Economic Effects of Free Trade Zones based on Panel Data and
Synthetic Control Methods
QIANWEI TAN
Lyceum of the Philippines University Manila Campus,
Manila 1002,
PHILIPPINES
Abstract: - To analyze the impact of the establishment of FTZs on economic growth, the study quantitatively
evaluates the economic effects of FTZs through the synthetic control method, taking Fujian FTZs as the unit of
policy intervention, and combining the panel data with the placebo test method. The mechanism of economic
growth effect of the establishment of FTZ is analyzed through the mechanism of mediating effect. The results
show that the estimation results of the research method are better and more robust. The difference between the
true and composite values of the control variable is relatively small, with the true and composite values of
openness being 15.9226 and 15.4561, respectively, with a difference of 0.4665. In 2015, the difference in GDP
between Fujian Province and synthetic Fujian Province reached 0.1, and the difference has significantly
increased thereafter. After adding control variables, the trend of this difference over the years is consistent with
that before adding control variables, i.e., the establishment of FTZ in Fujian Province can significantly promote
economic growth. There is a 1/26 probability that other provinces and cities can obtain the same policy effect
as Fujian Province. Compared to other synthesized results, the maximum root mean square error ratio in Fujian
Province is 6.0, which is 2.8 higher than that in Jiangsu Province, indicating the maximum processing effect.
Among them, the root-mean-square error ratios of Shandong Province and Chongqing Municipality are 2.2 and
2.3, respectively, while the root-mean-square error ratio of Fujian Province is 6.0. The research method can
effectively analyze the economic effects of FTZ in Fujian Province and deepen the comprehensive
understanding of the economic effects of FTZ.
Key-Words: - Free trade zone; Economic effects; Panel data; Synthetic control method; Economic growth;
Fujian Province; Robustness check
Received: May 11, 2023. Revised: October 17, 2023. Accepted: October 29, 2023. Published: November 10, 2023.
1 Introduction
In the increasingly complex world economic
environment, countries are looking for new ways of
economic growth, trying to find a power engine that
can sustainably promote high-quality economic
development, [1], [2], [3]. Against this background,
China has established pilot free trade zones (FTZs)
based on its national and international conditions.
As a strategic development plateau, FTZs can not
only provide convenient conditions for the
development of foreign-funded industrial enterprises
in the region but also give full play to their overall
effectiveness and become an endogenous force and
driving engine for economic growth, [4], [5], [6].
Under the circumstances that the industrial chain
and supply chain are gradually anti-globalization,
China's FTZ while playing the role of innovation
and other platforms and comprehensive effects,
should also pay attention to the internal
construction, and promote the informationization
and modernization of economic development.
Fujian FTZ, as one of China's FTZs, has carried out
financial reform and other related attempts. Among
them, it is of great practical significance to analyze
the economic effects of Fujian FTZ. The synthetic
control method, as a policy effect evaluation
method, can reduce the error caused by subjective
choice. In this regard, the study takes Fujian FTZ as
the research object and chooses the synthetic control
method to understand the economic effects of Fujian
FTZ policies. The study is divided into four parts.
The first part is a literature review, which introduces
the research of scholars at home and abroad on the
FTZ and the synthetic control method. The second
part analyzes the impact of Fujian FTZ on the
regional economy through the synthetic control
method, conducts a placebo test, and conducts a
mediation effect test. The third part conducts
empirical analysis and mechanism tests. The fourth
part summarizes the research methods and so on,
WSEAS TRANSACTIONS on BUSINESS and ECONOMICS
DOI: 10.37394/23207.2023.20.213
Qianwei Tan
E-ISSN: 2224-2899
2487
Volume 20, 2023
and proposes the future research direction based on
the research deficiencies.
2 Related work
FTA, as the high ground of the current policy,
analyzing its impact on economic growth has more
important practical significance. In the analysis of
FTAs, many scholars have carried out relevant
discussions. Eteria takes Georgia and Moldova's
FTA agreement with the European Union as the
object of study, and to understand the relationship
between it and the corresponding trade performance,
the relationship is explored based on relevant data
analysis. The results show that during the agreement
period, the relevant trade shows positive
development, but the performance in export is flat,
[7]. [8], take the China-ASEAN FTA as the research
object, to explore the situation of its trade creation
effect, firstly, we carry out the selection of samples
and construct the trade gravity model. After
empirical analysis, it is found that the increase in
GDP of trading countries has a positive effect on
trade flows, [8]. [9], take COMESA FTA as the
research object, and to analyze the relationship
between NTBs and it, they carry out a case study
design to explore the relationship. From the results
of the correlation analysis, it can be seen that the
elimination of this barrier is facilitated by the
provision of bilateral agreements. [10], take the
Chinese FTA as the object of research, to study the
relationship between it and economic growth,
choose the double-difference-in-differences model
(DID), and carry out the correlation analysis. From
the results of the study, it can be seen that there is a
positive correlation between FTZs and economic
growth. [11], take FTZs as the research object, and
through analyzing the relevant panel data, they
found that the policy of FTZs is conducive to the
reduction of income disparity among provinces, and
this effect is more obvious in the provinces with
relatively low incomes.
[12], analyzed the economic effects of the
Spanish government choosing the synthetic control
method and improving it, and after a comparative
analysis, it was found that the results of the
improved synthetic control method were more
accurate. [13], used the police department in a
business district of a city in New Jersey as a study to
explore its effects on crime. The construction of
relevant control groups was carried out through the
synthetic control method. The results show that,
unlike the synthetic control area, the relevant crimes
in the target area appear to be significantly
improved. [14], face the problem of effective
formulation of market policies, take the carbon
trading policy as the research object, analyze its
implementation, study the factors affecting carbon
emissions, and explore the implementation of the
pilot policy through the synthetic control method.
After analysis, under the role of carbon trading
policy, it can promote carbon emission reduction.
To study the impact of the new coronary pneumonia
interventions on the epidemic, [15], chose the
synthetic control method and analyzed it, and the
introduction of an epidemiological house room
model was carried out based on analyzing and
observing the data. By comparing the results of the
relevant assessments, it was found that earlier
implementation of interventions favored the control
of the epidemic, while mild interventions, which
may favor the control of the epidemic at an early
stage, had less social damage.
To summarize, in the FTZ research, most of the
research methods are case studies and so on, which
have multiple treatment groups and do not apply to
the research of policy evaluation with only one
treatment group, whereas the synthetic control
method meets this requirement and can be used in
the policy evaluation with only one treatment group.
Therefore, the study adopts the synthetic control
method in the exploration of the relationship
between the construction of the FTZ and the
economic dynamics of the Fujian region. Compared
to previous studies, the study's method is apt and
can be applied to policy evaluation with only one
treatment group.
3 Application of Synthetic Control
Method in the Analysis Of
Economic Effects of FTZs
To understand the situation of economic effects of
Fujian FTZ, the study chooses the synthetic control
method and placebo test method to quantitatively
assess the economic effects of Fujian FTZ. The
mechanism of economic growth effect of the
establishment of Fujian FTZ is analyzed through the
mechanism of mediation effect.
3.1 Application of Synthetic Controls and
Placebo Tests in the Analysis of the
Economic Effects of FTAs
The establishment of the FTZ is conducive to the
market-oriented operation of China's economy and
promotes the coordination of internal economic
WSEAS TRANSACTIONS on BUSINESS and ECONOMICS
DOI: 10.37394/23207.2023.20.213
Qianwei Tan
E-ISSN: 2224-2899
2488
Volume 20, 2023
development strategies and external strategies, [16],
[17].
Selected 26
provinces and
cities
Control
unit
Policy
intervention unit
Yunnan
Province Beijing Jilin
Province
Sichuan
Province
Anhui
Province
Shandong
Province
Shanxi
Province
Jiangsu
Province
Jiangxi
Province
Hebei
Province
Heilongjiang
Province
Inner
Mongolia
Autonomous
Region
The Ningxia
Hui
Autonomous
Region
The Guangxi
Zhuang
Autonomous
Region
The Xinjiang
Uygur
Autonomous
Region
Henan
Province
Zhejiang
Province
Hainan
Province
Hubei
province
Hunan
Province
Gansu
Province
Shaanxi
province
Guizhou
Province
Liaoning
Province
Chongqing
City
Fujian
Province
Fig. 1: 26 provinces and cities
Taking the Fujian FTZ as an example, the
economic effects of the establishment of this FTZ
are analyzed. In this regard, the study adopts the
synthetic control method to make relevant
assessments. Based on the basic principles of the
method, Fujian Province is set as a policy
intervention unit. Since Fujian Province is not the
first region to implement the FTZ, the provinces and
cities in Fujian Province that have established FTZs
before are excluded from the sample, thus obtaining
one policy intervention unit and 25 control units in
Fujian Province. Among them, the details of these
26 provinces and cities are shown in Figure 1.
In Figure 1, there are 26 provinces and cities
such as Yunnan Province and Beijing. Based on the
selected 26 provinces and cities, the construction of
the relevant model is carried out, and its
mathematical expression is shown in Equation (1).
0
1
1 1 1 1 1
1
,0
N
t t t t t
tT
Y D Y D other
(1)
In Equation (1), the policy intervention effect is
denoted as
1t
,
t
,
0
T
for time, and the time-varying
policy intervention effect is set to
0
1 1 1
,,
tt
,
N
for quantity, and
1Yt
variable. To obtain
1t
, it is
necessary to estimate
1
N
t
Y
when
0
tT
denotes the
outcome variable for observable policy intervention
individuals as
1
1t
Y
. In the absence of a policy point in
time, the corresponding individual outcome variable
for the policy intervention
1Yt
can be denoted as
1
N
t
Y
, in which case
1
N
t
Y
is unobservable. The
synthetic control approach, centers on performing
the construction of the counterfactual
1
N
t
Y
of
1Yt
.
The relevant assumption is made that the
normalized weight vector of potential comparison
group objects is assumed to be
*
2 26
,,W W W
such that for each
W
vector value there exists a
potential synthetic control variable, which is
denoted as
26
2j jt
jWY
. In the meantime, there exists
an optimal weight vector
* * *
2 26
,,W W W
such
that
26
2j jt
jWY
can be used as an approximate
estimator of
1
N
t
Y
. On this basis, an estimator of
1t
can be obtained, and the relevant mathematical
expression for this estimator is shown in Equation
(2).
26
11
1 1 1 0
2,
t t t j jt
j
Y Y W Y t T
(2)
In Equation (2), the estimate of
1t
is denoted as
1t
, and the estimate of
1
1t
Y
is denoted as
1
1t
Y
.
Conduct the selection and description of variables.
Based on the research subjects selected for the
study, Fujian Province was set as the treatment
group. For the control group, to make its conditions
as identical as possible to those of Fujian Province
except for the explanatory variables, the years of the
provincial panel data selected for the study are from
WSEAS TRANSACTIONS on BUSINESS and ECONOMICS
DOI: 10.37394/23207.2023.20.213
Qianwei Tan
E-ISSN: 2224-2899
2489
Volume 20, 2023
2000 to 2018, and the provinces and cities that do
not meet the requirements, such as the relatively low
data, are excluded, and the data of 26 provinces
(municipalities) are selected.
Control
variable
Fixed assets
investment
Industrial
structure
Openness to the
outside world
Foreign direct
investment
Consumption
level
Mc
Open
FDI Is
consumption
Regional residents'
consumption expenditure
Total regional
fixed assets
investment
Proportion of Tertiary sector
of the economy/Secondary
sector of the economy
Actual amount of foreign
investment utilized by the
region
Total amount
of regional
entry and exit
Fig. 2: Relevant explanations of control variables
The year of the start of the policy is set according
to the time of the establishment of the Fujian FTZ,
which is set to 2015. The explanatory variable is
economic growth, and the level of economic growth
is expressed through the real gross domestic product
(GDP) of each province and city, taking its natural
logarithm. Among the explanatory variables, the
dummy variable treat is selected whether or not the
FTZ is established, and it indicates the FTZ reform.
The value of treat is 1 if it is established in the
current year and beyond; otherwise, it is 0. The
control variables and their relevant definitions are
shown in Figure 2.
In Figure 2, there are five control variables,
whose variable symbols are derived from the
abbreviation of the variable name or a part of it, and
the meanings represented by the symbols of
different control variables, such as FDI for foreign
direct investment, which is the amount of foreign
capital utilized in the region. According to the
synthetic control method, a synthetic Fujian
Province is obtained, which contains the provinces
(cities) of Liaoning Province, Ningxia Hui
Autonomous Region, Jiangsu Province, Zhejiang
Province, and Hainan Province, and the weights of
the corresponding provinces (cities) are 0.621, 0.045,
0.067, 0.168 and 0.098, respectively, where the
synthetic Fujian Province is the same as the real
Fujian Province in other cases. The synthetic effect
of synthetic Fujian Province is tested to analyze the
real and synthetic values of its control variables, as
well as the economic growth of Fujian Province and
synthetic Fujian Province, to explore the impact of
the establishment of FTZ on economic growth.
Robustness test. For economic growth, there are
many factors affecting it, and there are only a few
control variables selected for the study. To avoid
interference with the research estimation results due
to the omission of variables, the addition of control
variables is carried out, adding variables such as
financial market size, regional urbanization level,
etc., and the reconstruction of synthetic Fujian
Province is carried out through the synthetic control
method. The provinces (cities) included in the
synthetic Fujian Province are Jiangsu Province,
Zhejiang Province, and Hainan Province, and their
weights are 0.433, 0.266, and 0.301, respectively.
The values of the control variables of Fujian
Province and the synthetic Fujian Province are
analyzed to study the changes of GDP in different
years, and to judge the robustness of the estimation
results of the study.
To avoid the variable measurement error
affecting the estimation results of the study, the
regional economic growth level is expressed
through the per capita GDP, and the synthetic Fujian
Province obtained through the synthetic control
method changes, as shown in Figure 3.
WSEAS TRANSACTIONS on BUSINESS and ECONOMICS
DOI: 10.37394/23207.2023.20.213
Qianwei Tan
E-ISSN: 2224-2899
2490
Volume 20, 2023
Synthetic
Fujian
Province
Hainan
Province
Liaoning
Province
Zhejiang
Province
The Ningxia
Hui
Autonomous
Region
0.109
0.704
0.184
0.004
Weight
Fig. 3: Relevant explanations of control variables
In Figure 3, the provinces (cities) it contains are
Liaoning Province, Ningxia Hui Autonomous
Region, Zhejiang Province, and Hainan Province,
and the weights of the corresponding provinces
(cities) are 0.704, 0.004, 0.184, and 0.109,
respectively, with Liaoning Province taking the
largest weight. On this basis, the correlation values
of the control variables of Fujian Province and
synthetic Fujian Province are analyzed to study the
changes in their GDP under different years and to
judge the robustness of the estimation results of the
study. Compared with the DID method, although the
synthetic control method will select the control
group more strictly, the estimated economic growth
effect of the study cannot be fully guaranteed to be
affected by the FTZ because the interregional
economy is affected by more unobservable policies,
so a placebo test is needed to accurately assess the
estimated effect of the policies. The study takes 25
provinces (municipalities) other than Fujian
Province as treatment groups respectively, and
estimates them through the synthetic control method,
comparing the net effect of the policies they
obtained with the baseline results, and analyzing the
effect of the existence of the FTZ and economic
growth in Fujian Province. The results of the
multiple synthetic control method are compared
with the results of Fujian Province as the treatment
group to analyze the effect of the establishment of
FTZ in Fujian Province on economic growth. Root
Mean Square Error (RMSE) was selected to
improve the accuracy of the test by calculating the
ratio of RMSE before and after the establishment of
FTZ in each province (city) and conducting a
placebo test. That is to say that setting Fujian
Province as affected by the FTZ will result in a large
treatment effect, and the out effect on other
provinces (cities) when synthesized are all relatively
small. The RMSE ratios of Fujian Province and
other provinces (cities) are calculated to analyze the
robustness of the estimation results of the research
method.
3.2 Analysis of the Economic Growth Effect
Mechanism of the Establishment of the
Fujian FTZ
After the empirical analysis of the research method,
the mechanism of the impact of the establishment of
Fujian FTZ on economic growth is studied.
Referring to the existing research, the mechanism of
the establishment of FTZ affecting economic growth
is summarized, which can be divided into three
kinds, namely, improving the level of technological
innovation, improving industrial agglomeration, and
improving the efficiency of resource allocation.
Based on this, the study adopts the mediation effect
model when conducting the impact mechanism
research. First, the mediation effect test of the
innovation drive is carried out. In the test,
technological innovation is set as a mechanism
variable, and in the process of testing the impact of
the establishment of FTZ on economic growth
through regression analysis, the first relevant
econometric modeling is carried out, in which the
relevant mathematical expression for testing the
impact of the establishment of FTZ on economic
growth is shown in Equation (3).
it 0 1 2
lngdp _ it it it it
t treat X
(3)
In Equation (3), the set of economic growth is set
to
lngdp
. The dummy variable for the establishment
of FTZ is denoted as
_it
t treat
, and the set of values
for all the years before the establishment of FTZ in a
certain region is set to 0, and in the year of the
establishment of the FTZ in the region and beyond
is set to 1. According to the year of the start of the
policy of FTZ in Fujian Province in the study, the
lagged value of the province in 2015 is set to 1, and
the value of other provinces and cities is set to 0.
The set of control variables is set to
it
X
. To make
the estimation results reliable, the selection of
relevant control variables is the same as the above
selection.
1
The regression coefficients are
2
and
3
, and the parameters are
0
,
it
and
it
. The
relevant mathematical expressions involved in
testing the impact of the establishment of FTZ on
technological innovation are shown in Equation (4).
WSEAS TRANSACTIONS on BUSINESS and ECONOMICS
DOI: 10.37394/23207.2023.20.213
Qianwei Tan
E-ISSN: 2224-2899
2491
Volume 20, 2023
it 0 1
2
lnrd _ it
it it it
t treat
X
(4)
In Equation (4), the level of technological
innovation is expressed as
lnrd
and is measured by
the number of patents granted in several zones
lnrd
.
The mediating effect of technological innovation in
FTZs affecting economic growth is tested, involving
the relevant mathematical expressions shown in
Equation (5).
it 0 1
2 it 3
lngdp _
lnrd
it
it it it
t treat
X
(5)
Based on the above regression equation, the
innovation-driven mediation effect test is carried out
to determine whether technological innovation is the
mediating mechanism for the FTZ's effect on
economic growth from the relevant regression
coefficients obtained. The establishment of the FTZ
has brought many conveniences to the companies in
the region, such as tax breaks and exemptions. If
companies in the FTZ can get more policy support
for import and export and lower prices for imported
goods, then for companies in other industries, the
establishment of a branch in the FTZ can help them
reduce costs and improve their productivity, thus
enhancing their market competitiveness. Therefore,
compared with regions without FTZs, regions with
FTZs will have a certain "siphoning" effect on
enterprises in the neighboring regions, thus
attracting more enterprises to come. To analyze
whether the level of industrial agglomeration is the
intermediary mechanism of the establishment of
FTZ affecting regional economic growth, the same
method as that of the innovation-driven
intermediary effect test is adopted to carry out the
correlation analysis. Among them, the relevant
mathematical expressions involved in the
calculation of industrial agglomeration are shown in
Equation (6).
31 31
11
it it it it it
aggl ind ind lngdp lngdp
(6)
In Equation (6), the index of industrial
agglomeration is set to
it
aggl
, and the proportion of
the output value of the secondary industry to the
total output value of the secondary industry of the
province is expressed as in the
n
it it
i
ind ind
t
th
year of the
i
province (city). In the year
t
of
i
province (city), the proportion of GDP to the total
GDP of the sample province (city) after the GDP is
expressed as
ln ln
n
it it
i
gdp gdp
. For
aggl
, the
higher value means a higher level of industrial
agglomeration. On this basis, the test of mediation
effect driven by industrial agglomeration is carried
out, so that the results obtained can be judged
relevantly. The establishment of FTZ will break the
barriers to the flow of production factors,
commodities, etc. Based on this, the study explores
the impact of the establishment of FTZ on economic
growth from the perspective of resource allocation
efficiency. Total factor productivity is calculated
through deap2.1, and in the selection of input
variables, fixed capital stock, and employment are
selected, and GDP is taken as the output variable.
The Malmquist index was calculated assuming
constant returns to scale. Using this index, the
efficiency of resource allocation is measured. It is
worth mentioning that regions with higher total
factor productivity possess relatively higher
resource allocation efficiency. The mediation effect
driven by resource allocation efficiency is tested by
the method of mediation effect test described above.
Based on the regression estimation results, relevant
judgments are made.
4 Empirical Analysis and Mechanism
Test of Economic Effects of Fujian
Free Trade Zone
The synthetic control method and placebo test are
chosen to quantitatively assess the economic effects
of Fujian FTZ. Through regression analysis, the
mediating effect test of innovation-driven, industry
agglomeration-driven, etc. is carried out.
4.1 Benchmark Regression Results and
Robustness Testing Analysis
The provincial panel data selected for the study
ranged from 2000 to 2018. Provinces and cities with
relatively low data that did not meet the
requirements were excluded, and data from 26
provinces (cities) were selected. According to the
synthetic control method, a synthetic Fujian
province is obtained, which includes provinces
(cities) such as Liaoning Province, Ningxia Hui
Autonomous Region, Jiangsu Province, Zhejiang
Province, and Hainan Province. The relevant
weights are shown in the previous text. The
synthetic effect of synthetic Fujian Province under
the synthetic control method is examined to analyze
the real and synthetic values of control variables
under synthetic Fujian Province, and the specific
results are shown in Figure 4.
WSEAS TRANSACTIONS on BUSINESS and ECONOMICS
DOI: 10.37394/23207.2023.20.213
Qianwei Tan
E-ISSN: 2224-2899
2492
Volume 20, 2023
9.3214 8.6875
0.8335 0.0472
15.9226
Variable value
Variable
0.0000
2.0000
4.0000
6.0000
8.0000
10.0000
12.0000
14.0000
16.0000
18.0000
Consumption
level
Fixed assets
investment
Industrial
structure
Foreign direct
investment
Openness to
the outside
world
(a) Control variable true value
9.3061 8.6759
0.9855 0.0457
15.4561
0.0000
2.0000
4.0000
6.0000
8.0000
10.0000
12.0000
14.0000
16.0000
18.0000
(b) Composite value of control variables
Consumption
level
Fixed assets
investment
Industrial
structure
Foreign direct
investment
Openness to
the outside
world
Variable value
Variable
9.0134 8.4944
1.0228 0.0219
14.3900
Variable value
0.0000
2.0000
4.0000
6.0000
8.0000
10.0000
12.0000
14.0000
16.0000
Consumption
level
Fixed assets
investment
Industrial
structure
Foreign direct
investment
Openness to
the outside
world
(c) Control variable, mean of all provinces and cities
Variable
Fig. 4: Relevant explanations of control variables
In Figure 4(a), the real values of different control
variables are different, the real value of the degree
of external openness is the largest, while the real
value of foreign direct investment is the smallest.
The real value of external openness is 15.9226,
which is 6.6012 higher than the consumption level,
while the latter's real value is 9.3214; the real value
of industrial structure is smaller at 0.8335. In Figure
4(b), the synthetic value of external openness is the
largest, followed by the consumption level. The
synthetic value of the degree of openness to the
outside world is 15.4561, which is 0.4665 smaller
than its true value; the synthetic value of the
consumption level is 9.3061, which is 0.0153
smaller than its true value. Comparing Figure 4(a)
and Figure 4(b), it can be seen that the difference
between the true value of the control variables and
the synthetic stone is small. In Figure 4(c), the
average value of all provinces and cities of the other
party's openness is 14.3900, which is 5.8956 larger
than the fixed asset investment. To study the
economic growth of Fujian Province and synthetic
Fujian Province, the specific results are shown in
Table 1.
Table 1. GDP of Fujian Province and composite Fujian Province
Time (year)
2000
2001
2002
2003
2004
2005
Fujian's (real) GDP
8.23
8.31
8.40
8.51
8.66
8.79
Composite GDP
8.19
8.28
8.38
8.50
8.63
8.80
Difference
0.04
0.03
0.02
0.01
0.03
-0.01
Time (year)
2006
2007
2008
2009
2010
2011
Fujian's (real) GDP
8.93
9.13
9.29
9.41
9.60
9.77
Composite GDP
8.95
9.13
9.32
9.42
9.61
9.80
Difference
-0.02
0.00
-0.03
-0.01
-0.01
-0.03
Time (year)
2012
2013
2014
2015
2016
2017
2018
Fujian's (real) GDP
9.89
9.99
10.05
10.19
10.37
10.48
10.56
Composite GDP
9.90
9.99
10.05
10.09
10.15
10.21
10.26
Difference
-0.01
0.00
0.00
0.10
0.22
0.27
0.30
WSEAS TRANSACTIONS on BUSINESS and ECONOMICS
DOI: 10.37394/23207.2023.20.213
Qianwei Tan
E-ISSN: 2224-2899
2493
Volume 20, 2023
In Table 1, the difference between the GDP of
Fujian Province and synthetic Fujian Province under
different years is different, the difference is very
small before 2015, and after 2015, and the
difference becomes bigger gradually. When the time
is 2001, the real GDP value of Fujian Province was
8.31, which is 0.03 larger than that of synthetic
Fujian Province; when the time is 2007, the GDP of
Fujian Province and synthetic Fujian Province were
both 9.13; when the time is 2015, the GDP of Fujian
Province and synthetic Fujian Province are 10.19
and 10.09, respectively, and the difference between
them is 0.10; and when the time is 2018, the real
GDP value is 10.56, which is 0.30 larger than that of
synthetic Fujian Province. It can be seen that
compared with the control group for the
establishment of FTZ, the GDP of Fujian Province
is significantly improved, and the establishment of
FTZ can significantly promote economic growth.
For the robustness test, after adding the control
variables such as financial market size, the control
variables and the true value of the synthetic Fujian
Province and the GDP of Fujian Province and the
synthetic Fujian Province are shown in Figure 5.
8.0000
6.000
4.000
2.0000
0.0000
Control variable value
10.0000
True value
Synthetic value
(a) Comparison of real and synthetic values of control variables
Level of
consumptio
n
Investment
in fixed
assets
Structure
of
industry
Foreign
direct
investment
Degree of
openness
Urbanizatio
n rate
Financial
market
size
12.0000
14.0000
Average of all
provinces (cities)
Variable
10.00
9.50
9.00
8.50
8.00
2000 2005 2010 2015 2020
lngdp
10.50
Year
Fujian
(b) Economic growth in Fujian Province and Fujian combined
Synthetic Fujian
Auxiliary line
Fig. 5: Control variables and GDP of Fujian
Province and composite Fujian Province
In Figure 5(a), except for individual control
variables, the synthetic values of other variables are
closer to the true values. The real value of industrial
structure is 0.8335, which is 0.3200 smaller than the
synthetic value, which is 1.1535; the synthetic and
real values of consumption level are 9.3476 and
9.3214, respectively; and the real value of
urbanization rate is 0.5459, which is 0.0069 smaller
than the synthetic value. In Figure 5(b), the trend of
the change in GDP of Fujian Province and synthetic
Fujian Province over time is the same as that of
Table 1 It is the same. When the time is 2005, the
real GDP value of Fujian Province is 8.79, which is
0.05 smaller than that of synthetic Fujian Province,
which is 8.84; when the time is 2017, the GDP of
Fujian Province and synthetic Fujian Province are
10.48 and 10.33, respectively. Thus, it can be seen
that the estimation results of the research method are
robust. To avoid the impact of variable
measurement error on the estimation results, the
replacement of the explanatory variables is carried
out, which is replaced by GDP per capita, and thus
the control variables and GDP of Fujian Province
and synthetic Fujian Province are shown in Figure 6.
0.0000
2.0000
4.0000
6.0000
8.0000
10.0000
12.0000
14.0000
16.0000
18.0000
Control variable value
True value Synthetic value
Variable
Consumption
level
Fixed assets
investment
Industrial
structure
Foreign direct
investment
Openness to
the outside
world
(a) Control variable true value
11.00
10.50
10.00
9.50
2000 2005 2010 2015 2020
Per capita GDP value
11.50
Year
Fujian
(b) Per capita GDP of Fujian Province and
Synthetic Fujian Province
Synthetic Fujian
Auxiliary line
Fig. 6: Control variables and GDP of Fujian
Province and composite Fujian Province
WSEAS TRANSACTIONS on BUSINESS and ECONOMICS
DOI: 10.37394/23207.2023.20.213
Qianwei Tan
E-ISSN: 2224-2899
2494
Volume 20, 2023
In Figure 6(a), the synthetic and real values of
the control variables are closer to each other, the
synthetic and real values of the degree of openness
to the outside world are 15.5184 and 15.9226
respectively, with a difference of 0.4042; and the
synthetic and real values of the foreign direct
investment are the same, both are 0.0472.
0.2
0.1
0.0
2000 2005 2010 2015 2020
GDP difference
0.3
Year
(a) Net effect
GDP difference
Auxiliary line
0.2
-0.2
-0.6
2000 2005 2010 2015 2020
GDP difference
0.4
Year
(b) Sorting test
-0.4
0.0
Fujian
The Guangxi Zhuang
Autonomous Region
Jiangsu Province
Hunan Province
Chongqing City
Sichuan
Hubei province
Auxiliary line
Fig. 7: Net effect and ranking test
From the trend of the graph in Figure 6(b), the
trend of the change is still the same as that of Table
1, and after 2015 the real GDP per capita increases
significantly, and when the time is 2016, the GDP
per capita of Fujian Province and synthetic Fujian
Province are 11.23 and 11.18 respectively. Further
proving that the estimation results of the research
method are robust. A placebo test is conducted to
obtain the net effect of the FTA and the partial
ranking test after the multiple synthetic control
method is shown in Figure 7.
Figure 7(a) shows the difference between the
GDP of the treatment group and the control group,
and it can be seen from the trend of the graph that
the net effect before 2015 stays near 0, and the net
effect increases significantly after 2015, which
indicates that there is a promotion effect of the FTZ
on economic growth. In Figure 7(b), the results of
the synthetic control method in different provinces
and cities differ, and by comparing their graphs, it
can be seen that compared with other provinces and
cities, the difference between the real GDP and the
synthetic group GDP in Fujian Province is more
obvious, and other provinces and cities have 1/26
probability of obtaining the same policy effect as
that of Fujian Province. The RMSR ratio test is
carried out, which is shown in Figure 8.
0.0
1.0
2.0
3.0
4.0
5.0
6.0
Yunnan
Inner
Beijing
Jilin
Sichuan
Ningxia
Anhui
Shandon
Shanxi
Guangxi
Xinjiang
Jiangsu
Jiangxi
Hebei
Henan
Zhejiang
Hainan
Hubei
Hunan
Gansu
Fujian
Guizhou
Liaoning
Chongqi
Shanxi
Heilongji
Province
RMSE ratio
Fig. 8: Net effect and ranking test
WSEAS TRANSACTIONS on BUSINESS and ECONOMICS
DOI: 10.37394/23207.2023.20.213
Qianwei Tan
E-ISSN: 2224-2899
2495
Volume 20, 2023
In Figure 8, Fujian Province has the largest
RMSR ratio of 6.0, followed by Guangxi Zhuang
Autonomous Region, whose RMSR ratio is 3.4.
Jiangsu Province has an RMSR ratio of 3.2, which
is 0.5 larger than that of Hunan Province, which is
2.7, and Shandong Province and Chongqing
Municipality have RMSR ratios of 2.2 and 2.3,
respectively. It can thus be seen that Fujian Province
receives the largest treatment effect, and the study
estimation results are robust.
4.2 Empirical Analysis of the Economic
Effects of the Fujian Free Trade Zone
The mechanism of Fujian FTZ affecting economic
growth is analyzed, and based on regression
analysis, the innovation-driven mediation effect test
is carried out, and the relevant results are shown in
Table 2.
Table 2. Correlation regression results of technological innovation
Variable
Equation (3)
Equation (4)
Equation (5)
lngdp
lnrd
lngdp
_t treat
0.438###
0.462###
0.429###
lnrd
/
/
0.031##
lnxf
0.522###
0.630###
0.505###
lnmc
0.192###
-0.026
0.192###
isl
0.048###
-0.067
0.050###
lnopen
0.043###
0.286###
0.035##
Note: ### indicates P < 0.01 and #### indicates P < 0.001.
In Table 2, the regression coefficient of FTZ
establishment on economic growth is significantly
positive, which is consistent with the above
estimation results. The coefficient in column 3 is
significantly positive (P < 0.001), which means that
FTZ can positively and significantly affect
technological innovation. The coefficients of and
in column 4 are significantly positive (P < 0.001),
which means that technological innovation is the
mediating mechanism of FTZs affecting economic
growth effects. The test of the mediating effect
driven by industrial agglomeration is conducted, as
shown in Table 3.
Table 3. Relevant inspection results
Variable
aggl
lngdp
_t treat
0.369###
0.365###
aggl
/
0.476##
lnxf
0.339###
0.361###
lnmc
0.281###
0.058###
isl
0.004
0.046###
lnopen
0.112###
-0.010
Note: ### indicates P < 0.001.
In Table 3, the regression coefficient of the
establishment of FTZ on industrial agglomeration is
significantly positive, indicating that the
establishment of FTZ has a certain positive effect on
regional industrial agglomeration. The coefficients
in column 3 are significantly positive (P < 0.001),
which implies that the level of industrial
agglomeration is the intermediary mechanism of the
FTZ's effect on economic growth. The test of the
mediating effect driven by resource allocation
efficiency is conducted, which is shown in Table 4.
Table 4. Correlation regression results
Variable
tfpch
lngdp
_t treat
0.033#
0.430###
tfpch
/
0.367###
lnxf
0.041##
0.508###
lnmc
0.016##
0.186###
isl
0.024###
0.039###
lnopen
-0.003
0.044###
Note: #, ###, #### indicate P < 0.05, P < 0.01, P <
0.001, respectively.
Table 4, denotes the resource allocation
efficiency, and the coefficients, in column 3 are
significantly positive (P < 0.001), which implies that
the resource allocation efficiency is the intermediary
mechanism of the FTZ affecting the economic
growth effect. From the analysis results in Table 2,
Table 3, and Table 4, it can be concluded that the
establishment of free trade zones affects three
mechanisms of economic growth: firstly, to improve
the level of technological innovation; Secondly,
improving industrial agglomeration; The third is to
improve the efficiency of resource allocation.
WSEAS TRANSACTIONS on BUSINESS and ECONOMICS
DOI: 10.37394/23207.2023.20.213
Qianwei Tan
E-ISSN: 2224-2899
2496
Volume 20, 2023
5 Discussion
Analyze it based on empirical results. Overall, the
establishment of the Fujian Free Trade Zone has a
significant promoting effect on economic
development, industrial agglomeration, and other
aspects of development. The establishment of the
Fujian Free Trade Zone has played a positive role in
promoting economic growth, investment, financial
openness, and trade in Guangdong Province. That is
to say, the establishment of the Fujian Free Trade
Zone has promoted financial openness through a
series of institutional innovations, improved the
level of investment and trade liberalization and
facilitation, and thus promoted local economic
growth. However, in relatively developed regions,
its policy effect is not sustainable. Therefore, the
Fujian Free Trade Zone should further deepen
institutional innovation, deepen reforms, and
formulate new strategic goals based on the actual
situation. From this, we can draw relevant policy
insights. Firstly, we should focus on maintaining the
vitality and effectiveness of economic and trade
policies, continue to deepen the facilitation process
of promoting cross-border trade and investment and
continue to seek innovative development. Optimize
the economic and trade environment to "safeguard"
the continued expansion of opening up to the
outside world. The Fujian Free Trade Zone can
adopt the policy of a "pre-admission national
treatment+negative list". Under this institutional
framework, the approval system for foreign
investment admission will be transformed into a
record-keeping system, and the threshold will be
relaxed while strengthening management, focusing
on both sides. In addition, it is possible to gradually
improve the optimization and reform of the financial
system, promote the upgrading of the industrial
structure of the free trade zone, seek impetus for
long-term economic growth within the province,
and achieve healthy, stable, and sustainable
development of the economy in the previous period.
6 Conclusion
To explore the situation of economic effects of
Fujian FTZ, the study takes Fujian FTZ as the
research object, chooses the synthetic control
method, selects 25 provinces and cities as the
control unit, and evaluates the economic effects of
the establishment of Fujian FTZ based on the annual
panel data, and conducts the placebo test and the
mediation effect test such as innovation drive. The
results show that after the synthetic control method,
the difference between the real and synthetic values
of the control variables is small, and the synthetic
effect is better. The real value of the degree of
openness to the outside world is the largest at
15.9226, which is 0.4665 larger than its synthetic
value, which is 15.4561. Before 2015, the difference
between the GDP of Fujian Province and the
synthetic Fujian Province is smaller; after 2015 its
difference becomes larger, and the real GDP of
Fujian Province is larger. When the time is 2018,
the real GDP value of Fujian Province is 10.56,
which is 0.30 larger than that of the synthetic Fujian
Province. In the robustness test, the estimation result
of the research method is robust, and the synthetic
and the real values of the consumption level are
9.3476 and 9.3214 respectively after adding the
control variables. Compared with the other synthetic
results, Fujian Province has the largest RMSR ratio,
which is 6.0, which is 2.8 larger than that of Jiangsu
Province. Technological innovation, industrial
agglomeration level, and resource allocation
efficiency are the mediating mechanisms of FTZs
affecting the economic growth effect. Among them,
the coefficient of is significantly positive (P <
0.001). This shows that the application of the
research method is better. The refinement of
research data and variable selection needs to be
optimized, and in the future, the economic effect can
be analyzed from more subdivided perspectives.
References:
[1] Pasara M T, A WITS-SMART simulation
analysis of trade creation, diversion and
welfare effects of the African tripartite
region, International Journal of Trade and
Global Markets, Vol.14, No.1, 2021, pp. 31-
47.
[2] Kouty M, Implementing the African
continental free trade area (AFCFTA): the
effects of trade procedures on trade flows,
Research in Applied Economics, Vol.13,
No.1, 2021, pp. 15-31.
[3] Ijaiya T A, Ijaiya T, Ijaiya M A. African
continental free trade area (AFCFTA) and
poverty reduction in Nigeria: an exploration
of the possibilities, Ilorin Journal of
Economic Policy, Vol.8, No.3, 2021, pp. 21-
35.
[4] Toropygin A V, Economic and political
aspects of the Serbia-EAEU free trade area,
Vol.14, No.2, 2021, pp. 120-131.
[5] Jacobsz T, Rules of origin. Notes on
challenges for the African continental free
trade area (AFCFTA), Global Trade and
WSEAS TRANSACTIONS on BUSINESS and ECONOMICS
DOI: 10.37394/23207.2023.20.213
Qianwei Tan
E-ISSN: 2224-2899
2497
Volume 20, 2023
Customs Journal Vol.15, No.3/4, 2020,159-
163.
[6] Rélouendé Zidouemba P, Jallab M S, The
African continental free trade area and the
trade facilitation agreement: some regional
macroeconomic impacts, International
Journal of Trade and Global Markets,
Vol.14, No.3, 2021, pp. 325-337.
[7] Eteria E, Impact of the EU deep and
comprehensive free trade area agreements on
trade performance of Georgia and Moldova,
CES Working Papers, Vol.12, No. 3, 2020,
pp. 198-212.
[8] Han J, Tian Z, The effect of trade creation in
China-Asean free trade area based on the
gravity model. Journal of Intelligent & Fuzzy
Systems, Applications in Engineering and
Technology, Vol.38, No.6 y, 2020, pp. 7061-
7065.
[9] Sibangilizwe M, Jeffrey K, The implications
of non-tariff barriers to trade on COMESA
free trade area: the case of Zimbabwe and
Zambia, Canadian Social Science, Vol.15,
No.2, 2019, pp. 34-43.
[10] Zhao T, He F, Does the pilot free trade zone
promote the quality of urban economic
growth: an empirical research based on
quasi-natural experiment? Sustainability,
Vol.14, No.12, 2022, pp. 1-14.
[11] Cheng H, Zhang Y I, Hongyi L I, Effects of
free trade zone policy on China's income
disparity, The Singapore Economic Review,
Vol.67, No.2, 2020, pp. 709-732.
[12] Albalate D, Bel G, Mazaira-Font F A,
Decoupling synthetic control methods to
ensure stability, accuracy and
meaningfulness, SERIEs, Vol.12, No.4, 2021,
pp. 549-584.
[13] Piza E L, Wheeler A P, Connealy N T, Feng
SQ, Crime control effects of a police
substation within a business improvement
district: a quasi -experimental synthetic
control evaluation, Criminology and Public
Policy, Vol.19, No.2, 2020, pp. 653-684.
[14] Qi X, Han Y, How carbon trading reduces
China's pilot emissions: an exploration
combining LMDI decomposition and
synthetic control methods, Polish Journal of
Environmental Studies, Vol.29, No.5, 2020,
pp. 3273-3284.
[15] Tian T, Luo W, Tan J, Jiang Y, Zhang H,
The timing and effectiveness of
implementing mild interventions of COVID-
19 in large industrial regions via a synthetic
control method, Statistics and its Interface,
Vol.14, No.1, 2021, pp. 3-12.
[16] Yazbek H A, Surriya F, Khan S U, Jan N,
Marinkovic D, A novel approach to model
the economic characteristics of an
organization by interval-valued complex
pythagorean fuzzy information, Journal of
Computational and Cognitive Engineering,
Vol. 2, No.1, 2023, pp. 75-87.
[17] Ajibo C C, African continental free trade
area agreement: the euphoria, pitfalls and
prospects, Journal of World Trade, Vol. 53,
No.5, 2019, pp. 871-894.
Contribution of Individual Authors to the
Creation of a Scientific Article (Ghostwriting
Policy)
The article is written and revised by the only author
Qianwei Tan.
Sources of Funding for Research Presented in a
Scientific Article or Scientific Article Itself
No funding was received for conducting this study.
Conflict of Interest
The authors have no conflict of interest to declare.
Creative Commons Attribution License 4.0
(Attribution 4.0 International, CC BY 4.0)
This article is published under the terms of the
Creative Commons Attribution License 4.0
https://creativecommons.org/licenses/by/4.0/deed.en
_US
WSEAS TRANSACTIONS on BUSINESS and ECONOMICS
DOI: 10.37394/23207.2023.20.213
Qianwei Tan
E-ISSN: 2224-2899
2498
Volume 20, 2023
