritize heating and cooling as the lowest concern in their household budgets.

Regarding support mechanisms, households that own their property without loans apply for support more often than those in greater need, suggesting a budget shortage and educational gap. The main reasons for not applying for support were excessive bureaucracy, uncertainty about necessary home improvements, and process complexity. When asked about monthly repayments for home improvements, very few households were willing to pay a reasonable amount. This indicates to policymakers that subsidy schemes should not require upfront costs from low-income households, and the application process must be simplified, as well as the need to provide information and increase awareness about energy efficiency investments. Analyzing the results of the self-declared comfort and support mechanisms, it can be concluded that energy poverty is not an exclusive problem of low-income households, since other socio-economic groups are affected by low comfort levels and high energy bills. Considering Portugal’s social policies, the middle class faces greater difficulties paying bills than the very poor, who receive government support and incentives, due to the middle class’s declining purchasing power.

The reality of energy poverty was observed using different options of indicators. Using the 10% indicator, we determined that 12% of surveyed households spend over 10% of their net income on energy services, rising to 29% among lower-income households. The survey shows that 8.4% of households had arrears on energy bills in the last 12 months. Such a number is higher than the national average of 4.5% and the EU average of 6.2% [43]. Additionally, not being in arrears often masks real energy poverty, as many restrict energy use to avoid disconnection. In such a context, the LILEE indicator shows that 9.3% of respondents are low-income households with homes with very low energy efficiency, increasing to 17.2% when medium-efficiency homes are included. The LIHC indicator reveals that 15.4% of households face high risk due to low income and high energy costs, rising to 23.2% when medium-income households with high energy costs are considered. When comparing these numbers with other studies [44] and, mainly with the statistics, they clearly show a higher percentage of households with energy poverty risk, since the most recent statistics by Eurostat identify 17.5% of the households in Portugal and 9.3% in the EU as being in energy poverty risk [1].

As a global conclusion, at the national level, there is a need to work on developing and improving existing poverty indexes and indicators considering the several dimensions of poverty, targeted to the regions (most inner regions have stronger impacts), and then design tailor-made effective policies, close to meeting real needs. The pandemic, the rising inflation rate, and the high migration rates intensified inequalities and increased the number of families living with economic constraints. Furthermore, in a scenario of a huge housing crisis in the larger cities in Portugal, energy poverty is going far beyond the usual definition, bringing the complexity of poverties to bear: energy poverty, digital poverty, education poverty, food poverty, etc. The analysis shows evidence that energy poverty is also related to geographical location, since in winter, those living in the inner parts of Portugal must make a larger effort to pay their energy bills. Establishing regional indicators when policies are being designed for the country seems to be logical from a social perspective of equity and justice.

Based on the findings of the survey, capacity building and education play an important role, as do the local action taken and the support offered by municipalities, which know the real needs of their citizens well. Thus, the strategies advocated should focus on the following:

• Analyzing existing energy rehabilitation techniques and adapting them to the region, in a process of multidisciplinary collaboration and the co-creation of innovative and effective solutions in conjunction with local players and stakeholders under a logic of non-invasive intervention and life cycle analysis, complying with the legal and normative framework defined by the regulation of the energy performance of buildings.
• Incentivizing the installation of heat pumps with a high coefficient of performance, particularly in existing buildings, to boost energy efficiency and enhance thermal comfort, as their use in new constructions is already widespread but not yet common in retrofits.
• Accelerate the implementation and promotion of passive houses and other building standards, installing solar thermal systems for heating sanitary water, and promoting services provided by renewable energy communities as the municipalities are keen on renewable energy communities and are open to innovative schemes.
• “One-stop shops” that provide information, guidance, and rehabilitation services to vulnerable households. These are emerging in the market in association with European projects, but more need to be established in association with municipalities to enroll vulnerable households in financing programs to improve energy efficiency, health, and comfort conditions in homes.
• Actions aimed at promoting awareness through less formal activities involving the population and the exchange of knowledge, training, and coaching in order to promote the development of skills and combat energy illiteracy, as these are crucial for the success of this concerted action.