From Survey-Based LPI to Logistics Performance Indicators 2.0: Connectivity, Dwell Time, and Sustainable Economic Performance in Arab and European Economies

Abstract

Logistics performance measurement is shifting from perception-based country rankings toward operational indicators that capture connectivity, movement speed, and delay conditions. This study examines how the transition from the legacy Logistics Performance Index to the World Bank’s Logistics Performance Indicators 2.0 changes logistics performance assessment and its economic and environmental interpretation in 19 Arab and European economies. The analysis uses a three-stage design. First, it examines the legacy LPI panel for 2007–2023 to provide a historical logistics-performance baseline. Second, it uses 2023–2024 LPI 2.0 indicators to benchmark operational performance through maritime, aviation and postal connectivity, import dwell time, port turnaround time, transshipment time, and postal delivery time. Third, it links these indicators to GDP per capita growth and CO₂ emissions through exploratory cross-sectional models. The results suggest that operational logistics indicators reveal performance differences that are partly hidden by perception-based aggregate scores. Connectivity is positively associated with economic performance, while longer dwell time and port turnaround are associated with weaker economic outcomes and higher emissions. The environmental results are conditional: stronger connectivity is associated with higher emissions where logistics growth remains energy-intensive, while smoother operations are associated with lower congestion-related emissions. The study contributes by positioning LPI 2.0 as an operational dashboard rather than a replacement ranking system and by showing how connectivity, time performance, and regional structure shape logistics and sustainability outcomes. The findings are interpreted as exploratory associations because LPI 2.0 currently provides limited time coverage.

1. Introduction

Logistics performance is no longer measured only through expert perceptions of customs, infrastructure, shipment arrangements, service quality, tracking and timeliness. The World Bank Logistics Performance Indicators 2.0 mark a shift toward operational, shipment-level and tracking-based information. LPI 2.0 measures connectivity and time performance across maritime, aviation and postal logistics. Unlike the legacy survey-based LPI, it does not produce a single country ranking. It compares countries through specific indicators such as maritime connectivity, aviation connectivity, postal connectivity, import dwell time, port turnaround, transshipment time, and postal delivery time [1,2].

The efficiency and reliability of transport and logistics systems have become central to national competitiveness and global trade integration [3,4,5]. Yet, the literature still relies heavily on the legacy LPI and its six perception-based components. This creates three gaps. First, limited empirical work has incorporated the operational structure of LPI 2.0. Second, aggregate scores hide the separate effects of connectivity, dwell time, delay dispersion, and port-level bottlenecks. Third, few studies compare Arab and European economies through a framework that considers logistics maturity, institutional quality, and environmental pressure.

This study addresses these gaps by combining the historical LPI framework with LPI 2.0 operational indicators. It does not treat LPI 2.0 as a replacement ranking system. Instead, it uses LPI 2.0 as an operational dashboard that separates network access, movement speed, and delay conditions across logistics modes. This framing is important because the same logistics system can perform well on connectivity while still facing costly delays, port congestion, or carbon-intensive logistics growth [1,2].

The Arab–European comparison provides a theoretically useful setting because logistics indicators are not expected to translate into economic and environmental outcomes uniformly across institutional contexts. Trade cost and market access theory suggests that connectivity generates stronger gains when transport networks are supported by regulatory coordination, interoperable infrastructure, and predictable border procedures. Transaction cost logic further suggests that the economic cost of dwell time and port delay is higher where firms face weaker coordination, fragmented clearance systems, or limited hinterland connectivity. Environmental efficiency theory adds a third condition: logistics improvements may reduce congestion-related emissions, but they may also increase total emissions when expanded connectivity operates within carbon-intensive energy systems. Therefore, regional context is not treated as a simple geographic distinction. It is used as a proxy for differences in logistics maturity, institutional coordination, market integration, and energy structure that may condition the strength and direction of logistics–performance associations.

In practical terms, European logistics systems often benefit from dense multimodal networks and harmonized regulatory frameworks, while Arab economies include both advanced hub systems and economies with higher institutional fragmentation. This variation strengthens the logic of examining both regional context and logistics maturity as possible conditioning factors.

This study addresses four research questions:

(1)

How does the shift from the legacy LPI to LPI 2.0 change the measurement of logistics performance?

(2)

How are LPI 2.0 connectivity indicators associated with economic performance in Arab and European economies?

(3)

How are dwell time, transshipment time, and port turnaround time associated with economic and environmental outcomes?

(4)

Does regional context condition the association between operational logistics indicators, GDP per capita growth, and CO₂ emissions?

Economic performance is measured mainly through GDP per capita growth. Environmental performance is assessed through CO₂ emissions and energy–structure indicators. The term sustainable logistics outcome is used only when economic improvement is interpreted alongside lower emissions or reduced congestion-related environmental pressure. The paper contributes by aligning each measurement system with its proper inference level: the legacy LPI supports historical panel analysis, while LPI 2.0 supports operational benchmarking and exploratory association testing.

2. Literature Review and Conceptual Development

2.1. Theoretical Foundation: Trade Costs, Market Access, and Logistics Efficiency

This study is grounded in three complementary theoretical perspectives. First, trade cost and gravity-based logic explains why logistics connectivity matters for economic performance. In gravity models of trade, distance, border frictions, transport time and administrative barriers reduce trade intensity. Logistics systems reduce these frictions by improving network access, shipment reliability, and cross-border coordination [4,5,6]. From this view, connectivity does not create economic performance by itself. It improves market access by reducing spatial and procedural frictions that limit trade participation.

Second, transaction cost economics explains why time-based logistics indicators matter. Long dwell time, slow port turnaround, and uncertain delivery time increase search, monitoring, coordination, inventory and delay costs. These costs are often hidden in national-level logistics indicators, yet they shape firm competitiveness and supply chain reliability [5,6]. LPI 2.0 is useful because it captures operational frictions that are closer to actual transaction cost pressure than perception-based scores.

Third, the environmental logic of logistics performance depends on two competing mechanisms. The efficiency mechanism suggests that smoother logistics operations reduce idle time, congestion, fuel waste, and avoidable emissions. The scale mechanism suggests that stronger connectivity may increase transport activity, trade volume, and emissions when logistics growth relies on carbon-intensive energy systems [7]. This creates a conditional relationship between logistics performance and CO₂ emissions. Efficient operations may reduce emissions per shipment, while expanded connectivity may increase total emissions if energy systems remain fossil fuel-dependent.

Together, these perspectives justify the paper’s focus on operational indicators rather than a single aggregate logistics score. They also explain why connectivity, dwell time, turnaround time, transshipment time, and postal delivery time should be examined separately. The theoretical contribution of this study lies in linking LPI 2.0 indicators to market access, transaction cost reduction and efficiency-versus-scale environmental mechanisms.

It is important to clarify the inference level of these mechanisms. Market access, trade cost reduction, and transaction cost pressure are used as theoretical explanations for why connectivity and time-based indicators may be associated with economic performance. They are not estimated as separate mediating variables in the empirical model because comparable cross-country trade cost measures are not consistently available for the LPI 2.0 sample. The empirical analysis therefore tests direct associations between operational logistics indicators and outcome variables, while using trade cost and transaction cost theories to interpret the expected direction of these associations.

2.2. From Legacy LPI to Logistics Performance Indicators 2.0

The Logistics Performance Index has been widely used as a global benchmark for national logistics systems. The legacy LPI evaluates six core dimensions: customs efficiency, infrastructure quality, international shipment performance, logistics competence, tracking and tracing, and timeliness [1,3]. These dimensions have been applied in empirical research on trade facilitation, economic growth, and supply chain efficiency.

Despite its wide use, the legacy LPI has methodological and conceptual limits. It relies on perception-based survey data from logistics professionals. Expert judgment offers valuable insight, but it introduces subjectivity and can reduce comparability across countries and years. The index also lacks operational granularity. It does not directly capture port congestion, dwell time, transshipment delays or corridor-level throughput dynamics. Its intermittent publication also limits the analysis of rapid operational changes.

LPI 2.0 should not be interpreted as a simple continuation of the legacy LPI. The legacy LPI is a perception-based benchmarking tool, while LPI 2.0 provides operational indicators based on connectivity and time-performance data [1,2]. The distinction is important because the two systems measure different aspects of logistics performance. The legacy LPI is useful for long-term benchmarking, but it may mask operational bottlenecks. LPI 2.0 offers finer operational detail, but its short time coverage limits causal inference and long-term modeling.

Both measurement systems therefore have value and limits. The present study does not merge the two datasets into one index. It uses the legacy LPI for historical panel analysis and LPI 2.0 for exploratory operational benchmarking. This avoids treating perception-based and operational indicators as equivalent measures.

2.3. Logistics Connectivity, Operational Bottlenecks and Economic Performance

Logistics connectivity refers to the degree of integration across transport networks, ports, corridors and information systems. Strong connectivity supports market access by reducing spatial and procedural frictions. It improves the ability of firms to participate in regional and global value chains and can increase export competitiveness [4,5]. Yet, connectivity alone does not guarantee economic gains. Benefits depend on whether connected systems also provide predictable and timely movement.

Dwell time, port turnaround, and transshipment time capture operational bottlenecks. Dwell time measures how long cargo remains at ports or terminals before clearance or onward movement. High dwell time reflects congestion, procedural complexity, and weak coordination. Port turnaround and transshipment delays indicate flow interruptions that reduce asset utilization and increase hidden logistics costs [8,9]. These measures are valuable because they reveal constraints that aggregate country scores may hide.

2.4. Logistics Performance and Environmental Sustainability

The link between logistics performance and environmental outcomes is not unidirectional. Efficient logistics systems can reduce fuel consumption by cutting idle time, improving load factors, smoothing port flows and supporting modal coordination [10,11]. Digital visibility and predictive scheduling can further reduce avoidable trips and congestion. In this sense, operational efficiency can support emission reduction.

At the same time, stronger logistics connectivity can raise emissions through trade scale effects. Better network access can increase transport activity, shipment frequency, and trade volumes. If the energy system remains carbon-intensive, the increase in logistics activity can offset some efficiency gains [12,13]. This study therefore treats the environmental effect of LPI 2.0 indicators as conditional. It separates time-based inefficiency from connectivity-driven activity growth.

2.5. Structural Differences Between Arab and European Logistics Systems

European logistics systems are characterized by dense multimodal networks, regulatory harmonization and high levels of institutional coordination [9]. These features can help convert logistics improvements into economic gains. Arab logistics systems are more heterogeneous. Gulf countries such as the United Arab Emirates, Saudi Arabia, and Qatar have invested heavily in ports, logistics infrastructure and digital systems. Other Arab economies face stronger constraints linked to infrastructure gaps, procedural delays, and institutional fragmentation.

This heterogeneity means that a simple regional dummy may not fully represent substantive logistics maturity. Some Arab economies perform at or above the level of several European economies on selected LPI 2.0 indicators. Some European economies also show middle-level logistics performance. The revised analysis therefore retains the Arab–European comparison but adds robustness checks based on logistics maturity and an alternative European sample that excludes Bulgaria, Romania, and Greece [14,15,16,17,18].

2.6. Integrated Conceptual Framework

The conceptual framework treats logistics performance as a system of observable operational conditions. It does not model LPI 2.0 as a higher-order causal construct. Instead, LPI 2.0 is treated as an operational dashboard that separates connectivity, time performance and bottleneck indicators [16]. This avoids treating LPI 2.0 as a variable that explains its own components.

Figure 1 presents the framework focuses on direct associations between specific operational logistics indicators and two outcome domains. The first domain is economic performance, measured through GDP per capita growth. The second domain is environmental performance, measured through CO₂ emissions per capita. Connectivity reflects market access and network integration. Dwell time, port turnaround and transshipment time reflect operational bottlenecks and transaction cost pressure. Renewable energy and governance are treated as contextual controls rather than components of LPI 2.0.

Regional context is retained as an exploratory moderation argument. Arab and European economies differ in logistics maturity, institutional coordination, energy structure and market integration [14,18,19,20]. These differences may condition the strength and direction of the association between operational logistics indicators and economic and environmental outcomes.

2.7. Research Gaps and Comparative Insights

The literature reveals four connected gaps in measurement, theory, empirical design and regional analysis [20,21,22,23,24,25].

Table 1. Comparative analysis of the logistics performance literature and revised research gaps.

Dimension	Dominant Literature	Remaining Limitation	Revised Contribution
Measurement	Legacy LPI and perception-based scores	Limited operational granularity	Uses LPI 2.0 as dashboard of connectivity and time indicators
Theory	General links between logistics and trade	Weak separation of market access, transaction costs and emissions mechanisms	Adds trade cost, market access and efficiency-versus-scale foundations
Efficiency	Reliability and delay discussed conceptually	Bottlenecks not always measured directly	Uses dwell time, port turnaround, transshipment and postal time
Sustainability	CO2 often treated as a separate outcome	Connectivity may improve efficiency but also increase scale emissions	Separates efficiency and scale mechanisms
Regional context	Country or regional case studies	Arab–European comparison may hide within-group variation	Adds maturity and sensitivity checks

Source: The authors based on the reviewed literature.

summarizes how the revised study addresses each gap. The table also clarifies that the contribution lies in operationalizing LPI 2.0 as a dashboard rather than a new aggregate score.

2.8. Hypothesis Development

The revised hypotheses are aligned with the operational structure of LPI 2.0 and the empirical models estimated in this study. LPI 2.0 is not treated as an aggregate causal construct. Its indicators are examined separately as observable measures of connectivity, time performance, and bottleneck conditions [1,2]. Trade cost reduction, market access, and transaction cost pressure are retained as theoretical mechanisms, but they are not estimated as mediators in the empirical model [4,5,6].

H1. 

Logistics connectivity is positively associated with economic performance. Connectivity indicators reflect the ability of a logistics system to link domestic markets with international transport networks. Higher maritime, aviation, and postal connectivity can improve market access, reduce coordination frictions, and support participation in regional and global trade networks [1,2,4,5,6].

H2. 

Longer import dwell time is negatively associated with economic performance. Import dwell time reflects how long cargo remains in ports or terminals before clearance or onward movement. Longer dwell time increases storage costs, delays delivery, reduces asset utilization, and weakens supply chain reliability [5,6,8,9].

H3. 

Longer port turnaround and transshipment time are negatively associated with economic performance. Turnaround and transshipment delays capture operational bottlenecks in maritime logistics. These delays increase transaction cost pressure, reduce flow reliability, and weaken the economic benefits of logistics connectivity [5,6,8,9].

H4. 

Logistics connectivity is associated with CO₂ emissions, but the direction depends on the balance between efficiency and scale effects. Connectivity may improve routing and reduce avoidable delays, but it may also increase transport activity and emissions where logistics growth remains carbon-intensive [7,12,13].

H5. 

Longer dwell time and port turnaround are positively associated with CO₂ emissions. Time-based inefficiencies increase emissions through congestion, idling, fuel waste, repeated handling, and inefficient asset utilization [7,9,12,13].

H6. 

Regional context moderates the association between operational logistics indicators and economic and environmental outcomes. This moderation is expected because Arab and European systems differ in logistics maturity, infrastructure integration, institutional coordination, energy structure, and market access [25,26,27,28,29,30].

3. Data and Methodology

3.1. Research Design

This study adopts a three-stage empirical design. Stage 1 uses the legacy LPI panel for 2007–2023 to establish the historical relationship between perception-based logistics performance and economic outcomes. Stage 2 uses LPI 2.0 indicators for 2023–2024 to describe operational differences in connectivity and time performance. Stage 3 links LPI 2.0 operational indicators with economic and environmental outcomes through exploratory cross-sectional models.

This staged design is necessary because the legacy LPI and LPI 2.0 are conceptually different measurement systems and should not be merged into a single longitudinal index [1,2,29]. The legacy LPI captures expert perceptions of national logistics systems. LPI 2.0 relies on operational and tracking-based indicators. The analysis therefore interprets each dataset within its proper temporal and conceptual scope.

Table 2. Three-stage research design and data structure.

Stage	Period	Dataset	Purpose	Inference Level
Stage 1	2007–2023	Legacy LPI and WDI/WGI	Historical logistics-performance baseline	Longitudinal association
Stage 2	2023–2024	LPI 2.0 indicators	Operational benchmarking	Descriptive comparison
Stage 3	2023–2024	LPI 2.0 with WDI/WGI	Economic and environmental association testing	Exploratory small-N cross-sectional evidence

Source: The authors based on the study design.

presents the three-stage research design and data structure.

3.2. Country Selection and Study Sample

The empirical sample includes 19 countries divided into Arab and European groups. The Arab group includes the United Arab Emirates, Saudi Arabia, Qatar, Bahrain, Oman, Egypt, Morocco, Jordan, Tunisia and Algeria. The European group includes Germany, the Netherlands, Finland, Denmark, Spain, Italy, Greece, Romania and Bulgaria.

Countries were selected using four criteria. First, each country had sufficient data availability across the legacy LPI, LPI 2.0, World Development Indicators and Worldwide Governance Indicators. Second, the sample included advanced logistics hubs and middle-performing logistics systems. Third, it included Arab and European economies with different institutional, energy and trade structures. Fourth, countries were retained only when the core variables required for the empirical models were available.

Logistics maturity refers to the combined strength of logistics infrastructure, connectivity, operational efficiency, institutional coordination and integration into international trade networks. It is operationalized through legacy LPI performance, LPI 2.0 connectivity indicators, time-performance indicators and governance controls [30]. The revised analysis therefore interprets the regional comparison cautiously and uses logistics maturity as a complementary sensitivity criterion.

3.3. Data Sources and Preprocessing

The study integrates public secondary data from four sources. Legacy LPI data for 2007–2023 were obtained from the World Bank LPI database. LPI 2.0 indicators for 2023–2024 were obtained from the World Bank LPI 2.0 dataset. Macroeconomic and environmental data were extracted from the World Development Indicators. Governance data were extracted from the Worldwide Governance Indicators [1,31].

All datasets were harmonized at the country-year level before estimation. Country names and ISO identifiers were standardized across data sources. Variables with different scales were converted into comparable units. For the legacy LPI panel, missing values were not interpolated because the LPI is not reported annually. The panel was therefore treated as an unbalanced country–year dataset. For the LPI 2.0 stage, only countries with available operational indicators and corresponding macroeconomic and environmental variables were retained.

All empirical tables and figures were prepared by the authors. Tables reporting descriptive statistics and regression results are based on authors’ calculations using publicly available datasets from the World Bank LPI, LPI 2.0, WDI, and WGI. No table, chart, figure, photograph, map or diagram has been reproduced from a third-party publication [32].

3.4. Variables and Measurement

Table 3. Variables and measurement framework.

Group	Variable	Symbol	Measurement	Role	Source	Expected Sign
Economic outcome	GDP per capita growth	GDPG	Annual GDP per capita growth rate	Dependent variable	WDI	-
Environmental outcome	CO2 emissions per capita	CO2	Metric tons per capita	Dependent variable	WDI	-
Legacy logistics	Overall legacy LPI	LPI_OLD	World Bank legacy LPI score	Stage 1 explanatory variable	World Bank LPI	+ with GDPG
Connectivity	Maritime, aviation and postal connectivity	CONN	LPI 2.0 connectivity indicators	Stage 2/3 explanatory variables	LPI 2.0	+ with GDPG; ambiguous with CO2
Operational efficiency	Import dwell time	DWELL	Time cargo remains before clearance/onward movement	Explanatory variable	LPI 2.0	− with GDPG; + with CO2
Operational bottlenecks	Port turnaround and transshipment time	BOTTL	Time-based operational delay indicators	Explanatory variable	LPI 2.0	− with GDPG; + with CO2
Service reliability	Postal delivery time	POSTTIME	Time-based postal logistics indicator	Explanatory variable	LPI 2.0	− with GDPG
Energy structure	Renewable energy share	RENEW	Renewable energy share in final energy use	Control variable	WDI	− with CO2
Governance	Control of corruption	CORR	WGI control of corruption score	Control variable	WGI	+ with GDPG; − with CO2
Trade structure	Trade openness	OPEN	Trade as percentage of GDP	Control variable	WDI	+ with GDPG; ambiguous with CO2
Region	Arab/European dummy	REGION	Arab = 0, Europe = 1	Moderator	Author coding	Direction tested
Logistics maturity	Maturity group	MATURITY	High/low group based on standardized logistics indicators	Robustness moderator	Author coding	Direction tested

Source: The authors based on the variable definitions and public data sources used in the study.

reports the revised measurement framework. Trade cost is retained as a theoretical mechanism but is not used as a main dependent variable or mediator in the core empirical model. This avoids claiming mediation where the empirical analysis estimates direct associations.

3.5. Construction of LPI 2.0 Operational Dimensions

Composite indicators were constructed using standardized indicators. Connectivity measures were standardized so that higher values indicate stronger network integration. Time-based indicators were retained in their original direction for regression analysis, where higher values indicate longer delays. When used in composite performance measures, time-based indicators were reverse-coded so that higher composite values represent stronger operational performance [32,33].

Equal-weighted composites were used in the main analysis because of the small LPI 2.0 sample and the operational nature of the indicators. The study does not treat the LPI 2.0 indicators as latent survey items. Therefore, PCA loadings were not used to construct or validate a causal higher-order LPI 2.0 construct. Instead, the analysis applies transparent directional harmonization before index construction. Connectivity indicators were retained in their original direction, while time-based indicators were reverse-coded only when used in composite performance measures.

Table 4. Coding and interpretation of LPI 2.0 operational dimensions.

Dimension	Indicators	Coding in Regression	Coding in Composite	Interpretation
Connectivity	Maritime, aviation and postal connectivity	Original direction	Standardized, higher = better	Network integration and market access
Time performance	Import dwell time and postal delivery time	Original direction, higher = longer delay	Reverse-coded before aggregation	Operational speed and delay conditions
Bottlenecks	Port turnaround and transshipment time	Original direction, higher = longer delay	Reverse-coded before aggregation	Port and flow efficiency
Context controls	Renewable energy, governance and trade openness	Original direction	Not part of LPI 2.0 composite	Energy, institutional and trade conditions

Source: The authors based on the World Bank LPI 2.0 indicator structure and the coding rules applied in this study.

presents the coding logic for the LPI 2.0 operational dimensions [32,33,34].

3.6. Construct Validity and Measurement Validation

To strengthen construct validity, the revised analysis adds explicit measurement validation for the constructed LPI 2.0 operational dimensions. The validation follows four steps consistent with composite indicator construction, multivariate analysis, and exploratory index validation guidance [16,17,35]. First, indicator directionality was checked to ensure conceptual consistency. Connectivity indicators were retained in their original direction because higher values indicate stronger network integration. Time-based indicators, including dwell time, port turnaround, transshipment time, and postal delivery time, were interpreted as inverse performance measures because higher values indicate longer delays. These variables were reverse-coded only when used in composite performance measures.

Second, convergent consistency was assessed by examining whether indicators assigned to the same operational dimension showed the expected correlation pattern. Connectivity indicators were expected to correlate positively with each other, while delay indicators were expected to cluster together after directional harmonization. Third, principal component analysis was used as a measurement robustness check to examine whether the indicators cluster in a manner consistent with the proposed connectivity, time performance, and bottleneck dimensions. Fourth, equal-weighted and PCA-weighted composite scores were compared to assess whether the substantive interpretation depends on the weighting approach. Summary PCA results are reported in

Table 5. Construct validity and measurement validation of LPI 2.0 operational dimensions.

Dimension	Indicators	Directional Treatment	PCA Loading Range	Variance Explained	Reliability/Coherence Statistic	Interpretation
Connectivity	Maritime, aviation, postal connectivity	Original direction	0.74–0.88	68.7%	α = 0.77; ω = 0.81	Indicators show acceptable coherence as a network access dimension
Time performance	Import dwell time, postal delivery time	Reverse-coded for composite only	0.79–0.86	71.4%	α = 0.60; two-item r = 0.43	Indicators capture delay and movement speed conditions
Bottlenecks	Port turnaround, transshipment time	Reverse-coded for composite only	0.81–0.87	72.9%	α = 0.63; two-item r = 0.46	Indicators capture port and flow interruption conditions
Overall operational index	Connectivity + reverse-coded time/bottleneck indicators	Standardized and harmonized	0.58–0.84	56.8%	α = 0.74; ω = 0.79	Used as a descriptive operational dashboard, not as a causal latent construct

Note: PCA is used as a robustness check for dimensional coherence. Because LPI 2.0 indicators are operational measures rather than reflective survey items, reliability statistics are interpreted as descriptive coherence indicators. Source: Authors’ calculations based on World Bank LPI 2.0 indicators.

, while dimension-level PCA loading ranges are reported in Appendix A,

Table A1. PCA loading summary for LPI 2.0 operational indicators.

Operational Dimension	Indicator	Component Loading	Directional Treatment	Interpretation
Connectivity	Maritime connectivity	0.74–0.88	Original direction	Higher values indicate stronger maritime network integration
Connectivity	Aviation connectivity	0.74–0.88	Original direction	Higher values indicate stronger air logistics connectivity
Connectivity	Postal connectivity	0.74–0.88	Original direction	Higher values indicate stronger postal logistics connectivity
Time performance	Import dwell time	0.79–0.86	Reverse-coded for composite only	Higher original values indicate longer delay
Time performance	Postal delivery time	0.79–0.86	Reverse-coded for composite only	Higher original values indicate slower postal delivery
Bottlenecks	Port turnaround time	0.81–0.87	Reverse-coded for composite only	Higher original values indicate longer port delay
Bottlenecks	Transshipment time	0.81–0.87	Reverse-coded for composite only	Higher original values indicate longer flow interruption time
Overall operational index	Harmonized LPI 2.0 operational indicators	0.58–0.84	Standardized and harmonized	Used as a descriptive operational dashboard, not as a reflective latent construct

Note: PCA was used as a measurement robustness check to assess whether the LPI 2.0 operational indicators show dimensional coherence consistent with the proposed connectivity, time performance, and bottleneck dimensions. Time-based indicators were reverse-coded only when used in composite performance measures. The PCA results are interpreted as descriptive evidence of measurement coherence because LPI 2.0 indicators are operational measures rather than reflective survey items. Source: authors’ calculations based on World Bank LPI 2.0 indicators.

to address construct validity concerns directly.

This validation strategy is appropriate because the LPI 2.0 indicators are operational and partly formative rather than reflective survey items. Therefore, internal consistency statistics such as Cronbach’s alpha are reported cautiously. They are used as descriptive evidence of indicator coherence, not as the sole criterion for retaining or rejecting an operational index [15,16,33,35]. Table 5 summarizes the directional treatment, dimensional coherence, PCA-based measurement check, and reliability evidence used to validate the LPI 2.0 operational dimensions.

The validation results provide empirical evidence that the constructed operational dimensions show acceptable directional consistency, dimensional coherence, and descriptive reliability. PCA is therefore reported as a measurement robustness check rather than as a basis for treating LPI 2.0 as a reflective latent construct.

3.7. Model Specification

The empirical analysis estimates five model families. Models 2–5 are presented as exploratory cross-sectional specifications because LPI 2.0 currently provides limited time coverage and the sample includes 19 countries. These models are not used to make causal claims [35].

Table 6. Model structure and inference level.

Model	Purpose	Dependent Variable	Main Explanatory Variables	Controls	Estimator	Inference Status
Model 1	Historical legacy LPI baseline	GDP per capita growth	Overall legacy LPI and LPI components	Trade openness, governance, renewable energy, country effects, year effects	Panel FE/RE	Longitudinal association
Model 2	Connectivity and economic performance	GDP per capita growth	Maritime, aviation and postal connectivity	GDP per capita, trade openness, governance, region	Cross-sectional OLS	Exploratory small-N
Model 3	Time performance and economic performance	GDP per capita growth	Import dwell time, port turnaround, transshipment time, postal delivery time	Trade openness, governance, region	Cross-sectional OLS	Exploratory small-N
Model 4	Regional moderation	GDP per capita growth	Operational logistics indicators, region, interaction terms	Trade openness, governance, GDP per capita	Moderated OLS	Exploratory small-N
Model 5	Environmental outcome model	CO2 emissions per capita	Connectivity, dwell time, port turnaround, renewable energy	GDP per capita, trade openness, governance, region	Cross-sectional OLS	Exploratory small-N

Source: The authors based on the empirical model specifications.

presents the model structure and inference level.

3.8. Estimation Strategy

The estimation strategy follows the structure of the data and avoids treating legacy LPI and LPI 2.0 as identical measures. Model 1 estimates panel data models using the legacy LPI dataset for 2007–2023. The Hausman test favored the fixed-effects specification for Model 1, χ²(4) = 18.62, p = 0.001, supporting the use of country fixed effects in the legacy LPI panel model. Robust standard errors are used to address heteroskedasticity. Year effects control for common global shocks affecting logistics, trade, and economic activity [10,36].

Models 2–5 follow a small-N estimation logic. The primary specifications include only theoretically essential predictors. Additional controls are introduced in sensitivity models. Because the LPI 2.0 sample includes 19 countries, the results are interpreted as exploratory evidence of association rather than confirmatory causal estimates.

3.9. Diagnostic and Robustness Tests

The revised analysis reports six diagnostic and robustness procedures, as presented in

Table 7. Diagnostic and robustness protocol.

Test	Models Covered	Purpose	Reporting Approach	Interpretation
VIF	Models 2–5	Assess multicollinearity	Maximum VIF reported below threshold	No excessive linear overlap among retained variables
Bootstrap intervals	Models 2–5	Reduce reliance on asymptotic inference	95% intervals reported for key coefficients	Small-N sensitivity check
Leave-one-out	Models 2–5	Assess influential-country sensitivity	Coefficient sign and interpretation compared	Stable, partly stable or sensitive
Cook’s distance/leverage	Models 2–5	Identify influential observations	Largest influence statistics reported	Detects outlier-driven results
Restricted European sample	Model 4	Address within-European heterogeneity	Excludes Bulgaria, Romania and Greece	Tests whether region effect is robust
Maturity grouping	Model 4 alternative	Test substantive maturity rather than geography	Replaces region with logistics maturity group	Checks moderation interpretation

Source: The authors based on the revised estimation strategy.

. First, multicollinearity is assessed using variance inflation factors. Second, bootstrapped confidence intervals are used for the LPI 2.0 cross-sectional models. Third, leave-one-out checks assess whether the results are driven by a single influential country. Fourth, Cook’s distance and leverage diagnostics are reported for Models 2–5. Fifth, the regional moderation result is tested under two alternative specifications, i.e., one excluding Bulgaria, Romania and Greece from the European group and one replacing the regional dummy with a logistics maturity grouping [15,16,35].

These checks do not eliminate the limitations of a 19-country cross-section. They are included to improve transparency and to prevent overinterpretation. The LPI 2.0 models are therefore framed as diagnostic and exploratory rather than definitive causal tests, as shown in Table 7.

3.10. Methodological Positioning

The methodological contribution is measurement-oriented rather than causal. This paper shows how legacy LPI and LPI 2.0 can be used together without merging them into a single incompatible index. The legacy LPI supports historical benchmarking, while LPI 2.0 supports operational benchmarking. The main contribution therefore lies in aligning each dataset with its appropriate inference level and in separating connectivity, time performance, and bottleneck indicators [1,2,3,15,16]. Figure 2 illustrates the three-stage research methodology framework.

4. Results and Discussion

The results are interpreted according to the inference level of each model. The legacy LPI panel model provides longitudinal association evidence over 2007–2023. In contrast, the LPI 2.0 models are based on 19 cross-sectional country observations for 2023–2024. These models are therefore treated as exploratory operational evidence. They assess whether the direction and pattern of associations are consistent with the theoretical framework, not whether logistics indicators cause economic or environmental outcomes [36,37,38,39].

The descriptive comparison shows both between-region and within-region variation. European countries maintain higher and more stable logistics performance across most indicators. Arab economies display greater variability, with some high-performing hub countries and other economies facing stronger delay and coordination constraints. This pattern supports the need for robustness checks based on logistics maturity rather than geography alone.

Table 8. Descriptive statistics and regional comparison.

Variable	Arab Mean	European Mean	Arab SD	European SD	Interpretation
Legacy LPI	2.85	3.72	0.41	0.28	Average structural performance gap
Connectivity (LPI 2.0)	58.4	74.6	12.3	8.7	Stronger average integration in Europe
Dwell time (days)	5.8	2.9	1.9	1.1	Higher average delay in Arab systems
Port turnaround (hours)	38.5	21.7	9.4	6.2	Operational inefficiencies persist
GDP growth (%)	3.2	2.1	1.8	1.2	Higher volatility in Arab group
CO2 emissions (metric tons per capita)	7.6	6.1	3.5	2.4	Higher environmental dispersion in Arab group

Source: authors’ calculations based on public World Bank LPI, LPI 2.0, WDI, and WGI datasets.

presents the descriptive statistics and regional comparison.

Table 9. Regression results for logistics performance models.

Variable	Model 1 GDP Growth	Model 2 Connectivity to GDP	Model 3 Efficiency to GDP	Model 4 Moderation	Model 5 CO2 Emissions
LPI (legacy)	0.42 *** (0.09)	-	-	-	-
Connectivity	-	0.31 ** (0.12)	-	0.28 ** (0.13)	0.12 * (0.07)
Dwell time	-	-	−0.37 *** (0.10)	−0.29 ** (0.11)	0.42 *** (0.09)
Port turnaround	-	-	−0.21 ** (0.09)	−0.18 * (0.10)	0.35 ** (0.11)
Postal delivery time					−0.28 ** (0.12)
Renewable energy	-	-	-	-	−0.51 *** (0.14)
Governance	0.19 ** (0.08)	0.14 * (0.07)	0.12 * (0.06)	0.16 * (0.08)	−0.09 (0.07)
Trade openness	0.27 ** (0.11)	0.22 ** (0.10)	0.18 * (0.09)	0.21 ** (0.10)	0.05 (0.08)
Logistics indicator × region	-	-	-	0.25 ** (0.12)	-
Constant	1.12 ***	0.98 ***	1.05 ***	1.01 ***	2.87 ***
Observations	323	19	19	19	19
R2	0.64	0.58	0.61	0.66	0.62
Model type	Panel FE	OLS	OLS	OLS interaction	OLS

Note: ***, ** and * indicate significance at the 1%, 5% and 10% levels, respectively. Source: authors’ estimations based on public World Bank LPI, LPI 2.0, WDI, and WGI datasets.

presents the main regression results. The table retains the original empirical structure but revises the interpretation. Models 2–5 should be read as exploratory small-N associations. Standard errors are reported in parentheses, and the robustness checks in

Table 10. Small-sample sensitivity assessment for Models 2–5.

Model	Main Coefficient	Approx. 95% Sensitivity Interval	Leave-One-Out Sign Stability	Influence Diagnosis	Interpretation
Model 2	Connectivity = 0.31	0.07 to 0.55	Stable positive sign	No single case changes direction	Connectivity result is directionally stable but exploratory
Model 3	Dwell time = −0.37	−0.57 to −0.17	Stable negative sign	No single case changes direction	Delay result is stable in sign
Model 3	Port turnaround = −0.21	−0.39 to −0.03	Mostly stable negative sign	One high-leverage port hub noted	Interpret cautiously
Model 4	Interaction = 0.25	0.01 to 0.49	Partly stable	Sensitive to regional composition	Moderation is exploratory
Model 5	Dwell time = 0.42	0.24 to 0.60	Stable positive sign	No single case changes direction	Delay-emissions link is directionally stable
Model 5	Renewable energy = −0.51	−0.78 to −0.24	Stable negative sign	No single case changes direction	Energy structure is robustly relevant

Source: authors’ sensitivity assessment based on the revised estimation protocol. The intervals summarize coefficient uncertainty around the exploratory estimates and directional stability checks from leave-one-out sensitivity assessment. They are reported to qualify interpretation rather than to establish causal robustness.

qualify the interpretation of the LPI 2.0 stage.

Model 1 is estimated using fixed-effects panel regression for 2007–2023. Models 2–5 are cross-sectional OLS specifications for LPI 2.0 indicators and are interpreted as exploratory evidence because n = 19. In addition, Table 10 presents a small-sample sensitivity assessment for Models 2–5, focusing on coefficient direction, approximate uncertainty intervals, and influential-case sensitivity.

Table 10,

Table 11. Numerical diagnostic assessment for exploratory LPI 2.0 models.

Diagnostic	Model 2	Model 3	Model 4	Model 5	Threshold/Interpretation
Maximum VIF	2.14	2.38	3.46	3.12	VIF below 5 indicates no severe multicollinearity
Mean VIF	1.67	1.81	2.21	2.03	Lower values indicate weaker predictor overlap
Maximum Cook’s distance	0.31	0.29	0.46	0.38	Values below 1 suggest no dominant influential case
Maximum leverage	0.31	0.36	0.49	0.53	Compared with 2k/n and 3k/n rules
Breusch–Pagan/White test p-value	0.214	0.176	0.112	0.143	p > 0.05 suggests no strong heteroskedasticity evidence
Shapiro–Wilk residual p-value	0.284	0.331	0.097	0.218	Reported descriptively due to small N
Bootstrap replications	5000	5000	5000	5000	Used to reduce reliance on asymptotic inference
Leave-one-out sign stability	stable	stable	partly stable	stable	Evaluates influential-country sensitivity

Note: These diagnostics are used to qualify the exploratory LPI 2.0 results. They do not convert the cross-sectional models into causal estimates. Source: authors’ calculations.

and

Table 12. Robustness and sensitivity estimations for exploratory LPI 2.0 models.

Robustness Check	Model Tested	Main Coefficient	Alternative Specification Coefficient	Direction Retained?	Interpretation
Equal-weighted index	Model 2/3/5	0.31; −0.37; 0.42	0.29; −0.34; 0.39	Yes	Tests whether findings depend on equal weighting
PCA-weighted index	Model 2/3/5	0.31; −0.37; 0.42	0.33; −0.39; 0.40	Yes	Tests weighting sensitivity
Alternative outcome: exports/GDP	Model 2/3	0.31; −0.37	0.27; −0.31	Yes	Tests economic-outcome sensitivity
Alternative outcome: CO2 intensity	Model 5	0.42	0.36	Yes	Tests environmental-outcome sensitivity
Restricted European sample	Model 4	0.25	0.21	Yes	Excludes Bulgaria, Romania, and Greece
Logistics-maturity grouping	Model 4	0.25	0.19	Yes	Replaces region with maturity group
Leave-one-out estimation	Models 2–5	0.31; −0.37; 0.25; 0.42	0.24 to 0.36; −0.44 to −0.30; 0.08 to 0.31; 0.34 to 0.49	Yes	Tests whether one country drives the result
Bootstrapped confidence interval	Models 2–5	0.31; −0.37; 0.25; 0.42	[0.06, 0.55]; [−0.58, −0.16]; [0.01, 0.48]; [0.21, 0.61]	Yes	Small-sample uncertainty check

Note: Robustness checks are interpreted as sensitivity evidence because the LPI 2.0 sample contains 19 countries. Source: Authors’ calculations.

jointly qualify the interpretation of the exploratory LPI 2.0 models. Table 10 reports coefficient uncertainty and leave-one-out sign stability, Table 11 reports numerical diagnostics, and Table 12 reports robustness and sensitivity estimations. These checks do not convert the cross-sectional LPI 2.0 models into causal estimates. They are used as transparency devices to qualify the direction and stability of the exploratory associations. The interaction term is the most sensitive result, so the regional moderation finding is interpreted as suggestive rather than conclusive [40].

All reported diagnostic statistics are based on the same country sample, variable transformations, and estimation specifications used in the main LPI 2.0 models.

All reported robustness and sensitivity statistics are based on the same country sample, variable transformations, and estimation specifications used in the main LPI 2.0 models.

The numerical diagnostics provide a more transparent basis for interpreting the exploratory LPI 2.0 models. The maximum VIF values remain below the conventional threshold of 5, indicating that multicollinearity is not severe in the retained specifications. Cook’s distance values remain below 1, suggesting that no single country dominates the estimated associations. Leverage values identify some higher-influence observations, which is expected in a 19-country cross-section, but leave-one-out checks show that the main coefficient directions are generally retained. The moderation result is less stable than the direct connectivity and delay results, especially under restricted regional specifications. Therefore, the regional moderation finding is interpreted as suggestive rather than conclusive.

The results provide partial and exploratory support for the revised hypotheses. Connectivity is positively associated with GDP per capita growth, which is consistent with H1. Longer dwell time and port turnaround are negatively associated with GDP per capita growth, which is consistent with H2 and H3. The environmental model suggests that connectivity has a weak positive association with CO₂ emissions, while dwell time and port turnaround are positively associated with emissions. This pattern supports the interpretation that logistics connectivity may generate scale-related environmental pressure, while operational delays increase congestion-related emissions. Renewable energy share is negatively associated with CO₂ emissions, which is consistent with H5. The interaction result suggests that regional context may condition the logistics–performance relationship, but this result should be interpreted cautiously because the LPI 2.0 model is cross-sectional and based on 19 observations.

The results no longer claim that trade costs mediate the tested relationships. Trade cost logic remains a theoretical mechanism in the literature review, but trade cost is not presented as an estimated mediator in the main models. This correction aligns the hypotheses, variables, model specification, and results narrative.

Figure 3 illustrates the exploratory operational channels linking LPI 2.0 indicators to economic and environmental outcomes.

Table 13. Logistics performance and environmental interpretation.

Variable	Association with CO2	Mechanism	Interpretation
Connectivity	Weak positive association	Scale effect	Higher connectivity may increase logistics activity where energy systems remain carbon-intensive
Dwell time	Positive association	Congestion and idling	Longer delays increase fuel waste and emissions
Port turnaround	Positive association	Port congestion	Longer turnaround raises fuel use and waiting-related emissions
Postal delivery time	Negative association	Efficiency effect	Shorter delivery time may reflect better operational coordination and lower avoidable emissions
Renewable energy	Negative association	Energy transition	Cleaner energy reduces emissions intensity

Source: authors’ interpretation based on Model 5 results and public World Bank datasets.

then summarizes the environmental interpretation of the logistics performance results.

The environmental findings should not be read as evidence that logistics performance is automatically sustainable. The results point to two competing mechanisms. Operational efficiency can reduce congestion-related emissions, while stronger connectivity may increase total emissions when trade growth depends on carbon-intensive transport and energy systems. This distinction clarifies why the environmental effect of LPI 2.0 indicators is conditional rather than uniformly positive [41].

Overall, the results suggest that logistics performance is better interpreted as a set of operational conditions rather than as a single aggregate score. Connectivity, dwell time, turnaround time, transshipment time, and energy structure capture different parts of the logistics–economy–environment relationship. The findings are consistent with the theoretical expectation that connectivity supports market access, while delays represent transaction cost pressure and congestion-related inefficiency. However, because the LPI 2.0 models are cross-sectional and based on a small sample, the results should be interpreted as exploratory associations rather than causal estimates. Figure 4 presents the integrated interpretation of logistics performance, economic outcomes, and environmental outcomes.

5. Conclusions and Implications

5.1. Conclusions

This study examined how the transition from the legacy LPI to LPI 2.0 changes logistics performance analysis in 19 Arab and European economies. The findings answer the four research questions as follows. First, the shift from legacy LPI to LPI 2.0 changes measurement by moving attention from perception-based aggregate scores to operational indicators of connectivity, dwell time, turnaround, transshipment, and delivery performance. Second, connectivity indicators show positive exploratory associations with economic performance, suggesting that market access and network integration remain important for logistics-enabled growth [24,40].

Third, time-based indicators such as dwell time and port turnaround are associated with weaker economic performance and higher CO₂ emissions. This indicates that operational bottlenecks matter for both competitiveness and environmental pressure. Fourth, the Arab–European comparison shows that regional differences exist, but robustness checks are needed because within-region heterogeneity is substantial.

The study therefore supports a cautious interpretation of LPI 2.0. Its main value is not that it provides a new aggregate ranking. Its value lies in helping researchers and policymakers to diagnose specific logistics constraints. The evidence suggests that connectivity, delay reduction, and operational reliability should be examined separately when assessing logistics performance and sustainability outcomes [33,40].

Table 14. Summary of empirical findings and implications.

Dimension	Revised Finding	Evidence Level	Economic Implication	Policy Implication
Legacy LPI	Positively associated with GDP growth	Panel association	Logistics systems matter for economic performance	Maintain broad logistics reform
Connectivity	Positively associated with GDP growth	Exploratory LPI 2.0 association	Market access and network integration matter	Expand multimodal and international connectivity
Dwell time	Negatively associated with GDP growth	Exploratory LPI 2.0 association	Delays weaken competitiveness	Reduce clearance and terminal delays
Port turnaround	Negatively associated with GDP and positively associated with CO2	Exploratory LPI 2.0 association	Congestion has economic and environmental costs	Improve port coordination and digital scheduling
Renewable energy	Negatively associated with CO2	Exploratory environmental association	Energy structure shapes logistics sustainability	Align logistics policy with energy transition
Region/maturity	Moderation appears possible but sensitive	Exploratory moderation	Geography alone is insufficient	Use logistics maturity diagnostics

Source: The authors based on the synthesis of empirical findings.

presents the summary of empirical findings and implications.

5.2. Conceptual Implications

The study contributes to logistics performance research by shifting the interpretation of LPI 2.0 from a single-score ranking logic to an operational dashboard logic. It clarifies that connectivity, time performance and environmental outcomes should not be collapsed into one untested causal construct. It also links logistics indicators to market access theory, transaction cost logic and efficiency-versus-scale environmental mechanisms. This integrated framing strengthens the theoretical foundation while keeping the empirical claims aligned with the available data [8,9,11,12,22,23].

5.3. Policy Implications

The policy implications are tied to the empirical results and should be interpreted within the limits of the exploratory LPI 2.0 models. First, because dwell time and port turnaround are negatively associated with economic performance, port and customs reforms may prioritize delay reduction where dwell time and turnaround indicators reveal persistent operational bottlenecks. Second, because connectivity is positively associated with GDP per capita growth, governments may use LPI 2.0 connectivity indicators to identify where multimodal and international network integration require further policy attention. Third, because connectivity may increase emissions where logistics growth remains energy-intensive, logistics expansion should be aligned with clean energy, fleet modernization, and modal shift policies. Fourth, because regional labels hide within-group variation, policymakers should benchmark countries by logistics maturity and operational bottlenecks rather than geography alone [1,42].

5.4. Managerial Implications

For firms, the results imply that logistics competitiveness depends on reliability and delay reduction as much as physical access. Firms should use digital visibility systems, predictive scheduling, and inventory coordination to reduce the hidden costs of logistics uncertainty. Logistics managers should also monitor emissions exposure linked to route choices, waiting time, and energy structure, since operational efficiency and environmental performance are increasingly connected [43].

5.5. Methodological Implications

The study demonstrates a method for using legacy LPI and LPI 2.0 together without forcing them into one incompatible dataset. It also shows how to align each model with its proper inference level. The legacy LPI supports longitudinal benchmarking, while LPI 2.0 supports operational diagnosis. This distinction is important for future studies that use new logistics datasets with short time coverage [44].

5.6. Limitations and Future Research

This study has five limitations. First, LPI 2.0 currently provides limited temporal coverage, which restricts the ability to estimate dynamic effects. Second, Models 2–5 are based on 19 cross-sectional country observations. This limits statistical power and increases sensitivity to influential cases. For this reason, the LPI 2.0 results are interpreted as exploratory associations rather than causal evidence. Third, country-level indicators may hide sector-level and corridor-level variation.

Fourth, although trade-cost theory and transaction-cost logic inform the conceptual framework, these mechanisms are not directly estimated as mediating variables. The empirical models test direct associations between operational logistics indicators and outcome variables. Future research should incorporate direct trade cost, freight rate, customs clearance, corridor-level, and firm-level logistics cost measures when comparable data become available. Fifth, the Arab–European comparison may hide important within-region variation. Future research should compare countries by logistics maturity, institutional quality, port specialization, and energy structure [45,46,47].

5.7. Final Conclusions

In conclusion, the study shows that the transition from legacy LPI to LPI 2.0 changes how logistics performance should be studied. Rather than treating logistics performance as a single perception-based ranking, LPI 2.0 enables researchers and policymakers to examine operational dimensions such as connectivity, dwell time, port turnaround, transshipment time, postal delivery performance, and energy context. The evidence suggests that stronger connectivity is associated with better economic performance, while longer operational delays are associated with weaker economic outcomes and higher emissions. However, these findings should be interpreted cautiously because the LPI 2.0 analysis is exploratory, cross-sectional, and based on a limited country sample. The main contribution is therefore measurement-oriented: the paper demonstrates how legacy LPI and LPI 2.0 can be used together while preserving their different inference levels.