Multi-Criteria Assessment of Transport Sustainability in Chosen European Union Countries: A Dynamic Approach

Abstract

The main aim of this article is to dynamically evaluate the sustainable development of transportation as an important economic sector in each member state of the European Union. Furthermore, the authors tried to identify underdeveloped spatial areas and indicate related trends in particular countries. To address this research topic, a multivariate-order statistical measure was implemented. The data sources of the study were Eurostat databases. The rankings of the chosen European Union countries for transport sustainability and its individual components (pillars and orders) were obtained for 2016–2019. This allowed the underdeveloped space regions and their individual pillars in the field of transportation sustainability to be identified in an appropriate manner. Then, the total (general) synthetic measures applicable to the entire period of analysis were constructed. It should be noted that the initial set of diagnostic variables and its classification in certain sequences were implemented. Furthermore, the taxonomic method applied with Weber’s multivariate median was first used to dynamically assess aspects of traffic sustainability. Such synthetic methods allow for analysis of the interaction of different areas of complex transportation systems and allow distortions of the diagnostic variables.

1. Introduction

Transport plays a key role in everyday life, being considered one of the main factors of socioeconomic development. Furthermore, it is vital for social welfare, living standards, and quality of life. Without accessibility, people would not be able to physically access jobs, health resources, education, and other important necessities. Consequently, their quality of life would be negatively affected [1].

This article is a continuation of the research carried out in the article entitled “Quantitative Analysis of Sustainable Transport Development as a Support Tool for Transport System Management: Spatial Approach” [2]. The published article evaluates the sustainability of transport using two multidimensional research methods (i.e., classical and positional at the microlevel) in the form of individual Polish provinces for a selected period of time (years).

This study describes the international aspect in the form of selected countries from the European Union. The superiority of the positional method for constructing a synthetic measure using A. Weber’s median was demonstrated, and it was decided to narrow the study to this type of statistical tool. (The same method was used, so the description of the method is very similar to the description in the previous article.) In addition, a dynamic approach (4 years covered) is included in the present work. In addition to static assessments, an assessment was also made for the entire study period by constructing spatiotemporal models for both the phenomenon of transport sustainability and its individual pillars: environmental, social, and economic.

From this point on, transport is considered the basis for both normal functioning territories, such as the European Union, and their development [3,4]. In addition, it is considered one of the main links introduced into the process of the movement of people, raw materials, and goods from the place of origin to the destination. As a result, transport is considered one of the cornerstones of globalization [5].

However, all transportation activities generate positive and negative externalities. The first orientation includes positive aspects, such as achieving economic objectives, namely producing a wide range of products that consumers need and delivering them to the market. In addition, the essence and advantage of each highly developed modern economy is well-developed and effective transport. This applies to all types of transport: road, rail, and air. In other words, a high level of transport is the healthy bloodstream of the developing economies of the modern world. As a result, transportation is considered to be a fundamental factor in macroeconomic and microeconomic development. The second area brings a wide range of problems, including greenhouse gas emissions [6], which results in climate change [7], such as global warming. Moreover, other changes are also suggested, including the effects of climate change on terrestrial, marine, and freshwater biological systems, the impact on human health caused by excess heat, and changes in infectious disease vectors [8]. Additionally, it should be noted that annual air pollution causes more than 0.4 million premature deaths [9]. Consequently, this has implications for human health and affects the public health system. Therefore, sustainable mobility systems are vital to the health of citizens and economies, allowing them to move freely and safely while respecting the environment [10]. All this calls for policies to reduce pollution, such as carbon dioxide and noise emissions, as well as oil consumption. However, key activities in that area include so-called decarbonization, which is understood by the European Union as the elimination of carbon dioxide emissions [11].

The area of the European Union is very diversified due to both different levels of socioeconomic development and pollution from the natural environment. Thus, all activities must be balanced so as not to undermine the economy of the member states and help the environment and society of specific countries. All resources, among them being financial funds, are limited and should be used most efficiently. Furthermore, making decisions has many drawbacks. This includes biased decisions, time-consuming analyses, and an unpredictable future [12]. Moreover, the evaluation of sustainable transport today faces the following challenges: the first is to understand and define its indicators, and the second is to quantify and operationalize them [13]. In general, however, transport is considered to be the main factor in reducing spatial and social differences [14]. Therefore, a correct diagnosis of transport sustainability is needed in particular areas of the European Union. Moreover, it can be treated as a support tool for infrastructure management and should be allowed to implement the most efficient support policy.

The main purpose of this article is to evaluate the sustainable development of transport as an important branch of the economy in select countries of the European Union dynamically. Moreover, the authors tried to identify underdeveloped spatial areas and indicate the related trends in particular countries.

This paper focuses on a broadly based analysis of the dynamics of transportation development and the application of economic and sustainable development policy principles. The specific objective is to evaluate the sustainability of transportation development in certain years. Then, a total (dynamic measurement) was built for the sustainability of transport and its specific pillars (e.g., the environment, society, and economy). This phenomenon is identified as an important factor contributing to the integration of the world economy. The first part discusses select aspects of sustainable transport development. Then, sustainable transport development measures are considered on the basis of research experience. Examples of sustainable transport indicators and indicators developed for transport management needs are provided. This article also examines the potential use of tools identified in business practice and outlines directions for future research. Furthermore, the synthetic indicators for individual years of analysis formed the basis for calculating the total synthetic indicators (gross) for the entire period.

In the investigation of research problems, select multivariate statistical tools were introduced into the investigation process. It should be noted that the ordinary taxonomic method was introduced using Weber’s multidimensional median. This synthetic construction allowed us to consider interactions in various areas of the transport system. Additionally, the use of this median, except for the interactions included in the study, makes the analysis resistant to distortions in diagnostic variables. The source of information in the research is data drawn from the Eurostat database from 2016 to 2019.

2. Background of the Literature

Research shows that sustainable transportation concepts are well known and have been used in the scientific literature for many years. However, it remains a subject of scientific debate, and new ways and tools for effective development management at all levels are constantly being explored.

First, sustainable transport development must be seen as a broader approach, such as a narrow field of sustainable development. However, the concept of sustainable development was developed as a response to the need to combat environmental degradation and quickly incorporates the social and economic aspects of development [15]. Since then, it has been enriched by input from a variety of scientific disciplines. It will be an important model for all programs and policies, as well as international organizations, national governments, and local authorities for creating development strategies. Its main objective is still to ensure the improvement of the long-term quality of life of current and future generations by rational proportional arrangements between different types of capital: the economy, humanity, society, and nature [16]. It should also be noted that the concept of sustainability applies to all sectors of public life and to all economic sectors, such as construction [17] and transportation [18].

Therefore, transportation is one of many strategic directions in sustainable development policy. First, it determines the competitiveness of the country’s economy [19,20]. Due to the efficiency of transport services and infrastructure, the potential to use the economic potential in a particular region is determined by efficiency [21]. Second, it is a source of many important external costs negative to both society and to the environment [22,23]. The negative impacts of transportation on the basic elements of sustainable development (e.g., the elements of the environment, society, and the economy) have already been recognized in the literature [24]. The first is the reduction of air and water pollution, loss of habitat, and water effects and the eradication of non-renewable resources. The second area involves the following phenomena: the influence of inequality, the lack of mobility, the impact on human health, the interaction and responsibility of communities, and aesthetics. The last area, therefore, is closely linked to traffic jams, mobility barriers, accident damage, the cost of facilities and consumers, or the decline of non-renewable resources. Therefore, transportation today is facing many changes in terms of sustainability.

The sector has developed sustainable policies and practices in response to the negative situation in the transport sector. Therefore, modern transport systems must be constructed and formulated according to the needs of sustainable development strategies [25]. Literature research has provided a wide range of strategies to increasing the sustainability of transport and introducing the process of decarbonization. All activities are based on a number of strategic objectives, including reducing energy emissions from transportation, reducing fuel costs, reducing carbon dioxide, and increasing energy intensity [26,27,28]. It is worth mentioning that most of the energy consumed for transport purposes comes from petroleum, which leads to the limitation of natural fuel resources. Therefore, changes in the structure of fuel production can be observed, which results in alternatives such as electromobility, hybrid technologies, biohydrocarbons, or biofuels such as methyl esters or bioethanol [29,30,31,32,33]. It is worth mentioning autonomous driving with artificial intelligence technologies, which is under scientific research [34], and its impact on both traffic and the environment. However, from the perspective of the examined subject matter, it is worth putting additional attention in Poland on alternatives to fossil fuels and their current situation. The current market for alternative fuels in Poland can be considered as underdeveloped. This is due to the fact that there were only over 9000 electric cars on the roads at the end of 2019. Moreover, the lack of charging points is observed as well. Furthermore, there were only approximately 4000 passenger cars using compressed natural gas [35]. Due to the fact that agriculture plays an important role in Poland’s energy security, the promotion of energy crop farming can grow dynamically [36]. Nevertheless, the share of biofuels in Poland can be considered insufficient. According to the analyses published in the literature [37], liquid biofuels are still used on a small scale. Nevertheless, the production of bioethanol as a biocomponent in traditional fuels is growing. Moreover, the structure of raw materials for bioethanol has already changed. Many technological processes use organic waste or byproducts from agriculture production.

In the scope of the research subject, it is worth mentioning hydrogen. Huge potential for its use as an alternative to conventional transport fuels has been observed. According to many predictions, the share of hydrogen in the structure of global consumption will be dominant at the end of the century. Therefore, Poland may be the user of blue hydrogen, which is received by its separation from natural gas. This kind of fuel could be transported through gas pipelines such as the Baltic Pipe form Norway, which is focused on its separation from gas, from which carbon dioxide is removed by means of CCS technology [38].

Overall, the transportation system balancing process must be balanced and carried out with the same intensity in all dimensions of the transport system. Therefore, in any research activity, it seems that an appropriate definition of sustainable transportation is essential. However, studies in the literature have shown that the definition of sustainable transport, which is universally accepted, is not yet theoretically developed. Yet, many attempts have been made to define the concept of transportation sustainability. Sustainable transportation can be considered a system of transport of goods and passengers, taking into account environmental, social, and economic factors at the same time, such as access to human and work centers at a reasonable price and transport options, at affordable prices and by socially acceptable means [39]. Further studies have shown that it is possible to distinguish a narrow approach to the definition of transport sustainability from a broad approach [24]. However, a narrow approach highlights the environmental aspects of the equilibrium of transport. The most important conditions for implementing a sustainable transport system are (1) the protection of human health and maintenance of ecosystem integrity, (2) compliance with health and environmental restrictions, (3) the prevention and reduction of emissions, (4) sustainable use of renewable and nonrenewable resources, and (5) the prevention of anthropogenic change in the world’s ecosystem [40]. In a narrow sense, sustainable transportation is consistent with the well-accepted objectives of environmental health and quality and being compatible with the integrity of the ecosystem. It does not exacerbate adverse global events such as climate change or the loss of ozone in the stratospheric ozone. However, in general, the evaluation of sustainable transport is carried out within the framework of the integrated system [41], although it should be noted that there is an emerging consensus that the issues of transportation system sustainability should be incorporated in transportation planning to have any policy-based effect on decision making [42,43].

Considering the above considerations, the objective of transport sustainability is summarized in the following aspects, such as the environment, society and the economy [44,45,46,47]. This process leads to a global order of transportation systems and their relationship to the environment. Integrated order means that a transport system is in its process of development, and the level of the individual partial order (social, economic, and environmental) has reached the desired level. All of these are very coherent, giving it a new qualitative dimension: integration into the systemic dimension, and not only in part [48]. Such a complete order transportation system can only be regarded as modern. It meets the needs of the current generation in the transport sector, focusing on the environment, economy, and society, and meets the environmental, economic, and social expectations of the current generation.

In the literature review, sustainability of transport is proven to be a continuous process of change, and achieving complete results requires careful management and control. Consequently, to implement a sustainable transport development policy, appropriate measures must be selected in the evaluation process. It should be noted that the requirements for sustainable transport make it difficult to measure sustainability. Furthermore, the multidimensional nature of transport phenomena causes many differences in the statistical study of spatial objects with the help of various diagnostic variables.

Two approaches should be taken into account to assess spatial objects in the field of transportation [49]. First, one-dimensional analysis should be performed in which the level of development of transport should be evaluated separately depending on the variables. Second, taxonomic tools are being introduced to enable the management and classification of research objects that are characterized by many variables to support a logistics policy. However, reviews of the literature have shown that sustainable transportation is already under scientific study [50,51,52,53,54]. Based on the literature reviews, the following areas of social transport, ecological transport, economic transport, and innovative transport should be highlighted as a methodology to develop smart indicators [55]. Recent studies have already been published in the field of sustainable transport, implementing various research methods in the process, including taxonomic measures. The results of the reviews of the collected literature indicate that various research methods, such as fuzzy logic, PCA, and equal weighting, are used to construct composite indices [56], or a three-step procedure is applied [57]. Regarding the review of the literature, it should be recalled that the same weight of diagnostic variables is often applied in such an analysis [58,59]. The literature study proved that there are many methods which can be used in order to assess transport sustainability (e.g., the report delivered by the Transport Department of Canada proposed 42 research methods). There are, among the proposed methods, projects for evaluation, cost–benefit analysis, impact assessment, multicriteria determination analysis (MCDM), the Delphi method, surveys, and indexing [60]. All in all, studies in the literature proved that there is no standard method for transport sustainability assessment. Most methods are based on multidimensional evaluations of economic, environmental, and social impacts. Such a multidimensional nature of sustainability proves that multicriteria methods would be more appropriate for transport sustainability assessment than single-criterion methods [42,43]. Therefore, one of the most popular methods in the transportation sustainability assessment process is the construction of a composite index [61]. Such a composite index in taxonomy science is generally called a synthetic measure.

Spatial research on transport development with the implementation of taxonomic methods has already been implemented in the scientific investigation process [62,63]. The problem of including no directly observed relationships (interactions) in the transportation research process is very important, as has already been demonstrated in studies in the literature [64,65,66]. Furthermore, it should be noted that the most recent research on the implementation of linear order in urban transport has been carried out [67].

It should be noted that empirical and spatial evaluations of transportation development have already been performed using taxonomy methods, which take into account the interactions of the set of diagnostic variables [68,69,70,71]. All analyses relate to a general assessment of transportation development and do not relate to the issue of transportation sustainability.

The first research on transport sustainability and the non-direct observation of relations took place in Polish and international aspects [2,72]. However, the above analysis is static; that is, it does not take into account the changes in the phenomena over time. It is worth analyzing the impact of time on the analyzed phenomena of transportation sustainability, taking into account not only the relationships that cannot be directly observed but also the analysis that is worth considering.

In summary, regarding the considerations considered in the field of transport sustainability, it should be noted that development problems are widely known in the scientific literature. However, it remains the subject of scientific debate and research aimed at finding more effective methods to implement spatial object assessment. However, existing transportation sustainability analyses do not simultaneously include the interaction of a set of diagnostic variables and dynamic aspects. This requires a comprehensive approach because there are already links and connections in the field of research. Due to the many and difficult-to-quantify effects of these relationships, knowledge of the impact of sustainable development policies in the field of transfer is not complete.

3. Presentation of the Research Method and Collection of Potential Data Sets

The proposed research method is focused on assessing the sustainability of transport with the implementation of the synthetic measure in the evaluation process. It can be built in two ways. The first method uses standard methods that use statistical indicators such as averages and standard deviations. The construction of this type of synthetic measurement is well described in the literature [73]. The last sequential method introduces a multidimensional median vector and insanity. This method has been applied in the research processes of Polish statisticians [74].

Diagnostic variables describing the statistically investigated phenomenon are standardized according to this equation [75]:

$$z_{ij}=\frac{x_{ij}-{\theta }_j}{1.4826\times mad\left(X_j\right)}$$

(1)

where $x_{ij}$ is the value of the variable j for the i-th member state, i = 1, 2, …, n for the i-th member state, n is the number of chosen member states, $j=1,2,\mathrm{\dots }, m$ for the diagnostic variable j, m is the number of diagnostic variables, ${\theta }_j$ is a particular value of a multidimensional medium vector $\Theta$ (border or Weber), 1.4826 is the constant coefficient estimated in empirical studies, and $mad\left(X_j\right)$ is the median absolute deviation of the j-th variable.

As far as the $\Theta$ vector is concerned, its partial value (element) ${\theta }_j$ is considered to be a multidimensional middle vector, such as border or Weber. In contrast, the cutting vector average estimates are based on each variable’s unique analysis median calculation. These medians are then considered individual elements of a specific median vector. Weber’s average value is calculated by removing the following process:

$$T\left(\Theta ,R^m\right)=argmi{n}_{\mathit{\Theta }\in R^m}\left\{{\displaystyle \sum }_{i=1}^n{\left[{\displaystyle \sum }_{j=1}^m{\left(x_{ij}-{\theta }_j\right)}^2\right]}^{1/2}\right\}$$

(2)

where m is the number of diagnostic variables, $R^m$ is the m-dimensional space of real numbers, ${\theta }_j$ is the particular values of multidimensional Weber median $\Theta$, $j=1,2,\dots ,m$ for the j-th diagnostic variable, $\Theta =\left({\theta }_1,{\theta }_2,\dots ,{\theta }_m\right)$ for the multidimensional Weber median, $x_{ij}$ is the value of the variable j for the i-th member state, i = 1, 2, …, n for the i-th member state, and n is the number of member states included in the investigation.

When Weber’s formula is mentioned above, the distance between the points of the multidimensional space is minimal, so the total distance between these points and other objects of the space must be determined. Such multidimensional medium vectors resist distortion and take into account interactions. It should be noted that when synthetic measurements are created, such relationships are very important from the perspective of taxonomic analysis of complex phenomena such as socioeconomic development, living standards, spatial cohesion, or sustainability of a transportation system.

In relation to the median absolute deviation (mad), it should be noted that its particular values are received with the following equation:

$$mad\left(X_j\right)=\underset{i=1,2,\mathrm{\dots },n}{med}\left|x_{ij}-{\theta }_j\right|$$

(3)

where med is the median, $x_{ij}$ is the value of the variable j for the i-th member state, i = 1, 2, …, n for the i-th member state, n is the number of member states included in the investigation, ${\theta }_j$ is the values of the multivariate median vector $\Theta$ (border or Weber), and $j=1,2,\mathrm{\dots },m$ for the j-th diagnostic variable.

In addition, attention should be paid to the methods of dust normalization in the literature, which have been successfully introduced into the research process with classical statistical indicators [76,77,78].

Furthermore, median vectors have already been successfully applied in the process of normalizing data sets in consumption research [79] or transport assessment [68,69].

Finally, the synthetic measure of normalized variables, in the case of the order method, is estimated with the following formula:

$$M{P}_i=1-\frac{d_i}{med\left(D\right)+2.5mad\left(D\right)}$$

(4)

where d_i is the individual values of the distance vectors D, D = (d₁, d₂, …, d_n) is the distance vector, n is the number of member states included in the investigation, med(D) is the median of the distance vector D, mad(D) is the median absolute deviation of the distance vector D, and 2.5 is a constant value (immune threshold value).

Individual elements of the distance vector can be obtained using an urban metric, also known as a taxi, according to the following formula:

$$d_i={\displaystyle \sum }_{j=1}^m\left|z_{ij}-{\varphi }_j\right|$$

(5)

where ${\varphi }_j$ are considered elements of this development model $\mathsf{\Phi }=({\varphi }_1,{\varphi }_2,\dots ,{\varphi }_{m)}$. Its individual values were determined according to the method provided by A. Mlodak [80]; that is, the maximum value of the standard variable of the stimulants, the minimum value of the stimulants, and the nominal value of the nominants were assumed. The synthetic measure allows us to prepare a ranking of objects that are under investigation in the context of transport sustainability as well as divide them into similar groups. To create a group of the same member states, three median methods can be applied according to the following formula:

$$\begin{array}{l}\mathrm{I} \mathrm{group}: \left\{iϵ\left(1,2,\dots ,n\right):M{P}_i>me{d}_1\left(MP\right)\right\}\\ \mathrm{II} \mathrm{group}: \left\{iϵ\left(1,2,\dots ,n\right):med\left(MP\right)<M{P}_i\le me{d}_1\left(MP\right)\right\}\\ \mathrm{III} \mathrm{group}: \left\{iϵ\left(1,2,\dots ,n\right):me{d}_2\left(MP\right)<M{P}_i\le med\left(MP\right)\right\}\\ \mathrm{IV} \mathrm{group}: \left\{iϵ\left(1,2,\dots ,n\right):M{P}_i\le me{d}_2\left(MP\right)\right\}\end{array}$$

(6)

where $M{P}_i$ is the value of the order synthetic measure for the i-th member state, $med\left(MP\right)$ is the median value of all orders synthesized in the system, $me{d}_1\left(MP\right)$ is the median value of the order of the synthetic measure not less than $med\left(MP\right)$, and $me{d}_2\left(MP\right)$ is the median for the values of the order synthetic measures not greater than or equal to $med\left(MP\right)$.

The basis for every proper multidimensional analysis is a set of diagnostic variables. Therefore, it is worth mentioning T. Litman, whose scientific work on appropriate indicators brought about a great deal of knowledge in relation to the process of evaluating sustainable transports [81,82,83]. Furthermore, a very interesting approach in the scientific literature is related to the construction of intelligent transportation indicators in the following areas: social, environmental, economic, and innovative [84]. However, the diagnostic variable group must include scientific principles, objectives, systems, and comparable and feasible principles [85].

The investigation process is based on data obtained from the European Statistics Database (Eurostat) 2016–2019. It should be noted that the construction process has two main stages in a diagnostic variable set. The first is based on the theories of observed phenomena and creates a set of diagnostic variables that can potentially be determined. Subsequently, it applies basic statistical measures to obtain the final diagnostic variables.

Consequently, the three main areas of sustainable transportation development, measured from the point of view of the integrated order or pillar, were taken into account.

The first of the indicated areas (i.e., environmental order) is presented by the following variables: X₁ is the average CO₂ emissions per km of new passenger cars (g CO₂ per km), X₂ is the exposure in urban areas to air pollution by particulate matter <10 µm (µg/m³), X₃ is the share of renewable energy in the final energy consumption (%), X₄ is the share of electricity consumption in the total energy consumption of the transport sector (%), X₅ is the total length of the electrified railway track (%), X₆ is the share of natural gas consumption in the total energy consumption of the transport sector, X₇ is the carbon dioxide emission in the area of transportation and storage (tons per capita), X₈ is the proportion of new passenger cars registered as a whole (%), X₉ is the emission of nitrous oxide in the area of transportation and storage (grams per capita), and X₁₀ is the emission of methane in the area of transportation and storage.

The second area, social order, is represented by the following variables: X₁₁ is the number of passenger cars per 1000 citizens, X₁₂ is the number of deaths in road accidents per 100,000 citizens, X₁₃ is the motor coaches, buses, and trolley buses per 1000 citizens, X₁₄ is the number of deaths in rail accidents per 1 million citizens, X₁₅ is the share of trains in total passenger transport (%), X₁₆ is the share of the population living in households considering that they suffer from noise (%), X₁₇ is the exposure in urban areas to air pollution by particulate matter < 2.5 µm (µg/m³), X₁₈ is the share of motor coaches, buses, and trolley buses in total passenger transport (%), X₁₉ is the number of airline passengers per one citizen, and X₂₀ is the number of people killed in commercial air transport per 1 million passengers.

The third area (i.e., economic order) is presented by the following variables: X₂₁ is the goods transported by rail in thousands of tons per 1000 citizens, X₂₂ is the length of railway lines per 1000 km², X₂₃ is the share of environmental taxes in the total tax revenues (%), X₂₄ is the air transport of goods in tons per 1000 citizens, X₂₅ is the share of renewable energy in the transport sector (%), X₂₆ is the goods transported by roads in thousands of tons per 1000 citizens, X₂₇ is the number of lorries and road tractors per 1000 citizens, X₂₈ is the share of rails in the total freight transport (%), X₂₉ is the share of roads in the total freight transport (%), and X₃₀ is the total length of two or more railway lines (%).

Therefore, 30 variables were collected for further taxonomic analysis. This shows that the number of indicators that are essential to capture the multidimensionality aspects of transportation is quite large [86]. Nevertheless, not all member states of the European Union were included in the research process. This was due to the problem of data availability, especially in the case of international research. Therefore, three member states were not included in the transport sustainability assessment. However, two of them (i.e., Cyprus and Malta) do not have railway transport, which is considered one of the most environmentally friendly menaces of transport and should not be excluded from research. Furthermore, Belgium was also omitted due to significant incomplete data. However, to not exclude some countries from the research process, the missing data were completed by time series extrapolation and interpolation, but this was only in a few cases.

It should be noted that collecting data for the member states of the European Union in the form of a set of diagnostic variables for different periods of time brings many difficulties. In many cases, such analysis can significantly reduce both the number of diagnostic variables and the number of research objects (i.e., member states). This is due to the fact that the statistical data in the case of Eurostat are published with a long delay and are supplemented over time in the case of various European Union member states. Therefore, including the latest information into the research process often becomes a very difficult task and sometimes can be impossible. Furthermore, it is worth mentioning that the researcher must consider cost effectiveness during data collection [42,43].

4. Statistical Verification of the Potential Data Set and Choice of Total Taxonomic Model

According to the theory of synthetic measure construction, every set of potential diagnostic variables should be subjected to statistical investigation. Usually, statistical verification of potential variables contains two main stages: variation and correlation analysis.

The first was conducted in order to check the variation and make a choice for the type of the total (dynamic) model. Therefore, to accomplish this, the classic variation coefficient which is based on both location and differentiation measures, such as the arithmetic mean and standard deviation, is commonly used. However, proper implementation of such measures should be preceded by examination of the skewness of particular variables. The results of such analysis for the three particular pillars (orders) of sustainable transport development—environmental, social, and economic—are presented in

Table 1. Chosen statistical measures for particular pillars of transport sustainability.

Variable	2016				2017				2018				2019
	As	$\overline{\mathit{x}}$	MB	MW	As	$\overline{\mathit{x}}$	MB	MW	As	$\overline{\mathit{x}}$	MB	MW	As	$\overline{\mathit{x}}$	MB	MW
	Environmental
X1	−0.3	118.9	120.8	119.3	−0.3	119.4	120.7	119.7	−0.4	120.9	121.9	120.9	−0.7	122.9	124.0	123.4
X2	0.5	21.7	20.7	22.1	0.3	21.8	22.3	22.9	0.2	22.3	21.5	22.9	0.1	20.2	20.3	20.4
X3	0.7	31.2	29.5	29.7	0.6	32.2	31.3	31.4	0.7	33.2	32.8	32.0	0.7	34.9	33.8	33.4
X4	0.5	1.4	1.3	1.5	0.5	1.4	1.2	1.4	0.6	1.4	1.2	1.4	0.7	1.4	1.2	1.4
X5	−0.1	47.6	48.7	47.6	−0.3	49.4	55.0	51.1	−0.3	49.5	55.1	51.4	−0.3	49.9	55.1	51.8
X6	2.2	1.2	0.5	1.4	2.3	1.3	0.5	1.0	2.1	1.2	0.6	1.0	2.2	1.2	0.7	1.2
X7	2.7	1.5	0.8	1.0	2.6	1.6	0.8	0.9	2.6	1.6	0.8	1.0	2.5	1.7	0.8	1.0
X8	0.6	6.1	6.2	5.8	0.6	6.3	6.1	6.1	0.5	6.4	6.2	6.2	0.6	6.3	6.1	6.0
X9	2.6	52.0	31.1	36.2	2.6	52.6	35.5	36.7	2.5	54.5	36.2	37.0	2.4	54.6	34.1	35.9
X10	4.0	0.6	0.1	0.7	4.0	0.6	0.1	0.5	3.9	0.6	0.1	0.5	3.9	0.6	0.1	0.6
	Social
X11	−0.2	479.3	479.5	479.6	0.0	489.3	487.0	488.4	0.1	501.1	503.0	504.8	0.1	513.8	511.5	515.3
X12	0.6	5.8	5.4	5.1	0.7	5.4	5.1	5.1	0.4	5.5	5.3	5.2	0.5	5.3	4.9	4.9
X13	0.3	2.1	1.9	1.7	0.3	2.1	2.0	1.8	0.2	2.1	2.0	1.9	0.3	2.2	2.0	1.8
X14	1.4	2.7	2.1	1.7	1.5	3.1	2.0	2.3	1.4	2.7	2.1	2.0	1.6	2.4	1.6	1.5
X15	0.1	6.0	5.9	7.0	0.1	6.1	5.7	6.0	0.1	6.1	6.0	5.7	0.1	6.2	6.3	6.2
X16	0.4	15.7	15.4	17.7	0.6	15.5	13.9	17.9	0.6	15.6	14.7	17.8	0.5	15.3	14.1	16.6
X17	0.0	14.0	13.8	12.0	0.1	14.1	13.4	11.9	0.1	14.0	13.3	12.1	−0.1	12.1	11.9	11.7
X18	−0.1	12.3	12.1	9.8	0.0	11.9	12.1	9.6	0.0	11.8	11.9	9.9	0.0	11.7	11.9	9.9
X19	0.7	2.7	2.3	3.1	0.6	3.0	2.5	3.6	0.6	3.2	2.6	3.7	0.6	3.4	2.7	4.0
X20	4.9	0.1	0.0	0.0	4.2	0.0	0.0	0.0	4.6	0.0	0.0	0.0	3.2	0.0	0.0	0.0
	Economic
X21	1.5	6.3	5.0	5.7	1.3	6.5	4.9	6.0	1.5	6.8	4.8	6.2	1.3	6.3	4.9	5.9
X22	0.9	54.0	46.3	50.0	0.9	53.9	46.3	49.8	0.9	54.0	46.3	50.1	0.9	54.1	46.4	50.5
X23	0.1	7.6	7.8	7.8	0.1	7.4	7.6	7.6	0.1	7.2	7.3	7.4	0.1	7.1	7.0	7.3
X24	4.8	76.6	11.9	17.4	4.9	82.8	13.4	18.6	4.9	82.1	14.2	18.9	4.8	77.3	11.7	18.4
X25	2.8	6.5	6.3	6.6	2.4	7.2	6.6	7.3	3.1	8.1	7.0	8.1	2.9	9.1	7.9	9.1
X26	1.9	32.4	31.0	29.1	1.2	33.0	31.8	30.3	1.3	33.4	32.1	30.4	1.6	34.2	33.7	31.0
X27	0.4	4.6	4.2	4.6	0.5	4.9	4.4	4.9	0.6	5.2	4.8	5.2	0.8	5.4	4.8	5.4
X28	1.3	24.1	22.5	23.9	1.2	24.3	22.0	24.1	1.3	24.6	24.0	24.3	1.3	23.8	23.4	23.7
X29	−0.6	70.1	72.1	71.2	−0.6	70.2	72.6	71.2	−0.6	70.1	71.6	71.2	−0.6	70.4	71.8	71.4
X30	0.6	32.0	27.2	30.6	0.6	32.1	27.2	30.6	0.6	31.6	26.1	30.2	0.7	32.0	27.2	30.4

Notation: A_S = skewness, $\overline{x}$ = arithmetic mean, M_B = border median, and M_W = Weber median. Source: own studies.

.

The results presented in Table 1 show that some diagnostic variables were heavily biased for each test order. First, the environmental area of transport sustainability was characterized by the following four highly skewed variables: X₆, X₇, X₉, and X₁₀. Their values were outside the usually acceptable interval (−2.2). Second, the social order had one highly skewed variable: X₂₀. Third, the economic area included variables such as X₂₄ and X₂₅ which could be considered highly skewed. Furthermore, it should be mentioned that this kind of strong asymmetry was observed in every year of research.

Table 1 shows that some diagnostic variables were highly biased in all sequences studied. First, the environmental area of transport sustainability was characterized by the following four highly skewed variables: X₆, X₇, X₉, and X₁₀. Their values were outside the usually acceptable interval (−2.2). Second, the social order had one highly skewed variable: X₂₀. Third, the economic area included variables such as X₂₄ and X₂₅ which could be considered to be highly skewed. Furthermore, it should be mentioned that this kind of strong asymmetry was observed in every year of research.

Overall, the implantation arithmetic mean seems to be insufficient for such analysis. Consequently, the classical method of synthetic measure construction should not be implemented either. Therefore, in order to bring the analysis to reality, a median, which is considered the order substitute of the arithmetic mean, was introduced in two forms: border and Weber versions. Analysis of their values showed that in the case of variable X₂₀, the border median was equal to zero. This was due to the extreme asymmetry of the empirical distribution of that variable, which made it impossible to calculate the value of the order variation coefficient. Therefore, the authors decided to implement a multidimensional Weber median.

Finally, in order to facilitate the analysis of both forms, the variation coefficient was implemented (i.e., the classic version and its order version with only the Weber median). Their particular values for each of the pillars analyzed for sustainable transport development, as well as each year of analysis, are presented in

Table 2. Variation coefficients of diagnostic variables.

Variable	2016		2017		2018		2019
	VS	VW	VS	VW	VS	VW	VS	VW
	Environmental
X1	7.11	5.66	6.81	5.81	6.83	6.04	7.73	6.48
X2	31.51	17.60	33.55	17.91	28.50	18.08	28.28	19.40
X3	58.71	45.78	57.65	43.89	55.86	40.44	53.81	39.10
X4	63.01	47.21	61.52	41.46	62.56	40.94	62.35	35.91
X5	51.29	35.70	49.74	37.24	49.44	37.98	48.12	35.74
X6	142.39	78.54	143.88	77.02	130.09	73.68	127.51	73.16
X7	130.87	35.75	126.99	35.00	127.65	34.26	125.66	36.40
X8	42.60	24.49	42.89	21.29	41.41	24.13	40.35	28.09
X9	108.16	46.72	106.26	39.35	107.84	41.68	108.03	38.99
X10	258.21	95.70	261.84	93.66	255.65	93.37	255.18	94.33
	Social
X11	19.46	11.02	18.95	13.12	18.25	11.69	17.86	11.80
X12	33.03	23.61	36.48	26.73	32.99	24.45	36.43	28.20
X13	41.01	39.33	40.48	37.39	40.66	33.19	41.03	36.94
X14	91.11	75.27	82.75	48.45	81.59	55.87	93.40	61.51
X15	57.44	40.55	58.45	51.32	57.82	56.21	58.41	55.42
X16	30.87	24.49	33.37	27.35	35.17	22.58	35.59	22.32
X17	37.61	27.04	40.28	40.08	35.66	32.17	32.49	22.72
X18	36.51	35.75	38.22	38.19	36.96	29.83	36.60	33.04
X19	63.34	44.60	60.64	41.84	58.83	37.84	58.39	38.45
X20	465.91	100.00	381.05	100.00	404.95	100.00	300.33	100.00
	Economic
X21	100.18	63.60	98.46	66.66	102.32	66.39	95.51	68.54
X22	56.22	42.48	56.26	42.22	56.33	40.95	56.25	41.37
X23	25.86	18.51	25.80	19.38	25.32	19.96	25.57	21.69
X24	366.53	74.21	368.47	72.40	365.33	69.88	362.82	72.39
X25	78.14	23.51	76.35	18.57	67.58	19.21	61.88	15.74
X26	48.89	30.45	44.81	30.02	44.68	30.47	47.41	31.95
X27	59.41	47.51	60.49	46.79	62.74	48.16	65.28	47.49
X28	76.32	40.55	75.76	42.65	76.01	43.35	76.49	47.44
X29	28.10	19.95	27.85	21.08	27.91	20.79	27.86	20.43
X30	51.66	37.68	51.91	37.20	52.06	36.23	52.06	36.58

Notation: V_S = classic variation coefficient and V_W = order variation coefficient based on Weber median. Source: own studies.

.

The analysis of the presented data showed that only one variable should be removed from further analysis: X₁. This is due to the fact that the values of both variation coefficients were below 10%.

In the scope of the transport sustainability evaluation process, over the entire period of time which was put under statistical investigation, the total synthetic measure should be implemented. In relation to the literature study, there are three types of taxonomic models—spatial-time, spatial-spatial, and aggregate—which can be used to evaluate the spatial object over the entire period of time [80]. Research processes based on variance coefficients have shown that spatial time models should be integrated into the analysis process.

The second phase of statistical verification of possible diagnostic variables is the verification of the second stage; in other words, correlation analysis is carried out using the reverse matrix method of the Pearson’s correlation coefficient [87]. To check which variables carried the same information, three separate inverted matrixes were calculated in relation to the sets of diagnostic variables in the area of particular orders of transport sustainability. The results of the correlation analysis in the form of the main diagonals of the inverted Pearson correlation matrices are presented in

Table 3. Main diagonals of inverted matrices of Pearson’s correlation coefficients.

Variable	2016		2017		2018		2019
	I	II	I	II	I	II	I	II
	Environmental
X1	1.76	-	1.73	-	1.59	-	1.66	-
X2	1.21	-	1.26	-	1.29	-	1.32	-
X3	2.55	-	2.88	-	2.87	-	2.86	-
X4	1.81	-	2.15	-	1.96	-	1.98	-
X5	3.39	-	1.02	-	3.57	-	2.01	-
X6	9.84	-	8.15	-	7.80	-	8.11	-
X7	2.05	-	2.08	-	1.78	-	1.77	-
X8	7.60	-	7.24	-	6.77	-	6.63	-
X9	2.92	-	3.42	-	3.09	-	1.85	-
X10	1.76	-	1.73	-	1.59	-	1.66	-
	Social
X11	3.14	-	3.00	-	1.59	-	2.39	-
X12	10.01	-	6.88	-	4.91	-	5.98	-
X13	5.68	-	3.76	-	3.43	-	3.74	-
X14	2.45	-	2.23	-	2.25	-	2.82	-
X15	2.64	-	2.22	-	2.02	-	2.46	-
X16	2.01	-	1.93	-	2.24	-	2.84	-
X17	4.92	-	4.96	-	3.55	-	3.62	-
X18	3.12	-	3.42	-	2.91	-	2.92	-
X19	2.74	-	2.72	-	2.01	-	2.21	-
X20	1.38	-	1.43	-	1.60	-	1.61	-
X21	11.33	3.61	11.10	4.82	12.87	4.74	18.51	5.22
	Economic
X22	1.86	1.85	2.76	2.76	2.51	2.50	2.31	2.14
X23	2.59	2.54	3.33	3.29	2.56	2.48	2.64	2.30
X24	3.92	3.85	3.16	2.99	3.40	3.05	3.62	3.38
X25	2.44	2.34	4.50	4.39	3.12	3.09	2.56	2.55
X26	4.58	4.19	4.84	4.71	4.38	4.35	2.38	5.38
X27	2.44	2.40	2.58	4.82	2.59	2.47	2.36	2.34
X28	16.45	-	15.59	-	18.99	-	27.19	-
X29	5.15	3.84	6.00	2.47	6.40	4.77	7.01	4.50
X30	1.81	1.64	2.15	4.55	1.85	1.67	2.09	1.86

Notation: I = stage one and II = stage two. Source: own studies.

.

The investigation process proved that in both areas (i.e., the environmental and social pillars), there were no highly correlated variables. This is because the values located on the main diagonals of the inverted matrices were less than 10. It should be noted that only in the case of variable X₁₂ in 2016 was its value above the customary threshold value. However, with a slight deviation from the rule of 10, this variable can be successfully included in further analysis throughout the period. This is because the values of the inverted matrices in other years of the analysis confirmed the fact that this variable should not be excluded from the synthetic measure construction process. Further analysis of the correlation in the economic order demonstrated that only the X₂₈ variable should be left out of the research every year. Then, the Pearson correlation matrices for the economic order were inverted again. The new calculated values proved that other variables in this area could be included for further research. However, this kind of procedure for eliminating highly co-related variables has already been discussed in the literature [88].

In summary, regarding a statistical test of a possible set of diagnostic variables, it should be mentioned that 28 diagnostic variables were a base for synthetic measure construction for both the spatial and dynamic orders.

5. Assessment of Transport Sustainability and Discussion of the Research Results

To bring the final set of diagnostic variables to comparability, the normalization process was introduced. This transformation was based on standardization of the order with the Weber median. On the one hand, it should be mentioned that its version with a border median was omitted because this kind of median is only immune to skewness and does not consider interactions among diagnostic variables, which is very important from the point of view of research. On the other hand, implementation of a multidimensional Weber median allowed us to consider directly unobtrusive relationships between diagnostic variables in the process of designing a synthetic measurement of transport stability and made the analysis robust to distortion of the selected variables. Therefore, the multidimensional Weber vector was the basis for normalization in every year of analysis and in every pillar of the sustainability of transport, as well as in the integrated order evaluation. Furthermore, this type of median was implemented to normalize the variables in the assessment of the process of the pillars of transport sustainability and its integrated measure throughout the period.

The procedure to construct synthetic measures requires the nature of the variables to be established. Therefore, each of the three subgroups of transport sustainability was divided into three subgroups according to the final diagnostic variables. The first, the stimulants, included the following variables: X₃–X₆, X₈, X₁₃, X₁₅, X₁₈, X₂₁, and X₂₂. The second, nominates, consisted of the following features: X₁₁, X₁₉, X₂₄, X₂₆, X₂₇, and X₂₉. The last, the destimulants, included X₉, X₁₀, X₁₂, X₁₄, X₁₆, X₁₇, X₂₀, X₁₃, X₂₃, X₂₅, and X₃₀.

According to the methodology for compiling synthetic indicators, the implementation of the normalized variables of the three final sets made it possible to construct a synthetic indicator for each of the analyzed pillars (environmental and socioeconomic) separately for each year of the analysis. Consequently, the total measures for sustainable development (integrated order) were constructed for each year of the analysis, as well as throughout the period of the analysis. Their values for the chosen European countries are presented in

Table 4. Values of synthetic measures according to both particular orders and time periods for the countries analyzed.

Country	2016	2017	2018	2019	2016–2019	2016	2017	2018	2019	2016–2019
	Environmental					Social
Bulgaria	0.13	0.13	0.15	0.07	0.22	0.24	0.22	0.19	0.15	0.35
Czechia	0.28	0.27	0.28	0.28	0.45	0.40	0.44	0.36	0.35	0.64
Denmark	−0.34	−0.37	−0.43	−0.46	−0.56	0.46	0.50	0.41	0.30	0.68
Germany	0.32	0.30	0.33	0.35	0.52	0.28	−0.84	0.13	0.21	0.00
Estonia	0.18	0.18	0.21	0.20	0.32	0.67	0.58	0.56	0.67	1.00
Ireland	0.11	0.06	0.04	0.02	0.12	0.53	0.56	0.55	0.53	0.87
Greece	0.08	0.10	0.12	0.10	0.19	0.34	0.37	0.33	0.28	0.55
Spain	0.35	0.31	0.34	0.32	0.53	0.48	0.45	0.39	0.28	0.65
France	0.38	0.37	0.40	0.39	0.61	0.42	0.44	0.37	0.32	0.63
Croatia	0.32	0.29	0.36	0.35	0.52	0.19	0.16	0.12	0.17	0.29
Italy	0.47	0.46	0.50	0.49	0.75	0.27	−2.99	0.32	0.12	−0.72
Latvia	0.10	0.09	0.06	−0.02	0.12	0.15	0.16	0.12	0.09	0.24
Lithuania	0.16	0.14	0.09	−0.01	0.18	0.27	0.25	0.28	0.32	0.48
Luxembourg	−0.44	−0.43	−0.46	−0.49	−0.64	0.32	0.34	0.21	0.37	0.52
Hungary	0.38	0.42	0.43	0.42	0.65	0.16	0.14	0.17	0.11	0.26
Netherlands	0.25	0.23	0.25	0.25	0.40	0.37	0.36	0.23	0.23	0.49
Austria	0.66	0.63	0.64	0.65	1.00	0.31	0.41	0.36	0.33	0.59
Poland	0.35	0.30	0.31	0.30	0.50	0.20	0.25	0.13	0.11	0.30
Portugal	0.44	0.39	0.41	0.39	0.65	0.28	0.26	−0.12	0.12	0.24
Romania	0.26	0.20	0.22	0.22	0.37	0.05	0.13	0.06	0.03	0.14
Slovenia	0.28	0.28	0.30	0.31	0.47	0.16	0.25	0.21	0.23	0.37
Slovakia	0.43	0.38	0.38	0.42	0.64	−1.89	0.30	0.28	0.23	−0.33
Finland	0.24	0.24	0.26	0.25	0.41	0.55	0.55	0.52	0.57	0.88
Sweden	0.57	0.53	0.54	0.57	0.86	0.49	0.55	0.48	0.49	0.82
	Economic					Integrated
Bulgaria	0.13	0.14	0.12	0.11	0.24	0.20	0.25	0.23	0.07	0.25
Czechia	0.32	0.31	0.29	0.31	0.58	0.50	0.55	0.52	0.44	0.60
Denmark	0.28	0.28	0.25	0.29	0.52	0.15	0.20	0.13	−0.03	0.17
Germany	0.47	0.45	0.45	0.45	0.86	0.57	0.08	0.53	0.48	0.50
Estonia	0.13	0.09	0.12	0.13	0.23	0.45	0.41	0.48	0.47	0.54
Ireland	0.19	0.23	0.17	0.22	0.39	0.38	0.43	0.41	0.34	0.47
Greece	0.02	0.05	0.05	0.06	0.09	0.14	0.23	0.24	0.14	0.25
Spain	0.22	0.23	0.23	0.26	0.45	0.51	0.52	0.53	0.38	0.58
France	0.37	0.36	0.33	0.35	0.67	0.61	0.65	0.62	0.50	0.69
Croatia	0.05	0.04	0.04	0.12	0.13	0.21	0.23	0.25	0.24	0.31
Italy	0.23	0.21	0.21	0.25	0.43	0.47	−1.04	0.56	0.37	0.17
Latvia	0.15	0.12	0.14	0.17	0.28	0.14	0.17	0.16	0.02	0.19
Lithuania	0.16	0.17	0.15	0.15	0.30	0.25	0.28	0.27	0.15	0.31
Luxembourg	−1.87	−1.86	−1.94	−1.70	−3.43	−1.53	−1.52	−1.55	−1.27	−1.53
Hungary	0.27	0.25	0.24	0.26	0.49	0.39	0.45	0.47	0.34	0.50
Netherlands	0.36	0.38	0.41	0.47	0.77	0.50	0.55	0.52	0.45	0.59
Austria	0.42	0.39	0.37	0.38	0.74	0.73	0.81	0.78	0.68	0.86
Poland	0.22	0.20	0.21	0.22	0.41	0.36	0.40	0.35	0.24	0.42
Portugal	0.18	0.18	0.17	0.20	0.35	0.42	0.45	0.24	0.29	0.42
Romania	0.11	0.13	0.11	0.14	0.24	0.15	0.23	0.19	0.09	0.23
Slovenia	0.16	0.16	0.19	0.24	0.36	0.26	0.36	0.38	0.33	0.41
Slovakia	0.31	0.30	0.28	0.32	0.57	−0.64	0.54	0.53	0.45	0.29
Finland	0.24	0.40	0.34	0.43	0.66	0.50	0.65	0.64	0.63	0.70
Sweden	0.52	0.54	0.52	0.55	1.00	0.85	0.93	0.90	0.84	1.00

Source: own studies.

.

Furthermore, the values of the synthetic measures in three different pillars of transport sustainability were treated as the basis for the construction of the integrated order, which is considered the measure of the development of sustainable transport.

Then, the synthetic means provided were arranged in a monotonous manner. This allowed us to prepare a ranking for each of the analyzed orders as well as for sustainable transport development. On the one hand, the maximum value of the synthetic indicator for each consideration order indicates that the most developed countries are based on specific components and transportation sustainability. On the other hand, the minimum values of the synthetic measures pointed to the least developed countries. The results of this investigation are presented in

Table 5. Positions of countries in the rankings according to particular orders in time.

Country	2016	2017	2018	2019	2016–2019	2016	2017	2018	2019	2016–2019
	Environmental					Social
Bulgaria	19	19	18	19	18	17	18	17	18	15
Czechia	12	13	13	13	13	8	8	8	6	7
Denmark	23	23	23	23	23	6	5	5	10	5
Germany	11	9	10	8	10	13	23	19	16	22
Estonia	17	17	17	17	17	1	1	1	1	1
Ireland	20	22	22	20	21	3	2	2	3	3
Greece	22	20	19	18	19	10	10	10	11	10
Spain	8	8	9	10	8	5	6	6	12	6
France	6	7	6	7	7	7	7	7	9	8
Croatia	10	11	8	9	9	19	19	22	17	17
Italy	3	3	3	3	3	16	24	11	20	24
Latvia	21	21	21	22	22	22	20	21	23	19
Lithuania	18	18	20	21	20	15	15	13	8	13
Luxembourg	24	24	24	24	24	11	12	15	5	11
Hungary	7	4	4	4	4	21	21	18	21	18
Netherlands	15	15	15	15	15	9	11	14	13	12
Austria	1	1	1	1	1	12	9	9	7	9
Poland	9	10	11	12	11	18	17	20	22	16
Portugal	4	5	5	6	5	14	14	24	19	20
Romania	14	16	16	16	16	23	22	23	24	21
Slovenia	13	12	12	11	12	20	16	16	15	14
Slovakia	5	6	7	5	6	24	13	12	14	23
Finland	16	14	14	14	14	2	4	3	2	2
Sweden	2	2	2	2	2	4	3	4	4	4
	Economic					Integrated
Bulgaria	19	18	20	22	19	18	16	20	21	19
Czechia	6	7	7	8	7	6	5	9	9	5
Denmark	8	9	9	9	9	19	20	23	23	22
Germany	2	2	2	3	2	4	22	7	5	9
Estonia	20	21	19	20	21	10	12	11	6	8
Ireland	14	12	15	15	14	13	11	13	12	11
Greece	23	22	22	23	23	22	18	18	19	18
Spain	12	11	11	11	11	5	8	6	10	7
France	4	6	6	6	5	3	4	4	4	4
Croatia	22	23	23	21	22	17	19	17	16	16
Italy	11	13	13	12	12	9	23	5	11	23
Latvia	18	20	18	17	18	21	21	22	22	21
Lithuania	17	16	17	18	17	16	15	16	18	15
Luxembourg	24	24	24	24	24	24	24	24	24	24
Hungary	9	10	10	10	10	12	9	12	13	10
Netherlands	5	5	3	2	3	8	6	10	7	6
Austria	3	4	4	5	4	2	2	2	2	2
Poland	13	14	12	14	13	14	13	15	17	13
Portugal	15	15	16	16	16	11	10	19	15	12
Romania	21	19	21	19	20	20	17	21	20	20
Slovenia	16	17	14	13	15	15	14	14	14	14
Slovakia	7	8	8	7	8	23	7	8	8	17
Finland	10	3	5	4	6	7	3	3	3	3
Sweden	1	1	1	1	1	1	1	1	1	1

Source: own studies.

.

Furthermore, the values of the synthetic measures can be implemented in the process of identifying similar areas according to the level of sustainable transport development and its particular pillars. It should be noted that this study can be performed both statically and dynamically. The results of such analysis, according to three-median method, are presented in

Table 6. Groups of countries according to particular orders in time.

Country	2016	2017	2018	2019	2016–2019	2016	2017	2018	2019	2016–2019
	Environmental					Social
Bulgaria	IV	IV	III	IV	III	III	III	III	III	III
Czechia	II	III	III	III	III	II	II	II	I	II
Denmark	IV	IV	IV	IV	IV	I	I	I	II	I
Germany	II	II	II	II	II	III	IV	IV	III	IV
Estonia	III	III	III	III	III	I	I	I	I	I
Ireland	IV	IV	IV	IV	IV	I	I	I	I	I
Greece	IV	IV	IV	III	IV	II	II	II	II	II
Spain	II	II	II	II	II	I	II	I	II	I
France	I	II	I	II	II	II	II	II	II	II
Croatia	II	II	II	II	II	IV	IV	IV	III	III
Italy	I	I	I	I	I	III	IV	II	IV	IV
Latvia	IV	IV	IV	IV	IV	IV	IV	IV	IV	IV
Lithuania	III	III	IV	IV	IV	III	III	III	II	III
Luxembourg	IV	IV	IV	IV	IV	II	II	III	I	II
Hungary	II	I	I	I	I	IV	IV	III	IV	III
Netherlands	III	III	III	III	III	II	II	III	III	II
Austria	I	I	I	I	I	II	II	II	II	II
Poland	II	II	II	II	II	III	III	IV	IV	III
Portugal	I	I	I	I	I	III	III	IV	IV	IV
Romania	III	III	III	III	III	IV	IV	IV	IV	IV
Slovenia	III	II	II	II	II	IV	III	III	III	III
Slovakia	I	I	II	I	I	IV	III	II	III	IV
Finland	III	III	III	III	III	I	I	I	I	I
Sweden	I	I	I	I	I	I	I	I	I	I
	Economic					Integrated
Bulgaria	IV	III	IV	IV	IV	III	III	IV	IV	IV
Czechia	I	II	II	II	II	I	I	II	II	I
Denmark	II	II	II	II	II	IV	IV	IV	IV	IV
Germany	I	I	I	I	I	I	IV	II	I	II
Estonia	IV	IV	IV	IV	IV	II	II	II	I	II
Ireland	III	II	III	III	III	III	II	III	II	II
Greece	IV	IV	IV	IV	IV	IV	III	III	IV	III
Spain	II	II	II	II	II	I	II	I	II	II
France	I	I	I	I	I	I	I	I	I	I
Croatia	IV	IV	IV	IV	IV	III	IV	III	III	III
Italy	II	III	III	II	II	II	IV	I	II	IV
Latvia	III	IV	III	III	III	IV	IV	IV	IV	IV
Lithuania	III	III	III	III	III	III	III	III	III	III
Luxembourg	IV	IV	IV	IV	IV	IV	IV	IV	IV	IV
Hungary	II	II	II	II	II	II	II	II	III	II
Netherlands	I	I	I	I	I	II	I	II	II	I
Austria	I	I	I	I	I	I	I	I	I	I
Poland	III	III	II	III	III	III	III	III	III	III
Portugal	III	III	III	III	III	II	II	IV	III	II
Romania	IV	IV	IV	IV	IV	IV	III	IV	IV	IV
Slovenia	III	III	III	III	III	III	III	III	III	III
Slovakia	II	II	II	II	II	IV	II	II	II	III
Finland	II	I	I	I	I	II	I	I	I	I
Sweden	I	I	I	I	I	I	I	I	I	I

Source: own studies.

.

The analysis showed that the classification of the chosen member states of the European Union into groups differed according to particular pillars of transport sustainability. Moreover, their positions measured in terms of the integrated order depended on the time. Thus, research into the application of the standardized measurement of the order in a dynamic analysis with the Weber median yielded some interesting results. The results of the present study allow us to take into account static and dynamic interactions in complex areas of transport sustainability, while the availability of data limited the study.

However, analyses have shown that certain trends can be observed over a period of time. The correct assessment of transport sustainability should start with a detailed analysis of its individual pillars: environmental, social, and economic.

The first area, economic order, was occupied by member states such as Austria, Sweden, and Italy, which are considered the most developed. However, the position of Hungary in the area of the environmental pillars may be considered surprising. This member state moved from the seventh position in 2016 to a permanent fourth position in the rankings later. The leading and constant positions of the countries mentioned above are observed in the constructed rankings for particular years of the analysis. This phenomenon was confirmed by the final (i.e., spatial-time) values of the synthetic measure in the area of the environmental order. Therefore, the countries mentioned above were in the first group of countries analyzed in terms of environmental order. However, the very weak positions of Luxemburg and Denmark, which are perceived as old European Union countries, can be considered very surprising in the environmental area. Furthermore, as a third underdeveloped member state in that area, Latvia is generally considered a new member state of the European Union.

The second area, the social pillar, is occupied by different member states in relation to the previous order of transport sustainability. Therefore, the rankings in relation to the social order look completely different. On the one hand, countries such as Finland, Ireland, and Sweden can be considered to be among the most developed in that pillar of transport sustainability. Slight changes in their positions in the rankings of the social order were observed in particular years. However, these changes were limited to one place in the rankings. Therefore, it can be concluded that relatively stable positions were maintained by these countries, and the spatial-time model can be considered to be very credible. Nevertheless, Estonia’s first and constant position over time can be considered to be very surprising. All these countries were in the first group according to the social order of sustainability of transportation. Additionally, Denmark and Spain were included in the first group of European Union states according to the social order. Therefore, it should be noted that the social aspect clearly dominated in these countries. However, member states of the European Union such as Slovakia and Romania were located in the last group according to the social pillar of transport sustainability. However, the very week positions of Italy and Germany in the constructed rankings for particular years of analysis were very surprising. It seems that in the case of their positions in the form of a spatial-time model in the social area, differences for individual years had a great influence.

Therefore, this situation shows that social issues in these countries do not play a significant role. However, Poland’s position in this ranking does not differ from the countries which are in the middle of the rankings. In 2016–2019, 16 (sixteen) places from 24 countries included in the investigation were taken.

The last area, the economic pillar, was characterized by totally different member states, which could be considered the most developed in that area. On the one hand, there are Sweden, Germany, the Netherlands, Austria, France, and Finland, which are considered the most developed member states. These countries are in the first group according to the economic order of transport sustainability. It is worth mentioning that they are generally called the old European Union. In most cases, slight changes in their positions in the economic order rankings were observed. However, these changes were limited to one or two maximum places in the constructed rankings over time. Sweden has both a permanent position and the highest position throughout history. Hence, this member state of the European Union was located as the first object in the spatial-time model of the economic pillar. In the scope of the worst group, according to the economic order of transport sustainability, two old member states of the European Union were found: Luxembourg and Greece. Furthermore, countries such as Croatia, Estonia, Romania, and Bulgaria were in the same group. When analyzing this situation, it should be stated that the countries of eastern Europe dominated in this group of countries. Luxembourg was an exception here. Poland was in the 13th position in this classification. This shows that economic development in Poland is not too bad.

The analysis of all three areas of transport sustainability mentioned above— environmental, social, and economic—provides a solid basis for a comprehensive assessment of transport sustainability throughout the study period. Taking into account the integrated order, which consisted of the three areas of transport sustainability mentioned above, the following trends can be observed with respect to the positions of the countries included in the research.

On the one hand, Sweden, Austria, and Finland were located at the top of the sustainable transport development rankings over the entire period of time. It should be noted that Sweden took first place in this ranking. This member state was at the forefront of all three aspects of sustainable transport development. Furthermore, Austria can be considered as second. However, this country is currently in the ninth position in the social area. In the others, it is at the forefront. All in all, the leadership positions of the mentioned countries of all the member states analyzed by the European Union are based on the comprehensive order of transport stability, which is measured in the area of transport stability. However, it should be noted that the slight differences among these aforementioned countries relate to the economic pillar of transport sustainability, especially for Austria and Finland. Furthermore, the high positions of France, Czechia, and the Netherlands, which held the fourth, fifth, and sixth positions, respectively, can be considered a bit surprising. This probably shows that great efforts to promote sustainable and ecological development are emphasized in these countries.

However, countries such as Luxembourg, Italy, and Denmark remain in the opposite extreme position with respect to sustainable development over time. These countries took the 24th, 23rd, and 22nd positions, respectively. The spatial time model of transport sustainability has been included in all countries. It should be mentioned that these countries belong to the so-called kernel of the European Union, and their positions do not fully reflect the possibilities of sustainable transport development in the three examined pillars.

It should be stated in relation to Poland that its position in the constructed rankings was more or less in the middle of the member states of the European Union included in the study.

All in all, the research carried out proved that there are differences between the analyzed member states of the European Union. Furthermore, the positions taken by individual countries in the area of particular pillars affect the total position, measured as the integrated order. Additionally, their final positions are strictly affected by changes throughout all years of analysis.

6. Discussion

The results obtained in the dynamic research on transport sustainability in the selected European Union countries delivered several conclusions concerning both synthetic measure construction and the issues of transport sustainability as part of the European Union logistics system.

The former proved that a potential set of diagnostic variables should be investigated statically with both classic and order measures, namely the mean arithmetic, the standard deviation, and the mean and mean absolute deviation. Implementation of orderly statistical measures is essential due to the high level of skewness of the chosen variables. In relation to the entire set of diagnostic variables, the strong skewness of the empirical distribution was noted in every data set of diagnostic variables describing the examined pillars of transport sustainability (i.e., environmental, social, and economic). This phenomenon relates to all the years that were taken during the research process. Therefore, the implementation of not only classic variation coefficients should be considered essential for the correctness of statistical verification of the potential data set. Furthermore, using Weber’s median order variation coefficients confirmed that some variables were accepted or rejected for further analysis during the period.

Examination of the variation coefficients in all years of analysis showed that the spatial-time taxonomic model should be used. Therefore, the first static taxonomic models were built for each year of the analysis. Then, on the basis of the synthetic measures obtained, space-time models were constructed for both the individual pillars of sustainable transport and the integrated measures.

In relation to correlation analysis, through implementation of a Pearson correlation coefficients inverted matrix, the static verification process of potential data sets could take into account unobserved relationships in the three analysis pillars of transport sustainability in order to take into account the relationship that was not directly observable. As a result of this kind of analysis, one diagnostic variable in the area of economic order has been removed for all the times.

The significance of including interactions in the process of synthesized measure construction was implemented by using a Weber median in the final diagnostic variable renormalization process. It should be mentioned that in the case of the chosen variables throughout the analysis period, the meanings differed from the shape of the border.

Due to the strong distortion and interaction of diagnostic variables in each set of diagnostic variables, the order method for the construction of synthetic measurements with Weber medians should be applied throughout the analysis period.

Additionally, the choice of one method of synthetic measure construction (i.e., its order form) resulted from the number of constructed measures. A large number of synthetic measures resulted from the use of a dynamic approach, as well as the information capacity of the sustainable transport phenomenon. It should be noted that the implementation of the classic taxonomic measure would double the number of synthetic measures obtained. This situation could have caused several difficulties in the stage of discussion of the research results and in drawing conclusions. This could especially occur in the case of dynamic analysis of transport sustainability.

7. Conclusions

The latest observations, based on the results of the study, allowed us to draw some synthetic conclusions about the level of sustainability of transport and its specific components in individual member states of the European Union. These comments are related to specific EU member states that were included in the research process. It should be mentioned that spatial evaluation of transport sustainability in the selected countries can be considered a complex task. The research targeted from 2016 to 2019 proved both the disparity in transport sustainability and its changes between the 24 member states. However, three member states of the European Union were not included in the research process. This was due to the lack of adequate data (indicators) related to transport sustainability. The problem mentioned above especially grows in the case of the dynamic approach.

Furthermore, it can be noticed that European Union member states can be divided into two main groups, taking into account the constructed rankings. This is due to the traditional approach to the division of the European Union into two main areas. The first is usually called the old Europe and includes member states considered the founders of the European Union. The latter is treated as the rest of the European Union, or Central and Eastern Europe. However, it can be observed that significant differences in the area of transport sustainability have not been eliminated so far.

On the one hand, taking into account the total measure (i.e., the spatial time of transport sustainability), Sweden, Austria, and Finland occupied higher and more stable positions. However, the research showed that countries such as Luxemburg, Italy, and Czechia took the last three positions in the rankings over the entire period of time.

To sum up the considerations, it should be mentioned that a certain regularity of the spatial development of transport sustainability and its trends can be indicated. Nevertheless, significant spatial disproportions in the area of transport sustainability in the analyzed European Union member states can be observed. The results presented in the research, in some part, can be treated as a reflection of the level of socioeconomic development of particular countries. Both the issues presented and discussed in the article and the implemented methodology of taxonomic methods can be applied by responsible organs for planning transportation policy in the European Union. Furthermore, the proper monitoring of the development level of sustainable transport, as well as its individual pillars, can facilitate decision-making processes and optimize the use of limited financial resources. Due to the lack of data on variables and certain objects in the Eurostat database, it was impossible to create an ideal indicator system that reflected the entire area of transportation sustainability. Further research should focus on improving the sets of statistical variables implemented in the description process of particular pillars of transport sustainability. However, the latest data could be a base for further analysis, especially in the context of the COVID-19 pandemic.