The Impact of Sino-US Clean Energy Trade Complementarity on

China’s Clean Energy Consumption

Yaowu Dong, Ruoxi Hu and Denghui Duan

Guizhou University of Commerce, Guiyang, China 550014

Keywords: Energy Structure, Markov Chain.

Abstract.

This paper analyzes the dynamic relationship between Sino--US clean energy trade complementary index and

China’s clean energy consumption by using autoregressive distributed lagged model (ARDL) based on the

sample data from 1992 to 2017. The results show that there is a long-term cointegration relationship among

Sino—US clean energy trade complementary index, China’s clean energy consumption, economic growth,

and energy intensity. In the long run, the complementarity of clean energy trade between China and the United

States has a negative impact on the growth of China’s clean energy consumption, proves that the

complementary relationship between U.S. clean energy export growth and China’s import growth is only

reflected in theory, and there is no actual trade complementarity ,reflected in the trade between the two

countries. In the long run, the increase of per capita GDP and the decrease of energy intensity will be

accompanied by the increasing clean energy in China, which indicates that China is in the process of energy

structure optimization, and with the economic growth, the proportion of clean energy utilization is also

increasing. Besides, in the short term, the industrial structure has a positive impact on China’s clean energy

consumption, and clean energy consumption in the first lag stage also promotes the clean energy consumption

in the current period, while other factors do not significantly contribute to clean energy consumption in China.

Finally, on the basis of empirical analysis, the corresponding countermeasures and suggestions are put

forward.

1 INTRODUCTION

With the change of global climate and the destruction

of human living environment caused by the massive

use of traditional energy, global warming and

environmental pollution have become difficult

problems that all countries need to deal with together,

and the countries began to pay attention to the

development and utilization of clean energy. As the

world’s second largest economy, China is facing

internal and external pressure in the fields of climate

and environmental improvement for its annual energy

consumption and CO2 emissions are ranking first in

the world. The energy consumption in China—the

biggest increase among all countries in more than a

decade—rose by nearly half from 2007 to

2017(BP,2018). In 2017, the energy consumption of

coal and oil were top 2 of the world, which accounted

for 60.4 % and 18.8 % respectively in China.

Proportion of coal consumption is relatively high, and

put great pressure on energy conservation and

emission reduction. Based on that point, the Chinese

government has pledged to accelerate the use of clean

energy, which will reach to 35 % of total energy

consumption in China by 2030. Compared with

China, the United States is rich in clean energy

resources and has advantages in both experience and

technology in developing clean energy, while China

has just focused on strengthening international trade

cooperation and clean energy technologies in recent

years.

According to the economic development of

various countries, there is a correlation relationship

between the economic growth and the energy

consumption(Yemane, 2004; DolgpolovalI, et

al,2014; Ergin& Simbarashe, 2019) . Statistically,

there is a great difference in thermal efficiency among

different energy resources. The thermal efficiency of

natural gas is above 75%, that of oil is about 65% and

coal is 40%-60%. Therefore, based onthe energy

consumption structure of a country, the energy

efficiency will be low if the proportion of traditional

energy consumption (such as coal)is high (Meng&

Zhou,

2014). As inefficient energy consumption

Dong, Y., Hu, R. and Duan, D.

The Impact of Sino-US Clean Energy Trade Complementarity on China’s Clean Energy Consumption.

DOI: 10.5220/0011104000003355

In Proceedings of the 1st International Joint Conference on Energy and Environmental Engineering (CoEEE 2021), pages 10-17

ISBN: 978-989-758-599-9

Figure 1. Annual Primary Energy Consumption in Mtoe for the US and China from 2007 to 2017.

drives economic growth, it will inevitably generate

large amounts of carbon emissions and cause

pollution to the environment.The pollution, in

turn,will dampen down economic growth(Tiba&

Omri ， 2017). Technological innovation plays a

significant role in promoting energy efficiency both

in the short and the long term(Xiongfeng, et al, 2019).

As the traditional energy, like coal, is gradually

replaced by clean energy, and energy technology is

improved, the energy consumption structure will be

optimized and the energy efficiency will be

improved. It means that ifeconomic

growthaccompanied by improvements in energy

efficiency, it will lead to a slowdown or reduction in

total energy consumption.Therefore, improving

energy efficiency, changing the structure of energy

consumption and reducing the intensity of energy

consumption are keys to reduce CO2 emissions,

which should be important means used to promote the

development of China's low-carbon economy in the

future (Wu&Zeng, 2013; Xiaojun ，et al, 2019). It

suggests that advanced energy technologies and the

increase of clean energy consumption enhance the

energy efficiency. In this case, economic growth does

not necessarily enlarge energy consumption. Figure 1

illustrates the annual primary energy consumption for

the US and China from 2007 to 2017. Since 2007, the

consumption of China has risen continuously and

there is a great increase during the succeeding years.

Meanwhile, the trend of America’s energy declined

slightly and nearly remained stable in recent years. It

is clear that the growth of the US economy has not

been accompanied by an increase in total energy

demand, which has been driven by improvements in

energy efficiency. By contrast, China's energy

efficiency is still relatively low.

For a long time, there has been a huge trade deficit

between the United States and China. The clean

energy cooperation between the United States and

China enables the United States to take advantage of

its energy and technology, increase its exports of

clean energy products, technologies and equipment to

China, and reduce the trade deficit. In the cooperation

of clean energy, if the United States can appropriately

reduce the export restrictions onhigh-tech products to

China, China will also moderately widen the market

in the investment field, the cooperation between the

two sides will surely promote win-win situation under

the trade balance. For example, China has been faced

with insufficient natural gas supply every winter in

recent years, while the United States is the largest

producer of natural gas. Therefore, increasing the

trade of natural gas between China and the United

States can help solve the problem of natural gas

shortage in China.

In order to analyze the impact of Sino-U.S. clean

energy trade cooperation on China’s clean energy

consumption, this paper will take Sino-U.S. clean

energy trade complementary index (TCI) , China’s

clean energy consumption ratio (CE) , energy

intensity (TEG) , per capita output (PG) and industrial

structure (STRU) as the objects, construct an ARDL

model, analyze the dynamic relationship among

them, and try to put forward the corresponding

recommendations.

2 METHODS AND DATA

2.1 Methods

In this section, we develop an econometric model to

estimate the dynamic relation betweenSino—U.S.

clean energy trade complementarity and China’s

clean energy consumption. We know that the premise

of foreign trade is that the export products have

comparative advantages, such as relatively rich

resources, high-quality products and technology. If a

country’s products have a certain export

competitiveness, there should be a potential

cooperation with other countries. Where there is a

certain trade potential, there is a potential economic

1000

2000

3000

4000

2007

2008

2009

2010

2011

2012

2013

2014

2015

2016

2017

U.S

CHINA

The Impact of Sino-US Clean Energy Trade Complementarity on China’s Clean Energy Consumption

effect. Trade Completeness Index (TC) in this paper

measures the complementarity of clean energy trade

between China and the United States. The TC

between countries k and j is defined as(WTO,2017):

= 100(1 – sum(|m

– x

| / 2))

where xij is the share of good i in global exports of

country j and mik is the share of good i in all imports

of country k. The index is zero when no goods are

exported by one country or imported by the other and

100 when the export and import shares exactly match.

In order to reflect the influence of the Sino-U.S.

clean energy trade complementarityon China’s clean

energy consumption, this paper adopts the proportion

of clean energy consumption in total energy

consumption (CE) to show the change of China’s

clean energy consumption.

At the same time, considering other factors in

clean energy consumption, this paper introduces the

per capita output (PG), industrial structure (STRU)

and energy intensity (TEG) as control variables. PG

is the real GDP per capita, STRU is the ratio of

industrial output to GDP, and TEG is the ratio of total

energy consumption to GDP. In view of the empirical

strategies provided by theprior literatures (Bas and

Ledezma, 2010; Bustos, 2011), the empirical

framework can take thefollowing form.

C𝐸

𝑡

= (C𝐸

𝑡

,TC

𝑡

,STRU

𝑡

,TEG

𝑡,

,𝜀

𝑡

) (1)

As that time series are always non-stationary, it is

effective to specify the econometric model following

the process of the data generation. Results in table (2)

show the dependent variable is not of integration, so

the equation (1) can be represented by a structural

ARDL model. The ARDL model is a standard least

squares model. The ARDL model is to determine

whether there is cointegration relationship between

variables through the boundary test method, and to

estimate the correlation Coefficient between

variables on the basis of this. First proposed by

Charemza&Deadman(Charemza&Deadman, 1992),

then perfected and extended by Pesaran et al.(Pesaran

et al, 2001). Different from the traditional co-

integration test model, the ARDL model does not

require the same order single integer when testing the

long-term relationship among variables.Even under

small samples, the estimations of ARDL model

would be stable enough. The ARDL model takes the

form shown as follows (2):

C𝐸

𝑡

=α

∑









C𝐸

𝑡-i

∑









𝑡-i

∑









STRU

𝑡-

∑









TEG

𝑡-i

∑









𝑡-i

+𝜀

𝑡

(2)

Furthermore, the unit root test also imply that all

the explanatory variables are not of integration higher

than I (1), thus explanatory variables might be co-

integrated with the dependent variable. Then, the

equation (1) can fit a structural ARDL model in error-

correction form as follows.

∆C𝐸

𝑡

=α

+α

C𝐸

𝑡-1

+α

𝑡-1

+α

STRU

𝑡-1

+α

TEG

𝑡-

+α

𝑡-1

∑

𝛾









∆C𝐸

𝑡-i

∑

𝛾









∆TC

𝑡-i

∑

𝛾









∆STRU

𝑡-i

∑

𝛾









∆TEG

𝑡-i

∑

𝛾









∆PG

𝑡-i

+𝜀

𝑡

(3)

where α



is the long-term Correlation Coefficient of

the variable, 𝛾



is the short-term Correlation

Coefficient of the variable, 𝑞



is the maximum lag

order of the model, and 𝑞



is the White noise of the

normal distribution.

2.2 Data

Based on the classification of HS codes, this paper

selects the representative clean energy trade products

of the United States and China from 1992 to 2017 as

the research sample, including Solar Energy, wind

energy, biomass energy, water energy, natural gas,

nuclear power, and selected both import and export

data from the UN comtrade Database. Other data

involved in the study, such as China’s GDP per

capita, industrial structure, and clean energy

consumption, are collectedfrom China Energy

Statistical Yearbook and China National Bureau of

Statistics.

Among them, the per capita real GDP (PG) is

calculated by the China's real GDP (which is adjusted

by price deflator as the base year of 1978) divided by

China's total population at the end of the year; The

value of Energy consumption intensity (TEG) is

calculated by dividing the total energy

consumptionby the real GDP. Table 1 shows the

descriptive data of the values above after natural

logarithm. Figure 2 shows the trend of CE and TC. It

can be seen that CE has been rising steadily since

1992, but clearly there is still room to achieve the goal

of 35% of total consumption by 2030; At the same

time, TC has been in a high position for many years,

which shows that there is a strong complementarity in

clean energy trade between China and America in

theory. However, the dynamic relationship between

TC and CE needs to be further tested by the ARDL

model.

CoEEE 2021 - International Joint Conference on Energy and Environmental Engineering

Table 1. Descriptive Statistics.

LNCE LNPG LNSTRU LNTC LNTEG

Mean 2.388465 -0.346857 3.669433 4.087292 0.928757

Maximum 3.034953 0.655577 3.738473 4.278937 1.389892

Minimum 1.916923 -1.455002 3.511376 3.728934 0.520496

Std. Dev. 0.320811 0.658869 0.062195 0.136925 0.223413

Figure 2. General Trends for CE and TC (1992-2017).

3 RESULTS AND DISCUSSION

3.1 Result of Unit Root Test

In this paper, the variables above are tested by the

Augmented dickey-fuller test statistics. The test

results are shown in Table 2. It can be seen from Table

2 that the variable LNPG is stationary and the

variables LNCE, LNSTRU, LNTC and LNTEG are

non-stationary, but they are stationary after the first-

order difference. Therefore, the variables used in this

paper do not exceed I(1), and the cointegration

relationship among variables can be further analyzed

by using ARDL model.

Table 2. Unit Root Tests.

Var ia bl es

Augmented Dickey-

Fuller

LNCE

Level 0.512

First Difference -4.582***

LNPG

Level

-2.164（drift）**

First Difference

LNSTRU

Level 0.502

First Difference -3.812***

LNTC

Level -2.672*

First Difference -7.162***

LNTEG

Level -1.063

First Difference

-1.771（drift）**

3.2 Cointegration Analysis

In this paper, the ARDL model is tested by eviews10,

and the cointegration relationship between variables

is tested by calculating the value of the corresponding

F-statistic. The lag order is determined by AIC and

SC information criterion. The cointegration test

results are shown in Table 3. Referring to the research

of Pesaran,et al.[13], Table 3 illustrates the F-statistic

values of I (0) and I (1) processes at the significance

levels of 1%, 5% and 10% respectively through

ARDL-Bounds Test. The test results show that when

CE is the explained variable, there is a cointegration

relationship among variables at the significance level

of 1%. The optimal model is determined as ARDL

(2,0,2,2,2). The estimations of ARDL model are

The Impact of Sino-US Clean Energy Trade Complementarity on China’s Clean Energy Consumption

shown in Table 3, and conditional error correction

regressionand the coefficient of long-term

equilibrium are shown in Table 4.

From Table 4, it can be found that in the long term,

the impact coefficient of LNTC on LNCE is -

0.652176 at the significant level of 10%. It indicates

that Sino--US clean energy trade complementarity

has negative impact on the growth of China’s clean

energy consumption, which is not conducive to the

optimization of China energy consumption structure.

This proves that although China’s clean energy

imports and U.S. clean energy exports have formed a

synchronous growth trend, there is rare relationship

between the growth of China’s import of clean energy

and theUS’clean energy export.In other words, the

growth of China’s clean energy imports is not driven

by the growth of clean energy exports of the United

States. By comparing Figure 3 and Figure 4, it can be

seen that the overall China’s clean energy import

shows an upward trend, and the import of clean

energy has accelerate since 2010; however, after

2011, the trend of clean energy export of China is

opposite to the overall clean energy export of the

United States. The total export of the United States

has steadily increased, while the export to China has

begun to slow down. This shows that the growth of

U.S. clean energy export and that of China’s clean

energy import are not complementary, which is not

reflected in the trade between the two countries. From

table 4, the coefficient of LNTEG is also significantly

negative(-1.835918). It shows that the change of

energy intensity and the change of the proportion of

clean energy consumption are reverse. To some

extent, it is related to China’s strengthening pollution

control, limiting CO2 emissionsand increasing the

use of clean energy. As a result, while the energy

consumption per unit GDP decrease, the proportion

of clean energy increase. The coefficient of LNPG is

significantly positive(1.5614.6), that is, the increase

of per capita GDP is in direct proportion to the growth

of clean energy consumption. Furthermore, it shows

that with China’s economic growth, the use of clean

energy is also increasing, and the energy consumption

structure is constantly optimized.

Table 3. ARDL Estimation (dependent variable LNCE).

Variable Coefficient Std.Error t-Statistic Prob.*

LNCE(-1) 0.146732 0.227824 0.644059 0.534

LNCE(-2) -0.365591 0.228609 -1.599199 0.1409

LNPG 1.903133 0.778937 2.443244 0.0347

LNSTRU 0.22887 0.828776 0.276154 0.7881

LNSTRU(-1) 0.050416 1.030798 0.048909 0.962

LNSTRU(-2) -1.305904 0.740789 -1.762855 0.1084

LNTC -0.251401 0.133551 -1.882435 0.0892

LNTC(-1) -0.291565 0.149303 -1.95284 0.0794

LNTC(-2) -0.251944 0.138546 -1.818491 0.099

LNTEG -2.262761 0.863481 -2.620512 0.0256

LNTEG(-1) 1.714517 1.181713 1.450874 0.1774

LNTEG(-2) -1.689481 0.825431 -2.046787 0.0679

C 14.77916 4.031182 3.666209 0.0043

@TREND -0.167481 0.075814 -2.209114 0.0516

S.E. of regression 0.045135 Akaike info criterion -3.06714

Sum squared resid 0.020371 Schwarz criterion -2.379942

ARDL-BoundsTest Value Signif. I(0) I(1)

F-statistic 5.787697 10% 3.03 4.06

5% 3.47 4.57

1% 4.4 5.72

Table 4. Long-term relationship coefficient of ARDL model.

Variable Coefficient Std. Erro

t-Statistic Prob.

LNPG 1.561406 0.652661 2.392370 0.0378

LNSTRU -0.842278 0.577513 -1.458457 0.1754

LNTC -0.652176 0.294462 -2.214809 0.0511

LNTEG -1.835918 0.667798 -2.749212 0.0205

CoEEE 2021 - International Joint Conference on Energy and Environmental Engineering

Figure 3. Clean energy imports and exports for US and China (1992-2017).

Figure 4. General Trends for bilateral trades between US and China (1992-2017).

After the long-term relationship coefficient of the

cointegration equation is estimated, error correction

model (ECM) regression based on ARDL model can

be further established toanalyze the dynamic

relationship between the short-term fluctuation of

independent variables and dependent variable . The

results are shown in table 5. The coefficients of

D(LNCE(-1)) and D(LNSTRU(-1)) are significantly

positive, indicating that clean energy consumption and

the industrial structure in the first lag stage have a

positive impact on the increase of the clean energy

consumption in the current period. D(LNTC)

coefficient is significantly negative, while D(LNTC(-

1)) coefficient is significantly positive, which

indicates that although Sino--US clean energy trade

complementary index is unfavorable to the increase of

China’s clean energy consumption in the current

period, it has a positive effect in the first lag stage. The

coefficient of D(LNTEG) is significantly negative,

while the coefficient of D(LNTEG(-1)) is significantly

positive. Generally speaking, the impact of Sino--US

clean energy trade complementary index and energy

intensity on China’s clean energy consumption in the

short term is limited.The coefficient of the error

correction term CointEq(-1) is significantly negative

at the significance level of 1%, which conforms to the

negative feedback mechanism, that is, when the short-

term change is out of the long-term equilibrium level,

it can be adjusted to the long-term equilibrium state

through the error correction model.

Table 5. ECM regression.

Variable Coefficient Std. Erro

t-Statistic Prob.

C 14.77916 2.324173 6.358890 0.0001

@TREND -0.167481 0.026788 -6.252120 0.0001

D(LNCE(-1)) 0.365591 0.158065 2.312908 0.0433

D(LNSTRU) 0.228870 0.550510 0.415741 0.6864

(

LNSTRU

(

-1

))

1.305904 0.563589 2.317122 0.0430

(

LNTEG

)

-2.262761 0.474407 -4.769662 0.0008

(

LNTEG

(

-1

))

1.689481 0.514627 3.282920 0.0082

D(LNTC) -0.251401 0.087067 -2.887450 0.0162

D(LNTC(-1)) 0.251944 0.096313 2.615903 0.0258

CointE

-1

)

* -1.218858 0.191492 -6.365051 0.0001

The Impact of Sino-US Clean Energy Trade Complementarity on China’s Clean Energy Consumption

4 CONCLUSION AND

SUGGESTION

Based on the annual data from 1992 to 2017, this

paper studies the relationship between Sino—U.S.

clean energy trade complementarity and China’s

clean energy consumption. The results show that

there is a long-term co-integration relationship among

Sino—US clean energy trade complementary index,

China’s clean energy consumption, economic

growth, and energy intensity. In the long run, the

complementarity of clean energy trade between

China and the United States has a negative impact on

the growth of China’s clean energy consumption,

which is not conducive to the optimization of energy

consumption structure. This proves that the

complementary relationship between U.S. clean

energy export growth and China’s import growth is

only reflected in theory, and there is no actual trade

complementarity, which is not reflected in the trade

between the two countries. In the long run, the

increase of per capita GDP and the decrease of energy

intensity will both boost clean energy consumption

and optimize the energy consumption structurein

China. Besides, in the short term, the industrial

structure has a positive impact onclean energy

consumption, and clean energy consumption in the

first lag stage also promotes the consumption of clean

energy in the current period, while other factors do

not significantly contribute to clean energy

consumption in China.

In recent years, China has formulated policies to

save energy and reduce emissions, and to increase the

proportion of clean energy. At present, China should

speed up the adjustment of industrial structure, the

development and application of clean energy

technology, and improve the awareness of energy

conservation of enterprises and residents. From the

perspective of clean energy development goals and

policy guidance of China and the United States, both

China and the United States attach importance to the

clean energyconsumption. The United States has

advancedexperience and technology in clean energy,

and it needs to expand new markets, while China is

just in the early stage of development and has great

market demand. Obviously, bilateral trade

cooperation is beneficial to both countries.

Theoretically, there is a strong complementarity, and

there is a large space for win-win cooperation in

practice. At present, China and the United States

should reach a consensuson the intellectual property

system and related legal system as soon as possible.

In terms of patent application and protection, it is

necessary to jointly establish transparent

management measures and valid examination system

to ensure the standardization of application and

examination, strengthen cooperation and

communication between China and the United States,

so as to crack down on cross-border intellectual

property violations and crimes, and maintain the

bilateral trade order. In the process of specific

implementation, the United States should gradually

reduce the technical restriction, and China should also

accelerate the liberalization of market access. Only in

this way will be benifit to the clean energy trade

cooperation which is caused by different division of

labor in the industrial chain, that is, the high-end

products of the United States enter the Chinese

market, while the medium and low-end products of

China enter the American market. We should

improve the mechanism of capital access and exit,

ensure the legalization and transparency of capital

investment, and protect the legitimate rights and

interests of enterprises. In addition, we should

establish an effective bilateral communication and

coordination mechanism, actively listen to the

opinions involved in trade divergence of enterprises,

and strengthen bilateral communication and

consultation when formulating relevant trade

policies, so that we can solve the problems existing in

Sino--US cooperation in a timely manner, and

promote win-win cooperation between the two sides.

REFERENCES

BP, Statistical Review of Word Energy, 2018

Yemane W R., 2004. Disaggregated industrial energy

consumption and GDP: the case of Shanghai. Energy

economics, 26(2), 9-75.

DolgpolovalI, et al. 2014. Energy consumption and

economic growth: evidence from non-OPEC oil

producing states. Quality & quantity, 48(2), 887-898.

ErginA ， Simbarashe H., 2019. Carbon

emissions, energy use, real GDP per capita and trade

matrix in the Indian economy-an ARDL

approach.Energy, 168, 1081-1093.

Meng F Y, ZhouP., 2014. Ineffciency and Congestion

Assessment of Mix Energy Consumption in 16 APEC

Countries by Using DEA Window Analysis. Energy

Procedia,61 ( 36), 2518-2523．

TibaS ， Omri A., 2017, Literature survey on the

relationships between energy

，

environment and

economic growth. Renewable and Sustainable Energy

Reviews, 69, 1129-1146.

Chen P Y， et al., 2016. Modeling the global

relationships among economic growth, energy

consumption and CO2 emissions. Renewable &

Sustainable Energy Reviews, 65, 420-431.

CoEEE 2021 - International Joint Conference on Energy and Environmental Engineering

Xiongfeng P，et al., 2019. Dynamic relationship among

environmental regulation, technological innovation

and energy efficiency based on large scale provincial

panel data in China. Technological Forecasting and

Social Change, 2019, 144, 428-435.

Wu L Y，Zeng W H., 2013. Research on the Contribution

of Structure Adjustment on SO2 Emissions Reduction-

Case Study of Shijingshan District Beijing. Procedia

Environmental Sciences, 18( 114), 849-855.

Xiaojun M，et al., 2019. Carbon emissions from energy

consumption in China: Its measurement and driving

factors. Science of the Total, 648, 1411-1420.

WTO. (2017, 19 June). WTO. Retrieved from Trade

Indicators:http://wits.worldbank.org/wits/wits/witshelp

/Content/Utilities/e1.trade_indicators.htm.

Charemza W W, Deadman D F., 1992. New Directions in

econometric practice: General to specific modelling,

cointegration and vector autoregression. Cheltenham:

Edward Elgar,.

Pesaran M H，Shin Y, Smith R J.， 2001. Bounds testing

approaches to the analysis of level relationships.

Journal of Applied Econometrics, 16(3), 289-326.

The Impact of Sino-US Clean Energy Trade Complementarity on China’s Clean Energy Consumption