IMPACTS OF LINER SHIPPING CONNECTIVITY AND GLOBAL COMPETITIVENESS ON LOGISTICS PERFORMANCE: THE MEDIATING ROLE OF THE QUALITY OF PORT AND INFRASTRUCTURE

■ LP is crucial both for national and international economy; ■ carried-out research showed the impact of LSC and global competitiveness on LP in Asia; ■ Asian policy-makers need a guide to focus on how to improve LP; ■ the study will serve as the theory of strategies highlighting the dimension of the factors to be concentrated in this regard. Article


Introduction
This study is relevant to delineate the effects of LSC and global competitiveness on LP in mediating of port and infrastructure quality.Modern logistics being strongly influenced by globalization and internationalization (Beysenbaev, Dus 2020), efficient logistics services are considered as the comparative advantages and vital catalysts of the global logistics hub (Erkan 2014;Önsel Ekic et al. 2016;Chen, Hasan 2020).It is very crucial both to national and international economic development (Karaman et al. 2020;Önsel Ekici et al. 2019;Sy et al. 2020;Wong, Tang 2018).The opportunity of bilateral trade depends on access to the high quality of logistics network (Önsel Ekic et al. 2016).LP in Asia still varies, particularly in the quality of trade and transport-related infrastructure.The performance score of Asia over the last few years measured by the WB was satisfactory but the score in the South Asian region was the second-lowest, 2.51 (out of 7) in 2018, compared with others, while 3.24 for Central Asia (Ahmed, Khan 2020).Since Asian transport-logistics network is essential for regional connectivity, an innovative and smart logistics approach must adopt to improve the combined LP (Su et al. 2011).Compared to the average score (2.40) of low-income countries, South Asia did slightly well in 2018, though it was considerably lower than the average scores of lower middle (2.60) and upper middle income countries (2.70) (Ahmed, Khan 2020).However, all previous publications focused on the LP of each country or on an economic group.Few of them have tried to combine maritime connectivity with global competitiveness.This created a gap in the development of regional LP for decision-makers.This gap and its importance lead us to conduct this empirical study to show a roadmap to improve the performance of logistics in Asia.
The qualities of logistics services as well as the port and infrastructure influence the transportation of goods between countries (Sy et al. 2020).The time that the world passes through is for mutual benefit and the facilitation of global trade between countries, creating opportunities to win, which is equally important for combined economies (Acar et al. 2015).The PORT as well as flexible regulations between countries has created opportunities but challenges for advanced and intelligent logistics networks.The performance of logistics in every economy now revolves around the public and private sector services such as policies, intervention and the implementation of good governance.However, characteristics such as the development of transport infrastructure, political dimension, intra-connectivity, public-private partnerships, global competitiveness, port and infrastructure quality serve as a podium for the intelligent evolution paths of the LP (Pettit, Beresford 2009).Here again, the efficient logistics capacities of individual companies constitute a competitive advantage over their competitors in the market (Sy et al. 2020).
Now the question for global logistics hub is how to increase LP, especially for the Asian economy, the ones that triggered the core of the stimulation of regional logistics development (Arvis et al. 2018;Bookbinder, Tan 2003) and a major transhipment market in the world (Yap, Zahraei 2018).The WB's LP index published since 2007 has raised concerns about the importance of LP among the countries.The index is performed worldwide on the basis of the survey of operators in the field.The index evaluates the performance of 6 basic indicators: ECCP, QTTINFRT, CPS, CQLS, FS, and CT.These elements of LP could be improved by combining them with indicators of global competitiveness, port quality, and LSC.Because the competitiveness of a nation is defined as the sets of institutions, market development, education and training, infrastructure, labour market efficiency, and business sophistication (Önsel Ekici et al. 2016), while LSC depicts how well countries are connected to global shipping networks in terms of number of ships, container-carrying capacity, vessel size, services and the number of containers deploying port companies.In addition, LP is a complex sequence (Önsel Ekici et al. 2016) and coordinated activities of port facilities as well as shipping connectivity as a whole.Again, it is generally accepted that the quality of port performance today plays a big role not only in cargo handling, but also maintains the aspects of LP (Jouili 2019).It has the potentiality to shape and sharpen the social and environmental performance of transportation systems as well (Gonzalez-Aregall et al. 2018;Hua et al. 2020).As shipping connectivity reduces trade costs (Hoffmann et al. 2020), the structure of LSC, port infrastructure endowments, and LSBC will accelerate freight rate among the countries (Wilmsmeier, Hoffmann 2008).LSBC is an extension of the country-level LSC index and is based on an appropriate bilateral transformation of transhipments, common connections and direct connections between 2 countries.LSC indicates the level of integration into the global shipping network.Therefore, integrated shipping connectivity, the quality of port facilities and global competitiveness works together as catalysts of improving LP.
The PORT contribute greatly to the development of LP.Port activities are the provision of logistics services for national and international businesses.Port logistics facilities take great advantages over the countries those who don't have these facilities at all.These lead us to select 29 economies those have the port facilities in Asia.Many countries are now planning to build up a regional port hub to facilitate and hasten SCM activities (Munim, Schramm 2018).Meanwhile, port-city relation has changed the urban and coastal structure.That's why, Wan et al. (2018) underlined the important role of the port in international shipping, which inevitably influences the global trade, as better port facilities and infrastructure are considered to be crucial nodes and catalyst of better logistics network (Song, Van Geenhuizen 2014;Wan et al. 2018).Lun & Hoffmann (2016) pointed out that the influential importance of intra-trade on the development of LSC may have an impact on the performance of the logistics network as well.
Poor transport connectivity and inefficient LP continue to reduce the access of smaller and weaker economies to global markets, especially for developing countries, leading them to underperformance (UNCTAD 2017).
Empirical studies on Asian have intensified the importance of LP.Most of the works have focused on the multidimensional role of LP on trade and economy, while few have studied the interactions between LP and GCI.The studies conducted previously are related to the direct and indirect or mediating effects of LP on a particular economy.Works of Bookbinder et al. (2003);Önsel Ekici et al. (2019); Kabak et al. (2020); Khan et al. (2019); Liu et al. (2018b); Sharapiyeva et al. (2019); Clark et al. (2004) can be taken as the testimony.This created a research deficit and very few of them tried to blend shipping connectivity, PORT with global competitiveness to see the impacts on LP in Asia.Our study attempted to minimize this gap undertaking 28 Asian economies and 1 special administrative region (Hong Kong).The reason for selecting 29 economies is the lack of port facilities and activities in other economies.The study will sort out the combined impacts of shipping connectivity and global competitiveness on LP.It will pay particular attention to the implication of global competitiveness indicators and LSC for improving the LP in Asia.Furthermore, it will intensify the necessity of port facilities among the nations.

Literature review and hypothesis development
All studies done previously have focused on the impacts of the PORT, global competitiveness, and LSC at economic level.They showed how LSC and the PORT can improve economic growth at national level.But no research has yet empirically attempted to see the impacts on LP in mediating the PORT.Hence, our study is the first and foremost effort to observe the influence of global competitiveness and LSC on LP.
Global competitiveness is a widely used term that deal with the ability to achieve certain outcomes, such as a high standard of transportation, technological readiness, productivity and economic growth of a country (Sargsyan 2017).Most of the countries continue to trade using shipping nodes in the global shipping networks to reduce bilateral trade cost (Wilmsmeier, Hoffmann 2008).The competitiveness indicators have a degree of free and fair market of goods and services (Sargsyan 2017), those stimulate the structure of LSC, port infrastructure endowment (Wilmsmeier, Hoffmann 2008).Schøyen et al. (2018) conducted research on the contribution of logistics service and container liner connectivity on port efficiency.They used DEA to measure port efficiency in the context of liner connectivity.Wilmsmeier & Hoffmann (2008) investigated the interconnection between LSC and port infrastructure in the Caribbean, which has resulted in competition between shipping lines that makes shipping services less expensive and improves connectivity.Port characteristics have been studied by Wilmsmeier et al. (2006) and found the quality of ports and infrastructure has a significant and positive impact on LSBC.In other words, shipping connectivity, the PORT and global competitiveness are interconnected and impact improving LP.
Önsel Ekici et al. (2019) assessed how to improve LP by reforming the GCI.The study used PLS-SEM approach to test the significant impacts of the pillars of global competitiveness and suggested some policy implications related to LP. Global competitiveness indicators are the economic indicators of the country (Çemberci et al. 2015), and illustrate the major factors and institutions to determine the long-term growth and competitiveness of the countries (Önsel Ekici et al. 2019).The success of improving the LP of any country is highly dependent on competitiveness (Önsel Ekici et al. 2016).Because competitiveness has also a positive influence on the country's transportation, economy, and logistics services (Önsel Ekici et al. 2016;Chen, Hasan 2020).The implementation of flexible policies, public and private investments, the development of a regulatory regime for transport services, improvement of business environment and the expansion of national and international markets can bring great change to LP. Kabak et al. (2018) highlighted the importance of LP for both the national and international economy using global competitiveness indicators.They used BN to investigate the relationship between logistics network that mostly depends on the policies, services and investments, and global competitiveness.It was found that pillars of competitiveness directly and indirectly affect the success of LP and the competitiveness of a country.Thus, we hypothesized: ■ H1a: global competitiveness has a positive impact on LSC; ■ H1b: global competitiveness has a positive impact on LP; ■ H1c: global competitiveness has a positive impact on PORT; ■ H1d: global competitiveness has a positive impact on LP in mediation of LSC; ■ H1e: global competitiveness has a positive impact on LP in mediation of PORT; ■ H1f: global competitiveness has a positive impact on LP in the mediation of LSC and PORT.Lam et al. (2011) examined the liner shipping network and port connectivity of 4 major ports in East Asia.Their study identified shipping capacity, trade routes, and geographic regions are connected to the ports and shipping network, suggesting policies and research implications related to smart port planning and the development of supply chain systems.Jiang et al. (2015) developed 2 scientific models to compute port connectivity, which helped improving overall LP.The importance of LP has been identified using a novel scenario based analysis by Kabak et al. (2018).Maritime relations of bilateral LSC and LSC are crucial determinants of bilateral trade (Fugazza, Hoffmann 2017).The predominance of bilateral trade and maritime transhipment operations in hub ports has accelerated the intensification of smart LP.Lin et al. (2020) demonstrated a spatial analysis of LSC, which made it possible to identify the geographic ripple of transport infrastructure.However, the growth of trade and the need for competitiveness in the global market significantly force ports to facilitate SOLP (Pavlic et al. 2014;Rashidi, Cullinane 2019).Therefore, we hypothesized here: ■ H2a: LSC has a positive impact on LP; ■ H2b: LSC has a positive impact on PORT; ■ H2c: LSC has a positive impact on LP in the mediation of PORT.
Beysenbaev & Dus (2020) developed a proposal to modify LP by applying qualitative and quantitative assessment of logistics efficiency of 159 countries.Gordon et al. (2005) added that combination of port facilities like govt.support, more investments, flexible policies, and excellence in technological readiness can bring sustainable port competitiveness, which leads to outstanding performance.Yeo et al. (2008) established that PORT, connectivity and port accessibility significantly contribute to the development of LP.Using the gravity method, port's connectivity was studied to access the density of connectivity and the importance of accelerating policy and management of marine transportation of national logistics (Zaman et al. 2015).The efficiency of ports for bilateral trade, trade liberalization and infrastructure reduce artificial barriers (Clark et al. 2004;Dare et al. 2019) and logistics infrastructure has been playing an increasingly significant contribution to economic development (Qi et al. 2020), because global supply chain activities have increased pressure on the maritime haul, port operations and inland freight distribution (Notteboom, Rodrigue 2005).Port maritime connectivity is essential to accelerate and expand the strength of national connectivity that will rebuild the logistics network and quality infrastructure (Rumaji, Adiliya 2019).Sharapiyeva et al. (2019) adopting SEM studied the impact of port transport-logistics infrastructure of 37 landlocked countries and found that the countries without access to the sea are inferior to other countries in terms of port infrastructure, logistics efficiency, and economic growth.So, it can be hypothesized: ■ H3: PORT has a positive impact on LP.

Conceptual framework
We adopted the mediation path diagram (Figure 1) method using 3-stage integrated PLS-SEM with smart-PLS software.First, reliability and stability of latent factors are checked applying CFA.Next, we verified the direct effects of competitiveness, LSC and the PORT on our target variable.The mediating effects of the variables were checked in the second step, and an IPMA was added in the final step to clarify the importance of the factors that will determine the most influential factors for improving LP in Asia.

Data sources and methodology
This study focused on 28 Asian countries and 1 special administrative region (Hong Kong) (Appendix A).The reasons of selecting 29 economies are port facilities and the availability of data.We collected data on different variables from the WB database.Data on the GCI were extracted from annual publications by the WEF.
The LP consists of 6 indicators based on the customs clearance process, QTTINFRT, competitively priced shipments, quality of logistics services, ability to track and trace and FS.Erkan (2014); Chen & Hasan (2020); Munim & Schramm (2018); Liang & Liu (2020) and Çemberci et al. (2015) used the same datasets in their study.The overall performance score reflects perceptions of a country's LP ranging from very low (1) to high ( 5).The performance score is obtained through surveys conducted by the WB.
The GCI presents the report of competitiveness on the individual country every year.It assesses the capacity of countries to provide a high level of welfare, using the resources they have (Sharapiyeva et al. 2019).It presents the information on the competitiveness score combining 12 indicators of competitiveness such as macroeconomic stability, ICT adaptation, labour market efficiency, market openness, financial market development, infrastructure, technological readiness, education and training and so on.The score is measured on a Likert scale 1…7; 1 -for low and 7 -for the high.Çemberci et al. (2015); Hasan & Chen (2020) and Sharapiyeva et al. (2019) used the same set of variable in their study as well.
Transport connectivity is a crucial determinant of LP.To measure shipping connectivity, we adopted LSC index and LSBC index.LSC score shows how well countries are connected to global shipping networks.The index is calculated by UNCTAD on the basis of 5 components of maritime transport sectors: the number of ships, container transport capacity, maximum vessel size, services and the number of companies that deploy container ships in a country's ports.The maximum values are 100 for the country with the highest average index (UNCTAD 2020).On the other hand, LSBC index indicates a country pair's integration level into global liner shipping networks.However, the index is an extension of UNCTAD's country-level shipping connectivity and it is based on a proper bilateral transformation.LSBC is measured on the basis of 5 components-number of transhipments, direct connections common to both countries, common connections per pair of countries with one transhipment, level of competition in services and size of the largest ship on the weakest route connecting one country to another (UNCTAD 2020).A correlation matrix of latent variables and Pearson's heat map is presented in Appendix B and Appendix C, respectively.
The PORT measures business executives' perception of their country's port facilities.Data are measured on Likert scale 1…7, where 1 represents the lower PORT and 7 -for the best.As smart port and infrastructure are interrelated to the facilities of communication infrastructure and transportation (Liang, Liu 2020), the NIU (% of population), BBS (per 100 people), MOBS (per 100 people) were also taken into account in our study.
As the variables of the study such as LP, PORT and LSC are hard to observe directly, it is necessary to use multiple explicit observations to measure the latent constructs indirectly (Liang, Liu 2020;Munim, Schramm 2018).Most of the previous studies used a set of observed variables in the empirical analysis and this lead them to use SEM (Green et al. 2008;Liang, Liu 2020;Munim, Schramm 2018;Sharapiyeva et al. 2019).SEM is a powerful multivariate technique increasingly used in scientific experiments to test and evaluate multivariate causal relationships (El-Sheikh et al. 2017;Fan et al. 2016;Hox, Bechger 1998).However, we also used latent constructs along with their observed factors, which motivated us to adopt a complex PLS-SEM mediation path diagram analysis.SEM is different from other statistical techniques because it can measure the direct and indirect effect on putative causal relationship using a path diagram (Fan et al. 2016).Since it is difficult to measure the direct impacts of the latent constructs, this motivated us to use complex PLS-SEM composed of several observed factors as well.In addition, PLS-SEM allows estimation of very complex models with many constructs and indicator variables, especially when prediction is the goal of analysis (Sarstedt et al. 2017).It allows flexibility regarding the data requirements and the specification of relationships between the constructs and observed indicators, which result in an excellent approximation of common factor models, while factor-based SEM can't do it due to its methodological limitations (Sarstedt et al. 2017).A research framework is figured outlying the steps involved in the methodology (Figure 2).The PLS-SEM path diagrams will measure the direct and indirect effects of latent factors on LP.PLS-SEM works well with small sample size (Wu et al. 2012) and it not only estimates the total effects of every individual factors score, but also decomposes the total effect into direct and indirect effects (Önsel Ekici et al. 2019).SEM is often visualized by a graphical diagram (Hox, Bechger 1998).The reasonable sample size for a complex SEM should be greater than 100 (Kline 2023) or about 200 (Hox, Bechger 1998).The sample size used in this study was pooled from 6 years (i.e., 2007, 2010, 2012, 2014, 2016, 2018), generating 174 observations, except for observations with missing values.A list of the latent variables presented in Table 1 and the data were transformed into log form to standardize the measurement.After estimating the significant path coefficients for the inner and outer models with their relevant t-values, we will continue to perform IPMA in the final stage to find the most important factors affecting target variable.

Validity and reliability test
This section presents the result of the stability and reliability tests.Composite reliability, convergent validity, standardized residuals, discriminant validity, modification indices, and average variance are used to test the validity and reliability of datasets (Koufteros 1999).Descriptive statistics as well as Cronbach a score (Cronbach 1951) presented in Appendix D.
Cronbach a score is used to check the homogeneity and internal consistency of the latent variables.In general, an a scores greater than 0.6 is regarded as the effective reliability.Overall a score we obtained is 0.84, which is higher than the recommended level (0.6).It clarifies the data has reached a good standard.We have further checked normality by Q-Q plot presented in Appendix E.
In addition, KMO and Bartlett's test of sphericity were checked.KMO test is used to assess the sample adequacy and evaluates the correlations and partial correlations to validate whether the observations are likely to coalesce on components (Hasan, Chen 2020;Liang, Liu 2020).The measured KMO value is 0.91 and the c 2 value is 2012.19.Since KMO value is greater than 0.70, this satisfies the level of adequacy and validity of the observations.Bartlett's test of sphericity is significant at 5% level, which is required for validation and necessary to adapt FA.PLS-SEM assessment generally follows 2 steps, the 1st step is to check the reliability and validity of the measurements and an assessment evaluate the structural model estimates in the 2nd step (Hair et al. 2011).Therefore, the rest of the paper is designed according to the recommendation of Hair et al. (2011).

CFA
SEM is a combination of CFA and path analysis (Fan et al. 2016;Hox, Bechger 1998).It is necessary to perform CFA for all latent constructs involved in the model before running SEM (Awang 2012) to obtain both model parameter estimates and the predicted factor scores (Liu et al. 2018a).In order to determine how well the predictor variables explain the direct and indirect path impacts on the outcome variable in mediation analysis, a CFA was conducted.Standardized factor loadings, model fit indices and GFI are considered to be the key statistical tools for an acceptable measurement model (Koufteros 1999).
To verify the unidimensionality of the measurement items, FA is adopted.Table 2 summarizes the result of CFA.The values of the squared multiple correlations, R 2 for all indicators except LnQTTINFRT, LnLSBC and LnMOBS, are above the recommended level of 0.50 (Bollen 1989;Koufteros 1999;Lu et al. 2007;Sharapiyeva et al. 2019).All standardized factor loadings are greater than or close to the recommended level 0.70 and statistically significant as well, although the values for 3 indicators are not up to the mark.TLI, CFI, better known as NNFI and NFI (Ainur et al. 2017;Bentler, Bonett 1980), RMSEA and SRMR were considered for the best model fit of CFA as well (Hox, Bechger 1998;Munim, Schramm 2018;Sharapiyeva et al. 2019).Both NFI and NNFI adjust for complexity of the model (Hox, Bechger 1998) by Schermelleh-Engel et al. (2003).Both TLI and CFI are 1, NFI = 0.99, NNFI = 1; while RMSEA = 0.00 and SRMR = 0.01 indicate the good model fitness.RMSEA and SRMR should be less than 0.05 for the best fit (Hox, Bechger 1998).GFI and MFI are 0.99; while Bollen's RFI is 0.94, which means the estimated model predicates between 94…99% of the variance and covariance in the observed data.Finally, the result demonstrates a good fit according to the model fit indices of c = < 2 2.63 3 df within the required recommended level (Bollen, Long 1993).
Statistical significance of the factors loadings are examined to test for convergent validity through their z-values.Acceptable threshold levels of z-values should be greater than 2 or less than -2 as rules of thumb (Koufteros 1999).The z-values presented in Table 2 greater than 2 signifying all factors measure their respective construct and confirm the unidimensionality and convergent validity (Anderson, Gerbing 1988).

PLS-SEM path and mediation analysis
As the measurement model and the validity and reliability tests confirm the validity of the possible relationship in the system of the factors, we proceed with SEM, which follows a prioritize assessment -IPMA of the latent constructs.A consistent PLS algorithm was used to test the significance for the goodness-of-fit measures.Henseler et al. (2014) introduced the SRMR as a goodness of fit measure for PLS-SEM that can be used to avoid model misspecification.
The path coefficients between the latent constructs as well as their p-values (in parenthesis) and the loads of the outer model on individual latent factors are shown in Figure 3.The results show that indicators of LP, PORT, global competitiveness and LSC produce enough loads to be accepted.The loads of the 14 factors in the outer model of the latent construct (except LnQTTINFRT 0.65) are greater than 0.70 (Figure 3) and statistically significant as well.The t-values/z-values for all latent constructs are above the recommended level, t-values > 2 or t-values < 2. R 2 for LP is 0.70, which indicates that the estimated model can analyse 70% variance in the model.SRMR should be less than 0.08 for a good model fit and the NFI should be 0.90 or greater than 0.90 (Önsel Ekici et al. 2019).The SRMR for our estimated model is 0.06 and NFI = 0.90, those are the adequate testimony of a good model fit.
After confirming the fitness of the estimated model and outer model evaluation on individual latent constructs, we estimate the path coefficients of the hypothesized relationship as well as the significant level in the mediation analysis presented in Table 3.Before establishing mediation relationship among the indicators (hypotheses H1d, H1e, H1f, H2c), the direct path coefficients of the observed variables were measured.
According to the statistical significance of the path coefficients shown in Figure 3, the study determined the direct effects of hypotheses H1a, H1b, H1c, H2a, H2b and H3 to be accepted.Thus, LSC (LSC→LP, hypothesis H2a), global competitiveness (GCI→LP, hypothesis H1b) and PORT (PORT→LP, hypothesis H3) have positive and statistically significant impacts on LP.The effects of LSC and global competitiveness on the PORT are positive and meaningful as well.
a scores of all latent constructs are above the recommended level 0.60.However, ρ is a better measure of reliability than Cronbach's a in SEM because it is based on the loadings rather than the correlations between the observed variables (Demo et al. 2012).Accepted level of ρ must be greater than 0.70 (Ravand, Baghaei 2016).The values of our model are higher than the accepted level (Figure 5, Appendix F).Composite reliability is also high, while the AVE for all constructs are greater than the recommended level 0.50 (Chin 1998;Fornell, Larcker 1981) demonstrated in Appendix F. Collinearity statistics (VIF) are lower than the recommended level, 10 (García et al. 2015;Miles 2005;O'Brien 2007).Therefore, the validity and reliability results of the latent construct incurred from PLS-SEM path analysis provide sufficient evidence for the model to be reasonably valid.In addition, we verified the discriminate validity, which is a prerequisite for the evaluation of PLS-SEM model (Hamid et al. 2017;Henseler et al. 2015).HTMT ratio, FLC and cross-loadings are 3 universal scales for determining the discriminate validity.HTMT ratio of correlations is a new and superior approach compared to 2 other methods (Hamid et al. 2017), which can achieve higher specificity and sensitivity rates (97…99%) (Henseler et al. 2015).HTMT ratio less than 1 is required (Figure 6) for the model to be accepted, although a threshold of 0.85 (Henseler et al. 2015;Kline 2023) has been suggested as rules of thumb.HTMT ratio and FLC is presented in Appendix G, and none of the score exceeds the recommended level signifying model validation.

IPMA
The IPMA is an extension of the PLS-SEM path model that displays the performance of each latent construct.The aim of IPMA is to sort out the most crucial factors affecting the target factor (LP) (Önsel Ekici et al. 2019).In this regard, the IPMA uses the importance (total effects TE) and the performance P to prioritize the factors.
The factors are ranked based on the overall score.Since the major concern of IPMA is to improve the performance of the factors those have relatively high importance but relatively low performance, the factors with lower overall scores have more importance.The factors are ranked in ascending order with respect to their overall scores (Kabak et al. 2020;Önsel Ekici et al. 2019).Importance performance score is obtained by dividing P by TE presented in Table 4 and Figure 7.

Discussion and policy implication
A discussion on integrated results is presented in this section.We conducted this research in order to observe the impacts of LSC and global competitiveness on LP in mediation of the PORT in Asian.The development of smart LP is the demand of modern trade and economic activities, as growth in Asia largely dependent on the LP and advanced infrastructure level (Tang, Abosedra 2019).Asian countries have not yet been able to use their potential due to the lack of regional connectivity (Su et al. 2011).So, the study has the novelty to guide improving LP.Path analysis in mediation of PORT has shown that shipping connectivity and global competitiveness has statistical significance and positive effects of improving LP.Empirical evidence derived from the SEM path diagram reveals that the direct and indirect effects of competitiveness and port and infrastructure accelerate operational and strategic change in LP as well, leading to increase competitiveness and yielding better connectivity.The study demonstrated that 70% variations of LP can be explained by latent factors.Therefore, PORT, global competitiveness indicators such as business sophistication, financial market development, goods market expansion, labour market efficiency, technological advancement, economic satiability and macroeconomic development have become the central theme for both developed and developing countries in Asia.Capabilities of competitively priced shipments, ability to track and trace consignments, quality transport and timeliness have changed the mode of logistics network and connectivity.Port infrastructures, performance of logistics and transportation facilities have increased bilateral trade among the nations.The loadings of all observations in the outer model have high significant values (Figure 3), which are the indications of the importance of latent factors to explain LP.Direct impacts of path coefficients of GCI→LSC (0.488 *** ), GCI→LP (0.386 *** ), GCI→PORT (0.52 *** ), LSC→LP (0.27 *** ), LSC→PORT (0.31 *** ) and PORT→LP (0.30 *** ) within the inner model are highly significant (note: *** p < 0.001).Again, path coefficients of the mediation effects of GCI→PORT→LP (0.16 *** ), GCI→LSC→PORT→LP (0.048 * ), and LSC→PORT→LP (0.098 * ) are important to be taken as dimension of the factors to be concentrated to improve LP as well (notes: * p < 0.05, *** p < 0.001).
Our findings are also similar to the study of Acar et al. (2015);Önsel Ekici et al. (2016Önsel Ekici et al. ( , 2019) ) and Çemberci et al. (2015).It is clear from the analysis that LSC and global competitiveness in mediation of port infrastructure serve as the most influential factors.Results proved that shipping connectivity can connect countries, business, people and markets, enabling them to buy and sell goods on a scale; and today, it transports goods approximately onethird of the total value of global trade (WSC 2020).LSC is the most efficient mode of transportation for bilateral trade, and the most carbon-efficient mode of transportation that a single large container ship can handle over 200000 containers of cargo in a year, while an individual ship container transports as many as 10000 containers of goods in a single voyage (WSC 2020).This can minimize transportation costs, which are essential for efficient performance.This means that LP is greatly influenced by the LSC and Asian countries can focus on bilateral and integrated shipping connectivity.Again, the factors of GCI are crucial for LP in Asia, as the factors like technology, infrastructure, and labour market efficiency reduce shipment time and risk enabling consignments tracing and tracking flexibility (Chen, Hasan 2020).The flexible labour market in Asia promotes productivity and productivity leads to the development of trade-and transport-related connectivity.Extension of PLS-SEM; IPMA reveals important findings regarding latent factors.It confirms that more attention and importance should be given on GCI factors.The same result was obtained by Önsel Ekici et al. (2019) and Chen & Hasan (2020) highlighting technological readiness, market size, higher education and training, infrastructure and innovation to facilitate the improvement of LP.The IPMA specifies that global competitiveness factors are at the top of importance while LSC is second and the PORT stands at the end of the map.IPMA signifies that global competitiveness pillars like technological adoption assess the agility, infrastructure and connectivity brings productivity by connecting economic agents, easing distance and time, facilitating the flow of information, reducing costs, and expanding markets in global value chains.In addition, the qualities of port and infrastructure were found to be crucial in boosting LP, as port facilities constitute vital economic activity in the hinterland of coastal areas (Chen, Hasan 2020).Therefore, policy-makers must focus on the PORT and LSC to fasten and improve logistics capability.
But policy-makers in Asia need to think about vicious circle of high transport costs and low services, which discourage trade and inter-connectivity.However, the mediating role of port and infrastructure leads to smart LP, as connectivity guarantees more choice, lower costs, higher speed and frequencies and direct and indirect income for port users.In other words, Asian policy-makers can utilize the corroborated benefits of shipping connectivity and global competitiveness in mediation of PORT.Countries like Bangladesh, Sri Lanka, India, Vietnam and Pakistan need to improve governance of port authorities, strengthen private sector participation and create a more competitive environment in port sector.Obviously, the policy-makers must focus on customs procedure, infrastructure building, shipping and port services, tariffs and terminal handling charges, minimize the costs associated with delays, loss of markets, customer choices as well.Individual country should strengthen the port community system, connect landlocked countries, enhance of regional connectivity and improve the performance of container ports.Efficient use of port facilities, as well as improved scale of operations is expected to increase in South Asia.Since LSC and bilateral shipping connectivity focus on the number of ships, maximum vessel size, container-carrying capacity, number of services, and number of companies that deploy container ships in a country's ports, the policy-makers and stakeholders have to deal with a wider range of policies, processes and procedures.Their concerns in the region should be spatial planning, training skills and resources, environmental and socio-economic sustainability of supply chain.
The contributions of the study are twofold.It will theoretically expand the frontier of knowledge in logistics research in Asian region and guide policy-makers together with the government to adopt pragmatic and effective policies-cum-measures to improve LP.For practical application, the findings of the study are vital in integrating national, social and economic policies for Asia.However, this will be challenging task to implement and focus on the facilitation of technological readiness, customs procedure, expansion of market size, infrastructure building, provide higher education and training and ensure the quality of port and services at a time.IPMA will lead them to take a quicker decision where to put more efforts and ensure sustainable and resilience LP in Asia.

Concluding remarks
In the context of rapid transformation of global trade, the importance of smart LP is increasing day by day.The LP of Asian countries is getting more attention in the regional logistics-hub.This paper delineated how LP in Asian can be improved using global competitiveness indicators and LSC in mediation of PORT.The empirical evidence revealed that LP of Asia in mediation of PORT is greatly influenced by global competitiveness indicators and LSC.The analysis showed that global competitiveness indicators (as GCI) are at the top of the importance-performance map.In other words, LP can be increased if the countries pay close attention to the 12 global indicators of WEF.However, shipping connectivity among the countries as well as the PORT needs to be focused on.
The study will increase the knowledge regarding regional LP both theoretically and empirically in a system by integrating and relating validity test.Secondly, the methodology, PLS-SEM and IPMA, applied to analyse the effects of global competitiveness indicators and LSC on LP in mediation of PORT will illuminate new roads for Asian policy-makers giving suggestions to be implemented of LP development.
The integrated results of this study are not applicable to each individual country; rather it is applicable only for the combined countries having port facilities.As a further suggestion, it is possible to see individual country LP by clustering them and recommend policies to be implemented.Therefore, our future study will investigate individual country performance and rank them in accordance to their score.

Figure 1 .
Figure 1.Framework of PLS-SEM mediation path diagram analysis

Figure 3 .Figure 4 .
Figure 3. Result of PLS-SEM path analysis diagram with latent construct

Table 1 .
List of observed variables

Table 2 .
Summary of CFA

Table 3 .
Result of PLS-SEM path analysis

Table 4 .
Result of importance performance analysis