Sustainable Production and Consumption 33 (2022) 28–37

Contents lists available at ScienceDirect

Sustainable Production and Consumption

journal homepage: www.elsevier.com/locate/spc

Analysis of emission reduction effects of carbon trading: Market

mechanism or government intervention?
Boqiang Lin ⁎, Chenchen Huang
School of Management, China Institute for Studies in Energy Policy, Collaborative Innovation Center for Energy Economics and Energy Policy, Xiamen University, Fujian, 361005, China

a r t i c l e i n f o a b s t r a c t

Article history: Carbon trading is a crucial policy to combat climate change and China's carbon market is considered one of the
Received 29 April 2022 carbon markets to consider. Some studies have shown that China's carbon trading scheme promotes carbon re-
Received in revised form 18 June 2022 duction. However, others have shown that low carbon prices and poor liquidity characterize China's carbon mar-
Accepted 20 June 2022
ket. How can China's carbon trading mechanism achieve the expected emission reduction when the market
Available online 24 June 2022
mechanism has not been fully established? To answer it, this article collected 30 Chinese provinces' panel data
Editor: Prof. Yue-Jun Zhang from 2005 to 2019 and explored the relationship between the implementation of carbon trading policies and car-
bon emission reduction through a multi-period difference-in-differences model. In addition, a series of robust-
ness tests were performed. The results indicate that carbon trading does effectively curb carbon emissions. But
Keywords: this policy effect is not achieved through the market mechanism. In the pilot phase of China's carbon market, gov-
Carbon trading ernment intervention played a significant role in reducing carbon emissions. The results also suggest that the car-
Carbon market bon trading mechanism promotes carbon emission reduction through energy consumption rather than industrial
Government intervention structure. This study enriches the empirical evidence of carbon trading and provides valuable references for pol-
Multi-period difference in difference model
icymakers and the construction of carbon markets.
Carbon emission reduction
© 2022 Institution of Chemical Engineers. Published by Elsevier Ltd. All rights reserved.

1. Introduction Therefore, research on the effectiveness of the carbon market in the

pilot stage would help understand the market mechanism and explore
What is carbon trading? Carbon trading, also known as the carbon its deficiencies. It provides a valuable reference for the perfection and
emission trading scheme (ETS), is an important policy tool to promote further development of China's national carbon market.
carbon emission reduction by using a market mechanism. Thus, it's a Recently, carbon trading has gradually heated up and aroused a hot
place where carbon emission rights are traded. Carbon trading is playing debate. Existing studies have discussed the relationship between carbon
an increasingly important role in addressing climate change and achiev- trading and carbon emissions, carbon intensity, technological innova-
ing carbon neutrality and has been widely concerned and recognized by tion, etc. (Golpîra and Javanmardan, 2022; Goulder et al., 2022; Zhang
the international community. At present, several carbon trading mar- et al., 2020b). Among them, the extent to which carbon trading pro-
kets have been established globally, including the European Union, motes carbon emission reduction has attracted attention. Theoretically
California of the United States, and New Zealand. In 2011, China speaking, for the subject of emission control, if an enterprise's emissions
approved carbon trading trials in seven provinces and cities, including exceed the carbon quota allocated to the enterprise, the enterprise is ex-
Beijing, Tianjin, Shanghai, Chongqing, Hubei, Guangdong, and pected to purchase carbon emissions right in the carbon market. Other-
Shenzhen. Since 2013, seven pilot carbon markets have been open for wise, the enterprise will be subjected to administrative punishment. If
trading. On July 16, 2021, China's national carbon emission trading mar- the emission is lower than the carbon quota, the enterprise can sell
ket opened, which is regarded as the largest carbon market. As an the excess carbon quota in the carbon market for profit gains. If the car-
emerging energy financial market, the carbon market is expected to bon market is mature enough, enterprises will face high emission costs.
realize carbon emission reduction effectively. However, the national Enterprises with relatively low emission reduction costs and economic
carbon market is still in its infancy and faces many problems in practice. incentives will take the lead in emission reduction, and sell the excess
carbon emission rights to enterprises with relatively high emission re-
duction costs and gain additional income (Guo et al., 2020). Therefore,
⁎ Corresponding author at: School of Management, China Institute for Studies in Energy
Policy, Collaborative Innovation Center for Energy Economics and Energy Policy, Xiamen
carbon trading can promote green technology innovation in enterprises,
University, Fujian, 361005, China. eliminate high-polluting enterprises, promote industrial transformation
E-mail addresses: [email protected], [email protected] (B. Lin). and reduce carbon emissions (Yao et al., 2021; Zhang et al., 2021). Some

https://doi.org/10.1016/j.spc.2022.06.016
2352-5509/© 2022 Institution of Chemical Engineers. Published by Elsevier Ltd. All rights reserved.
B. Lin and C. Huang Sustainable Production and Consumption 33 (2022) 28–37

existing studies support China's carbon trading encourages carbon re- carbon emission reduction will be explored from two aspects of energy
duction (X. Yu et al., 2021; Zhang et al., 2020a). According to Zhang consumption structure and industrial structure.
et al. (2020a), carbon trading can significantly reduce carbon emissions Overall, this article makes three contributions.
from China's industrial sectors. W. Zhang et al. (2020) proved that car- First, this study finds that the reduction effect of China's carbon mar-
bon trading significantly reduced industrial carbon dioxide emissions ket is achieved through government intervention rather than a market
in carbon pilot areas through the difference-in-differences (DID) mechanism. Most of the existing studies only focus on whether carbon
model. Cao et al. (2021) pointed out that coal consumption of regulated trading achieves a carbon emission reduction effect. Besides exploring
coal-fired power plants in ETS pilot areas was significantly reduced, the carbon emission reduction effect, this paper further analyzes the
proving the effectiveness of carbon trading in reducing carbon emis- reasons why the carbon emission reduction effect still exists under the
sions from the power sector. Y. Hu et al. (2020) showed that compared condition of an immature market. The results show that China's pilot
with non-pilot areas, carbon dioxide emissions in pilot areas decreased carbon market has indeed achieved carbon emission reduction under
by 15.5 %. They believed that ETS achieved significant energy saving and the premise of the immature carbon market. However, this policy effect
emission reduction effects. is achieved through government intervention rather than the market
But the issue is controversial. The effectiveness of pilot carbon mar- mechanism. This finding provides new evidence and a new perspective
kets seems inadequate. According to the actual performance of China's for related carbon trading research from the macro level. Under the
carbon market, the pilot carbon market is characterized by low carbon background of China's dual carbon goal, this paper provides a reference
prices, low activity, and low liquidity. These facts cast doubt on the prac- for accelerating the construction of China's national carbon market and
tical effects of carbon trading. Wen et al. (2020) believe that carbon giving full play to the role of the carbon trading mechanism, which is
trading has little impact on industrial carbon emissions. The reasons helpful to the realization of China's dual carbon goal.
may be due to the inadequate environmental supervision of local gov- Second, unlike previous studies, this paper adopts the multi-period
ernments and unreasonable allocation of carbon quotas. Lyu et al. DID model for analysis, which is more consistent with the actual situa-
(2020) insisted that China's carbon market was not active and had de- tion of China's carbon trading pilot construction. Most studies applied
fects in the stability of volatility. Zhang et al. (2019) constructed an eval- the two-stage DID model to evaluate the policy effect. The year 2011
uation index system of carbon market maturity, and the research results proposed by the carbon trading pilot policy or the year when launched
suggested that China's carbon market was far from mature and practi- the first pilot trading, was adopted as the policy implementation year in
cal. S. Zhang et al. (2020) believed that most pilot carbon markets did almost all relevant studies. However, none of the pilot projects started
not reach the weak efficient market in any period. Fan et al. (2019) trading in 2011, and China's carbon trading pilot projects started trading
pointed to the widespread inefficiencies in the pilot markets. Thus, if phase by phase. This paper enriches the application of DID method in
the market is immature, a suitable price mechanism is not formed. the study of the carbon trading mechanism, to provide more scientific
Then the carbon market participants will not have enough incentive to results.
implement carbon emission reduction. A lower carbon price means a Third, the study provides valuable references for policymakers and
lower cost for those who emit excess carbon. Carbon trading is not carbon market construction. Combined with provincial panel data
enough to change their emissions behavior (Lin and Jia, 2020). At the from 2005 to 2019, this study explored the relationship between carbon
same time, it seems that reducing emissions cannot gain much profit. trading and carbon emissions through a multi-period DID model, and
Hence, carbon trading cannot motivate emission control subjects to re- analyzed the transmission mechanism. The results passed a series of ro-
duce carbon emissions through innovation and green technology. bustness tests. Finally, some targeted policy suggestions are put for-
Based on the analysis of existing literature, the relationship between ward.
carbon trading and carbon emissions remains to be clarified. Hence, we The remaining structure of this paper is as follows: Section 2 is a lit-
test if carbon trading can help achieve the desired effect of carbon re- erature review; Section 3 introduces methods, research ideas, and data
duction, and in instances, market mechanisms fail to operate effectively, sources. Section 4 analyses and discusses empirical analysis; the last
what drives this reduction. Existing research does not seem to have a section is the conclusion and policy recommendations.
good answer to this question. This study aims to answer the above ques-
tions based on the background that the excellent price mechanism of
the carbon market has not been formed in the pilot stage. 2. Literature review
Considering the dominant position of government intervention in
China's traditional environmental regulation and the fact that most of The question to be studied in this paper is whether China's emission
the enterprises covered by the carbon trading pilot are state-owned en- trading mechanism effectively reduces carbon emissions. The impact of
terprises (J. Hu et al., 2020). This paper proposes a reasonable guess: ETS on carbon emissions has attracted a great deal of attention, with
market mechanism and government intervention synergistically pro- most studies presenting ideas and evidence that ETS contributes to car-
mote carbon emission reduction. The subject of emission control can re- bon reduction.
alize the benefits through the carbon market and, at the same time, At the provincial level, several studies provide evidence. Dong et al.
fulfill the obligations and avoid punishment. Therefore, this paper raises (2019) used the DID model to prove that ETS could achieve a significant
two research questions: does the pilot phase of carbon trading achieve carbon emission reduction effect in both the short and long term. Chen
the expected carbon emission reduction effect? Is the emission reduc- et al. (2020) suggest that ETS can reduce carbon emissions by 13.39 %. Qi
tion effect of carbon trading produced through the synergistic effect of et al. (2021) applied DID model to carry out the empirical study, and the
market mechanism and government intervention? results showed that ETS significantly reduced carbon emissions without
To answer the above questions, this paper will use a multi-period sacrificing economic development. Zhang and Zhang (2019) also ob-
DID model to analyze the carbon emission reduction effect generated tained the result that ETS helped promote the development of a low-
by implementing a carbon trading policy. In this paper, panel data of carbon economy through the DID model. Yi et al. (2020) applied the
30 provinces and cities in China from 2005 to 2019 will be used for em- DID model to analyze the emission reduction effect of carbon trading
pirical analysis, and a series of robustness tests will be conducted. What pilot projects, and the results demonstrate that ETS was effective in
is different from the existing research is that this paper deeply discusses some pilot projects. According to Dong et al. (2020), the emission reduc-
the mechanism behind China's ETS to achieve carbon emission reduc- tion effect of ETS is gradually enhanced. Wang et al. (2021) used the DID
tion, and not only proves the existence of the carbon emission reduction to prove that ETS achieved significant carbon emission reduction
effect. The heterogeneous emission reduction effect of market mecha- through adjusting energy consumption structure and technological in-
nism and government intervention is analyzed. Moreover, the path of novation. Yang et al. (2022) proved the effect of ETS on reducing carbon

29
B. Lin and C. Huang Sustainable Production and Consumption 33 (2022) 28–37

emissions by changing energy consumption structure through the DID point when four carbon trading pilots officially began trading. In 2014,
model. however, the remaining carbon trials began trading. The carbon trading
There is also a lot of discussion at the city level, industry level and en- pilot was not started at the same time, so it is difficult to accurately eval-
terprise level. Y. Zhang et al. (2020) applied a DID model to study city-level uate the policy effect of ETS using the two-stage DID model.
panel data from 2004 to 2015 and found that ETS reduced carbon emis-
sions in pilot areas by about 16.2 %. Gao et al. (2020) analyzed data from 3. Method and data
28 industries and found that ETS contributed to regional and industrial car-
bon reduction. Wen et al. (2020) applied a DID model to evaluate the im- 3.1. Methods
pact of ETS and believed that the implementation of ETS had little effect on
industrial carbon emissions. Taking industry as the research object, W. The DID model is a quasi-experimental research design, which is
Zhang et al. (2020) verified that ETS significantly reduced industrial car- more and more popular in policy effect evaluation. This approach has
bon emissions by using DID model. A study of the power industry also several advantages (Fanghella et al., 2022; Harju et al., 2022). First,
suggested that ETS could help control carbon emissions from the industry this method can largely avoid the problem of endogeneity (P. Yu et al.,
(X. Yu et al., 2021). Peng et al. (2021) conducted empirical research on the 2021). Second, compared with traditional methods, the model setting
data of various industries in China, and the results showed that China's car- of the DID is more scientific and can estimate the policy effect more ac-
bon trading helped to reduce carbon emissions. Y. Hu et al. (2020) applied curately. Third, the principles and setting of the DID are simple and easy
the DID model to analyze industrial data, and the results imply that carbon to understand and apply. Therefore, based on the advantages of the DID
emissions in pilot areas were reduced by 15.5 %. Tan et al. (2022) method and the research problems of this paper, this paper chooses the
õconducted an analysis based on the data of enterprises in the Hubei DID model as the core model. Further, combined with the actual trading
pilot program, and the results proved that ETS may help reduce carbon situation of the pilot carbon market in China, this paper analyzes the im-
emissions caused by energy consumption. Zheng et al. (2021) suggest plementation effect of carbon trading policy by using the multi-period
that ETS reduced corporate carbon emissions by promoting technological DID model (Huang et al., 2021; Sun and Li, 2021).
innovation. Shen et al. (2020) found evidence that ETS reduced carbon A benchmark regression model was constructed based on the multi-
emissions in a sample of listed companies. period DID model, as shown in formula (1). In the formula, the fixed ef-
The above studies are based on actual data, and some studies use fects of year and province are controlled. Subscripts i and t represent re-
simulation techniques and scenario analysis to analyze the carbon re- gion and year, respectively. Yit is the dependent variable, and in this
duction effect of ETS. Jin et al. (2020) found that the carbon emission re- paper, it is carbon emission. treati is a dummy variable. If the sample is a
duction effect is stronger when the output-based allocation is fewer. treatment group, make treati equal to 1, otherwise equal to 0. periodit is
The scenario analysis results of Fang et al. (2020) show that mature car- also a dummy variable, set to 1 if there is carbon trading in the current
bon trading can effectively control carbon emissions. Zhang et al. year, or 0 otherwise. μi represents province fixed effect, λt represents
(2020a) simulated three carbon trading schemes, and the results indi- year fixed effect and εit represents error term. Other variables are
cate that carbon trading has great potential for energy conservation control variables, and the specific information is shown in Table 1.
and emission reduction. Some studies also analyze the realization of
emission reduction effect from the top-level design. Qin et al. (2020) Y it ¼ α þ β1 treat i  periodit þ β2 Lnpopit þ β3 Lnecoit þ β4 pdit ð1Þ
suggest that giving negative carbon quotas in regions with a large pro- þ β5 sisit þ β6 tisit þ β7 urbit þ β8 rsit þ β9 ppsit þ β10 ouit
portion of clean energy will encourage carbon emission reduction. Lin
and Jia (2020) found that the coverage of the carbon market and the car- þ β11 Lngpit þ β12 presit þ μ i þ λt þ ε it
bon price will affect the effect of carbon emission reduction. Zhang et al.
(2022) prove that the scheme combining carbon trading with carbon Since treati × periodit = DIDit, formula (1) can be shortened to
tax achieves higher economic efficiency and emission reduction effect. formula (2). DIDit is the core explanatory variable. Xit represents the
In short, there is more and more research on carbon trading; espe- control variable in formula (1).
cially the influence of the carbon market on carbon emissions is a hot
topic. Carbon markets are attracting more and more attention as one Y it ¼ α þ β1 DIDit þ βx X it þ μ i þ λt þ εit ð2Þ
of the most effective policies to deal with climate change. However,
the existing literature has the following limitations. On the one hand, The overall process of the study can be divided into three parts. The
the articles about whether ETS has a positive effect on carbon emission flow chart of the study is shown in Fig. 1. The first part is baseline regres-
mainly discuss whether the emission reduction effect exists, but ignore sion. This part also includes tests to be met using the DID, namely paral-
discuss why the emission reduction effect can exist. As mentioned lel trend tests and a series of robustness tests. The second part is the
above, numerous studies have demonstrated the carbon reduction ef- analysis of the coordinated emission reduction mechanism. First, the
fect of ETS. But they are invariably coy about how the market works. Un- role of the market mechanism is analyzed, and then the role of govern-
doubtedly, based on the actual performance and trading data of China's ment intervention. The third part analyzes the path of carbon emission
pilot carbon market, it is hard not to doubt the effectiveness of the pilot reduction and examines whether the energy consumption structure or
phase of the carbon market. Existing studies do not solve these doubts industrial structure has an intermediary effect.
and give rise to the illusion that market-based trading mechanisms pro-
mote carbon reduction. This is not conducive to the optimization of the 3.2. Data description
ETS mechanism. Therefore, it is necessary to explore whether and why
the carbon market achieves the desired effect against the background The explained variable of this paper is carbon dioxide emissions, and
of poor market performance. the core explanatory variable is whether to start carbon trading, namely
On the other hand, most studies apply the two-stage DID model to the DID variable.
analyze the policy effect of ETS. That is, the policy implementation According to existing studies, carbon emissions are closely related to
point is set as a specific year. For example, the implementation period population size, economic development, urbanization level, industrial
of the policy is set to 2011, because that is the year China officially estab- structure, resource endowment, technological level, environmental reg-
lished carbon trading pilots. However, the fact is that this year has only ulation, and other factors (Jiang et al., 2021; Li et al., 2021; Lin and Liu,
been the establishment of pilot projects. However, the pilots did not 2015; Safi et al., 2021). To make the experimental and control groups'
open the actual transaction. So, 2011 as a point of policy impact is not carbon emissions comparable, the relevant characteristics of the treat-
convincing enough. Others take 2013 as a policy implementation ment group and the control group must be controlled. Therefore, this

30
B. Lin and C. Huang Sustainable Production and Consumption 33 (2022) 28–37

Table 1
Summary statistics.

Symbol Variables Definition Obs Mean Std. Min Max

Dev

LnCO2 Carbon emission Carbon dioxide emissions (ton)(log)a 450 5.43 0.77 2.80 6.84
DID DID dummy variable Whether to launch a carbon market 450 0.09 0.28 0.00 1.00
Lnpop Population size The total population (million persons)(log)b 450 3.58 0.75 1.69 4.83
Lneco Economic level The ratio of real GDP to population (yuan per person)b 450 10.20 0.57 8.56 11.67
pd The population density The ratio of population to land area (persons per 10 square meters)b 450 0.43 0.53 0.01 3.01
sis The industrial structure The ratio of the output value of the secondary industry to GDPb 450 0.43 0.08 0.16 0.62
tis The industrial structure The ratio of the output value of the tertiary industry to GDPb 450 0.46 0.09 0.30 0.84
urb Urbanization level The ratio of urban population to total populationb 450 0.54 0.14 0.27 0.94
re Resources endowment The ratio of raw coal production to coal consumptionc 450 0.67 0.72 0.00 3.25
pps Power production The ratio of thermal power generation to total power generationc 450 0.76 0.23 0.08 1.00
structure
ou Level of opening-up The ratio of total imports and exports to GDPc 450 0.45 0.49 0.00 2.16
Lngp Level of green innovation Number of green patents(log)d 450 7.01 1.63 2.40 10.44
pres Carbon reduction The difference between the current year's carbon reduction target and the actual results of last 450 −0.96 5.66 −71.11 30.60
pressure year's carbon reduction (%)e
conc Transaction The ratio of the trading volume in the 30 days before the performance period to the trading volume 450 0.03 0.14 0.00 1.00
concentration in the performance yearf
effe Percentage of valid The ratio of non-zero trading days to total trading daysf 450 0.04 0.16 0.00 1.00
trading days
c
ecs Energy consumption The ratio of coal consumption to total energy consumption 450 0.50 0.17 0.10 0.97
structure
a
CEADs https://www.ceads.net.cn/.
b
National Bureau of Statistics.
c
CEIC Database.
d
Intellectual property search platform http://www.patviewer.com/.
e
Government work reports on the websites of provincial and municipal governments.
f
Carbon market analysis platform http://k.tanjiaoyi.com/.

paper selects several control variables that may affect regional carbon group due to their late start, limited data, and limited influence on the
emissions, including population size, economic level, population den- study. Finally, the treatment group consisted of 6 regions (Beijing,
sity, industrial structure, urbanization level, resource endowment, Shanghai, Guangdong, Tianjin, Hubei, Chongqing), and the control
power production structure, opening-up level, green innovation level, group consisted of the remaining 24 regions. The policy implementation
and carbon emission reduction pressure. point of Beijing, Shanghai, Guangdong, and Tianjin was in 2013, and that
Moreover, this paper introduces two indexes of trading concentra- of Hubei and Chongqing was in 2014. The specific information of the
tion and the percentage of valid trading days as proxy variables of the data is shown in Table 1.
market mechanism. The energy consumption structure variable is also
introduced to assist the empirical analysis. 4. Results and discussion
Considering the availability of data, this paper collected panel data
from 30 provinces in China from 2005 to 2019, with 450 samples. As it 4.1. Effect of ETS on carbon emission reduction
is provincial panel data, the Shenzhen carbon market will not be dis-
cussed separately, and will be incorporated into the Guangdong carbon The benchmark regression results of the multi-stage DID model is
market for discussion. Fujian and Sichuan were regarded as the control shown in Table 2. Column (1) of Table 2 does not add control variables,

Start

The benchmark Parallel trend test and

Robustness test
return dynamic effect analysis
Placebo test

PSM-DID

Analysis of transmission Collaborative emission

End
mechanism reduction mechanism analysis Eliminate other policy
interference

Elimination of special
Energy consumption samples
The market mechanism
structure
Change the sample time
The industrial structure Government intervention window

Fig. 1. Research flow chart.

31
B. Lin and C. Huang Sustainable Production and Consumption 33 (2022) 28–37

Table 2 Although low carbon prices and poor liquidity characterize the pilot
The result of baseline regression. carbon market in China, it still shows carbon emission reduction effects.
LnCO2 One of the reasonable guesses for this result is that government inter-
(1) (2)
ventions play a role. Carbon trading itself is a market-based environ-
mental regulation measure, and administrative instruments will come
DID −0.2037*** −0.1309***
into play when the carbon market price does not provide sufficient in-
(0.0258) (0.0279)
Lnpop 1.0425*** centives. Moreover, most of the participants in the Chinese carbon mar-
(0.2834) ket are state-owned enterprises (SOEs) (Wang et al., 2019). Compared
Lneco −0.0258 to non-SOEs, the government has stronger control over SOEs, and the
(0.0724) administrative measures will be more effective. Evidence for the
pd −1.1314***
(0.2635)
above speculation is that the Chinese carbon market is characterized
sis −0.4339 by a high compliance rate and low liquidity (Yi et al., 2018). In addition,
(0.332) the observed emission reduction effect may also be due to the spillover
tis −0.5008 effect of carbon trading (Yu and Li, 2021). Although the pilot phase
(0.4377)
covers a limited number of industries, enterprises in other sectors in
urb 0.6992**
(0.3135) the pilot area have also carried out emission reduction actions due to
rs 0.0515 the expected policy effects.
(0.0362)
pps 0.2937*** 4.2. Parallel trend test and dynamic effect analysis
(0.1005)
ou −0.0328
(0.054) One of the preconditions for using DID method is that the treatment
Lngp 0.0847*** group and the control group satisfy the parallel trend hypothesis (Yang
(0.022) et al., 2021). The parallel trend test model is shown in formula (3). Di is
pres 0.0034*
the dummy variable for the year. This study takes 2005 as the base period
(0.0017)
constant 5.4484*** 1.6629 to conduct a parallel trend test, and sets fourteen different dummies
(0.0062) (1.0853) (D−7-D+6). The test results are shown in Fig. 2, where phase 0 represents
Province FE Yes Yes the first year of policy impact. According to the results, the 95 % interval of
Year FE Yes Yes the corresponding coefficients from 2006 to 2012 contained 0, all of
N 450 450
adj. R-square 0.9777 0.9837
which failed the significance test, indicates that the experimental and
control groups' variation trend met the parallel trend test.
Note: Standard errors in parentheses; * p<0.1, ** p<0.05, *** p<0.01.
6
Y it ¼ α þ β1 ∑ Di DIDit þ β2 X it þ μ i þ λt þ εit ð3Þ
i¼  7
and column (2) adds control variables. Both the year effect and province
effect were controlled. The robust standard error reports are reported in According to the results, it can be seen that the carbon abatement ef-
parentheses. According to the calculation results, the DID coefficient is fect starts to appear in period D1, and then strengthens year by year. In
−0.1309 and significant at the 1 % level. The results indicate that carbon the last period of the observation period, the abatement effect is not ob-
trading significantly promotes carbon emission reduction, which is the vious. This may be because, after several years of practice, the enter-
same as X. Yu et al. (2021) and Y. Zhang et al. (2020). prises in the carbon market have made some technological progress,

Fig. 2. Results of parallel trend tests.

32
B. Lin and C. Huang Sustainable Production and Consumption 33 (2022) 28–37

while the quantity constraint of carbon quotas is relatively loose and the Table 3
level of a carbon price is low, and the enterprises do not have enough in- Results of robustness tests.

centives to implement further emission reduction behavior. LnCO2

(1) (2) (3) (4) (5)

4.3. Robustness tests
DID −0.2189*** −0.1998*** −0.1375*** −0.0764** −0.0738**
(0.0511) (0.0577) (0.0288) (0.0344) (0.0322)
4.3.1. The Placebo test Lnpop 2.2647*** 1.5416*** 1.1055*** 0.6121* 2.1294***
The placebo test can further rule out the influence of other policy or (0.4182) (0.3060) (0.2910) (0.3186) (0.4319)
randomness factors on experimental results (Jia et al., 2021). This paper Lneco 0.5508*** −0.1153 −0.0069 0.1755* 0.0903
randomly sampled all provinces and policy times without repetition. Six (0.1112) (0.0857) (0.0751) (0.1005) (0.1179)
pd −5.7971*** −1.4090*** −1.1355*** −0.7935*** −2.0111***
virtual treatment groups and corresponding policy implementation pe- (1.0620) (0.2861) (0.2682) (0.2154) (0.6596)
riods were randomly selected each time, and this operation was re- sis −1.6713*** −0.1363 −0.5002 −0.2225 −1.1761**
peated 500 times. This test hypothesizes that, when carbon trading (0.3449) (0.3688) (0.3418) (0.3952) (0.4836)
has a significant carbon emission reduction effect in the pilot area, the tis −0.5675 0.4423 −0.5459 −0.0474 −0.3869
(0.6096) (0.5079) (0.4563) (0.4811) (0.6846)
DID coefficient obtained by baseline regression should be located at
urb 0.4704 0.9647* 0.7417** 0.8369** −0.1024
the lower tail of the distribution function. The results of the placebo (0.4449) (0.5010) (0.3205) (0.3257) (0.4464)
test are shown in Fig. 3. The results suggested that the baseline regres- rs 0.0728 −0.034 0.0535 0.0929*** −0.1688***
sion results passed the placebo test, which further proved the robust- (0.0815) (0.0465) (0.0367) (0.0345) (0.0414)
ness of the baseline regression results. pps −0.0875 0.6837*** 0.2455** 0.2663** 0.3794***
(0.1289) (0.1086) (0.1140) (0.1192) (0.1365)
ou −0.0152 −0.0772 −0.0397 −0.0434 −0.0700
4.3.2. PSM-DID (0.0859) (0.0709) (0.0546) (0.0651) (0.0649)
The DID method requires that the treatment and control group have lngp 0.0491* 0.0756*** 0.0876*** 0.0799*** 0.0316
homogeneity, considering that China may have sample selection bias (0.0260) (0.0237) (0.0225) (0.0244) (0.0222)
pres 0.0018 0.0023 0.0033* 0.0069*** 0.0037***
due to some factors when selecting carbon trading pilots. Therefore,
(0.0019) (0.0017) (0.0018) (0.0014) (0.0010)
the propensity score matching-difference-in-differences (PSM-DID) constant −6.1474*** 0.0558 1.3334 0.6631 −1.7757
method was applied to test the robustness further (Yang and Wang, (1.4345) (1.1833) (1.1106) (1.5110) (1.8716)
2021). The specific approach is to carry out year-to-year propensity Province Y Y Y Y Y
score matching for the data from 2005 to 2010, reserving only the sam- fixed
effects
ple points within the common value range in each matching year, and Year fixed Y Y Y Y Y
conducting the multi-period DID model on these samples. The results effects
are shown in column (1) of Table 3. The calculated DID coefficient is Observations 135 285 420 330 240
−0.2189, which is significant at the 1 % level, proving the robustness adj. R- square 0.9957 0.9843 0.9840 0.9891 0.9922
of the benchmark regression results in this paper. Note: Standard errors in parentheses; * p<0.1, ** p<0.05, *** p<0.01.

4.3.3. Excluding the interference of other policies

Since 2007, the Ministry of Finance, the Ministry of Environmental Chongqing, Shaanxi, Hebei, and Henan. Emission trading policies may
Protection, and the National Development and Reform Commission also affect carbon emissions in pilot areas and interfere with identifying
have approved the pilot trading of emission rights in 11 regions, includ- carbon emission reduction effects. The specific approach eliminates the
ing Jiangsu, Zhejiang, Tianjin, Hubei, Hunan, Inner Mongolia, Shanxi, areas covering the pilot trading of emission rights and then carries out

Fig. 3. Results of the placebo test.

33
B. Lin and C. Huang Sustainable Production and Consumption 33 (2022) 28–37

the DID model again. The results are shown in column (2) of Table 3. The Table 4
calculated DID coefficient is −0.1998, and is significant at the 1 % level, Results of synergistic mechanism testing.

further proving the robustness of the benchmark regression results in LnCO2

this paper. (1) (2) (3)

DID −0.1301*** −0.1156*** −0.1556***

4.3.4. Deleting special samples
(0.0316) (0.0357) (0.0270)
In 2016, the Sichuan carbon market and the Fujian carbon market DID*conc −0.0031
opened one after another. These two samples may affect the results, (0.0481)
so they were removed, as the DID result was calculated. The results DID*effe −0.0345
are summarized in column (3) of Table 3. The calculated DID coefficient (0.0548)
DID*pres −0.0084***
is −0.1375 and is significant at the 1 % level, proving the robustness of (0.0015)
the benchmark regression results in this paper. pres 0.0072***
(0.0013)
4.3.5. Shortening of the sample time window Controls Y Y Y
Province fixed effects Y Y Y
In 2015, the Chinese government submitted to the United Nations on
Year fixed effects Y Y Y
June 30, 2015, the “Enhanced Action on Climate Change – China's Nation- Observations 450 450 450
ally Determined Contribution”, which set a voluntary action target by adj. R-square 0.9836 0.9837 0.9844
2030: carbon dioxide emissions will peak around 2030 and strive to Note: Standard errors in parentheses; * p<0.1, ** p<0.05, *** p<0.01.
peak as soon as possible. The incident could lead to tighter government
controls after 2015. Therefore, the sample time was shortened from
2005 to 2015. The calculated results are shown in column (4) of Table 3, do not consider the illiquid nature of the carbon market. Chen et al.
and the estimated coefficient of DID is −0.0764, which is significant at (2020) only analyzed the mediating mechanism and heterogeneity of
the 5 % level. At the same time, considering the time limitation before the carbon emission reduction effect. Dong et al. (2019) and Qi et al.
and after the policy, the sample window period was shortened to three (2021) consider the synergy between carbon emission reduction and
years before and after the policy, namely 2010–2017, and the regression economic development. The lack of analysis of market mechanisms in
was conducted again. The calculation results are presented in column these studies creates the illusion that carbon trading is effectively pro-
(5) of Table 3. The coefficient of DID is −0.0738 and is significant at the moting carbon reduction. Yi et al. (2020)'s study is slightly different.
5 % level. The above results all prove the robustness of the benchmark re- They prove that there is heterogeneity in the emission reduction effects
gression results. of different pilot programs. To be specific, the pilot programs in Beijing,
Shanghai and Hubei have emission reduction effects, the pilot programs
4.4. Analysis of coordinated emission-reduction mechanism in Guangdong promote carbon emissions, and the pilot programs in
Tianjin have no significant impact on carbon emissions. Nevertheless,
4.4.1. Measuring the emission reduction effect of the market mechanism Yi et al. (2020) still do not answer the question. Pilots in Beijing,
The interaction term of marketit and DIDit is introduced into the Shanghai and Hubei also suffer from low liquidity. The results of this sec-
baseline regression to examine whether the market mechanism is tion show that the market mechanism does not play an effective role, and
effective, as shown in formula (4). In this paper, marketit is an how the emission reduction effect is realized will be further analyzed.
indicator to measure the performance of the carbon market.
Specifically, it is expressed by the concentration of trading volume 4.4.2. Measuring the emission reduction effect of government intervention
(conc) and the proportion of effective trading days (effe). In this way, Here, the interaction term of marketit and DIDit is replaced by the
the carbon emission reduction effect is divided into two parts. θ cross-product term of govit and DIDit to analyze further whether
measures the carbon emission reduction effect caused by the market government intervention produces a carbon emission reduction effect,
mechanism, while β1 represents the carbon emission reduction effect as shown in formula (5). govit is an indicator measuring the degree of
caused by the non-market mechanism. government intervention.

Y it ¼ α þ β1 DIDit þ θDIDit  market it þ β2 X it þ μ i þ λt þ εit ð4Þ Y it ¼ α þ β1 DIDit þ θg DIDit  govit þ β2 X it þ μ i þ λt þ εit ð5Þ

It's important to note that, because DIDit × marketit = marketit, Given the content of this paper, government intervention here
marketit is not added to the formula (4). should measure government control of the carbon market. Therefore,
The results are displayed in columns (1) and (2) of Table 4. For conc common indicators such as fiscal dependence and the proportion of
or effe, the coefficient θ of the interaction term is not significant. While state-owned enterprises cannot help better measure the government
the β1 is both significantly negative at the 1 % level. The results intervention in this study. Based on the above analysis, this paper uses
suggest that the market mechanism does not inhibit substantially carbon emission reduction pressure (pres) to represent govit.
carbon emissions. The carbon trading promotes carbon emission Theoretically, the greater the pressure to cut carbon, the more likely
reduction through channels rather than the market mechanism. The governments will intervene in carbon markets.
possible reasons are that the carbon market in the current stage is not In formula (5), pres is not added to the equation because the control
mature, the market mechanism is not perfect, the number of free variable contains pres. θg measures the carbon emission reduction effect
carbon quotas is too large, the carbon price is too low, and the main caused by the government intervention.
body of emission control is more passive to fulfill the requirements of The results are shown in column (3) of Table 4. According to the re-
the implementation of the treaty, leading to the failure of the market gression results, the θg is significant at the 1 % level, and the β1 is also
mechanism to play an influential role. For example, Ji et al. (2021) significant at the 1 % level. The results demonstrate that, during the
pointed out that there were problems of oversupply of quota and low observation period, government intervention significantly strengthened
transaction prices in the pilot projects. the carbon emission reduction effect of carbon trading in the pilot area,
This is a new finding compared to existing research. Similar studies and hence the stronger the government intervention, the stronger the
include Chen et al. (2020), Dong et al. (2019), Qi et al. (2021) and Yi emission reduction effect of the pilot area. Similar conclusions can be
et al. (2020), all of which use provincial panel data and the DID model found in other articles. For example, Y. Hu et al. (2020)'s study suggests
to prove the carbon emission reduction effect of ETS. But these studies that pilot areas with stricter environmental law enforcement performed

34
B. Lin and C. Huang Sustainable Production and Consumption 33 (2022) 28–37

better in carbon emission reduction. Therefore, a reasonable guess is that Table 5

although the market mechanism failed to act as an incentive when the Results of mediating effect test.

pilot carbon market was still in its infancy, the government was under LnCO2 ecs sis LnCO2
pressure to adopt a series of measures to ensure compliance and thus (1) (2) (3) (4)
reduce carbon emissions. Local governments face the pressure of
DID −0.1284*** −0.0303** 0.0017 −0.0895***
performance appraisal and the central government has an incentive to
(0.0278) (0.0138) (0.0044) (0.0208)
ensure high compliance with the carbon market. Meanwhile, local ecs 1.3003***
governments are facing the pressure of popular opinion. Popular (0.0908)
opinion plays an important role in implementing carbon emission sis 0.2722
reduction targets (Moore et al., 2022). Since the carbon market pilot (0.1767)
Controls Y Y Y Y
project was launched, the carbon market has received extensive Province fixed effects Y Y Y Y
attention from the media and the public. Carbon market trading Year fixed effects Y Y Y Y
information is subject to public supervision. Popular opinion has Observations 450 450 450 450
prompted the government to pay attention to the construction of the adj. R-square 0.9837 0.9091 0.9304 0.9907
carbon trading pilot to a certain extent. Therefore, the government will Note: Standard errors in parentheses; * p<0.1, ** p<0.05, *** p<0.01.
take a series of measures to ensure a high compliance rate and carbon
emission reduction effect. For example, the measures taken include
extending the performance period, setting up the correction period, analysis. The regression equations are shown in formulas (6)–(8). Zit
publicizing the list of non-performance, using the disciplinary mechanism represents the possible intermediary variable to be tested.
for dishonesty to punish, conducting carbon quota auction, carrying out
performance training, and holding talks with critical units, etc.
Y it ¼ α þ β1 DIDit þ β2 X it þ μ i þ λt þ εit ð6Þ
Based on the above results, in the pilot stage of China's carbon mar-
ket, government intervention, not market mechanisms, achieved the
main reduction effect. This result answers the question of why the car- Z it ¼ α þ β3 DIDit þ β4 X it þ μ i þ λt þ εit ð7Þ
bon emission reduction effect still exists under the premise of the low
liquidity of carbon market. Different from relevant papers (Chen et al.,
2020; Dong et al., 2019; Qi et al., 2021; Yi et al., 2020), this paper focuses Y it ¼ α þ β5 DIDit þ β6 Z it þ β7 X it þ μ i þ λt þ εit ð8Þ
on the causes of carbon emission reduction effect, rather than just the
proof of emission reduction effect and the analysis of intermediary
mechanism. This paper combined with the actual operation of the car- According to relevant research, this paper takes energy consumption
bon market to better explain the role of ETS on carbon emissions. structure (ecs) and industrial structure (sis) as possible mediating vari-
It is worth noting that although the current ETS does achieve carbon ables and tests the mediating effect (Lin and Fei, 2015; Lin and Zhang,
emission reduction, it does not mean that government intervention is 2016; Ouyang et al., 2020).The regression results are shown in Table 5.
optimal. The core of carbon trading is to cost the environment and According to the results, the energy consumption structure plays a me-
transform it into a paid factor of production with the help of market diating role, but the industrial structure does not.
forces. This paper does not deny that government intervention should Some studies have pointed out that the step-by-step test has minor
play a role, but market mechanisms must play a leading role in achiev- statistical efficacy and may miss the actual mediation impact
ing sustainable carbon trading and emission reduction goals. Fang et al. (MacKinnon et al., 2007). Therefore, the Sobel-Goodman test was also
(2020) pointed out that carbon trading is sensitive to government inter- used to analyze the mediating effect (Fritz and MacKinnon, 2007). The
vention. The boundary between the government and the market must Sobel-Goodman test results are presented in Table 6, suggesting that
be well grasped if the sustainable development of carbon trading is re- energy consumption structure has a mediating impact, while industrial
alized. Only under the action of the market supply and demand mecha- structure does not pass the test.
nism and competition mechanism can a reasonable carbon price be Both tests indicate that carbon trading reduces carbon emissions in
formed in the carbon market to effectively and rationally allocate emis- pilot areas by adjusting energy consumption structure rather than in-
sion reduction resources. Through market competition, the survival of dustrial structure. This result is similar to Wen et al. (2020), and they
the fittest will be realized. Enterprises will be encouraged to save energy also believe that ETS can affect carbon emissions through the energy
and reduce emissions at the lowest cost, which will force industrial structure effect. The reason why energy consumption structure plays a
upgrading and thus contribute to the realization of carbon neutrality. mediating role may be that most of the industries covered by the pilot
If only relying on government intervention, it can only ensure the com- areas are the electric power industry, and the proportion of coal in the
pliance rate of the carbon market, resulting in low market activity, poor power generation end is reduced. This speculation can be supported
liquidity, and market failure, resulting in insufficient motivation for en- by Cao et al. (2021). They found that changes in coal-fired efficiency of
terprises to reduce emissions. At present, China has completed the first ETS regulated coal-fired power plants had no effect, but coal consump-
performance cycle of the national carbon market. The national carbon tion was significantly reduced, and the output of non-coal-fired power
market seems to be facing the same dilemma as the pilot carbon market. plants in ETS increased significantly. The reasons for the failure of indus-
According to the public data, the completion rate of the first perfor- trial structure to play a significant intermediary effect may be that the
mance cycle is as high as 99.5 %, but the average carbon price is only carbon market is immature, the low carbon price fails to stimulate in-
42.80 Yuan/ton of CO2. The carbon market is a long-term market mech- dustrial adjustment effectively, and the industry scope covered by the
anism. If we want to speed up the construction of the national carbon carbon market is limited.
market, we must balance the relationship between the market mecha-
nism and government intervention to give full play to the role of the car-
Table 6
bon trading mechanism and stimulate market activity. Sobel test results.

Variables Coefficient z p > |z| intervening The proportion of total

4.5. Analysis of transmission mechanism and discussion
variable effect that is mediated

Further, explore the way of carbon emissions decline. The mediation ecs −0.0421 −2.142 0.0322 Yes 0.3220
sis 0.0001 −0.2138 0.8307 No –
effect model put forward by Baron and Kenny (1986) is used for

35
B. Lin and C. Huang Sustainable Production and Consumption 33 (2022) 28–37

5. Conclusions and policy recommendations reduction cost of thermal power enterprises cannot be transmitted
downstream, and end-users cannot be effectively encouraged to reduce
Based on China's provincial panel data from 2005 to 2019, this paper emissions. Energy-intensive industries must be included in the carbon
explores whether carbon trading can effectively promote carbon emis- market to form a supply-demand linkage, force industrial transforma-
sion reduction through a multi-period difference-in-differences model. tion and improve market efficiency.
In addition, the proxy variables of market mechanism and government The research design in this paper can more accurately estimate the
intervention are innovatively introduced into the benchmark model, effects of carbon trading policies, as the regression results passed a se-
the synergistic effect of government intervention and market mecha- ries of robustness tests. However, this article has some limitations. In
nism is discussed, and the problem of why the emission reduction effect the benchmark regression, the use of provincial panel data does not
can still be achieved under the premise of the immature carbon market consider the carbon market of Shenzhen, which may be classified as
is solved. The conclusions are as follows. the carbon market of Guangdong. At the same time, the carbon markets
First, carbon trading promotes carbon emission reduction, but the in Fujian and Sichuan were not regarded as the control group due to
market mechanism alone cannot significantly achieve the effect. The their late opening and small trading volume in baseline regression. Al-
new finding of this paper is that the carbon emission reduction effect though these two markets are not pilots and have been excluded from
is realized through a non-market mechanism. The reason may be that the robustness test, it is possible to bias the baseline regression results.
the current stage of the carbon market is immature, the market mecha- In the future, more micro-level data should be used to evaluate the ef-
nism is not perfect, and the scope of industry coverage is limited. In ad- fect of policies to obtain more scientific results.
dition, there are many free carbon quotas, and the carbon price is too
low. Many reasons may lead to the failure of market mechanisms to CRediT authorship contribution statement
play the role of incentive control and emission reduction effectively.
Second, government intervention has significantly enhanced the Boqiang Lin: Conceptualization, Methodology, Software, Data
role of carbon trading in promoting carbon reduction. The more govern- curation, Writing – original draft. Chenchen Huang: Methodology, Soft-
ment intervention, the more carbon trading will curb emissions. This ware, Data curation, Writing – original draft.
may be related to China's traditional environmental regulation means
and control, as discharge subjects are primarily state-owned enter-
prises. In the initial stage of carbon market construction, the emission Declaration of competing interest
reduction behaviors of the controlling and emission subjects were
more constrained by the obligations of compliance and punishment The authors declare that they have no known competing financial
measures than by the active participation in market transactions. interests or personal relationships that could have appeared to influ-
Third, carbon trading reduces carbon emissions by adjusting the en- ence the work reported in this paper.
ergy consumption rather than the industrial structure. The reason may
be that the current carbon market is immature, with low carbon prices Acknowledgments
and an oversupply of quotas failing to incentivize the industry to adjust.
In addition, the current carbon market covers a limited range of indus- This paper is supported by National Natural Science Foundation of
tries, principally the electric power industry, so carbon trading has little China (Key Program, No 72133003).
impact on the industrial structure.
According to the research conclusions, this paper puts forward three References
policy suggestions. Baron, R.M., Kenny, D.A., 1986. The moderator-mediator variable distinction in social psy-
First, we should further develop and improve market mechanisms chological research. Conceptual, strategic, and statistical considerations. J. Pers. Soc.
and give full play to the role of the invisible hand. In the operation of Psychol. 51, 1173–1182. https://doi.org/10.1037/0022-3514.51.6.1173.
Cao, J., Ho, M.S., Ma, R., Teng, F., 2021. When carbon emission trading meets a regulated
the carbon market in the whole country, based on the pilot experience,
industry: evidence from the electricity sector of China. J. Public Econ. 200, 104470.
accelerate to carry out all kinds of supporting system construction, de- https://doi.org/10.1016/j.jpubeco.2021.104470.
sign reasonable carbon quota allocation system, to ensure that the Chen, S., Shi, A., Wang, X., 2020. Carbon emission curbing effects and influencing mecha-
total moderately tight quotas, form suitable price mechanism, great car- nisms of China’s emission trading scheme: the mediating roles of technique effect,
composition effect and allocation effect. J. Clean. Prod. 264, 121700. https://doi.org/
bon financial products, improve the carbon financial service system, to 10.1016/j.jclepro.2020.121700.
give full play to the incentive effect of the market mechanism. Dong, F., Dai, Y., Zhang, S., Zhang, X., Long, R., 2019. Can a carbon emission trading scheme
Second, we should grasp the boundary between the government generate the porter effect? Evidence from pilot areas in China. Sci. Total Environ. 653,
and the market and give fair play to the role of the visible hand. Al- 565–577. https://doi.org/10.1016/j.scitotenv.2018.10.395.
Dong, Z.Q., Wang, H., Wang, S.X., Wang, L.H., 2020. The validity of carbon emission trading
though government intervention is needed in the early stage to pro- policies: evidence from a quasi-natural experiment in China. Adv. Clim. Chang. Res.
mote the implementation of the regulation, in the long run, if the 11, 102–109. https://doi.org/10.1016/j.accre.2020.06.001.
future national carbon market wants to play the role of market- Fan, X., Lv, X., Yin, J., Tian, L., Liang, J., 2019. Multifractality and market efficiency of carbon
motivated environmental law, the market mechanism must play a lead- emission trading market: analysis using the multifractal detrended fluctuation tech-
nique. Appl. Energy 251, 113333. https://doi.org/10.1016/j.apenergy.2019.113333.
ing role in promoting the active participation of the control and emis- Fang, G., Lu, L., Tian, L., He, Y., Yin, H., 2020. Research on the influence mechanism of car-
sion players in the market transaction. China must pay attention to bon trading on new energy—a case study of ESER system for China. Phys. A Stat.
this problem in accelerating the construction of the national carbon Mech. Appl. 545, 123572. https://doi.org/10.1016/j.physa.2019.123572.
market. Only by taking effective measures to improve the dominant po- Fanghella, V., D’Adda, G., Tavoni, M., 2022. Evaluating the impact of technological renova-
tion and competition on energy consumption in the workplace. J. Environ. Econ.
sition of the market mechanism can we improve the market activity and Manag. 102662. https://doi.org/10.1016/j.jeem.2022.102662.
realize the sustainable emission reduction effect. The government Fritz, M.S., MacKinnon, D.P., 2007. Required sample size to detect the mediated effect.
should correctly understand its role in carbon trading and play a good Psychol. Sci. 18, 233–239. https://doi.org/10.1111/j.1467-9280.2007.01882.x.
position in supervision and management. Finally, the goal of carbon Gao, Y., Li, M., Xue, J., Liu, Y., 2020. Evaluation of effectiveness of China’s carbon emissions
trading scheme in carbon mitigation. Energy Econ. 90, 104872. https://doi.org/10.
emission reduction can be achieved effectively through the synergistic 1016/j.eneco.2020.104872.
effect of government intervention and the market mechanism. Golpîra, H., Javanmardan, A., 2022. Robust optimization of sustainable closed-loop supply
Third, we should gradually expand the scope of the carbon trading chain considering carbon emission schemes. Sustain. Prod. Consum. 30, 640–656.
https://doi.org/10.1016/j.spc.2021.12.028.
scheme to include more energy-intensive industries. The first industry
Goulder, L.H., Long, X., Lu, J., Morgenstern, R.D., 2022. China’s unconventional nationwide
included in the national carbon market was coal-fired power genera- CO2 emissions trading system: cost-effectiveness and distributional impacts.
tion. Due to China's electricity price mechanism setting, the emission J. Environ. Econ. Manage. 111, 102561. https://doi.org/10.1016/j.jeem.2021.102561.

36
B. Lin and C. Huang Sustainable Production and Consumption 33 (2022) 28–37

Guo, J., Gu, F., Liu, Y., Liang, X., Mo, J., Fan, Y., 2020. Assessing the impact of ETS trading Tan, X., Sun, Q., Wang, M., Se Cheong, T., Yan Shum, W., Huang, J., 2022. Assessing the ef-
profit on emission abatements based on firm-level transactions. Nat. Commun. 11, fects of emissions trading systems on energy consumption and energy mix. Appl. En-
2078. https://doi.org/10.1038/s41467-020-15996-1. ergy 310. https://doi.org/10.1016/j.apenergy.2022.118583.
Harju, J., Kosonen, T., Laukkanen, M., Palanne, K., 2022. The heterogeneous incidence of Wang, F., Sun, J., Liu, Y.S., 2019. Institutional pressure, ultimate ownership, and corporate
fuel carbon taxes: evidence from station-level data. J. Environ. Econ. Manage. 112, carbon reduction engagement: evidence from China. J. Bus. Res. 104, 14–26. https://
102607. https://doi.org/10.1016/j.jeem.2021.102607. doi.org/10.1016/j.jbusres.2019.07.003.
Hu, J., Pan, X., Huang, Q., 2020. Quantity or quality? The impacts of environmental regu- Wang, C., Shi, Y., Zhang, L., Zhao, X., Chen, H., 2021. The policy effects and influence mech-
lation on firms’ innovation–Quasi-natural experiment based on China’s carbon emis- anism of China’s carbon emissions trading scheme. Air Qual. Atmos. Heal. 14,
sions trading pilot. Technol. Forecast. Soc. Change 158, 120122. https://doi.org/10. 2101–2114. https://doi.org/10.1007/s11869-021-01081-z.
1016/j.techfore.2020.120122. Wen, Y., Hu, P., Li, J., Liu, Q., Shi, L., Ewing, J., Ma, Z., 2020. Does China's carbon emissions
Hu, Y., Ren, S., Wang, Y., Chen, X., 2020. Can carbon emission trading scheme achieve en- trading scheme really work? A case study of the Hubei pilot. J. Clean. Prod. 277,
ergy conservation and emission reduction? Evidence from the industrial sector in 124151. https://doi.org/10.1016/j.jclepro.2020.124151.
China. Energy Econ. 85, 104590. https://doi.org/10.1016/j.eneco.2019.104590. Yang, F., Wang, C., 2021. Clean energy, financial development, and economic growth: ev-
Huang, J., Shen, J., Miao, L., Zhang, W., 2021. The effects of emission trading scheme on in- idence from spatial spillover effects and quasi-natural experiments. J. Clean. Prod.
dustrial output and air pollution emissions under city heterogeneity in China. J. Clean. 322, 129045. https://doi.org/10.1016/j.jclepro.2021.129045.
Prod. 315, 128260. https://doi.org/10.1016/j.jclepro.2021.128260. Yang, Z., Shao, S., Yang, L., 2021. Unintended consequences of carbon regulation on the
Ji, C.J., Hu, Y.J., Tang, B.J., Qu, S., 2021. Price drivers in the carbon emissions trading performance of SOEs in China: the role of technical efficiency. Energy Econ. 94,
scheme: evidence from Chinese emissions trading scheme pilots. J. Clean. Prod. 105072. https://doi.org/10.1016/j.eneco.2020.105072.
278, 123469. https://doi.org/10.1016/j.jclepro.2020.123469. Yang, Z., Yuan, Y., Zhang, Q., 2022. Carbon emission trading scheme, carbon emissions re-
Jia, R., Shao, S., Yang, L., 2021. High-speed rail and CO2 emissions in urban China: a spatial duction and spatial spillover effects: quasi-experimental evidence from China. Front.
difference-in-differences approach. Energy Econ. 99, 105271. https://doi.org/10. Environ. Sci. 9, 824298. https://doi.org/10.3389/fenvs.2021.824298.
1016/j.eneco.2021.105271. Yao, S., Yu, X., Yan, S., Wen, S., 2021. Heterogeneous emission trading schemes and green
Jiang, M., An, H., Gao, X., Jia, N., Liu, S., Zheng, H., 2021. Structural decomposition analysis innovation. Energy Policy 155, 112367. https://doi.org/10.1016/j.enpol.2021.112367.
of global carbon emissions: the contributions of domestic and international input Yi, L., Li, Z., Yang, L., Liu, J., Packianather, M., 2018. Evaluation on the development degree
changes. J. Environ. Manag. 294, 112942. https://doi.org/10.1016/j.jenvman.2021. of China's seven pilot carbon markets. Chin. J. Popul. Resour. Environ. 16, 28–35.
112942. https://doi.org/10.1080/10042857.2018.1433809.
Jin, Y., Liu, X., Chen, X., Dai, H., 2020. Allowance allocation matters in China’s carbon emis- Yi, L., Bai, N., Yang, L., Li, Z., Wang, F., 2020. Evaluation on the effectiveness of China's pilot
sions trading system. Energy Econ. 92, 105012. https://doi.org/10.1016/j.eneco.2020. carbon market policy. J. Clean. Prod. 246, 119039. https://doi.org/10.1016/j.jclepro.
105012. 2019.119039.
Li, R., Wang, Q., Liu, Y., Jiang, R., 2021. Per-capita carbon emissions in 147 countries: the Yu, D.J., Li, J., 2021. Evaluating the employment effect of China's carbon emission trading
effect of economic, energy, social, and trade structural changes. Sustain. Prod. policy: based on the perspective of spatial spillover. J. Clean. Prod. 292, 126052.
Consum. 27, 1149–1164. https://doi.org/10.1016/j.spc.2021.02.031. https://doi.org/10.1016/j.jclepro.2021.126052.
Lin, B., Fei, R., 2015. Regional differences of CO2 emissions performance in China's agricul- Yu, P., Cai, Z., Sun, Y., 2021. Does the emissions trading system in developing countries ac-
tural sector: a Malmquist index approach. Eur. J. Agron. 70, 33–40. https://doi.org/10. celerate carbon leakage through OFDI?Evidence from China. Energy Econ. 101,
1016/j.eja.2015.06.009. 105397. https://doi.org/10.1016/j.eneco.2021.105397.
Lin, B., Jia, Z., 2020. Does the different sectoral coverage matter? An analysis of China’s car- Yu, X., Dong, Z., Zhou, D., Sang, X., Chang, C.Ter, Huang, X., 2021. Integration of tradable
bon trading market. Energy Policy 137, 111164. https://doi.org/10.1016/j.enpol.2019. green certificates trading and carbon emissions trading: how will chinese power in-
111164. dustry do? J. Clean. Prod. 279, 123485. https://doi.org/10.1016/j.jclepro.2020.123485.
Lin, B., Liu, H., 2015. China’s building energy efficiency and urbanization. Energy Build. 86, Zhang, Y., Zhang, J., 2019. Estimating the impacts of emissions trading scheme on low-
356–365. https://doi.org/10.1016/j.enbuild.2014.09.069. carbon development. J. Clean. Prod. 238, 117913. https://doi.org/10.1016/j.jclepro.
Lin, B., Zhang, Z., 2016. Carbon emissions in China×s cement industry: a sector and policy 2019.117913.
analysis. Renew. Sust. Energ. Rev. 58, 1387–1394. https://doi.org/10.1016/j.rser.2015. Zhang, F., Fang, H., Song, W., 2019. Carbon market maturity analysis with an integrated
12.348. multi-criteria decision making method: a case study of EU and China. J. Clean. Prod.
Lyu, J., Cao, M., Wu, K., Li, H., Mohi-ud-din, G., 2020. Price volatility in the carbon market in 241, 118296. https://doi.org/10.1016/j.jclepro.2019.118296.
China. J. Clean. Prod. 255, 120171. https://doi.org/10.1016/j.jclepro.2020.120171. Zhang, S., Jiang, K., Wang, L., Bongers, G., Hu, G., Li, J., 2020. Do the performance and effi-
MacKinnon, D.P., Fairchild, A.J., Fritz, M.S., 2007. Mediation analysis. Annu. Rev. Psychol., ciency of China’s carbon emission trading market change over time? Environ. Sci.
593–614 https://doi.org/10.1146/annurev.psych.58.110405.085542. Pollut. Res. 27, 33140–33160. https://doi.org/10.1007/s11356-020-09168-3.
Moore, F.C., Lacasse, K., Mach, K.J., Shin, Y.A., Gross, L.J., Beckage, B., 2022. Determinants of Zhang, W., Li, J., Li, G., Guo, S., 2020. Emission reduction effect and carbon market effi-
emissions pathways in the coupled climate–social system. Nature 603, 103–111. ciency of carbon emissions trading policy in China. Energy 196, 117117. https://doi.
https://doi.org/10.1038/s41586-022-04423-8. org/10.1016/j.energy.2020.117117.
Ouyang, X., Fang, X., Cao, Y., Sun, C., 2020. Factors behind CO2 emission reduction in chi- Zhang, Y., Li, S., Luo, T., Gao, J., 2020. The effect of emission trading policy on carbon emis-
nese heavy industries: do environmental regulations matter? Energy Policy 145, sion reduction: evidence from an integrated study of pilot regions in China. J. Clean.
111765. https://doi.org/10.1016/j.enpol.2020.111765. Prod. 265, 121843. https://doi.org/10.1016/j.jclepro.2020.121843.
Peng, H., Qi, S., Cui, J., 2021. The environmental and economic effects of the carbon emis- Zhang, Y.J., Liang, T., Jin, Y.L., Shen, B., 2020a. The impact of carbon trading on economic
sions trading scheme in China: the role of alternative allowance allocation. Sustain. output and carbon emissions reduction in China’s industrial sectors. Appl. Energy
Prod. Consum. 28, 105–115. https://doi.org/10.1016/j.spc.2021.03.031. 260, 114290. https://doi.org/10.1016/j.apenergy.2019.114290.
Qi, S., Cheng, S., Cui, J., 2021. Environmental and economic effects of China's carbon mar- Zhang, Y.J., Shi, W., Jiang, L., 2020b. Does China’s carbon emissions trading policy improve
ket pilots: empirical evidence based on a DID model. J. Clean. Prod. 279, 123720. the technology innovation of relevant enterprises? Bus. Strateg. Environ. 29,
https://doi.org/10.1016/j.jclepro.2020.123720. 872–885. https://doi.org/10.1002/bse.2404.
Qin, Q., Liu, Y., Huang, J.P., 2020. A cooperative game analysis for the allocation of carbon Zhang, S., Wang, Y., Hao, Y., Liu, Z., 2021. Shooting two hawks with one arrow: could
emissions reduction responsibility in China's power industry. Energy Econ. 92, China’s emission trading scheme promote green development efficiency and regional
104960. https://doi.org/10.1016/j.eneco.2020.104960. carbon equality? Energy Econ. 101, 105412. https://doi.org/10.1016/j.eneco.2021.
Safi, A., Chen, Y., Wahab, S., Ali, S., Yi, X., Imran, M., 2021. Financial instability and 105412.
consumption-based carbon emission in E-7 countries: the role of trade and economic Zhang, Y., Qi, L., Lin, X., Pan, H., Sharp, B., 2022. Synergistic effect of carbon ETS and carbon
growth. Sustain. Prod. Consum. 27, 383–391. https://doi.org/10.1016/j.spc.2020.10. tax under China’s peak emission target: a dynamic CGE analysis. Sci. Total Environ.
034. 825. https://doi.org/10.1016/j.scitotenv.2022.154076.
Shen, J., Tang, P., Zeng, H., 2020. Does China’s carbon emission trading reduce carbon Zheng, Y., Sun, X., Zhang, C., Wang, D., Mao, J., 2021. Can emission trading scheme im-
emissions? Evidence from listed firms. Energy Sustain. Dev. 59, 120–129. https:// prove carbon emission Performance?Evidence From China. Front. Energy Res. 9,
doi.org/10.1016/j.esd.2020.09.007. 759572. https://doi.org/10.3389/fenrg.2021.759572.
Sun, L., Li, W., 2021. Has the opening of high-speed rail reduced urban carbon emissions?
Empirical analysis based on panel data of cities in China. J. Clean. Prod. 321, 128958.
https://doi.org/10.1016/j.jclepro.2021.128958.

37