Digital Skills and Digital Transformation Performance in the EU-27: A DESI-Based Nonparametric and Panel Data Study

Abstract

Digital skills represent a key dimension of digital transformation, shaping the innovation potential, competitiveness, and long-term sustainability of the European economy. The aim of this paper is to compare the development of digital skills in EU-27 countries from 2018 to 2024 and identify the strengths and weaknesses within the European context. The analysis is based on secondary data from the Digital Economy and Society Index (DESI). From the total of 36 indicators included in DESI, 12 variables were selected, with an emphasis on 3 core digital-skills metrics: Internet use, ICT specialists, and ICT graduates. To assess their interrelationships and linkages with overall digital transformation performance, non-parametric correlation analyses (Kendall’s Tau and Spearman’s rank correlation) were applied. Furthermore, across-year nonparametric tests (Friedman ANOVA with Kendall’s coefficient of concordance, W) were used to evaluate year-to-year differences and the stability of country rankings over 2018–2024. The empirical results confirmed that higher levels of digital skills are associated with stronger digital transformation performance among EU member states, while significant cross-country disparities persist. Germany and the Nordic economies (Finland, Sweden, and Denmark) achieved the best results, while Southern and Eastern European countries such as Bulgaria, Portugal, and Greece lagged behind. These findings highlight the strategic role of digital education, ICT specialization, and lifelong learning initiatives in promoting sustainable digital transformation and competitiveness across Europe. In addition, panel regression analysis confirmed that digital infrastructure, particularly FTTP coverage and Very High Capacity Networks, is a key driver of digital skills development, whereas the effects of business digitalization appear indirect or delayed. The outcomes provide relevant implications for broadband deployment and user-centric digital public services to support the objectives of the EU Digital Decade 2030. The study contributes to a deeper understanding of the determinants of digital skills and digital transformation performance, providing evidence-based guidance for targeted digital policies aimed at reducing the digital divide and strengthening digital transformation performance within the European Union.

1. Introduction

Currently, communication skills for working with technology and information are used at work. Processes change rapidly due to globalization, and therefore, an immediate response to these changes is necessary. Digital skills are also needed to understand and manage these changes (Leahy & Wilson, 2014). Digital technologies are, according to Brolpito (2018), the main driving force of a country’s growth, productivity, competitiveness, and innovation capacity. Digital technologies are a challenge for the labour market, impacting both existing jobs and newly created ones that involve working with technology as well as routine tasks. We can also look at digital technologies as an opportunity to create new jobs. In addition to the inherent positives for a country’s economy and employees, the further development of the digital economy brings new internet technologies and a proliferation of digital technology applications, effectively reducing energy intensity and energy consumption per unit of output (Gao et al., 2024).

Digitization represents great opportunities for the economic growth of a country (Bacho et al., 2019) as well as the improvement of working conditions for its employees. At the same time, it brings with it new challenges (Xholo et al., 2025). Digitization is a general term made up of many areas. Depending on the context, it is viewed from technical, economic, financial, and social perspectives. The term encompasses ICT, AI, the Internet of Things, platform and blockchain technologies, virtual reality, etc., among others (Mura & Donath, 2023).

A major priority is the new requirement for workers to have the skills that are necessary for this field. To facilitate the digital transition and reap its benefits, people will need a broad set of skills (Morandini et al., 2020). The point is to have a basic computer awareness, which is essential for all employees. Additional technical skills are a requirement for those involved in the development and maintenance of systems (Leahy & Wilson, 2014).

The COVID-19 pandemic has shown the importance of digitalization, especially in the European region. It is therefore essential for countries in Europe to follow this trend. Digitalization can be a crucial tool to ensure the functioning of everyday life, as well as business and work. The European path to a digitalized economy and society is about solidarity, prosperity, and sustainability. The main tool for identifying the level of digitalization in a population is the DESI. Correct identification of the results of this index can help countries improve.

2. Literature Review

Digital transformation is a topical issue worldwide, of major importance for all companies in all sectors (Zaoui & Souissi, 2020). Digital transformation is a key and strategic orientation in the current era of the digital economy. It is becoming a key competitive advantage for countries (D. Wang & Shao, 2024). It is about adopting disruptive technologies to increase productivity. As a result, the overall social well-being of the residents will be ensured. Many governments base their long-term policies on this principle. Digital transformation is predicted to have high annual growth and rapid penetration. According to El Awady et al. (2025), digital transformation has a positive and immediate impact on the Sustainable Development Goals and thus on the country itself, mainly through technological progress. This uplifts the country and enriches it with new knowledge. But we must not forget that this process has, according to Ebert and Duarte (2018), certain limits. Barriers that slow down its spread include, e.g., company culture, ROI visibility and also lack of skills and skilled labour (Fenech et al., 2019), missing or insufficient infrastructure, and poor access to finance (Ebert & Duarte, 2018).

The development of a global knowledge-based society and the rapid integration of ICT (Mazorodze, 2025) require the acquisition of the new digital skills necessary to carry out our current activities (Van Laar et al., 2017). Digital transformation requires people with knowledge, skills, and motivation to use ICT. Human capital is seen as key to effective digital transformation (Xholo et al., 2025), as it can drive sustainable development when people use ICT in both their work and personal lives. If there is no equal access to ICT or possibilities to deepen or to create digital skills and knowledge for all, the digital transformation can also deepen the existing inequalities between citizens and in the labour market (Qureshi, 2023). Along with changes in labour markets, digital skills are seen as essential skills of the 21st century, especially for searching for and evaluating information, solving problems, and developing ideas in a digital context (Van Laar et al., 2017).

The imperative for digital transformation is becoming increasingly universal across various industry sectors (Díaz-Arancibia et al., 2024). The potential benefits of digitization for companies are diverse, and include, among others, increased productivity, innovation in value creation, and new forms of interaction with customers. Employees that have some processes facilitated by the influence of digital technologies can work more efficiently. Ultimately, business models can be completely reshaped by the impact of digitization (Matt et al., 2015). Digital transformation, i.e., the integration of digital technologies into business, has changed the way companies communicate with their customers, how they manage their operations, and how they set up their business model (Fenech et al., 2019). Whilst digitalization impacts day-to-day HR practices and procedures, particularly with the use of human resources information systems, there is less emphasis on the role of HR in contributing to the strategy of digitalization (Fenech et al., 2019).

Digital transformation is of great importance for the manufacturing industry. This transformation will lead to many new business models and products that will increase the power of manufacturing and exporting companies (Y. Wang et al., 2024). Digital technologies play an increasingly influential role in both the working lives of employees and human resource management. Technology plays a dominant role in defining which skills are considered important. It has become a key concept in the discussion of what kind of skills and understanding citizens must have in the knowledge society (Van Laar et al., 2017). The integration of digital technologies into business processes has become very important to for companies to survive and build a competitive advantage in the market (Fenech et al., 2019). Digital skills are important and should form part of educational policy (Stofkova et al., 2022). By implementing enterprise digitalization, it is possible to significantly reduce production costs, enhance enterprise research and development efforts, increase innovation potential, and promote high-quality growth in the manufacturing sector (Y. Wang et al., 2024).

In order to take full advantage of technology, a person must demonstrate a wide range of complex cognitive, motor, sociological, and emotional skills. Without these skills, there is a risk of exclusion. It is important that people know how to use technology safely (Leahy & Wilson, 2014).

In this context, the concept of digital literacy arises. This term is attributed to Gilster (1998) and is the ability to understand and use information from various digital sources. Digital literacy must be more than just the ability to use digital resources effectively (Van Laar et al., 2017).

Van Laar et al. (2017) consider mastering the digital skills of the 21st century to require the following:

• Mastering ICT applications for solving cognitive tasks at work.
• skills that are not technology-based, as they do not involve the use of any software programme.
• Skills that support higher order thought processes.
• Skills related to cognitive processes that support the continuous learning of employees.

Mastering ICT applications for solving cognitive tasks at work can push the work envelope higher. An important finding, according to Lissitsa et al. (2017), is that digital skills are positively correlated with income. Skills supporting higher order thought processes need to be developed over a long period of time, and education plays a key role in this area. Several researchers (Edwards, 2016; Troussas et al., 2018; Mytra et al., 2021; Senadjki et al., 2024) are devoted to this area and the importance of education. According to Edwards (2016) the students themselves are aware of the advantages and disadvantages of using technology as a learning tool in the classroom. They later bring these positives with them to the labour market. Iordache et al. (2017) pointed out, that, in the future, there will be a need for more attention, reflection, and integration of the notion of support networks and the ability to share different knowledge and resources.

Several approaches have been developed to compare the use and attainment of digital skills. It is important to constantly progress, and a country or a company can progress correctly if it knows its current situation, sees progress or stagnation during a certain period, and can take the right attitude towards it. Since 2014, the Digital Economy and Society Index DESI has been one of the key tools for monitoring and measuring the digital development of the European economy and society. The indicators track the evolution and development of the digital transformation performance of EU Member States (Kovács et al., 2022). The index is published annually, based on data for the previous period. The DESI is designed to measure the readiness and progress of digital transformation (Russo, 2020; Jenčová, 2021). The practicality of this index is in capturing the trend of the constantly advancing socio-economic digital transformation. It will create a comprehensive picture of the digital system and enable comparison between EU Member States.

Since 2014, the European Commission has annually monitored the progress and level of development of Europe’s digital transformation performance in individual member countries using the DESI, which uses a combination of several indicators sorted into several main dimensions as well as sub-dimensions. Since changes are constantly taking place in the world, the European Commission also proceeded to change the methodology of this index in order to reflect the current situation as best as possible. Over the past ten years, the European Commission has made several changes to the number of indicators within the DESI, as well as reducing the number of dimensions. In 2022, the number of dimensions of the DESI was reduced and the 33 indicators were structured into the following 4 main dimensions: Human capital, Connectivity, Integration of digital technologies, and Digital public services.

In 2023, the names of the individual dimensions were be changed, and their designations were changed to the following: Digital skills, Digital infrastructures, Digital transformation of business and Digitalization of public services. At the same time, the number of indicators was reduced to 32, while in the following year the dimensions were not changed, and the number of indicators was increased to 36.

In

Table 1. Development of the structure of the DESI for the years 2022, 2023, and 2024.

Period	Name of Dimensions	Number of Subdimensions	Number of Indicators
2022	DIM_1—Human Capital	2	8
	DIM_2—Connectivity	4	8
	DIM_3—Integration of digital technology	3	11
	DIM_4—Digital Public services	1	5
	Summary	10	33
		2023 and 2024	2023	2024
2023 and 2024	DIM_1—Digital Skills	2	7	6
	DIM_2—Digital Infrastructures	2	7	11
	DIM_3—Digital transformation of business	3	10	11
	DIM_4—Digitalisation of public services	2	8	8
	Summary	9	32	36

Source: Own processing according reports of DESI (European Commission, 2022, 2023, 2024).

, we present a mutual comparison of the development of the DESI for the years 2022, 2023, and 2024 from the point of view of the number of dimensions, sub-dimensions and the number of indicators.

Currently, secondary data for the period from 2018 to 2022 are available, categorized into four main dimensions (Human Capital, Connectivity, Integration of Digital Technologies, and Digital Public Services) and digital transformation performance is evaluated using 33 input indicators. As of 2023, and in line with the Digital Decade Policy Programme 2030, the DESI is now integrated into the State of the Digital Decade Report and used to monitor progress towards digital targets. In the following years (2023 and 2024), compared to previous years, the European Commission has modified the indicators within the DESI. As a result, the overall DESI score and dimension scores are not available, only the values of individual indicators.

The paper is organized into the following sections. First, a literature review was conducted on the topics of digitalization, digital transformation, and digital skills, as well as on the most significant tool for assessing the digital transformation performance of European countries, known as the Digital Economy and Society Index (DESI). The next section briefly presents the aim of the paper, the partial research objectives, the methodology, and the secondary data used. The following empirical section focuses on conducting the research and presenting the findings. The final section summarizes the significant findings and provides key recommendations to support digital transformation with a focus on enhancing the competitiveness of the European Union member states’ economies.

3. Data and Methodology

The main goal of this paper is to compare the development of digital skills levels in EU member states over the period of 2018–2024 and identify the strengths and weaknesses of the analyzed countries within the European area.

As part of our analyses, we used secondary data represented by the Digital Economy and Society Index (DESI) given the availability of secondary data for the years 2018 to 2024 for the 27 member states of the European Union (European Commission 2018, 2019, 2020, 2021, 2022, 2023, 2024). From the total number of 36 indicators making up the DESI, we have chosen the following 12 variables, an overview of which is provided in the following

Table 2. List of selected indicators from DESI.

Name of Indicator	Description of the Indicator	Unit of Measure	Source
DIM_1a_01	Internet use, all individuals (aged 16–74)	% of individuals	Eurostat
DIM_1b_05	ICT specialists	thousands of individuals	Eurostat
DIM_1b_06	ICT graduates	% of graduates	Eurostat
DIM_2c_04	Fixed Very High-Capacity Network coverage	% of households	Broadband coverage in Europe (BCE)
DIM_2c_05	Fibre to the Premises (FTTP) coverage	% of households	BCE
DIM_3g_10	SMEs selling online	SMEs selling online	Eurostat
DIM_4h_02	Digital public services for citizens	Score (0 to 100)	e-Government Benchmark Reports (eGBR)
DIM_4h_03	Digital public services for businesses	Score (0 to 100)	eGBR
DIM_4h_04	Pre-filled Forms	Score (0 to 100)	eGBR
DIM_4h_05	Transparency of service delivery, design and personal data	Score (0 to 100)	eGBR
DIM_4h_06	User support	Score (0 to 100)	eGBR
DIM_4h_07	Mobile friendliness	Score (0 to 100)	eGBR

Source: Own processing.

.

We obtained data on selected indicators from the DESI. The sources from which DESI obtains its primary data are listed in the Table 2.

The basic output of descriptive statistics within selected indicators from the DESI is presented in the

Table 3. Descriptive statistics of the selected indicators from DESI.

Variable	Mean	Median	Min	Max	Std.Dev.	Skewness	Kurtosis
DIM_1a_01	85.5003	86.8800	60.7800	98.9200	8.0858	−0.7254	0.1100
DIM_1b_05	313.4143	166.7000	9.4000	2114.0000	422.5823	2.3418	5.6988
DIM_1b_06	4.5571	4.4000	1.0000	10.1000	1.7861	0.5015	0.1293
DIM_2c_04	60.0275	64.3620	0.0000	100.0000	26.5166	−0.5227	−0.5776
DIM_2c_05	50.9648	54.2447	0.0000	95.5864	24.9432	−0.2884	−0.9259
DIM_3g_10	19.2889	17.9000	5.5000	37.7000	7.6303	0.4886	−0.6628
DIM_4h_02	75.4307	75.8343	44.2418	100.0000	13.4899	−0.3223	−0.6259
DIM_4h_03	84.3819	86.1968	42.2688	100.0000	12.1383	−1.2343	1.5503
DIM_4h_04	63.1209	68.0019	5.5500	100.0000	21.8146	−0.4401	−0.6412
DIM_4h_05	65.1863	66.5896	30.5687	98.2653	15.2258	0.0249	−0.5541
DIM_4h_06	82.2649	82.5397	45.0397	100.0000	11.4295	−0.5255	0.0001
DIM_4h_07	82.2248	88.3651	31.3438	100.0000	16.7183	−0.9717	−0.0389

Source: Own processing.

. The analysis was carried out for the period of 2018 to 2024.

The main goal was achieved through the following partial objectives, which were logically defined as follows:

Partial Objective 1:

to analyze the development and evaluation of the EU-27 countries using three selected DESI indicators, focusing on digital skills, for the period of 2018–2024.

Partial Objective 2:

to identify the existence of relationships between digital skills indicators and selected DESI indicators for the EU-27 countries for the period of 2018–2024.

Partial Objective 3:

to test whether the levels of the three digital-skills indicators (DIM_1a_01, DIM_1b_05, DIM_1b_06) differ across the years 2018–2024 in the EU-27 and to quantify the concordance of country rankings over time.

Partial Objective 4:

to identify the determinants of digital skills in the EU-27 countries using panel regression analysis for the period of 2018–2024.

For Partial Objectives 2, 3, and 4, we formulated and tested the following hypotheses:

Hypothesis 1.

We assume the existence of statistically significant relationships between the selected digital skills indicators and selected DESI indicators for the EU-27 countries for the period of 2018–2024.

Hypothesis 2.

We assume that, for each indicator (DIM_1a_01, DIM_1b_05, DIM_1b_06), the yearly ranks differ significantly across 2018–2024.

Hypothesis 3.

We assume that country rankings exhibit statistically significant concordance over time within each indicator.

Hypothesis 4.

We assume that selected indicators representing digital infrastructure, business digitalization, and digital public services have a statistically significant impact on the level of digital skills in the EU-27 countries over the period of 2018–2024.

To verify the established hypothesis, we applied non-parametric correlation analysis using both the Kendall Tau coefficient (τ) and the Spearman rank correlation coefficient (ρ) in Statistica software. The choice of non-parametric methods was justified by the results of normality tests (Kolmogorov–Smirnov, Lilliefors, and Shapiro–Wilk tests), which confirmed that some of the analyzed indicators did not follow a normal distribution. In such cases, Kendall’s Tau and Spearman’s rank correlation are appropriate, as they do not require normality of the data and measure the strength and direction of monotonic relationships between the variables.

The analysis was performed using the Spearman correlation coefficient due to the non-parametric nature of the data and the presence of ordinal scales within DESI indicators. Correlations were calculated for each year from 2018 to 2024, and only statistically significant relationships (p < 0.05) were considered in further analysis.

To evaluate whether the three digital-skills indicators (DIM_1a_01, DIM_1b_05, DIM_1b_06) differed across the years 2018–2024 and to quantify the concordance of country rankings over time, we applied Friedman ANOVA by ranks (countries treated as blocks) and reported Kendall’s coefficient of concordance (W) as an effect size. Following significant omnibus tests, we conducted pairwise year comparisons using the Nemenyi critical-difference approach based on the average ranks exported from Statistica 14. The critical difference was CD = 1.734 for k = 7 years and n = 27 countries (α = 0.05), and year pairs with |Δ rank| > CD were deemed significant. As a robustness check, we also computed pairwise Wilcoxon signed-rank tests across years, which yielded conclusions consistent with the Nemenyi procedure. Throughout, we use two-tailed p-values with α = 0.05 and the following abbreviations: Kendall Tau (τ), Spearman rank (ρ), Kendall’s W (W), and Nemenyi critical difference (CD); all post hoc comparisons were derived from Statistica’s exported average ranks and calculated in Microsoft Excel.

Despite using official DESI data, potential methodological inconsistencies across years due to indicator restructuring may affect the comparability of results. Therefore, all correlations should be interpreted as indicative rather than causal.

As part of the analytical procedure, a panel regression analysis was conducted with the aim of identifying the determinants of digital skills in the European Union member states. The panel dataset consisted of 27 EU countries (cross-sectional dimension) over the period of 2018–2024 (time dimension). The dependent variables represented three digital-skills indicators included in DESI: Internet use, all individuals (aged 16–74) (DIM_1a_01), ICT specialists (DIM_1b_05), and ICT graduates (DIM_1b_06). The explanatory variables were selected to represent key areas of digital development, specifically Fixed Very High Capacity Network coverage (DIM_2c_04), Fibre to the Premises (FTTP) coverage (DIM_2c_05), SMEs selling online (DIM_3g_10), Digital public services for citizens (DIM_4h_02), and Mobile friendliness (DIM_4h_07).

Given the nature of the data, panel econometric techniques were applied, including a sequential testing procedure to determine the suitability of fixed-effects or random-effects model specifications. The Breusch–Pagan test (to verify the existence of random effects) and subsequently the Hausman test (to assess estimator consistency) were used for model selection. Robust standard errors were applied in all models to account for the heteroskedasticity and autocorrelation typical for economic panel data.

4. Results

To achieve the set partial goals in our contribution, we performed the following research analysis, which was devoted to the assessment of the development of digital skills of the EU member states for the period 2018 to 2024 using the following indicators:

• Internet use, all individuals (aged 16–74)—DIM_1a_01,
• ICT specialists—DIM_1b_05 (KPI),
• ICT graduates—DIM_1b_06.

4.1. Analysis of the Evaluation of Digital Skills of EU Member Countries

The first analyzed variable (DIM-1a_01), focusing on digital skills, was the percentage of Internet users aged 16 to 74, which was available for the years 2018 to 2024. The processed average data is presented in the following figure.

Denmark achieved the best results, with an average of 96.27% of internet users, followed by countries like Luxembourg, Sweden, the Netherlands, and Finland, with averages around 95%. The EU member states’ average was 85.50%, and 11 countries fell below this average. At the end of the ranking is Bulgaria, where the average was 70.61%. Based on the results for individual years, we can state that Denmark ranked first for four years (2019–2021, 2023), Luxembourg was the leader in 2018, Ireland in 2022, and in 2024, the Netherlands reached the highest value with 98.92% of internet users.

Figure 1 illustrates the persistent digital divide between EU countries. For Slovakia, this means that despite high internet penetration, there is still space for improvement to reach (and surpass) the EU average. These data are key in the DESI assessment, where connectivity is one of the main pillars.

The second analyzed variable (DIM_1b_05) was the number of ICT specialists, which, according to the strategic intentions of the EU, was included among the key performance variables, and the achieved average values are presented in the following Figure 2.

The highest average number of ICT specialists for the period of 2018 to 2024 was in Germany (1.862 million), followed by France with 1.176 million. In third place in terms of ICT specialists was Italy, where there was an average of 0.849 mil. ICT specialists. The worst was Malta, where there were only 12,000 ICT specialists on average. Cyprus registered 16,000 ICT specialists, while the annual average of EU countries was at the level of 313,000 ICT specialists. The total number of ICT specialists for the EU had a growing trend for each individual year and ranged from 7180 million to 9789 million over the six-year period. During the entire analyzed period, the leader was Germany, which registered 1556 million in 2018 of ICT specialists, and their number had increased by 551,700 by 2024, representing a 26.18% increase. On the contrary, Malta achieved the worst results when looking at individual years, as the number of ICT specialists ranged from 9400 to 13,900. In 2020, Cyprus ranked last, with 11,200, which has steadily improved over the following years and is registered at 24,700 ICT specialists in 2024.

In the development of the average number of ICT specialists, we see a large disparity between EU countries. Only 1/3 of the studied countries have values above the EU average, and the rest of the countries achieve very low values for this indicator.

The last analyzed variable (DIM_1b_06) is the number of ICT graduates, which is expressed as a % of graduates, and we again present the average values in graphic form (Figure 3).

The highest average was achieved by Estonia, where more than 8% were ICT graduates, followed by Ireland (7.86%) and Finland (7.29%), whose averages were higher than 7%. The EU average was 4.56%, and 14 economies were below this average, with the worst results being for Italy, which registered only 1.29% ICT graduates. If we look at the results for individual years, we can conclude that Italy was always placed at the bottom of the list, and Malta (2018, 2019), Ireland (2020, 2022), and Estonia (2021, 2023, 2024) took the first place.

For the indicator average of ICT graduates, it is possible to see the proportionality of the achieved results across EU countries. Half of the EU-27 countries have values above the average. The differences in the achieved values between countries are not as significant as in the previous indicators.

At the end of our analysis, we also processed the ranking of individual EU-27 countries according to all indicators of digital skills and compiled the resulting ranking of the analyzed countries based on the average values of the indicators for the period 2018–2024.

Based on the results in the previous

Table 4. Ranking of countries according to selected indicators of digital skills.

Country	avg DIM_1a_01	avg DIM_1b_05	avg DIM_1b_06	Summary	Ranking
DE	6	1	9	16	1
FI	5	12	3	20	2
SE	3	7	12	22	3
DK	1	15	7	23	4
ES	9	4	15	28	5
IE	10	16	2	28	6
CZ	14	9	8	31	7
EE	8	23	1	32	8
LU	2	25	5	32	9
NL	4	6	23	33	10
AT	11	10	14	35	11
FR	12	2	21	35	12
BE	7	8	25	40	13
RO	25	11	6	42	14
HU	19	14	11	44	15
PL	21	5	19	45	16
MT	16	27	4	47	17
LV	15	24	10	49	18
IT	22	3	27	52	19
SK	17	18	18	53	20
HR	24	20	13	57	21
SI	18	22	17	57	22
BG	27	17	16	60	23
LT	20	21	20	61	24
PT	23	13	26	62	25
CY	13	26	24	63	26
EL	26	19	22	67	27

Source: Own processing.

, we can conclude that the European leader in the analyzed digital skills is Germany, followed by Nordic countries such as Finland, Sweden, and Denmark. At the opposite end of the digital skills scale are countries such as Bulgaria, Lithuania, Portugal, Cyprus, and Greece with the worst digital skills rankings.

The Nordic countries’ success in this area is the result of a combination of several factors that reinforce each other. On the one hand, there is education and digital literacy, which is established in citizens from children to the elderly through lifelong learning. On the other hand, access to the internet is everywhere, coverage is very good, and technology is becoming a common part of everyday life. Also, part of this is the high digitalisation of public services and e-government.

On the other hand, for countries with weaker performance in DESI, it is clear that their weaker results are usually the result of persistent socio-economic and structural problems. In many rural areas, high-capacity network coverage and quality are lower than in cities, which limits access to online education and modern digital services. For certain groups of the population (especially the elderly or low-income households), cost can be a barrier. These countries are trying to reverse the situation but overcoming these deep-rooted structural and cultural differences is a long-term process.

4.2. Results of Correlation Analysis

In the following part of our contribution, we tried to identify the existence of mutual relations between the three indicators of digital skills and selected indicators of the DESI for the EU-27 countries for the period of 2018 to 2024 in the form of the following hypothesis:

H1. 

we assume the existence of statistically significant relationships between the selected digital skill indicators and selected DESI indicators for the EU-27 countries for the period of 2018–2024.

To verify the hypothesis, we conducted a non-parametric correlation analysis in Statistica using Kendall’s Tau and Spearman’s rank coefficients; employing both enabled us to validate the consistency and robustness of the identified relationships. The results of the correlation analysis are presented in the following section.

Table 5. Results of Kendall’s tau correlation coefficient analysis.

Kendall Tau Correlations (Selected Data DESI 2018_24 in Workbook) MD Pairwise Deleted Marked Correlations Are Significant at p < 0.05
Description and Abbreviation of the Variable		DIM_1a_01	DIM_1b_05	DIM_1b_06
DIM_2c_04	Fixed Very High-Capacity Network coverage	0.3400	0.0441	0.2166
DIM_2c_05	Fibre to the Premises (FTTP) coverage	0.1765	−0.0184	0.1769
DIM_3g_10	SMEs selling online	0.3571	0.0867	0.1631
DIM_4h_02	Digital public services for citizens	0.4400	−0.0840	0.2828
DIM_4h_03	Digital public services for businesses	0.3279	−0.1313	0.2037
DIM_4h_04	Pre-filled Forms	0.3131	−0.0926	0.1499
DIM_4h_05	Transparency of service delivery, design, and personal data	0.2841	−0.0952	0.1994
DIM_4h_06	User support	0.3402	−0.1610	0.2041
DIM_4h_07	Mobile friendliness	0.4940	0.1104	0.2300

Source: Own processing in Statistica.

presents the results of the Kendall’s Tau non-parametric correlation analysis for the EU-27 over the period of 2018–2024. The coefficients (τ) indicate the strength and direction of the monotonic relationships between the analyzed indicators, and statistically significant values (p < 0.05) are highlighted in red.

From the achieved results of the correlation, it is clear that at the level of significance α = 5% Kendall Tau reached positive values in the interval from 0.1104 to 0.4940, which confirmed the positive relationship between the indicator DIM_1a_01 (Internet use) and DIM_1b_06 (ICT graduates) by all nine selected indicators from the DESI. Between the variable DIM_1b_05 (ICT specialists) a positive relationship was confirmed only with the variable DIM_4h_07 (Mobile friendliness) in the amount of 0.1104, and a negative relationship was confirmed with two variables, namely DIM_4h_03 (Digital public services for businesses) and DIM_4h_06 (User support). We found the highest Kendall Tau value between the indicators DIM_1a_01 (Internet use) and DIM_4h_07 (Mobile friendliness) in the amount of 0.4940, and the lowest positive value was between the variable DIM_1b_05 (ICT specialists) and DIM_4h_07 (Mobile friendliness) in the amount of 0.1104.

Table 6. Results of Spearman’s rank correlation coefficient analysis.

Spearman’s Rank Correlation Coefficient (Selected Data DESI 2018_24 in Workbook) MD Pairwise Deleted Marked Correlations Are Significant at p < 0.05
Description and Abbreviation of the Variable		DIM_1a_01	DIM_1b_05	DIM_1b_06
DIM_2c_04	Fixed Very High Capacity Network coverage	0.4858	0.0478	0.3057
DIM_2c_05	Fibre to the Premises (FTTP) coverage	0.2520	−0.0518	0.2500
DIM_3g_10	SMEs selling online	0.4761	0.1159	0.2404
DIM_4h_02	Digital public services for citizens	0.6159	−0.1257	0.4130
DIM_4h_03	Digital public services for businesses	0.4507	−0.2206	0.3052
DIM_4h_04	Pre-filled Forms	0.4492	−0.1545	0.2229
DIM_4h_05	Transparency of service delivery, design and personal data	0.4058	−0.1403	0.2940
DIM_4h_06	User support	0.4944	−0.2493	0.3069
DIM_4h_07	Mobile friendliness	0.6833	0.1555	0.3328

Source: Own processing in Statistica.

presents the results of Spearman’s rank correlation analysis, which confirmed a similar pattern of relationships as observed with Kendall’s Tau, thereby supporting the robustness of the identified associations.

Based on the results of Spearman’s rank correlation analysis, it is evident that at the significance level of α = 0.05, the coefficients reached positive values ranging from 0.2229 to 0.6833, confirming positive relationships between the indicators DIM_1a_01 (Internet use) and DIM_1b_06 (ICT graduates) with all nine selected indicators from the DESI. In the case of DIM_1b_05 (ICT specialists), a positive relationship was confirmed only with DIM_4h_07 (Mobile friendliness) at a value of 0.1555, while negative relationships were identified with DIM_4h_03 (Digital public services for businesses), DIM_4h_04 (Pre-filled Forms), and DIM_4h_06 (User support). The highest value of the Spearman correlation coefficient was recorded between DIM_1a_01 (Internet use) and DIM_4h_07 (Mobile friendliness) at 0.6833, indicating a strong positive relationship, whereas the lowest positive correlation was observed between DIM_4h_04 (Pre-filled Forms) and DIM_1b_06 (ICT graduates) at 0.2229.

At the end of this section, it can be stated that when verifying Hypothesis 1, a statistically significant relationship between digital skills (represented by three indicators) and digital transformation performance (measured by selected indicators from the DESI) of the EU-27 countries for the period of 2018–2024 was confirmed through correlation analysis at a significance level of α = 0.05. Both Kendall’s Tau and Spearman’s rank correlation coefficients revealed predominantly positive and statistically significant associations, indicating that higher levels of digital skills are generally accompanied by stronger digital transformation performance across EU member states. The results of both correlation methods were highly consistent, with only minor variations in the magnitude of the coefficients. Spearman’s rank correlations tended to show slightly higher values, which is consistent with its greater sensitivity to nonlinear monotonic relationships, while Kendall’s Tau provided a more conservative estimate of the strength of association. Overall, the similarity of results obtained from both methods confirms the robustness, reliability, and stability of the identified relationships, thus providing strong empirical evidence for the rejection of the null hypothesis and confirming the existence of dependence between digital skills and digital transformation performance within the DESI framework.

To extend beyond correlational evidence (H1) and to evaluate year-to-year differences and concordance over time (H2 and H3), we additionally applied Friedman ANOVA by ranks with Kendall’s W and post hoc multiple comparisons. These results are reported in Section 4.4 after the dynamic correlation analysis in Section 4.3.

4.3. Dynamic Analysis of Average Significant Correlations

To capture the evolution of the relationships between digital skills and digital transformation performance over time, a dynamic correlation analysis was conducted. This approach enabled the identification of long-term trends and changes in the strength of associations between key digital skills indicators and DESI sub-dimensions, reflecting how the role of human capital has evolved in the context of digital transformation.

Table 7. Temporal evolution of average significant Spearman correlation coefficient (p < 0.05) between digital skills indicators and DESI performance (2018–2024).

Year	DIM_1a_01 (Internet Use)	DIM_1b_05 (ICT Specialists)	DIM_1b_06 (ICT Graduates)
2018	0.4868	−0.0561	0.2828
2019	0.5396	n. s.	n. s.
2020	0.5644	n. s.	n. s.
2021	0.5258	n. s.	n. s.
2022	0.5150	n. s.	0.4124
2023	0.5468	n. s.	0.4392
2024	0.4812	n. s.	0.4338

Note: Values represent averages of all significant (p < 0.05) Spearman correlations between digital skills indicators and DESI sub-dimensions for each year, n. s. = not significant, p ≥ 0.05. Source: Own processing.

presents the dynamic evolution of the average significant Spearman correlation coefficients (p < 0.05) between the selected digital skills indicators (DIM_1a_01—Internet use, DIM_1b_05—ICT specialists, DIM_1b_06—ICT graduates) and the DESI sub-dimensions for the period 2018–2024.

The results show that the Internet use indicator (DIM_1a_01) maintained a moderately strong and stable correlation with DESI performance across all observed years. The coefficients varied only slightly, suggesting that citizens’ online engagement continuously supports national digital transformation performance and remains one of the most stable determinants of digital readiness.

By contrast, the ICT specialist’s indicator (DIM_1b_05) did not exhibit statistically significant correlations in most years, which implies that the quantitative presence of ICT professionals alone does not guarantee higher levels of digital transformation performance. This finding highlights the importance of qualitative factors within the ICT workforce, such as innovation capacity and upskilling, over pure numerical representation.

The ICT graduate’s indicator (DIM_1b_06) shows a clear upward trend in correlation strength beginning in 2022, indicating the increasing importance of digital education and human capital formation in driving digital transformation performance. The strengthening relationship between ICT graduates and DESI outcomes reflects the positive impact of formal education and skills-based training on the development of digital ecosystems.

Overall, the time-based analysis confirms a gradual shift in the determinants of digital transformation performance, from general digital inclusion (internet use) toward education, and competence-based human capital. This pattern underlines the strategic importance of investing in digital literacy and skills development as fundamental pillars of sustainable digital transformation in the European Union.

4.4. Non-Parametric Inference Across Years (2018–2024)

To assess whether levels of the three digital skills indicators differ across years and how stable country rankings are over time, we applied Friedman ANOVA by ranks with Kendall’s W and performed post hoc multiple comparisons (Nemenyi CD = 1.734). The tests confirmed significant year-to-year differences for all three indicators and high to moderate concordance of rankings (EU-27, 2018–2024), as shown in

Table 8. Friedman ANOVA and Kendall’s W for digital-skills indicators (EU-27, 2018–2024).

Indicator	χ2 (df = 6)	p-Value	W	N	Significant Pairs (Count)
DIM_1a_01	127.0952	<0.001	0.7845	27	13
DIM_1b_05	135.3298	<0.001	0.8354	27	12
DIM_1b_06	95.5219	<0.001	0.5896	27	11

Note: Post hoc significance determined by the Nemenyi critical difference, CD = 1.734. Pairwise contrasts were verified with Wilcoxon signed-rank tests. Source: Own processing.

.

For DIM_1a_01 (Internet use) the Friedman test was significant, χ²(6) = 127.0952, p < 0.001, with Kendall’s W = 0.7845 (N = 27). This indicates high concordance of country ordering and a clear upward shift in ranks from 2018 to 2024.

For DIM_1b_05 (ICT specialists) the Friedman test was significant, χ²(6) = 135.3298, p < 0.001, with Kendall’s W = 0.8354 (N = 27). This shows very high concordance and an almost monotonic increase over time. Wilcoxon checks confirm the same pattern except for the adjacent pair 2023–2024, which is not significant.

For DIM_1b_06 (ICT graduates) the Friedman test was significant, χ²(6) = 95.5219, p < 0.001, with Kendall’s W = 0.5896 (N = 27). This reflects moderate concordance and stronger changes in later years. Wilcoxon results show that early adjacent pairs 2019–2020 and 2020–2021 are not significant while later contrasts are typically significant.

Taken together, the omnibus Friedman tests, the magnitude of Kendall’s W, and the corroborating Wilcoxon checks provide convergent evidence that the distributions of yearly ranks differ across 2018–2024 while the ordering of countries remains substantively consistent over time. These results confirm Hypothesis 2 and Hypothesis 3. Year-to-year differences are statistically significant for all three digital-skills indicators, whereas concordance is high for Internet use and ICT specialists and moderate for ICT graduates. The pattern is stable across alternative non-parametric procedures and aligns with the descriptive trends reported earlier, which strengthens the internal validity of the findings and supports their interpretation as robust evidence on temporal change with persistent cross-country ordering. For context, Kendall’s W values around 0.30 are commonly viewed as low, around 0.50 as moderate, and at or above 0.70 as high, which is consistent with the magnitudes observed here. Post hoc decisions relied on the Nemenyi critical difference computed from average ranks, and ties were handled according to the software’s ranking procedures. Wilcoxon signed-rank checks yielded the same substantive picture, with only the adjacent pair 2023–2024 for ICT specialists not significant, which supports the stability of the results. The sample size is N = 27 for all yearly contrasts and no countries were dropped, so the inference reflects the full EU-27 panel for 2018–2024.

4.5. Panel Regression Results

A panel regression analysis was conducted to identify the determinants of digital skills in the EU-27 countries over the period of 2018–2024. The dependent variables represented three indicators of digital skills: Internet use (DIM_1a_01), ICT specialists (DIM_1b_05), and ICT graduates (DIM_1b_06). The explanatory variables included indicators of digital infrastructure, business digitalization, and digital public services. Robust standard errors were applied in all models to correct for heteroskedasticity and autocorrelation in panel data. The choice between fixed-effects and random-effects specifications was based on the Breusch–Pagan and Hausman tests.

The general specification of the three panel models was defined as follows:

Y_it = β₀ + β₁DIM_2c_04_it + β₂DIM_2c_05_it + β₃DIM_3g_10_it + β₄DIM_4h_02_it + β₅DIM_4h_07_it + u_i + λ_t + ϵ_it

where Y_it is the dependent variable for country i and year t

• (DIM_1a_01; DIM_1b_05; DIM_1b_06),
• u_i represents the country-specific effect (fixed or random),
• λ_t is the time effect (included only in Model 1),
• ϵ_it it is the error term.

Breusch–Pagan tests confirmed the presence of panel effects in all cases (p < 0.001). The Hausman test supported a fixed-effects model specification for DIM_1a_01 (p = 0.003), and the F-test confirmed the joint significance of time effects (p = 0.0146). For DIM_1b_05 and DIM_1b_06, random-effects models were found to be consistent (p = 0.357 and p = 0.7206, respectively).

The regression results are reported in

Table 9. Panel regression results.

Variable/Model	Model 1: DIM_1a_01	Model 2: DIM_1b_05	Model 3: DIM_1b_06
Specification	FEM + Time Effects	REM	REM
DIM_2c_04	−0.0154 (p = 0.7327)	1.5765 * (p = 0.0000)	−0.0066 (p = 0.1927)
DIM_2c_05	0.1124 * (p = 0.0393)	−0.3973 (p = 0.2606)	0.0150 * (p = 0.0150)
DIM_3g_10	0.0397 (p = 0.4129)	−3.1206 (p = 0.0680)	−0.0059 (p = 0.7389)
DIM_4h_02	−0.0538 (p = 0.2892)	1.9973 (p = 0.2514)	0.0169 (p = 0.0701)
DIM_4h_07	0.1332 * (p = 0.0103)	−0.1488 (p = 0.8016)	0.0177 * (p = 0.0008)
Constant	83.8292 (p = 0.0000)	−58.1614 (p = 0.7457)	1.9192 (p = 0.3565)

Note: * p < 0.05. Source: Own processing.

, summarizing the estimated coefficients and p-values.

• Model 1: DIM_1a_01—Internet use

The fixed-effects model with time effects was identified as the preferred specification. The results revealed a statistically significant positive effect of FTTP coverage (DIM_2c_05; p = 0.0393) and mobile-friendly digital public services (DIM_4h_07; p = 0.0103) on Internet use. This implies that modern broadband infrastructure and accessible online services contribute to higher levels of digital inclusion.

• Model 2: DIM_1b_05—ICT specialists

The random-effects model was found appropriate for this indicator. A significant positive relationship was confirmed for Very High Capacity Network coverage (DIM_2c_04; p < 0.001), suggesting that technologically advanced infrastructure supports job creation and demand for ICT professionals. Other covariates were not statistically significant, pointing to the dominant role of network infrastructure in shaping the ICT labour market.

• Model 3: DIM_1b_06—ICT graduates

The random-effects model was preferred for this indicator. Statistically significant positive effects were identified for FTTP coverage (DIM_2c_05; p = 0.0150) and mobile-friendly digital services (DIM_4h_07; p = 0.0008). These findings indicate that high-quality digital environments stimulate interest in ICT-oriented studies and contribute to developing skilled human capital. The variable digital public services for citizens (DIM_4h_02) were marginally significant (p = 0.0701), suggesting a possible delayed effect.

The findings consistently indicate that digital infrastructure is the key determinant of digital skill development across the EU-27. Its influence extends to digital participation of citizens, the structure of the ICT labour market, and the motivation to pursue ICT-focused education. Business digitalization does not yet emerge as a direct and consistent driver, potentially due to indirect or lagged impacts. The results of the panel regression analysis therefore provide empirical support for Hypothesis 4, confirming that selected DESI indicators significantly affect the development of digital skills in the European Union.

5. Discussion

The goal of this study was to compare the development of digital skill levels in EU member states over the period of 2018–2024 and identify the strengths and weaknesses of analyzed countries within the European area.

Our results show that the European leader in the analyzed digital skills is Germany, followed by Nordic countries such as Finland, Sweden, and Denmark. At the opposite end of the digital skills scale are countries such as Bulgaria, Lithuania, Portugal, Cyprus, and Greece with the worst digital skills rankings. Denmark achieved the best results, with an average of 96.27% of internet users and Bulgaria had the lowest average of 70.61%. Based on the results for individual years, we can state that Denmark ranked first for four years (2019–2021, 2023), Luxembourg was the leader in 2018, Ireland in 2022, and in 2024, the Netherlands reached the highest value with 98.92% of internet users. The highest average number of ICT specialists for the period from 2018 to 2024 was in Germany (1.862 million) and the worst was Malta, where there were only 12,000 ICT specialists on average. The highest average was achieved by Estonia, where more than 8% were ICT graduates, and the opposite side was Italy, which registered only 1.29% of ICT graduates.

There are many research and studies (Jenčová, 2021; Russo, 2020; Kovács et al., 2022; etc.) investigating the issue of digital transformation and digital skills. Our attention was directed to the analysis of the DESI in EU countries.

Similarly, like the study of Borowiecki et al. (2021), DESI can be considered an effective tool for creating a comprehensive picture of digital transformation processes in today’s Europe, but at the same time we also see certain limits in the methodology of this index. Despite this, together with Borowiecki et al. (2021), we identified Denmark, Finland and Sweden as the strongest countries.

The more complex study made by Mărginean and Orăștean (2017), which evaluated the digital level of countries based on the results of DESI and also achieved in other indices such as NRI, DEI, IMD, IDI, also evaluated the countries Sweden, Denmark and the Netherlands as the best. Here we see the difference with our results, when in our evaluation Finland was in 2nd place and in this study, it was in 5th place among EU countries.

Furthermore, the correlation analyses conducted using Kendall’s Tau and Spearman’s rank coefficients provided additional insights into the relationships between digital skills and digital transformation performance. Both methods confirmed statistically significant and predominantly positive associations between the selected indicators, suggesting that improvements in digital competences and ICT education are closely linked to a higher level of digital transformation performance measured by DESI. These findings are consistent with the conclusions of previous research emphasizing the key role of human capital and digital skills in shaping national digital transformation performance (Mărginean & Orăștean, 2017; Russo, 2020).

Beyond the correlations, non-parametric tests show that year to year differences are statistically significant, while rank concordance is high for Internet use and ICT specialists and moderate for ICT graduates, indicating stable cross-country ordering alongside temporal improvement.

The panel regression analysis enabled a deeper examination of which digital development factors statistically significantly influence the level of digital skills across the 27 EU member states. The empirical results confirmed the dominant role of digital infrastructure, as wider FTTP coverage and the availability of Very High Capacity Networks significantly support Internet use, the expansion of the ICT workforce, and interest in ICT-oriented education. Conversely, the digitalisation of small and medium-sized enterprises did not emerge as a direct determinant of digital skills in the short term, which may indicate mediated or lagged effects. These findings complement the correlation results and provide a stronger analytical basis for the conclusion that modern digital infrastructure represents the primary prerequisite for digital skills development in the EU-27.

The results of this exploratory study also point to some possible avenues for further study. First, the DESI methodology itself is causing the biggest debate. Even though certain changes have been made in the composition and naming of the dimensions as well as the number of indicators included in them, it is still an area that requires constant monitoring and adjustments. Therefore, for further research, it would be appropriate to focus exclusively on this area.

What may be a problem in the future is the DESI methodology itself. Frequent changes in the indicators that make up this index are not the most suitable if you plan to compare countries over a longer period. This can cause misinterpretation of the time series. Another problem we have identified with DESI is its data lag. This can cause the data no longer reflect the current state and impact on countries.

Another relevant topic for future research would be to examine the impact of inequalities in ICT education and access on digital development, as well as other forms of inequalities related to the digital divide. Understanding how these disparities affect a country’s DESI performance and its specific dimensions would contribute to more inclusive digital policies.

And finally, the analysis of citizens’ digital competences deserves further attention. It would be appropriate to map and study various initiatives (policy measures, projects, services) aimed at actively supporting citizens in the use of digital public services and further develop courses, lecturer activities, online education to strengthen them.

It would also be valuable to map and analyzed national and EU-level strategies aimed at enhancing digital literacy and inclusion, with particular attention to the role of lifelong learning and adult education programs.

6. Conclusions

The issue of digital transformation and digital skills is very extensive. Our study on a sample of EU countries pointed to findings that are characteristics of this region. Based on the analysis of the DESI, we have identified which of its indicators have a significant impact. In conclusion, we can state that Germany is the European leader in three indicators of digital skills, followed by Nordic countries such as Finland, Sweden, and Denmark. At the opposite end of the digital skills scale are countries like Bulgaria, Lithuania, Portugal, Cyprus, and Greece with the worst digital skills scores. To improve performance in the mentioned areas, it is necessary for countries such as Bulgaria to consider internet coverage and to involve more citizens on the internet. It is also necessary for citizens to know how to work with the internet. It might be appropriate to go the route of short training courses aimed at older citizens or disadvantaged groups. This could be beneficial for them both in terms of better socialization and in terms of improving their status. Another suggestion for countries with lower values in the indicators number of ICT specialists and average ICT graduates is that these areas should be supported either by training or specialization in studies so that there are enough of these professionals with the given skills.

From the achieved results of the correlation analysis, at the level of significance α = 5% Kendall Tau reached positive values in the interval from 0.1104 to 0.4940, which confirmed the positive relationship between the indicator Internet use and ICT graduates with all nine selected digital transformation performance variables of the DESI. For the variable ICT specialists, the relationship was confirmed with only three variables of the digital transformation performance of the DESI.

Similarly, Spearman’s rank correlation analysis produced consistent results, showing predominantly positive and statistically significant relationships between digital skills and digital transformation performance indicators. The consistency of both correlation coefficients confirmed the robustness of the findings and led to the acceptance of Hypothesis 1, verifying the existence of a statistically significant dependence between digital skills and digital transformation performance within the EU-27.

Complementing these correlations, non-parametric tests show statistically significant year to year differences across 2018 to 2024 for all three digital skills indicators, while Kendall’s W indicates high rank concordance for Internet use and ICT specialists and moderate concordance for ICT graduates, thereby confirming H2 and H3 and pointing to stable cross-country ordering alongside temporal improvement.

The extended results of the panel regression analysis enabled the identification of statistically significant digital development factors influencing the level of digital skills in the EU-27 countries. Wider FTTP coverage and the availability of Very High Capacity Networks represent essential prerequisites for strengthening digital inclusion, expanding the ICT workforce, and increasing interest in ICT-oriented fields of study. These findings provide clear implications for EU policymakers, who should place greater emphasis on the deployment of modern broadband networks and the enhancement of user-centric digital public services. At the same time, there is a need for further research into indirect and delayed effects of SME digitalisation, which may materialize only over a longer time horizon.

DESI serves as the main monitoring tool for the ambitions and targets set out in the Digital Compass. This index provides concrete data that helps the European Commission and Member States identify priority areas where investment in and acceleration of digitalization efforts are needed to achieve the goals set out in the Digital Compass 2030. The DESI is therefore a tool with which Europe seeks to achieve these goals and lead Member States towards digital sovereignty by 2030.

It is worth paying attention to the DESI. It is a well-constructed index, suitable for comparing countries. Of course, it has its limits and shortcomings, but by eliminating them and especially by processing the indicators into the index in a timely manner, it could be a valuable tool for countries and an indicator of future directions.

Although this issue is very current, it is impossible to cover it all. The work has limitations that we have outlined as possible suggestions for future studies. We believe that this study has contributed at least partially to the creation of the image of digitization in the EU.

From a policy perspective, the results highlight the need to strengthen digital education systems, invest in ICT-related study programs, and develop targeted strategies to reduce the digital divide among EU member states. Moreover, enhancing citizens’ digital literacy through lifelong learning initiatives may represent a key factor in improving overall digital transformation performance.