Macroeconomic and Energy Drivers of Sustainable Logistics: Evidence from the Baltic Sea Region

Abstract

This study examines the impact of macroeconomic and energy factors on the sustainable development of the logistics sector in eight Baltic Sea Region (BSR) countries from 2008 to 2023. A synthetic logistics sustainability index (SD), ranging from 0.54 (Lithuania, 2009) to 0.93 (Germany, 2023), was constructed to capture economic, social, and environmental dimensions. The analysis employed country-level regressions, fixed-effects panel models, and a one-step dynamic GMM estimator. Results show that higher GDP per capita (β ≈ +0.35, p < 0.05) significantly supports sustainable logistics, while higher energy intensity (β ≈ −0.41, p < 0.01) constrains it. Across the region, GDP per capita increased by 45% on average, and energy intensity (EI) declined by 18%, contributing to a steady rise in SDI, particularly in Finland, Germany, and Denmark. Renewable energy (RES) has heterogeneous effects: it promotes sustainability in Germany, Finland, and Latvia, but negatively affects Sweden, where rapid energy transition and high electricity costs temporarily reduce logistics efficiency. Electrification rate (RE) also shows a short-term adverse effect in Sweden and Finland, where investment speed exceeds infrastructure adaptability. Labour productivity (LP) and unemployment (UR) exhibit inconsistent effects. Overall, the findings confirm GDP per capita and energy efficiency as dominant drivers of sustainable logistics, while structural and policy differences explain cross-country heterogeneity in sustainability outcomes. These insights provide practical guidance for policymakers by emphasising the need to balance energy transition speed with infrastructure readiness and to tailor sustainability strategies to national economic and energy profiles.

1. Introduction

The logistics sector represents a fundamental pillar of the European economy, being the foundation of the functioning of the single market and a catalyst for trade with third countries [1,2,3]. Its significance in the Baltic Sea Region (BSR) is particularly high due to geographical location and well-developed economic integration [4,5]. This paper focuses on eight countries making up the BSR: Denmark, Estonia, Finland, Germany, Latvia, Lithuania, Poland, and Sweden. They consist of a heterogeneous mix of advanced Nordic economies (Denmark, Finland, and Sweden), one of the foremost European economies (Germany), and Central and Eastern European transition economies (Estonia, Latvia, Lithuania, and Poland) [6,7]. This heterogeneity makes the BSR a unique research sample for analysing how macroeconomic and energy-related determinants shape the sustainable development of logistics in different economic profiles.

The BSR countries act as transport hubs between Northern and Central and Eastern Europe, and handle a significant portion of trade with third countries. Logistics contributes to economic growth and employment, but also generates high environmental costs in the form of emissions [8,9,10]. For this reason, the logistics sector is becoming a priority area in the transition towards sustainable development.

The period covered by this paper, from 2008 to 2023, contains intense economic and social challenges. During this period, countries in the region grappled with the global financial crisis, launched economic recovery measures, and in 2020–2021 faced the challenges of the COVID-19 pandemic. Moreover, after 2019, the energy transition became increasingly important as part of the European Green Deal and the Fit for 55 packages. The paper, therefore, covers a period in which the transition from traditional logistics models to more sustainable solutions is particularly evident.

Previous research points to the importance of economic growth, energy intensity, and the share of renewable energy sources in the sustainable transformation of transport and logistics [11,12]. However, the results are inconsistent. Some authors emphasise the dominant role of economic factors [13,14,15] while others emphasise the crucial importance of social issues and resource efficiency [16,17]. This is particularly true for countries with lower levels of development. This creates a research gap in a comprehensive assessment of the determinants of sustainable logistics development in economic, social, and environmental dimensions. In particular, no up-to-date studies focus on the BSR at the macroeconomic and energy level, making this analysis a novel contribution that fills the regional gap. This gap is especially significant when comparing Nordic with Central and Eastern European economies, which differ in the pace and nature of transformation.

This research aims to assess the impact of six macroeconomic and energy variables on the sustainable development of the logistics sector in eight countries of the BSR from 2008 to 2023. It focuses on such explanatory variables as the energy intensity of the economy (EI), the share of renewables in end-use energy consumption (RES), the unemployment rate (UR), GDP per capita in purchasing power parity (GDP_pps), the rate of growth of labour productivity in the logistics sector (LP), and the rate of electrification (RE). These variables were chosen because they represent key levers through which economic and energy transitions influence three dimensions of logistics sustainability. For instance, LP affects logistics companies’ efficiency and environmental performance by shaping resource use and operational intensity, while EI and RES capture structural energy constraints and opportunities.

The methodology uses a synthetic logistics sustainability indicator (SD), combining three dimensions: economic, social, and environmental. The study also uses a dynamic model based on panel models (Fixed Effects Losses, OLS) and GMM models. Compared to previous studies on the BSR logistic sector [18,19], this research uniquely integrates both macroeconomic and energy determinants at the regional level. This attitude allows static comparison between countries and adjustment processes over time. In this way, the study contributes to the literature by presenting a comprehensive assessment of the macroeconomic and energy factors shaping logistics sustainability in the BSR.

The paper offers critical theoretical insights and practical contributions. The results can support decision-makers in shaping transport and energy policies and help logistics companies adapt their development strategies. Recognising the key barriers and drivers makes it possible to better explain the divergence between the Nordic countries and those of Central and Eastern Europe, pointing to the specific domains where more decisive action is required to accelerate the low-emission transition.

2. Theoretical Background

The Baltic Sea Region (BSR) is an excellent example of studying economic and energy factors that affect sustainable logistics. It includes Germany, Sweden, Poland, Finland, and the Baltic states. It is an important trade link connecting Western Europe with Scandinavia and Central and Eastern Europe [20]. Logistics is a key sector in these economies, contributing to GDP and employment, yet the region shows considerable economic diversity [21,22]. While Scandinavian countries lead in green logistics practices, Central and Eastern European countries face challenges in meeting EU climate objectives, e.g., decarbonisation and sustainable transport development [23,24,25]. Geopolitical tensions, changes in energy supply, and new environmental rules also affect logistics in the BSR. They influence investment, costs, and how competitive economies are [26].

BSR countries share similarities due to EU climate regulations (Fit for 55 package, climate neutrality strategy by 2050) and standard pressure to transition to low-carbon energy systems [27]. However, Nordic countries and Germany benefit from diversified energy mixes and long-term strategies, whereas Poland and the Baltic states face structural and financial barriers [18]. This divergence shapes the pace of sustainable logistics development, with progress supported by green energy and digitalisation in some economies and constrained by high energy intensity and delayed transitions in others.

Energy systems play a crucial role in sustainable logistics development. The Nordic countries, Sweden, Finland, and Denmark, stand out for their high shares of RES, including wind, biomass, and hydropower [28,29,30]. Germany pursues the Energiewende, a dynamic and intensive decarbonisation strategy. In contrast, Poland and the Baltic states remain more dependent on fossil fuels, although Estonia has actively diversified energy sources through LNG and wind. By 2016, Latvia, Estonia, and Lithuania were already close to achieving their RES and greenhouse gas reduction targets for 2020 [31,32].

Previous research highlights the importance of economic factors, such as GDP and labour productivity, and energy factors, including RES share and energy intensity, in sustainable transport and logistics [33]. However, the existing subject literature lacks a comprehensive approach. Most analyses focus on individual aspects of the transition, economic or environmental, rather than on integrating energy and logistics [34]. Some studies emphasise the positive impact of renewable energy investment in highly developed economies [35,36]. Others underline social factors, such as employment structure and labour market stability, particularly in less developed BSR countries [37]. Moreover, some previous studies have explored approaches such as strategic management in transport and logistics or green enterprise logistics systems within circular economy frameworks [38,39,40,41,42]. However, these analyses do not address the combined macroeconomic and energy-related determinants of sustainable logistics development. This study fills this gap by focusing on these factors and their influence on the sustainability of logistics across heterogeneous BSR economies. Thus, it demonstrates that the energy economy is one of the most significant factors influencing the pace and direction of sustainable logistics development.

3. Research Questions Development

Several external and internal factors shape the sustainable development of the logistics sector [43,44]. Due to the nature and purpose of this study, it focused on economic growth and energy transition. Previous studies emphasise that higher GDP per capita supports eco-innovation in the logistics sector [45,46]. Economically more developed countries have higher levels of education and understanding of environmental protection [47]. Furthermore, countries with higher levels of socioeconomic development are better prepared to implement ESG standards [48,49]. This, therefore, accelerates the transition to sustainable logistics. Consequently, a positive correlation between economic growth and the SD is expected.

At the same time, energy intensity is one of the main barriers to sustainable development. A high unit energy consumption per GDP indicates low energy efficiency, with higher greenhouse gas emissions. Research has shown that higher energy intensity in the logistics sector decreases the climate’s potential, limiting the green transition [50,51]. Therefore, it is predicted that energy intensity will negatively impact the sustainability of logistics.

The share of RES in ultimate energy consumption is one of the deciding factors for SD. The growing importance of RES directly adds to lower greenhouse gas emissions and the environmental costs of logistics activities [52,53,54]. Moreover, RES reduces dependency on fossil fuels, which is particularly beneficial for energy security and the volatility of world fuel markets. Previous research shows that a greater share of renewable energy in countries leads to faster development towards greening logistics [55,56]. A higher share of renewable energy in final energy consumption is assumed to positively influence the sustainable development of logistics in the BSR.

Based on the theoretical and empirical background, the study’s objectives are formulated as two focused research questions to capture the specific characteristics of the BSR:

• RQ1: How do GDP per capita, energy intensity, and the share of renewable energy shape sustainable logistics development in the Baltic Sea Region, and which factor is most influential in each national context?
• RQ2: In the Baltic Sea Region, how do electrification, labour productivity, and unemployment moderate the effects of economic growth and energy transition on logistics sustainability across countries?

These research questions allow us to examine the dual role of economic growth and energy transition in shaping sustainable logistics while explicitly considering country-specific differences and structural heterogeneity between Nordic and Central and Eastern European economies.

4. Research Methodology

4.1. Scope of the Study and Data Sources

This study focuses on eight countries of the BSR: Denmark, Estonia, Finland, Germany, Latvia, Lithuania, Poland, and Sweden. The study period is 2008–2023. The time span was chosen due to data availability and the relevance of this time’s economic and social events. The analysis considers the effects of the financial crisis, economic recovery processes, the COVID-19 pandemic, and the initial phase of the European Green Deal and the Fit for 55 packages. The principal source of information is the Eurostat database. The data are all yearly and constitute a balanced panel.

4.2. Construction of the Dependent and Explanatory Variables

The dependent variable is the synthetic indicator of sustainable logistics development (SD). This indicator was constructed based on three economic, environmental, and social dimensions, each represented by a set of standardised diagnostic variables. The economic dimension included: the number of transport enterprises, turnover or gross premiums, production value, value added at factor cost, gross operating surplus, total purchases of goods and services, personnel costs, and the investment rate. The environmental dimension comprised methane, nitrous oxide, sulphur oxides, carbon dioxide, and ammonia emissions. The social dimension covered wages and salaries, social security costs, the number of employees, apparent labour productivity, and gross value added per employee. All variables were normalised using the binary method to ensure comparability across countries and years. The sub-indices for each dimension were calculated as arithmetic means, then aggregated into a composite SD index, with all variables and dimensions assigned equal weights.

Thus, the composite SD index served as the dependent variable in the panel models, including the dynamic GMM specification described later, where lagged variables were used to address potential endogeneity.

This study focuses on six independent variables, reflecting key economic and energy conditions:

-

EI—the energy intensity of the economy. It expresses energy use efficiency in production and transport processes, where higher values indicate lower efficiency and greater CO₂ emissions. Elevated EI limits environmental sustainability and contradicts the principles of resource-efficient logistics, as it increases the environmental burden of transport operations [57,58,59].

-

RES—the share of renewable energy sources in final energy consumption reflects progress toward a low-carbon economy. A higher RES share supports decarbonisation and reduces dependence on fossil fuels, aligning with sustainable logistics and circular economy principles [59,60,61,62]. Empirical evidence shows that countries with ambitious RES targets, such as Sweden, achieve faster emission reductions and more advanced green transport systems [63].

-

UR—the unemployment rate reflects the social dimension of sustainability by capturing labour market stability and inclusion. Stable employment enhances adaptive capacity and supports investments in sustainable technologies [64,65]. Conversely, high unemployment weakens these capacities and limits resources for green investments, while excessive labour shortages may hinder the pace of the green transition [66].

-

GDP_pps—GDP per capita in purchasing power parity reflects economic prosperity and investment capacity. Within the framework of sustainable growth theory, higher GDP per capita enhances innovation potential and facilitates the adoption of clean technologies and sustainable logistics infrastructure [67,68,69]. However, when GDP growth relies on resource-intensive or fossil-fuel-based production, it may counteract sustainability goals and increase environmental pressures [70].

-

LP—the growth rate of labour productivity in the logistics sector reflects the efficiency and innovativeness of logistics systems. Higher labour productivity supports competitiveness, cost efficiency, and the more sustainable use of resources in transport and logistics operations [71]. Maintaining low or stagnant productivity can reduce returns on investment and limit the potential for sustainable growth in the transport sector. Therefore, productivity growth is an economic and environmental driver of sustainable logistics development [72].

-

RE—the electrification rate represents the share of electricity in total final energy consumption and indicates progress in decarbonising energy use. It is a crucial technological driver of the energy transition, particularly relevant to transport and logistics, as it can substantially reduce emissions from fleets and logistics operations [73]. Research demonstrates that electrification offers significant environmental benefits by eliminating tailpipe emissions and lowering noise levels, which are vital in densely populated or environmentally sensitive areas [74,75]. However, the literature notes operational challenges during the transition, particularly in long-haul transport, where infrastructure readiness and climatic conditions may affect efficiency and costs. Therefore, adaptive policy interventions, such as expanding high-power charging networks and supporting R&D in battery innovation, are essential to ensure the sustainable electrification of logistics systems [76].

These variables were selected to capture the relationships between economic development, the labour market, energy transition, and environmental performance.

4.3. Research Stages

The analysis was conducted in several stages:

-

Descriptive analysis—descriptive statistics and correlation coefficients were calculated for all variables.

-

OLS models for individual countries—linear regressions were performed in each country:

SD_it= β₀ + β₁ EI_it + β₂ RES_it + β₃ UR_it + β₄ GDP_pps_it + β₅ LP_it + β₆ RE_it + ε_it

(1)

where

i—country index, i = 1, …, N.

t—time index (year), t = 2008, …, 2023.

β₀—intercept (constant term) in pooled OLS.

β_k (for k = 1, …, 6)—slope coefficients measuring the marginal effect of each explanatory variable on SD.

ε_it, u_it—idiosyncratic error terms (disturbances) with zero mean; in FE u_it is the within-country residual after controlling for fixed effects.

-

Fixed-effects (FE) panel models—these were used to control for unobserved, country-specific factors that remain constant over time:

SD_it = α_i + γ_t + β₁EI_it + β₂RES_it + β₃UR_it + β₄GDP_pps_it + β₅LP_it + β₆RE_it + u_it,

(2)

where

α_i—country fixed effect (time-invariant unobserved heterogeneity specific to country i, e.g., geography, institutions).

γ_t—time fixed effect.

-

Dynamic panel model (GMM, one-step)—in the final stage, generalised moments estimation was used, which allowed for the inclusion of a lagged dependent variable and mitigated the problem of endogeneity:

SD_it = ρSD_i,t−1 + β₁EI_it + β₂RES_it + β₃UR_it + β₄GDP_pps_it + β₅LP_it + β₆RE_it + α_i +η_t + ε_it.

(3)

where

ρ—coefficient on the lagged dependent variable capturing persistence/inertia in SD.

SD_i,t−1—one-period lag of the dependent variable.

η_t—time effect in the dynamic specification.

The dynamic model employed internal instruments following the Arellano–Bond approach to address potential endogeneity. Lagged dependent variable (SD) and lagged endogenous regressors—GDP_pps, EI, and RE—were instrumented in levels, and their first differences were instrumented with two- and three-period lags. The instrument matrix was constrained to avoid over-identification and reduce endogeneity between energy, growth and sustainability variables.

4.4. Diagnostic Tests

All models were subjected to diagnostic tests:

-

Before estimating the OLS models, all key diagnostic tests were conducted: the White test and the Breusch–Pagan test for heteroskedasticity, the Durbin–Watson test for autocorrelation, the Jarque–Bera test (χ²-based) for normality of residuals, and the Variance Inflation Factor (VIF) test for multicollinearity. Also, the Lagrange Multiplier (LM) and LMF tests listed in the OLS tables were used for random effects and testing across specifications. The LM test follows the Breusch–Pagan approach and tests whether the variance across entities is zero. The LMF test tests whether fixed effects are jointly significant compared to the pooled OLS baseline. Both tests support the robustness of the model selection process.

-

In the FE models, the F test was used to confirm the joint significance of variables and the coefficients of determination (R²) were calculated.

-

In the dynamic GMM model, Arellano–Bond tests (AR(1), AR(2)) were performed to detect autocorrelation of residuals.

-

The Wald test was used to verify the joint significance of variables.

-

The Sargan test assessed the validity of the instruments used.

The obtained results confirm the high explanatory power of the models. However, the Sargan test (p = 0.0449) indicated borderline instrument validity (typical for small macro panels). To mitigate this issue, the number of instruments was minimised. This was accomplished by reducing the matrix and imposing a lag depth constraint. Therefore, the results should be interpreted cautiously but remain statistically significant.

4.5. Justification for the Choice of Methods

OLS, fixed-effects, and dynamic panel models allowed for capturing cross-country variation and adjustment processes over time. Applying this combination of methods increased the robustness of the results and provided a more comprehensive perspective on the macroeconomic drivers of sustainable logistics development in the BSR.

5. Results

Table 1. Sustainable Logistics Development (SD) indicators in BSR Countries and Descriptive Statistics.

Year	Denmark	Estonia	Finland	Germany	Latvia	Lithuania	Poland	Sweden
2008	0.77	0.67	0.79	0.68	0.67	0.55	0.62	0.76
2009	0.7	0.65	0.77	0.66	0.64	0.54	0.58	0.71
2010	0.75	0.61	0.77	0.67	0.61	0.55	0.59	0.77
2011	0.72	0.69	0.79	0.69	0.65	0.58	0.62	0.82
2012	0.72	0.71	0.79	0.66	0.7	0.6	0.63	0.88
2013	0.75	0.7	0.79	0.68	0.7	0.67	0.63	0.86
2014	0.76	0.71	0.81	0.71	0.71	0.69	0.66	0.85
2015	0.75	0.75	0.84	0.75	0.72	0.67	0.66	0.85
2016	0.73	0.77	0.84	0.76	0.73	0.66	0.67	0.83
2017	0.77	0.81	0.86	0.78	0.73	0.67	0.64	0.86
2018	0.75	0.81	0.87	0.79	0.78	0.69	0.71	0.83
2019	0.77	0.88	0.87	0.82	0.79	0.66	0.76	0.84
2020	0.82	0.84	0.87	0.86	0.75	0.61	0.79	0.87
2021	0.87	0.84	0.84	0.93	0.77	0.64	0.87	0.86
2022	0.88	0.86	0.83	0.92	0.77	0.68	0.89	0.84
2023	0.87	0.85	0.84	0.93	0.77	0.66	0.88	0.85
Mean	0.77	0.76	0.82	0.77	0.72	0.63	0.7	0.83
Std. Dev.	0.06	0.08	0.04	0.1	0.05	0.05	0.1	0.05
Max	0.88	0.88	0.87	0.93	0.79	0.69	0.89	0.88
Min	0.7	0.61	0.77	0.66	0.61	0.54	0.58	0.71

Source: own study based on https://ec.europa.eu/eurostat/databrowser/product/view/sbs_sc_ovw?category=bsd.sbs.sbs_ovw (accessed on 16 September 2025).

shows the sustainable development indicator of the logistics sector (SD) from 2008 to 2023 for eight countries in the BSR: Denmark, Estonia, Finland, Germany, Latvia, Lithuania, Poland, and Sweden.

The data show that the highest mean SD values were recorded in Sweden (0.83) and Finland (0.82). In these countries, sustainable logistics development is stable and strongest. The lowest mean values were achieved in Lithuania (0.63) and Poland (0.70), with relatively high variability visible in both cases (standard deviations of 0.05 and 0.10, respectively).

Most considerable SD fluctuations occurred in Poland and Germany (from 0.58 to 0.89; σ = 0.10 and from 0.66 to 0.93; σ = 0.10). This indicates dynamic transformation processes in these countries during the period under review. The most stable results were observed in Finland (σ = 0.04). Analysis of minimum and maximum values shows that countries in the region generally improved their SD scores after 2015. In the 2020–2023 period, SD levels were significantly higher than in the initial years of observation. The highest SD value was recorded in Germany in 2021 and 2023 (0.93), while the lowest was in Lithuania in 2009 (0.54).

The results show variation in SD levels across countries. Sweden and Finland lead the way, with dynamic growth observed in Germany, Estonia, and Poland and the lowest SD levels in Lithuania and Latvia.

Table 2. Descriptive Statistics of the explanatory variables.

		EI	RES	UR	GDP_pps	LP	RE
Denmark	max	119.61	44.40	7.80	48,266.20	11.40	2.13
	min	57.79	18.54	4.50	30,595.90	−13.80	0.70
	Mean	89.64	30.92	6.21	37,319.47	−0.07	1.04
	Std. Dev.	19.30	7.74	1.07	5266.59	7.50	0.41
Germany	max	141.57	21.56	7.50	45,253.10	9.80	2.90
	min	69.58	10.07	2.90	28,855.70	−12.50	1.74
	Mean	110.26	15.42	4.56	35,978.97	−0.64	1.99
	Std. Dev.	20.75	3.36	1.46	4544.12	4.69	0.31
Estonia	max	281.86	40.95	16.60	30,583.40	15.50	1.09
	min	105.81	18.81	4.50	15,454.00	−19.10	0.31
	Mean	193.57	29.08	8.49	22,643.36	0.60	0.67
	Std. Dev.	49.77	5.69	3.49	4621.79	8.93	0.23
Latvia	max	178.63	43.72	19.70	27,063.70	21.20	1.19
	min	91.23	29.81	6.30	12,506.00	−11.80	0.78
	Mean	136.51	37.82	11.29	18,329.38	1.68	0.93
	Std. Dev.	24.66	4.14	4.42	4197.92	8.58	0.13
Lithuania	max	206.71	31.93	17.80	33,029.20	16.60	0.45
	min	76.62	17.82	6.00	13,722.50	−13.60	0.29
	Mean	126.82	24.31	10.13	22,233.83	1.53	0.37
	Std. Dev.	33.59	3.79	3.65	5701.75	6.67	0.05
Poland	max	181.36	16.63	10.60	29,572.10	13.80	1.93
	min	86.97	7.69	2.80	14,261.40	−14.00	1.24
	Mean	135.83	12.47	6.61	20,506.54	1.36	1.51
	Std. Dev.	25.12	2.86	2.96	4518.07	7.20	0.21
Finland	max	234.04	50.75	9.40	40,029.20	7.90	3.04
	min	150.12	31.05	6.80	28,678.60	−18.60	1.46
	Mean	196.01	39.04	8.06	32,455.64	−1.08	1.73
	Std. Dev.	23.90	5.70	0.73	3244.48	6.48	0.42
Sweden	max	176.02	66.39	8.90	42,892.00	7.30	5.84
	min	102.88	43.92	6.30	30,274.70	−16.00	2.71
	Mean	146.35	53.66	7.69	35,394.23	−0.75	3.45
	Std. Dev.	21.87	6.70	0.77	3324.33	6.24	0.85

Source: own study based on https://ec.europa.eu/eurostat/data/database (accessed on 16 September 2025).

presents descriptive statistics of explanatory variables of SD in the BSR countries. The results reveal huge differences between countries. Denmark, Sweden, and Finland are characterised by a high percentage of renewable energy, a relatively high per capita GDP, and low unemployment. In contrast, Poland, Germany, Lithuania, Latvia, and Estonia have a lower share of renewable energy and a higher energy intensity of their economies. Estonia, Latvia, and Lithuania also experience higher unemployment rates and greater variability in labour productivity.

The results confirm that the Nordic countries are achieving better results in terms of sustainable energy development. The economies of Central and Eastern Europe are lagging, especially in terms of energy transition.

Table 3. Correlation between SD and explanatory variables.

	Variable	EI	RES	UR	GDP_pps	LP	RE
Denmark	Correlation with SD (r)	−0.809	0.779	−0.698	0.917	0.038	0.897
	p-value	0.00015	0.00037	0.00266	0.0000006	0.889	0.0000024
Germany	Correlation with SD (r)	−0.943	0.954	−0.781	0.944	0.242	0.683
	p-value	0.00000004	0.000000011	0.00035	0.000000039	0.366	0.0035
Estonia	Correlation with SD (r)	−0.921	0.843	−0.888	0.952	0.005	−0.807
	p-value	0.00000042	0.000041	0.0000045	0.000000014	0.984	0.00016
Latvia	Correlation with SD (r)	−0.906	0.892	−0.960	0.867	0.078	−0.693
	p-value	0.0000014	0.0000034	0.0000000039	0.000013	0.773	0.0029
Lithuania	Correlation with SD (r)	−0.779	0.694	−0.793	0.615	0.113	−0.621
	p-value	0.00037	0.0029	0.00025	0.011	0.677	0.010
Poland	Correlation with SD (r)	−0.907	0.918	−0.845	0.964	0.078	−0.796
	p-value	0.0000013	0.00000052	0.000038	0.0000000018	0.773	0.00023
Finland	Correlation with SD (r)	−0.689	0.746	−0.335	0.566	−0.217	0.334
	p-value	0.0031	0.00090	0.204	0.022	0.420	0.206
Sweden	Correlation with SD (r)	−0.514	0.541	−0.010	0.539	−0.023	0.344
	p-value	0.042	0.030	0.970	0.031	0.932	0.192

Source: own study based on https://ec.europa.eu/eurostat/data/database (accessed on 16 September 2025).

shows the correlation coefficients between sustainable development (SD) and explanatory variables in eight Baltic Sea nations. The results indicate the inverse correlation between EI and SD in all countries. Decreasing the energy intensity of the economy, therefore, promotes greater sustainability. UR is inversely correlated with SD as well, and the correlations are particularly high for Latvia (r = −0.960, p < 0.001) and Estonia (r = −0.888, p < 0.001). Increased unemployment, therefore, erodes the impact of sustainable development.

SD, RES, and GDP_pps have a positive and strongly significant relationship, thus demonstrating that renewable energy development and economic growth are essential in SD. Extremely significant impacts were evident in Germany (RES: r = 0.954; GDP_pps: r = 0.944) and Poland (RES: r = 0.918; GDP_pps: r = 0.964).

The RE findings are inconclusive. In Denmark, the relationship is highly positive and strong (r = 0.897, p < 0.001), while in Estonia and Poland, there are strong negative relationships (r = −0.807, p < 0.001; r = −0.796, p < 0.001). The results confirm the different structural conditions of the economies.

In addition, LP is not statistically significantly correlated with SD in any countries examined (p > 0.05).

The results of the OLS estimation indicate heterogeneous determinants of logistics sector development in the Baltic Sea countries (

Table 4. The OLS results.

Country		Coefficient	Std. Error	t-Statistic	p-Value
Denmark	const	0.0447549	0.194780	0.2298	0.8218
	EI	0.00171158	0.000870839	1.965	0.0711
	GDP_pps	1.54499 × 10−5	3.19081 × 10−6	4.842	0.0003
	R2 = 0.876798; Durbin-Watson = 1.692833
	LM = 7.34888; p = P(Chi-square(5) > 7.34888) = 0.195959
	Chi-square(2) = 0.574789; p = 0.750216
	LMF = 0.0938855; p = P(F(1. 12) > 0.0938855) = 0.764546
Germany	const	0.348809	0.0360146	9.685	<0.0001
	RES	0.0271400	0.00228229	11.89	<0.0001
	R2 = 0.909916; Durbin-Watson = 0.934564
	LM = 1.23106; p = P(Chi-square(2) > 1.23106) = 0.540354
	Chi-square (2) = 2.13343; p = 0.344137
Estonia	const	0.550381	0.0670002	8.215	<0.0001
	UR	−0.00775615	0.00272008	−2.851	0.0136
	GDP_pps	1.21149 × 10−5	2.05638 × 10−6	5.891	<0.0001
	R2 = 0.942256; Durbin-Watson = 1.046832
	LM = 2.86939; p = P(Chi-square(5) > 2.86939) = 0.720113
	Chi-square(2) = 2.41052; p = 0.299614
	LMF = 3.81521; p = P(F(1. 12) > 3.81521) = 0.0745038
Latvia	const	−0.0416483	0.0819521	−0.5082	0.6198
	RES	0.0180487	0.00215658	8.369	<0.0001
	UR	0.00686144	3.82450 × 10−6	1794.	<0.0001
	R2 = 0.999999; Durbin-Watson = 1.639702
	LM = 5.01652; p = P(Chi-square(5) > 5.01652) = 0.413868
	Chi-square(2) = 0.50466; p = 0.776988
	LMF = 0.103493; p = P(F(1. 12) > 0.103493) = 0.753213
Lithuania	const	1.25266	0.0949018	13.20	<0.0001
	EI	−0.00178427	0.000312355	−5.712	<0.0001
	UR	−0.0137652	0.00242222	−5.683	0.0001
	GDP_pps	−1.14425 × 10−5	2.25779 × 10−6	−5.068	0.0003
	R2 = 0.903943; Durbin-Watson = 1.350481
	LM = 13.5028; p = P(Chi-square(9) > 13.5028) = 0.141142
	Chi-square (2) = 2.89371; p = 0.235309
	LMF = 1.12755; p = P(F(1. 11) > 1.12755) = 0.311055
Poland	const	0.263106	0.0327089	8.044	<0.0001
	GDP_pps	2.12692 × 10−5	1.55769 × 10−6	13.65	<0.0001
	R2 = 0.930154; Durbin-Watson = 0.941021
	LM = 0.0873766; p = P(Chi-square(2) > 0.0873766) = 0.957252
	Chi-square(2) = 3.97172; p = 0.137262
	LMF = 2.26734; p = P(F(1. 13) > 2.26734) = 0.156033
Finland	const	0.598448	0.0316475	18.91	<0.0001
	RES	0.00893691	0.00131824	6.779	<0.0001
	RE	−0.0719935	0.0177841	−4.048	0.0014
	R2= 0.804148; Durbin-Watson = 0.933739
	LM = 2.89309; p = P(Chi-square(5) > 2.89309) = 0.716464
	Chi-square(2) = 0.728363; p = 0.694765
	LMF = 3.54887; p = P(F(1. 12) > 3.54887) = 0.084038
Sweden	const	170.277	3.21807	52.91	<0.0001
	EI	−0.527028	0.0196351	−26.84	<0.0001
	RES	−1.71914	0.0465774	−36.91	<0.0001
	R2 = 0.995302; Durbin-Watson = 1.399009
	LM = 4.9579; p = P(Chi-square(5) > 4.9579) = 0.421039
	Chi-square(2) = 1.90987; p = 0.384837
	LMF = 0.67977; p = P(F(1. 12) > 0.67977) = 0.425751

Source: own study based on https://ec.europa.eu/eurostat/data/database (accessed on 16 September 2025).

). In Denmark, GDP per capita positively and significantly affects SD (1.54 × 10⁻⁵; p < 0.001), while energy intensity (0.0017; p = 0.071) is near significance. In Poland, a similar effect is observed for GDP per capita (2.13 × 10⁻⁵; p < 0.001; R² = 0.93). In Estonia, the positive effect of GDP per capita on SD (1.21 × 10⁻⁵; p < 0.001) is combined with an adverse effect of the unemployment rate (−0.0078; p < 0.05).

Renewable energy is a significant development factor in Germany, Latvia, and Finland. In Germany, the coefficient value is 0.0271 (p < 0.001) and in Finland, 0.0089 (p < 0.001), with resource efficiency also having a negative impact (−0.0720; p < 0.01). In Latvia, the model indicates powerful positive associations with both renewable energy (0.0180; p < 0.001) and the unemployment rate (0.0069; p < 0.001), resulting in an almost perfect fit (R² ≈ 1).

Energy intensity has a negative impact in most cases, particularly in Lithuania (−0.0018; p < 0.001) and Sweden (−0.5270; p < 0.001). In Lithuania, all the variables considered—GDP per capita (−1.14 × 10⁻⁵; p < 0.01), unemployment (−0.0138; p < 0.001), and energy intensity—have a negative impact. In Sweden, both energy intensity and renewable energy (−1.7191; p < 0.001; R² = 0.995) have a significant and negative impact.

In all countries, the coefficients of determination exceed 0.80, confirming the stability of the obtained results and the correctness of the adopted model specifications.

Table 5. Panel model estimation results (Fixed Effects).

Variable	Model 1 (all Variables)	Model 2 (Model with Statistically Significant Variables)
const	0.6608 (0.1016) ***	0.5922 (0.0857) ***
EI	–0.00061 (0.00031) *	–0.00055 (0.00028) **
RES	–0.00059 (0.00166)	–
UR	–0.00277 (0.00240)	–
GDP_pps	8.57 × 10−6 (2.30 × 10−6) ***	9.86 × 10−6 (1.83 × 10−6) ***
LP	0.00047 (0.00051)	–
RE	–0.0170 (0.0131)	–0.0279 (0.0094) ***
R2 (LSDV)	0.832	0.828
R2 (within)	0.697	0.691
F-test	43.33 (p < 0.001)	56.50 (p < 0.001)
N	128	128

* p < 0.10; ** p < 0.05; *** p < 0.01; Source: own study based on https://ec.europa.eu/eurostat/data/database (accessed on 16 September 2025).

shows the panel model estimation results. Model 1 contains a complete set of explanatory variables. It indicates a statistically significant impact of the economy’s energy intensity (EI) and GDP per capita in PPS (GDP_pps) on SD. High energy intensity negatively impacts, while increasing per capita income promotes SD improvement. The model is characterised by a high fit (LSDV R² = 0.83), but autocorrelation of residuals was observed, suggesting certain limitations in interpreting the results.

Model 2 includes only significant variables (EI, GDP_pps, RE) and confirms the negative impact of energy intensity and the positive role of economic development. Model 2 also showed a good fit (R² = 0.83); all variables were significant at p < 0.05.

Table 6. Dynamic one-step panel estimation, using 112 observations.

	Coefficient	Std. Dev.	Z	p-Value
SD(−1)	0.00414863	0.179980	0.02305	0.9816
EI	−0.000594228	0.000211909	−2.804	0.0050
UR	−0.00625159	0.00359272	−1.740	0.0818
GDP_pps	6.94833 × 10−6	2.68158 × 10−6	2.591	0.0096
RE	−0.0212943	0.0116549	−1.827	0.0677
AR(1): z = 0.881816 [0.3779]
AR(2): z = −1.46598 [0.1427]
Sargan test: Chi-square(13) = 22.738 [0.0449]
Wald (joint): Chi-square(5) = 1896.77 [0.0000]

Source: own study based on https://ec.europa.eu/eurostat/data/database (accessed on 16 September 2025).

shows the results of the one-step dynamic panel model based on 112 observations. The lagged dependent variable SD(−1) is statistically insignificant. Energy intensity (EI) has a significant negative impact. This means that higher energy intensity reduces sustainable development in the logistics sector. GDP_pps has a positive and significant impact. It confirms the supporting role of economic growth. The UR and RE are negative and close to significance. They point to possible limits from the labour market and resource use.

Diagnostic tests confirm the reliability of the model. There is no first- or second-order autocorrelation (AR(1) p = 0.3779, AR(2) p = 0.1427). The Wald test rejects the null hypothesis of joint insignificance (p < 0.001). The model has strong explanatory power. The Sargan test (p = 0.0449) shows doubts about the validity of the instruments. Results should be interpreted with caution.

6. Discussion

The analyses indicate that energy intensity (EI) is the main barrier to sustainable logistics development, especially in Sweden and Lithuania. In contrast, countries with a more diversified energy mix, like Denmark and Finland, display higher resilience and stability. GDP per capita (GDP_pps) consistently supports logistics sustainability, reflecting that higher income enables investments in infrastructure, modernisation, and energy-efficient solutions, which facilitate the logistics transition [68,70].

The moderate Sargan statistic is a residual endogeneity threat, mainly from two-way relationships between GDP growth, energy intensity, and logistics sustainability. Lagged internal instrumenting reduced this bias, and qualitative trends concurred with the fixed-effects results.

The electrification rate (RE) negatively affects several countries, including Finland, Sweden, Estonia, and Poland. This likely reflects that the pace of electrification can temporarily exceed the adaptive capacity of logistics systems, raising operational pressures and investment costs. The effect of RE is context-dependent: Denmark benefits from electrification due to well-prepared infrastructure and energy systems, highlighting the importance of national readiness for achieving positive outcomes [77].

Renewable energy sources (RES) produce mixed effects across the BSR. While RES contributes positively to sustainable logistics in Germany, Finland, and Latvia, it negatively affects Sweden. Structural and transitional factors drive this: Sweden relies heavily on hydropower, biomass, and wind, with electricity central to industry and transport. Rapid renewable energy expansion entails high transition costs, grid balancing challenges, regional energy disparities, and the need for storage and flexible infrastructure, which can temporarily reduce energy efficiency and increase logistics costs. Energy-intensive sectors, such as steel, paper, and chemicals, face higher costs during decarbonisation, lowering short-term productivity [78,79]. Policies aiming for climate neutrality by 2045 accelerate electrification and infrastructure development, requiring substantial investment. Spikes in electricity prices after 2020 further intensified these pressures.

Labour productivity growth (LP) and unemployment rate (UR) have inconsistent effects across the region. Country-specific models show that higher unemployment constrains SD in Estonia, Latvia, and Lithuania, indicating that labour market conditions can limit logistics sustainability locally. However, these effects do not form a uniform regional pattern.

The results also provide clear answers to the research questions. GDP per capita, energy intensity, and RES shape sustainable logistics differently across the BSR. High-income Nordic countries achieve stable and high SD due to diversified energy, robust infrastructure, and supportive policies. In contrast, Central and Eastern European countries face challenges from higher EI, lower renewable energy shares, and labour market constraints. The dominant factor varies by country: GDP per capita in Poland and Denmark, energy intensity in Sweden and Lithuania, and RES in Germany and Latvia (positive) versus Sweden (negative).

Structural and transitional factors, electrification, labour productivity, and unemployment, moderate the effects of economic growth and energy transition in a country-specific manner. Electrification accelerates sustainability where infrastructure and adaptability allow, but can reduce SD in countries with rapid transitions or energy-intensive sectors [77]. Labour market constraints particularly affect the Baltic states, and variations in electrification and renewable integration explain the heterogeneous outcomes across countries.

Overall, macroeconomic and energy factors strongly shape sustainable logistics development in the BSR, with observed heterogeneity reflecting the differing pace of economic and energy transitions between Nordic and Central and Eastern European countries [80,81]. The findings underline that energy mix stability, policy design, and the adaptive capacity of logistics systems are key to maximising the benefits of growth and renewable energy integration.

7. Conclusions

This study aimed to assess the impact of macroeconomic and energy-related factors on the development of the logistics sector in eight BSR countries from 2008 to 2023. The analysis confirms that sustainable logistics development in the BSR is strongly determined by macroeconomic and energy-related factors, with GDP per capita and energy intensity emerging as the most consistent drivers. At the same time, the results highlight significant national differences, ranging from RES-driven leaders such as Germany, Finland, and Latvia to income-based growth models in Poland, Denmark, and Estonia, and more complex, paradoxical outcomes in Lithuania and Sweden. Static and dynamic panel models reinforce hypotheses, confirming the positive role of GDP per capita, the negative role of energy intensity, and, to a lesser extent, the adverse effect of the electrification rate. Labour market variables, including unemployment, proved relevant in selected national cases but not at the regional level. These findings underline the heterogeneity of sustainable logistics pathways across the region.

This study has several limitations. First, the scope was narrowed to macroeconomic and energy-related determinants, excluding institutional, technological, and regulatory dimensions. Second, while the time span of this research (2008–2023) captured major shocks (financial crisis, COVID-19, Green Deal), the long-term effects of the 2022 energy crisis remain only partially reflected. Third, the relatively small sample of eight economies limits generalisability. Fourth, GMM model results should be interpreted cautiously due to some doubts about instrument validity. These methodological limitations and structural heterogeneity among BSR economies may influence the robustness of cross-country comparisons.

In practical terms, the findings suggest that policymakers should tailor energy transition policies to national contexts, ensuring that electrification and renewable expansion proceed at a pace compatible with logistics sector adaptability. Strengthening infrastructure resilience and improving energy efficiency can mitigate transition costs and support long-term sustainability.

Future research could extend the scope by incorporating governance and technological innovation. It could broaden the geographical scale by comparing the BSR with the Mediterranean or Black Sea regions. It may also refine the methodology using nonlinear or spatial econometric techniques to capture spillover effects. In addition, a greater focus on labour market adaptation would provide valuable insights.