Economic Effect of Ecological Reconstruction in Poyang Lake
Yong Zheng
1
, Yu e Fu
2
, Hua Yin
3
1
College of Economics, Sichuan Agricultural University, Chengdu, China
2
College of Economics, Sichuan Agricultural University, Chengdu, China
3
Institute of Shanghai Development Research, Shanghai University of Finance and Economics, Shanghai, China
181764131@qq.com, 491452540@qq.com, 412953666@qq.com
Keywords: Poyang Lake Ecological Reconstruction, Economic Effect, Counterfactual Method.
Abstract: It is difficult to show the economic value of environment improvement. By using the counter factual policy
evaluation method, we evaluated the economic value of the ecological reconstruction project in Poyang
Lake. Our results show that the economic effect of the implementation of environmental projects is
significant. After the implementation of the project, Jiangxi Province economic growth rate increased by 3
percentage points from the 1 quarter of 2014 to the 4 quarter of 2013.
1 INTRODUCTION
With the improvement of people's living standard
and the enhancement of environmental protection
consciousness, people put forward higher and higher
request to the environment quality. Scientific
evaluation of the economic benefits of improving
environmental quality has attracted wide attention of
policy makers and researchers. As early as in 2002,
China has promulgated the "People's Republic of
China Environmental Impact Assessment Law",
which requires strict assessment of the economic
value of environmental changes.
Although the improvement of the ecological
quality is very important for improving the welfare
of the residents, but it is difficult to show the
economic value. Therefore, in order to help policy
makers to better formulate the environmental policy,
it is necessary to actively explore a reasonable
environmental assessment methods, and construct
the system of cost - income evaluation.
By using "hedonic price method", researchers
have studied the influence of air quality
improvement on the housing price in Qingdao
(Yongwei Chen and Lizhong Chen, 2012). They
found that air pollution 1 indices decrease,
consumers are willing to pay 99.785 yuan more per
square meter. There are others use the "hedonic price
method" to study the impact of air quality on the real
estate price (Bender et al., 1980; Brucato et al. 1990)
by using the method of DID and transformation of
distribution function, researchers have studied the
influence of environmental improvement of power
plant relocation on the housing market (Guoying
Deng et al., 2013). They found that after the
relocation of power plants, housing transaction
volume in that district increased by 54.6% compared
to other areas, the average transaction price
significantly increased by 6.8%-10.3%. To study the
economic effect of air pollution, some others used
"natural experiment" method (Chay and Greenstone,
2005). The results show that during 1970-1980, the
U.S. government's governance of air quality has
brought about 45 billion dollars of housing asset
premium. In addition, scholars have studied the
coordination between carbon emission and GDP
growth in the process of urbanization (Boqiang Lin
and Xiying Liu, 2010). They examined the long-term
equilibrium relationship between carbon dioxide
emissions and the main economic variables by using
cointegration method.
All of the above studies show that improving
environmental quality is conducive to the
improvement of economic efficiency, but the above
research is not perfect, the main problem is: first of
all, although "Hedonic price model" can measure the
impact of various properties on housing prices, it is
inevitable that there is endogeneity problem in the
empirical analysis, and the "hedonic price model"
tends to ignore the correlation of house price in
space, which results in the error of the estimation.
Second, the use of "natural experiments" requires
detailed microscopic data, but the data are often not
available. Last, the above research has not made an
36
36
Fu Y., Zheng Y. and Yin H.
Economic Effect of Ecological Reconstruction in Poyang Lake.
DOI: 10.5220/0006443300360040
In ISME 2016 - Information Science and Management Engineering IV (ISME 2016), pages 36-40
ISBN: 978-989-758-208-0
Copyright
c
2016 by SCITEPRESS Science and Technology Publications, Lda. All rights reserved
accurate assessment of the impact of environmental
quality improvement on economic growth, but only
in a certain part of the economic benefits, such as
housing market.
In view of the above mentioned problems, this
paper will introduce a new method of policy
evaluation to make an accurate assessment of the
economic effect of Poyang Lake ecological
construction. Our method can well solve the above
three problems mentioned.
2 MODEL SETTING
Based on the appraisal method for counterfactual
policies (
Cheng, H., Ching, H. S., and Shui, K. W.,
2012), this paper will construct reasonable
comparison targets of the experimental group,
overcoming the difference problem between the
experimental group and the control group. In this
process, we need to seek a reasonable control group
to appraise the counterfactual GDP growth rate of
Jiangxi Province. The developments of different
economy are mutually connected, and influenced by
some common factors, such as economic cyclical
changes and technological progress, etc. Thus, we
can construct the counterfactual state of the economy
of Jiangxi Province by using the linear combination
of the growth rate of the economy that aren’t
influenced by the transformation of Poyang Lake.
This counterfactual state can be regarded as the
economy that Jiangxi Province hasn’t carried out the
transformation of Poyang Lake.
2.1 Model Settings
y

represent the GDP growth rate of region i at time
t without Poyang Lake reconstruction, and y

represents the GDP growth rate while the
reconstruction being not implemented by region i at
time t. Since y

and y

cannot be observed and
recorded simultaneously, the actual-observed data
y

can be recorded as:


=


+(1−

)

(1)
d

=1 represents the GDP growth rate of the
Poyang Lake reconstruction being implemented by
area i at time t, and d

=0 represents the GDP
growth rate of the Poyang Lake reconstruction not
being implemented by area i at time t.
Generally, i=1 represents Jiangxi Province. From
the first quarter of 2013, Poyang Lake reconstruction
began in Jiangxi Province, labeled as T
+1.
Therefore,


=

, t = 1,..., T
1
(2)

=

, t = T
1
+1, …,T (3)
Since other provinces (or cities) have not
implement the Poyang Lake reconstruction, thus,

=

, i =2,…,N, t = 1,…,T (4)
If y

and y

can be observed simultaneously,
then, at time t, the effect of the Poyang Lake
reconstruction on the economic growth in Jiangxi
Province can be recorded as:

=

−

,t = T
1
+1, …,T (5)
The question is how to estimate

. Because
y

cannot be observed afterT1 +1 , we cannot
calculate

directly.
As discussed above, the economic growth of
provinces (cities) are affected by common factors,
such as the business cycle, technological progress,
etc. Thus, according to Gregory and Head (1999)
and Forni and Reichlin(1998), y

is determined by
a factor model

= b
i
'
f
t
+
+

,i=1,…,N, t=1,…,T (6)
Here ,ft is the unobservable and observable
common factor changing with time in the K×1
dimension; b'i is the coefficient of the 1 ×K
dimension that varies with individuals; b
≠b
,
implies that the influence of common factors on each
economy can be different; α
is the heterogeneity of
individual characteristics; ε

is the random
heterogeneous component of individual i, and
E(ε

)=0
We could rewrite (6) as a vector form:
y
t
0
= Bf
t
+ α + ε
t
, (7)
where y
t
0
=(y

,…,y

)
α=(α
,…,α
)
ε
t
=(ε

,…,ε

)
, and ε
~I(0) E( ε
)=0
E(ε
ε
)=VE(ε
t
f
t
)=0. B=(b
,…,b
)
is a N×K
dimensional factor loading matrix, and ||b
||=c<∞,
Rank(B)=K, allowing the common factor dimension
being less than the number of observed individuals.
There is a very important hypothesis, E(ε

|d

)=0
where j≠i, implying that the heterogeneity of
the individual j is independent of d

. Economically,
it ensures that other province economy will be
independent of the economic changes after the
implementation of the Poyang Lake reconstruction in
Jiangxi Province.
Further, supposing vector a
=(1, −a
) is an
element in the zero-dimensional space of matrix,
namely , a
B =0. Multiply a
for both sides of
equation (7):

=+
y
-1t
+

−
̃
(8)
Economic Effect of Ecological Reconstruction in Poyang Lake
37
Economic Effect of Ecological Reconstruction in Poyang Lake
37
Where α=a
α y
-1t
=(y

,…,y

)
ε
=(ε

,…,ε

)
.There is a clear linear correlation
between ε
and y
-1t
, so we rewrite (8) as:

=+
*
y
-1t
+

(9)
Where a * =a (IN-1 – cov( ε
, y
-1t
)var( y
-1t
)-1),
ε

=a ε
+a cov( ε
, y
-1t
)var( y
-1t
)-1 y
-1t
. Now,
there is no longer linear correlation between ε

and
y
-1t
. According to Hsiao et al. (2012), it means
E(ε

−a
ε
|y
-1t
) is a linear function of y
-1t
, and
E( ε

|y
-1t
) =0 can be guaranteed. The OLS
estimation parameters of (9) keep consistent. The
same discussion about this hypothesis can be found
in Bai et al. (2014) and Ouyang and Peng (2015).
Because other provinces economy are not
affected by the Poyang Lake reconstruction, we
use the GDP growth rates y
-1t
of other provinces to
substitute the common factor ft ,throughout T period,
in order to estimate the situation of Jiangxi Province
without the implement Poyang Lake reconstruction
during the time t.
According to (9), we estimate

as the
following steps: First, before time T
, we conduct be
the regression of y

for y
-1t
=
(y

,…,y

)
following (9), to get parameter
estimatesα
and a
∗
.Then, after the timeT
+1, we
use these parameter estimates and y
-1t
=
(y

,…,y

)
following (9) to construct the GDP
growth rate of Jiangxi Province y

+a
∗
y
-1t
t≥T
+1under the counterfactual situation, and
finally we get

=y

−y

t≥T
+1.Under
normal conditions,

and

keep consistent.
Since there may be serial correlation in

. we
can use Box-Jenkins method to create an ARMA
model:
(
)

=+
()
, (10)
Furthermore, we can use α
(
L
)

μ to express the
long-run effect following Poyang Lake
reconstruction, and use t-test to check whether the
effect is significant.
2.2 Model Selection
Step1: Starting from j=1, select totally j components
from N-1 provinces, C

combinations can be
obtained. Conducting the regression of y

t≤T
for each combination according to (9), so as to select
the combination with the highest R2 and then label it
M(j)
.
Step2: Select a combination from M (1)
M(2)
M(N 1)
that minimizes the AIC or
AICC information criterion1, and label it M(m)
,
where:
p is the number of provinces in the control group and
e is the residual of the OLS regression.
Step3: Using the optimal M(m)
to construct
y

t≥T
+1, we get the GDP growth rate of
Jiangxi province in the counterfactual situation.
Step4: According to

=y

−y

t≥T
+1
, we can get result of the economic effect on Jiangxi
Province following the Poyang Lake reconstruction.
3 EMPIRICAL RESULT
We use 25 other province economy to construct the
counterfactual economic growth rate of Jiangsu
province: Peking, Tientsin, Hebei, Shanxi, Liaoning,
Jilin, Heilongjiang, Shanghai, Jiangsu, Zhejiang,
Anhui, Fujian, Shangdong, Henan, Hubei, Hunan,
Guangdong, Hainan, Chongqing, Sichuan, Guizhou,
Yunnan, Shanxi, Gansu, Qinghai, which are not
influences by the alteration of Poyang Lake. Those
provinces and cities are inextricably linked with
Jiangxi Provinces in the aspect of economy and
geography. The data of quarter nominal GDP and the
CPI is from Ner Statistics Bureau and the period is
from the first quarter(January) in 2005 to the fourth
quarter (December) in 2014.
There are tow ways to calculate the quarter
growth rate of actual GDP, one is on year-on-year
basis the other is on quarter-on-quarter. In this paper,
we adopt the second one. One one hand, it can avoid
the seasonal adjustment. On the other hand ,it is
conducive for us to estimate the long-term effect of
the transformation of Poyang Lake on economy of
Jiangxi Province.
According to the modeling strategy and the
AICC criteria, Shangxi, Hubei, Hunan, Guangdong,
Shanxi, Qinghai, are selected. Then, we can use the
GDP growth rate of these areas to construct the
counterfactual growth rate of GDP in Jiangxi.
The T value in table 1 shows each coefficient is
significant, R2 is 0.97. The coefficient only refers
to the correlation of economic growth determined
by common factors between different countries or
regions, which has no cause and effect significance.
In figure 2, solid line represents the GDP growth
charts of Jiangxi Province from the first quarter in
2006 to the fourth quarter in 2012, the dotted line
represents the GDP growth charts of Jiangxi
Province predicted by regression equation. It can be
seen directly from figure 2, the GDP growth rate
data of selected province and region well fit the
)2(2)ln()(AIC
1
'
1
++= m
T
ee
Tp
2)1(
)3)(2(2
)2(2)ln()(AICC
11
'
1
+
++
+++=
mT
mm
m
T
ee
Tp
ISME 2016 - Information Science and Management Engineering IV
38
ISME 2016 - International Conference on Information System and Management Engineering
38
Jiangxi Province's GDP growth rate. Therefore, in
the forecast period, the provinces' GDP growth rate
can well help to construct counterfactual GDP
growth rate in Jiangxi Province.
Table 1: AICC: Regression equation of optimal control
group
2006:Q1-2012:Q4
Control group Coefficient SD T-Value
Intercept -0.0231 0.0244 -0.9471
Shanxi 0.5684 0.1080 5.2614
Jilin -0.8576 0.1780 -4.8186
Jiangsu 1.1274 0.2514 4.4844
Shandong -0.6234 0.1604 -3.8856
Hubei 0.2390 0.0879 2.7189
Hunan 0.4373 0.1179 3.7103
Guangdong -0.2980 0.1310 -2.2742
Shanxi 0.8294 0.1588 5.2235
Qinhai -0.4434 0.1599 -2.7735
R
2
=0.9675
Figure 1 : Actual path vs predicted path of GDP growth
rate in JiangXi 2006:Q1-2012:Q4
Table 2 compares the differences of real GDP
growth rate and encounter-factual growth rate in
condition of reconstruction of the Poyang lake in
Jiangxi Province and no reconstruction of Poyang
lake in Jiangxi Province. The second line in table is
the real GDP growth rate in Jiangxi Province during
the first quarter in 2006 and the fourth quarter in
2012. The third line in table is the real GDP growth
rate in Jiangxi Province, namely the GDP growth
rate when there is no transformation of Poyang lake
in Jiangxi Province. The fourth line in table is the
transformation effect of Poyang lake, namely the
difference of actual value and the encounter-factual
value.
From the chart 2 , it can be seen that if Jiangxi
Province did not remould Poyang Lake ,its
counterfactual grow rate of average GDP would be
4.10% ,which was lower than the growth rate of
effective GDP,from first quarter in 2013 to fourth
quarter in 2014.The reform of Poyang Lake has
considerablely promoted the economic growth of
Jiangxi ,increasing 3.04%averagely.Comparing with
this situationg , the average rate of Poyang Lake
transformation effet reached 3.04% , during the same
period.
Table 2: AICC: Economic effect of ecological
improvement in Poyang Lake 2013:Q1-2014:Q4
Time Actual
value
Anti-factual
value
Economic
effect
2013Q1 0.0776 0.0419 0.0357
2013Q2 0.0580 0.0479 0.0101
2013Q3 0.0722 -0.0277 0.0999
2013Q4 0.0927 0.0510 0.0417
2014Q1 0.0560 0.0061 0.0499
2014Q2 0.0704 0.0524 0.0180
2014Q3 0.0643 0.0697 -0.0054
2014Q4 0.0802 0.0871 -0.0069
Mean 0.0714 0.0410 0.0304
In the line chart 2,the solid line is the growth rate
of actual GDP after alteration of Poyang Lake in
Jiangxi while the dotted line indicates the trend of
counterfactual growth rate of GDP.The point line
representes the confidence intervals which the rate is
at the siginifcance level of 95%.If the actual path is
not in the confidence intervals ,it means that in the
meaning of statistics,the transformation of Poyang
Lake has notable effect on economy in Jiangxi.
4 CONCLUSION
The construction of the Poyang Lake eco-economic
zone dose not solely focuses on the amelioration of
the ecological environment; its economic effect can
not be ignored, However, so far there are no closely
related researches on it or only one aspect of
economic effects is involved. In this paper, the
economic effect of the ecological environment
construction in Poyang Lake was evaluated
Economic Effect of Ecological Reconstruction in Poyang Lake
39
Economic Effect of Ecological Reconstruction in Poyang Lake
39
Figure 2 AICC: Actual path vs anti-factual path of GDP
growth rate in JiangXi 2013:Q1-2014:Q4
comprehensively employing the counter-factual
policy evaluation approach (Hsiao et al, 2012). With
the study of the regional economic correlation, a
control group was reasonably set, and the economic
growth of Jiangxi Province was modeled under the
condition of no Poyang Lake ecological construction
project. The simulation results and the data of the
actual economic growth in Jiangxi Province were
compared so as to analyze the effect of the
implementation of the ecological transformation
project. Results reveal that the Poyang Lake
eco-transformation Project exerts great influence on
the economy of Jiangxi Province. From the first
quarter of 2013 to the fourth quarter of 2014, the
economic growth rate of Jiangxi Province increased
by 3 percentage points on quarter-on-quarter basis,
with the economic effect being remarkable. The
analyses of our results show that the Poyang Lake
ecological construction project not only improved
the Jiangxi ecological environment, but also
produced significant economic effects. On the one
hand, the economic effect could be attributed to the
increase of the government expenditure. On the
other, the improvement of ecological environment
will also enhance the utility of the consumers and the
income of the producers, resulting in "indirect
market value". Over the past few decades, China's
economy has achieved great success, but in many
other respects, challenges remain, in particular,
environmental and ecological problems. On the way
to further development, we must change the mode of
economic development, and make efforts to realize
the coordinated development of ecological
environment and the sustainable economic growth.
REFERENCES
Bender, B., Gronberg, T. J., & Hwang, H. S. (1980).
Choice of functional form and the demand for air
quality. Review of Economics & Statistics, 62(4),
638-43.
Boqiang Lin and Xiying Liu, (2010), Carbon emission in
China's urbanization stage: influencing factors and
reduction strategies, Economic Research, 08, .66-78
Brucato, P. F., Murdoch, J. C., & Thayer, M. A. (1990).
Urban air quality improvements: a comparison of
aggregate health and welfare benefits to hedonic price
differentials. Journal of Environmental Management,
30(3), 265-279.
Chay, K. Y., & Greenstone, M. (2004). Does air quality
matter? evidence from the housing market. Social
Science Electronic Publishing, 113(2), 376-424.
Cheng, H., Ching, H. S., & Shui, K. W. (2012). A panel
data approach for program evaluation: measuring the
benefits of political and economic integration of hong
kong with mainland china. Journal of Applied
Econometrics, 27(5), 705–740.
Guoying Deng, Shu Xu and Shaoyang Zhao, (2012),
Economic value of Environmental Governance:
evidence from the real estate market, World Economy,
09, 143-160.
Yongwei Chen and Lizhong Chen, (2012),“Pricing for
clean air: Empirical Evidence from Qingdao”, World
Economy, 04, 140-160.
ISME 2016 - Information Science and Management Engineering IV
40
ISME 2016 - International Conference on Information System and Management Engineering
40