Spatial Network Structure of Global Logistics Service Trade

Tao Luo

School of Economics and Management, Guangdong Construction Polytechnie, 638 Guanghua Second Road, Guangzhou,

China

Keywords: Spatial Network Structure, Social Network Analysis, QAP, Distance.

Abstract: Research purpose: Logistics service trade is an important part of international service trade, but the spatial

network structure of logistics service trade and its influencing factors is not clear. Research methods: This

paper uses the WIOD world input-output table, constructs a world logistics service trade matrix, establishes

relational data, and uses social network analysis to characterize. On this basis, it uses QAP regression methods

to study the factors affecting trade in logistics services. Study found: The logistics service trade network has

a small-world network effect. The overall cyberspace structure of global logistics service trade is getting

closer and closer. Developed countries are at the core. China's status in developing countries is gradually

rising. The network density of the global logistics service trade network shows an upward trend, but the

change of the network is a gradual process. The control of each country in the global logistics service trade

network is declining. Economic distance, geographical distance, proximity and trade distance have significant

effects on global logistics service trade.

1 INTRODUCTION

This paper attempts to analyse the spatial network

structure of global in logistics services trade and its

influencing factors. The development of the logistics

industry has significantly reduced the cost of

international trade, increased the efficiency of global

trade, reduced the impact of geographical distance on

global trade, and connected the various links of the

global value chain. It can effectively facilitate the full

integration of goods, technology, knowledge and

services across borders. Developed countries not only

occupy the middle and high end of global R&D and

design value chains, but also dominate in logistics

services. The World Bank report "Connecting to

Compete: Trade Logistics in the Global Economy

2018" shows that developed economies continue to

lead in global trade logistics and that most countries

in the world are actively developing innovative

logistics services. Logistics services are not only an

important component of global trade, but also an

important part of international trade and a highly

profitable link in the global value chain. Therefore,

logistics services trade plays an important role in

international trade in services and is a further focus of

international trade in services.

Compared to other productive services, logistics

services trade is more widespread, extensive and

intimate. The links are no longer mere geo-

adjacencies, but rather a multi-threaded and complex

network structure. The spatial network structure of

logistics services trade provides a better measure of

the position and evolution of logistics services in

global trade across countries. Spatial network

structure characteristics can reflect the comparative

advantage of countries in logistics services trade.

Most studies have focused on producer services trade.

Few studies have been conducted on the sub-sectors

of logistics services trade. There are no systematic

answers to the characteristics of the vertically linked

network structure of logistics services trade and the

factors influencing it. Logistics services trade is very

different from other productive service industries.

Using typical studies of other productive service

trades is not conducive to an accurate understanding

of the changing characteristics of logistics services

trade and its influencing factors. The article uses the

world input-output tables provided by the WIOD

database to construct a logistics service trade matrix

and uses social network analysis to describe the

characteristics of the spatial network structure of

logistics service trade. To solve the problem that

traditional statistical methods cannot quantitatively

Luo, T.

Spatial Network Structure of Global Logistics Service Trade.

DOI: 10.5220/0011187700003440

In Proceedings of the International Conference on Big Data Economy and Digital Management (BDEDM 2022), pages 483-489

ISBN: 978-989-758-593-7

483

study "relational" data, the QAP regression analysis

method is used to study the influencing factors of the

vertical network structure of logistics service trade.

2 LITERATURE REVIEW

Logistics services exist in the production, marketing,

consumption and recycling of global value chains

(Bai, 2010). Logistics services are an important

component of the global value chain, connecting all

aspects of the value chain. Logistics service providers

are required in the global value chain to provide

services beyond "basic" production (Bair, 2005).

Microscopically, leading companies in the global

value chain can outsource non-core services such as

logistics services to focus on their core competencies,

optimize their internal value chains and industry

chains, and thereby enhance their competitiveness

(Memedovic, Ojala, Rodrigue, et al, 2008).

Macroscopically, fierce competition has reduced

corporate profit margins. Logistics services are the

result of an effective division of labor in the value

chain, which contributes to the rationalization of

industrial division of labor and industrial structure,

and improves the production efficiency of complex

production networks in global value chains, which in

turn can enhance the overall economy. Innovation

and competitiveness (Rodrigue, 2010). Research in

the apparel industry shows that the development of

logistics and supply chain management helps to the

development of the apparel industry in the global

value chain (Cammett, 2013). The above research has

shown that logistics services are the result of the

international division of labor and an important

subsystem of the global value chain, connecting all

aspects of the global value chain. Logistics services

can improve the core competitiveness of enterprises

in the global value chain on a micro level. They can

enhance the innovation and competitiveness of the

economy on a macro level, promote the progress of

global value chains, and form a complex global value

chain production network.

How to measure the logistics service network

with the appropriate method is an issue we must

solve. Previous studies have concentrated on

micro-enterprise logistics operations networks or

product distribution networks (MD, Haijema,

Bloemhof, et al, 2015). Less discussion of logistics

service trade networks from a global perspective.

Moreover, existing research mainly adopts

geography methods, ESDA spatial analysis

techniques or regression analysis methods. These

methods form the basis for quantitative research on

this issue. However, traditional statistical methods

mainly deal with "attribute" data, and cannot process

data with obvious "relationship" characteristics.

Therefore, existing research lacks quantitative

research on the spatial network structure and

evolution of logistics service trade, especially the

lack of quantification of the space network from the

network "relationships".

Social network analysis is just an effective

method to study "relationship" data. It has been

implemented in logistics network research and

provides new methods and ideas for researching

logistics networks. The first use of social network

analysis methods to study logistics networks, and

believe that the various nodes of logistics trade and

transportation can be continuously optimized in the

network (Phillips, Phillips, 1998). Social network

analysis has changed the way of relying on surveys in

the field of logistics and supply chain management.

In particular, it can study the model of binary

relationship in logistics and supply chain, and catch

up with the lack of attention to traditional research

methods relationship" (Carter, Ellram, Tate, 2010).

Therefore, based on the relationship data and

network perspective, using the logistics service

export data of 43 major economies (including China

Taiwan) provided by WIOD in 2000-2014, the

logistics service trade matrix is constructed, and its

spatial network structure is analyzed by means of

social network analysis. And the influencing factors

were studied. This study reflects the overall

characteristics and evolution of the spatial network of

logistics service trade by measuring network density

and the overall network structure. Through central

analysis, the status and role of each economy in the

spatial network of logistics service trade are

examined. Finally, QAP regression analysis is

utilized to study study the impact of economic

distance, geographical distance, proximity and trade

distance on the spatial network of logistics service

trade.

3 RESEARCH METHODS AND

DATA

3.1 Model Building

The logistics input service network model is

constructed by using the world input-output table

provided by the WIOD database. According to the

gravity model, the two basic factors affecting

international trade are economic size and

BDEDM 2022 - The International Conference on Big Data Economy and Digital Management

484

geographical distance. The efficient scale is

positively related to international trade and has a

negative correlation with geographical distance.

Therefore, the factors affecting logistics service trade

choose GDP, geographical distance, and proximity as

variables. Considering the demand for logistics

services, influenced by the scale of global trade, the

value added of trade is added as a variable to the

model. The model is built as following:

Logtra=f (GDP, Dist, Contig, VAX) (1)

In equation (1), Logtra represents the logistics

service trade matrix. GDP represents economic

distance. Dist represents geographical distance.

Contig represents proximity. VAX represents trade

distance.

3.2 Variable Selection and Meaning

The equations are an exception to the prescribed

specifications of this template. You will need to

determine whether or not your equation should be

typed using either the Times New Roman or the

Symbol font (please no other font). To create

multileveled equations, it may be necessary to treat

the equation as a graphic and insert it into the text

after your paper is styled.

Logtra means logistics service trade. The WIOD

database provides a world input-output table built by

the world's major economies, using public data from

the logistics industry code H49-H53 to merge, and 43

countries (or regions) as nodes (remove ROW data).

Logistics service trade matrix. Because the

determinants of the international competitiveness of

the production services sector is mainly intermediate

needs and technical levels (Carter, Ellram, Tate,

2010). The logistics service trade matrix constructed

is a power matrix. According to the social network

analysis, the UCINET6.212 software is used to

convert the appropriate matrix into a weightless

binary matrix.

GDP means Economic distance. Economic

distance is characterized by GDP differences.

Select the GDP data of each country provided by the

IMF to construct a GDP difference matrix. With each

country as a node, the GDP difference matrix is

constructed by subtracting the GDP of other countries

of the GDP of one country. That is, the GDP of

country i is subtracted from the GDP of country j,

which is 0 on the diagonal. The calculation method is

presented in Equation 2.

GDP

=GDP

-GDP

(2)

GDP means Geographic distance. A geographic

distance matrix is constructed by selecting

geographic distance data between capitals of various

countries in the CEPII database. The geographical

distance is the farther, the higher the cost and risk of

providing logistics services between the two

countries, and the less likely it is to trade logistics

services.

Contig means Proximity relationship.

Considering the neighbor relationship on land

between countries, the neighbor relationship matrix is

constructed. "0" indicates that there is no adjacent

relationship between the two countries on the land,

and"1” indicates that the two countries have adjacent

relations on land. If the two countries have adjacent

relations on land, it is more convenient to trade

logistics services, especially between landlocked

countries and neighboring coastal countries

(Evangelista, 2000).

VAX means Trade distance. The value-added

export matrix produced in the final stage of the

WIOD database builds a trade distance matrix. The

value-added exports produced in the final stage

reflect the mutual production and value added of

trade between the two countries, which better reflects

the trade gap between the two countries, especially

the trade level between the two countries. With each

country as a node, the VAX of one country is

subtracted from the VAX of other countries to

construct a trade distance matrix. That is, the VAX of

the country i is subtracted from the VAX of the

country j, and the diagonal is 0. The calculation

method is shown in Equation 3.

VAX

=VAX

-VAX

(3)

According to the international trade

gravity model,

the relationship between logistics service trade and

explanatory variables is expected as follows:

Expectation1: The economic distance is in line

with the symbol of the international logistics service

trade matrix.

Expectation2: The geographical distance and the

symbol of the international logistics service trade

matrix are negative.

Expectation3: The symbol of the proximity

relationship and the international logistics service

trade matrix is positive.

Expectation4: The trade distance and the symbol

of the international logistics service trade matrix are

positive.

Spatial Network Structure of Global Logistics Service Trade

485

3.3 Data Sources

Logistics service trade data comes from the WIOD

database. National GDP data comes from the World

Monetary Fund (where Taiwan data comes from

Taiwan's monthly statistical report). Geographic

distance and proximity data are derived from the

CEPII database. The added value export data for the

final stage production is derived from the WIOD

database. The data sources and descriptions are

shown in Table I.

Table 1: Data source description.

Va ri a bl e

symbol

Variable meaning Data source

Logtra International

Logistics Services

Trade

http://www.wiod.org

Dist Geographical

distance

http://www.cepii.fr

Contig Proximity http://www.cepii.fr

GDP Economic distance http://data.un.org/IMF

VAX Trading Distance http://www.wiod.org

4 INTERNATIONAL LOGISTICS

SERVICE TRADE NETWORK

ANALYSIS

4.1 Small World Characteristic

Logistics service trade efficiency has increased.

Cooperation has become closer and closer, and

geographical distance has been narrowed. The length

of the feature path reflects the efficiency of

information transfer for each node in the network.

The larger the value, the lower the efficiency;

otherwise, the higher the efficiency. Table 2 shows

that the length of the characteristic pathway

decreased from 1.317 in 2000 to 1.122 in 2014. The

characteristic path length is much smaller than the

network size of 43 and is also significantly lower than

the 6 steps mentioned in six-degree separation

inference (Milgram's,1967). Explain that the logistics

service trade network has obvious characteristics of

small world networks. Moreover, the length of the

feature route is declining from 2000 to 2014,

indicating that the logistics service trade has a

phenomenon of bridging and bridging. The

characteristics of the group show that the cooperation

between countries in the world is getting closer and

closer, and mutual trust is getting higher and higher.

Bridging shows that the geographical distances of

countries around the world are getting closer, and

countries are more likely to get logistics services

from other countries.

Table 2: Average Shortest Path and Aggregation Coefficient Table.

Variable 2000 2005 2010 2014

Average Shortest Path 1.317 1.171 1.127 1.122

Aggregation Coefficient 0.825 0.877 0.900 0.903

Logistics service trade is highly aggregated. The

clustering coefficient indicates the case where the

points of the network are connected to each other. The

value of the set class coefficient is [0, 1]. When the

set class coefficient is 1, it means that there is a

connection between any two points in the network.

When the set class coefficient is 0, it means that the

network does not. There are cases where three points

are completely connected to each other. Table 2

shows that the clustering coefficient of logistics

service trade increased from 0.825 to 0.903 in 2014

from 2000 to 2014, and the clustering coefficient is

getting closer to 1. It shows that the logistics service

trade network is highly concentrated, and the number

of sides is also very large. This means that in the

world, goods can be reached in a short path from one

country to another. It proves that logistics service

trade has obvious Small World Characteristic

characteristics.

4.2 Overall Spatial Network Structure

Netdraw, a visualization tool of UCINET6.212

software, is used to map the logistics service trade

network. Due to space limitations, only the logistics

service trade network structure maps for 2000 and

2014 are given. According to the network structure

chart, the following conclusions can be brought: (1)

BDEDM 2022 - The International Conference on Big Data Economy and Digital Management

486

The logistics service trade space network has obvious

structure and strict space, and all countries can carry

out logistics service trade with other countries. There

is a link between countries, indicating the existence

of logistics services trade between the two countries.

The size of the network node is precisely proportional

to the centrality of a country's trade network. The

larger the node, the higher the centrality of the

country's logistics service trade network. (2) The

relationship between countries is closer. Observing

the graph, it can be seen that 2014 was more dense

than the 2000 network. Explain that the relationship

between countries in the network is closer with time.

(3) Developed countries are still at the core of the

network. Developed countries such as the United

States, the Netherlands, France, Germany, Denmark,

the United Kingdom, and Spain are mostly at the core

of the network. (4) China gradually approaches the

center from the edge in the network. In 2014, China's

position in the network was closer to the network

center than countries such as Japan, South Korea and

Canada, indicating that China's position in the

network is gradually rising.

Figure 1: Structure of the international logistics service

trade network in 2000.

Figure 2: Structure of the international logistics service

trade network in 2014.

4.3 Overall Spatial Network Structure

The logistics links between countries are getting

closer and closer and showing a steady growth trend.

Social network analysis uses density to measure the

ratio of the total number of relationships actually

present in a network to the total number of possible

theoretical relationships. Network density can be

used to assess the tightness of connections between

nodes in a network. According to Table III, it can be

observed that the density of logistics service trade

network increased from 0.6827 in 2000 to 0.8776 in

2014, and there was no obvious sudden drop or

upward trend. It shows that the network density of

logistics service trade network shows an overall

upward trend, reflecting the increasingly precise

logistics trade between countries. It also demonstrates

that the logistics service trade between countries has

shown a steady growth state, and the relationship is

getting closer and closer.

Table 3: Global Logistics Service Trade Network Density

Table 2000-2014.

Year 2000 2005 2010 2014

Density 0.6827 0.8295 0.8726 0.8776

5 QAP REGRESSION RESULTS

AND ANALYSIS

Using the constructed model, QAP regression

analysis was carried out on the spatial network matrix

of logistics service trade and the matrix of various

influencing factors from 2000 to 2014, and the

number of random replacements was selected 10,000

time. The observation value 1806 means that the

composition between 43 countries or regions is a

matrix of 43 × 43, and the observation value is 1,806

except for the diagonal. Table

IV reports the

regression results. The adjusted R2 in 15 years is

between 0.488 and 0.556, and the significance

probability value is 0.000, indicating that the

difference in economic distance, geographical

distance, proximity and trade distance can account for

this logistics service trade. Spatial network changes

from 48.8% to 55.6%.

First of all, the economic distance has a

significant effect on the logistics service trade

network, and the symbol is positive. The logistics

services trade network has passed at least a 10%

significant test in all years except 2003. It is shown

that the size of the economy can significantly affect

the relevance of logistics service trade. The smaller

Spatial Network Structure of Global Logistics Service Trade

487

the economic distance, the higher the possibility of

logistics services trade, the higher the correlation

between logistics service trade between the two

countries. That is to say. The logistics service trade

between countries is more likely to occur between

countries with similar economic development levels.

This conclusion is in line with the gravity model. The

closer the economic distance is, the more frequent the

logistics service trade between countries and the

stronger the network relationship.

Secondly, geographical distance has a significant

effect on the logistics service trade network, and the

symbol is negative. The logistics service trade

network did not pass the significance test before

2006, but passed the significance test of at least 10%

in the years after 2006. The significance of recent

years has increased substantially, and all symbols are

negative. It shows that geographical distance has a

significant impact on logistics services trade between

countries, i.e. the greater the geographical distance

between countries, the less likely it is that logistics

services trade will take place. This phenomenon is

consistent with the relationship between the median

distance of gravity model and trade. This is explained

by the fact that the farther away the two countries are,

the higher the cost of logistics service trade will be,

and the probability of risk increases. It is more

inclined to choose countries with closer distances as

trading partners of logistics services. The less the

geographical distance, less international logistics

services trade, and the weaker the network

relationship.

Thirdly, the proximity relationship has a

significant impact on the logistics service trade, and

the symbol is positive. Except for 2000, the proximity

relationship between the two countries has passed at

least 10% significance test, and the symbols are

positive. Explain that proximity can promote the

trade of logistics services between countries. The

main reason is that logistics service trade between

neighboring countries can make up for the lack of

logistics capacity between countries to meet the needs

of import and export. It is more reasonable to choose

logistics services from neighboring countries, which

can effectively reduce the cost and risk of logistics

service trade between countries. More international

logistics services trade between countries with

neighboring relationships, the stronger the network

relationship.

Finally, trade distance significantly affects the

logistics service trade network, and the symbol is

positive. The trade distance between countries has a

1% significant test on the impact of logistics services

trade, and the symbols are all conclusive. It shows

that the greater the trade distance between countries,

the more likely the countries are to trade logistics

services. This conclusion is consistent with

international trade theory. Because logistics services

are the cause of demand, when goods trade between

the two countries generates demand for logistics

services, the more frequent the trade of goods

between countries, the larger the scale, the greater the

demand for logistics services. Moreover, it is more

likely to be consistent with the flow of goods to trade,

that is to say, when a country exports goods, it may

also bring export of logistics services. The trade

distance is farther, the less logistics services trade

between countries and the weaker the network

relationship.

Table 4: Global Logistics Service Trade Weighted Network

Qap Regression Results.

Year 2000 2014

GDP 0.051** 0.083**

Dist -0.002 -0.046*

Contig 0.006 0.066 ***

VAX 0.742 *** 0.675 ***

0.556 0.488

Adj R

0.555 0.488

Probability 0.000 0.000

Number of observation 1806 1806

Number of permutations 10000 10000

6 CONCLUSIONS

As a downstream link of the global value chain,

logistics services are increasingly valued by various

countries. Developed countries not only occupy the

upstream links of global value chains, but also occupy

the dominant position of logistics services in global

trade. Utilizing the world input-output table provided

by WIOD database, constructing logistics service

trade matrix, using UCINET to study the spatial

network structure and evolution of logistics service

trade, and using QAP analysis to study the

influencing factors of logistics service trade, the

following conclusions are obtained:

First, the logistics service trade network has a

small world network effect. The length of the

characteristic route of logistics service trade has been

decreasing year by year, and the logistics service

trade between countries is increasing. The clustering

coefficient is getting bigger and bigger, and the

logistics service relationship between countries is

getting closer and closer.

Second, the overall cyberspace structure of

logistics service trade is getting closer and closer, the

BDEDM 2022 - The International Conference on Big Data Economy and Digital Management

488

developed countries are right on the core, and the

status of China in developing countries is gradually

rising. In 2014, the logistics service trade network

was more dense than the 2000 network, indicating

that the relationship between countries in the logistics

service trade network is closer. Developed countries

remain at the core of logistics service trade. China is

moved from the edge to the center in the logistics

service trade network, and its position in the logistics

service trade network has gradually increased.

Third, the network density of the logistics service

trade network is on the rise, but the change of the

network is a gradual process. Since 2000, the network

density of the logistics service trade network has

generally shown an upward trend, and the trade links

between countries in the network have become more

and more close.

Fourth, economic distance, geographical distance,

proximity and trade distance have a significant effect

on logistics service trade. The economic distance has

a significant effect on the logistics service trade

network. The closer the economic distance is, the

more noticeable the interaction of the logistics service

trade and the stronger the network relationship.

Geographical distance has a negative effect on the

logistics service trade network. The farther the

geographical distance is, the weaker the worldwide

logistics service trade network relationship.

Neighboring relations have a significant influence on

logistics service trade. The more frequent worldwide

logistics service trade between countries with

neighboring relationships, the stronger the network

relationship. The trade distance substantially affects

the logistics service trade network. The closer the

trade distance is, the more frequent the logistics

service trade and the stronger the network

relationship. In developing the economy and

increasing exports, developing countries have

gradually narrowed the gap with developed countries.

ACKNOWLEDGEMENT

This research was funded by Guangdong

Construction Polytechnie Subject (grant number

KY2021-26).

REFERENCES

Bai X J. Research for logistics outsourcing based on value

chain promotion. IEEE, 2010.

Bair, Jennifer. Global Capitalism and Commodity Chains:

Looking Back, Going Forward[J]. Competition &

Change,2005,9(2):153-180.

Cammett M. Development and the Changing Dynamics of

Global Production: Global Value Chains and Local

Clusters in Apparel Manufacturing[J]. Competition &

Change,2013,10(1):23-48.

Carter C R, Ellram L M, Tate W.THE USE OF SOCIAL

NETWORK ANALYSIS IN LOGISTICS

RESEARCH[J]. Journal of Business Logistics, 2007,

28(1).

Evangelista, Rinaldo. Sectoral Patterns of Technological

Change in Services[J]. Economics of Innovation &

New Technology, 2000, 9(3): 183-222.

MD Keizer, Haijema R, Bloemhof J M, et al. Hybrid

optimization and simulation to design a logistics

network for distributing perishable products[J].

Computers & Industrial Engineering, 2015, 88(OCT.):

26-38.

Memedovic O, Ojala L, Rodrigue J P, et al. Fuelling the

global value chains: what role for logistics capabilities?

[J]. International Journal of Technological Learning

Innovation &Development,2008,1(3):353-374.

Milgram S. The Small World Problem[J]. Psychology

today, 1967, 2(1).

Phillips D M, Phillips J K. A social network analysis of

business logistics and transportation[J]. International

Journal of Physical Distribution & Logistics

Management, 1998,28(5):328-348.

Rodrigue J P. Transportation and the Geographical and

Functional Integration of Global Production

Networks[J]. Growth & Change, 2010,37(4):510-525.

Spatial Network Structure of Global Logistics Service Trade

489