Progressing Sustainable Development Goal 7 via Energy Access: Results from the 27 EU Member States

Abstract

Energy plays an undeniable role in socioeconomic development. The energy issue is addressed in Agenda 2030 as one of its 17 goals. The article presents research on the statistical evaluation of the level of implementation of Sustainable Development Goal 7, that is, ‘ensure access to affordable, reliable, sustainable and modern energy for all’ in the 27 EU member states. Seven indicators measuring different areas of energy accessibility monitored by Eurostat were applied to evaluate the EU countries. A multivariate comparative analysis was used to create a synthetic measure, i.e., the zero unitarisation method. The empirical analysis resulted in a ranking of the 27 EU countries in terms of the level of SDG 7 in 2015–2023. In addition to the ranking, a classification of countries into groups of similar countries in terms of the level of SDG 7 in 2015 and 2023 was presented. Furthermore, countries with the most significant advancement and those with the greatest declines in achieving SDG 7 in 2023 compared to 2015 were proposed. A sensitivity analysis was applied to evaluate a robustness of the composite indicator. Research confirmed a downward differentiation of EU countries in terms of the degree of implementation of SDG 7. Denmark, Romania, and Sweden were the top three countries in 2015, and Denmark, Sweden, and Estonia in 2023. The countries that appeared the weakest in the implementation of SDG 7 were in 2015, Cyprus, Belgium, and Luxembourg, and in 2023, Belgium, Lithuania, Cyprus, and Luxembourg. Luxembourg, Ireland, and Malta made the greatest progress toward SDG 7. At the same time, the largest decline was observed in Spain. The research results highlighted the problems in implementing SDG 7 not only for the less developed countries, mainly in Central, Eastern and Southern Europe, but also for the highly developed Western countries. The applied research procedure may help to identify areas for improvement needed in the effective implementation of SDG 7 in the EU member countries. The procedure, however, has limitations. These include, among others, using a linear approach, taking into account only variables measured by Eurostat or an ambiguous effect of energy consumption indicators on SDG 7 evaluation.

1. Introduction

Overcoming challenges and taking advantage of the many opportunities in today’s world requires access to energy. Modern energy is fundamental to human development. It initiated the industrial revolution over 200 years ago and has played a pivotal role in fostering the almost uninterrupted global economic expansion experienced since that time [1]. Energy is necessary to do work, provide security, combat climate change, produce food, or strive to increase national income. It is at the centre of environmental and economic issues [2]. Energy is a fundamental component of both economic development and ecological sustainability. Insufficient energy availability impedes both economic growth and human progress [3].

Energy is essential to achieve the 2030 Agenda for Sustainable Development, which was adopted in 2015 by all 193 Member States of the United Nations (UN) and is expected to be achieved by 2030 [4,5,6]. The 17 Sustainable Development Goals (SDGs), along with their 169 related targets, provide a comprehensive framework designed to direct the initiatives of both governmental bodies and non-state entities. These tackle numerous worldwide issues, such as poverty, inequality, energy, climate change, environmental degradation, peace, and justice [7,8].

The seventh global SDG is to ‘ensure access to affordable, reliable, sustainable, and modern energy for all’. Three targets anchor Goal 7: ensuring universal access to energy services (7.1), increasing the share of renewables in the energy mix (7.2), and improving energy efficiency (7.3). The priorities for implementing SDG 7 are to enhance international cooperation and promote investment (7.a) and to expand infrastructure and upgrade technology in developing countries (7.b) [9]. However, the latest 2022 Tracking SDG 7: The Energy Progress Report, collectively issued by the International Energy Agency (IEA), the International Renewable Energy Agency (IRENA), the United Nations Statistics Division (UNSD), the World Bank and the World Health Organisation (WHO), declares that at the current rate of progress, the world is not on track to achieve the SDG 7 by 2030 [10].

Progress towards the accomplishment of the 17 goals is tracked, assessed, and analysed through various sets of indicators [11]. The UN applies a global set of indicators, while Eurostat provides a set of indicators for the European Union (EU). Moreover, every nation tracks its priorities using distinct indicators, customised according to the unique conditions of the country or region [12]. However, it is important to stress that there is no single perfect method for measuring the advancement of the SDGs [13]. Furthermore, research suggests that current indicators may not fully capture the complexities of energy sustainability [11,14].

In the article, we propose measuring progress in the implementation of SDG 7 using multivariate comparative analysis (MCA). This approach has advantages over single measures because it creates a synthetic measure (SM) that combines information from the various areas to be assessed. This allows comparisons between countries and assessments of the advancement of the phenomenon being studied. Composite measures can provide a more accurate and comprehensive understanding of energy sustainability, guide better policy decisions, and evaluate progress toward achieving SDG 7. Therefore, the purpose of this article is to assess the level of implementation of SDG 7 in the 27 EU countries from the introduction of the 17 SDGs (2015) to 2023, that is, the point in time when the latest statistics were available. Evaluation was carried out using a zero unitarisation method (ZUM) belonging to the MCA. Employing a linear approach enabled the creation of a ranking for countries, a feat unattainable through non-linear methods, which merely group similar objects without offering a ranking option [15].

Seven indicators were used to evaluate countries, measuring different areas of energy accessibility monitored by Eurostat [16]. Based on this, a ranking of countries was presented in terms of the implementation of this goal for the period 2015–2023, as well as their classification into groups of similar countries in terms of the implementation of SDG 7.

Furthermore, while there are studies on the implementation of SDG 7 in less developed countries [10,17,18], there are limited results from highly developed states [19,20,21,22,23]. These countries, despite using the latest technologies, often face problems of excessive energy consumption. Therefore, it is important to identify areas for improvement to ensure the effective implementation of SDG 7. This article fills a perceived research gap by providing results from the 27 member countries of the EU. Empirical analyses were performed for the years 2015–2023. This made it possible to assess countries’ level of advancement of SDG 7 in the period studied.

The article raises the following research questions.

RQ1: Which of the 27 EU countries were the best in achieving SDG 7?

RQ2: Which of the 27 EU countries were the weakest in achieving SDG 7?

RQ3: Which countries made the most significant advancement in terms of achieving SDG 7 in 2023 compared to 2015?

RQ4: Which countries made the biggest declines in terms of achieving SDG 7 in 2023 compared to 2015?

RQ5: How did classifying the 27 EU countries into groups of similar countries look in terms of achieving SDG 7 in 2015, and 2023?

The structure of the article is as follows. After the Introduction, the paper presents a Literature Review Section. This Section consists of two parts: the first part is devoted to presenting the results of research on the SDGs, in particular, Goal 7, and the second presents the characteristics of areas measured by the Eurostat. The next Section, Data and Method, presents diagnostic variables, the method of their selection, and the fundamentals of further statistical analysis. The Results Section is devoted to a detailed presentation of the research results. The Discussion Section is dedicated to comparing the results of research with those available in the subject literature. The article ends with the Conclusions Section, which contains research limitations.

2. Literature Review

2.1. Research Results on SDGs

Research on the implementation of the SDGs is being undertaken by institutions and international organisations (e.g., UN, World Bank, Eurostat) as well as individual countries. In this regard, the number of studies is large, with research topics on the 17 SDGs. The situation is different for scientific studies. Publications on evaluating the implementation of all the SDGs are rare. Scientific articles generally focus on the issue of assessing the implementation of selected SDGs. The SDGs are interrelated, and the realisation of one cannot occur without the realisation of the others.

Access to clean and affordable energy for all is a fundamental enabler of economic activities, included in SDG 8. Goal 8 promotes inclusive and sustainable economic growth, employment, and decent work for all. A comparative analysis of EU countries from 2015 to 2020 shows positive changes in the implementation of SDG 8, despite setbacks due to the COVID-19 pandemic. In 2020, 12 EU countries improved their position in the SDG 8 ranking, while 12 countries fell [11]. The problem of assessing the level of achievement of Goal 8 was addressed in the study by [12]. According to the results, Denmark, Finland, The Netherlands, and Sweden were the countries where the achievement of SDG 8 was the highest, while the lowest was in Greece, Italy, Romania, Slovakia, and Spain. The study also showed that countries that joined the EU in 2004 generally demonstrated a much lower degree of implementation of SDG 8 compared to well-developed Western Europe.

To address SDGs, the adoption of an innovative approach and new technologies is needed. Research was carried out that evaluated the level of EU countries in terms of building stable infrastructure, promoting sustainable industrialisation, and fostering innovation, the main areas of Goal 9 of Agenda 2030 [24]. The results show that the 27 EU countries varied significantly in implementing SDG 9 of Agenda 2030 during the analysed period (2015–2020). The results also provided an opportunity to trace changes in the value of the designated index in individual countries and groups of countries of the ‘old’ and ‘new’ EU. Ref. [25] also conducted an evaluation of the implementation of SDG 9 in the 27 member states. Results achieved with the application of multidimensional analysis underscored significant challenges in digital transformation (DT), particularly within Central, Eastern, and Southern European nations. Ref. [26] applied a multi-criteria statistical analysis that evaluated the phenomenon of sustainable industrial development in Poland and Czechia using indicators identified by the UN under SDG 9. The study emphasised the importance for the examined nations to foster a competitive economy, generate employment, and safeguard the environment by efficiently utilising non-renewable resources.

The problem of achieving Goal 7 is addressed primarily to developing and underdeveloped countries. Research results from these countries are numerous and accurately illuminate the problem of energy access in these countries. For example, ref. [10] explained the problem of energy poverty and the challenges that threaten progress toward SDG 7. It was shown that automation in the workplace significantly increases household energy poverty in China. Ref. [17] stressed that Ethiopia has invested heavily in the power sector to achieve SDG 7, however, peace is needed. Ref. [18] offers a significant review and examination of the influence of energy access initiatives across eight nations in Africa, Asia, and Latin America. Research underlined the crucial role of aligning international, national, and regional policies to achieve the SDGs [18]. The global analysis of SDG 7 was conducted by [27]. Research evaluated the national performance of SDG 7 through 29 indicators capturing environmental and techno-economic aspects relevant to the sustainability of the energy sector. The framework was applied to 176 countries. It was stated that 52 countries from Africa, and Asia could show performance below their threshold according to SDG 7.

Although many theoretical and empirical works are available on assessing the SDGs, there is a dearth of scientific studies on measuring the level of advancement of EU countries in implementing SDG 7. In particular, there is a paucity of scientific studies that provide results on the progress of all EU countries concerning the implementation of this goal. Despite available studies in EU countries in this area [19,20,21,22,23], there are no recent results (for 2023).

Ref. [19] assesses the progress of the implementation of SDG 7 for 2010–2021 and the distance from the 2030 target. The smallest distance observed about the target set for SDG 7 was for Sweden, Denmark, Estonia, and Austria. The problem of the progress of EU countries in implementing SDG 7 was also raised in [20]. The empirical analysis covered the period 2019–2020. The study showed a disparity in the implementation of SDG 7 between the ‘old’ and ‘new’ EU countries. Ref. [21] broadens the problem of assessing the progress of EU countries in SDG 7 to include the relationship between this level and social, economic, and environmental factors. The analysis underlined the crucial role of ecological factors in driving clean energy development. Similarly, ref. [22] evaluated the advancement of EU countries in adopting SDG 7 in 2010–2017. Sweden was at the top of the ranking, and Luxembourg was at the bottom. The strong position of Sweden in terms of the implementation of Goal 7 was also confirmed by [23]. Another country that was high in the ranking was Austria. This empirical analysis covered three time points, i.e., 2010, 2015 and 2020.

As presented above, the latest results on the advancement of SDG 7 in EU countries are from 2021. This article fills the perceived research gap by providing the latest results on the level of progress in the implementation of SDG 7 in the 27 EU countries between 2015 and 2023.

2.2. SDG 7 Areas

The first area of SDG 7 analysed by Eurostat [16] is primary energy consumption. Ensuring a stable supply of primary energy is essential for energy security, which is the cornerstone of SDG 7 [28]. Countries with higher primary energy consumption often have better infrastructure and resources to ensure widespread electrification and access to clean fuels. However, countries with high industrial energy consumption need robust frameworks to integrate renewables and reduce the dependence on fossil fuels [29]. They must manage their primary energy consumption by increasing the share of renewables in their energy mix [30]. However, primary energy consumption reflects overall energy consumption and is weakly related to SDG 7. Thus, it has limited relevance in the assessment of SDG 7 [27].

The final energy consumption in EU countries is influenced by various factors, including improvements in energy efficiency, economic activities, and the adoption of renewable energy sources. Economic growth and increased activity in the financial sectors lead to higher energy consumption. Developed countries have a higher energy consumption per capita than underdeveloped states [31]. The contribution of this indicator is strongly dependent on other indicators related to demographic evolutions, energy efficiency, or renewable shares [27]. This indicator, as the previous one, is rather weakly related to SDG 7.

The final energy consumption in households is closely related to the Gross Domestic Product (GDP) per capita. A higher GDP per capita generally leads to higher energy consumption, although this relationship varies between EU countries [32]. The share of renewable energy consumption in households is increasing, while the share of fossil sources is decreasing [33]. An interpretation of this indicator is similar to the interpretation of final energy consumption in the evaluation of SDG 7.

Energy productivity is the next area that impacts the implementation of SDG 7. It is defined as the ratio of GDP to energy consumption. Energy productivity is a critical measure for evaluating the efficiency with which countries use energy to generate economic output [34]. The EU has been actively working to improve energy productivity in its member states through various policies and directives; however, there are significant discrepancies in energy efficiency between EU countries [35]. Generally, older member states show higher efficiency compared to newer ones. This means the newer member states have more space for improvement in this aspect [36]. From the point of view of the assessment of SDG 7, the higher the level of energy productivity, the better. This indicator directly addresses energy efficiency, i.e., the 7.3 target of SDG 7.

The share of renewable energy in the final gross energy consumption by sector is similar in nature. The EU has set ambitious targets to reduce greenhouse gas emissions and increase the share of renewable energy. It has established a target for 2030, which states that the share of renewable energy in gross energy consumption must be 32% [37,38]. However, the revised Directive on Renewable Energy EU/2023/2413 raised the EU’s binding renewable target for 2030 to a minimum of 42.5%, up from the previous target of 32%, to achieve 45% [38]. Renewable energy share varies significantly between sectors and countries within the EU. Furthermore, a positive correlation exists between GDP per capita and the share of renewable energy, indicating that more developed countries tend to consume more renewable energy [36]. This indicator directly supports target 7.2 on increasing the share of renewables in the energy mix within SDG 7.

Increasing the values of the variables that measure the following two areas of SDG 7 has a negative impact on its assessment. They are the dependency of energy imports by products and the population being unable to keep their homes adequately warm due to poverty status. The dependence on energy imports varies significantly by product type, and understanding these dependencies is crucial for energy security and policy making. The EU is highly dependent on energy imports, with a primary energy import dependence of approximately 53% expected to reach or exceed 70% in the next 20–30 years [39]. This dependency includes various energy products, such as oil, natural gas, and coal. Increasing the share of renewable energy can reduce dependence on energy imports. For example, a share of renewable energy at 32% can stabilise import dependency, and higher shares can further reduce it [40]. This indicator is intended to support measuring the ‘reliability’ and ‘accessibility’ of the energy services reflected in target 7.1 through the concept of energy security [27].

The last area examined is the population unable to keep their homes adequately warm due to poverty status. It is a critical indicator of energy poverty. Low income, high energy costs, and energy-inefficient homes influence energy poverty. Collectively, these factors contribute to the inability to maintain adequate warmth in homes [41,42]. Addressing these factors through comprehensive policies can help reduce the number of households unable to keep their homes adequately warm, thus improving overall health and well-being [43]. This indicator is intended to cover the ‘affordable’ nature of energy in target 7.1 of the SGD 7. However, this indicator is also dependent on climate zones (e.g., areas where needs would favour keeping the home cold in summer time) [27].

On the one hand, Eurostat measures a limited set of data to evaluate the level of implementation of SDG 7. On the other hand, these relatively long time series data are available for all EU countries.

Given the incompleteness of the existing indicators used to measure SDG 7, it is necessary to revise them. Researchers suggest modifying indicators to better represent reality, such as the level of electrification, energy prices, and the share of renewable and nuclear energy [44]. Similarly, ref. [45] emphasise that the SDG 7 index, which is constructed using the UN’s indicators [46], does not adequately capture energy sustainability. This is because it overlooks critical facets, such as emissions, clean energy use, and energy security, which are integral to the fundamental aims of SDG 7. The results presented confirm that more relevant SDG 7 indicators are needed.

3. Materials and Methods

3.1. Materials

We conduct a comparative study of the 27 EU countries focusing on their progress in SDG 7, utilising the variables outlined by Eurostat [16]. As of now, the seven indicators established in 2017 to track advancements in SDG 7 remain unchanged [19]. They are systematically monitored and evaluated by Eurostat. These indicators are as follows:

X1—Primary energy consumption—The indicator measures the total energy needs of a country excluding all non-energy use of energy carriers. It is measured in million tons of oil equivalent (TOE) per capita.

X2—Final energy consumption—The indicator measures the end-use of energy in a country, excluding all non-energy use of energy carriers. It is measured in million tons of oil equivalent (TOE) per capita.

X3—Final energy consumption in households per capita—The indicator measures how much energy every citizen consumes at home, excluding the energy used for transportation. It is measured in kg of oil equivalent.

X4—Energy productivity—The indicator measures the amount of economic output produced per unit of gross available energy. Its unit of measure is chain-linked volumes (2010) in EUR per kg of oil equivalent and Purchasing Power Standards (PPS) per kg of oil equivalent.

X5—Share of renewable energy in gross final energy consumption by sector—The indicator measures the share of renewable energy consumption in gross final energy consumption according to the Renewable Energy Directive. Its unit of measure is % by sectors.

X6—Energy import dependency by products—The indicator shows the share of total energy needs of a country met by imports from other countries. It is measured in % of imports in total energy consumption.

X7—Population unable to keep home adequately warm by poverty status—The indicator measures the share of population who are unable to afford to keep home adequately warm. The unit of measure is % of the population.

The variables listed are stimulants (X4, X5), destimulants (X6, X7) and nominants (X1–X3). Stimulants are variables that, when their values increase, have a positive impact on the phenomenon under investigation. Destimulants are variables whereby increasing values yield adverse effects [47]. Nominants, on the other hand, are variables whose growing values can affect both favorably and unfavorably on the phenomenon under study. Therefore, the optimal value of a variable is the nominal value, while any deviation from this value negatively affects the evaluated phenomenon [48]. Since nominal values are generally not known, researchers either avoid the use of nominants or classify them into stimulants or destimulants.

In this study, the second approach was used, i.e., the nominants were classified into destimulants. This was justified by the fact that the study covered 27 EU countries, that is, countries that are mostly highly developed and of the relatively high level of energy consumption. Further increases in energy consumption were considered to be negative, adverse effects for the assessment of SDG 7 [19].

3.2. Methods

The MCA is employed in empirical research concerning complex phenomena. Inherent in this concept is the issue of diagnostic variables. The most essential requirements for diagnostic variables [49] are as follows: (a) they must play a crucial role in detailing the phenomenon being examined, (b) they must be accessible, (c) if feasible, they ought to be assessed using strength scales, (d) they ought to have a minimal correlation with one another to avoid repeating the information that other variables already contain.

Because diagnostic variables should effectively discriminate classified objects, we first assess their variability using the coefficient of variation (CV). Among the possible diagnostic features, we eliminate those with a CV derived from the equation:

$$v_j={\displaystyle \frac{{SD}_j}{\left|{\overline{x}}_j\right|}}(j=\mathrm{1,2},\dots ,n),$$

(1)

where:

${SD}_j$—standard deviation of j-th variable

$\left|{\overline{x}}_j\right|$—absolute value of the mean of j-th variable is less than 10%.

The next stage of statistical verification of variables is the elimination of highly correlated variables, that is, variables that carry similar information about the objects under study. The study uses Pearson’s inverse linear correlation coefficient matrix method to assess the degree of correlation between variables. We first determine the correlation matrix (R) and then the inverse matrix (R⁻¹). If the variables do not show many close interdependencies, the diagonal elements of the R⁻¹ matrix are variance inflation factors (VIF) of the variables and are given by equation [50]:

$$r_{jj}^{(-1)}={VIF}_j={\displaystyle \frac{1}{1-R_j^2}},$$

(2)

where $R_j^2$—the regression coefficient of determination of the j-th variable against the others variables, (j = 1, 2, …, p), p—the number of variables in the model after variation verification. If $X_j$ variable is orthogonal to the other variables, then the diagonal element $r_{jj}^{(-1)}$ of R⁻¹ matrix equals 1. In the case of non-orthogonality $r_{jj}^{(-1)}\mathsf{ϵ}$ (1, +∞).

When a variable is highly correlated with the other variables, the inverse matrix (R⁻¹) diagonal elements are significantly higher than 1—e.g., higher than 10. This means poor numerical conditioning of the R⁻¹ matrix, that is, excessive correlation of a given variable with the other variables. If there is only one such variable in the data set, it is eliminated. If there are more variables with this feature, it would be possible to eliminate all the variables, but this would generally lead to an excessive reduction in the available information resource. It is usually sufficient to eliminate some variables so that the diagonal elements of the inverse correlation matrix of the remaining variables are sufficiently low. The critical value of the diagonal elements of the inverse correlation matrix was assumed to be 10.

The article uses the ZUM to compare multiple objects using selected criteria. These criteria can be expressed in different quantities. The ZUM aims to standardise the criteria being evaluated. The normalisation of the variables is performed based on formulas [51]:

For stimulants

$$z_{ij}={\displaystyle \frac{x_{ij}-\underset{i}{\min }\left\{x_{ij}\right\}}{R_j}},$$

(3)

For destimulants

$$z_{ij}={\displaystyle \frac{\underset{i}{\max }\left\{x_{ij}\right\}{-x}_{ij}}{R_j}},$$

(4)

where $R_j$ is the range calculated from the equation:

$$R_j=\underset{i}{\max }\left\{x_{ij}\right\}-\underset{i}{\min }\left\{x_{ij}\right\}.$$

(5)

The next step in constructing the SM is to assign weights to the normalised variables. This procedure, however, sparks a vast array of discussions. In most empirical work in which the SM is constructed, an assumption is made about the equal weights of all selected diagnostic variables e.g., [11,25,51,52]. In any scenario, assigning equal weights does not mean the absence of weights; rather, it implicitly suggests that all weights are the same [53]. This is mainly the result of the lack of information on the circumstances that affect the differential importance and role of diagnostic variables [49]. Furthermore, the introduction of weights very rarely results in differential results [54]. With this in mind, we assume equal weights for the diagnostic variables in this study.

The SM according to the ZUM method is calculated from the equation:

$${ZUM}_i={\displaystyle \frac{1}{m}}{\sum }_{j=1}^m{z}_{ij},$$

(6)

where: ${ZUM}_i$—SM for the i-th object, $z_{ij}$—normalised values of the variables. The measure takes values between 0 and 1. The closer the measured value is to 1, the more favorable the condition of the evaluated object regarding the examined phenomenon.

The subsequent phase of the statistical analysis involves grouping objects into categories. Objects are categorised in the following manner:

$$\mathrm{Group}1:{SM}_i\ge \overline{{SM}_i}+{SD}_i\hspace{1em}\mathrm{high} \mathrm{level}$$

(7)

$$\mathrm{Group}2:\overline{{SM}_i}+{SD}_i>{SM}_i\ge \overline{{SM}_i}\hspace{1em}\mathrm{medium-high} \mathrm{level}$$

(8)

$$\mathrm{Group}3:\overline{{SM}_i}>{SM}_i\ge \overline{{SM}_i}-{SD}_i\hspace{1em}\mathrm{medium-low} \mathrm{level}$$

(9)

$$\mathrm{Group}4:{SM}_i<\overline{{SM}_i}-{SD}_i\hspace{1em}\mathrm{low} \mathrm{level}$$

(10)

where: $\overline{{SM}_i}—$mean value of the SM, ${SD}_i$—standard deviation of the SM.

The last step in the study’s empirical analysis is an uncertainty and sensitivity analysis, as indicated in [52]. This task should be carried out to evaluate the strength of the composite indicator. Considerations include the process for adding or removing indicators, the normalisation strategy, how missing data is handled, the selection of weights, and the method of aggregation [53].

The study used sensitivity analysis to diagnose the necessity of including or excluding the examined indicators. It was carried out according to the following algorithm: for each country in the EU and each year, the SM values were calculated, each time eliminating one of the explanatory (independent) variables. The absolute value (modulus of the number) was then calculated from the difference between the obtained value and the initially calculated value of the SM for all independent variables. The obtained result was divided by the initial value of the SM and expressed as a percentage.

4. Results

4.1. Selection of Variables

The initial stage in choosing diagnostic variables (X1–X7) was to evaluate their variability (

Table 1. Evaluation of the variability of diagnostic variables.

2015
Statistics	X1	X2	X3	X4	X5	X6	X7
SD	1.23	1.15	180.34	3.23	11.58	24.88	10.25
Mean	3.11	2.31	560.15	6.95	20.35	56.71	11.22
CV	0.40	0.50	0.32	0.46	0.57	0.44	0.91
2023
Statistics	X1	X2	X3	X4	X5	X6	X7
SD	0.94	0.82	164.45	4.80	12.91	22.18	5.98
Mean	2.83	2.16	523.93	9.01	27.13	57.18	9.47
CV	0.33	0.38	0.31	0.53	0.48	0.39	0.63

SD—standard deviation, CV—coefficient of variation.

). For this purpose, we applied the CV. The variation in the variables under study reached values ranging from 0.32 for variable X3 to 0.91 for variable X7 in 2015. However, the variation in variables in 2023 ranged from 0.31 for X3 to 0.63 for X7. Based on the criterion of adequate variability, all diagnostic variables were retained without any being eliminated from the set.

At the subsequent phase, the degree of correlation among the variables was assessed (

Table 2. Correlation matrix (2015, 2023).

2015	X1	X2	X3	X4	X5	X6	X7
X1	1.0000	0.9474	0.7805	0.2997	0.0187	0.0542	−0.5644
X2	―	1.0000	0.7567	0.4043	−0.0074	0.2025	−0.5055
X3	―	―	1.0000	0.3932	0.2850	−0.2213	−0.6963
X4	―	―	―	1.0000	−0.0651	0.2610	−0.2906
X5	―	―	―	―	1.0000	−0.4931	−0.0965
X6	―	―	―	―	―	1.0000	0.2847
X7	―	―	―	―	―	―	1.0000
2023	X1	X2	X3	X4	X5	X6	X7
X1	1.0000	0.9245	0.7329	0.1461	0.2151	−0.1156	−0.5151
X2	―	1.0000	0.6751	0.3188	0.1571	0.0723	−0.5184
X3	―	―	1.0000	0.1178	0.4869	−0.4552	−0.6666
X4	―	―	―	1.0000	−0.0475	0.3091	−0.1245
X5	―	―	―	―	1.0000	−0.6303	−0.0935
X6	―	―	―	―	―	1.0000	0.2186
X7	―	―	―	―	―	―	1.0000

). The values of Pearson’s correlation coefficients should confirm the relationships between the SDG 7 areas presented in the theoretical part of the article. Stimulants should generally be negatively correlated with destimulants and positively correlated with each other. However, the presented results of the correlation analysis (for 2015 and 2023) show that these assumptions were not often met (e.g., correlations between X1 and X7, X2 and X7, X3, and X7). This was likely related to the ambiguous effect of the variables X1–X3 on the assessment of SDG 7.

Very strong correlations were observed between variables X1 and X2. This was valid for 2015 and 2023. These correlations were confirmed by the inverse correlation matrix (

Table 3. Inverse matrix (2015, 2023).

2015	X1	X2	X3	X4	X5	X6	X7
X1	13.8172	−12.7234	−1.1791	1.3061	0.8556	1.4809	0.5858
X2	―	15.3239	−1.9750	−1.0385	−0.5401	−2.7362	−0.3844
X3	―	―	5.0900	−0.7959	−0.8782	1.0005	1.2793
X4	―	―	―	1.5401	0.0897	−0.4645	0.2465
X5	―	―	―	―	1.5576	0.7371	−0.4351
X6	―	―	―	―	―	2.3457	−0.5825
X7	―	―	―	―	―	―	2.2225
2023	X1	X2	X3	X4	X5	X6	X7
X1	10.7544	−10.0048	−1.5427	1.2066	1.1022	1.7623	−0.8076
X2	―	12.1289	−0.8375	−1.3306	−1.1835	−2.9558	0.9459
X3	―	―	4.8228	−0.2917	−1.0004	1.1881	1.5968
X4	―	―	―	1.3923	−0.0797	−0.3744	−0.0150
X5	―	―	―	―	2.2078	1.3511	−0.8115
X6	―	―	―	―	―	3.0155	−0.4118
X7	―	―	―	―	―	―	2.1511

).

Elements on the main diagonal of the inverse matrices calculated for 2015 and 2023 exceeded the critical value of 10, which was the case for variables X1 and X2 (Table 3). The highest value was obtained for the variable X2 (2015—15.32 and 2023—12.13, respectively). This determined the removal of variable X2 from the set of diagnostic variables.

After removing the variable X2, the correlation and inverse matrices looked as in

Table A1. Correlation matrix after excluding X2 variable (2015, 2023).

2015	X1	X3	X4	X5	X6	X7
X1	1.0000	0.7805	0.2997	0.0187	0.0542	−0.5644
X3	―	1.0000	0.3932	0.2850	−0.2213	−0.6963
X4	―	―	1.0000	−0.0651	0.2610	−0.2906
X5	―	―	―	1.0000	−0.4931	−0.0965
X6	―	―	―	―	1.0000	0.2847
X7	―	―	―	―	―	1.0000
2023	X1	X3	X4	X5	X6	X7
X1	1.0000	0.7329	0.1461	0.2151	−0.1156	−0.5151
X3	―	1.0000	0.1178	0.4869	−0.4552	−0.6666
X4	―	―	1.0000	−0.0475	0.3091	−0.1245
X5	―	―	―	1.0000	−0.6303	−0.0935
X6	―	―	―	―	1.0000	0.2186
X7	―	―	―	―	―	1.0000

and

Table A2. Inverse matrix after excluding X2 variable (2015, 2023).

2015	X1	X3	X4	X5	X6	X7
X1	3.2530	−2.8190	0.4438	0.4072	−0.7909	0.2666
X3	―	4.8355	−0.9297	−0.9478	0.6479	1.2298
X4	―	―	1.4697	0.0531	−0.6499	0.2205
X5	―	―	―	1.5386	0.6407	−0.4487
X6	―	―	―	―	1.8571	−0.6511
X7	―	―	―	―	―	2.2129
2023	X1	X3	X4	X5	X6	X7
X1	2.5017	−2.2336	0.1090	0.1260	−0.6758	−0.0274
X3	―	4.7650	−0.3836	−1.0821	0.9841	1.6621
X4	―	―	1.2463	−0.2095	−0.6987	0.0887
X5	―	―	―	2.0923	1.0627	−0.7192
X6	―	―	―	―	2.2951	−0.1813
X7	―	―	―	―	―	2.0774

(Appendix A).

The employed variable selection method resulted in choosing variables X1 and X3–X7 as the definitive set of diagnostic variables.

4.2. Descriptive Statistics

We start by presenting the variables from their descriptive statistics (

Table 4. Descriptive statistics of diagnostic variables.

2015
Indicator	Character of Indicator	Maximum	Minimum	Mean	CV	CA
X1	D	7.27 (Luxembourg)	1.55 (Romania)	3.11	39.57%	1.66
X3	D	904.00 (Finland)	180.00 (Malta)	560.15	32.19%	−0.08
X4	S	16.48 (Ireland)	2.18 (Bulgaria)	6.95	46.41%	1.29
X5	S	52.22 (Sweden)	4.99 (Luxembourg)	20.35	56.91%	0.85
X6	D	97.32 (Cyprus)	11.18 (Estonia)	56.71	43.88%	−0.07
X7	D	39.20 (Bulgaria)	0.90 (Luxembourg)	11.22	91.30%	1.31
2023
Indicator	Character of indicator	Maximum	Minimum	Mean	CV	CA
X1	D	5.61 (Finland)	1.57 (Romania)	2.83	33.16%	1.71
X3	D	991.00 (Finland)	206.00 (Malta)	523.93	31.39%	1.98
X4	S	26.16 (Ireland)	2.94 (Bulgaria)	9.01	53.24%	1.35
X5	S	66.39 (Sweden)	11.62 (Luxemburg)	27.13	47.59%	−0.30
X6	D	97.55 (Malta)	3.47 (Estonia)	57.18	38.79%	0.96
X7	D	20.80 (Bulgaria)	2.10 (Luxemburg)	9.47	63.15%	1.98

CV—coefficient of variation, CA—coefficient of asymmetry.

). The destimulant variable X1 took the highest values for Luxembourg in 2015 and Finland in 2023. This means that at this moment in time, primary energy consumption was the highest among the 27 member states. The lowest primary energy consumption was recorded in Romania. The member states were strongly differentiated in terms of this variable, with most countries achieving values below the average. This was valid for 2015, and 2023. The final consumption of energy per capita (X3) in 2015, and in 2023 was highest in Finland and lowest in Malta. EU countries were rather poorly differentiated in terms of this variable. The variable values were distributed above (in 2015) and below the mean (in 2023). The variable X4, which was a stimulant, had the highest values in 2015 and 2023 for Ireland. The lowest values in 2015, and 2023 were recorded for Bulgaria. The level of differentiation among the 27 EU countries in terms of energy productivity was at a comparable level in the years examined. The EU countries were quite differentiated in terms of this variable. Most of the values for measured energy productivity were below average. The highest share of renewable energy in gross final energy consumption by sector (X5) was recorded in 2015 and 2023 in Sweden. The X5 variable had the lowest values in 2015 and 2023 for Luxembourg.

Based on the values of the CV, it can be concluded that the disparities between EU countries in terms of this variable are decreasing (2015 compared to 2023). On the other hand, the asymmetry between the values of this variable is increasing: i.e., the number of countries with below-average values has increased over time. The highest dependence on energy imports by products (X6) was recorded in 2015 in Cyprus and 2023 in Malta. In contrast, Estonia, which is one of the most energy-independent countries in the EU, took the highest values for this variable in 2015 and 2023. The negative asymmetry observed in 2015 meant that more values of this variable were above the EU average. The positive asymmetry observed in 2023 meant that more values of this variable were below the EU average. X6 and X7 variables were destimulants. In the examined years, the X7 variable was highest for Bulgaria. This means that among the EU countries, Bulgaria had the highest percentage of the population unable to keep their homes adequately warm due to poverty status. The best situation in this regard prevailed both in 2015 and 2023 in Luxembourg. The variability of this variable, measured by the CV, dropped significantly in 2023 compared to 2015. The positive asymmetry for X7 showed that most values of this variable were below the EU average.

4.3. SM Values and Ranking of the 27 EU Countries

The SM values determined according to the ZUM for the countries studied are summarised in

Table 5. SM values in the EU 27 countries (2015–2023).

Country	2015	2016	2017	2018	2019	2020	2021	2022	2023	2023 vs. 2015 Change
Austria	0.5498	0.5578	0.5565	0.5622	0.5369	0.5276	0.5101	0.4711	0.5179	−0.0319
Belgium	0.3679	0.3933	0.4008	0.3820	0.4032	0.3811	0.3772	0.3786	0.3837	0.0158
Bulgaria	0.4374	0.4304	0.4256	0.4274	0.4313	0.4138	0.3898	0.3872	0.4052	−0.0322
Croatia	0.5724	0.5824	0.5779	0.5674	0.5650	0.5496	0.5302	0.5004	0.5092	−0.0632
Cyprus	0.3753	0.3977	0.4008	0.4025	0.3916	0.3739	0.3803	0.3556	0.3699	−0.0054
Czechia	0.4963	0.5012	0.4983	0.4915	0.4951	0.4735	0.4489	0.4468	0.4491	−0.0472
Denmark	0.7032	0.7207	0.7333	0.6977	0.6808	0.6204	0.6560	0.6284	0.6303	−0.0729
Estonia	0.5922	0.5534	0.5691	0.5749	0.5992	0.5629	0.5993	0.5619	0.5900	−0.0021
Finland	0.4667	0.4627	0.4674	0.4529	0.4635	0.4401	0.4256	0.4194	0.4244	−0.0423
France	0.5242	0.5329	0.5373	0.5289	0.5248	0.5128	0.4900	0.4610	0.4591	−0.0650
Germany	0.4933	0.5047	0.5127	0.5097	0.5038	0.4526	0.4731	0.4351	0.4515	−0.0418
Greece	0.4507	0.4584	0.4723	0.4774	0.4883	0.4566	0.4431	0.4113	0.4239	−0.0267
Hungary	0.4825	0.4846	0.4800	0.4811	0.4670	0.4630	0.4467	0.4461	0.4489	−0.0335
Ireland	0.5237	0.5806	0.6139	0.5928	0.5971	0.5832	0.5744	0.5508	0.5646	0.0408
Italy	0.4937	0.5142	0.5125	0.4984	0.5067	0.5025	0.4707	0.4638	0.4768	−0.0169
Latvia	0.5674	0.5926	0.6021	0.5924	0.5907	0.5844	0.5826	0.5751	0.5834	0.0160
Lithuania	0.4346	0.4409	0.4411	0.4055	0.4002	0.3860	0.3581	0.4016	0.3800	−0.0547
Luxembourg	0.2778	0.3072	0.3099	0.3059	0.3105	0.2897	0.3057	0.3271	0.3362	0.0584
Malta	0.4595	0.4999	0.4947	0.4842	0.4791	0.4748	0.4709	0.4647	0.4767	0.0172
The Netherlands	0.4986	0.5106	0.5093	0.4878	0.4883	0.4741	0.4735	0.4415	0.4671	−0.0315
Poland	0.5541	0.5466	0.5356	0.5121	0.5297	0.5141	0.4962	0.4836	0.4874	−0.0667
Portugal	0.5530	0.5728	0.5611	0.5473	0.5431	0.5481	0.5278	0.5028	0.4698	−0.0832
Romania	0.6569	0.6453	0.6434	0.6319	0.6238	0.5957	0.5538	0.5275	0.5579	−0.0990
Slovakia	0.5037	0.5056	0.4971	0.4933	0.4782	0.4835	0.4665	0.4339	0.4400	−0.0637
Slovenia	0.5416	0.5449	0.5492	0.5447	0.5582	0.5497	0.5402	0.5284	0.5368	−0.0049
Spain	0.5529	0.5750	0.5719	0.5447	0.5578	0.5357	0.4815	0.4475	0.4290	−0.1239
Sweden	0.6510	0.6428	0.6589	0.6469	0.6598	0.6435	0.6529	0.6374	0.6144	−0.0366

.

In 2015, we observed the lowest SM for Luxembourg (0.28) and the highest for Denmark (0.70). The difference between the highest and the lowest value was 0.43 in 2015. In turn, the 2023 range resulted in 0.29, with the highest and lowest SM values recorded for the same countries as in 2015. High SM variability between years was observed for most countries, e.g., Spain, Romania, Lithuania, Denmark, Germany, France or Portugal. Low variability was observed in Estonia, Belgium, Malta, Sweden, or Slovenia. The use of ZUM results in higher variability of the SM values than with other ranking methods, e.g., TOPSIS (Technique for Order Preference by Similarity to Ideal Solution).

Analysing changes in the value of the SM over time, it should be noted that there was a steady trend for several EU countries (e.g., Belgium, Estonia, Slovakia, Sweden). An increasing trend was observed for Luxembourg and decreasing for, e.g., Finland, Lithuania, Poland, Romania, Slovakia or Spain.

By comparing the SM for the countries studied in 2015 and 2023, the countries with the largest increase/decrease in the SM were highlighted (Table 5). The countries with increases in the value of the SM turned out to be Luxembourg, Ireland, Malta, Latvia, Belgium. The country with the largest decrease in the value of the SM when comparing 2023 with 2015 was Spain (−0.1239). Other countries with decreases in the value of the SM were, among others, Romania, Portugal, Denmark, and Poland. Values oscillating around zero were recorded for Estonia, Slovenia, and Cyprus.

Based on the SM values calculated for the 27 EU countries (Table 5) the ranking of EU countries was made for 2015–2023 (

Table 6. Ranking of the 27 EU countries in terms of achievement of the SDG 7 (2015–2023).

Country	2015	2016	2017	2018	2019	2020	2021	2022	2023
Austria	10	9	10	8	11	11	10	11	8
Belgium	26	26	25	26	24	25	25	25	24
Bulgaria	23	24	24	23	23	23	23	24	23
Croatia	5	5	6	7	7	8	8	9	9
Cyprus	25	25	26	25	26	26	24	26	26
Czechia	16	18	17	17	16	18	19	16	17
Denmark	1	1	1	1	1	2	1	2	1
Estonia	4	10	8	6	4	6	3	4	3
Finland	20	21	22	22	22	22	22	21	21
France	12	13	12	12	13	13	12	14	15
Germany	18	17	14	14	15	21	15	19	16
Greece	22	22	21	21	18	20	21	22	22
Hungary	19	20	20	20	21	19	20	17	18
Ireland	13	6	4	4	5	5	5	5	5
Italy	17	14	15	15	14	14	17	13	11
Latvia	6	4	5	5	6	4	4	3	4
Lithuania	24	23	23	24	25	24	26	23	25
Luxembourg	27	27	27	27	27	27	27	27	27
Malta	21	19	19	19	19	16	16	12	12
The Netherlands	15	15	16	18	17	17	14	18	14
Poland	7	11	13	13	12	12	11	10	10
Portugal	8	8	9	9	10	9	9	8	13
Romania	2	2	3	3	3	3	6	7	6
Slovakia	14	16	18	16	20	15	18	20	19
Slovenia	11	12	11	11	8	7	7	6	7
Spain	9	7	7	10	9	10	13	15	20
Sweden	3	3	2	2	2	1	2	1	2

).

The first in the rankings for 2015 and 2023 was Denmark; however, in 2020 and 2022 it took second place. Another Scandinavian country that ranked highly was Sweden. This country was third for the first two years of the analysis (2015–2016). It alternated with Denmark in second or first place for the next years. Much lower in the rankings was another Scandinavian country, namely Finland. It started the ranking from the 20th position in 2015, and finished it on the 21st place.

The highest position among the so-called ‘new’ member states was occupied by Estonia. In 2021 and 2023, Estonia was third. Its lowest ranking position was in 2016 when it ranked tenth. Latvia, similar to Estonia also ranked high. It moved from sixth position in 2015 to fourth in 2023. Latvia’s neighbour, Lithuania, performed much worse in the rankings. It dropped from the twenty-fourth position in 2015 to the twenty-fifth in 2023.

Relatively high in the ranking made for 2015–2023 was one of the poorest countries in the EU, Romania. It dropped from a high second place held in 2015 to sixth in 2023. The worst in the ranking were Poland, Slovenia, Slovakia, Czechia, and Hungary, which are better developed economically than Romania and also represent the so-called ‘new’ EU member states.

4.4. Classification of the 27 EU Countries

Based on the SM (Table 5), the EU member states were classified into four groups of countries similar to each other in terms of the level of SDG 7 (

Table 7. Classification of the 27 EU countries in terms of the achievement of SDG 7 (2015, 2023).

Group No.	2015	2023
High level of SDG 7: Group 1	Denmark, Romania, Sweden	Denmark, Sweden, Estonia, Latvia, Ireland, Romania, Slovenia
Medium-high level of SDG 7: Group 2	Estonia, Croatia, Latvia, Poland, Portugal, Spain, Austria, Slovenia, France, Ireland	Austria, Croatia, Poland
Medium-low level of SDG 7: Group 3	Slovakia, The Netherlands, Czechia, Italy, Germany, Hungary, Finland, Malta, Greece, Bulgaria, Lithuania	Italy, Malta, Portugal, The Netherlands, France, Germany, Czechia, Hungary, Slovakia, Spain, Finland, Greece, Bulgaria
Low level of SDG 7: Group 4	Cyprus, Belgium, Luxembourg	Belgium, Lithuania, Cyprus, Luxembourg

). The classification was made for 2015 and 2023.

A detailed characterisation of Group 1 was based on data presented in

Table 8. High level of advancement of SDG 7: Group 1 (2015, 2023).

2015
Country	X1	X3	X4	X5	X6	X7
Denmark	2.96	782	14.34	30.47	13.08	3.6
Romania	1.55	372	4.61	24.79	16.69	13.1
Sweden	4.47	756	8.46	52.22	30.07	1.2
Mean	2.99	636.67	9.14	35.83	19.95	5.97
SD	1.19	187.45	4.00	11.82	7.31	5.14
CV	0.40	0.29	0.44	0.33	0.37	0.86
2023
Country	X1	X3	X4	X5	X6	X7
Denmark	2.58	698	18.81	44.92	38.87	6.9
Sweden	3.93	669	10.07	66.39	26.39	5.9
Estonia	3.00	683	4.64	40.95	3.47	4.1
Latvia	2.27	574	5.66	43.22	32.73	6.6
Ireland	2.65	479	26.16	15.25	77.9	7.2
Romania	1.57	396	6.35	25.76	27.86	12.5
Slovenia	2.79	490	7.72	25.07	49.27	3.6
Mean	2.68	569.86	11.34	37.37	36.64	6.69
SD	0.62	102.42	7.00	14.67	19.89	2.52
CV	0.23	0.18	0.62	0.39	0.54	0.38

.

In 2023, the average final energy consumption (X1) in Group 1 decreased from 2.99 to 2.68 TOE per capita compared to 2015. The year 2023 also brought a decrease in the average value of final energy consumption in households per capita (X3). Its lowest level was recorded for Romania and the highest for Denmark (in 2023). The year 2023 saw improvements in energy productivity (X4) in this group. There was also a slight increase in the average share of renewable energy in gross final energy consumption by sector (X5) in 2023 compared to 2015. The highest share in 2023 was recorded in Sweden, significantly above average, and the lowest in Ireland. The energy import dependency by products (X6) increased about twofold in 2023 compared to 2015 in Group 1. The highest import dependency was recorded in 2015 in Sweden and in 2023 in Ireland. The least dependent on energy imports was Estonia in 2023 (3.47%). The year 2023 revealed an increase in the share of those unable to keep their homes adequately warm (X7). The lowest percentage was recorded in 2015 in Sweden (1.2%) and the highest in Romania (13.1%). In 2023, the most affected by being unable to keep home adequately warm were citizens of Romania and the least were citizens of Slovenia. In Group 1, the highest level of variability in 2015 was recorded for the variable X7 (0.86), with the highest value of X7 for Romania and the lowest for Sweden. In 2023, the CV was highest for the variable X4, with the highest X4 value recorded for Ireland and the lowest for Estonia. In both 2015 and 2023, countries classified into Group 1 were the least differentiated in terms of the variable X3. The variability measured by the CV for this variable was 0.29 and 0.18 in 2015 and 2023, respectively.

Group 2—with a medium-high level of achievement of SDG 7—contained in 2015, ten countries, and in 2023, only three countries (

Table 9. Medium-high level of advancement of SDG 7: Group 2 (2015, 2023).

2015
Country	X1	X3	X4	X5	X6	X7
Estonia	3.62	652	3.43	28.99	11.18	2.0
Croatia	1.92	585	5.36	28.97	48.79	9.9
Latvia	2.16	559	4.54	37.54	51.18	14.5
Poland	2.37	501	4.39	11.88	29.85	7.5
Portugal	2.09	265	7.09	30.51	76.29	23.8
Spain	2.55	329	8.30	16.22	72.74	10.6
Austria	3.66	767	9.19	33.50	60.38	2.6
Slovenia	3.08	568	5.60	22.88	49.26	5.6
France	3.67	599	8.04	14.80	45.89	5.5
Ireland	2.96	605	16.48	9.08	88.65	9.0
Mean	2.81	543.00	7.24	23.44	53.42	9.10
SD	0.65	140.19	3.57	9.37	21.51	6.07
CV	0.23	0.26	0.49	0.40	0.40	0.67
2023
Country	X1	X3	X4	X5	X6	X7
Austria	3.21	721	10.93	40.84	61.05	3.9
Croatia	2.21	577	6.68	28.05	55.72	6.2
Poland	2.55	542	5.69	16.50	48.02	4.7
Mean	2.66	613.33	7.77	28.46	54.93	4.93
SD	0.42	77.46	2.27	9.94	5.35	0.95
CV	0.16	0.13	0.29	0.35	0.10	0.19

). Average primary energy consumption (X1) decreased slightly in 2023 compared to 2015. However, on average, the final energy consumption in households per capita (X3) increased in 2023 compared to 2015. Austria’s highest final energy consumption per capita in household was recorded for both examined years. The year 2023 brought an increase in average energy productivity (X4) in this group of countries. The highest energy productivity was recorded in 2015 in Ireland, and the lowest in Estonia. In 2023, the variable X4 took the highest value for Austria and the lowest for Poland. The average share of renewable energy increased in 2023 compared to 2015. The leaders in this variable (X5) in 2015 were Latvia and Austria. The average dependence on energy imports (X6) increased slightly in 2023 compared to 2015. The countries most dependent on energy imports were Ireland and Austria in 2015 and 2023, respectively. The least energy import-dependent country in 2015 was Estonia, and 2023, Poland. The largest share of the population unable to keep their homes adequately warm (X7) in 2015 was in Portugal, and 2023, Croatia. The best performing country in this regard was Estonia in 2015 and Austria in 2023.

In 2015, Group 2 countries were the most differentiated, similar to Group 1 countries, due to variable X7 (0.67). Its highest value was recorded for Portugal, and the lowest for Estonia. In 2023, the level of differentiation of countries classified into Group 2 was highest for variable X5, with the highest value of this variable recorded for Austria and the lowest for Poland. The lowest level of the CV for countries classified into the medium-high level groups in 2015 and 2023 was recorded for variables X1 and X6, respectively.

In Group 3 in 2015 and in 2023, the highest value of the X1 variable, primary energy consumption (TOE per capita), was in Finland (

Table 10. Medium-low level of advancement of SDG 7: Group 3 (2015, 2023).

2015
Country	X1	X3	X4	X5	X6	X7
Slovakia	2.91	474	4.62	12.882	58.013	5.8
The Netherlands	3.78	563	7.6	5.714	48.368	2.9
Czechia	3.74	642	4.11	15.07	32.089	5
Italy	2.48	540	9.92	17.525	77.03	17
Germany	3.62	673	8.87	14.901	62.132	4.1
Hungary	2.38	609	4.38	14.495	53.875	9.6
Finland	5.69	904	5.7	39.23	47.901	1.7
Malta	1.69	180	3.94	5.119	97.296	14.1
Greece	2.16	412	7.02	15.69	71.047	29.2
Bulgaria	2.57	314	2.18	18.261	36.446	39.2
Lithuania	1.99	467	4.7	25.748	75.452	31.1
Mean	3.00	525.27	5.73	16.79	59.97	14.52
SD	1.09	183.46	2.24	8.91	18.35	12.45
CV	0.36	0.35	0.39	0.53	0.31	0.86
2023
Country	X2	X3	X4	X5	X6	X7
Italy	2.29	468	11.8	19.56	74.81	9.5
Malta	1.65	206	4.52	15.08	97.55	6.8
Portugal	1.96	285	8.98	35.16	66.87	20.8
The Netherlands	3.01	431	10.46	17.15	70.45	7.1
France	3.07	526	10.23	22.28	44.87	12.1
Germany	2.85	633	11.84	21.55	66.38	8.2
Czechia	3.27	591	5.23	18.59	41.68	6.1
Hungary	2.31	560	5.69	17.36	62.06	7.2
Slovakia	2.85	483	5.48	16.99	57.73	8.1
Spain	2.28	287	9.94	24.85	68.42	20.8
Finland	5.61	991	6.1	50.75	29.57	2.6
Greece	1.91	362	8.94	25.27	75.60	19.2
Bulgaria	2.57	317	2.94	22.58	39.72	20.7
Mean	2.74	472.31	7.86	23.63	61.21	11.48
SD	0.95	195.72	2.86	9.28	17.61	6.28
CV	0.35	0.41	0.36	0.39	0.29	0.55

). The lowest value of the X1 variable in the examined years was observed in Malta. The final energy consumption in households per capita (X3) was the highest in Finland in 2015 and 2023, and the lowest in both years examined in Malta. Among the countries in the third group, the highest energy productivity (X4) was in Italy in 2015 and, in Germany in 2023. The highest value of the X5 variable was in Lithuania in 2015, and the lowest was in Malta. In 2023, this variable’s highest and lowest values were in Finland and Malta, respectively. The highest energy import dependency (X6) was observed in 2015 and in 2023 in Malta. The lowest share of those unable to keep their homes adequately warm (X7) was in Finland in 2015 and 2023.

In 2015, the highest variability was observed for the X7 variable (0.86) in countries in Group 3. This variable was highest for Bulgaria and lowest for Finland. In 2023, as in 2015, the highest variability was recorded for variable X7. Its value in the countries of the medium-low level group was the highest for Spain and Portugal, and the lowest for Czechia. The lowest variability in 2015 and 2023 was for the variable X6.

Group 4 presents the countries with the lowest level of achievement of SDG 7 (

Table 11. Low level of advancement of SDG 7: Group 4 (2015, 2023).

2015
Country	X1	X3	X4	X5	X6	X7
Cyprus	2.68	382	7.04	9.90	97.32	28.3
Belgium	4.05	730	6.46	8.06	84.23	5.2
Luxembourg	7.27	894	11.29	4.99	95.96	0.9
Mean	4.67	668.67	8.26	7.65	92.50	11.47
SD	1.92	213.47	2.15	2.03	5.88	12.03
CV	0.41	0.32	0.26	0.27	0.06	1.05
2023
Country	X1	X3	X4	X5	X6	X7
Belgium	3.57	591	7.84	14.74	76.10	6.0
Lithuania	2.21	537	6.02	31.93	68.04	20.0
Cyprus	2.72	363	9.48	20.21	92.21	16.9
Luxembourg	5.48	686	15.08	11.62	90.62	2.1
Mean	3.50	544.25	9.61	19.62	81.74	11.25
SD	1.24	117.46	3.39	7.74	10.10	7.41
CV	0.36	0.22	0.35	0.39	0.12	0.66

). The highest primary energy consumption (X1) was in 2015, and in 2023 in Luxembourg. The final energy consumption in households per capita (X3) was highest in 2015 and 2023 again in Luxembourg. In Luxembourg, there was the highest energy productivity (X4) among the Group 4 countries, while the lowest was in Belgium (in 2015) and Lithuania (in 2023). The maximum share of renewable energy in gross final energy consumption by sector (X5) was in 2015 in Cyprus and 2023 in Lithuania. The minimum of the X5 variable was observed in Luxembourg in 2015 and 2023. The most dependent on energy imports was Cyprus in 2015 and 2023. In 2015, the largest share of the population unable to keep their homes adequately warm (X7) was in Cyprus, and in 2023, in Lithuania.

Countries with low levels of achievement of Goal 7 noted the highest variability in 2015 for the variable X7. Of the countries classified into this group in 2015, the highest value was recorded for Cyprus and the lowest for Luxembourg. Additionally, in 2023, countries classified into the group with the lowest level of advancement in Goal 7 recorded the highest variability for this variable, with its highest value for Lithuania and the lowest for Luxembourg. The countries in Group 4 were the least differentiated in 2015 and 2023 for the variable X6.

As a next step of the empirical analysis, a sensitivity analysis was applied. Basic descriptive statistics, such as mean, minimum, maximum, and SD, were calculated. The results of the analysis are presented in

Table 12. Sensitivity analysis results (2015–2023).

Descriptive Statistics	X1 Excluded	X3 Excluded	X4 Excluded	X5 Excluded	X6 Excluded	X7 Excluded
Results for absolute values
Mean from the mean	0.4598	0.4929	0.5433	0.5429	0.5130	0.4550
Mean from the minimum	0.2967	0.3044	0.2578	0.3617	0.3495	0.1739
Mean from the maximum	0.6909	0.7509	0.7223	0.7057	0.6612	0.6291
Mean from SD	0.0940	0.1104	0.0988	0.0815	0.0787	0.0968
Mean from CV	20.44%	22.40%	18.18%	15.02%	15.34%	21.28%
Results for unitised values
Mean from the mean	15.6%	8.9%	9.9%	9.7%	7.1%	16.7%
Mean from the minimum	1.2%	0.4%	1.1%	0.3%	0.3%	1.6%
Mean from the maximum	43.0%	28.4%	22.5%	16.7%	17.6%	78.3%
Mean from SD	10.5%	7.8%	5.0%	4.9%	4.9%	14.6%

.

The sensitivity analysis of the modulus values of average deviations revealed that the average difference calculated for all examined years was the biggest when the variable X7 was excluded. The smallest difference occurred when the X5 variable was excluded. It is worth noting that the analysis of the average of the maximum differences indicates very large differences when X7 and X1 were excluded (78.3% and 43%, respectively). The analysis also shows differences of more than 16% for the exclusion of any of the analysed explanatory variables.

Given the results of the sensitivity analysis, an additional classification of countries was carried out each time eliminating one of the variables (X1, X3–X7). The application of this procedure led to obtaining another six classifications for 2015 and 2023. Classifications are shown in

Table A3. Classification of the 27 EU countries in terms of the level of SDG 7—sensitivity analysis results (2015, 2023).

Countries	Classification in 2015 Excluding the Selected Variables						Classification in 2023 Excluding the Selected Variables
	X1	X3	X4	X5	X6	X7	X1	X3	X4	X5	X6	X7
Austria	2	2	2	3	2	2	2	2	2	3	2	3
Belgium	4	4	4	4	4	4	3	3	4	3	3	4
Bulgaria	4	4	3	3	4	2	4	4	3	3	4	2
Croatia	2	2	2	2	2	2	2	2	2	2	2	2
Cyprus	4	4	4	4	3	3	4	4	4	4	3	3
Czechia	2	2	2	2	3	3	3	3	3	3	3	3
Denmark	1	1	1	1	1	1	1	1	1	1	1	1
Estonia	1	1	1	2	3	2	1	1	1	1	2	2
Finland	2	2	3	4	3	3	2	2	3	4	4	4
France	2	2	2	2	3	3	3	3	3	3	3	2
Germany	3	2	3	2	3	3	3	3	3	3	3	3
Greece	4	3	3	3	3	2	4	3	3	3	3	2
Hungary	3	3	2	3	3	3	3	3	3	2	3	3
Ireland	2	2	4	2	1	2	1	2	3	1	1	2
Italy	3	3	3	3	2	2	3	3	3	2	2	3
Latvia	2	2	2	3	2	2	1	1	1	1	1	1
Lithuania	4	3	3	4	3	2	4	4	4	4	4	3
Luxembourg	4	4	4	4	4	4	3	4	4	4	4	4
Malta	4	4	3	2	2	3	3	3	2	2	1	3
The Netherlands	2	3	3	2	3	3	3	3	3	2	2	3
Poland	2	2	2	1	3	2	3	2	2	2	3	3
Portugal	2	3	2	2	1	1	3	3	3	3	2	1
Romania	1	1	1	1	1	1	2	2	1	1	2	1
Slovakia	3	3	2	2	3	3	3	3	3	3	3	3
Slovenia	2	2	2	2	2	2	2	2	2	1	2	2
Spain	2	3	2	2	1	2	3	4	3	3	3	2
Sweden	1	1	1	2	1	1	1	1	1	2	1	1

(Appendix A).

Denmark was classified in Group 1 in each of the models. Sweden, which was classified in the same group, also confirmed its strong position. The exceptions were the models (for 2015 and 2023) with the X5 variable excluded, in which Sweden was qualified for Group 2.

In each of the six models determined for 2015, Romania ranked among the SDG 7 leaders. However, the results for 2023 indicated a weaker position for the country, which was three times classified in Group 1 and three times in Group 2. Estonia and also Latvia in 2023 scored highly in the classification. However, Ireland and Slovenia ranked lower than in classifications for all diagnostic variables (Table 7).

The composition of the group of countries with the lowest SDG 7 levels in 2015 and 2023 in the models with excluded variables was similar to the classifications obtained for all variables. The research confirmed the weak position of Luxembourg, Lithuania, Cyprus and Belgium (particularly in 2015) in SDG 7.

On the one hand, the results of the sensitivity analysis suggested the accuracy of the obtained classifications for all diagnostic variables, while on the other hand, they indicated a relativity of the results obtained in Table 7. The exclusion or addition of a variable impacts the value of the synthetic variable, which affects the country’s ranking position and classification group.

5. Discussion

The next step in the study was to compare the research results with the results of other authors on the raised issue. To our knowledge, these studies were carried out in [19,20,21,22,23].

The first study demonstrated a consistent advancement toward SDG 7 from 2010 to 2021, with a noticeable reduction in disparities among various EU countries. The smallest distance was observed in relation to the target set for SDG 7 for Sweden, Denmark, Estonia, and Austria. By far, Malta made the greatest advancement in the same period, and it was significant for Cyprus, Latvia, Belgium, Ireland, and Poland [19].

The decreasing differences between the countries studied in the implementation of SDG 7 were also suggested by our research. This is evidenced by the decreasing values of the CV determined for the variables studied. The CV for 2023 was lower than for 2015 for variables X1, X3, and X5–X7. The exception was variable X4, that is, energy productivity. In this sense, the variation between EU countries in 2023 was greater than in 2015. The levelling of differences between the EU countries was also pointed out by the reduced range between the largest and smallest values for the SM calculated in 2015 and 2023.

The results of our study in terms of the countries that did the best in implementing SDG 7 in the years studied were generally in line with the results of [19]. We confirmed the high position of Denmark, Sweden, and Estonia in the survey. Furthermore, according to our research, Luxembourg made the greatest advancement in implementing SDG 7, followed by Ireland and Malta. However, our results were not so optimistic for Poland. In our opinion, this country did not see an increase, but a decrease in implementing SDG 7.

The second research referenced sought to assess how the 27 EU countries are progressing in transforming their economies and energy systems in line with the SDGs. It was assumed that the actions taken in connection with the implementation of SDG 7 should unify and integrate the 27 EU countries and that differences between the ‘old’ and ‘new’ EU countries could arise [20]. Studies indicated a considerable disparity among member states regarding the implementation of SDG 7. Furthermore, in the period 2000–2019, this differentiation increased. Differences primarily appeared in the energy mix of the member countries, energy efficiency, reliance on imports, and the percentage of the population unable to keep their homes adequately warm due to the poverty status indicator. Differences were observed in the case of the ‘old’ and ‘new’ EU countries [20]. It is important to note that our research covered the period of 2015–2023, that is, it was compiled on the latest available statistics. The discrepancies in cross-country variation observed between our results and those of [20] are most likely due to the different study periods in which the analyses were performed.

At the top of our ranking, were both better and less developed member countries. In particular, Romania, classified in Group 1, ranked very high in the 2015 ranking and high in 2023 (second and sixth place, respectively). Romania’s high scores may have been significantly influenced by the fact that the country had the lowest level of primary energy consumption per capita (X1) in 2015 and 2023 in the whole EU. Our results for this country are mostly consistent with the literature. In the rankings based on the SM created of economic variables (unemployment rate and expenditure of households on final consumption), Romania ranked ninth in 2015 and sixth in 2023 [21].

Although the EU has made progress towards SDG 7, there are still significant differences between individual countries. For example, Sweden leads in energy sustainability, while Luxembourg has areas that need improvement [22]. According to our research, these areas are: reducing primary energy consumption per capita (X1) and increasing the share of renewable energy in gross final energy consumption (X5). It should be noted that Luxembourg recorded the highest primary energy consumption per capita in the entire Union in 2015, and the lowest renewable energy use in 2015 and 2023.

The share of renewable energy in the final gross energy consumption (X5) and the energy productivity (X4) are the determinants of innovation and modernity in contemporary economies. In both 2015 and 2023, Sweden and Ireland had the best performance in this regard, respectively. Ref. [23] demonstrated that efficient energy utilisation is predominantly characteristic of North-Western European countries, with Sweden and Austria consistently leading the rankings.

Countries whose development is based on traditional energy sources do not fare well in this regard. Among them is one of the ‘new’ member countries, Poland. Its energy mix is currently highly dependent on fossil fuels, particularly coal. However, this country has made notable progress in increasing the share of renewable energy in the energy mix [29].

The low position of Belgium (Group 4 in 2015 and 2023), which is among the highly developed EU countries, also requires a comment. Its position was likely due to the high primary energy consumption per capita (X1) and the low energy productivity (X4). Although Belgium, in our study, was classified among the least advanced countries in SDG 7, ref. [19] showed an improvement in the situation regarding the implementation of SDG 7 in the period 2010–2021. The observed discrepancies may have been because the compared study [19] did not include the variable X1 in the evaluation.

An indicator of the progress of EU countries in SDG 7 is the maintenance of the low percentage of population unable to keep their homes adequately warm due to poverty status (X7). In both compared years, this level was highest for Bulgaria. As the research results suggest, the level of this variable is closely related to the country’s GDP [43]. Bulgaria faces significant energy poverty due to energy-inefficient homes and lower incomes [41]. The problem of energy poverty also affects Southern European countries (Greece, Portugal, Spain), which, despite milder winters than in the rest of the EU, have a large percentage of the population at risk of energy poverty. In contrast, cold Scandinavian countries have the lowest levels of energy poverty [54]. Examples of solutions to this problem include increasing the energy efficiency of homes occupied by households or diversifying the sources of energy used for this purpose. However, this cannot happen without increased investment in upgrading household buildings.

The results of studies show that SDG 7 is crucial for the EU, as it addresses key aspects of energy security [32], environmental sustainability [14,55] and economic growth [56]. The commitment to this goal is evident through strategic frameworks and policies promoting clean energy and reducing carbon emissions. However, achieving SDG 7 requires continued efforts, innovation and cooperation among member states to overcome existing differences and ensure a sustainable energy future for all.

The comparative analysis presented in the paper draws attention to the need for research on the progress of countries in implementing the SDGs. This is essential for the successful implementation of the concept of sustainable development embodied in Agenda 2030. Agenda 2030 is a plan that takes into account the diverse requirements, distinct starting points, and unique national characteristics of each EU member state.

This article addresses the issue mentioned above about developed countries. These countries were evaluated in terms of the level of implementation of SDG 7 on access to affordable and clean energy for all. It turned out that difficulties in achieving SDG 7 could be experienced by less developed EU countries and very well-developed member states.

6. Conclusions

Implementing sustainable development requires actions in social and economic areas. These activities correspond to 17 SDGs, including the seventh goal i.e., ‘ensure access to affordable, reliable, sustainable and modern energy for all’. SDG 7 integrates the two fundamental principles concerning the EU’s energy sector. Firstly, this refers to providing energy in a quantity suited to the needs of the recipients. Secondly, it is crucial that this energy is environmentally friendly, meaning its generation should not cause harm to nature. Energy is the basis for society’s development, and its consumption level largely reflects civilisation and technological progress. It is an essential factor in activities for sustainable development.

This paper provides a proposal for assessing the level of implementation of SDG 7 in the 27 EU countries in 2015–2023. The assessment was carried out using the MCA, that is, the ZUM. As a result, the composite measure of SDG 7 was built.

Based on the SM, we proposed a ranking of the 27 EU countries in terms of the degree of implementation of SDG 7 in 2015–2023. In addition to the rankings, we classified EU countries into groups of countries similar in terms of implementation of SDG 7 in 2015 and 2023. The statistical analysis was also the basis for determining the countries that were the most and least advanced in implementing SDG 7. Furthermore, it was pointed out which countries made the greatest progress (decline) in SDG 7 adoption by comparing the results of 2023 with 2015.

The research suggested the downward differentiation of EU countries in terms of the degree of implementation of SDG 7. The top three countries in the implementation of SDG 7 in 2015 were Denmark, Romania, and Sweden and in 2023, Denmark, Sweden and Estonia. The countries that appeared to be weakest in this regard were, in 2015, Cyprus, Belgium, and Luxembourg, and in 2023, Belgium, Lithuania, Cyprus, and Luxembourg. Among the 27 EU countries, Luxembourg, Ireland, and Malta made the greatest progress toward SDG 7, while Spain experienced the largest decline. The research results highlighted obstacles in implementing SDG 7 not only for the less developed countries, mainly in Central and Eastern and Southern Europe, but also for the highly developed Western countries. The research procedure employed could assist in pinpointing necessary enhancements for the successful execution of SDG 7 within EU member states.

In the study, we used data obtained from the Eurostat database for 2015–2023. It allowed the evaluation of the changes in energy access in the 27 EU countries. The application of a multidimensional measurement method, with the help of the SM, allowed a relatively simple assessment of the implementation of SDG 7.

Based on statistical analysis, the development of theoretical and managerial implications was possible. This article contributes to theory by extending the domain of sustainable development. Among the managerial implications, the proposed multicriteria assessment allowed us to compare EU countries using only one SM. A further recommendation pertains to the management cadre. Given the established rankings, an opportunity arises to pinpoint effective practices within the relevant countries and implement corrective and remedial actions. Furthermore, the applied research procedure may have implications for the continued successful implementation of SDG 7 by individual countries and the EU as a whole. The proposed multidimensional method can be used to assess the level of advancement of SGD 7 and other SDGs in non-EU countries.

However, the presented procedure has its limitations. Emphasising the complexity of the concept of sustainable development is crucial. It incorporates economic, social, and environmental aspects. These aspects should be reflected in the variables used to measure. To assess the level of advancement of countries, we used a ready-made set of indicators proposed by Eurostat, which does not adequately reflect all of the issues included in SDG 7. As indicated in the article, the first three variables had an ambiguous impact on the evaluation of SDG 7. However, we ultimately classified them into destimulants. When discussing the limitations of the study, it is important to mention the high variability in the value of the SM within the countries. This is due to the zero unitarisation method application for creating the SM. Another limitation associated with the presented study is the linear character of the applied method. In the linear approach, there is a difficulty in interpreting the distances between objects whose ranking positions are determined on the basis of the SM.

Given the possibility of using the SM to conduct surveys among all countries that implement the SDGs, for which data are available, further research is highly recommended. However, it should be stressed that measuring progress in sustainability requires special attention, and there is no ideal method to evaluate it.