water

Article
The Impact of Trade Facilitation on Cross-Border E-Commerce
Transactions: Analysis Based on the Marine and Land
Cross-Border Logistical Practices between China and
Countries along the “Belt and Road”
Yingying Liang 1 , Liangliang Guo 1 , Jianlu Li 2, *, Shuang Zhang 1 and Xiangyun Fei 1

1 School of Economics and Management, Liaoning University of Technology, No. 169, Shi Ying Street,
Guta District, Jinzhou 121000, China; [email protected] (Y.L.); [email protected] (L.G.);
[email protected] (S.Z.); [email protected] (X.F.)
2 College of Business Administration Maharishi International University (MIU), 1000 N. 4th Street,
Fairfield, IA 52557, USA
* Correspondence: [email protected]; Tel.: +86-18004067107

Abstract: In the post-epidemic era, cross-border e-commerce has become a new growth point for
global foreign trade. Unlike traditional trade, which is dominated by marine transport, cross-border
e-commerce transactions have high requirements for both marine and land transport, and the scale
of their trade is accordingly limited by the level of trade facilitation i.e., the convenience of cross-



border logistics in bilateral trading countries. Based on transaction cost theory, this article takes


cross-border e-commerce transactions between China and countries along “The Belt and Road” as
Citation: Liang, Y.; Guo, L.; Li, J.;
the core of the study. From the perspective of marine and land transport timeliness, the theoretical
Zhang, S.; Fei, X. The Impact of Trade
framework is constructed using the marine and land logistics infrastructure, customs clearance
Facilitation on Cross-Border
environment, government–governance environment, and cross-border logistics services as the main
E-Commerce Transactions: Analysis
Based on the Marine and Land
influencing paths; the GMM method is then applied in order to conduct a study on the impact of
Cross-Border Logistical Practices trade facilitation on the scale effect of cross-border e-commerce. The study finds that marine and
between China and Countries along land transport infrastructure has the strongest impact, with customs clearance environment and
the “Belt and Road”. Water 2021, 13, government–governance environment having the second strongest and comparable impact. The
3567. https://doi.org/10.3390/ findings of the study further clarify the differences in the application of different cross-border logistics
w13243567 facilitation measures, and provide a theoretical basis for improving the timeliness of cross-border
e-commerce transactions and reducing trade costs as well as a reference for the realization of land–sea
Academic Editor: Shuhong Wang integration and land–sea interconnection under “The Belt and Road” initiative.

Received: 29 October 2021

Keywords: cross-border logistics facilitation; marine transport; cross-border e-commerce; China
Accepted: 6 December 2021
railway express; SYS-GMM estimation
Published: 13 December 2021

Publisher’s Note: MDPI stays neutral

with regard to jurisdictional claims in
1. Introduction
published maps and institutional affil-
iations. The current global multi-billion-level consumer market tends to be e-commerce. At
the same time, “The Belt and Road” initiative put forward by the Chinese government
has deepened the integration of countries along the route with China’s ocean economy,
and has promoted cross-border e-commerce to become a new driving force for global
Copyright: © 2021 by the authors.
trade growth. In traditional goods trade, marine transportation is the mainstream mode of
Licensee MDPI, Basel, Switzerland.
transportation due to large transaction volumes. However, marine transportation is easily
This article is an open access article
affected and restricted by natural conditions such as waterway hydrological conditions
distributed under the terms and and meteorological conditions. In cross-border e-commerce, due to the small transaction
conditions of the Creative Commons volume of a single order and the higher requirements of the consignee on with respect to
Attribution (CC BY) license (https:// the timeliness of transportation, marine–land transportation is the mainstay for ensuring
creativecommons.org/licenses/by/ that goods can be delivered to customers on time. To this end, some companies rely
4.0/). on border ports and cross-border logistics channels to set up border warehouses, or set

Water 2021, 13, 3567. https://doi.org/10.3390/w13243567 https://www.mdpi.com/journal/water

 Water 2021, 13, 3567 2 of 17

Water 2021, 13, 3567 2 of 16

that goods can be delivered to customers on time. To this end, some companies rely on
border ports and cross-border logistics channels to set up border warehouses, or set up
overseas warehouses in destination countries to better meet customer consumption
up overseas warehouses in destination countries to better meet customer consumption
experience. Because overseas warehouses are restricted by the political environment of a
experience. Because overseas warehouses are restricted by the political environment of
country’s market, their establishment is not suitable for all countries along the “Belt and
a country’s market, their establishment is not suitable for all countries along the “Belt
Road”. Correspondingly, the development of cross-border e-commerce transactions will
and Road”. Correspondingly, the development of cross-border e-commerce transactions
also be affected and restricted by a country’s trade facilitation environment. In the past
will also be affected and restricted by a country’s trade facilitation environment. In the
eight years, China’s cross-border e-commerce transaction scale has reached RMB 12.5
past eight years, China’s cross-border e-commerce transaction scale has reached RMB
trillion, an increase of 11.6% year-on-year, and the transaction scale accounted for 38.86%
12.5 trillion, an increase of 11.6% year-on-year, and the transaction scale accounted for
(Figure 1). Driven by the “The Belt and Road” policy, the General Administration of
38.86% (Figure 1). Driven by the “The Belt and Road” policy, the General Administration
Customs
of CustomsofofChina
Chinahas
hassuccessively
successivelyissued
issuedaa series
series of trade facilitation
of trade facilitation measures
measureswhich
which
havecontinuously
have continuouslyoptimized
optimizedthethemarine
marineeconomic
economicenvironment
environmentand andeffectively
effectivelyreduced
reduced
transaction costs and expenses such as cross-border shipping and land transportation. ItIt
transaction costs and expenses such as cross-border shipping and land transportation.
isis particularly
particularly necessary
necessary to
toexplore
explorethetheimpact
impactmechanism
mechanismofofcross-border
cross-bordere-commerce
e-commerce
transaction scale growth against the trend of cross-border logistics facilitation,
transaction scale growth against the trend of cross-border logistics facilitation, as this as is
this
theis
the core of the new era around the practice of marine and land cross-border
core of the new era around the practice of marine and land cross-border logistics. logistics.

15
12.5
10.5
10 9
8.06
6.7
5.4
5 4.2
3.15

0
2013 2014 2015 2016 2017 2018 2019 2020
Unit: Trillion RMB

Figure 1.
Figure 1. Cross-border
Cross-bordere-commerce
e-commercetransaction
transactionscale inin
scale China, 2013–2020.
China, DataData
2013–2020. source: China
source: China
Network Economy Platform Data Center.
Network Economy Platform Data Center.
The
Theessence
essenceofoftrade is information
trade is information flow, capital
flow, flowflow
capital and logistics. In terms
and logistics. of infor-
In terms of
mation flow, compared with traditional foreign trade transactions cross-border
information flow, compared with traditional foreign trade transactions cross-border e- e-commerce
transactions use various use
commerce transactions forms of carriers
various formstoofcollect,
carrierspublish and exchange
to collect, publish andinformation
exchange
through e-commerce platform websites so as to complete transaction
information through e-commerce platform websites so as to complete transaction negotiations in the
most convenient way, thus realizing efficient transmission of information
negotiations in the most convenient way, thus realizing efficient transmission of flow. Comparing
capital flow with
information flow.traditional
Comparing trade settlement
capital flow with methods, its complexity
traditional is moremethods,
trade settlement focused on its
dealing with the remittance of different currencies of different overseas
complexity is more focused on dealing with the remittance of different currencies of buyers and the
compatibility of various
different overseas types
buyers andofthe
different localizedofpayment
compatibility various methods, mainly involving
types of different localized
settlement
payment methods, mainly involving settlement and billing issues; however at such
and billing issues; however at present, cross-border payment platforms as
present,
Stripe, PayPal, Amazon Pay, Payoneer, Alipay, etc., have realized the settlement of payment
cross-border payment platforms such as Stripe, PayPal, Amazon Pay, Payoneer, Alipay,
for goods on both sides. It can be seen that in terms of information flow and capital flow,
etc., have realized the settlement of payment for goods on both sides. It can be seen that
cross-border e-commerce transactions can bring lower transaction costs than traditional
in terms of information flow and capital flow, cross-border e-commerce transactions can
trade. However, in terms of logistics, cross-border e-commerce transactions face bottlenecks
bring lower transaction costs than traditional trade. However, in terms of logistics, cross-
in development similar to traditional trade transactions. Examples include cumbersome
border e-commerce transactions face bottlenecks in development similar to traditional
customs clearance procedures, poor infrastructure at marine logistics ports, low quality of
trade transactions. Examples include cumbersome customs clearance procedures, poor
land transport services, low efficiency of port crossings, and other cross-border logistics
infrastructure at marine logistics ports, low quality of land transport services, low
non-facilitation environmental issues. In view of this, the WTO promulgated and imple-
efficiency of port crossings, and other cross-border logistics non-facilitation
mented the Trade Facilitation Agreement in 2017, aiming to achieve simplification and
environmental issues. In view of this, the WTO promulgated and implemented the Trade
transparency of trade procedures in member countries to facilitate the free flow of goods
and services.
At present, the world’s economic and trade development is in a very sensitive period;
unilateralism and trade protectionism are expected to rise further, leading to an increase in
Water 2021, 13, 3567 3 of 16

non-tariff barriers in the future and a more severe export trade situation; coupled with the
secondary impact of the epidemic, this results in downward pressure on traditional trade
transactions and many negative factors overlapping [1]. At this time, the flexibility and
convenience of cross-border e-commerce online transactions highlight the advantages of
the future may usher in a new round of opportunities and challenges. How the government,
e-commerce enterprises and consumers can rely on the optimization of the trade facilitation
environment to carry out cross-border e-commerce transactions is a pressing issue worthy
of in-depth consideration and exploration.
In view of this, the present paper constructs a conceptual model of the impact of trade
facilitation on cross-border e-commerce transactions by taking cross-border e-commerce
transactions between China and the countries along the “Belt and Road” as the research
object. Based on this model, we use panel data to examine the differences in the impact of
trade facilitation measures on cross-border e-commerce transactions in terms of infrastruc-
ture, customs clearance environment, government–governance capacity, and cross-border
logistics services on the scale of cross-border e-commerce transactions in the countries
along the “Belt and Road”.

2. Literature Review
In domestic and international academic circles, trade facilitation, cross-border e-commerce
transactions, trade scale and other research themes have been fruitful areas of study in
recent years; however, the results of exploring the impact of trade facilitation environment
on the scale of cross-border e-commerce transactions are very limited. The existing studies,
mainly along the lines of trade facilitation and the construction of its indicator system, the
impact of trade facilitation on trade scale, etc., gradually deepen and unfold [2].

2.1. Review on the Connotation and Evaluation of Trade Facilitation

Trade facilitation connotes the enhancement and improvement of trade effects through
the simplification of trade procedures and formalities, harmonization of applicable laws and
regulations, and standardization and improvement of infrastructure, i.e., by preventing
and controlling trade risks and reducing trade costs [3]. The OECD estimates that for
every 1% reduction in trade costs, trade can increase by USD 40 billion. Regarding the
measurement of trade facilitation indicators, traditional studies focus on the simplification
and harmonization of international trade procedures, i.e., the reduction of trade costs that
occur at the “border”, mainly related to the transparency of the rules of the customs control
system, customs clearance efficiency, tariff system, etc. The contemporary trade facilitation,
on the other hand, extends from “at the border” to “within the border”, i.e., the main content
adds logistics infrastructure, communication facilities, logistics performance, government
governance capacity, and related service measures to the traditional trade facilitation [4]. At
present, the WTO, APEC and the World Bank have introduced different indicator systems
for evaluating and measuring trade facilitation. Since the evaluation system of trade
facilitation indicators is more complex and the evaluation of its measurement is not unified
in academic circles, the authors believe that before carrying out this study, it is necessary
to combine the research subjects and make a scientific assessment of the trade facilitation
evaluation system [5].

2.2. Review of Research on the Impact of Trade Facilitation on Cross-Border E-Commerce Transactions
Combing domestic and foreign research results, the research on trade facilitation can
be divided into two major aspects: first, the analysis of the impact of trade facilitation on
cross-border e-commerce transactions. Essentially, trade facilitation is the reduction of
transaction costs through “institutional arrangements” involving multiple fields, which in
turn generates trade creation and trade expansion effects, promotes the development of
international trade, and further enhances the overall welfare of society [6,7]. Second, in
terms of research methodology, the Global Trade Analysis Model (GTAP) and Computable
General Equilibrium Model (CGE) are usually used to analyze and verify the effects of
Water 2021, 13, 3567 4 of 16

enhancing trade facilitation on trade flows, trade costs, e-commerce, capital investment,
business opportunities, etc. Existing results carry out research on its contribution to trade
growth from different perspectives, such as facilitation of customs clearance procedures,
infrastructure construction standards, non-tariff barriers, port operation efficiency, domestic
environmental regulations, and construction of free trade zones, laying an important
foundation for the study in this paper [8,9].
In summary, there is a great deal of literature on the impact of trade facilitation on
cross-border e-commerce transactions, which provides important theoretical and factual
basis for this study; however, results on cross-border e-commerce trade effects as the main
body of research are not very common, and there is still the possibility of further in-depth
excavation and expansion in the research content and research methods. Specifically, the
construction of the evaluation system of trade facilitation first needs to be improved in
terms of relevance. Second, the research subject is not micro-specific enough, and results of
research on the impact of the scale effects of cross-border e-commerce trade are relatively
rare. Third, the design of the research structure is not sufficiently considered, and the
perspective is singular, without a comprehensive analysis of the impact and constraints
from the perspective of differences in trade facilitation in bilateral trade countries.
In view of this, the present paper intends to first, carry out the design of a trade
facilitation evaluation index system for bilateral countries based on transaction cost the-
ory; second, construct a conceptual model of the impact of trade facilitation environment
on cross-border e-commerce transactions; and third, use a dynamic panel model with
interaction terms to verify the effects of trade facilitation measures such as cross-border
transportation costs, cross-border customs clearance costs, political transaction costs and
time costs on cross-border e-commerce trade Finally, the effects of trade facilitation mea-
sures such as cross-border transportation costs, cross-border customs clearance costs,
political transaction costs and time costs on the scale of cross-border e-commerce trade will
be further verified using the dynamic panel model to further explore the extent of the role
of government–governance capacity on the customs clearance environment in the process
of cross-border e-commerce transactions [10,11].

3. Construction of a Conceptual Model of the Impact of Trade Facilitation Based on

Cross-Border Logistics Costs on the Scale Effect of Cross-Border E-Commerce Trade
In the fields of new institutional economics and trade economy, transaction costs
are usually interpreted as economic exchange costs, which can be broken down into two
parts, “transaction” and “cost” [12]. Transaction refers to “the conversion of ownership
of goods between individuals”; correspondingly, transaction costs are understood as
the costs incurred by both parties to establish a relationship. Williamson believes that
transaction costs are all the resources needed to negotiate, obtain business information, and
execute contracts; that is, they include manager transaction costs incurred in the exercise
of power within the company, market transaction costs incurred in the external market
of the company, and operation and maintenance Political transaction costs incurred by
the government’s institutional framework can also be included. Although cross-border
e-commerce is a new mode of international trade, transaction costs represented by cross-
border logistics occupy an important position. Compared with traditional trade, the biggest
advantage of cross-border e-commerce transactions lies in the saving of transaction costs,
that is to say, the cost of information flow between the two parties in terms of business
information acquisition, sharing, and exchange has been greatly reduced. However, cross-
border e-commerce transactions are often sporadic and small items and groceries are mainly
purchased in the form of daily consumer goods transactions; thus, higher requirements
are often placed on the flow of goods. [13,14] In this case, the cross-border logistics cost
of related goods in cross-border e-commerce business is directly related to the logistics
infrastructure, customs environment, and government–governance contexts of bilateral
trade countries. These will directly affect the confidence of cross-border e-commerce
traders and the scale of transactions via such paths as transport costs, time costs, political
transaction costs and information costs [15,16].
Water 2021, 13, 3567 5 of 16

In this way, information flow and capital flow for cross-border e-commerce transac-
tions can be completed online, and their impact on transaction costs is relatively weakened,
while logistics needs to be realized offline, and has become the bottleneck that restricts the
trade cost of cross-border e-commerce transactions. Logistics costs include a wide range of
content. According to the practical needs of cross-border e-commerce transactions, this arti-
cle mainly discusses transportation costs, customs clearance costs, government–governance
costs, and time costs.

3.1. Analysis of Infrastructure Impact Based on Cross-Border Transportation Costs

The cross-border transportation cost usually involves all the expenses for completing
the transportation within a certain period of time. In cross-border e-commerce logistics,
this cost is closely related to the quality of the national logistics infrastructure of both
parties to the transaction [17]. Infrastructure refers to the quality and service capabilities
of domestic and foreign ports and port infrastructure. Generally, it can be divided into
two aspects: hardware and software infrastructure services in international trade, usually
including the quality of seaports, airports, roads, railways and other infrastructures, and
the capacity of loading and unloading and transshipment of transportation services. In
practice, infrastructure is mainly reflected through facility quality, and facility quality as a
qualitative concept can be quantified through a series of indicators [18].
Taking Khorgos port in Xinjiang as an example, as an important transportation node
between China and Central Asian countries, the state has issued a series of supporting
development policies for this port in recent years. With the improvement of logistics
facilities, the efficiency of port functions has been rapidly improved, greatly enhancing
the port’s competitive advantage in cross-border trade and international transportation.
With the increase in the number of China–Europe express trains, the volume of customs
clearance trade has increased year by year (see Table 1).
Table 1. Statistics on the customs clearance trade volume at Khorgos port and the number of China–Europe trains opened
from 2011 to 2019.

Time 2011 2012 2013 2014 2015 2016 2017 2018 2019
Customs clearance trade
71 110 144 146 120 240 386 498 –
volume (US$ billion)
Number of China–Europe
17 42 80 308 815 1702 3673 6300 8225
Class Trains Opened

For cross-border e-commerce transactions, the quality of infrastructure plays a more

important role in the transfer of cargoes between the two sides; better logistics infrastructure
in the bilateral trading countries results in stronger operational capacity, higher operational
efficiency [19], higher quality of logistics services provided in the transaction process, and
accordingly lower transportation costs, which help improve the overall quality and scale
of a country’s cross-border e-commerce transactions and realize the positive impact of
the scale of cross-border e-commerce transactions [20]. Accordingly, this paper proposes
hypothesis 1 as follows.
Hypothesis 1 (H1). Infrastructure development in countries along the route (“along the route”
refers to the countries along the “Belt and Road”, the same below.) has a positive effect on the scale
of cross-border e-commerce by saving cross-border transportation costs.

3.2. Analysis of the Impact of Customs Clearance Condition Based on the Cost of Cross-Border
Customs Clearance
Cross-border customs clearance is one of the necessary aspects of international trade
activities, and all inbound and outbound goods must enter into circulation under customs
supervision within a specified period and at a specified location [21,22]. The customs clear-
ance environment is mainly an assessment of the customs regulatory environment, which
Water 2021, 13, 3567 6 of 16

is closely related to customs supervision and clearance management as well as commodity

inspection, quarantine, and the origin regulatory environment, including specifically the
complexity and transparency of customs clearance procedures management, the number
of export documents, turnaround time, etc. Since the establishment of the first batch of
cross-border e-commerce pilot zones in China in 2015, a total of 105 zones have been
established in five batches, aiming to reduce the costs of logistics and customs clearance for
cross-border e-commerce transactions in all aspects and to promote the experience and poli-
cies of cross-border e-commerce transactions. With innovative regulatory measures such
as “advance declaration, goods inspection and release, simplified declaration, clearance
and release, summary statistics”, the scale of cross-border e-commerce transactions has
been promoted [23,24]. According to Guangzhou Customs data, in the first four months of
2020 alone, the total import and export commodities supervised through the cross-border
e-commerce management platform in the customs area was RMB 11.8 billion, an increase
of 7% over the same period last year and accounting for 22.6% of the country.
In practical business, the quality of the environment of customs control in bilateral
countries has a direct impact on e-commerce transactions, and the optimization of the cus-
toms control environment depends on the administrative control of both sides, i.e., on the
preference for a transparent regulatory system and trade approval procedures. Therefore, it
is very important to correctly assess the effectiveness of import and export procedures, the
transparency of border management, the average number of days for customs clearance,
the number of major document types and other import and export process efficiencies. In
summary, the quality of the customs clearance environment will have a significant impact
on the cross-border e-commerce trade effect. Therefore, this article proposes hypothesis 2,
as follows.
Hypothesis 2 (H2). The cross-border customs clearance environment of coastal countries positively
affects the scale of cross-border e-commerce by reducing the cost of cross-border customs clearance.

3.3. Analysis of the Impact of Government–Governance Capacity in Importing Countries Based on

Political Transaction Costs
Political transaction cost refers to the various resources consumed in the exchange of
rights in the political sphere; such costs often arise due to information asymmetry, oppor-
tunism and asset specificity. Because the two sides in cross-border e-commerce transactions
are located in countries with different political and institutional environments, the two
sides of the transaction can easily generate asymmetric information and subsequently
incur political transaction costs [25]. At this point, the government–governance environ-
ment of bilateral countries plays an important role in influencing the political transaction
costs [26,27]. A good government–governance environment can introduce adaptive insti-
tutional policies, which can better maintain the order of national market operations, and
helps domestic manufacturers to reduce production costs and transaction costs so that their
products can have a competitive advantage in the international market competition and
promote the development of national trade [28,29].
Government–governance capacity is also reflected in the control of corruption, and
there is an important correlation between the level of government corruption and a coun-
try’s foreign trade and economic situation [30]. It has been shown that in some less
developed countries, unscrupulous businessmen influence the formulation and implemen-
tation of policy measures and even laws by, for example, bribing government officials. The
deeper the level of corruption, the more likely it is that trade intervention will occur to
protect and facilitate those who pay bribes [31]. In countries with a higher degree of democ-
racy, government administration is often widely monitored and constrained by the public,
and thus policy formulation, laws and regulations are relatively open and transparent. This
environment is more conducive to trade liberalization and trade openness, which helps
the implementation of liberal policies, thus making a country’s trade environment more
open [32,33].
Water 2021, 13, 3567 7 of 16

In addition, according to classical free trade theory and institutional economics trans-
action cost theory, a more fair, transparent and open trading environment is conducive to
the development of bilateral trade. Therefore, the effectiveness, transparency and reliability
of trade policies and regulations of bilateral trade between countries can provide facilitation
and support for trade promotion, and ensure the reduction of uncertainties in a country’s
customs clearance environment. Therefore, government–governance environment itself has
a positive impact on the scale of cross-border e-commerce transactions while at the same
time indirectly promoting the scale of cross-border e-commerce transactions through the
continuous influence and improvement of the customs clearance environment. Therefore,
this article proposes hypotheses H3a and H3b, as follows.
Hypothesis 3a (H3a). The government–governance capabilities of countries along the route
promote the expansion of cross-border e-commerce trade by saving political transaction costs.
Hypothesis 3b (H3b). The governance capacity of the countries along the route promotes the
impact of customs clearance on the scale of cross-border e-commerce.

3.4. Analysis of the Impact of Cross-Border Logistics Services Based on Time Costs
In transaction cost theory, time cost refers to the time that customers must spend to
get the desired goods or services, converted into a cost; this cost may be the time cost
of reaching an agreement, and may also be the transaction process due to the handover
conversion inefficiency brought about by waiting, or caused by missed market opportuni-
ties. According to Khorgos Customs statistics, the average whole train changeover time
of inbound and outbound China–Europe trains at Khorgos port has been reduced from
an average of 6 h in 2017 to about 3.5 h now. In addition, the inbound and outbound
China–Europe trains via Khorgos railroad port in 2019 amounted to 3403 trains, an increase
of 65.60% year-on-year, with a freight volume of 3,135,100 tons, an increase of 71.86% year
over year. For the January–May 2020 period, outbound China–Europe trains increased by
28.95% year over year, driving the cumulative export of cross-border e-commerce parcels
at Alashankou Port to exceed 10 million pieces during the same period.
As can be seen, for cross-border e-commerce merchants it is very important to complete
the delivery of high-value goods in a timely manner in order to minimize inventory backlog
and improve circulation efficiency in order to ensure corporate cash flow and give full
play to comparative cost advantages. For time costs, which are often hidden, they need to
be evaluated in combination with order elapsed time, stock preparation time, shipment
cycle and delivery time to customers. The essence of efficient cross-border e-commerce
transactions in terms of time lies in the use of synchronized online and offline operations;
however, if the exchange of goods between customers still needs to use offline logistics
to complete the actual transmission, then the time cost remains subject to cross-border
logistics services. Cross-border logistics services can be analyzed from the perspective,
of either capacity or efficiency [25]. For capacity, the impact of the freight sector on
cross-border logistics parcel tracking queries and freight service quality is examined;
efficiency, specifically logistical efficiency, is usually explained as the ratio of input to
output of logistics elements within a certain period, and represents the cost and efficiency
of the whole process of freight logistics [32]. Therefore, this article proposes hypothesis 4,
as follows.
Hypothesis 4 (H4). Cross-border logistics services along the route have a positive effect on the
scale of cross-border e-commerce trade by saving time and costs.
In summary, based on the four transmission paths under transaction costs, this paper
constructs a conceptual model of the impact of trade facilitation environmental factors on
the scale of cross-border e-commerce transactions represented by infrastructure, customs
clearance environment, government–governance capabilities, and cross-border logistics
services, and intends to introduce the interaction terms between government–governance
capabilities and customs clearance environment into the model in order to verify the
 In summary, based on the four transmission paths under transaction costs, this paper
constructs a conceptual model of the impact of trade facilitation environmental factors on
the scale of cross-border e-commerce transactions represented by infrastructure, customs
Water 2021, 13, 3567 clearance environment, government–governance capabilities, and cross-border logistics8 of 16
services, and intends to introduce the interaction terms between government–governance
capabilities and customs clearance environment into the model in order to verify the
possible differences in the impact of different trade facilitation measures on the scale of
possible differences in the impact of different trade facilitation measures on the scale of
cross-border e-commerce transactions (see Figure 2).
cross-border e-commerce transactions (see Figure 2).

Figure 2.
Figure A path
2. A path analysis
analysis framework
framework for
for the
the impact
impact of
of trade
trade facilitation
facilitation on
on the
the scale
scale effect
effect of
of cross-border
cross-border e-commerce.
e-commerce.
4. Models and Data
4. Models and Data
4.1. Variable Selection and Data Source Description
4.1. Variable
4.1.1. Selection
Explained and Data Source Description
Variables
4.1.1.This
Explained
paperVariables
selects the trading partner countries along the “Belt and Road” with
which China
This conducted
paper cross-border
selects the e-commerce
trading partner transactions
countries between
along the “Belt and2013 and
Road” 2019
with (The
which
sampleconducted
China period is set at 2013–2019
cross-border for two main
e-commerce reasons: first,
transactions data2013
between availability;
and 2019 second,
(The
this period
sample coincides
period with the transition
is set at 2013–2019 for two mainof China’s
reasons: cross-border e-commerce
first, data availability; transac-
second, this
tions from the growth phase to the maturity phase)as the sample countries
period coincides with the transition of China’s cross-border e-commerce transactions from (The sample
countries
the growthare selected
phase to theas the main
maturity destination
phase)as countries
the sample of China’s
countries export countries
(The sample cross-borderare
e-commerce
selected in main
as the the 2020 “B2C Cross-border
destination countries ofE-commerce Platform
China’s export Overseas
cross-border Research Re-
e-commerce in
port”,
the 2020including: Russia, France,
“B2C Cross-border the United
E-commerce Kingdom,
Platform Brazil,
Overseas Canada,
Research Germany,
Report”, Japan,
including:
South Korea,
Russia, France,India, Thailand,
the United Malaysia,
Kingdom, Singapore,
Brazil, Canada, Indonesia,
Germany, Kuwait,
Japan, the UAE,
South Qatar
Korea, and
India,
other 16 countries), and conducts a study on the impact of trade facilitation
Thailand, Malaysia, Singapore, Indonesia, Kuwait, the UAE, Qatar and other 16 countries), on the trade
effect
and of cross-border
conducts a study on e-commerce.
the impact of trade facilitation on the trade effect of cross-border
e-commerce.
4.1.2. Explanatory Variables
Trade facilitation level indicators are the core explanatory variables in this paper. For
the measurement of trade facilitation environment, this paper relies on the trade assessment
system proposed by Wilson in 2003 based on four aspects of trade facilitation in each partner
country, namely infrastructure, customs clearance environment, government–governance
environment and cross-border logistics service capacity. The specific index selections and
correspondences are shown in Table A1.
In addition, this article selects the number of postal international express orders (PIE),
cross-border Internet payments (CBP), per capita GDP of trading partner countries (GDPE),
broadband Internet users (ITP), etc., (Source of data for control variables: World Bank
Database) as control variables in order to carry out empirical verification and analysis.

4.2. Econometric Model Construction

Based on transaction cost theory in combination with the previous analysis, it can
be seen that the impact of trade facilitation environment on cross-border e-commerce
transactions unfolds through different transmission paths. To ensure the conciseness
of the empirical analysis process, this paper first uses principal component analysis to
calculate the quantitative values of each path, such as infrastructure, customs clearance
environment, government–governance environment, and cross-border logistics service
capacity, and then uses them as explanatory variables in the model regression. Before the
model regression, all variables are taken as natural logarithms to effectively reduce the
heteroskedasticity of the model. Considering that the growth of cross-border e-commerce
transaction scale has a certain time continuity, i.e., the current period’s cross-border e-
Water 2021, 13, 3567 9 of 16

commerce transaction scale is influenced by that of the previous period, and considering
the endogeneity problem caused by the possible omission of variables, the lagged period
of the explained variables is introduced as an explanatory variable in the model, i.e., the
dynamic panel data econometric regression model (Equation (1)) is constructed as follows:

LnTradeit = α0 + ρLnTradeit−1 + β 1 LnLFit + β 2 LnCEit + β 3 LnGEit

+ β 4 LnLSit + Xit + ε it (1)
Xit = θ1 LnPIE + θ2 LnCBP + θ3 LnGDPEit + θ4 LnITPit
where the subscript i represents the country and t represents the year. Lntrade is the
explanatory variable, indicating the scale of cross-border e-commerce transactions in each
country; Lntradei,t−1 is the period after the explanatory variable; LnLF, LnCE, LnGE, and
LnLE represent the level of infrastructure, customs clearance environment, government
governance capacity, and cross-border logistics service capacity, respectively; X is the
control variable; and ε is the random error term, which varies simultaneously with time
and individuals.
Equation (1) is the base regression model of this paper, taking into account that trade
facilitation is one of the multiple factors that promote the effect of cross-border e-commerce
trade, and the existing trade facilitation evaluation measurement data are subjective data,
which cannot measure absolute amounts in the same way as factual data; in order to better
verify the relationship between them, it is therefore appropriate to use the amount of change
to measure the analysis at the same time, in order to eliminate the individual effects in the
model and to alleviate the effects of multicollinearity between the explanatory variables
both in Equation (1) (This article first standardizes the indicators under the five paths,
and then uses the principal component analysis method to select the first two principal
components that cumulatively explain 60% or more of the change in the above difference
to perform a weighted average so as to obtain the core indicators for measuring each path.
Due to space limitations, the calculation results of each principal component score and
contribution rate are not provided here) and the common time trend between variables, etc.
Thus, perform log differencing on Equation (1) to get regression Equation (2).

∆LnTradeit = ρLnTradeit−1 + β 1 ∆LnLFit + β 2 ∆LnCEit + β 3 ∆LnGEit + β 4 ∆LnLSit + ∆Xit + ∆ε it (2)

At the same time, in order to verify for the influence of government–governance
capacity in the customs clearance environment on the scale of cross-border e-commerce
transactions, the model introduces the interaction term between government–governance
capacity and customs clearance environment to obtain the regression Equation (3).
∆LnTradeit = β 0 LnTradeit−1 + β 1 ∆LnLFit + β 2 ∆LnCEit + β 3 ∆LnGEit
(3)
+ β 4 LnLSit + β 5 ∆LnGEit × ∆LnCEit + ∆Xit + ∆ε it

4.3. Estimation Method Determination

In Equation (1), the explanatory variable is lagged by one period as the explanatory
variable; this model construction belongs to the typical dynamic panel model. The advan-
tage of a dynamic panel is that it better solves the endogeneity problem that tends to exist
in the traditional fixed-effects and random-effects model panel data models; however, the
model is vulnerable to the interference of heteroskedasticity in the model. In order to avoid
bias in the regression results, this paper proposes to adopt a two-step systematic general-
ized moment estimation, i.e., a systematic GMM approach, in the selection of estimation
methods in order to obtain consistent and valid GMM model estimates. The model also
needs to be tested for the validity of instrumental variables, i.e., the Hansen test and the
correlation test of residual terms.
In addition, it should be noted that the model includes five main explanatory vari-
ables: trade facilitation indicators and their interaction terms, the first-order lags of the
explanatory variables, and control variables such as cross-border Internet payments, GDP
per capita, and broadband Internet users. Most of these explanatory variables are closely
Water 2021, 13, 3567 10 of 16

associated with changes in the size of a country’s cross-border e-commerce transactions;

see the Pearson correlation coefficient matrix in Table A2 for details.

4.4. Model Estimation and Analysis of Results

The regressions of Equations (1)–(3) were conducted using Stata15 software; the em-
pirical results are shown in Table A3. In particular, SYS-GMM is the core analytical model
used to facilitate comparison and to carry out robustness tests of the model. The estimation
results of mixed OLS regression, fixed effects regression, random effects regression and
differential GMM are also presented in Table A3.
Looking at the overall regression results in Table A3, the model first estimates the two
key tests of validity in the systematic GMM estimation method. In particular, the p-value
for the AR(2) test is 0.310, indicating that there is no second-order serial correlation in the
residual terms. The p-value for the Hansen test statistic is 0.438, which is greater than
the commonly set critical value of 0.05; thus, all instrumental variables are valid. The R2
values of the regression models are 0.7466, 0.7587, 0.7707 and 0.8318 respectively, with
generally low R2 values. The main reasons for this analysis may be due first to the data of
the explanatory and explained variables containing both factual and subjective data, as the
simultaneous occurrence of such different types of data will likely lead to this result, and
second to the trade facilitation evaluation indicators in the explanatory variables, as the
human and social capital factor included in the explanatory variables is difficult to measure
and quantify, and subjective evaluation data must be used as a substitute. However, with
the introduction of the cross-sectional term, the R2 value increases significantly, indicating
that the explanatory power of the model has been enhanced. In terms of the regression
coefficients of the control variables, the explanatory variables with a lag of one period are
significantly positive and the regression coefficients are less than one, which is consistent
with the actual situation and helps to ensure the overall explanatory power of the model.
In addition, all three control variables also show a significant positive contribution to the
growth of cross-border e-commerce transaction size, which is more in line with both the
theoretical and practical situations.
From the regression results of the DIFF-GMM model, the overall quality of infrastruc-
ture, customs clearance environment, government–governance capacity, and cross-border
logistics services in trade facilitation measures all have a significant positive effect on the
scale of cross-border e-commerce transactions in each path, i.e., hypotheses H1, H2, H3a
and H4 are all supported. Meanwhile, in the SYS-GMM regression, infrastructure, customs
clearance environment and government–governance capacity show a stronger influence.
This suggests that for a new trade model such as cross-border e-commerce, the smooth
conduct of transactions depends first and foremost on the maturity of logistics support
facilities in bilateral countries. While cross-border e-commerce transactions are growing
at a rapid pace, the smooth flow of every node in the supply chain should be ensured.
Secondly, cross-border e-commerce transactions also require high efficiency in customs
clearance, as the efficiency and construction level of customs in terms of the time taken to
clear goods, the frequency of inspection and clearance methods, and the intelligence and
informatization of ports directly affect the cost of clearance for cross-border e-commerce
transactions. This means that there is a need to create a more streamlined customs clearance
environment and regulatory environment for cross-border e-commerce.
At present, the country is vigorously developing the construction of free trade zones
and comprehensive pilot zones for cross-border e-commerce, providing a higher level,
more preferential policies and a more simplified one-stop service platform for cross-border
e-commerce through preferential trade arrangements in free trade zones and providing
solutions and professional information support for the global supply chain by, i.e., optimiz-
ing the customs clearance environment at ports, improving the efficiency of cargo clearance
and reducing the cost of customs clearance for enterprises to achieve substantial Support.
This is all premised on the improvement of the government’s governance capacity,
which is further evidenced by the regression results of the interaction term in the SYS-GMM
Water 2021, 13, 3567 11 of 16

model. The positive coefficient of the interaction term indicates that governance capacity
not only contributes to the smooth running of cross-border e-commerce transactions on
its own, but also brings about the expansion of cross-border e-commerce transactions
by influencing the customs clearance environment. The stronger and more efficient the
governance capacity, the more likely it is in practice to provide more preferential and
transparent trade policies and convenient trade measure arrangements for importers and
exporters; thus, hypothesis H3b holds and the trade promotion effect of government–
governance capacity on cross-border e-commerce transactions is evident and empirically
verified by the data.
It should be noted that while the coefficients of the above explanatory variables can
indicate the impact of the trade promotion effect on the scale of cross-border e-commerce
transactions from their respective paths and perspectives and clarify the important role of
government governance capacity in the impact path, they have not yet been explored in
terms of the total impact of this trade facilitation condition on cross-border e-commerce
transactions. To this end, this paper will further examine the overall impact by specifically
examining the impact government–governance capacity on the scale of cross-border e-
commerce transactions by calculating the value of the bias effect in the model. According
to Equation (3) in this paper, the partial derivative can be obtained from the bias effect
Equation (4), as follows:

∂∆LnTradeit
= β 3it + β 5 ∆LnCEit (4)
∂∆LnGEit

The partial effect of government governance capacity on the trade promotion effect of
cross-border e-commerce transactions can then be calculated according to Equation (4) (In
model (4), ∆lnCE is calculated using the mean of the difference variables of the logarithm
of the change in customs environment for the full sample of countries, and the calculation
represents a bias effect based on the mean level); from the calculation results in Table A3,
it can be seen that the partial effect of government–governance capacity on the trade
promotion effect of cross-border e-commerce transactions is 4.0147 (∆lnCE uses the mean
of the difference variables of the logarithm of the quantile values of the full sample, and
therefore represents a bias effect at the mean level.). These calculation results indicate that
government–governance capacity has a significant trade promotion effect on cross-border
e-commerce transactions, and the trade promotion effect is high and very clear.

4.5. Robustness Tests

From the regression results of each econometric model in Table A3, it can be seen
that although the explanatory variables have slightly different regression results and
significance in each model, the significance and coefficient signs of cross-border logistics
services, customs clearance environment and government–governance indicators are the
same, and there are differences only in the degree of significance and coefficient size; thus,
the above regression results have clear robustness. To further prove the robustness of the
empirical validation conclusion, we regressed again with the relative proportion of the
cross-border e-commerce export (import) transaction scale as the explanatory variable. In
the regression results of the SYS-GMM model of the robustness test, there is no second-
order serial autocorrelation in the AR(2) test, and the p-values of the instrumental variable
validity test statistics are all higher than 5%; thus, the estimation results of the model
are considered to be reasonably valid. The regression results show that although there
are still slight differences in the estimated results of each explanatory variable in each
model, the two indicators of infrastructure and government–governance still show high
significant effects among the four trade facilitation environment indicators, which indicates
that the regression results of the robustness test model with the proportion of cross-border
e-commerce transaction scale as the explanatory variable also have clear robustness.
Water 2021, 13, 3567 12 of 16

5. Conclusions and Policy Recommendations

5.1. Conclusions
In this paper, we have conducted a study on the impact of trade facilitation on the
trade scale effect of cross-border e-commerce transactions, using dynamic panel data based
on interaction terms around four types of trade facilitation environmental factors: logistics
facilities, customs clearance environment, government–governance environment, and
logistical efficiency. The research returned the following results.
First, the trade facilitation environment is constrained and influenced by four paths,
transport costs, time costs, customs clearance costs and political transaction costs, which
affect the expansion of a country’s cross-border e-commerce transactions. Second, among
the different paths and factors affecting the trade facilitation environment the impact
of infrastructure quality varies greatly, which shows that the improvement of logistics
infrastructure plays an important role in the expansion of cross-border e-commerce trade on
both sides. This further demonstrates that China is right to give top priority to infrastructure
connectivity in “Belt and Road” construction. Third, the skewed effect shows that the level
of governance has significantly contributed to the improvement of the customs clearance
environment and trade facilitation.

5.2. Policy Recommendations

Under the current situation of increasingly complex international supply chains,
designing effective trade facilitation environment improvement measures is a realistic need
to adapt to the sustainable development of cross-border e-commerce. The findings of this
paper will provide a reference for the development and improvement of relevant measures
to optimize the trade facilitation environment.
First, it is necessary to further accelerate the improvement of port infrastructure
construction and to promote the interconnection of infrastructure.
The government should continue to increase the construction and investment of do-
mestic port infrastructure, and especially accelerate the construction of free trade zones
around the region as a reference point, as well as strongly support inter-regional inter-
connection. Strengthening financial and technical support for the construction of ports,
land and railroad infrastructure with countries in Southeast Asia and West Asia in order to
realize the interconnection and interoperability of infrastructure with countries along the
route is also desirable.
Second, the customs environment should be optimized, promoting customs clearance
facilitation and intelligence and strengthening cross-border inter-customs cooperation.
Continuously deepening bilateral and multilateral customs clearance cooperation
under the leadership of the government and strengthening docking and customs clearance
rules with ASEAN countries and countries along “The Belt and Road” through the construc-
tion of cross-border e-commerce pilot zones and free trade zones will further accelerate
the construction of the port management sharing platform and improve the intelligence of
port management.
Third, on the basis of the existing government administrative management advantages,
improving governance capacity will contribute to the excellence of the trade environment.
On the one hand, the government should take advantage of current administration
to comprehensively strengthen the combining of global supply chains, cut administrative
intervention in trade policies, and appropriately reduce and decentralize power, while at
the same time resolutely resisting corruption in order to create service-efficient governance.
On the other hand, through an appropriately relaxed foreign exchange control policy,
the government should relax the management of the virtual accounts of foreign payment
institutions for cross-border e-commerce and simplify filing, account opening and reporting,
etc., as reducing the complexity of these procedures and processes facilitates the operation
of enterprise funds.
Water 2021, 13, 3567 13 of 16

5.3. Research Limitations and Prospects

Based on the analysis of trade costs, this article takes countries along the “Belt and
Road” as the main body and cross-border e-commerce transactions between China and the
countries along the route as the core in exploring the influence path of trade facilitation
measures in the expansion of the scale of cross-border e-commerce transactions between
China and countries along the route and drawing preliminary research conclusions. The
findings of the study have laid an important theoretical foundation for the study of the
trade expansion effects of trade facilitation, trade costs, etc., and the scale of cross-border
e-commerce transactions. On the basis of this paper, subsequent studies can be considered
which will go deeper and expand on the following two aspects. First, the analysis of the
impact of government capacity on the influence mechanism of cross-border e-commerce
trade expansion can be deepened. This study has analysed and verified the mechanism
of the role and influence effect of government–governance capacity; however, there is no
specific node at which the government’s ability to govern plays an effective role. In the
future, if the data are available, a similar analysis and validation can be carried out by
expanding the sample of countries. It would also be worthwhile to explore the direction of
the differential economic development level of the target countries.

Author Contributions: Conception and design of trade facilitation mechanisms, Y.L.; Methodology
and model construction, L.G. and J.L.; Data collection, S.Z.; Literature synthesis, X.F.; Writing review
and editing, Y.L. All authors have read and agreed to the published version of the manuscript.
Funding: This paper is the result of Liaoning Provincial Higher Education Innovation Talent Support
Program; The project of Liaoning Provincial Social Science Foundation: Study on High Quality Direct
Investment by Enterprises in Countries along the Belt and Road from the Perspective of Human
Destiny Community (L21BGJ005).
Institutional Review Board Statement: Not applicable.
Informed Consent Statement: Not applicable.
Data Availability Statement: The data presented in this study is available on request from the
corresponding author.
Conflicts of Interest: The authors declare no conflict of interest.

Appendix A

Table A1. Description of indicators for each explanatory variable.

Type of Indicator Tier 1 Indicator Name Tier 2 Indicator Name Explanation of Indicators Data Sources
Net Energy’s Cross Border
Cross-border e-commerce Import and
Explanatory variables – E-Commerce Research
transaction size export transactions
Centre Database
Freight terminal
TEU (number of TEUs) EPS database
throughput
Cross-border H1:
Quality of infrastructure 1–5 (1 = low, 5 = high) EPS database
transport costs Infrastructure (LF)
National highway
infrastructure construction 1–7 (1 = low, 7 = high) WEF_Global Compet
along the Belt and Road
Burden of WEF_Global
1–7 (1 = low, 7 = high)
customs procedures Competitiveness Report
H2: Cargo turnaround time Number of days EPS database
Cross-border customs Customs clearance
clearance costs Number of
environment (CE) Number EPS database
export documents
Average time for customs
Number of days EPS database
clearance of exports
Water 2021, 13, 3567 14 of 16

Table A1. Cont.

Type of Indicator Tier 1 Indicator Name Tier 2 Indicator Name Explanation of Indicators Data Sources
The burden of WEF_Global
1–7 (1 = low, 7 = high)
government regulations Competitiveness Report
H3a: Government policy WEF_Global
Political 1–7 (1 = low, 7 = high)
Government–governance transparency Competitiveness Report
transaction costs capacity (GE)
Account for at least one
bribe payment request
Incidence of bribery EPS database
experienced by
the company (%)
Cross-border logistics
tracking query 1–5 (1 = low, 5 = high) EPS database
Cargo capacity
Cross-border logistics
H4:
services Capacity 1–5 (1 = low, 5 = high) EPS database
Time costs Cross-border logistics
and quality
services (LS)
The frequency of
cross-border
transportation of goods at 1–5 (1 = low, 5 = high) EPS database
the scheduled time to
reach the consignee
Statistics of China
Payment and Clearing
Cross-border internet Association and “Research
– RMB billion
payment volume Report on Third-Party
Cross-border
Control Payment Industry”
variable Number of postal
Universal Postal Union
– international Number of pieces
Postal statistics
express orders
– Internet users Number of people World Bank Datebase
– GDP per capita US$ million World Bank Datebase

Table A2. Correlation coefficient matrix of independent variables.

LnLF LnCE LnGE LnLS LnPIE LnCBP LnGDPE LnITP

LnLF 1.000
LnCE 0.404 ** 1.000
LnGE 0.358 * 0.865 ** 1.000
LnLS 0.425 ** 0.620 * 0.183 * 1.000
LnPIE 0.782 * 0.341 *** −0.298 * −0.185 ** 1.000
LnCBP −0.595 * −0.356 ** −0.387 ** −0.457 * −0.158 ** 1.000
LnGDPE 0.575 *** 0.609 ** 0.398 *** 0.308 * 0.236 ** −0.288 * 1.000
LnITP 0.386 *** 0.277 * 0.262 * 0.669 ** −0.209 * −0.335 ** 0.304 * 1.000
Note: *, **, *** denote significant at the 10%, 5%, and 1% significance levels; the correlation calculation includes data from all 16 sample
countries from 2013 to 2019.

Table A3. Model estimation results of the effect of trade facilitation on the trade effect of cross-border e-commerce.

Independent FE RE DIFF-GMM SYS-GMM

Hybrid OLS (1)
Variable (2) (3) (4) (5)
L. lnTRADE 0.1346 ** 0.1801 * 0.1524 * 0.2002 ***
0.0017 0.0035
∆LnLF 3.0211 2.9014 * 2.0357 ** 2.1170 ** 2.8557 *
(−1.56) (−2.47) (−2.78) (−2.06) (1.95)
∆LnCE −1.9014 * −0.2241 ** −0.1897 * −0.2041 * −2.3507 ***
(2.37) (1.98) (2.47) (2.06) (-2.17)
∆LnGE 2.0109 * 2.3056 ** 0.5874 ** 0.7013 *** 2.3058 ***
(−2.14) (−2.36) (2.45) (1.96) (−2.03)
Water 2021, 13, 3567 15 of 16

Table A3. Cont.

Independent FE RE DIFF-GMM SYS-GMM

Hybrid OLS (1)
Variable (2) (3) (4) (5)
∆LnLS 0.1851 ** 2.2381 ** 2.2307 * 2.2001 * 0.1943 ***
(5.99) (2.69) (2.03) (2.51) (0.96)
C 0.1874 0.1667 0.1790 0.5774 ** 0.9730 **
(3.23) (4.46) (3.81) (−1.88) (2.30)
∆LnPIE 0.1670 * 0.9311 * 0.705 * 0.3931 ** 0.2784 ***
(0.43) (2.56) (1.96) (2.37) (2.59)
∆LnCBP 0.1868 0.9041 0.7780 * 0.5033 ** 0.1859 ***
(0.86) (1.99) (1.87) (1.09) (2.38)
∆LnGDPE 0.4104 0.1752 0.1492 * 0.1998 ** 0.0307 ***
(−1.87) (−1.98) (0.57) (0.68) (−2.08)
∆LnITP 1.5807 1.2098 1.2049 1.4507 * 1.3368 *
(−0.91) (−1.47) (−2.05) (−2.06) (−2.87)
t −0.0004 −0.0006 −0.0004 −0.0001 −0.0005
(−2.22) (−2.38) (−2.51) (−0.96) (−0.08)
Constant term 12.0328 15.523 * 17.3667 * 10.0014 ** 12.3640 **
(3.38) (5.64) (−4.02) (−1.88) (5.21)
R2 0.4947 0.7466 0.7587 0.7707 0.8318
F/Wald 187.43 160.29 (1.333 601.75 799.80
(P) (0.0000) (0.0000) (0.0000) (0.0000) (0.0000)
AR(1) test 0.0000 0.0000 0.0000 0.0000
AR(2) test 0.404 0.436 0.598 0.293
Hansan test 0.368 0.401 0.398 0.438
Observed values 1344 1344 1344 1232 1232
∆LnGEit bias effect 4.0147
Note: *, **, *** denote significance at the 10%, 5%, and 1% significance levels, respectively, and the values in parentheses are the t-statistics
of the corresponding estimates; “C” is the crossover term, C = ∆LnCE × ∆LnGE; FE is estimated using the clustering robust standard error
method, and DIFF-GMM and SYS-GMM are estimated using the two-step robust standard error method; The DIFF-GMM and SYS-GMM
differential equations include the first-order lags of the explanatory variables and the first- to third-order lags of all explanatory variables as
instrumental variables, respectively, the level equation of SYS-GMM introduces the first-order differences of all explanatory variables as
instrumental variables, and the Collapse function is used to control the instrumental variables to avoid biasing the estimation results due to
too many instrumental variables; Regarding the treatment of missing values in the sample, this paper uses the mean values of the previous
and subsequent years as data to add to the dataset.

References
1. Portugal-Perez, A.; Wilson, J.S. Export Performance and Trade Facilitation Reform: Hard and Soft Infrastructure. World Dev. 2012,
40, 1295–1307. [CrossRef]
2. Wang, S.; Lu, B.; Yin, K. Financial development, productivity, and high-quality development of the marine economy. Mar. Policy
2021, 130, 104553. [CrossRef]
3. Wang, S.; Zhao, D.; Chen, H. A spatial analysis of corruption, misallocation, and efficiency. Environ. Sci. Pollut. Res. 2020, 27,
36845–36856. [CrossRef]
4. Wang, S.; Yu, H.; Song, M. Assessing the efficiency of environmental regulations of large-scale enterprises based on extended
fuzzy data envelopment analysis. Ind. Manag. Data Syst. 2018, 118, 463–479. [CrossRef]
5. Beverelli, C.; Neumueller, S.; Teh, R. Export Diversification Effects of the WTO Trade Facilitation Agreement. World Dev. 2015, 76,
293–310. [CrossRef]
6. Wang, S.; Sun, X.; Song, M. Environmental Regulation, Resource Misallocation, and Ecological Efficiency. Emerg. Mark. Finance
Trade 2021, 57, 410–429. [CrossRef]
7. Bagwell, K.; Staiger, R.W. Multilateral trade negotiations, bilateral opportunism and the rules of GATT/WTO. J. Int. Econ. 2004,
63, 1–29. [CrossRef]
8. Saslavsky, D.; Shepherd, B. Facilitating international production networks: The role of trade logistics. J. Int. Trade Econ. Dev. 2013,
23, 979–999. [CrossRef]
9. Devlin, J.; Yee, P. Trade logistics in developing countries: The case of the Middle East and North Africa. World Econ. 2005, 28,
435–456. [CrossRef]
10. Wang, S.; Zhao, D.; Chen, H. Government corruption, resource misallocation, and ecological efficiency. Energy Econ. 2020, 85,
104573. [CrossRef]
11. Engman, M. The Economic Impact of Trade Facilitation. OECD Trade Policy Pap. 2005, 92–128. [CrossRef]
12. Kalnina, I.; Xiu, D. Nonparametric Estimation of the Leverage Effect: A Trade-off between Robustness and Efficiency. J. Am. Stat.
Assoc. 2015, 112, 384–396. [CrossRef]
Water 2021, 13, 3567 16 of 16

13. Munir, K.; Ameer, A. Effect of economic growth, trade openness, urbanization, and technology on environment of Asian emerging
economies. Manag. Environ. Qual. Int. J. 2018, 29, 1123–1134. [CrossRef]
14. Helble, M.; Mann, C.L.; Wilson, J.S. Aid-for-trade facilitation. Rev. WorldEcon. 2012, 148, 357–376. [CrossRef]
15. Wang, S.-H.; Song, M.-L.; Yu, T. Hidden Carbon Emissions, Industrial Clusters, and Structure Optimization in China. Comput.
Econ. 2019, 54, 1319–1342. [CrossRef]
16. Marti, L.; Puertas, R. The importance of export logistics and trade costs in emerging economies. Marit. Econ. Logist. 2017, 19,
315–333. [CrossRef]
17. Sakyi, D.; Villaverde, J.; Maza, A.; Bonuedi, I. The Effects of Trade and Trade Facilitation on Economic Growth in Africa? Afr. Dev.
Rev. 2017, 29, 350–361. [CrossRef]
18. Wang, S.; Song, M. China Environmental Economic Development Research Report 2019: Boosting the Development of the Blue Economy;
Science Press: Beijing, China, 2020.
19. Engman, M. The Economic Impact of Trade Facilitation. Sourceoecd Emerg. Econ. 2009, 2009, 92–128.
20. Blonigen, B.A. A Review of the Empirical Literature on FDI Determinants. Atl. Econ. J. 2005, 33, 383–403. [CrossRef]
21. Furuichi, M. Cross-border logistics practices, policies, and its impact. In Global Logistics Network Modelling and Policy; Elsevier:
Amsterdam, The Netherlands, 2021; pp. 47–69.
22. Dayal, R. Cross-border co-operation. Railw. Gaz. Int. 2019, 175, 23.
23. Lin, J.Y. “One Belt and One Road” and Free Trade Zones-China’s New Opening-up Initiatives. Front. Econ. China 2015, 10, 585–590.
24. Wang, S.; Chen, S.; Zhang, H.; Song, M. The model of early warning for China’s marine ecology-economy symbiosis security. Mar.
Policy 2021, 128, 104476. [CrossRef]
25. The World Bank. Belt and Road Economics: Opportunities and Risks of Transport Corridors; The World Bank: Washington, DC, USA, 2019.
26. Pan, H. Research on International Logistics Supply Chain Management Mode from the Perspective of Cross-border E-commerce.
In The International Conference on Cyber Security Intelligence and Analytics; Springer: Singapore, 2019; pp. 737–744.
27. Yang, J. Development and Innovation of the Belt and Road Cross-Border Logistics Service System. In Current Chinese Economic
Report Series; Springer: Singapore, 2019; pp. 161–187.
28. Basalim, S. An Analysis of Cross-Border Freight Mode Choice Using Stereotype Logistic Regression: A Case Study. Int. J. Adv.
Trends Comput. Sci. Eng. 2020, 9, 65–72. [CrossRef]
29. China Credit Insurance. Country Risk Analysis Report 2018-Country Risk Ratings, Sovereign Credit Risk Ratings and Risk
Analysis of 62 Key Countries. 2018. Available online: http://www.sinosure.com.cn/gywm/gsjj/xxpl/xxgg/index.shtml
(accessed on 29 October 2021).
30. National Information Center. “Belt and Road” Data Center, “Belt and Road” Trade Cooperation Data Report. 2018. Available
online: http://www.sinosure.com.cn/gywm/gsjj/xxpl/xxgg/index.shtml (accessed on 12 October 2020).
31. Cui, X.; Guo, L.; Li, F. Can trade facilitation promote inter-provincial trade through trade creation-Evidence from a study on trade
facilitation in China. Financ. Trade Econ. 2019, 40, 100–115.
32. Zhang, Y. Investment facilitation and China’s outward foreign direct investment choices under the “Belt and Road”—An empirical
study based on cross-country panel data and investment attraction model. Int. Trade Issues 2016, 42, 165–176.
33. Sheng, B.; Jin, C. Analysis of trade facilitation levels in countries along the “Belt and Road” and China’s countermeasures. Int.
Trade 2019, 38, 4–13.