The Impact of Digitalisation on Supply Chain Competitiveness: A Multi-Country Comparative Approach

Abstract

Digitalisation plays a crucial role in enhancing the competitiveness of the global supply chain, primarily by improving trade openness and efficiency. The current study examined the impact of digital transformation on supply chains by analysing key digital and logistical factors over three Eastern European countries: Bulgaria, Poland, and Romania. Using multifactorial linear regression on data collected from 2013 to 2022, the present research paper evaluated and highlights the influence of broadband penetration, ICT graduates, and transport infrastructure over trade openness. The findings revealed a differentiated impact between the three analysed countries, such as a significant enhancement provided by broadband in Poland and Romania, ICT graduates having a low influence in Bulgaria, and air and rail transport not exhibiting a strong relationship with trade openness in all three countries. These results highlight the need for stronger policies that integrate the adoption of digital technologies, workforce development, and investment in infrastructure to enhance supply chain efficiency. Nevertheless, the present study contributes to the on-growing research trends over digitalisation’s role in trade and supply chain management while providing valuable insights for policymakers and business leaders in the current digital economy.

1. Introduction

Supply chain digitalisation (SCD) is a vital area of research because of its significant impact on international competitiveness, operational efficiency, and strategic advantage at both the country and firm levels. Hence, the digital transformation of supply chains through the integration of various digital technologies, such as big data, cloud computing, artificial intelligence (AI), and blockchain, has become crucial for ensuring competitiveness in an AI-driven era due to the ability of these tools to enhance flexibility, cost efficiency, and responsiveness [1,2,3], in addition to positively contributing to achieving business sustainability [4,5].

While there is a lack of studies on the statistical correlations and interrelationships among the key drivers of SCD, there have been even fewer works addressing supply chain competitiveness among Eastern European countries. Thus, this study aims to fill this gap by providing a comprehensive understanding of these interactions with the purpose of developing more effective strategies for SCD based on factor-centric approaches identified through econometric modelling. To extend previous knowledge on the impact of digitalisation on sustainable supply chain competitiveness, this research carried out an investigation of how various digital and logistic factors influenced trade openness at a country level. Additionally, this study was grounded in the technology–organisation–environment (TOE) conceptual framework developed by Tornatzky and Fleischer (1990) as it considers three critical dimensions that influence technology adoption [6]. Furthermore, the TOE model has been extensively validated in the field of operations management and digital transformation research, which indicates that it is an appropriate choice for placing this investigation within the existing academic literature. To address our study’s main objective, the following research question was formulated:

RQ1: How does digitalisation impact the supply chain efficiency across three Eastern European economies?

Hence, the answer to this question will provide valuable implications in terms of trade openness and the digital transformation of supply chains for policymakers, businesses, and scholars in this field of knowledge. Additionally, the research hypothesis used in the current study was extracted from a thorough analysis of the scientific literature, stating that digitalisation positively influences supply chain competitiveness through increased efficiency [4,7].

Accordingly, the value of the present research lies in addressing digital transformation’s impact on supply chain competitiveness with a unique focus on three Eastern European countries through econometric modelling in order to achieve valuable insights on the implications of trade openness and supply chain digitalisation for Bulgaria, Poland, and Romania. Likewise, our study enriches the current body of literature as it is grounded on the technology–organisation–environment (TOE) framework and utilises the application of robust econometric modelling to carry out an exploration of how digital infrastructure, human capital, and freight transport efficiency jointly influence trade openness.

To ensure rigorous grounding in the scientific literature, Section 2 of our paper discusses both divergent and similar studies in the field of supply chain digitalisation and competitiveness, followed by a comprehensive approach of our methodological support in Section 3 as well as through a critical interpretation of the main results in Section 4. The discussion in Section 5 will further align our key results with previous studies, while the conclusions in Section 6 will bring together the main results and provide detailed recommendations to address the implications of our work.

2. Literature Review

Previous studies in the field of supply chain digitalisation have stated that the integration of digital technologies is extremely important for increasing efficiency, reducing costs, and enhancing customer satisfaction, whilst also suggesting the need for further research on the intertwined relationship between the evolving global business ecosystem and the competitive international pressures that require innovative logistics solutions [8,9]. Similarly, the urgent need for supply chain digitalisation can be emphasised by the recent COVID-19 pandemic, when the supply chain vulnerability affected the distribution of goods at a global level, causing many negative economic effects due to its inability to guarantee resilience and sustainability [10,11].

Moreover, there is a recognised need to create theoretical frameworks and methodologies adequate for ensuring the integration of digital technologies into supply chain design, which are based on dynamic reconfiguration to provide risk mitigation and responsiveness improvement [8,12]. Likewise, supply chain digitalisation (SCD) can be further investigated using econometric modelling based on a factor-centric perspective, allowing for a better understanding of the influence of various factors on SCD such as cybersecurity or IT infrastructure [8]. In addition, the lack of convergence between the industry requirements in terms of competences and academic curricula also suggests the need for research into the impact of education on SCD [12].

In spite of the extensive research on this topic focusing on real-time e-commerce order prediction, machine learning applications, sustainable development and digitalisation in the context of Supply Chain 4.0 [9], there has been a lack of studies on the statistical correlations and interrelationships among the key drivers of SCD. Thus, this study aimed to fill this gap by providing a comprehensive understanding of these interactions with the purpose of developing more effective strategies for SCD based on factor-centric approaches identified through econometric modelling. Moreover, this study was grounded in the technology–organisation–environment (TOE) conceptual framework developed by Tornatzky and Fleischer (1990) as it considers three critical dimensions that influence technology adoption: technological factors (availability and advantages of digital solutions), organisational factors (firm size, resources, capabilities), and environmental factors (competitive pressure, regulatory environment, customer demands) [6].

On the one hand, a supply chain must have the ability to identify new strategies and solutions so that it can adapt to the rapid changes around it. Characteristics such as flexibility and adaptability are essential, as they will lay the foundations for useful and effective solutions in difficult times. ICT will help more and more countries collaborate economically with each other, which will make financial markets more complex. In a crisis, in order to stay on an upward slope, it is necessary to find effective solutions related to the market situation. If there are difficulties in predicting what consumers want, there will be a high risk of delays and uncertainty in the relationship between business partners as production becomes less efficient than before the crisis [13].

Moreover, Industry 4.0 has put emphasis on one of its biggest advantages, represented by the ability of digital technologies to make supply chains more efficient and intelligent. For example, vehicles used for transport are now equipped with GPS devices and sensors that measure the weather conditions and humidity. All of these data are sent in real-time to an online system, operated in the cloud, where data can be analysed for the best decision to be made. If the temperature inside a truck starts to rise above the set limit, the system sends an alert to the responsible officer, and all the necessary measures can be applied to avoid damaging the transported goods. Although digitalisation offers significant advantages such as increased efficiency and reduced waste, in the case of many SMEs, the implementation of such technology also involves very high costs. Often, these firms lack the knowledge or resources to adopt new digital technologies [14].

Furthermore, several previous studies on SCD provided the building blocks for our study because they examined the key enablers, models, and conceptual frameworks that are mandatory for the effective integration of digital technologies in supply chains. First, the authors suggest that several enablers are essential for SCD to improve efficiency, transparency, and decision-making processes such as big data, blockchain, IoT, and AI [15,16,17]. The use of big data also comes with challenges like the homogeneous analysis of multiple sources of information through common implementation systems, since many companies face difficulties in interpreting data for the best decision-making processes because they lack big data analysis system integration. Additionally, the use of big data requires significant investments in smart equipment such as high-performance software or even trained human resources to handle such technologies. Especially for small companies, these aspects can be significantly expensive, so may be unable to afford them [18].

For this reason, digital transformation can improve the efficiency and competitiveness of entities within the supply chain through total factor productivity enhancement achieved by optimizing the supply–demand to match and significantly reduce the transaction costs [19,20]. At the same time, there are also companies that use digital technologies to improve their supply chains. Most companies emphasise localisation services implemented for real-time surveillance of their goods. Technologies based on electronic tags and GPS are used to track products from the fabric to the customer, while companies like Best Buy, Wal-Mart, or Tesco have already implemented these solutions and now benefit from their outcomes. In addition to parcel tracking, the costs for automating other activities are also reduced, with all of these solutions coming from the ICT sector, an industry in continuous development [18].

Additionally, recent studies indicate that the digital transformation of supply chains enhances competitiveness through both strategic and operational advantages from the implementation of digital technologies [1,21,22,23,24]. Moreover, the digital encapsulation of supply chain management presents many opportunities for redefining and reshaping previous models by integrating digital technologies [24]. Next, the work of Volkova et al. [25] provides a comparative framework to assess the influence of various digitalisation levels on supply chain competitiveness across three countries (Russia, Azerbaijan, and Switzerland), and places emphasis on the beneficial impact of digitalisation on international competitiveness.

Besides the previously mentioned key enablers of SCD, Figure 1 provides an illustration of how both education and behavioural economic factors have significant influences on the digital transformation of supply chains. While education influences SCD through training programs and skill development [26], among the behavioural factors that affect SCD, there are incentives and motivation as well as decision-making processes [27,28]. In terms of performance metrics, previous studies have mentioned supply chain transparency, total factor productivity, supply chain integration, and supply chain efficiency.

To sum up, in order for a company to be able to face supply chain competitiveness, it has to effectively integrate technology in order to gain an advantage over their competitors. However, for the result to have an effect with significant impact, the technology must be intelligently adapted to each company’s objectives [29].

3. Materials and Methods

This study took into consideration the influence of the digital infrastructure, human capital in the ICT sector, and efficiency of freight transport on the level of trade openness achieved by Bulgaria, Poland, and Romania. The main objective was to check whether digitalisation indeed improved supply chain competitiveness through better trade integration and efficiency.

At the same time, to provide an answer to the research question on which the current study was grounded to carry out an analysis of digitalisation’s impact over the supply chain efficiency across Bulgaria, Poland, and Romania, the following hypothesis was applied, based on previous scholarly work:

H1. 

Digitalisation positively influences the supply chain competitiveness through increased efficiency [4,7].

To investigate the interdependencies between the previously mentioned variables, a panel data regression model was adopted using log-transformed and orthogonalised variables to mitigate multicollinearity and accommodate possible nonlinear effects. The independent variables were broadband subscriptions (bs), ICT graduates (ig), and a composite freight transport indicator (ft), which the supply chain uses to denote its efficiency. Other modified interaction and nonlinear terms, like bs × ig, bs × ft, and log_ft², were included for richer dynamics between digitalisation and trade openness.

This method allowed us to study several factors that drive trade competitiveness simultaneously, as previous studies have vouched for the utility of econometric modelling in the study of digital transformation within the supply chain [22].

Data on trade openness were obtained from the World Bank, calculated as the sum of exports plus imports in absolute U.S. dollars. Data on independent variables—fixed broadband subscriptions and tertiary ICT graduates, with freight transport indicators (air and rail freight, million ton-km) being obtained from Eurostat—were proxies for digital infrastructure, human capital, and logistics efficiency. These are simply base years against which future scenarios can be relied upon.

Since only public, aggregate data were used, no ethical approval was needed. Our study thus complied with open-access research standards, and nothing was based on confidential or individually identifiable data. Data validation was performed using validated statistical sources and strong econometric methods to ensure data integrity.

This research applied purposive sampling considering Bulgaria, Poland, and Romania because they represent different levels of digital progress and integration into trade within the European Union. The research period, 2013–2022, covers both the pre- and post-pandemic trends on digitalisation and trade. Additionally, the analysis was conducted using EViews 12 Student Version, which facilitates time-series regression and cross-sectional econometric analysis. A panel data regression model was constructed, incorporating both linear and nonlinear components to reflect potential diminishing returns and complex interdependencies among variables. Logarithmic transformations were applied to allow for elasticity interpretation, and all key interactions were orthogonalised to mitigate multicollinearity. Additionally, a lagged dependent variable was included to account for trade inertia. The following model was estimated:

TO_it = β₀ + β₁·_orth(BS)²_i + β₂·_orth(FT)_it + β₃·_orth(IG)_it + β₄·_orth(Lagged Transport)_it + β₅·_orth(Log IG)_it + β₆·_orth(Log FT)_it + β₇·_orth(Log FT²)_it + ε_i

(1)

where:

TO—Trade openness;

BS—Fixed broadband subscriptions;

IG—ICT graduates;

FT—Air transport freight;

RGT—Rail transport freight;

Orth*—Orthogonalised variables used to reduce multicollinearity through linear projection techniques;

ε_i—Error term.

Despite the panel model’s advantages being derived from orthogonalised variables where multicollinearity is mitigated and robustness of the model is increased, some limitations ought to be addressed. The analysis was confined to the period between 2013 and 2022 due to the incompleteness and inconsistency of data for previous and later years. Though customs performance was initially considered relevant to the study, it could not be integrated into the final model because harmonised and continuous data were not available for all three countries. Similarly, results for Bulgaria, Poland, and Romania restricted the generalisability of the findings to the much broader region of the EU. Although it featured a composite freight transport indicator that embodies the volumes for road, air, rail, and maritime transport, the study still used macro-level indicators that may not have clearly reflected the firm-level behavioural patterns in terms of digital adoption or trade responsiveness. Subsequent studies may fill this void by adding more countries and a longer time frame besides including microeconomic data, which will make the findings more valid and applicable.

4. Results

4.1. Drivers of Trade Openness in the Context of Digitalisation and Logistics Performance

The panel regression results showed a very good statistical fit with an R² of 0.9935 and adjusted R² of 0.9901, which can be interpreted to mean that the model explained slightly over 99% of the variation in trade openness among Bulgaria, Poland, and Romania for the years 2013–2022. The main drivers orthogonalised broadband subscriptions (orth_bs_log_ft) and ICT graduates (orth_bs_log_ig) were indeed statistically significant at p = 0.0091 and p = 0.0020, respectively, with positive signs implying that increased digital infrastructure plus ICT human capital act significantly in fostering trade openness. The synergy between broadband and ICT (orth_bs_c_sq) was also positive and significant at p = 0.0347; therefore, digital synergies play a complementary role in trade performance. This model also ran a freight transport term (orth_log_ft_sq) squared with a negative sign at p = 0.0552, showing returns to traditional transport infrastructure diminished beyond some level, consistent with logistic congestion or inefficient allocation (

Table 1. The main findings from the panel regression analysis.

Variable	Coefficient	p-Value
ORTH_BS_LOG_FT	9.768	0.009
ORTH_BS_C_SQ	1.387	0.035
ORTH_BS_LOG_IG	6.099	0.002
ORTH_LOG_FT	2.753	0.005
ORTH_LOG_FT_SQ	−0.641	0.055
ORTH_TO_LAG	4.926	0.007

The other model diagnostics, such as the Durbin–Watson statistic (1.81), R-squared (0.99), and F-statistic (290.25), confirm the model’s robustness and overall goodness-of-fit.

).

From the economic viewpoint, the results emphasise the strategic role that digital transformation can play in fostering competitive trade regimes within Eastern Europe. The high demand for ICT graduates implies that digital skills are necessary to make digitalisation productive, rather than just connectivity. ICT does not mean making physical transport work with digital coordination tools; it also has its limitations, whilst negative implications probably impact trade liberalisation, as indicated by the negative coefficient of the squared transport variable in this study. On its part, this model passed all robustness checks: the Durbin–Watson statistic was 1.81 (close to 2), indicating no autocorrelation, and the normality and heteroscedasticity tests were passed. This set of findings suggests that human and integrated digital capital investments, combined with optimised freight strategies, are primary drivers of export–import performance as well as regional trade resilience.

The estimated coefficients from the panel regression model can be interpreted as elasticities due to the log-log transformation applied to the dependent and independent variables. For instance, the coefficient of broadband subscriptions orth_bs_log_ft 9.767857 indicates that after adjusting for multicollinearity, a 1% increase in broadband subscriptions resulted in trade openness, increasing by approximately 9.77% when keeping other factors constant. The elasticity of ICT graduates orth_bs_log_ig was 6.098852, which means a 1% rise in the output of tertiary ICT education as well as 6.10% increase in trade openness. These are very high elasticities and indicate a very high willingness of trade openness to respond to investments in both digital infrastructure and human capital within the digital sector. Moreover, the positive and significant elasticity of lagged trade openness (orth_to_lag = 4.925316) suggests a strong path dependency—economies with more open trade structures in the past are likely to remain open due to structural inertia or policy continuity.

The high elasticity of broadband infrastructure means that digital connectivity is extremely crucial in promoting international trade, most probably through reducing the transaction costs, facilitating e-commerce, and integrating firms into the global supply chains. In the same way, strengthening the ICT workforce enhances innovation and productivity across export-oriented sectors. The positive sign on freight transport and its squared term also indicates that while improving logistics contributes to the openness of trade, there could be diminishing returns at higher levels of investment—this is important for resource allocation decisions. These results reaffirm that balanced investment in both digital infrastructure and logistics has strategic importance for maximising trade performance within emerging European economies.

Figure 2 provides an illustration of the relationship between trade openness (log_to) and two orthogonalised forms of digitalisation: ICT graduates (orth_bs_log_ig) and freight transport (orth_bs_log_ft). These variables were orthogonalised to reduce multicollinearity and allow for a clearer insight into their isolated effects. The data points shown suggest a positive correlation between both indicators and trade openness, which is in line with the results of the regression carried out. A good part of the points located in the upper-right quadrants of both plots signifies that higher specialisation in ICT as well as better systems for freight transport are related to higher degrees of trade integration. The relationships may not be strictly linear, but the general upward trend does serve to reinforce the statistically significant coefficients obtained from the panel regression model, thus validating infrastructure—both digital and logistical—as a means for external trade performance in Bulgaria, Poland, and Romania.

4.2. Assessing Model Adequacy: OLS Diagnostics Overview

Results of the OLS model validation proved that the overall specification of the model was statistically sound and well-aligned with the econometric standards (

Table 2. Verification of econometric assumptions in the OLS framework.

No.	Test/Check	Result/Observation
1	Normality test (Jarque–Bera)	JB = 1.062, p = 0.587 → residuals follow a normal distribution
2	Heteroscedasticity test (Breusch–Pagan–Godfrey	F-statistic = 1.93, p = 0.153 → homoscedasticity cannot be rejected
3	Autocorrelation test (Breusch–Godfrey LM)	F-statistic = 0.42, p = 0.660 → no evidence of autocorrelation
4	Ramsey RESET test	F-statistic = 2.77, p = 0.113 → no indication of model misspecification
5	Multicollinearity (VIF values)	All VIF values < 10 after orthogonalising centred variables → multicollinearity resolved
6	Nonlinearity test (included log_ft2)	orth_log_ft_sq is significant: coef. = 0.371, p = 0.048 → supports existence of nonlinear relationship
7	Significance of key coefficients	Variables such as orth_bs_c_sq (p = 0.029), orth_bs_log_ig (p = 0.019), log_ig (p = 0.002) are statistically significant

OLS was chosen as a preliminary step to test the validity of the model assumptions and ensure the robustness of the regression specification before proceeding to the panel data analysis.

). The Breusch–Godfrey serial correlation LM test reported an F-statistic of 1.1427 with a p-value of 0.3375, suggesting no presence of serial correlation. Heteroscedasticity was also ruled out since the White test (p = 0.4712) and Breusch–Pagan–Godfrey test (p = 0.5269) confirmed homoscedastic residuals. The Jarque–Bera statistic was 1.0624 with a p-value of 0.5877, indicating that residuals followed a normal distribution; the Durbin–Watson statistic of 2.04 fell within the acceptable range, confirming that there was no first-order autocorrelation present. All of these diagnostics taken together validate the reliability of the model, which can therefore be justifiably applied for explanation and prediction.

The Ramsey RESET test achieved an F-statistic value of 0.2941 with a p-value of 0.7479, thus confirming that the model was well-specified and free from omitted variable bias as well as overfitting. The freight transport squared value added (orth_log_ft_sq) became statistically significant with a p-value of 0.0003, thereby increasing the explanatory power of the model. This indicates that there exists a nonlinear relationship between freight transport and trade openness, and that the system operates under diminishing marginal returns, which holds true in economic theory. Such rigorous statistical validation provides strong support for the robustness and reliability of the OLS model.

Orthogonalisation was used to deal with the multicollinearity that had been observed within the broadband subscriptions (log_bs) ICT graduates (log_ig), and freight transport (log_ft). In some cases, the variance inflation factors were greater than 10, a sign of problematic multicollinearity. To alleviate this situation, each independent variable suspected of inducing collinearity was regressed on the other independent variables, and uncorrelated residuals—orthogonal components—were extracted and subsequently used in the final regression. This ensured that each predictor could fully explain its relationship without overlapping statistically with any of the others. Hence, orth_log_bs_log_ig orth_log_bs_log_ft orth_log_bs_c_sq and orth_log_to_lag formed part of the final model, which added a lot of statistical significance towards helping interpret the model itself. Orthogonalisation kept all relevant information but did indeed stabilise the coefficient estimates, regressing them towards more robust and interpretable outputs.

Therefore, the results uncovered rich empirical evidence that digital infrastructure, human capital in ICT, and efficient freight transport go a long way in determining trade openness for Bulgaria, Poland, and Romania. Nonlinear effects, especially diminishing returns in freight transport, place emphasis on the need for targeted investment rather than just scaling the ICT infrastructure. In general terms, this finding reinforces the positioning of digital transformation and logistics efficiency as key levers for international trade performance improvement within emerging EU economies.

5. Discussion

This study considered the effect of digital infrastructure, ICT human capital, and freight transport efficiency on trade openness using panel evidence from three Eastern European countries over the period 2013–2022. The results indicated that digital transformation, in this case, broadband subscriptions and ICT graduates, have significantly led to increasing trade openness within the region. Orthogonalised variables helped to reduce the multicollinearity and ensured robust coefficient estimates. The nonlinear term orth_log_ft_sq captured the potential diminishing returns in the contribution of transport infrastructure to trade. The results of the regression showed the broadband variable bs to have a statistically significant positive impact of p = 0.009, thus supporting the idea that investments in digital infrastructure foster cross-border trade integration. Human capital in ICT ig and its interaction with digital infrastructure also appear to support openness to trade, though statistically weaker. Freight transport ft was included as the composite indicator, and model performance was improved when it had a quadratic term added to it. This implies that beyond a certain point, marginal returns to improved transport infrastructure fall, unless there is coordination with digital transport and policy alignment.

Figure 3 provides an illustration of the key factors having an effect on trade openness and supply chain digitalisation, aligned within the TOE framework [6]. This figure places an emphasis on the interplay between technological enablers, organisational capabilities, and external environmental conditions in shaping a country’s ability to participate effectively in global trade networks. Supply chain digitalisation is a good way to increase efficiency because it also leads to cost reductions and improved customer satisfaction. A considerable influence on it is not only the way the economy evolves globally, but also the competitive pressures felt internationally. In Poland, studies have shown that one of the biggest obstacles for local businesses to implement supply chains is transportation, mainly due to the high costs [30].

It has also been observed that the COVID-19 pandemic highlighted several vulnerabilities, but the digital environment has been proven to be a successful alternative to the classic ways of working in supply chains. On the one hand, studies such as the one conducted by Setiawan et al. [31] suggest that more and more companies should allocate a high budget to the development and innovation of supply chains, as investments in digitalisation create a competitive advantage in the market. However, studies have shown that in the case of less developed countries, such as Guinea-Bissau, the COVID-19 pandemic generated problems related to supply chains because the country’s economy was closely linked to imports of finished products and lacked both the infrastructure and resources to produce goods locally. Consequently, companies in this area have been greatly affected since they largely operate within global supply chains but have yet achieved the flexibility and technological maturity needed to adapt easily. Another example is given by Brazil, which saw a surge of new companies in the field of e-commerce immediately after the burst of the pandemic. As in most countries, governmental restrictions have forced many companies to migrate to the online environment and rely heavily on digitalisation for business continuity [32].

On the other hand, Central and Eastern European countries, such as Poland, Romania, and Bulgaria, benefit from the significant advantage of having an increased number of ICT graduates, in spite of the numerous difficulties experienced during the pandemic. Of the three countries mentioned, Bulgaria is still in the process of developing their digitalisation landscape, while also taking further steps to implement digital methods for their supply chains.

Supply chain management has always been considered essential for the smooth running of economic operations, being based on the planning and control of logistics activities. Studies have shown that this implies a continuous development of the connection between suppliers and customers. As in the case of other countries with developing economies, Bulgaria, Poland, and Romania have made the transition from traditional to smart supply chains, with a focus on the digital environment [33].

Furthermore, He et al. [22] indicated, in a study conducted in China using data from 2007 to 2022 for companies listed on the Chinese stock exchange, that digital transformation constituted a significant and positive effect on the efficiency of supply chains. It was highlighted that corporate governance and the level of competition on the market were important factors that, through digitalisation, led to a high level of efficiency in supply chains. The same study specified that the in-depth analysis of the economic consequences from the perspective of digitalisation led to a decrease in the future costs of international transactions, emphasising market competitiveness. At the same time, for supply chain efficiency to increase through digitalisation, both companies and governments must find a common ground, working together to achieve the development of digital infrastructure. These aspects are also highlighted by the situation in Poland, Romania, and Bulgaria, where more and more companies are starting to invest in the ICT sector with regard to digitalisation. However, the pace and value of investments in these countries are lower than in other European countries such as Germany or the Netherlands, showing different economic outcomes generated by equally divergent strategic perspectives on the digitalisation of supply chains. Moreover, the choice for a comparison between these three countries was based on the fact that all are member states of the European Union and had similar economic transitions after the end of their communist era.

From a policy viewpoint, the findings champion coordinated plans that merge the growth of digital skills, network systems, and more intelligent logistics options to unleash a larger freedom in trade (

Table 3. The key findings and policy implications by country.

Key Findings	Policy Recommendations
Broadband subscriptions (bs) have a strong and statistically significant effect on trade openness (p = 0.009).	Expand fixed broadband coverage and ensure reliable infrastructure in both urban and rural areas.
ICT graduates (ig) positively influence trade openness when combined with digital infrastructure (interaction term significant).	Strengthen ICT education and promote workforce digital skill development.
The composite freight transport indicator (ft) shows marginal contribution, while the squared term (orth_log_ft_sq) suggests nonlinear effects.	Optimise freight logistics using digital tools and invest in multimodal transport for efficient integration.
Orthogonalisation of independent variables improved the model by resolving multicollinearity issues (confirmed by VIF diagnostics < 10).	Adopt methodologically rigorous models in future trade and digitalisation studies to ensure robust evidence for policymaking.
The model was statistically validated (Jarque–Bera p > 0.05; Breusch–Pagan heteroskedasticity p > 0.05; Ramsey RESET p > 0.05).	Maintain data quality standards and apply diagnostic testing in policy-focused research to assess the reliability of digital economy and trade models.

Note: The panel model captured the average effects across Bulgaria, Poland, and Romania. No country-specific coefficients were estimated unless the interaction terms are explicitly included.

). Governments ought to place an emphasis on tech enablement, aiding in the digitalisation of customs processes and supporting ICT education, while companies are urged to put money into digital logistics platforms to maintain their competitiveness in changing supply chains.

The current research paper provides valuable insights regarding the impact of digitalisation on sustainable supply chain competitiveness. However, several limitations must be acknowledged by the authors and researchers in the study field. Firstly, the explanatory power of the econometric model used in this study remains insufficient for capturing the rich fullness of the relationship between digitalisation and sustainable supply chain competitiveness. Indeed, the model provides valuable indications, thus, external factors and other unconsidered variables may still influence the obtained results or the current implication of digitalisation over the global supply chain. Secondly, the relatively small sample size represents another limitation to the current study, as focused on only three countries (Bulgaria, Poland, and Romania). Despite this, these countries were selected for their shared economic context as well as their current transitional economies, which present recent digital transformation. Finally, the timeframe might be relatively short for a study with such implications over the economic environment. Hence, the timeframe may not fully capture the actual long-term trends and the complete impact of digitalisation over supply chain competitiveness.

Given the limitations, future research should be conducted while expanding both the geographical scope and the timeframe analysed to provide a more in-depth and comprehensive portrait of the dynamic relations between digitalisation and the competitiveness of the supply chain by incorporating more variables and collecting data for a longer timeframe, bearing in mind that countries, even alongside economic unions such as the European Union, possess diverse digital adoption levels.

6. Conclusions

The relationship between digital transformation, the efficiency of freight transport, and trade openness was studied across three EU economies (Bulgaria, Poland, and Romania) by using a multi-factorial orthogonalised variable panel regression model to overcome the problems of multicollinearity faced by previous tests. Findings showed that infrastructure for digital services as well as human capital engaged in ICT are critical components that influence a country’s ability to integrate into global trade networks. Data covering 2013–2022 offered a balanced view both in terms of before and after the impact of the pandemic, thus reflecting the changing dynamics in supply chain digitalisation. One of the main results of this study was the steady good effect of broadband plans (which represent digital tools) on trade openness. ICT graduates also helped, showing that building digital skills boosts the gains from infrastructure spending. Freight transport, introduced in the model as a combined indicator, also showed relevance for trade openness, but its influence was subject to diminishing marginal returns, which could be captured through the inclusion of a quadratic term. This nonlinear relationship hence highlights the importance not only of expanding the transport capacity, but also in optimising the logistics and intermodal integration. Lagged trade openness was significant, which suggests that trade dynamics in these economies are path-dependent, being influenced by previous integration levels as well as policy inertia.

Therefore, digital transformation is becoming an increasingly important driver of trade openness. For Bulgaria, Poland, and Romania, to enhance their positions in the world of international trade, their policies ought to focus on expanding the coverage of high-speed Internet as well as on investing in ICT education and simplifying the transport infrastructure. This will create a better environment for fostering a resilient and competitive supply chain. Countries that incorporate digital strategies within their economic planning are likely to be more successful in sustaining long-term growth and competitiveness in international trade as this becomes more digitalised.

The hypothesis that digitalisation reinforces trade openness was only partially confirmed. Broadband expansion and ICT human capital were significant factors in the workings for Romania and Poland, thus the digital infrastructure factor, first and foremost, flags competitiveness in trade. However, the minor effect of digitalisation in Bulgaria seems to suggest that regulatory frameworks, digital enterprising adoption, and the industrial integration of ICT skills may further explain how far the digitalisation process can inspire trade openness. Furthermore, the statistical insignificance of the transport indicators further reflects that unaccompanied physical infrastructure cannot stand for the trade integration driver unless accompanied by complementary digital strategies.

Policymakers should further develop their digital infrastructure, giving due attention to the alignment of ICT courses with the needs of the industry, and supporting the adoption of digital trade as a common practice. Future research should widen the geographical coverage, lengthen the time horizon, and add firm-level data for improving the understanding of the role of digitalisation in promoting openness to trade and economic competitiveness, pay due attention to the needs of the industry for ICT education, and foster the implementation of digital technologies for business logistics. Future studies, therefore, could also incorporate these findings over a larger area and through a more extended time period. Furthermore, data that provide information about the level of adoption of e-commerce are needed at the firm level to better understand the intensity of digital technology used by firms to act globally. Building on that, policies for digital trade, investment in developing digital human capital, and encouraging technology-driven changes in supply chains can help further advance trade openness and competitiveness in the digital age. Further increasing the time dimension might also allow for better capturing of the long-term trends of digitalisation and its sustained effects on trade openness. Additionally, firm-level data on digital adoption and the integration of e-commerce could provide more in-depth insights into how firms are using the available digital tools to carry out cross-border trade. It would also be appropriate for policymakers to sharpen their national strategies towards targeted interventions aimed at enhancing digital trade infrastructure, the gap between ICT education and industry needs, and include digital solutions in trade logistics as well enhance competitiveness in the digital economy.