Quality of Life and Energy Efficiency in Europe—A Multi-Criteria Classification of Countries and Analysis of Regional Disproportions

Abstract

Energy efficiency (EE) is an important driver of quality of life (QoL), which is an overarching goal of sustainable development (SD). The levels of these phenomena in the European Union (EU) vary. Previous analyses presented in the literature have focused mainly on a one-dimensional view of EE and QoL. The authors of this article, given the multidimensional nature of the phenomena under study, present both categories from a holistic perspective. The purpose of this study was to identify the level of QoL in the context of EE and to compare the results of the classification of EU countries in terms of the analyzed phenomena. The study was conducted using the ELECTRE Tri method, one of the advanced techniques of multi-criteria decision analysis (MCDA). The classification procedure used made it possible to assign countries to predefined decision-making categories on the basis of preference threshold values and dominance relations to reference profiles. The 27 EU member states were analyzed on the basis of empirical data from 2023, using a set of 20 indicators characterizing EE and QoL. Countries were assigned to one of five classes, differentiating the level of development in both analyzed areas. Optimistic and pessimistic approaches were used to assess the stability of the classifications. The analysis showed the presence of countries with consistent results (e.g., Poland and Germany), extreme countries (Ireland and the Netherlands—high QoL with low EE; Romania and Croatia—inversely), as well as non-unique cases (e.g., Malta, the Czech Republic/Czechia, and Finland). The spatial approach indicated regions requiring special support. The results of the study can be a useful tool to support the process of designing public policies aimed at integrating social, economic, energy, and environmental goals within SD.

1. Introduction

QoL is considered an overarching goal of the activities of states, various organizations, and individuals. In assessing it, the focus should not only be on the well-being of the population; health, the environment, social equality, and the general well-being of citizens are also important. A holistic view of this category increases feelings of happiness and leads to sustainable societies. In order to achieve high QoL, appropriate social policies focusing on investing in education, health, and access to housing should be proactively implemented. An important factor in ensuring its high level is to maintain a work–life balance and ensure the SD of societies through, among other things, an economy based on innovation, ecology, and EE. In addition, society should be involved in processes, decisions, and various types of initiatives, especially at the local level. Thus, it can be concluded that QoL is closely linked to SD, and the pursuit of a high level of it must go hand in hand with environmental protection and reductions in social inequality. This paper pays special attention to the indicated issues by referring to EE and QoL.

The purpose of the research undertaken was to identify the level of QoL in the context of EE. The study refers to the EU countries; although the original intention was to include other European countries, this proved impossible due to significant data gaps, especially with regard to EE.

The specific objectives of the study include a thorough assessment and measurement of QoL and EE, as well as a comparison of the results of the classification of EU countries in terms of the phenomena analyzed. In preparing the study, particularly at the stage of literature review and analysis of variables describing the studied phenomena, the authors formulated the following research questions:

• Are the distinguished classes of EU countries unambiguous in terms of QoL and EE?
• Do EU countries belong to similar classes in terms of QoL and EE?

Achieving the goal of the research was possible through the use of the multi-criteria decision analysis (MCDA) method, ELECTRE Tri. Information from the Eurostat database (including the EU-SILC survey), TheGlobalEconomy.com, and Our World in Data for 2023 was used to achieve the goal.

This study is innovative in nature and fills a gap in the assessment of QoL in the context of EE, both in terms of the proposed treatment of the categories analyzed and the research method used. Many researchers evaluate QoL and EE but rarely combine these two dimensions in a single comparative analysis. Our study shows whether and how these two areas are related, and whether higher EE translates into better living conditions. Due to the inclusion of all EU countries, the survey is comprehensive in nature, thus filling the contextual gap of EE assessment in relation to QoL. Our study is based on the use of the ELECTRE Tri method with a pessimistic and optimistic approach, one of the advanced techniques of multi-criteria decision analysis, which will significantly contribute to filling the gap in the methodology used in the assessment of the categories covered by the study.

The study presents results describing QoL and EE in one-dimensional and multidimensional terms, which allowed for a thorough evaluation of the assessment categories in the EU space and distinguishing classes of countries similar in terms of the level of the phenomena studied.

The article is structured as follows: the introduction presents the authors’ motivations and intentions for undertaking the study, the objectives of this study, data sources, and research methods. The second section contains a literature review covering QoL and EE issues and the use of the multi-criteria decision analysis (MCDA)—ELECTRE Tri method. The third section is devoted to the data used in the study and the research methods. The fourth section comprehensively describes the results of the study. This article ends with a discussion and conclusions.

2. Literature Review

2.1. Energy Efficiency as an Aspect of Sustainability and Quality of Life

This paper uses an environmental approach to QoL, specifically SD, since improving QoL is its primary goal. The UN document Our Common Future [1] states that SD is one in which the needs of the present generation can be met without diminishing the chances of future generations to meet them.

What distinguishes QoL from other economic categories is the wide range of problems and phenomena that make up its nature. It is expressed not only in the volume of consumption of material goods, but also in the possibility of satisfying needs that are related, for example, to the state of the environment, dignity, and respect for human rights, as well as the quantity and quality of free time and opportunities for self-realization [2]. In economic science, the concept of QoL began to gain prominence in the 1960s and 1970s. At that time, economists pointed out that GDP was not a sufficient measure of social wellbeing. Among others, Richard Easterlin—author of the Easterlin Paradox—was a pioneer in this field, showing that an increase in income does not always translate into a greater sense of happiness and higher QoL [3]. Other researchers have also emphasized the inadequacy of GDP to assess this phenomenon. The 1998 Nobel Prize winner in economics, Amartya Sen, in his proposed capabilities approach, stressed that QoL depends not only on income, but on people’s real capabilities in terms of education, health, and freedom of choice. According to Sen, the basis for assessing this category would be the ability of an individual to use the goods he or she owns in order to live with dignity. Therefore, access to goods and services alone does not necessarily improve the QoL of residents [4].

Also, Joseph Stiglitz and Jean-Paul Fitoussi criticized GDP as the main indicator of do-ability and proposed new measures of QoL. Together with Sen, in 2009, they produced a report for the French government criticizing GDP as the main indicator of well-being and invited new measures of this phenomenon [5]. According to Joseph Stiglitz, QoL is a much broader concept than traditional economic measures, such as GDP, which does not show how prosperity is distributed among people. The assessment of this category should take into account health, education, physical and economic security, social and political participation, the environment, and work–life balance. The report by the Commission on the Measurement of Economic Performance and Social Progress [5], also known as the Stiglitz–Sen–Fitoussi report, emphasizes the need to take into account the subjective assessment of life, the level of happiness, and satisfaction with various aspects of life. The document states that social progress cannot be sustainable without environmental policy. The report recommended not linking environmental measures to GDP, but rather creating separate indicators to monitor air, water, and soil quality.

The main arguments against GDP as a measure of QoL point out that GDP measures market value rather than well-being, does not take into account income inequality and the value of unpaid work and leisure time, does not reflect psychological and social well-being, and does not measure environmental impact. Accordingly, Stiglitz, Fitoussi, and Sen proposed a multidimensional approach to this phenomenon that takes into account health, education, environment, and social inequality. This approach to QoL is reflected in indicators such as the Human Development Index (HDI)—which takes into account life expectancy, education, and income; the Better Life Index (OECD)—which measures well-being in 11 areas (including health, education, environment, and work–life balance); and the Happiness Index—which takes into account psychological and social well-being. A very important element of QoL pointed out by Stiglitz, Fitoussi, and Sen is the inclusion of SD and environmental degradation in the scope of this category. QoL is inextricably linked to sustainable development, which provides future generations with a higher level of it by protecting natural resources and mitigating climate change, which in turn translates into improved population health.

Despite many years of research on QoL, no single definition of it has been developed. This is due to its complex and interdisciplinary nature. The concept of QoL has evolved towards a holistic view, from an approach that considers GDP as the most important element to approaches that capture health, living conditions, education, happiness, and, in recent years, SD. The most universal definition can be considered the WHO definition, which recognizes its multidimensional nature, including physical, psychological, social, environmental, and spiritual aspects. The WHO defines QoL as an individual’s perception of their position in life in the context of the culture and value systems in which they live and in relation to their goals, expectations, standards, and concerns [6]. The literature [7] emphasizes the advantages of this definition, namely: holistic approach, subjectivity, cultural context, universality, dynamism, and comprehensiveness. Thanks to these features, the WHO definition is widely used in scientific research, medicine, psychology, and social policy. An overview of the definitions, classifications, and measures that characterize QoL can be found in the works of many researchers in various scientific fields, who emphasize aspects related to the type of their scientific activity in assessing it [7,8,9,10,11,12,13].

In recent years, especially at a time of far-reaching climate change, SD has become a key element of QoL. This approach stems from and is linked to the assumption that prosperity cannot be achieved at the expense of future generations. SD is based on three pillars that must be in balance with each other: environment, economy, and society. In the case of the environment, emphasis is placed on protecting natural resources and ecosystems, through, for example, reducing CO₂ emissions, conserving water and energy, and protecting forests and biodiversity. In the area of the economy, stable growth without destroying resources is emphasized, such as the use of renewable energy sources, the application of the circular economy (recycling and reuse), and investigations into eco-innovation. In contrast, in the area of society, SD is understood as the well-being of the population and equality of opportunity, i.e., education for all, decent living and working conditions, fair wages, and access to health care.

Energy is inextricably linked to many aspects of human life, and energy supply shortages have always been an obstacle to human and economic development. Without it, it will not be possible to reduce poverty, end hunger, increase education, improve health, increase water supply, industrialize, or combat climate change [14]. The challenges of energy policy in Europe are becoming increasingly important, especially in the context of rising energy costs, limited availability of certain fuels, and climate change. On top of this, the EU, as the world’s largest energy importer, is struggling with the problem of energy dependence. As a result, it is becoming crucial to become less dependent on external markets by diversifying energy supply sources, but above all by raising EE, using alternative energy sources, and developing new technologies [15].

EE is an important element of SD. In the agenda for SD 2030 [16], the energy goal appears as the seventh: “Ensure access for all to sources of stable, sustainable and modern energy at affordable prices”. Tasks related to this goal include doubling the growth rate of global EE, significantly increasing the share of renewable energy sources in the global energy mix, investing in energy infrastructure, and clean energy technologies.

In EU legal acts, EE is defined as the ratio of outputs, services, goods, or energy obtained to energy input [17]. In the Polish legal system, it is presented as the ratio of the achieved magnitude of the utility effect (e.g., performing mechanical work and providing thermal comfort or lighting) of a given object, technical device, or installation, under typical conditions of their use or operation, to the amount of energy consumed by that object, technical device, or installation, or as a result of the performed service necessary to achieve that effect [18]. Thus, it should be considered that EE is a set of all activities, tools, and innovations that make it take less energy to produce a given product.

The European Green Deal Communiqué [19] states that EE must become a priority and identifies it as one of the key solutions in all sectors to achieve climate neutrality at the lowest possible cost [20]. The Energy Union Governance and Climate Action Regulation formulated the principle of “energy efficiency first” [21]. It mandates treating EE as a “first fuel”, a self-sustaining source of energy in which the public and private sectors can invest before investing in other, more complex or costly energy sources. It involves a shift from the traditional model of energy generation and consumption, based on large suppliers whose offerings are dominated by fossil fuels and passive consumers who accept imposed prices, toward a more flexible system that incorporates renewable technologies and focuses on actively engaged energy consumers. It is therefore important to recognize EE as a key element and an essential factor to be considered in future investment decisions on energy infrastructure in the EU [20], which requires the dissemination of this principle to government institutions at the regional and local levels, as well as to the private sector.

The main objective of the “energy efficiency first” principle is to consider EE and energy demand management measures on an equal footing with alternative measures to meet a specific need or objective, especially when it comes to investments in energy supply or energy infrastructure. As a consequence, this principle is expected to lead to the identification and implementation of cost-effective and energy-efficient solutions, while achieving the intended goals to reduce energy consumption and reduce energy losses as a result of technological, economic, and/or behavioral changes, i.e., concerning the behavior of all energy users [22].

An important role in this area is to raise the awareness of all EU citizens about the benefits of increased EE and to provide them with reliable information on methods to achieve it. It is important to access knowledge and develop an advisory system for residents, especially those living in low-income cities [23].

Applying the principle of “energy efficiency first” is important in combating energy poverty. Improving EE can reduce energy bills and have significant beneficial effects on the health and comfort of low-income households [20]. EU policies should therefore be inclusive and ensure the availability of EE measures for energy-poor consumers. Reducing consumer spending on energy should be achieved by helping consumers reduce their energy consumption by reducing energy demand in buildings and increasing the efficiency of appliances, which should be combined with the availability of low-energy demand transportation modes integrated with public transport and bicycling [24].

Another important aspect of EE is related to minimizing the negative environmental impact of energy extraction, storage, distribution, and use, promoting low-carbon technologies, and thus increasing the use of RES [25]. The biggest environmental threat from the exploitation of non-renewable energy resources is the emission of greenhouse gases, including carbon dioxide. Reducing these emissions brings benefits in the form of better air quality and job creation in RES-based sectors, among others. Protecting the environment in the operating conditions of a competitive energy market is therefore another major challenge for its participants. In this context, the exploitation of renewable raw materials is increasingly becoming an element of countries’ energy security strategies [26,27], with the process first increasing the pool of available solutions while maintaining the technological readiness of the entire sector [28].

EE is also an important pillar for ensuring the energy security of the state, understood as guaranteeing the supply of energy resources that provide the basic needs of the state even in a situation of crisis or international conflict [29]. It is also included as a state of the economy that allows for meeting current and prospective consumer demand for fuels and energy, in a technically and economically reasonable manner, while minimizing the negative impact of the energy sector on the environment and living conditions of society [30]. Stachowiak [31] also draws attention to socially acceptable prices and the preservation of political independence, while Klare [29] points to the diversification of energy sources and investment in environmentally friendly, renewable resources such as solar, biomass, and wind power plants.

Energy security is often analyzed as a component of a country’s economic security, which is a state in which a country’s economy has the supply of factors of production necessary for its functioning and development, including energy sufficiency [32]. The economic dimension of energy security concerns mainly the cost of obtaining energy and the continuity of supply. Energy is a specific product because it must be available continuously, including in situations of political or economic crises. Lack of such liquidity entails high costs for the country’s entire economy. Thus, the energy sector plays a fundamental role in shaping the efficiency and competitiveness of the economy, and directly and indirectly affects the QoL of citizens. Thus, energy resources are treated as a strategic product. Due to increased competition in the international raw materials market, the importance of the price of energy, which determines consumers’ living standards, industrial competitiveness, and economic growth, is increasing [33,34].

Energy security is of great importance for countries that are dependent on imports of energy carriers. Their position in the energy market is very weak, and they are susceptible to external pressure, not only economically but also politically. Dependence on a single producer threatens to undermine the liquidity of imports [35]. For countries that are importers of energy resources, energy security is a key element of their foreign policy and includes measures to reduce dependence on a single importer, i.e., diversification of raw material sources [34].

Improving EE throughout the energy chain, including during generation, transmission, distribution, and end use of energy, is a desirable phenomenon in the social, economic, and environmental aspects of a country’s operation. It is good for the environment, will result in improved air quality and public health, reduce greenhouse gas emissions, and improve energy security. Reducing dependence on energy imports, lowering energy costs for households and businesses, and helping to alleviate energy poverty will lead to greater competitiveness, increased employment, and a boost to the overall economy, which will raise the QoL for citizens. Improving this phenomenon can contribute to improved economic performance. Member states and the EU should strive to reduce energy consumption regardless of the level of economic growth [24]. The issue of EE is treated as a priority because progress in this area generates significant economic benefits and is important for the realization of all energy policy goals and most environmental and climate policy objectives. Reducing the energy intensity of the national economy increases the security of the energy supply. Stimulating investment in modern, energy-efficient technologies and products contributes to the growth of the economy’s innovativeness. Any effective energy-saving measures, therefore, have a significant impact on improving the efficiency of the national economy and increasing its competitiveness [36] and should thus be a priority in modernizing the country’s economy. Promoting EE reduces the need for additional investment in supply and storage infrastructure while reducing the rate at which existing infrastructure is operated, which reduces operating and maintenance costs [37].

2.2. The ELECTRE Tri Method in EE and QoL Assessment

Today’s challenges of energy transition and SD require the use of advanced methods, enabling a comprehensive assessment of the efficiency of energy systems and the impact of energy policies on society and the economy. One of the approaches used in such studies is the ELECTRE Tri method, which belongs to the ELECTRE family of techniques (Elimination et Choix Traduisant la Réalité) and utilizes the outranking relation. It makes it possible to classify objects into predefined categories, taking into account both measurable indicators and subjective preferences of decision-makers [38]. Its wide application is due to its ability to process multidimensional data and take into account complex decision-making processes.

The literature on EE and QoL is increasingly using MCDA methods, including ELECTRE Tri, which allow simultaneous consideration of social, economic, and environmental factors [39]. Traditional indicators, namely the Human Development Index (HDI) and Energy Efficiency Index (EEI), do not fully reflect the relationship between energy policy and socioeconomic processes. Instead of creating a unified ranking, the ELECTRE Tri method enables the classification of facilities according to established criteria, which allows detailed analysis of decision-making problems [40]. This makes it an effective tool to support decision-making under conditions of complex multidimensional relationships.

The use of ELECTRE Tri in EE analysis has been widely discussed in the scientific literature. One can highlight the work of Kartsonakis et al. [41], who developed the ELECTRE Tri model for assessing energy sustainability, ranking countries according to the Energy Trilemma Index and identifying significant differences in EE levels in Europe, the US, Japan, and Australia. For the EU, Cabeça et al. [42] used an expanded version of the method—the ELECTRE Tri-nC—to rank the EE management capabilities of EU member states, pointing to the need for tighter EU regulations for more effective energy policy implementation. In contrast, Baseer et al. [43] developed a probabilistic version of the ELECTRE Tri method (pELECTRE Tri), using Monte Carlo simulation to model uncertainty in residential retrofit decisions. Dell’Anna [44], in contrast, developed the ELECTRE Tri-B model, which was used to optimize building energy retrofits at the neighborhood level, supporting public policy management and financial resource allocation, which can reduce resource consumption and improve EE.

One of the most important areas of research is the impact of energy policy on CO₂ emissions and energy costs. Martins et al. [45] used the ELECTRE Tri-nC method to classify EU member states in terms of the effectiveness of policies supporting the development of electric vehicles, pointing to the importance of financial incentives and charging infrastructure in transforming the transportation sector. In contrast, Neves et al. [46] used the ELECTRE Tri to prioritize EE initiatives, highlighting its advantage over cost–benefit analysis by considering the full spectrum of impacts of actions. Antolínez [47], in contrast, used ELECTRE Tri-nc to evaluate local energy strategies related to photovoltaic energy, demonstrating that decisions made at the municipal level can have a significant impact on reducing CO₂ emissions and improving the efficiency of renewable energy sources (in the Spanish municipality of Rajadell). It is worth mentioning that Jovanović et al. [48] have developed an integrated approach to energy systems management, using MCDA, resource optimization, and energy demand management strategies. Bohra and Anvari-Moghaddam [49], in contrast, comprehensively reviewed the applications of MCDA methods in the power sector, taking into account their role in energy policy, the localization of renewable energy sources, and the optimization of grid load management. Regional differences in EE represent another area of research using ELECTRE Tri. It should be pointed out that Martins et al. [50] reviewed the energy modeling tools used in Europe’s smart cities, identifying those that best support EE and decarbonization strategies. At the same time, Hajduk and Jelonek [51] applied another (popular) MCDA method, TOPSIS, to classify smart cities in terms of EE, using ISO 37120 indicators and World Council on City Data, highlighting the importance of multi-criteria analysis in integrated city energy planning.

New approaches in ELECTRE Tri research focus on its integration with other MCDA methods in EE evaluation. It is worth noting that Rane et al. [52] conducted a comparative analysis of MCDA methods in SD topics, pointing out the advantages of ELECTRE and the potential benefits of using hybrid approaches combining it with AHP and TOPSIS. In contrast, Jong and Ahmed [53] conducted a systematic literature review of MCDA methods used in solar site selection, analyzing the relevant criteria and differences in rankings obtained by these methods, including ELECTRE, TOPSIS, AHP, and PROMETHEE. A similar review of MCDA methods was conducted by Azhar et al. [54], classifying them according to approaches based on superiority relations, pairwise comparisons, and distance methods, citing their wide applicability in various fields, including the energy sector. In turn, Mexis et al. [55] developed an innovative web-based tool based on MCDA methods to evaluate and finance EE investments through standardized benchmarks and compliance with the EU SD taxonomy.

Contemporary QoL and social policy research is also increasingly using MCDA methods. One example is ELECTRE Tri, used to assess public services, housing conditions, and the availability of social resources. In their study, Karakas et al. [56] applied MCDA to optimize EE and indoor environmental quality (IEQ) measures in English schools, emphasizing the co-efficiency of balancing energy savings with thermal comfort and student health. One of the most important application areas for ELECTRE Tri is the health sector. Therefore, Rocha et al. [57] assessed the quality of Portuguese public hospitals, using the ELECTRE Tri-nC model to assign facilities to specific categories, taking into account multiple performance criteria. Gregório et al. [58] extended this approach, using a modified version of the ELECTRE Tri-nC to analyze the relationship between healthcare accessibility and service quality. This made it possible to classify hospitals, with the results showing that the number of medical staff has no significant effect on the performance categories assigned to hospitals. The literature notes examples of other applications of the ELECTRE Tri method in various research areas. It is worth noting that Telles et al. [59] used it to evaluate the efficiency of eight agricultural cooperatives in the southwestern state of Paraná, Brazil. This analysis demonstrated the usefulness of the method in classifying entities in terms of their economic and social impact and also made it possible to identify areas in need of restructuring. It is important to mention that ELECTRE Tri has been used to assess SD. For example, Dias [60] showed that the method makes it possible to classify various entities—from companies to regions—based on multi-criteria SD indicators, while the low value of one criterion cannot be fully compensated for by the high values of others. In contrast, Madhooshiarzanagh and Abi-Zeid [61] developed an approach that enables the estimation of criteria weights in ELECTRE Tri-nC using incomplete decision-maker information, which expands its application in decision support, such as climate classification for tourism.

ELECTRE Tri is also used in analyses of housing policy and real estate revitalization. Barbaro et al. [62] developed a decision-making model based on this method to assess the feasibility of adaptive reuse of abandoned properties in depopulated regions of Sicily as part of a scattered-site social housing concept. In turn, Thebault et al. [63] used ELECTRE Tri to assess the potential of buildings for photovoltaic integration in greater Geneva, enabling the identification of the most promising sites for SD and supporting energy planning at the regional level. The importance of the ELECTRE family of methods in social structure research is also underscored by analyses of regional socioeconomic inequality. It should be mentioned that López-Parra et al. [64] used the ELECTRE III method to assess the marginalization of Mexico’s regions, analyzing the level of socioeconomic exclusion on the basis of educational indicators, housing conditions, and income. Górecka and Roszkowska [65] used modified MCDA methods (TOPSIS and BIPOLAR) to analyze the spatial differentiation of EU countries in terms of achieving SDG 11 (Sustainable Cities and Communities), showing clear disparities between northern and southern countries.

Contemporary research on EE and QoL is increasingly using MCDA methods, including ELECTRE Tri, but their application to the comprehensive analysis of these areas needs further exploration. The effectiveness of the ELECTRE Tri method in evaluating energy strategies is well documented, but its application to analyzing the impact of energy decisions on QoL requires further research. It seems that the integration of ELECTRE Tri with other MCDA techniques and hybrid models can enhance the comprehensive analysis of the socioeconomic consequences of energy policies.

3. Materials and Methods

The research conducted referred to the assessment of QoL and EE in the 27 EU countries in 2023. Information on indicators characterizing both categories was obtained mainly from the Eurostat database, as well as from Global Economy, World Economy, and Our World in Data. The selection of indicators is substantively justified and follows established standards set by Eurostat. This approach facilitates a meaningful comparison of QoL and EE across EU countries.

Given the complex nature of a category such as QoL, many areas of the population’s life, both objective and subjective, should be taken into account in its description and measurement. This is a category that is difficult to quantify due to its multispectral nature. It is composed of factors that are often difficult to measure and are subjective in nature. In this study, economic, social, and environmental indicators were included among the variables relating to QoL. The inclusion of indicators in the study was primarily related to the availability of information for EU countries. Among the variables of an economic nature, GDP per capita was taken into account, which reflects the average level of income in a country and provides an opportunity to compare objective QoL across EU countries. As mentioned in the literature review, GDP has disadvantages, one of which is that it does not take income inequality into account. To fill this gap, the study took into account the Gini coefficient, which measures income inequality. Among the indicators that inform about the level of health care, the infant mortality rate per 1000 live births, reflecting the level of health care, education, health awareness, and living conditions, plays an important role.

One of the basic human needs is access to clean water. Lack of access to safe water sources is a major risk factor for infectious diseases, and it also contributes to the risk of malnutrition and causes a high risk of death [66]. Clean water and sanitation are SDG 6 and call for ensuring universal access to safe and affordable drinking water, sanitation, and hygiene [67]. In the EU context, monitoring SDG 6 focuses on sanitation, water quality, and water scarcity. Due to the great importance of access to clean water, which is an indicator informing about the sanitation infrastructure, it was included among the variables describing QoL.

Various aggregate measures are used to assess the phenomenon under analysis: HDI, Better Life Index (BLI), or European Quality of Life Index (EQLI), which capture many aspects of a population’s life. The literature review indicated that QoL is often equated with happiness, so this study also included the Happiness Index [68], which measures the level of happiness and well-being of citizens in different countries. This index has an aggregate character and captures, among other things, GDP per capita, healthy life expectancy, social support, freedom to make life choices, generosity, and perceptions of corruption. A list of indicators related to QoL included in this study is presented in

Table 1. Indicators characterizing QoL.

Variable	Variable Name	Unit of Measure	Characteristics
Y1	GDP per capita [sdg_08_10]	USD	Gross domestic product (GDP) is a measure of economic activity. It refers to the value of the total output of goods and services produced by an economy, less intermediate consumption, plus net taxes on products and imports. GDP per capita is calculated as the ratio of GDP to the average population in a specific year.
Y2	Gini income inequality index	from 0 to 100	The Gini coefficient is defined as the relationship between the cumulative shares of the population arranged according to the level of equivalized disposable income and the cumulative share of the equivalized total disposable income received by them.
Y3	Unemployment rate [med_ps421]	%	The unemployment rate is the number of people unemployed as a percentage of the labor force. The labor force is the total number of people employed and unemployed.
Y4	Life expectancy [demo_mlexpec]	years	Life expectancy at a certain age is the mean additional number of years that a person of that age can expect to live, if subjected throughout the rest of his or her life to current mortality conditions.
Y5	Expected years of schooling	years	Expected years of schooling is the number of years a child of school entrance age is expected to spend at school, or university, including years spent on repetition.
Y6	Infant mortality rate [tps00027]	per 1000 live births	The infant mortality rate is defined as the ratio of the number of deaths of children under one year of age to the number of live births in the reference year.
Y7	Share of people having access to safe drinking water	%	Safe drinking water is defined as water from an improved water source, which includes household connections, public standpipes, boreholes, protected dug wells, protected springs, and rainwater collections [69].
Y8	Air Quality Index according to PM2.5	µg/m3	The Air Quality Index (AQI) for PM2.5 is a standardized measure used to indicate the level of fine particulate matter (PM2.5) in the air and its potential health effects. PM2.5 refers to airborne particles smaller than 2.5 μm, which can penetrate deep into the lungs and even enter the bloodstream.
Y9	Length of state, provincial, and municipal roads	km per 1 km2 of the country’s area	The length of state, provincial, and municipal roads per 1 km2 of a country’s area is a measure of road density, often expressed in kilometers of road per square kilometer of land area.
Y10	Population living in households, considering that they suffer from noise [sdg_11_20]	%	The indicator measures the proportion of the population who declare that they are affected either by noise from neighbors or from the street. Because the assessment of noise pollution is subjective, it should be noted that the indicator accounts for both the levels of noise pollution as well as people’s standards of what level they consider to be acceptable.
Y11	Happiness Index	0 (unhappy)–10 (happy)	The Happiness Index is a comprehensive survey instrument that assesses happiness, well-being, and aspects of sustainability and resilience. Zero points means unhappy, ten means happy [68].

. For indicators obtained from the Eurostat database, specific Eurostat database codes are given next to their name.

The selection of indicators used in the study was based on three primary criteria: consistency with the academic literature, availability of comparable data for all EU member states, and the relevance of these variables to public policy in the areas of QoL and EE. It should be noted that the ELECTRE Tri method does not require a formal variable selection procedure; the set of criteria is determined substantively, in direct relation to the study’s objectives and the nature of the decision problem under consideration. Given the exploratory nature of the research and the limited availability of regionally disaggregated data, methods such as the Delphi technique or principal component analysis (PCA) were not applied, although their potential usefulness has been identified as a possible direction for future, more in-depth analyses. In addition, no formal normalization procedure was applied, as all variables were used in indicator form, ensuring direct cross-country comparability and preserving interpretability in relation to real-world values.

The EE analysis included variables relating to energy production capacity, energy consumption, and energy demand. Also used to assess the phenomenon were data on the share of renewable energy in gross final energy consumption, the share of fossil fuels in primary energy consumption, and CO₂ emissions, which inform countries’ efforts to protect the environment and transition to zero-carbon economies. The study also included a variable related to energy security (energy import dependency) and energy poverty (population unable to keep their homes adequately warm). Eight of the ten variables describing EE were cost variables. Profitable variables included energy productivity and the share of renewable energy in gross final energy consumption. A list of variables describing EE, including database codes obtained from the Eurostat database, is presented in

Table 2. Indicators characterising EE.

Variable	Variable Name	Unit of Measure	Characteristics
X1	Final energy consumption per capita [sdg_07_11]	TOE per capita	This indicator measures a country’s energy end use, excluding all non-energy use of energy carriers (e.g., natural gas used not for combustion but for producing chemicals). Final energy consumption only covers the energy consumed by end users, such as industry, transport, households, services, and agriculture; it excludes energy consumption of the energy sector itself and losses occurring during the transformation and distribution of energy.
X2	Energy intensity [nrg_ind_ei]	kgOE/PPS	Energy intensity is one of the indicators that measure an economy’s demand for energy. It measures the amount of energy consumed per unit of Gross Domestic Product (GDP) adjusted for differences in price levels between countries (using purchasing power standards, or PPS). It is a measure of an economy’s EE that shows how much energy (usually in physical units) is needed to produce a unit of value in the economy, converted to purchasing power standards.
X3	Energy productivity [sdg_07_30]	PPS/kgOE	This indicator measures the amount of economic output that is produced per unit of gross available energy. Gross available energy represents the amount of energy products required to meet all the demands of entities in the geographic area under consideration. It shows how much value the economy can produce (in PPS) per unit of energy consumed (in KGOE). It is an indicator of EE, which shows how effectively a country uses its energy resources to generate economic value.
X4	Share of renewable energy in gross final energy consumption [sdg_07_40]	%	This indicator measures the share of renewable energy consumption in gross final energy consumption according to the Renewable Energy Directive. The gross final energy consumption is the energy used by end consumers plus grid losses and the self-consumption of power plants.
X5	CO2 emissions	t/person	This indicator illustrates how many tons of carbon dioxide are emitted per capita in a country in a year. Fossil fuel emissions measure the amount of carbon dioxide (CO2) emitted from the combustion of fossil fuels and directly from industrial processes, such as cement and steel production. Fossil CO2 includes emissions from coal, oil, gas, combustion, cement, steel, and other industrial processes. Fossil fuel emissions do not include land use change, deforestation, soils, or vegetation.
X6	Electricity demand	kWh/person	Annual average electricity demand per person, measured in kilowatt hours. This is electricity generation, adjusted for imports and exports.
X7	Electricity generation	kWh/person	Annual average electricity generation per person.
X8	Share of primary energy consumption from fossil fuels	%	Share of fossil fuels (coal, oil, and natural gas) measured as a percentage of primary energy.
X9	Energy import dependency [sdg_07_50]	%	The indicator shows the share of total energy needs of a country met by imports from other countries. It is calculated as net imports divided by the gross available energy.
X10	Population unable to keep their homes adequately warm [sdg_07_60]	%	This indicator measures the share of the population who are unable to keep their homes adequately warm. Data for this indicator are being collected as part of the European Union Statistics on Income and Living Conditions (EU-SILC) to monitor the development of poverty and social inclusion in the EU. It illustrates what proportion of the population has difficulty heating their homes due to financial problems. It is an important measure of energy poverty, which is caused by low income, high energy prices, and poor-quality housing.

.

In recent decades, a significant influence on the development of decision support methods has come from the so-called European, but more specifically French, school of operations research, initiated by Bernard Roy. This school emerged as a critique of overly simplified optimization models and aimed to create tools that reflect the complexity of real-world decision-making processes. Within this movement, a new philosophy within Multi-criteria decision analysis (MCDA) was formed, which moved away from purely compensatory models and focused on the outranking relationship [38,70]. It is worth noting that compensatory models assume that a poor evaluation in one criterion can be fully compensated by a high evaluation in another, which often proves insufficient under real decision-making conditions. In contrast, the foundation of the ELECTRE methods became the outranking relationship, which allows for modeling ambiguous, incomplete, or non-symmetric situations in which traditional comparisons of alternatives (objects, options) lose their effectiveness. This approach led to the development of a family of methods known as ELECTRE (ÉLimination Et Choix Traduisant la REalité—from the French, “elimination and choice translating reality”), which have found wide application in decision-making problems concerning selection, ranking of decision alternatives, and classification [71,72]. These methods are used in cases where data are incomplete, evaluations are imprecise, and the decision maker’s preferences cannot be clearly defined. In recent years, the development of the ELECTRE methods, including ELECTRE Tri and its variants (such as Tri-C, Tri-nC), has also focused on increasing the transparency of classification, which means enabling a better understanding and justification of the assignment of objects to classes [73,74]. Further research indicates the growing importance of these tools in the following areas: SD, risk management, and public policy analysis [41,44].

The ELECTRE family of methods employs the outranking relation, which is a binary relation whereby alternative $a$ outranks alternative $b$ if—based on the available information regarding the decision maker’s preferences—it can be considered that $a$ is at least as good as $b$, while at the same time there are no strong arguments that would contradict this assessment [38,75,76,77].

The ELECTRE Tri method belongs to the group of decision support methods dedicated to classification problems, in which the objective is to assign the evaluated objects to one of the ordered decision categories. A characteristic feature of this method is the use of so-called separating profiles, which are threshold vectors defining the boundaries between successive classes. Each alternative is evaluated against a set of such profiles, which makes it possible to determine its position within the class structure [75,78].

The fundamental input data in the ELECTRE Tri method include the evaluation matrix of alternatives with respect to a set of criteria, the criteria weights reflecting their importance, and three types of preference thresholds. The introduction of preference thresholds allows the model to account for the imprecision of evaluations and flexibility in perceiving differences between alternatives. Each threshold serves a different function, reflecting various aspects of the decision maker’s sensitivity to differences in evaluations. These thresholds are defined independently for each criterion, which enables a precise representation of the specificity and significance of individual decision-making aspects. Thus, we have the following:

(1)

Indifference threshold q_j—the maximum acceptable difference between the evaluation values of the alternative a and profile b_h, within which the decision maker considers both as equivalent from the perspective of a given criterion g_j; according to the formulation, if

$$\left|g_j\left(a\right)-g_j\left(b_h\right)\right|\le q_j$$

(1)

where

$g_j\left(a\right)$—the evaluation value of alternative $a$ with respect to criterion $j$,

$a$—the decision alternative,

$g_j\left(b_h\right)$—the evaluation value of the separating profile $b_h$ with respect to the same criterion $j$,

$b_h$—the boundary (separating) profile between adjacent decision categories,

$q_j$—the indifference threshold for criterion $j$,

then the difference in evaluations is considered insufficient to express a clear preference, the alternative and the profile are deemed indistinguishable from the perspective of that criterion.

(2)

Preference threshold p_j defines the minimum difference in evaluations above which the alternative a is clearly preferred over profile b_h with respect to criterion g_j, when

$$g_j\left(a\right)\ge g_j\left(b_h\right)+p_j$$

(2)

where

$g_j\left(a\right)$—the evaluation value of alternative $a$ with respect to criterion $j$,

$a$—the decision alternative,

$g_j\left(b_h\right)$—the evaluation value of the separating profile $b_h$ with respect to the same criterion $j$,

$b_h$—the boundary (separating) profile between adjacent decision categories,

$p_j$—the preference threshold for criterion $j$,

then it is assumed that alternative $a$ clearly outranks profile $b_h$ with respect to criterion $g_j$.

(3)

Veto threshold v_j is used to model situations in which even strong support from the majority of the criteria is not sufficient if there is a strong opposition on one of the criteria, when

$$g_j\left(b_h\right)-g_j\left(a\right)>v_j$$

(3)

where

$g_j\left(a\right)$—the evaluation value of alternative $a$ with respect to criterion $j$,

$a$—the decision alternative,

$g_j\left(b_h\right)$—the evaluation value of the separating profile $b_h$ with respect to the same criterion $j$,

$b_h$—the boundary (separating) profile between adjacent decision categories,

$v_j$—the veto threshold for criterion $j$,

this means that criterion $g_j$ prevents the establishment of the outranking relation, despite support from the remaining criteria, and the outranking relation $aS{b}_h$ does not hold.

The discussed thresholds enable the modeling of uncertainty and the decision maker’s tolerance toward differences in evaluations. The use of these thresholds makes it possible to reflect the actual way in which the decision maker perceives differences between alternatives, for example, by allowing for tolerance of minor differences or excluding solutions that are unacceptable for specific reasons (e.g., ethical or environmental) [76,79,80].

Another important element in the ELECTRE Tri procedure is the determination of the number of decision categories and the definition of the so-called separating profiles, which represent the boundaries between adjacent classes. Each decision alternative is compared individually with these profiles in order to determine its assignment to a given class. This process is based on conducting a concordance test and a discordance test, which allows for the assessment of the extent to which a given alternative outranks the separating profile [75,81].

Given a finite set of decision alternatives $A=\left\{a_1,a_2,\dots ,a_n\right\}$, a coherent family of criteria $G=\left\{g_1,g_2,\dots ,g_m\right\}$, and an ordered set of separating profiles $\left\{b_1,b_2,\dots ,b_{p-1}\right\}$, verifying the assertion that a given alternative $a$ outranks profile $b_h$ (or inversely) requires the fulfillment of the following two conditions:

(1)

Concordance for the outranking relation—the concordance test involves examining the strength of the so-called concordant coalition, meaning the set of criteria that support the assertion that a ≥ b_h; the majority of the criteria, taking into account their assigned weights, should support this claim.

(2)

Absence of strong discordance—the discordance test verifies whether there exists any criterion for which the advantage of profile b_h over alternative a is so significant (exceeding the veto threshold) that it can block the outranking relation, regardless of the support from the remaining criteria.

Formally, for each pair $(a,b_h)$, the credibility degree of outranking $\sigma (a,b_h)$ is calculated as the weighted sum of partial concordance indices. The relation $aS{b}_h$ is considered credible if

$$\sigma (a,b_h)\ge \lambda$$

(4)

where

$\sigma (a,b_h)$—the credibility degree of the outranking of $a$ with respect to profile $b_h$ expressed as the value of the overall concordance index,

$\lambda$—the cutting level, meaning the minimum value of $\sigma$, from which the outranking relation is considered credible; it is set by the decision maker and usually falls within the range [0.5;1]; an outranking degree of 1 means that all criteria fully support this assertion [74,75].

As a result of the comparisons, the following four possible relations between an alternative and a profile may occur:

$a>b_h$—the alternative is preferred over the profile;

$b_h>a$—the profile is preferred over the alternative;

$a~b_h$—the alternative and the profile are indistinguishable;

no relation—the alternative and the profile are incomparable [82].

Based on these relations, one of two classification procedures is applied:

(1)

Optimistic procedure (ascending)—begins with the comparison of the alternative with the lowest profile; the alternative is assigned to the first class C_h+1, for which the relation aSb_h holds, but the profile does not outrank the alternative;

(2)

Pessimistic procedure (descending)—the analysis starts with the highest profile; the alternative is assigned to class C_h if the first encountered profile satisfies the condition [83].

Both procedures can be applied in parallel, especially in sensitivity analyses, data uncertainty assessments, or multi-stage classification [78]. The use of concordance and discordance tests allows for the inclusion of both positive and negative preference information, making ELECTRE Tri a flexible tool that is robust to data imprecision and the subjectivity of the decision maker’s evaluations. The entire decision-making process—from the selection of input data, through the calculation of concordance and discordance indices, to the assignment of objects to classes—is illustrated in Figure 1 [75].

In recent years, the development of the ELECTRE Tri method has also included advanced extensions and adaptations that take into account, among other things, the modeling of multiple separating profiles (for example, the Tri-nB model), the use of artificial intelligence methods for automatic parameter estimation, and an in-depth analysis of the theoretical foundations of sorting models within the MCDA framework. For example, Bouyssou, Marchant, and Pirlot [84] conducted a formal analysis of the ELECTRE Tri-nB model, highlighting its structural properties and connections with other classification methods. In turn, Barros and Pereira [85] proposed an ELECTRE Tri algorithm that uses machine-learning techniques for the automatic calibration of parameters in the ELECTRE Tri-B model. These directions reflect the growing importance and potential of the ELECTRE Tri method in interdisciplinary applications.

The application of the ELECTRE Tri method was preceded by an evaluation of its suitability for analyzing complex, multidimensional phenomena that require classification rather than conventional ranking. The method was selected due to its non-compensatory evaluation mechanism, which prevents low scores in one criterion from being fully offset by high scores in others. This feature is particularly relevant when examining issues such as QoL and EE, where the considered dimensions—social, environmental, and economic—should not be treated as mutually substitutable. ELECTRE Tri also incorporates preference, indistinguishability, and veto thresholds, enabling the modeling of uncertainty, limited data precision, and nuanced relationships among the evaluated alternatives. Unlike ranking-based approaches such as TOPSIS or AHP, this method assigns alternatives to predefined, ordered categories without assuming full compensability or requiring a rigid weighting structure. Such a design reduces the risk of interpretive oversimplification caused by excessive aggregation and supports the generation of analytically robust and policy-relevant results. The model is better aligned with the nature of the socio-economic issues under study and the classificatory aims of the research.

4. Results

4.1. Analysis of the Structure of EU Countries in Terms of QoL and EE Indicators

QoL and EE in EU countries were characterized using basic parameters of descriptive statistics for indicators, presented in Table 1 and Table 2.

Table 3. Basic descriptive characteristics of indicators describing QoL and EE.

Variable	Mean	Median	Min	Max	Standard Deviation	Coefficient of Variation	Skewness
Y1	42,720.86	33,509.01	15,885.54	128,678.19	25,600.29	59.92	1.89
Y2	29.49	29.60	21.60	37.20	3.63	12.31	−0.10
Y3	5.75	5.59	2.59	12.14	2.21	38.46	1.10
Y4	80.75	81.70	75.80	84.00	2.59	3.21	−0.64
Y5	16.78	16.40	13.90	20.00	1.65	9.85	0.18
Y6	3.28	3.10	2.00	5.70	1.01	30.86	1.03
Y7	96.89	98.90	82.10	100.00	4.49	4.63	−2.09
Y8	10.59	9.90	4.70	17.40	3.46	32.66	0.03
Y9	1.54	1.28	0.12	8.99	1.71	110.89	3.27
Y10	16.94	15.50	6.70	31.30	7.35	43.42	0.50
Y11	6.59	6.49	5.46	7.74	0.53	8.10	0.24
X1	2.16	1.91	1.22	5.27	0.82	37.77	2.48
X2	86.25	84.13	34.19	150.12	24.25	28.11	0.62
X3	12.65	11.89	6.66	29.25	4.28	33.81	2.19
X4	27.22	22.55	14.36	66.39	12.77	46.92	1.40
X5	5.55	5.30	3.40	10.50	1.65	30.04	0.99
X6	6.44	5.82	2.67	14.74	2.68	41.56	1.78
X7	5.86	5.01	1.74	15.68	3.11	52.98	1.88
X8	69.11	71.81	25.90	88.25	14.88	21.54	−1.15
X9	57.18	61.05	3.47	97.55	22.18	38.79	−0.30
X10	9.47	7.10	2.10	20.80	5.98	63.15	0.96

shows the values of the basic parameters of the distribution of indicators characterizing QoL and EE.

EU countries are differentiated to varying degrees in terms of QoL variables. The strongest country differentiation was noted for the variables Y₉, Y₁, Y₁₀, and Y₃. Several indicators report weak variation among EU countries; these indicators include Y₄, Y₇, Y₁₁, and Y₅, as evidenced by variation coefficient values lower than 10%.

The distributions of indicators describing QoL are characterized by different directions and different strengths of asymmetry. The strongest asymmetry in the positive direction was noted for indicators numbered 9 and 1, while indicators 7 and 4 are characterized by at least moderate asymmetry in the negative direction. The positive direction of asymmetry means that countries with indicator values lower than the EU average dominate, which was noted in as many as 17 EU countries. The lowest GDP per capita is found in Bulgaria and Romania, and the highest in Luxembourg and Ireland. In addition, it was noted that in many EU countries (18), the quality of the transport structure expressed by length of state, provincial, and municipal roads in terms of km per 1 km² of the country’s area significantly deviates from the EU average. Such a situation occurs in Sweden, Croatia, and Portugal, where the lowest values of the indicator were recorded. In such countries as Finland, Romania, and Germany, the quality of transport infrastructure is high. In the case of three indicators, Y₇, Y₄, and Y₂, a negative direction of asymmetry was found, with at least moderately strong asymmetry for Y₇ and Y₄. The negative direction of asymmetry of the distributions of these indicators is a desirable phenomenon, meaning that most EU countries have a high percentage of residents with access to safe drinking water, and life expectancy is above the EU average.

In the case of such an important QoL indicator as the Gini income inequality index, the presence of a very weak negative asymmetry was noted. Thus, it can be concluded that the distribution of this index is characterized by symmetry. This is a favorable situation in relation to EU countries because it may indicate that the levels of income inequality in EU countries are similar and there is no strong deviation towards either very low or very high inequality. The Gini coefficient in the EU varies from country to country. The lowest is found in Slovenia, Germany, and Croatia, which indicates less income inequality. In contrast, the highest income inequality was recorded for Portugal, Spain, and Hungary. However, taking into account the symmetry of the distribution of the Gini coefficient, it can be concluded that the differences between countries are not very large, and neither countries with very low nor very high income inequality dominate.

An extremely important variable indicative of QoL is the infant mortality rate per 1000 live births. In this study, EU countries are strongly differentiated in terms of this indicator and have a strong asymmetry in the positive direction. Low levels of this indicator, signifying, among other things, good health care, were observed in countries such as Finland, Estonia, and Sweden, for example, while high levels were observed in Romania, Slovakia, and Malta.

The study also included the Happiness Index [66], which measures the level of happiness and well-being of citizens in various countries. This index is aggregate and captures, among other things, GDP per capita, healthy life expectancy, social support, freedom to make life choices, generosity, and perceptions of corruption. Happiness Index scores are presented on a point scale from 0 to 10, where zero indicates the lowest perceived happiness and 10 the highest. In this study, it was observed that the Scandinavian countries, Finland, Denmark, and Sweden, recorded the highest feelings of happiness, while Bulgaria, Greece, and Croatia recorded the lowest.

The indicators adopted in the study relating to EE are mainly characterized by right-sided asymmetry of at least high strength. The exceptions are the indicators numbered 8 and 9, whose distributions are characterized by negative asymmetry of high and weak strength, respectively. In the case of X₈, negative and strong asymmetry means that in most EU countries, the share of primary energy consumption from fossil fuels exceeds the EU average. The highest level of this indicator was recorded in Luxembourg, Poland, and Cyprus (above 87%), while the lowest was in Sweden and Finland (25.9% and 38.4%, respectively). The observed regularity shows that energy policy in many EU countries is still based primarily on the use of fossil fuels, which brings negative consequences for the environment, economy, and society in the form of, for example, climate change, air and water pollution, negative impact on health, and thus high costs associated with health care expenses. The use of RES can help reduce these negative effects. Analyzing the distribution of values of indicator X₄ showing the share of renewable energy in gross final energy consumption, it was noted that in many EU countries (18), the share of RES is at a low level, such a situation applies to: Luxembourg, Belgium, and Malta (about 14.5%). The highest share of renewable energy in gross final energy consumption was recorded in Sweden (over 66%), Finland (over 50%), and Denmark (over 44%). According to the recommendations of the International Energy Agency (IEA), the share of RES in global primary energy consumption should be a minimum of 35-40%. Another institution, the International Renewable Energy Agency (IRENA), recommends that by 2030, RES should account for 45-50% of primary energy consumption. EU countries, in order to meet the indicated recommendations, should implement specific measures; first of all, to make greater use of RES, which involves investments in obtaining energy from wind, solar, water, and hydrogen. An important indicator from the EE field in the context of QoL is also the population unable to keep their homes adequately warm. EU countries vary strongly on this variable, but in most countries (19), the percentage of the population unable to keep their homes adequately warm is lower compared to the EU average. The worst situation in terms of this indicator is in Spain, Portugal, and Bulgaria (over 20%), while the best situation is in Luxembourg and Finland (less than 3%).

All indicators adopted in the survey describing EE are characterized by strong variation, from 21.5% for X₈ to 63.2% for X₁₀.

4.2. Classification Study Results

The study focused on identifying EU countries in terms of QoL and EE in 2023. The results of the study made it possible to determine groups of countries with stable positions, extreme cases, and countries with ambiguous classification in terms of both QoL and EE. A multi-criteria decision analysis (MCDA) approach was applied in the classification study. MCDA methods allow for solving complex decision-making problems by taking into account multiple, often conflicting, evaluation criteria. The ELECTRE Tri method was used to classify EU countries, which enables the assignment of analyzed objects to predefined classes based on specific preference thresholds and adopted decision parameters. The study assumed a division into the following three classes: countries with the lowest, average, and highest performance in the analyzed areas. Calculations were carried out using the ELECTRE_v1-7.2 software, which is widely used in analyses applying the ELECTRE Tri method and allows for conducting classification using both pessimistic and optimistic approaches. The course of the research procedure is illustrated in Figure 2, which presents the successive stages of the analysis, from the selection of variables to the formulation of conclusions.

4.3. Methodological Assumptions of the Classification of EU Countries Using the ELECTRE Tri Method

The analysis included two sets of variables, each containing 10 indicators. Due to the aggregate nature of the Happiness Index (Y₁₁), it was not included in the classification study, as its components were already adopted as separate indicators. For both areas—QoL and EE—the selection of variables was carried out intentionally, without the use of statistical feature selection procedures. This approach is justified in the ELECTRE Tri method, where the evaluation of objects is based on predefined criteria. The applied variables have clearly defined substantive meaning and appropriate diagnostic value in the analyzed areas. They were selected based on a literature review and expert knowledge, which is consistent with the approach used in this method.

A division into the following three groups of countries was adopted, corresponding to different levels of the analyzed phenomena:

• Class 1—countries with the weakest results, characterized by an unfavorable situation in the analyzed area.
• Class 2—countries with average results, achieving values within the mid-range for the analyzed indicators.
• Class 3—countries achieving the best results, distinguished by high QoL and a high level of EE.

Two reference profiles were used to define the boundaries between classes, serving as benchmarks in the classification process. These profiles specified the minimum requirements that an EU country had to meet in order to be assigned to a higher class:

• Profile 1 (P1)—represented the boundary between Class 1 and Class 2. Its value corresponded to the 33rd percentile level. This means that countries with results not exceeding the specified value were assigned to the weakest group (Class 1). In contrast, the results above this threshold indicated at least an average level of performance, qualifying a given country for Class 2 or 3.
• Profile 2 (P2)—defined the boundary between Class 2 and Class 3 and was set at the 66th percentile level. This means that countries with results above this value were assigned to the group achieving the best outcomes (Class 3).

The adoption of the discussed profile values ensured an even distribution of the data, which is important in analyses involving diverse EU countries. The reference profiles were a key element of the ELECTRE Tri method, as they defined evaluation standards and determined the assignment of countries to the appropriate classes.

Following the determination of reference profiles P1 and P2, the range (difference in values) between them was calculated and used as a reference point for setting the values of the decision thresholds.

• Indifference threshold—set at 4.8% of the range for profile P1 and 5.4% for profile P2. The variation in these values resulted from the analysis of data distribution and the characteristics of variables with strong asymmetry.
• Preference threshold—set at 40% of the range in both cases. This value defines a significant difference between objects, sufficient to recognize the superiority of one country over another.
• Veto threshold—not applied. Tests showed that even at a high level (90% of the range), this threshold was overly restrictive, leading to the unjustified assignment of many countries to the lowest class. Omitting this parameter was justified, as the ELECTRE Tri method allows for effective classification even without it, especially since the remaining thresholds were appropriately selected.

A credibility threshold (λ) of 0.75 was adopted for the classification. This means that, in order for a country to be assigned to a higher class, it had to meet the preference thresholds with at least a 75% level of dominance of positive evaluations over negative ones. Adopting this value represented a compromise between an overly lenient approach (low λ) and an excessively strict one (high λ).

The study also applied equal weights to all variables in both analyzed areas. The adoption of equal weights was based on the assumption that all selected variables are equally important in describing the phenomena related to QoL and EE. This approach helped eliminate the risk of subjective manipulation of the results by, for example, assigning extreme weights to individual features.

The classification of EU countries was carried out in accordance with the procedure of the ELECTRE Tri method, which involves the application of two complementary approaches: pessimistic and optimistic. Both procedures play an important role in assessing the stability of the results and make it possible to identify countries with an uncertain or variable position within the group.

• Pessimistic procedure—assigned a country to the lowest class for which it met the thresholds. It was characterized by greater restrictiveness, which reduced the risk of misclassifying countries with uncertain characteristics into higher groups. Assignment to a higher class was more difficult and required meeting stricter criteria.
• Optimistic procedure—assigned a country to the highest class for which it met the thresholds. It focused on the country’s development potential and capabilities, adopting a more liberal approach. It allowed for the classification of countries that partially met the requirements of a higher class, thus highlighting their strengths and positive outlooks.

The application of both procedures within a single analysis enabled a more accurate assessment of the stability of the results, that is, the identification of countries with an established position within a given class. It also made it possible to distinguish objects sensitive to changes in decision criteria, whose classification depended on the chosen approach. This solution provided valuable insights into countries located at the boundaries between classes and the factors influencing their position in the hierarchy.

4.4. Classification of EU Countries Based on QoL Indicators

The classification of EU countries based on indicators describing QoL provided detailed insights into the dispersion of related factors across the analyzed states. The ELECTRE Tri method enabled the assignment of countries to three classes reflecting different levels of this aspect. The analysis incorporated two decision-making approaches: pessimistic and optimistic. The classification results obtained under both the pessimistic and optimistic approaches are presented in

Table 4. Classification of EU countries according to QoL indicators—pessimistic and optimistic approach.

Pessimistic Approach			Optimistic Approach
Class 1	Class 2	Class 3	Class 1	Class 2	Class 3
Bulgaria	Austria	Ireland	Bulgaria	Austria	Belgium
Croatia	Belgium	Netherlands	Croatia	Cyprus	Czechia
Finland	Cyprus		Romania	France	Denmark
Greece	Czechia			Germany	Estonia
Hungary	Denmark			Greece	Finland
Italy	Estonia			Hungary	Ireland
Latvia	France			Italy	Malta
Lithuania	Germany			Latvia	Netherlands
Luxembourg	Slovenia			Lithuania	Slovenia
Malta	Sweden			Luxembourg	Sweden
Poland				Poland
Portugal				Portugal
Romania				Slovakia
Slovakia				Spain
Spain

.

The next stage of the classification study involved supplementing the results with intermediate classes, the introduction of which allowed for more precise assignment of countries whose results were close to the boundary values. Countries with a stable position under both approaches remained in their previously assigned classes. This applied to Bulgaria, Croatia, Romania, Ireland, and the Netherlands, which were classified into the same group in both the pessimistic and optimistic procedures. For countries whose position differed depending on the decision-making procedure used, intermediate classes were applied. Countries assigned to Class 1/2 achieved better results than those in Class 1 but did not yet meet the requirements typical of Class 2. Meanwhile, Class 2/3 included countries with favorable results but not sufficient to qualify for Class 3. A distinct case was observed in the classification of Finland and Malta, whose positions proved to be ambiguous. Both countries obtained extreme results: they were assigned to the lowest class in the pessimistic procedure and to the highest class in the optimistic one. As a result, it was decided to temporarily assign them to both classes simultaneously (Class 1 and Class 3), emphasizing the uncertainty regarding their final classification. The introduction of intermediate classes allowed for a more accurate reflection of the actual positions of countries in the classification and helped reduce the risk of incorrect decisions resulting from discrepancies in the results obtained under both approaches. The classification results, including intermediate classes, are presented in

Table 5. Classification of EU countries in terms of QoL, including intermediate classes.

Classification Results Using the ELECTRE Tri Method (Including Intermediate Classes)
Class 1	Class 1/2	Class 2	Class 2/3	Class 3
Bulgaria	Greece	Austria	Belgium	Ireland
Croatia	Hungary	Cyprus	Czechia	Netherlands
Romania	Italy	France	Denmark	Finland
Finland	Latvia	Germany	Estonia	Malta
Malta	Lithuania		Slovenia
	Luxembourg		Sweden
	Poland
	Portugal
	Slovakia
	Spain

.

The classification results using intermediate classes provided important insights into the diversity of QoL-related conditions across EU countries. The inclusion of additional categories allowed for a more accurate representation of borderline cases and highlighted the specific characteristics of countries with unstable positions. As a result, the final division of countries became more detailed and more accurately reflected the actual differences in the level of QoL. This approach increased the credibility of the results and made it possible to identify countries whose situation requires further analysis. It also enabled the distinction of countries with extreme outcomes, namely Finland and Malta, whose positions remained ambiguous. The final classification results, taking into account both stable country positions and cases requiring assignment to intermediate classes, are presented in

Table 6. Final classification of EU countries in terms of QoL.

Classification Results Using the ELECTRE Tri Method
Class 1	Class 1/2	Class 2	Class 2/3	Class 3
Bulgaria	Greece	Austria	Belgium	Ireland
Croatia	Hungary	Cyprus	Czechia	Netherlands
Romania	Italy	France	Denmark
	Latvia	Germany	Estonia
	Lithuania		Finlandia
	Luxembourg		Malta
	Poland		Slovenia
	Portugal		Sweden
	Slovakia
	Spain

.

The final classification of EU countries allowed for a precise determination of the variations in conditions affecting QoL across the analyzed states. As a result of applying the ELECTRE Tri method, countries were assigned to five classes: three main ones (Class 1—low level, Class 2—average level, and Class 3—high level) and two intermediate ones (Class 1/2—moderately low and Class 2/3—moderately high). The largest group consisted of countries with a moderately low level of QoL (Class 1/2), which included 10 countries. Class 2/3 (moderately high) included eight countries. The next group was made up of countries with an average level of QoL (Class 2), including four countries, while the extreme classes—1 and 3—included three and two countries, respectively. This structure indicated a predominance of countries with moderately low or average results and a relatively small number of countries at the extremes in terms of QoL levels.

A comparison of the average values of variables within individual classes with the values for the entire group of countries provided additional insights into their specific characteristics.

• Class 1 (lowest level) was characterized by significantly lower average values in gain-type variables: GDP per capita, average life expectancy, expected years of schooling, access to safe drinking water, and road density. Additionally, this group showed noticeably higher average values in cost-type variables: the Gini index, infant mortality rate, Air Quality Index based on PM2.5, and exposure to noise.
• Class 2 (average level) was characterized by higher average values compared to the overall mean in terms of GDP per capita, average life expectancy, and access to safe drinking water. It also showed lower average levels for cost-type variables, such as the infant mortality rate, access to safe drinking water, and exposure to noise.
• Class 3 (highest level) stood out with favorable values for all gain-type variables and the lowest indicators for cost-type variables.
• Intermediate classes (1/2 and 2/3) were characterized by intermediate results—closer to the values of Class 1 or Class 3, respectively—which confirms the rationale for their introduction.

Countries with a stable position, such as Bulgaria, Croatia, Romania, Ireland, and the Netherlands, were assigned to the same classes in both the pessimistic and optimistic approaches. This phenomenon indicated a consistent position of these countries in relation to the analyzed indicators.

Particular attention was given to Finland and Malta, whose positions in the classification turned out to be ambiguous. In the pessimistic procedure, both countries were assigned to Class 1, while in the optimistic approach, they were placed in Class 3. This discrepancy resulted from the varied values of the analyzed indicators. Ultimately, this led to the assignment of Finland and Malta to the intermediate Class 2/3 (moderately high), because

• Finland achieved favorable results in terms of the Gini index, expected years of schooling, infant mortality rate, and air quality based on PM2.5. At the same time, the country showed unfavorable values in unemployment rate, length of national roads, and levels of household noise exposure. The combination of strong and weak results in key areas influenced the assignment of Finland to the intermediate class.
• Malta achieved good results in terms of unemployment rate, access to safe drinking water, average life expectancy, and the length of the road network. At the same time, it recorded unfavorable results in the Gini index, infant mortality rate, and noise exposure. This complex performance profile led to its assignment, similarly to Finland, to Class 2/3.

The cases of Finland and Malta show the importance of introducing intermediate classes, which made it possible to more accurately reflect the actual situation of countries with varied results. This approach increased the accuracy of the final classification, allowing for a clearer distinction in the level of QoL among EU countries. Taking into account the full set of indicators proved important for a reliable assessment of this aspect, especially in the case of countries with results close to the boundary values.

4.5. Classification of EU Countries Based on EE Indicators

The classification of EU countries based on EE indicators enabled a detailed analysis of the degree of rational energy management in individual states. The application of the ELECTRE Tri method allowed for the assignment of countries to three main classes, reflecting different levels of EE. Additionally, two decision-making approaches—pessimistic and optimistic—were introduced, making it possible to assess the stability of the results and identify countries with ambiguous positions. The classification results obtained under both approaches are presented in

Table 7. Classification of EU countries according to EE indicators: pessimistic and optimistic approach.

Pessimistic Approach			Optimistic Approach
Class 1	Class 2	Class 3	Class 1	Class 1	Class 3
Austria	Denmark	Croatia	Belgium	Bulgaria	Austria
Belgium	Germany	Portugal		Cyprus	Croatia
Bulgaria	Greece	Romania		Czechia	Denmark
Cyprus	Hungary			France	Estonia
Czechia	Italy			Germany	Finland
Estonia	Latvia			Greece	Hungary
Finland	Lithuania			Ireland	Latvia
France	Slovenia			Italy	Lithuania
Ireland	Spain			Malta	Luxembourg
Luxembourg				Netherlands	Portugal
Malta				Poland	Romania
Netherlands				Slovakia	Sweden
Poland				Slovenia
Slovakia				Spain
Sweden

.

The classification results revealed significant differences in the assignment of individual countries depending on the decision-making procedure applied. Countries assigned to the same groups under both approaches can be considered stable. In contrast, countries whose positions varied required additional analysis. To better reflect the situation of countries with results close to boundary values, intermediate classes were introduced. This solution allowed for a more accurate assignment of countries with ambiguous results. Intermediate classes also made it possible to account for cases where countries achieved good results in some indicators but weaker outcomes in other aspects of EE. This approach ensured a more reliable division of countries and a more accurate reflection of their actual position in the classification.

Countries whose positions were stable in both decision-making approaches (pessimistic and optimistic) retained their initial assignment to a specific class. This group included Belgium, Germany, Greece, Italy, Slovenia, Spain, Croatia, Portugal, and Romania. The stability of these countries’ positions indicates a consistent situation with respect to the analyzed EE indicators. For countries whose positions differed depending on the decision-making procedure adopted, intermediate classes were applied. Countries classified into the moderately low EE class showed more favorable results than those assigned to Class 1, but their performance was not sufficient to qualify them for Class 2. This means that these countries achieved partially positive results but still had significant areas requiring improvement. This group included Bulgaria, Cyprus, The Czech Republic/Czechia, France, Ireland, Malta, the Netherlands, Poland, and Slovakia. Similarly, countries assigned to the moderately high EE class obtained more favorable results than those classified in Class 2, but their performance was not high enough to fully meet the requirements of the best-performing class, Class 3. These countries stood out with solid results in several important areas, but certain indicators revealed significant limitations. This group included Denmark, Hungary, Latvia, and Lithuania.

The introduction of intermediate classes made it possible to account for borderline cases and increased the accuracy of the classification, allowing for a better representation of the variation in EE levels among EU countries. Particular attention was given to five countries: Austria, Estonia, Finland, Luxembourg, and Sweden, which were characterized by an ambiguous position. In the pessimistic procedure, these countries were assigned to the lowest class, while in the optimistic approach, they were placed in the highest class. This discrepancy resulted from significant differences in the values of individual EE indicators. As a result, these countries were assigned to extreme classes, reflecting the complexity of their situation. The classification results, including intermediate classes, are presented in

Table 8. Classification of EU countries in terms of EE, including intermediate classes.

Classification Results Using the ELECTRE Tri Method (Including Intermediate Classes)
Class 1	Class 1/2	Class 2	Class 2/3	Class 3
Belgium	Bulgaria	Germany	Denmark	Croatia
Austria	Cyprus	Greece	Hungary	Portugal
Estonia	Czechia	Italy	Latvia	Romania
Finland	France	Slovenia	Lithuania	Austria
Luxembourg	Ireland	Spain		Estonia
Sweden	Malta			Finland
	Netherlands			Luxembourg
	Poland			Sweden
	Slovakia

.

The final classification of EU countries in terms of EE confirmed a significant diversification of results. The outcomes obtained using the ELECTRE Tri method allowed for the division of countries into five groups: three main classes and two intermediate ones. The largest group consisted of countries with moderately low EE (Class 1/2), including nine countries. Class 2 (average) included six countries, while Class 2/3 (moderately high) comprised four countries. Among the extreme groups, Class 1 (lowest efficiency) contained sic countries, and Class 3 (highest efficiency) included eight EU countries. This distribution of results indicates a high level of variation in EE across EU countries, with a clear predominance of states with intermediate results and a significant number of countries achieving high EE.

As a result of comparing the average values of variables describing EE in individual classes with the values for the entire group of countries, information was obtained about the characteristics of each group.

• Class 1 (lowest level) was characterized by significantly lower average values in gain-type variables describing energy productivity and the share of renewable energy in final energy consumption. For cost-type variables, markedly higher average values were recorded for final energy consumption, energy intensity of GDP, CO₂ emissions, electricity demand, the share of fossil fuels in primary energy consumption, and energy import dependency.
• Class 2 (average level) was characterized by higher average values compared to the overall mean in terms of energy productivity and lower average levels of cost-type variables, particularly those describing final energy consumption per capita, energy intensity of GDP, electricity demand, and the share of fossil fuels in primary energy consumption.
• Class 3 (highest level) stood out with favorable values for most gain-type variables and the lowest indicators for cost-type variables. An exception was the variable concerning the percentage of the population unable to keep their home adequately warm due to poverty status. In this case, the cost-type average was higher than the overall mean.
• Intermediate classes (1/2 and 2/3) were characterized by intermediate results—closer to the values of Class 1 or Class 3, respectively—which confirms the rationale for their introduction.

Among the stable countries were Belgium, Germany, Greece, Italy, Slovenia, Spain, Croatia, Portugal, and Romania. The stability of these EU countries’ positions indicated consistent results obtained in both decision-making approaches. In contrast, Austria, Estonia, Finland, Luxembourg, and Sweden were characterized by an ambiguous position, as evidenced by their simultaneous grouping in the extreme classes. The analysis of variable profiles made it possible to identify the factors contributing to these discrepancies. The final classification results are presented in

Table 9. Final classification of EU countries in terms of EE.

Classification Results Using the ELECTRE Tri Method
Class 1	Class 1/2	Class 2	Class 2/3	Class 3
Belgium	Bulgaria	Austria	Denmark	Croatia
Luxembourg	Cyprus	Germany	Hungary	Portugal
	Czechia	Greece	Latvia	Romania
	Estonia	Italy	Lithuania
	Finland	Slovenia	Sweden
	France	Spain
	Ireland
	Malta
	Netherlands
	Poland
	Slovakia

.

Finland showed significant variation in its results. Its strengths included a high share of renewable energy in gross final energy consumption, a favorable structure of fossil fuel use, and relatively low dependence on energy imports. Additionally, a small percentage of the population suffering from energy poverty indicated good energy accessibility. In contrast, Finland was characterized by high final energy consumption per capita, high energy intensity of GDP, and low energy productivity. Moreover, high electricity demand per capita and low efficiency in electricity production represented important limitations. This combination of strong and weak aspects contributed to Finland being assigned to an intermediate class.

Austria also recorded results indicating clear disparities. Positive aspects of the country included a high share of renewable energy and a favorable fossil fuel structure in primary energy consumption. Like Finland, Austria stood out with a low percentage of people at risk of energy poverty. However, poor results in final energy consumption per capita, high electricity demand, and low efficiency in electricity production led to Austria being assigned to the second class in the final classification.

Estonia is a country that stood out positively in terms of its share of renewable energy, low dependence on energy imports, and limited energy poverty. On the negative side, the dominant issues were high energy intensity of GDP, low energy productivity, and high CO₂ emissions per capita. In addition, Estonia recorded low efficiency in electricity production and excessive use of fossil fuels, which lowered its position in the classification.

Luxembourg is a particularly complex case, as the country was predominantly characterized by negative indicators. Among the positive aspects were favorable results in energy intensity of GDP and energy productivity. In addition, the percentage of the population affected by energy poverty was low. However, Luxembourg showed high final energy consumption per capita, a low share of renewable energy, and high CO₂ emissions. Poor results were also recorded in electricity production per capita and significant dependence on energy imports, which ultimately led to Luxembourg being assigned to the first class.

According to the conducted study, Sweden is a country with ambiguous results. Among its strengths were low CO₂ emissions, a high share of renewable energy, and a favorable structure of fossil fuel consumption. As in the case of Finland and Austria, the percentage of the population at risk of energy poverty was low. However, negative aspects such as high final energy consumption per capita, high energy intensity of GDP, and low energy productivity influenced the assignment of the country to an intermediate class. It should be noted that all of the countries mentioned showed clear disparities between their positive and negative results due to the analyzed indicators. As a result, assigning them to intermediate classes stemmed from the need to account for both their strengths and significant limitations in the area of EE.

4.6. Comparison of Classification Results and Analysis of Group Consistency

The comparative analysis of the classification of EU countries based on QoL and EE indicators in 2023 revealed both similarities and significant differences among the countries. The results made it possible to distinguish groups of countries with a stable position in both analyzed aspects, as well as those whose place in the classification turned out to be ambiguous. The study applied the ELECTRE Tri method, which allowed for assigning countries to one of five classes representing levels of QoL and EE. The analysis identified a group of countries with consistent positions in both classifications. This group includes Poland and Slovakia, classified in Class 1/2 (moderately low position) in both aspects; Austria and Germany, assigned to Class 2 (average level); and Denmark and Sweden, which were placed in Class 2/3 (moderately high position). The consistency in the positions of these countries indicates a balanced level of QoL and EE.

The analysis also identified a group of countries that achieved high results in terms of QoL but received low scores in EE. This category includes Ireland and the Netherlands, classified in Class 3 in the QoL classification and Class 1/2 in the EE assessment. This disparity may result from uneven investments across different areas of development or from limiting factors affecting EE despite a high level of QoL. A different group consists of countries with low QoL but high EE. This group includes Croatia and Romania, assigned to Class 1 in the QoL classification and Class 3 in the EE classification. These countries, despite low QoL, achieved high EE, which may be attributed to intensive investments in energy infrastructure and limited spending on social aspects. It should be noted that countries with low QoL but high EE should aim to balance investments in social infrastructure, while countries with high QoL but low EE could benefit from technological solutions and support from pro-environmental policies.

The obtained classification results also revealed a group of countries with an ambiguous position, whose assignment to a specific class varied depending on the area analyzed. Hungary, Latvia, and Lithuania were placed in Class 1/2 in terms of QoL and Class 2/3 in terms of EE. Finland, Malta, The Czech Republic/Czechia, and Estonia were classified in Class 2/3 in the QoL classification and Class 1/2 in the EE assessment. The variability in the positions of these countries indicated significant differences between specific indicators and potential areas of imbalance that require further analysis.

The results obtained provide valuable insights into the diversity of QoL and EE conditions across EU countries. The analysis outcomes may serve as support for policymakers in planning actions aimed at improving EE and enhancing the QoL of citizens. Countries with high development potential can draw on the experiences of those with stable QoL. In turn, countries with results close to the boundary values between classes should implement strategies to reduce social and environmental inequalities. For countries with extreme outcomes, the introduction of integrated actions supporting both economic development and the improvement of social conditions is recommended. The spatial distribution of EU countries according to the classification results is presented in Figure 3, which shows areas with similar performance in both analyzed domains as well as regions marked by significant disparities. Mapping the results facilitates the identification of regions that require particular attention when planning actions related to QoL and EE.

5. Discussion

From the analysis of the results obtained in this study, the following can be concluded:

• In the classification of countries in terms of QoL, the most numerous group was made up of countries with a moderately poor level of this phenomenon, while the least numerous groups were countries belonging to the two extreme classes, with poor and high QoL.
• In the case of the classification of EU countries in terms of EE, the extreme classes showing weak and good levels of EE were also the least numerous.
• Some countries represent a very similar level, that is, they are consistent in terms of QoL and EE. These countries include Poland and Slovakia, included in the classes with moderately poor levels of both categories; Austria and Germany, assigned to the class with a mediocre level; and Denmark and Sweden, placed in the class with a moderately good position. The consistency of these countries’ positions is indicative of an even level of QoL and EE.

This study not only determines the unambiguous nature of the EU countries’ classes but also shows an indirect picture of the countries in terms of the phenomena studied and determines whether a country is moving toward improving or worsening its situation in terms of QoL and EE.

A review of the literature shows that the measurement and evaluation of QoL are carried out by many researchers representing different scientific disciplines. Some focus on theoretical considerations, definitions, and classifications of this phenomenon [86], while others attempt to study this category using various research tools and methods. As the literature review shows, due to its complex and interdisciplinary nature, QoL is analyzed in terms of both multiple aspects [8] and one selected according to the researcher’s interests, such as housing deprivation [87], housing conditions [9], higher education [88], or poverty [89]. With global changes, especially in the environment and climate, researchers are directing more and more attention to assessing QoL in the context of SD. Among the many issues addressed are climate change [90], energy transition [91], housing revitalization [92], or the degree to which the SDGs have been achieved [93]. The researchers concluded that there is a link between the issues identified and QoL. One of the conclusions is also the need to continue evaluating this category, especially in the context of SD. Our study is part of the theme of QoL research in relation to SD. Unlike previous studies, this study integrates the assessment of QoL and EE in EU countries using the ELECTRE Tri method, making it possible to identify both the current status and potential direction of each country. Moreover, this study is innovative in that it addresses a specific aspect of SD and sustainable energy, namely EE. There is still a lack of such studies that focus on EE, which is one of the most important elements of SD and, consequently, QoL. By targeting EE, less energy is consumed to produce a product. The literature review conducted in the study shows that the issue of EE is addressed by many authors. Researchers focus on different aspects of the phenomenon and present it in different thematic contexts, such as analyzing EE in heating [94], construction [95,96,97,98], urban development [99], or transportation [100]. They also point to the activities of various actors taking steps to increase EE [15]. They present the EE phenomenon not only internationally [101] and nationally [36,102,103] but also regionally and locally [15]. They refer to EE programs and their impact on the macroeconomic situation of countries [104,105,106,107]. Regardless of the perspective adopted, it should be recognized that the approaches presented link EE with the social, economic, and environmental components of SD and relate to the determinants of QoL. Indeed, improving EE is one of the main action areas in the context of implementing the SD principle [108] and contributes to improving QoL. In light of the 2030 agenda, the results obtained can support the monitoring of progress in the realization of selected SDG goals, especially SDG 3 (Good Health and QoL), SDG 7 (Clean and Accessible Energy), and SDG 13 (Climate Action). Our study provides new findings on the evaluation of EE, focusing on indicators from different areas that characterize the phenomenon. By taking into account variables from such areas as energy consumption, RES use, CO₂ emissions, the availability and stability of energy supply, and energy security, it was possible to conduct an in-depth assessment of this category and better understand the issues involved. It should also be emphasized that our survey captures the assessment of QoL in EU countries in addition to EE. The innovative nature of this survey is related to highlighting the importance of EE in the context of QoL through, among other things, indicators that provide a comprehensive picture of EE in EU countries. In addition, both categories’ assessments were conducted in unidimensional and multidimensional terms. The advantage of our study is that it primarily uses indicators included in the Eurostat database and, as a complement, uses information from Global Economy, World Economy, and Our World in Data for monitoring QoL and EE in the EU. Therefore, the selection of indicators is not random, but follows the established standards set by Eurostat and other databases. The results of our study cannot be compared with those of other researchers because they are based on a different set of indicators and are analyzed using a different method. This study used ELECTRE Tri, one of the MCDA methods, which is not often used in QoL research in the context of EE. The effectiveness of the ELECTRE Tri method in evaluating energy strategies is well documented, but its application to analyzing the impact of energy decisions on QoL requires further research. Using the ELECTRE Tri method, it is possible to determine not only the explicit nature of the classes of EU countries in terms of the analyzed phenomena but also to show an indirect picture of the countries and to determine whether a country is moving toward improving or worsening its situation. The strengths of the present study include the following:

• The identification of QoL in EU countries in the context of EE on the basis of a set of indicators, making it possible to cover EU countries in a comprehensive manner;
• The use of the ELECTRE Tri method, through which one can obtain a classification of objects into predefined categories, taking into account both measurable indicators and the subjective preferences of decision-makers.

Despite these many strengths, there were some limitations during the preparation of the study. Most often, these limitations were related to data availability. Before proceeding with the study, it was assumed that the survey would cover not only EU countries but also other European countries included in the Eurostat database. However, after careful analysis of the material, non-EU countries were dropped from the study, precisely because of significant data shortages. The European Commission recommends that the application of the “energy efficiency first” principle should be based on evidence, which requires proper verification, monitoring, and evaluation of the effects of decisions taken, particularly in terms of energy consumption. It also requires detailed and correct information and data. In many cases, energy information to make decisions based on more accurate knowledge is not available. Adequate resources should be allocated for data collection and compilation of statistics, to which the relevant units should have access. Decisions should also be evaluated from the perspective of future technological development and should encourage innovation that helps achieve the EU’s environmental, social, and economic goals [20]. The limitations indicated are useful for determining future research directions that take into account conducting analyses with the inclusion of other European countries, which include those that are applying for EU membership and those that are no longer members of the EU.

The created sets of indicators relating to QoL and EE, despite the limitations presented, provide an opportunity to identify and compare countries in terms of the phenomena studied.

The obtained research results presented in this article can be used by governments and EU institutions to monitor and compare QoL and EE in the countries accepted for the study. Having this information in EU countries allows for appropriate decisions to be made to reduce inequalities among regions or social groups. The classification carried out can support decision-makers in taking measures aimed at reducing regional disparities and supporting sustainable energy transitions. Mapping the position of countries stands now as a practical tool in cohesion policy and the management of EU funds.

6. Conclusions

The conducted study fits in and fills a gap in the discussion on QoL in the context of EE, supplementing it with a new methodological approach. The use of the ELECTRE Tri method allowed for the separation of three classes of EU countries, which corresponded to the following different levels of the analysed phenomena:

• Class 1—countries with the weakest results, characterized by an unfavourable situation in the analysed area.
• Class 2—countries with average results, achieving values within the average range for the analysed indicators.
• Class 3—countries achieving the best results, i.e., distinguished by high QoL and high EE.

In addition, the following two complementary approaches were used: pessimistic and optimistic, thanks to which it was possible to assess the stability of the results and identify countries characterized by an uncertain or changing position in the class.

Answering the first research question concerning the unambiguous nature of the classes of EU countries in terms of QoL and EE, it was shown that in the case of some countries, based on the adopted set of indicators, their nature cannot be unambiguously determined. For countries such as Malta and Finland, a stable situation in the classification in terms of QoL was not confirmed. In the pessimistic procedure, both countries were assigned to Class 1, while in the optimistic approach to Class 3. Finally, after an in-depth analysis of the values of the indicators, Finland and Malta were assigned to the intermediate Class 2/3 (moderately good). In contrast, countries such as Bulgaria, Croatia, and Romania, in the final classification of EU countries in terms of QoL, were included in the class with the lowest QoL, while Ireland and the Netherlands were included in the class with the highest QoL. The cases of Finland and Malta emphasize the importance of introducing intermediate classes, thanks to which it is possible to more accurately reflect the real situation of countries with different results.

In the case of classification in terms of EE, an ambiguous nature was observed in relation to countries such as Bulgaria, Cyprus, the Czech Republic/Czechia, Estonia, Finland, France, Ireland, Malta, the Netherlands, Poland, and Slovakia. Again, a thorough analysis of the indicator values allowed assigning these countries to the intermediate Class 1/2 (rather poor). In contrast, Denmark, Sweden, Hungary, Latvia, and Lithuania also did not show an unambiguous nature but were in the intermediate Class 2/3 (moderately good). The class with the lowest level of EE included Luxembourg and Belgium, while the class with the highest level of EE included Croatia, Portugal, and Romania.

Referring to the second research question, “Do EU countries belong to similar classes in terms of QoL and EE?”, it can be stated that the comparative analysis of the classification of EU countries showed similarities as well as differences between countries. The results made it possible to distinguish groups of countries with a stable position in both analyzed aspects and countries whose place in the classification turned out to be ambiguous. Similarity in terms of QoL and EE was noted for Poland and Slovakia, which were classified in the intermediate Class 1/2 in terms of both analyzed categories. Another similarity was noted for Austria and Germany, which were classified in Class 2 (average), as well as Denmark and Sweden, which belonged to Class 2/3 (moderately good). In the case of the indicated countries, a similar level of QoL and EE was noted. The conducted research showed between which countries there are differences in terms of the studied categories. Ireland and the Netherlands were classified in Class 3 in terms of QoL (high QoL) and Class 1/2 in terms of EE (rather low EE). A different situation was observed in the case of Croatia and Romania, which were assigned to Class 1 (low QoL) in terms of QoL and to Class 3 in terms of EE. Therefore, these countries, despite a low QoL, obtained a high EE. The analysis also showed countries with an ambiguous situation in terms of both QoL and EE. Countries such as Hungary, Latvia, and Lithuania were in Class 1/2 in terms of QoL and Class 2/3 in terms of EE (rather poor QoL and rather good EE). In contrast, Finland, Malta, the Czech Republic/Czechia, and Estonia were in Class 2/3 in the QoL classification and Class 1/2 in the EE assessment (rather good QoL and rather poor EE).

The obtained results provided valuable information on the diversity of QoL and EE conditions in EU countries. The results of the analysis can support policymakers in planning actions aimed at improving EE and raising the QoL of residents. Countries with low development potential can draw on the experience of countries with stable QoL. In turn, countries with results close to the borderline values between classes should implement strategies to reduce social and ecological inequalities. Directions for future research should include the analysis of dynamic changes over time and extending the geographical scope to countries neighboring the EU, including candidate countries and former members of the Community. Additionally, the use of alternative multi-criteria methods, such as PROMETHEE or AHP, could serve as a supplement to the obtained results and enable their comparison with other decision-making approaches. The results obtained can serve as a tool to support decision-making in the field of energy and social policy, enabling more effective targeting of activities with low EE and QoL. The classification of countries can also be a starting point for the development of differentiated regional strategies within the framework of EU cohesion policy.

For countries with extreme results, it is advisable to introduce integrated actions supporting economic development and improving social conditions. The applied research approach can serve as a practical tool for shaping policies oriented towards social cohesion and energy transformation, contributing to the implementation of SD goals on an EU scale.