The Well-Being-Related Living Conditions of Elderly People in the European Union—Selected Aspects

Abstract

One of the main demographic problems is the ageing of populations, especially in European countries. However, this process is not taking place equally in all countries and has a different impact on their socio-economic development. The aim of this research is to assess and compare the well-being-related living conditions of elderly people in European Union countries. We carried out the analysis for the years 2015 and 2020 on the basis of Eurostat data. We assessed the well-being-related living conditions of older people by applying the multidimensional scaling technique. On its basis, we constructed rankings of the EU countries in the two analysed years. In addition, by using the k-means method, we clustered the countries. The disparity between the well-being-related living conditions of elderly people in Western European countries and the post-communist ones has not declined. Six of the nine analysed indicators improved over the five years between 2015 and 2020, while three (the proportion of elderly people at risk of poverty, the disparity between the incomes of the richest and poorest groups of elderly people, and the proportion of elderly people in the total population) worsened. Socio-economic policies should therefore address these indicators to the greatest extent possible.

1. Introduction

Europe’s population is ageing. Over the period 2002–2022, the share of persons aged 65 and over increased in all member states from 16% to 21% (5 percentage points). In the group aged 80 and over, their share increased from 3.5% in 2002 to 6.1% in 2022. At the same time, the young population is decreasing. Over the period 2002–2022, the share of young people (aged 0–19 years old) decreased in all member states from 23% to 20% [1]. On the one hand, the extension of life expectancy has a positive impact on economic growth in European Union countries [2]. On the other hand, there is a negative demographic impact on the annual economic growth resulting from population onus. The problem of population ageing also affects other high-income countries and Central Asian countries [3]. After analysing demographic data and numerous articles in this area, the following research problems arise:

• The ageing of populations is one of the main demographic problems.
• This problem applies to highly developed European countries in particular.
• This problem is not taking place equally in all countries and has a different impact on their socio-economic development.

The ageing of the population can be considered in terms of economic and social aspects. These aspects are interpenetrating and encompass a number of factors. Ageing is connected with disability and increased health problems. However, many elderly people subjectively evaluate their health positively and see ageing as a positive life period with increased maturity and wisdom. A person’s well-being in later life is multidimensional and changes with time. As seniors age, they experience a decline in their sense of purpose in life and opportunities for personal growth, partly due to socio-economic factors [4]. Jivraj et al. looked at age-related changes in subjective well-being (SWB) later in life in their study. They showed that older cohorts had equivalent or better SWB than younger cohorts on each measure of well-being [5]. This research confirms the results of the analyses previously carried out by Horley and Lavery [6]. Lee [7] analysed social exclusion and subjective well-being among older adults in Europe. His research indicated that older adults from the Nordic countries were more likely to score low on social exclusion and high on subjective well-being. Older people from Central and Eastern European countries were more likely to score low on subjective well-being and high on social exclusion.

Authors often use the terms “quality of life”, “happiness”, “well-being”, and “life satisfaction” interchangeably in their research. These terms all have a lot in common, but we should be aware that there are some differences between them. The literature review section presents the terms used by the cited authors.

Improving the quality of life in society is one of the main goals of sustainable development. It is assessed using a variety of indicators, with the environment being one of the main factors influencing quality of life [8]. Research by Zhou et al. [9] has shown that the positive impact of age-friendly environments on the quality of life of older people is stronger for mental well-being than physical well-being. It is important to emphasise that there is no agreement on the definition of quality of life [10]. Every society is characterised by unique social, historical, and economic conditions. On the one hand, it is difficult to fully understand the living conditions of older people in a given society solely by comparing their quality of life with that of older people in another society [11]. On the other hand, comparative analyses allow us to identify differences in the quality of life, which allows for the creation of appropriate policy in international terms. European comparative studies show different priorities among elderly people in various countries. For instance, there is a bigger impact on the role of the family in southern Europe compared to northern Europe [12,13]. Another example of differences in Europe is the greater influence of objective living conditions on subjective quality of life in the post-communist bloc, such as Hungary and former East Germany, compared to more developed and prosperous European countries [14]. In general, in European countries, older people pay the most attention to their financial situation when assessing the condition of their household, followed by life satisfaction and professional position [15,16]. Developmental changes in old age can have a negative impact on the objective quality of life. But at the same time, there are internal changes that can improve the subjective quality of life [17].

Knowledge about the quality of life in old age has many gaps. In order to take effective action, it is important to correctly identify them and set research priorities [18]. As emphasised by Wahl et al. [19], it cannot be overlooked or forgotten that the final result of research on ageing is intended to serve seniors. Studies can be conducted on the basis of objective premises and statistical indicators created on the basis of data available in official statistics. In the case of research on quality of life, qualitative research based on the subjective feelings of older people provides a lot of interesting results. As Jürges and van Soest [20] point out, such studies raise the problem of subjective assessment of objective reality.

The aim of this research is the multivariate assessment and comparison of well-being-related living conditions of elderly people in European Union countries in the years 2015 and 2020. The study was conducted in an objective context, based on available indicators. We focused on the social and economic domains of well-being. There are many publications that analyse selected aspects of quality of life, where the rankings of countries are constructed. However, there is a shortage of publications that also compare these rankings over several years, which would allow us to assess the change in the situation of the analysed phenomenon in the multivariate aspects. Our research aims to fill this gap.

We put forward two research hypotheses:

Hypothesis 1. 

The well-being-related living conditions of elderly people are diversified across the EU countries.

Hypothesis 2. 

The well-being-related living conditions of elderly people in the new (after 2004) EU member states still lag behind the well-developed Western European ones.

The study was carried out with respect to two aspects: rankings of the EU countries and cluster analysis of the EU countries. We assessed the well-being-related living conditions of older people by using a synthetic variable created by applying the multidimensional scaling technique. On its basis, we constructed rankings of the EU countries in the two analysed years. In addition, by using the k-means method, we clustered the countries with respect to the values of the analysed indicators.

There are many studies that deal with the topic of quality of life. The added value of our study is to propose a set of indicators for the study of the well-being-related living conditions of elderly people.

The article is organised as follows: Section 2 contains a review of the literature. Section 3 presents the data used in the study and the research methods applied. Section 4 presents a discussion of the obtained empirical results. In Section 5, we present a discussion, and Section 6 contains conclusions, limitations of the study, and directions for future research.

2. Literature Review

The elderly population is growing much faster than the total population of the world. The world’s population is ageing, but the population of European Union countries is ageing faster than the rest of the world [21]. In 2016, Luxembourg, Denmark, France, and the Netherlands had the lowest rates of people at risk of poverty and social exclusion; Bulgaria, Estonia, and Latvia had the highest ones. People living in Northern Europe had generally better health and well-being than in other European countries. Ruggeri et al. [22] analysed multidimensional psychological well-being (MPWB) in 21 European countries between 2006 and 2012. Denmark had the highest well-being, and Bulgaria had the worst. In Europe, women exhibited lower MPWB scores than men. Seniors were typically characterised by lower MPWB scores compared to younger age groups in Europe.

One important research area is the assessment of active ageing. Research of this type is gaining momentum in connection with the pursuit of sustainable development [23]. In their research, the authors show that when aggregating indicators, all classifications should take into account the specificity of individual countries. Similar studies were conducted by Ramia and Voicu [24]. They investigated the effect of active ageing on subjective quality of life in 27 European countries in 2016 in a group of older people (65+). Their research shows that being active in later life does not necessarily lead to improvements in one’s own assessment of quality of life. Some may choose not to be active, while others may need to maintain certain levels and types of activity, which affects their well-being. Engaging in productive activities is beneficial to older people’s well-being [25,26]. Research by Steinmayr et al. [27] indicates that gender has an impact on active ageing. In Poland, Estonia, and the Czech Republic, but also in Denmark, on average, women score higher than men, while the opposite occurs in Austria, Belgium, and Switzerland. In Germany, Slovenia, Sweden, and France, no major gender differences were found. Spatial and cultural aspects of assessing the levels of individuals’ subjective quality of life are highlighted by Somarriba Arechavala and Zarzosa Espina [28]. Their analysis covered 28 European countries between 2011 and 2012. The ranking based on the Subjective Quality of Life Indicator was led by Denmark, Austria, and Finland. At the end of the ranking (looking from the end), there were Bulgaria, Latvia, and Greece.

Crisis situations can have a negative impact on the well-being and quality of life of older people. Ferreira et al. [29] analysed health-related anxiety levels and quality of life amongst Portuguese citizens under compulsory home quarantine due to the COVID-19 pandemic. Their research confirmed that one of the groups most at risk of the effects of quarantine were the elderly (60+). Garcia Diaz et al. [30] found that face-to-face interactions and support from family, friends, a healthcare provider, and the community alleviated the influence of social distancing restrictions at the time of the COVID-19 pandemic. The deterioration of the well-being of these support groups has had a negative impact on the well-being of older people during the pandemic. In contrast, a positive effect of group physical activity has been observed [31]. Research by Polinesi et al. [32] indicates that the European countries that experienced greater declines in well-being in the first year of the pandemic were France, Luxembourg, and Malta. In the second year of the pandemic, however, the Czech Republic, Spain, and Italy suffered the most. People living in northern Europe had better health and well-being overall than in other European countries [33]. The results obtained by Delhey et al. [34] indicate that in Germany and the United Kingdom, the developing pandemic did not increase inequality in well-being, and surviving the pandemic demanded psychological resources in the first place. A study by Socci et al. [35] showed that the active activities of older volunteers in Italy during the pandemic period had a positive impact on their overall well-being. In the Netherlands, younger individuals reported a significantly greater negative impact of the pandemic on physical activity and being active than the older participants [36]. Thus, the study shows that not all older people were equally susceptible to the effects of COVID-19, and some people even saw an improvement in their well-being [37].

The impact of socio-economic, health, and social factors on the quality of life among the senior population in the countries of southern Europe in the years 2014–2015 was analysed by Cantarero-Prieto et al. [38]. They found that these countries have a lower quality of life. They showed that older age, disability, and poorer health were associated with lower quality of life scores. The effect of gender was not significant. Secondary and tertiary education, being retired or employed, and living in rural areas contributed to higher quality of life ratings.

Healthy and active ageing can be considered in a socio-economic and cultural context. Factors that increase the well-being of the elderly include the opportunity to develop their interests and passions. Hence, universities of the third age are being established all over the world, enabling the further education of older people [39,40,41], and the senior travel market is growing [42,43,44,45]. Having a sense of life is one of the most important goals of older adults, whether they are formally affiliated with a religion or not [46]. Researchers also point to increased interaction with people or animals to improve the well-being of elderly people [47]. Public space, elderly population density, and services for seniors are positively connected with the well-being of seniors [48]. The connections between leisure consumption and well-being were also stronger for older people [49].

Research on happiness indicates that it changes with age and, in most cases, is an inverted U-shape. Analyses conducted by Laaksonen [50] have shown the existence of an age point when the happiness increase starts declining. It most often appears at a late age and is associated with the ageing process. European countries vary in size, which can affect the turning point. The fastest age point (under 60 years of age) occurs in Switzerland, Turkey, and the Netherlands, while France has the highest age point at 82.4 years. In countries recognised as stronger welfare states, the probability of living the happiest period declines more slowly with age than in countries recognised as weaker welfare states [51].

The subjective assessment of the quality of life of older people is influenced by their health status [52] and their financial situation [53]. The quality of life of older people, in particular those with disabilities, is improved by well-run and effective long-term care policies [54]. Improving the quality of life is often associated with the introduction and use of new innovative services and products based on new technologies. However, research indicates that older people rarely use such innovative products themselves [55]. This problem applies not only to industrial goods but also to food products. The vast majority of seniors indicate a variety of concerns about their quality and safety. However, in the case of the elderly, the process of education in this regard may be difficult due to well-established nutritional practices related to the purchase and consumption of products necessary for life [56]. The shopping behaviour of the elderly is influenced by the location of the store in relation to the senior’s home, easy access to the store, low prices and price reductions, and the impact of brand awareness. The importance of these factors increases with age [57]. Researchers also point to the impact of energy poverty on the poor health and well-being of older people in Europe [58].

Akdede and Giovanis [59] analysed connections between the net migration rates and the subjective and objective well-being of the senior natives in Europe during the period of 2004–2017. The authors proved a positive impact of migration on the subjective well-being and wages of elderly natives and second-generation immigrants in the Northern, Western, and Eastern European countries and a negative impact in the Southern region.

The aim of many studies is to answer the question: What factors affect the improvement of the living conditions and well-being of elderly people? Research conducted by Mária Sováriová Soósová [60] indicates the need to create opportunities for the development and maintenance of social contacts and to engage seniors in various forms of spending free time and in various types of programmes or volunteering. The screening for and treatment of depression and anxiety also improves the quality of life of the elderly. Van Leeuwen et al. [61] draw attention to the need for qualitative research. There is a lack of a systematic review of the opinions of older people themselves. Such knowledge is necessary to match the goals of care services to their expectations. An analysis of the life satisfaction indicator (LS) indicates that the debt of older people in Europe is significantly and positively related to low LS levels [62]. The opposite is true for the incomes of people aged 65 and over. Research by Neuberger and Preisner [63] indicates an ambiguous impact of parenthood on the quality of life in old age. It all depends on individual resources, the economy, and social service expenditures. In Europe, individual perceived autonomy, opportunity, and choice-enhancing societal conditions increase individual life satisfaction [64]. According to the cited study, six basic functionings (safety, friendship, health, financial security, leisure, and respect) exert a positive influence on individual life satisfaction. The relationship between these functionings and subjective well-being is attenuated by perceived autonomy and societal conditions. Objective income positions can be consistent with subjective self-perceptions, both good (well-being) and bad (deprivation), of people’s income situation [65]. However, the situation varies across the various countries. The authors stated that the discrepancy between subjective feelings and objective situations is particularly evident in Latvia, Lithuania, and Estonia. In these countries, the subjective sense of financial problems is higher than the objective situation. In Western and Northern European countries, the objective situation of individuals and their subjective assessment are better matched (particularly in Luxembourg, Germany, and Austria). Sometimes the objective situation of financial insecurity outweighs even a subjective sense of financial problems. This has especially been observed in Denmark, Sweden, and Switzerland.

3. Materials and Methods

The study was based on data available in the Eurostat database. The analysis covered the 27 countries of the European Union. Due to the availability of data, the observation period covered the years 2015 and 2020. On the basis of the available data and literature studies, it was decided to include 9 variables in the analysis. These are described in

Table 1. Variables applied to the assessment of well-being-related living conditions of elderly people in the European Union countries.

Variable Symbol	Variable Description
$x_1$	Mean consumption expenditure for persons aged 60+ years (in Euro, in purchasing power standard per adult equivalent)
$x_2$	Persons at risk of poverty or social exclusion aged 65+ years (percentage of the population aged 65+)
$x_3$	Income quintile share ratio S80/S20 for disposable income of people aged 65+ years (ratio)
$x_4$	Distribution of the population aged 65+ (percentage of the total population)
$x_5$	Distribution of the population aged 65+ assessing their health status as good or very good (percentage of the population aged 65+)
$x_6$	Housing cost overburden rate for the population aged 65+ (percentage of the population aged 65+)
$x_7$	Annual old age pension (in Euro, in purchasing power standard per inhabitant)
$x_8$	GDP at market prices (in Euro, in purchasing power standard per capita)
$x_9$	Activity rate of persons aged 65+ (percentage of the population aged 65+)

.

The study was carried out in two stages, as shown in Figure 1. The first step was to use multidimensional scaling to determine the distance between the pattern and the antipattern. Next, the composite variable was constructed using the multidimensional scaling and TOPSIS methods. This allowed us to build a ranking of the EU countries with respect to the well-being-related living conditions of elderly people. In the second stage of the analysis, we distinguished the homogeneous clusters of countries in terms of the well-being-related living conditions of elderly people by using the k-means method. Calculations were performed in the language R [66] with the use of libraries: clusterSim [67], ggforce [68], ggplot2 [69], ggrepel [70], grid [66], NbClust [71], plotrix [72], rworldmap [73], and stats [66].

3.1. Multidimensional Scaling

Similarity between countries was analysed and presented by means of multidimensional scaling. In order to graphically represent distances between the objects described by means of multiple variables, the dimensions need to be reduced to no more than three [74,75]. We can distinguish three main groups of multidimensional scaling techniques:

• Classic (metric) multidimensional scaling (cMDS), also called main coordinate analysis (PCoA)
• Non-metric multidimensional scaling (nMDS)
• Generalised multidimensional scaling (GMD)

Classic (metric) multidimensional scaling is conducted for metric data, and the results are contained in the Euclidean space. Non-metric multidimensional scaling is performed for non-metric, ordinal data, while generalised multidimensional scaling is the extension of the metric one where the target space is non-Euclidean.

The first step of any multidimensional (multivariate) analysis is the observation matrix X:

$$X=\left[x_{ij}\right]=\left[\begin{array}{cccc}{x}_{11}& x_{12}& \cdots & x_{1m}\\ x_{21}& x_{22}& \cdots & x_{2m}\\ ⋮& ⋮& \ddots & ⋮\\ x_{n1}& x_{n2}& \cdots & x_{nm}\end{array}\right]$$

(1)

where:

• x_ij—value of j-th variable in i-th object (i = 1, …, n, j = 1, …, m)
• m—number of variables
• n—number of objects

As variables have various units and magnitude orders, they must be normalised before further proceeding. All variables used in the study are measured on an interval scale. Therefore, among the many available normalisation procedures, we selected one of the quotient inversions:

$$z_{ij}=\frac{x_{ij}}{\sqrt{{\sum }_{i=1}^n{x}_{ij}^2}}$$

(2)

where:

• z_ij—normalised value of j-th variable in i-th object (i = 1, …, n, j = 1, …, m)

After normalisation, all variables are stripped of units, and their order of magnitude is unified. All variables can be of three types: stimulants, destimulants, and nominants. Stimulants are the types of variables for which the largest possible variables are desirable. The opposite holds for the destimulants. Nominants are the variables for which a specific value or interval is desirable. In the study, variables x₁, x₅, x₇, x₈, and x₉ were the stimulants. The remaining ones were the destimulants. There were no nominants in the set of indicators.

Knowing the best and worst values of the analysed indicators, we determined the hypothetical countries with the best and worst values of the indicators in both analysed years. They are further referred to as the pattern and the antipattern.

After all these preparations, we conducted the metric multidimensional scaling in the following steps:

• We calculated the distance matrix between the objects δ for the m-dimensional space by means of the Euclidean metric.
• We mapped the distance matrix δ to the matrix d for the q-dimensional space (q < m). In order to represent the results of multidimensional scaling graphically, q = 2.
• We calculated the loss of information when mapping a distance matrix in m-dimensional space to a distance matrix in q-dimensional space (stress) using the following formula:

$$Stres{s}_{\mathit{\delta }}\left(x_1,x_2,\dots ,x_m\right)={\left(\sum _{i\ne k=1}^m{\left({\delta }_{i,k}-‖x_i-x_k‖\right)}^2\right)}^{\frac{1}{2}}$$

(3)

where $‖x_i-x_k‖$ are distances between objects after a reduction in the space dimension from m to q.

Multidimensional scaling minimises the function of stress, and because of this, an optimal representation of the distance matrix in the m-dimensional space to the distance matrix in the q-dimensional space is obtained.

3.2. Application of the TOPSIS Method

We use the TOPSIS formula in order to calculate the distances of every object from the hypothetical best one (pattern) and the worst one (antipattern) in the q-dimensional space:

$$c_i=\frac{d_i^{-}}{d_i^{-}+d_i^{+}}$$

(4)

where:

• $d_i^{-}$—the distance of the i-th country from the antipattern
• $d_i^{+}$—the distance of the i-th country from the pattern

There are several reasons for using the TOPSIS (Technique for Order Preference by Similarity to Ideal Solution) method proposed by Hwang and Yoon [76]. The first one is that it considers distances from both pattern and antipattern objects. The second reason is that it is very easy and straightforward to use. Although it was meant to be used in decision theory, it could also be used in applications of multivariate statistical analysis [77,78,79,80]. There are other possible linear ordering methods that can be used in creating rankings of objects. One can use Hellwig’s method, which is based on the weighed distances of the objects from the pattern [81], or the patternless COPRAS (Complex Proportional Assessment) method [82].

3.3. The K-Means Method

The k-means method, a cluster analysis technique, was proposed by MacQueen [83]. It belongs to the group of methods that optimise the initial division of objects and is often used in the analysis of socio-economic phenomena [84,85,86]. We conducted the k-means clustering using the following steps:

• We created the observation matrix X.
• We normalised the data.
• We divided the set of objects into s clusters (s = 1, …, k, …, n).
• We calculated the centre of gravity in every cluster (centroid) and the distance of every object from it.
• We changed the assignment of objects to clusters with the closest centroid.
• We calculated the new centroids.
• We repeated steps 5–6 until the next relocation of objects ceased to improve the general distances of objects from the centroids.
• We repeated steps 4–7 for various numbers of clusters.

In order to select the optimal division of objects between clusters, we use the cost measure, which is the sum of squares of within-group distances from the centroids:

$$Cos{t}_P^{*}=\underset{P}{\mathrm{argmin}}\sum _{s=1}^k\sum _{z_{ijs}\in P_s}{\left(d_{is}-{\overline{d}}_s\right)}^2$$

(5)

where:

• P = {P₁, …, P_k}—set of homogeneous clusters
• $d_{is}$—distance of the i-th object from the centroid for the s-th cluster
• ${\overline{d}}_s$—centroid for the s-th cluster

We conducted the research in the following stages:

• For the years 2015 and 2020, we applied multidimensional scaling to represent the distances between the EU countries with respect to the well-being-related living conditions of elderly people.
• Using the TOPSIS formula, we created the rankings of EU member states with regard to the well-being-related living conditions of elderly people in the years 2015 and 2020.
• By using the k-means method, we distinguished the homogeneous clusters of the EU countries with respect to the well-being-related living conditions of elderly people.
• We analysed the mean values of indicators in the clusters for the years 2015 and 2020.

4. Results

4.1. Graphical Representation of Distances between the EU Countries in 2015 and 2020

We present the distances between the EU countries and the hypothetical pattern and antipattern ones in both analysed years together. As the pattern country (being the one with the best values of analysed indicators) is the best one, we also draw the isoquants that represent equal distances from it. We present them as circles with radii equal to ¼, ½, ¾, and 1 distance between the pattern and the anti-pattern. The countries that are located closer to the pattern “country” are considered to have a better situation with respect to the well-being-related living conditions of elderly people than the countries that are located further away. Therefore, we consider the graphical representation of distances between the EU countries and the pattern and the anti-pattern ones as an assessment of their situation with respect to the well-being-related living conditions of elderly people in both analysed years. The situation of the EU countries with respect to the well-being-related living conditions of elderly people in 2015 and 2020 is presented in Figure 2.

In both analysed years, the following countries: Luxembourg, Ireland, Austria, the Netherlands, Sweden, and Finland had the best situation with respect to the well-being-related living conditions of elderly people. In contrast, the worst situation was in the cases of Bulgaria, Romania, Greece, Latvia, Lithuania, and Croatia. In general, the well-developed Western European countries had a better situation than the post-communist Eastern and Middle European states. However, it is worth noting that in order to assess the situation of a given country, we must refer to both the pattern and the antipattern. For example, when we compare Denmark to Belgium, it can be noticed that both countries have a similar distance to the pattern. Belgium, however, is closer to the anti-pattern. Therefore, we may consider that Belgium has a worse situation with respect to the well-being-related living conditions of elderly people than Denmark.

The advantage of presenting the results of multidimensional scaling for both years is that we can observe not only the mutual relations between the countries but also state if their situation has improved or deteriorated. Therefore, when we compare the positions of the countries in both analysed years, it becomes visible that the situation of Cyprus and Spain deteriorated (the points representing these countries in 2020 were more distant from the pattern than in 2015), while the situation of Luxembourg, Belgium, and France remained more or less at the same level (in both analysed years, their distance from the pattern did not change much). The situation in the remaining countries improved. The biggest improvement was made by Greece, Romania, and Germany (the positions of these countries moved towards the pattern to the highest degree).

The obtained results confirm hypothesis H1—in both analysed years, the EU countries were diversified with respect to the well-being-related living conditions of elderly people. It is visible in Figure 2, as the multidimensional scaling preserves distances between objects. Although most countries were located between ¼ and ¾ of the distance between the pattern and the antipattern, some of them were closer to the pattern (Ireland and Luxembourg) and some were closer to the antipattern (Bulgaria, Greece, and Romania in 2015).

Also, hypothesis H2 was confirmed by means of multidimensional scaling. The distances between the well-developed Western European countries were quite small, and this group of countries was close to the pattern. Also, the distances between the new (after 2004) member states were quite small, and these countries were located further from the pattern. However, they were also quite far from the antipattern—the location of which was mostly influenced by the two worst countries (Bulgaria and Greece).

4.2. Rankings of EU Countries

The distance data presented in Figure 2 were then applied using the TOPSIS method to create the rankings of countries, as shown in

Table 2. Rankings of the EU countries with respect to the well-being-related living conditions of elderly people. Source: own elaboration.

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

.

In both analysed years, the rankings of the countries were very similar. This is confirmed by the high value of the Spearman’s rank coefficient, equal to 0.95. The best EU countries with respect to the well-being-related living conditions of elderly people were: Austria, Denmark, Finland, France, Ireland, Luxembourg, the Netherlands, Sweden, and Germany (in 2020). The worst EU countries with respect to the well-being-related living conditions of elderly people were: Bulgaria, Croatia, Greece, Romania, and the Baltic States. Among the post-communist countries, the best situation was in Slovenia and (quite surprisingly) Slovakia. Germany was the country whose position improved to the highest degree, while Cyprus and Spain were the countries whose positions deteriorated to the highest degree.

4.3. Cluster Analysis

In the third step of the research, we identified the homogeneous clusters of the EU countries with respect to the well-being-related living conditions of elderly people by means of the k-means method. The optimal number of clusters was obtained by means of the NbClust function in the NbClust R package. In both analysed years, we obtained the same results, as shown in Figure 3.

We obtained four clusters of EU countries:

• The first cluster (marked in orange): the Baltic States, Bulgaria, and Romania
• The second cluster (marked in red): Croatia, Cyprus, Czechia, Hungary, Italy, Malta, Poland, Portugal, Slovakia, Slovenia, and Spain
• The third cluster (marked in green): the Nordic countries, Austria, Belgium, France, Germany, Greece, and the Netherlands
• The fourth cluster (marked in blue): Ireland and Luxembourg.

Homogeneous groups of EU countries were quite clearly separated geographically. The first cluster was located in Eastern European, post-communist countries. The second one consisted of the Central European post-communist countries and the Southern European ones (with the exception of Greece). The third cluster consisted of most Western European countries, the Nordic countries, and Greece, while the fourth one included Ireland and Luxembourg.

4.4. Mean Values of Indicators

In the final stage of the analysis, we calculated the average values of indicators in every cluster in both analysed years, as shown in

Table 3. Mean values of indicators in analysed clusters. Source: own elaboration.

Cluster No.	x1	x2	x3	x4	x5	x6	x7	x8	x9
2015
1	7418.80	42.34	4.51	18.72	14.00	13.30	1091.27	17,690.12	7.76
2	13,472.27	18.55	3.76	17.18	24.55	6.13	1713.61	22,440.26	4.73
3	20,541.33	13.72	3.57	18.92	49.44	13.50	3216.88	32,132.02	5.68
4	28,423.00	12.10	3.79	13.30	55.85	3.10	2585.68	63,642.80	6.75
2020
1	8679.60	43.80	4.61	20.10	19.70	8.86	1461.97	22,431.94	9.82
2	13,659.91	21.27	4.00	19.26	30.89	4.63	1983.76	24,873.35	5.67
3	21,351.22	15.88	3.53	20.10	52.83	11.08	3623.31	34,472.17	6.41
4	28,400.00	13.10	3.96	13.70	61.70	3.80	2954.03	70,063.60	7.75

Note: The best values are marked in green, and the worst ones are marked in red.

.

Generally, for the six indicators, the best values were in the fourth cluster, and the worst values were in the first cluster. The third cluster generally has the second-best values of the analysed indicators, and the second cluster has the second-worst values. The results of the cluster analysis generally confirm the results of the rankings obtained by means of multidimensional scaling. The only exception is the membership of Greece in the third cluster, despite it being one of the worst countries with respect to the well-being-related living conditions of elderly people.

The values of most of the indicators in the clusters improved between 2015 and 2020. The three exceptions were:

• Percentage of persons aged 65+ years at risk of poverty or social exclusion
• Income quintile share ratio S80/S20 for disposable income of people aged 65+ years
• Percentage of population aged 65+ years

These results mean that, in general, the situation of elderly people in the EU with respect to well-being-related living conditions has improved. However, their share in the total population increased, their income inequality increased, and their risk of poverty and social exclusion also increased.

The cluster analysis also confirmed both research hypotheses. The well-being-related living conditions in the EU countries are geographically diversified, which confirms hypothesis H1—well-developed Western European countries created clusters (numbered 3 and 4) with the best average values of analysed indicators (with the exception of Greece). In contrast, the post-communist and Southern European countries created clusters (numbered 1 and 2) with the worst values of the analysed indicators. Their worse position in the rankings is confirmed by membership in the clusters with the worst values of the analysed indicators—so it confirms the research hypothesis H2.

5. Discussion

The presented study confirms the analyses carried out in previous years by other researchers. It follows that there is still a divide in the EU between countries with better and worse living standards for older people. This division broadly coincides with the division of countries into the “old” and “new” EU. However, even within the “old” EU, there are differences in the well-being-related living conditions of elderly people. Hence, the countries of Northern, Western, and Southern Europe are often additionally distinguished in the research. The countries of the “new” EU have a post-communist origin, which affects the general standard of living of older people. However, differences are starting to emerge in this group as well.

Research by Grané et al. [87] indicates that the population in Northern and Central European countries has the best well-being and health status, with the top three places occupied by Denmark, Sweden, and Switzerland. The worst situation in this respect is in Eastern European countries, such as Poland, Estonia, and Portugal. The rest of the countries are jumping between different levels. The analysis of activity participation and well-being among Europeans (excluding the countries of the communist bloc) for adults aged 65 years and older by Vozikaki et al. [26] showed that Northern Europe achieved a significantly higher position in all indicators reflecting high well-being, with these indicators being more than twice as high in northern countries compared to central countries and more than three times higher than in southern countries [26].

The results of our study coincide with the results of research on the activity of older people. Research by Przybysz and Stanimir [88] indicates that the most active seniors live in Denmark, Finland, the Netherlands, and Sweden. The least active seniors come from countries such as Croatia, Greece, Poland, and Romania. This is in line with our findings on countries with better and worse well-being-related living conditions for older people in the EU.

An adverse phenomenon affecting the standard of living of the elderly is social exclusion. Panek and Zwierzchowski [89] showed that Denmark, Sweden, France, and Germany were less affected by the social exclusion of older people than countries in Eastern Europe (the Czech Republic and Poland) and the Mediterranean (Italy and Spain).

On average, the inhabitants of the Nordic countries and Western Europe are more happy with their lives than those in southern and eastern Europe [90]. One explanation for the lower quality of life conditions in the Baltic States and Eastern Europe may be that older generations of inhabitants in these countries, which constitute a significant proportion of the population, have experienced dramatic events in terms of economic, social, and political change. Such events may have influenced their judgement and cognitive evaluation of certain areas of life. Southern European countries (and some Eastern European ones as well, such as Hungary) were severely affected by the economic crisis of 2008–2014, which had a negative impact on their financial situation and employment. An obvious polarisation in the dispersion of quality of life standards: the North and Centre as opposed to the South and East of Europe after the 2007–2011 crisis was also shown by Somarriba Arechavala et al. [91]. Another reason for the differences may be significant differences in the quality of life between urban and rural areas. Research by Shucksmith et al. [92] indicates that in Eastern and Southern European countries, rural areas are characterised by significantly lower levels of perceived well-being and quality of life. In the richest countries of the European Union, the differences between urban and rural areas are small. Despite this, the authors emphasise that subjective well-being is not significantly different. Many studies confirm the great impact of the built environment on the standard of living of older people. The built environment includes buildings (homes, workplaces, and schools), open spaces (recreational areas and parks), and infrastructure (transportation systems) [93]. In old age, the neighbourhood becomes particularly important, as many older people spend most of their time there. The quality of life of older people is also influenced by the development of infrastructure related to various aspects of life: public transport [94], social infrastructure [95], public health [96], and communication [97]. In this respect, the countries of Western and Northern Europe are more advanced than the countries of Eastern Europe, especially when it comes to the development of infrastructure in rural areas.

When analysing the Human Development Report 2021–22 [98], it should be noted that all EU countries have a very high Human Development Index (HDI). The HDI rank is open to all countries around the world. In 2021, almost all EU countries were classified as countries with very high human development. The exception is Bulgaria, which is classified as a country with high human development. Among the EU countries, Denmark, Sweden, and Ireland ranked the highest (6th, 7th, and 8th place, respectively), while Hungary, Romania, and Bulgaria were ranked the lowest (46th, 53rd, and 68th, respectively). A total of 19 EU countries recorded an increase in their position in the ranking in 2021 compared to 2020. Among the countries that recorded a drop in the ranking were Hungary, Romania, and Bulgaria. The ranking was made for people aged 15 and over, but it largely coincides with our results.

Multidimensional scaling has been used in the research of economic phenomena many times. In our research, a dynamic approach to the use of multidimensional scaling was first used in the analysis of population ageing in Polish provinces [99] and in the V4 countries [100]. It was also used in static analyses, for example, in research on the financial condition of medium-sized enterprises in Wielkopolska province in Poland [101] or in the analysis of the real estate market [102].

6. Conclusions

The well-being-related living conditions of elderly people are a very broad issue encompassing problems in the fields of economic development, social development, environmental protection, and political issues. The quality of life of people, including the elderly, is affected by the introduction of balance in all these aspects. Hence the link between the well-being-related living conditions of elderly people and each of the 17 Sustainable Development Goals.

The study covered the years 2015 and 2020. The differences in the well-being-related living conditions of elderly people between the European Union countries were significant but did not change significantly in 2020 compared to 2015. However, the general situation in most countries has improved. It follows that changes in the well-being-related living conditions of older people are a long-term process. Despite the fact that post-communist countries have been part of the EU for many years, the standard of living of the people, especially the elderly, is lower than in other EU countries.

The limitations of our study are related to the availability of up-to-date data. Not all the indicators that we intended to include in the study were available for all EU countries. Therefore, we had to limit ourselves to those in the Eurostat database. The availability of indicators for subsequent years will make it possible to extend the analysis to subsequent periods. Analysing the five-year time interval, it has been shown that changes in well-being-related living conditions do not occur quickly.

The directions for future research will largely aim to overcome the above-mentioned limitations. We will also try to evaluate the impact of the COVID-19 pandemic on the well-being-related living conditions of elderly people, possibly with the other set of indicators. It is important to monitor all phenomena related to human activity, especially social phenomena.

The study shows policy implications. The process of positive changes in people’s quality of life does not happen quickly. Governments must take into account the fact that improving well-being-related living conditions is a long-term process. This should be borne in mind, especially in the context of the life expectancy of older people. The social policy in the area of the well-being-related living conditions of elderly people in the EU should concentrate on decreasing the inequalities and the risk of poverty because the values of these indicators (percentage of persons aged 65+ years at risk of poverty or social exclusion and income quintile share ratio S80/S20 for disposable income of people aged 65+ years) deteriorated during the analysed period. The values of other indicators improved, so they do not need that much attention. In general, the well-being-related living conditions of elderly people depend directly on a country’s level of social and economic development—well-developed Western and Northern European countries have a much better situation than the less-developed Eastern and South European ones. Therefore, the basic method of improving the well-being-related living conditions of elderly people is to maintain steady and sustainable social and economic growth. Policymakers and authorities in CEE countries should improve ageing policies covering pensions and social and health care services. Such measures can contribute to reducing gaps in quality of life. Our recommendations are general. More detailed recommendations would require analysing every country separately and in detail, which can also be the direction for future research.