Energy Policy 183 (2023) 113803

Contents lists available at ScienceDirect

Energy Policy
journal homepage: www.elsevier.com/locate/enpol

Can carbon market policies achieve a “point-to-surface” effect?—

Quasi-experimental evidence from China
Xiaolin Yu a, b, *, Kai Wan c, Qunyang Du a, b
a
School of Economics, Zhejiang University of Technology, Hangzhou, 310000, China
b
Institute for Industrial System Modernization, Zhejiang University of Technology, Hangzhou, 310000, China
c
School of Economics, Hangzhou Dianzi University, Hangzhou, 310000, China

A R T I C L E I N F O A B S T R A C T

Keywords: The carbon market policy is an effective means to promote green and low-carbon development, and is important
Carbon market policy for China to achieve the carbon peaking and carbon neutrality goals. We manually collated a list of non-pilot
Green and low carbon development cities, neighbouring the regions implementing carbon market policies. With the panel data from 2006 to
External effects
2017, we examined the impact and mechanisms of the policy on non-pilot cities. The study found that the policy
Policy spillover
Number of industrial enterprises
of the pilot city not only worked in its region but also promoted green and low-carbon development in the
neighbouring non-pilot regions. This kind of external effect is self-reinforcing. Policy spillover effects and in­
dustrial relocation are important mechanisms for carbon market policies to exert externalities. However, the
“pollution refuge hypothesis” holds if polluting industries dominate neighbouring areas. Further analysis shows
that carbon market policies are more likely to have a point-to-surface effect in neighbouring cities if officials of
the pilot city are at a critical stage. Such effect is strongest in Beijing, followed by Shanghai, while it is not
effective in Hubei and Guangdong. Moreover, we found that carbon market policies were with synergistic effects
in terms of carbon and pollution reduction.

1. Introduction effect is not satisfactory (Yang et al., 2016). The development trend of
China’s carbon emissions trading market is taking a point-by-point
The problem of environmental degradation due to global warming is approach, which gradually transitions to a unified national carbon
of great concern to the world, and carbon dioxide (CO2) from human emissions trading market based on accumulated experience in pilot
activities has been a major contributor since the mid-twentieth century. areas. Then, does it work with the “pilot first, then replicate” model? Is
As the world’s largest emitter of CO2, China is actively addressing there a difference in the effect of the policy on neighbouring cities?
environmental issues. It has set the goal of carbon peaking by 2030 and As a market-based mechanism, carbon market policy topics have
achieving carbon neutrality by 2060. In 2011, the China Development been widely analyzed in environmental economics (Sandoff and Schaad,
and Reform Commission identified eight provinces and cities, to carry 2009). Earlier studies focused on assessing the effects of EU carbon
out carbon emissions trading pilot projects. It was officially launched in market policies, and since 2012 there has also been extensive debate in
2013. By 2020, China’s pilot carbon market has become the second China as well. The studies focused on China’s carbon market pilot pol­
largest in the world. With a cumulative achievement turnover of RMB icies, effects on carbon emissions (Zhang et al., 2017), carbon intensity
10.43 billion.1 It intuitively indicated that the carbon market policy has (Zhou et al., 2019), productivity (Yang et al., 2021), and technological
emerged as a main market-based emission reduction tool in China. The innovation (Chen et al., 2021). A review of the relevant literature reveals
carbon emissions trading market can theoretically optimize the distri­ that most studies affirm the effectiveness of pilot carbon market policies.
bution of emission rights and internalise the externalities of environ­ Cheng et al. (2015) took Guangdong Province as an example and argued
mental pollution through market incentives. However, due to the that carbon trading policies could reduce the cost of energy saving and
uneven distribution of quotas, the small scope of pilots, the imperfect emission reduction, which in turn could significantly reduce emissions
trading mechanism and the lagging legislation (Liu et al., 2015), the of CO2, sulphur dioxide and nitrogen oxides. Using carbon emission data

* Corresponding author. School of Economics, Zhejiang University of Technology, Hangzhou, 310000, China.
E-mail addresses: [email protected] (X. Yu), [email protected] (K. Wan), [email protected] (Q. Du).
1
The institutional background related to carbon market policies is reported in the Appendix section.

https://doi.org/10.1016/j.enpol.2023.113803
Received 31 October 2022; Received in revised form 22 August 2023; Accepted 30 August 2023
Available online 4 September 2023
0301-4215/© 2023 Elsevier Ltd. All rights reserved.
X. Yu et al. Energy Policy 183 (2023) 113803

from large-scale industries in Guangdong province, Wang et al. (2015) markets and the present study could make the first attempt to do so.
found that carbon trading policies could effectively curb carbon emis­ Discussing the relationship between carbon markets, energy efficiency
sions and carbon intensity of regional-scale industries. Previous studies and emissions reduction from a “grow out of nothing” perspective may
also concluded that carbon market policies could not only save energy lead to biased estimates. This paper not only solves the problem
and reduce emissions locally (Zhou et al., 2019; Huang et al., 2022), but mentioned above but also provides a new experience for the interactive
also had significant spatial spillovers (Yu and Zhang, 2021). However, behaviour of local governments.
other literature pointed out these policies might slow down economic Thirdly, the present study further examines the boundaries of the
growth due to the inadequacy of the market mechanism, and there were externalities of the carbon market pilot policies and finds that the
no substantial emission reductions (Zhang et al., 2019). Based on EU characteristics of the officials in charge and the nature of the cities are
data, Streimikiene and Roos (2009) found that carbon market policies important factors influencing the “point-to-surface” effect. The effects of
were ineffective in curbing the increase in carbon emissions. carbon market policies can play a synergistic role in reducing emissions,
The mainstream studies mainly focused on the effects of the policies providing new ideas on optimizing the construction of a unified China
where they were implemented, with little analysis of the impact on non- carbon market in next stage.
pilot cities from existing literature (Huang et al., 2022; Yu and Zhang,
2021). However, there was still some literature suggesting that the 2. Theoretical foundations and assumptions
policies could not only be effective in curbing carbon emissions in pilot
areas, but they could also lead to technological progress through 2.1. Green and low-carbon development of neighbouring non-pilot cities
knowledge spillovers (Dai et al., 2022; Li and Wang, 2022) and have a
positive impact on the environment and employment in neighbouring The core idea of the carbon market policy is to internalise the ex­
cities (Yu and Zhang, 2021; Yu and Li, 2021). For example, Li and Wang ternalities of pollution through a price mechanism (Mo et al., 2016).
(2021) found that technological progress driven by carbon market pol­ Under the constraints of carbon market trading rules, this paper argues
icies could generate spatial spillover effects through inter-regional ex­ that carbon markets can have external effects on neighbouring non-pilot
change and cooperation. Li et al. (2022) used spatial econometric cities in two ways. Firstly, the policies can drive green and low-carbon
modelling to examine the impact of the policies on transport carbon development in neighbouring non-pilot cities through technology and
emissions and found that they can produce spillover effects in the range knowledge cooperation and exchange. As a typical market-based envi­
of 250 km–650km while curbing carbon emissions. However, these ronmental regulation, the policies can provide favourable conditions to
studies only illustrated the spillover effects of the policies through realize the Porter Hypothesis (Borghesi et al., 2015). Companies with
spatial measures. They rarely focused on the relationship between pilot higher emissions would face more severe market pressures under the
carbon market policies and green low-carbon development in neigh­ rules. Therefore, they need to increase their R&D efforts and achieve
bouring non-pilot cities, which could easily lead to biased results in the changes in their production methods through technological innovation.
assessment. Due to the data limitations, research orientation and policy As knowledge and technology are non-competitive and exclusive, the
environment, most of the literature talked little about whether and how newly generated by firms in pilot regions would transfer to less devel­
the policies could promote green and low-carbon development in oped regions under the interaction of people and products. It compen­
neighbouring non-pilot cities. sates for the lack of innovation resources in other regions. The spillover
As an important tool to achieve“double carbon” goal in China, the of knowledge and technology is influenced by geographical distance,
carbon market pilot policy has two important meanings. First, policy and the diffusion of technology is distinctively localised (Bottazzi and
implementation could provide practical experience for the preparation Peri, 2003). Also, face-to-face communication is the only effective way
of national carbon market (Wu et al., 2017). Second, such policy could to transmit some tacit knowledge (Dong et al., 2020). In such circum­
generate a “demonstration effect”, being a more important aspect ac­ stances, the externalities of the policy are likely to increase with
cording to the present study. The effects of such policy in the pilot areas decreasing geographical distance. In other words, the spillover effects
could be disseminated through interactive exchanges and media are stronger in neighbouring non-pilot cities. When a city is close to a
coverage, which could inspire other cities. Therefore, scientifically pilot carbon market city, it has a certain degree of latecomer advantage.
evaluating and summarising the effects from the perspective of neigh­ It would have more opportunities to acquire cutting-edge technology
bouring non-pilot cities is an important task for the pilot regions to and development experience through visits, observations, learning ex­
implement the concept of low-carbon development and achieve a green changes and trade interactions for its green and low-carbon
and low-carbon transition. Furthermore, it has an important theoretical development.
and practical significance for summarising the Chinese experience of Secondly, carbon market pilot policies can have a “demonstration
“pilot-demonstration-promotion”, and for improving the key mecha­ effect” on neighbouring non-pilot cities. Specifically, pilot cities in the
nism of the “double carbon” goals. Under the active promotion of the carbon market have been given the honour of “first mover” in a
national carbon market, it is crucial to assess the combined effects more decentralised system where political and economic incentives coexist,
comprehensively. increasing the pressure on neighbouring non-pilot cities to make the
The marginal contributions of this paper are mainly in three aspects. green transition (Fredriksson and Millimet, 2002). This pressure for a
Firstly, it constructs a fundamental theoretical framework for examining green transition has two effects. For the first effect, officials from the
the relationship between carbon market pilot policies and green low- neighbouring non-pilot cities would adopt a more aggressive “upward
carbon development in neighbouring non-pilot cities. What’s more, it competition” strategy to gain an advantageous position under high
deepens the action mechanism of the externalities. This paper not only pressure of the political assessment. They would use the carbon market
broadens the evaluation of the effectiveness of carbon market pilot pilot model to raise environmental standards to reflect the imple­
policies but also compensates for the defect of the current literature, mentation of central policies (Vogel, 1997) to achieve green and
which mainly focused on policy implementation place. low-carbon development. For the second effect, the increased environ­
Secondly, this paper proposes a perfection index for carbon market mental pressure would also lead to intensified competition for resources.
mechanism, and explore its policy effects in two dimensions: carbon The neighbouring non-pilot cities failing to access carbon market pol­
market initiation and carbon market mechanism perfection. Further­ icies must be of shortcomings in environmental governance. The
more, we introduce government attention and urban wealth sources into shortcomings will undoubtedly reduce their environmental legitimacy
the analytical framework to explore the strategic choices of neigh­ with higher pollution costs and a disadvantage in the competition for
bouring non-pilot cities from the perspective of urban heterogeneity. resources (Alrazi et al., 2015). Thus, the only way for neighbouring
The existing studies neglected the incomplete development of carbon non-pilot cities to narrow the gap and achieve a green and low-carbon

2
X. Yu et al. Energy Policy 183 (2023) 113803

transition is to continuously learn from the advanced technologies and put competitive pressure on neighbouring non-pilot cities, leading
experiences of the pilot regions. Given this, this paper proposes hy­ neighbouring local governments to respond or adopt similar policies. In
pothesis H1. addition, as knowledge and technology are non-competitive and exclu­
sive, the implementation of the carbon market pilot policy can spill over
H1. Carbon market pilot policies can promote green and low-carbon
to neighbouring cities through the interaction of people and products
development in neighbouring non-pilot cities.
with reducing emission technologies and development experience.
The market mechanism is the core instrument for carbon market Therefore, this paper argues that the policy spillover effect of carbon
policies to exert the emission reduction effects. Therefore, the degree of markets on urban green and low-carbon development is reflected in two
perfection of the carbon market reflects the ease of factor flow from ways.
inefficient to efficient enterprises. According to the actual situation in Firstly, as “carbon reduction pioneers” approved by the central
the pilot regions of the Chinese carbon market, Huang et al. (2022) government, the pilot cities are more likely to become the targets of
found significant differences in both the price of carbon quota trading, policy emulation, providing the advanced experience for neighbouring
the scale of trading and market activity. With a specific analysis of non-pilot cities. When a local government adopts a corresponding pol­
carbon market pilot cities in China, Yu and Zhang (2021) find that the icy, it will put “horizontal competitive pressure” (Zhu and Zhang, 2016)
effects of the policy are more pronounced in Beijing, Tianjin and on governments in the same promotion-competing environment, lead­
Shanghai. Consequently, as the national carbon market effort becomes ing to similar policies for the neighbouring local governments. As
fully operational, the impact of policies on the green low-carbon “carbon reduction pioneers”, the pilot cities could use market mecha­
development of neighbouring non-pilot cities may vary depending on nisms to reduce the institutional transaction costs of polluting enter­
the degree of carbon market sophistication. When it was insufficiently prises through a coordinated package of environmental policies,
developed in pilot cities, some companies may seek monopoly gains on effectively achieving green and low-carbon development. Inspired by
the market by buying and hoarding large amounts of emission rights. It the pilot regions, neighbouring non-pilot cities would emulate the
would reduce the efficiency of carbon market operations (Goeree et al., advanced policies and systems, prepare detailed energy-saving and
2010). emission-reduction plans, and increase policy support for environmental
This inefficient market operation would make it difficult for the price protection to avoid lagging (Konisky, 2007). Guided by this develop­
mechanism of the carbon market policy to be truly effective. It would ment strategy, it will not only reduce the institutional transaction costs
not only affect the incentive of enterprises in the pilot regions to of the emitters in terms of compliance, distribution of rights and in­
participate in the trading of allowances (Yu and Zhang, 2021), but will terests, and transformation of technical achievements, but also make
also send negative signals to the emitters in neighbouring non-pilot re­ enterprises focus more on internal growth. In turn, it will lead to the
gions. Such a signal would lead neighbouring non-pilot regions to doubt green and low-carbon development of the whole city.
the effect, leading them to a state of just paying attention to compliance Secondly, the spillover can promote the green and low-carbon
rather than better low-carbon development strategies with the exter­ transition of neighbouring non-pilot cities through a learning mecha­
nalities. It would also largely reduce the “point-to-surface” effect of the nism. When a pilot region has gained relevant successful experience,
policy. Conversely, for regions with a more developed carbon market, governments in neighbouring regions may use it as a model. They would
there will be more participants in the market. It would lead to a further proactively and selectively learn from the relevant experience (Shipan
increase in market size, market liquidity and even a faster flow of factors and Volden, 2012), leading to the diffusion of policies from the pilot
because of the two effects. Firstly, the enhanced factor flows can region to neighbouring non-pilot regions. Furthermore, pilot policies are
accelerate knowledge and technology spillovers, helping to increase the not only a process of finding results and experiences, but also a process
level of matching frontier knowledge and technology with comple­ of identifying mistakes. As learners, the neighbouring governments can
mentary resources. In turn, it would enhance the innovation capacity of learn not only from the successes of the pilot areas but also from the
neighbouring non-pilot cities. Secondly, the flow of highly qualified failures (Volden, 2006). Then, they can get the more appropriate policy
personnel and other innovation factors increases the exchange and options through filtering. It shows that neighbouring non-pilot regions
cooperation between cities. It would not only help the spread of have a latecomer advantage. It can help them quickly accumulate
knowledge between different groups (Almeida and Kogut, 1999) but experience in energy efficiency and emission reduction. Furthermore, it
also generate network externalities through the interactive behaviour of can explore a new growth model to lay the foundation for a green and
peers. low-carbon transition. Because of this, this paper proposes hypothesis
Finally, it would lead to an exponential increase in productivity, H3.
innovation efficiency (Aizenman and Noy, 2007), acceleration of the
H3. Carbon market pilot policies would promote green and low-
green and low-carbon transformation of neighbouring non-pilot cities.
carbon development in neighbouring non-pilot cities through policy
In summary, regions with a well-developed market mechanism will
spillover effects.
further increase the enthusiasm for emission control and improvement
of the carbon market, . They would also accelerate the flow of factors to Polluting firms do not differ significantly from other firms in their
achieve a self-reinforcing “point-to-surface” effect. Therefore, this paper goal of minimising the cost, but rather in the negative environmental
proposes hypothesis H2. effects of their production. It leads to policy considerations regarding the
location of polluting firms (Keller and Levinson, 2002). The pollution
H2. As the carbon market mechanism improves, the externalities of
sanctuary hypothesis emphasises that the cost of environmental man­
carbon market pilot policies on neighbouring non-pilot cities have a self-
agement is an important factor in the location of firms. When a region
reinforcing function.
adopts a more environmentally regulated strategy, firms will reduce
their production costs by relocating (Fredriksson et al., 2003). Under the
2.2. Theoretical mechanisms influence of the carbon market pilot policy, neighbouring non-pilot
areas may adopt two strategies. Firstly, if neighbouring cities take
Under fiscal decentralisation system, the local government’s choices environmental protection more seriously, they will tend to adopt strict
of strategy are often influenced by other governments. In particular, environmental policies to avoid putting themselves at a competitive
geographically proximate regions are more likely to experience insti­ disadvantage and reduce the number of enterprises in industrial zones.
tutional emulation and policy spillovers due to similar resource en­ In this case, non-neighbouring non-pilot areas may maintain their
dowments and ease of information transfer (Shipan and Volden, 2012). existing environmental policies or relax standards due to less environ­
As the promotion of the carbon market policy, the spillover effect would mental pressure. They will attract investment through such

3
X. Yu et al. Energy Policy 183 (2023) 113803

environmental regulatory differences, achieving economic growth. 3.2.2. Independent variables

Therefore, polluting firms are prefer to choose corporate relocation to The implementation of carbon market policies in the neighbouring
earn the institutional difference due to cost-maximising conditions, and regions of cities is the independent variable (DID) of this paper, which
thus triggering the pollution migration phenomenon (Wu et al., 2017). takes the value of 1 if the neighbouring regions are Beijing, Shanghai,
Secondly, if neighbouring non-pilot areas heavily depend on Tianjin, Shenzhen, Guangdong, Chongqing, Hubei and Fujian as pilot
polluting industries, the number of industrial enterprises is more likely regions and 0 if not. Being consistent with the literature (Yu and Zhang,
to increase. Cities depending on polluting industries are likely to have a 2021), this paper assigns a value of 1 to the year in which the carbon
severe path dependency in their economic growth model and will be market policy is launched and each year thereafter, and 0 otherwise.
more inclined to attract firms with lenient environmental policies to However, due to the start-up time of the second batch of pilot cities
achieve economic benefits. Theoretically, the carbon market pilot policy being close to the end of 2012, it is difficult for the policy to have an
imposes hard environmental constraints on cities, pushing up the pro­ effect in the same year. Therefore, this paper treats it as a lagged period,
duction costs of enterprises. It would trigger the nearby relocation of so the non-pilot cities adjacent to the second batch of carbon market
industrial enterprises. Firms, as finite rational economic agents, are pilots start to be affected in 2013. Based on the list of carbon market
more inclined to relocate over short distances for profit maximisation pilot cities, after matching and filtering, a total of 46 carbon market pilot
(Conroy et al., 2016). Given this, this paper presents hypotheses H4a and cities, 63 neighbouring non-pilot cities and 174 non-neighbouring
H4b. non-pilot cities were obtained.2
H4a. Carbon market pilot policies promote green and low-carbon
3.2.3. Control variables
development in neighbouring non-pilot cities by reducing the
This paper refers to Cui et al. (2018) and selects control variables
number of industrial enterprises.
from the following aspects: level of industrial structure (Industry),
H4b. Carbon market pilot policies discourage green and low-carbon measured with the ratio of value added of secondary industry to GDP;
development in neighbouring non-pilot cities depending on polluting level of foreign openness (FDI), measured with the logarithm of foreign
enterprises by increasing the number of industrial enterprises. direct investment; degree of marketization (Market), measured with the
marketization index; the level of urban innovation (Innovation),
3. Model description of variables and data measured with the logarithm of the total number of patent applications;
the level of fiscal dependence (Financial), measured with the logarithm
3.1. Model construction of local general public budget revenue; and the level of urban greening
(Clean), measured with the logarithm of green space. To control the
The carbon market can be used as a “quasi-natural experiment”. impact of each province’s energy intensity target on the policy, this
Given the different policy start-up times of eight pilot regions, this paper paper introduces interaction terms between corresponding years and the
uses a multi-period double difference method to test the impact of car­ energy efficiency and emission reduction targets of each province during
bon market policies on the green and low-carbon development of the 11th, 12th and 13th Five-Year Plan periods, expressed as t1, t2 and
neighbouring regions. The model is set up as follows. t3.
Yi,t = α + β1 DIDi,t + β2 controli,t + γi +μt + εi,t (1)
3.3. Data sources
i and t denote area and time respectively, and Yi,t represents the
green and low-carbon development of neighbouring cities, including the This paper uses a sample of 283 cities in China from 2006 to 2017 to
scale of carbon emissions (CO2) and carbon sequestration by terrestrial assess whether pilot carbon market policies can achieve a “point-to-
vegetation (CST). DIDi,t is carbon market policy’s launch of neighbour­ surface” effect. The reasons are as follows.3 Carbon emissions and urban
ing regions. DIDi,t = (treati,t × posti,t ), treati,t is a dummy variable for the terrestrial vegetation sequestration data are from the China Carbon
presence of a carbon market pilot area in the neighbouring areas. postj,t is Emissions Database (CEADS). The carbon market trading volume,
a dummy variable for the start-up time of carbon market policy. controli,t trading value, number of trading days and carbon price are from the
is a control variable that affects the scale of carbon emissions or the China Carbon Emissions Trading Network. The night lighting data is
amount of carbon sequestered by terrestrial vegetation and varies with i from the National Oceanic and Atmospheric Administration (NOAA).
and t. γ i and μt are area-fixed effects and year-fixed effects respectively, The marketability index is from The China Marketability Index Report by
which control the effects of factors that vary with area and time on green Province (2016). Other regional-level data are from The China Urban
low-carbon development. εi,t is a random error term. Statistical Yearbook, China Regional Statistical Yearbook, China Energy
Statistical Yearbook and China Industrial Statistical Yearbook. The results
3.2. Description of variables of the descriptive statistics are shown in Table 1.

3.2.1. Dependent variables 4. Empirical results

This paper measures the green and low-carbon development of cities
in two dimensions: (1) scale of carbon emissions; (2) amount of carbon 4.1. Parallel trend test
sequestered by terrestrial vegetation. In terms of the carbon emission
scale, this paper refers to Hu et al. (2020), which uses the logarithm of To ensure the validity of the DID model estimation, the imple­
the total carbon emission of a city as a measurement. In terms of mentation of carbon market policies in the neighbouring areas is treated
terrestrial vegetation carbon sequestration, this paper draws on previous with six periods forward and six periods backwards respectively prior to
literature (Chen et al., 2020), and uses the particle swarm the models. The results are reported in Fig. 1. There is a significant
optimization-back propagation (PSO-BP) algorithm for annual synthe­
sis, reprojection, resampling and cropping of images from both data sets,
with outliers removed and time series corrected, to measure the amount 2
the map information downloaded from the standard map service website of
of terrestrial vegetation carbon sequestration (CST) in urban China, and the Ministry of Natural Resources of China with the review number GS (2020)
The CST was then logarithmically processed. 4636, this paper manually compiled the list of areas implementing carbon
market policies and their neighbouring non-pilot cities.
3
The reasons for choosing 2006–2017 as the sample for the study are
described in Appendix Note.

4
X. Yu et al. Energy Policy 183 (2023) 113803

Table 1 local carbon dioxide emissions, increase the amount of carbon seques­
Descriptive statistics. tered by terrestrial vegetation, and help promote the green and low-
Variables Obs Mean Std. Dev Min Max carbon development of the city. Columns (3) and (4) of Table 2 show
the regression results for the neighbouring non-pilot cities. We can find
CO2 3396 6.001 1.178 2.144 9.508
CST 3396 7.571 0.903 3.877 10.869 that the implementation can not only effectively reduce the scale of
DID 3396 0.067 0.250 0 1 carbon emissions in neighbouring cities, but also significantly improve
Industry 3396 0.496 0.122 0 0.909 the carbon sequestration capacity of ecosystems and achieve a green and
FDI 3396 10.802 2.388 0 16.887 low-carbon transition. Hypothesis H1 is verified.
Market 3396 9.896 2.475 3.037 17.473
Innovation 3396 6.854 1.775 1.609 12.019
Financial 3396 3.517 1.585 0.004 13.562 4.3. Analysis of the variability of carbon market completeness
Clean 3396 8.009 1.186 2.339 12.028
As the approach of Wei et al. (2010), we further compare the dif­
ferences in the impact of different levels of carbon market sophistication
on green and low-carbon development in neighbouring non-pilot areas
in three dimensions: market size, market liquidity and P/E ratio.4 Thus,
we use the 25% and 75% quartiles of market perfection as thresholds to
classify pilot cities into primary, secondary and tertiary levels. Then, we
construct city level dummy variables for carbon market policy to explore
the external effects of carbon markets with different degrees of perfec­
tion. The specific regression results are shown in Table 3. It shows that
regardless of whether the explanatory variable is carbon dioxide emis­
sions or carbon sequestered by terrestrial vegetation, the pilot cities of
the first, second and third levels of carbon markets can significantly
contribute to the green and low-carbon development of neighbouring
areas. The “point-to-surface” effect of the pilot cities of the first level is
significant, followed by the second level and the smallest regression
coefficient for the third level. It can be deduced that the improvement of
the carbon market mechanism in the pilot areas will accelerate the green
and low-carbon transition in the neighbouring areas, and there is a
self-reinforcing mechanism for the “point-to-surface” effect.

4.4. Robustness tests

Fig. 1. Results of the parallel trend test.

4.4.1. Approach of instrumental variables (IV)

negative impact after the carbon market pilot. The variability becomes
To address the issue of policy endogeneity (possible omitted vari­
more pronounced over time, consistent with the parallel trend
ables and reverse causality), this paper uses an IV approach for robust­
hypothesis.
ness testing, referring to Hering and Poncet (2014), selecting the air
circulation coefficient as the IV. Jacobson (2002) showed that pollution
4.2. Basic regression results dispersion may be influenced by two meteorological factors: wind speed
and mixed layer height. Faster winds contribute to the horizontal
The policy premise for the “point-to-surface” effect is the promotion dispersion of pollutants, and the mixed layer height may affect the
of local green and low-carbon development and the spillover of the local vertical dispersion of pollutants. The smaller the value of the air circu­
experience, technology and policies to the neighbouring areas. There­ lation coefficient, a product of the unit wind speed and the height of the
fore, this paper first tests the effectiveness of the carbon market policy in boundary layer, the lower the probability of pollutant dispersion, the
the pilot city and then verifies the “point-to-surface” effect. The greater the concentration of pollutants monitored, the more urgent the
regression results are shown in Table 2, where columns (1) and (2) need for environmental regulation, and the greater being selected as a
reflect the effect of the policy on the carbon market pilot areas as a pilot area. Therefore, the air circulation coefficient is negatively corre­
sample. It shows that the policy can significantly reduce the scale of lated with the probability of inclusion in the pilot area, satisfying the
correlation assumption of a valid instrumental variable. In addition, the
Table 2 air circulation coefficient depends on meteorological and geographical
Basic regression results. conditions, meeting the exogeneity requirement of the IV. The two-stage
(1) CO2 (2) CST (3) CO2 (4) CST least squares (2SLS) are used in this paper, and the results of the
DID − 0.218*** 0.196***
second-stage regression are reported in columns (1) and (2) of Table 4.
(0.039) (0.072) It can be found that carbon market policies have a significant enhancing
DID − 0.147*** 0.014** effect on the green and low-carbon development of neighbouring areas,
(0.031) (0.006) which validates hypothesis H1 again.
Control YES YES YES YES
City YES YES YES YES
Year YES YES YES YES 4.4.2. Exclusion of special samples
_Cons 0.529*** 8.394*** 0.422*** 7.469*** The presence of special samples may lead to biased regression
(0.093) (0.135) (0.104) (0.054)
Adj-R2 0.748 0.635 0.657 0.545
N 3396 3396 3396 3396 4
Specifically, the size of the carbon market is measured by two indicators:
Note: Control variable coefficients are not reported due to space constraints. trading volume and trading value. The market liquidity is measured by the
Values in brackets are robust standard errors; ***, ** and * indicate significance number of trading days in each carbon market, which is the same as the study
at the 1%, 5% and 10% levels respectively. Control variables, time and region by Zhao et al. (2016). The P/E ratio is calculated as follows: P/E ratio =
are all controlled for. The tables below are identical. ln(Carbon price yesterday) − ln(Carbon price today).

5
X. Yu et al. Energy Policy 183 (2023) 113803

Table 3
Test of variability in carbon market completeness.
(1) (2) (3) (4) (5) (6)

CO2 CST

DID*Perfect1 − 0.635*** (0.067) 0.164*** (0.043)

DID*Perfect2 − 0.054*** (0.013) 0.098*** (0.020)
DID*Perfect3 − 0.034*** (0.009) 0.041*** (0.004)
Control YES YES YES YES YES YES
City YES YES YES YES YES YES
Year YES YES YES YES YES YES
Adj-R2 0.756 0.757 0.752 0.646 0.547 0.686
N 3396 3396 3396 3396 3396 3396

Table 4
Robustness tests.
(1) CO2 (2) CST (3) CO2 (4) CST (5) CO2 (6) CST (7) CO2 (8) CST

DID − 0.013*** 0.018** − 0.157*** 0.075*** − 0.142*** 0.026** − 0.253*** 0.119** (0.058)
(0.002) (0.008) (0.041) (0.004) (0.042) (0.011) (0.039)
APAPPC*time 0.087** (0.037) − 0.105**
(0.052)
PDR* time 0.109*** (0.022) − 0.038 (0.034)
LCP* time − 0.199*** 0.277***
(0.029) (0.040)
Control YES YES YES YES YES YES YES YES
City YES YES YES YES YES YES YES YES
Year YES YES YES YES YES YES YES YES
Adj-R2 0.342 0.341 0.592 0.739 0.671 0.725 0.763 0.325
N 3396 3396 3396 3396 3396 3396 3396 3396

results. To effectively identify the externalities of the policies, cities with paper removed policies that may affect green and low-carbon develop­
strong economies such as Beijing, Shanghai and Shenzhen are excluded ment one by one. The main ones include APAPPC PDR and LCP.5 To
from the data. Considering that Chongqing, the only municipality avoid interference from above policies, this paper introduced interaction
directly under the central government in the western region among the terms between the Action Plan for Air Pollution Prevention and Control
eight pilot regions, may have more tasks to reduce emissions, this paper (APAPPC), Pilot Emissions Trading (PDR) and Low Carbon Pilot Policy
excludes it. In addition, Fujian Province officially launched its carbon (LCP) and time trends, respectively. The regression results are reported
market policy at the end of 2016, which may be difficult to reflect the in columns (7) and (8) of Table 4. It can be found that the regression
effect within the sample interval. Therefore, these two regions are results do not change significantly after controlling for the potential
excluded from this paper. The regression results are reported in column impact of contemporaneous policies, indicating that the conclusions of
(3) and column (4) of Table 4, and the conclusions of this paper are this paper are robust.
robust.
4.4.5. Placebo test
4.4.3. Propensity score matching (PSM)-DID estimation To further test the reliability of the regression results and to ensure
To further identify the net effect between carbon market policies and the carbon market pilot was not affected by the omitted variables, this
green low-carbon development in neighbouring non-pilot areas, this paper uses the practice of non-reference substitution tests to randomly
paper deals with the selectivity bias problem through PSM. As the effects select the experimental and control groups for spatial placebo tests,
of the carbon market pilot policy were already visible before 2011 and referring to Chetty et al. (2009) and Ferrara et al. (2012). Fig. 2 shows
the relevant variables in neighbouring areas may have changed with the the kernel density plot obtained by this method. The mean value of the
operation of the carbon market policy, this paper only matches the coefficient after randomisation is − 0.0001, which is very close to zero
sample from 2006 to 2010 before conducting a multi-period double compared to the baseline results, indicating that the baseline regression
difference test. The matched results are shown in the Appendix. The results of this paper can pass the robustness test.
standard deviations after matching are all less than 5% and the t-sta­
tistics became insignificant, indicating that there is no significant dif­ 5. Mechanistic analysis
ference between the matched treatment and control groups. It ensures
the accuracy of the matched results and the reliability of the estimation 5.1. Policy spillover
results. The regression results are reported in columns (5) and (6) of
Table 4. It can be found that carbon market policies can reduce CO2 To test the existence of a policy spillover effect, this paper measures
emissions, enhance the carbon sequestration of terrestrial vegetation in environmental policy by taking the logarithm of the amount of envi­
neighbouring areas and promote green and low-carbon development at ronmental investment in the city, drawing on the approach of Nie et al.
a significant level of 1%, indicating that the conclusions of this paper are (2015), The model is set as follows.
robust.
EIi,t = θ + η1 Peerj,t + η2 control1i,t + γi +μt + εi,t (2)
4.4.4. Cutting out policy interference
Where EIi,t represents the amount of environmental investment in the
To avoid the biased estimation results caused by other policies, this

5
Details of the policy can be found in the fourth point of note in the annex.

6
X. Yu et al. Energy Policy 183 (2023) 113803

model (1) to examine whether the mediating mechanism holds. The

regression results are reported in columns (1) to (3) of Table 6. It in­
dicates that the initiation of the pilot policy significantly reduces the
number of industrial enterprises in neighbouring areas. The coefficients
of the pilot policy regression results in columns (2) and (3) are signifi­
cantly lower, indicating that relocation of industrial enterprises is an
important way to promote green and low-carbon development in
neighbouring areas. Obviously, the neighbouring areas were influenced
and tended to set stricter conditions for environmental regulation in the
process of promoting the policy. In this situation, the production costs of
the “three high” enterprises6 are further pushed up and the enterprises
are forced to move out due to the constraints on their survival and
development. On the other hand, enterprises remaining in the city are
more actively exploring clean and green production technologies under
strict environmental policies, which in turn leads to the emergence of
low-pollution and high-tech enterprises. It effectively contributes to the
city’s transition to a green and low-carbon development model. Given
this, hypothesis H4a of this paper is confirmed.
Fig. 2. Placebo test chart.
Carbon market pilot policies have a limited effect on the green and
low-carbon development of neighbouring cities, which depends on
city i; Peerj,t is the amount of environmental investment in other cities polluting industries (Van der Ploeg, 2011). Thus, this paper uses a
adjacent to i; and the remaining variables are explained in the model (1). sample of such kinds of cities, neighbouring non-pilot cities with more
The regression results are reported in Table 5. The full sample regres­ than 50% of assets in polluting industries, to perform a mediation
sion results in column (1) show that the choice of government envi­ mechanism test. The regression results are reported in columns (4) to (6)
ronmental policies has a significant positive effect on neighbouring of Table 6. It can be observed that the launch of the policy significantly
areas, indicating that there is a significant imitation effect of environ­ increased the number of industrial enterprises in such kinds of cities,
mental policy-making between neighbouring areas. Columns (2) and (3) inhibiting the green and low-carbon transformation of these neigh­
show the regression results for the presence or absence of carbon market bouring non-pilot areas. It suggests that such cities have a weak envi­
pilots in neighbouring areas. It shows that the imitation effect of the ronmental foundation. In addition, the geographical distance may be a
system is significant in the sample with carbon market pilots, but not in critical factor in the proximity transfer of pollution (Conroy et al., 2016),
the sample without carbon market pilots. This suggests that carbon and given this, hypothesis H4b of this paper is confirmed.
market pilot policies can have an effective policy spillover effect on
neighbouring non-pilot areas, which in fact promotes their green and 6. Further discussion
low-carbon development.
6.1. Whether the principal officer is in a critical period for promotion
5.2. Relocation of industrial enterprises
The ability to effectively exploit the policy externalities of the carbon
market pilot of neighbouring areas is largely related to the tenure
Considering that industrial enterprises are the major sources for
characteristics of the officials in charge. If the officials are in a critical
pollution, this paper explains the mechanism of the impact of the carbon
period for promotion, they are more likely to increase their focus on
market pilot policy on neighbouring non-pilot areas from the perspec­
energy efficiency and emissions reduction for promotion (Li and Zhou,
tive of industrial enterprise relocation. We would use a mediating effects
2005). According to the Interim Provisions on the Tenure of Office of
model to test its effectiveness empirically. The specific model is set up as
Party and Government Leading Cadres and other relevant regulations,
follows.
officials are mostly in charge for five years. Therefore, this paper in­
Firmi,t = ρ + ρ2 DIDi,t + ηcontrol+μi,t +σ i,t + εi,t (3) troduces the mayors as the principal officials in charge and defines the
promotion-critical period if the officials are in the first three years,
Yi,t = ρ + ρ3 DIDi,t + λFirmi,t + ηcontrol+μi,t +σ i,t + εi,t (4) otherwise, it is a non-promotion-critical period. Based on this, we divide
the sample of neighbouring non-pilot areas. The regression results are
Where Firmi,t is the number of industrial enterprises. This paper collates reported in Table 7, which shows that the principal officials in a
the number of industrial enterprises according to The China City Statis­ promotion-critical period are more likely to respond positively to the
tical Yearbook to reflect the changes in the number of enterprises in each national energy conservation and emission reduction policies. Further­
city. Model (3) is used to test the impact of the policy on the relocation of more, they would utilize the positive spillover from the carbon market
industrial enterprises, and model (4) adds mediating variables based on pilot policies to promote the green and low-carbon development of their
cities.
Table 5
Tests of policy spillover mechanisms. 6.2. Regional heterogeneity
(1) (2) (3)
Studies have identified significant differences in the allocation of
Full Carbon market pilot Carbon Free Market
available Pilot allowances, trading conditions and market liquidity of the pilot carbon
markets in China (Zhang et al., 2020). Not all carbon markets have
Peer 0.018** 0.277** (0.049) 0.016 (0.009)
(0.008)
well-developed market mechanisms, which may affect their externalities
Control YES YES YES to neighbouring non-pilot areas. Based on the basic regression, the
City YES YES YES
Year YES YES YES
Adj-R2 0.591 0.417 0.587 6
The “three high” enterprises refer to those with the characteristics of “high
N 3396 552 2844
pollution, high energy consumption and high water consumption”.

7
X. Yu et al. Energy Policy 183 (2023) 113803

Table 6
Test of relocation mechanism of industrial enterprises.
Full Wealth from polluting industries

(1) (2) CO2 (3) CST (4) (5) CO2 (6) CST
Firm Firm

DID − 0.156** (0.070) − 0.149** (0.041) 0.013** (0.006) 0.046*** (0.004) 0.026*** (0.008) − 0.016** (0.006)
Firm − 0.030** (0.013) − 0.056*** (0.008) 0.037** (0.019) − 0.037*** (0.012)
Control YES YES YES YES YES YES
City YES YES YES YES YES YES
Year YES YES YES YES YES YES
Adj-R2 0.540 0.756 0.423 0.363 0.552 0.382
N 3395 3395 3395 1926 1926 1926

Table 7 Table 9
Test of whether the principal official is in a critical period for the promotion. Synergies of carbon market pilot policies.
Critical period for promotion Non-promotion critical period (1) SO2 (2) TSO2 (3) PM2.5

(1) CO2 (2) CST (3) CO2 (4) CST DID − 0.119*** (0.023) − 0.214*** (0.014) − 0.185*** (0.016)
Control YES YES YES
DID − 0.349*** 0.121*** 0.163 0.211
Adj-R2 0.421 0.332 0.512
(0.074) (0.016) (0.179) (0.342)
N 3396 3396 3396
Control YES YES YES YES
Adj-R2 0.761 0.464 0.763 0.323
N 1184 1184 2211 2211
market pilot policies have significant externalities while achieving a
green and low-carbon transition in the region. This conclusion still holds
interaction term between the region where the carbon pilot market is after a series of robustness tests.
located and the carbon market pilot policy is introduced in this paper. Second, the self-reinforcing effect of the analyzed policy is shown by
The explanatory variable is CO2 emission, being used to test whether the “point by point” effect, and the pilot carbon market policy can be
each carbon market pilot policy has a “point-to-surface” effect. The self-reinforcing through the improvement of the carbon market. The
regression results are reported in Table 8, which shows that Beijing has results are consistent with the findings of Bel and Joseph (2015) and
the strongest “point-to-surface” effect followed by Shanghai. others that the degree of carbon market mechanism improvement is an
important factor affecting the effect of emission reduction.
6.3. Synergies Third, this study finds that carbon market pilot policies not only
generate external effects through direct mechanisms such as technology
To further test the synergistic effect of carbon market policies on cooperation and exchange and demonstration effects, but also promote
carbon reduction and pollution reduction in neighbouring non-pilot green and low-carbon transformation in neighbouring non-pilot cities
regions, the explanatory variables are replaced with sulphur dioxide through intermediary mechanisms such as policy spillover effects and
emissions (SO2), sulphur dioxide intensity (TSO2) and particulate mat­ relocation of industrial enterprises. This study innovates from previous
ters (PM2.5). The regression results are shown in Table 9, and it can be literature which only analyze the spatial spillover effects of carbon
found that the carbon market pilot policy has the greatest inhibitory markets and do not fully explain the spillover mechanisms (Yu and
effect on PM2.5 in low-neighbouring areas, followed by SO2. It indicates Zhang, 2021; Yu and Li, 2021). However, if neighbouring areas are
that the policy can not only reduce carbon emissions and enhances dominated by polluting industries, they can inhibit green low-carbon
carbon sequestration in neighbouring non-pilot areas, but also has a development by attracting industrial firms, which provides more
significant inhibitory effect on other environmental pollutants. It has a empirical evidence for the pollution refuge hypothesis.
synergistic effect of reducing pollution and carbon, which is consistent Fourth, unlike the existing studies that analyze the heterogeneity of
with the findings of Hu et al. (2020). carbon market policy effects only from the perspective of geographical
location (Yu et al., 2022; Huang et al., 2022), this paper not only in­
7. Conclusion and policy implications corporates the characteristics of local government officials into the
analysis framework, but also comprehensively compares the external
7.1. Conclusion effects of carbon markets in “two provinces and five cities”. This paper
not only incorporates the characteristics of local government officials
This paper constructs a theoretical framework on how carbon market into the analytical framework, but also verifies the synergistic effects of
pilot policies affect the green and low-carbon development of neigh­ carbon market policies in reducing pollution and carbon emissions,
bouring non-pilot areas and uses a sample of 283 cities in China from providing a diversified perspective for existing studies.
2006 to 2017 as a research sample for empirical testing. The findings of
this paper show that, first, compared with Zhang et al.’s (2020) 7.2. Policy implications
conclusion that China’s carbon market is difficult to achieve energy
saving and emission reduction, this paper’s study finds that carbon The present study proposes several policy recommendations based

Table 8
Regional heterogeneity test.
Beijing CO2 Shanghai CO2 Tianjin CO2 Guangdong CO2 Chongqing CO2 Hubei CO2 Fujian CO2

DID − 0.770*** (0.051) − 0.537*** (0.046) − 0.154*** (0.011) − 0.353 (0.331) − 0.234* (0.124) − 0.115 (0.172) − 0.022 (0.151)
Control YES YES YES YES YES YES YES
Adj-R2 0.842 0.479 0.514 0.415 0.512 0.341 0.361
N 132 168 120 72 120 132 60

8
X. Yu et al. Energy Policy 183 (2023) 113803

on findings. First, we will continue to give full play to the external effects References
of the carbon market pilot regions and accelerate the construction and
improvement of China’s unified carbon market. It should start from the Aizenman, J., Noy, I., 2007. Prizes for basic research: human capital, economic might
and the shadow of history. J. Econ. Growth 12 (3), 261–282.
principle of wholeness, combine the pilot effect and external effect of the Almeida, P., Kogut, B., 1999. Localization of knowledge and the mobility of engineers in
pilot policy, smooth the mechanism of technical cooperation and ex­ regional networks. Manag. Sci. 45 (7), 905–917.
change and experience-sharing between regions, and drive the neigh­ Alrazi, B., De Villiers, C., Van Staden, C.J., 2015. A comprehensive literature review on,
and the construction of a framework for, environmental legitimacy, accountability
bouring cities to jointly realize green and low-carbon development and proactivity. J. Clean. Prod. 102, 44–57.
through policy spillover. At the same time, it is also necessary to Bel, G., Joseph, S., 2015. Emission abatement: untangling the impacts of the EU ETS and
accelerate the improvement of the trading mechanism, constraint the economic crisis. Energy Econ. 49, 531–539.
Borghesi, S., Cainelli, G., Mazzanti, M., 2015. Linking emission trading to environmental
mechanism and distribution mechanism of the carbon market, and innovation: evidence from the Italian manufacturing industry. Res. Pol. 44 (3),
accelerate the promotion of the construction of the national carbon 669–683.
market on the basis of summarising the experience of the pilot areas. Bottazzi, L., Peri, G., 2003. Innovation and spillovers in regions: evidence from European
patent data. Eur. Econ. Rev. 47 (4), 687–710.
Second, we should pay attention to policy spillovers and the relo­
Chen, J., Gao, M., Cheng, S., Hou, W., Song, M., Liu, X., Liu, Y., Shan, Y., 2020. County-
cation paths of industrial enterprises, and form an inter-regional model level CO2 emissions and sequestration in China during 1997–2017. Sci. Data 7 (1),
of joint prevention and control of emission reduction. On the one hand, 1–12.
future studies should comprehensively summarize the carbon market Chen, Z., Song, P., Wang, B., 2021a. Carbon emissions trading scheme, energy efficiency
and rebound effect–Evidence from China’s provincial data. Energy Pol. 157, 112507.
policy measures, paths, methods, and lessons learned to expand the Chen, Z., Zhang, X., Chen, F., 2021b. Do carbon emission trading schemes stimulate
scope and influence of the policy. On the other hand, future studies green innovation in enterprises? Evidence from China. Technol. Forecast. Soc.
should also promote the synchronization and coordination of environ­ Change 168, 120744.
Cheng, B., Dai, H., Wang, P., Zhao, D., Masui, T., 2015. Impacts of carbon trading scheme
mental regulations in neighbouring regions, to strengthen the industrial on air pollutant emissions in Guangdong Province of China. Energy sustain. develop.
division of labor and complementary advantages, weakening the 27, 174–185.
transfer of polluting enterprises to transfer power. Cities should also Chetty, R., Looney, A., Kroft, K., 2009. Salience and taxation: theory and evidence. Am.
Econ. Rev. 99 (4), 1145–1177.
clarify their own industrial strengths and the direction of environmental Conroy, T., Deller, S., Tsvetkova, A., 2016. Regional business climate and interstate
regulation, and reduce policy support for polluting transferring enter­ manufacturing relocation decisions. Reg. Sci. Urban Econ. 60, 155–168.
prises, in order to realize overall green and low-carbon development. Cui, J., Zhang, J., Zheng, Y., 2018. Carbon pricing induces innovation: evidence from
China’s regional carbon market pilots. In: AEA Papers and Proceedings, 108,
Third, rational and effective supporting policies are essential to pp. 453–457.
stimulate the externalities of carbon market policies. Government de­ Dai, S., Qian, Y., He, W., Wang, C., Shi, T., 2022. The spatial spillover effect of China’s
partments should continue to promote a multi-dimensional assessment carbon emissions trading policy on industrial carbon intensity: evidence from a
spatial difference-in-difference method. Struct. Change Econ. Dynam. 63, 139–149.
and evaluation system, increase the weight of ecological and environ­
Dong, Z., He, Y., Wang, H., Wang, L., 2020. Is there a ripple effect in environmental
mental assessment indicators in the promotion assessment, and effec­ regulation in China?–Evidence from the local-neighborhood green technology
tively stimulate the subjective initiative of local governments. Pilot innovation perspective. Ecol. Indicat. 118, 106773.
regions should give full play to their advantageous position. For Ferrara, E.L., Chong, A., Duryea, S., 2012. Soap operas and fertility: evidence from Brazil.
Am. Econ. J. Appl. Econ. 4 (4), 1–31.
example, Beijing, the capital city which has the potential to become the Fredriksson, P.G., Millimet, D.L., 2002. Strategic interaction and the determination of
center of China’s carbon market, should be prioritized when driving on environmental policy across US states. J. Urban Econ. 51 (1), 101–122.
the neighbouring regions due higher degree of perfection and better Fredriksson, P.G., List, J.A., Millimet, D.L., 2003. Bureaucratic corruption,
environmental policy and inbound US FDI: theory and evidence. J. Publ. Econ. 87
policy effects. (7–8), 1407–1430.
Goeree, J.K., Palmer, K., Holt, C.A., Shobe, W., Burtraw, D., 2010. An experimental study
CRediT authorship contribution statement of auctions versus grandfathering to assign pollution permits. J. Eur. Econ. Assoc. 8
(2–3), 514–525.
Hering, L., Poncet, S., 2014. Environmental policy and exports: evidence from Chinese
Xiaolin Yu: Conceptualization, Methodology, Writing – original cities. J. Environ. Econ. Manag. 68 (2), 296–318.
draft, Writing – review & editing. Kai Wan: Writing – review & editing, Hu, Y., Ren, S., Wang, Y., Chen, X., 2020. Can carbon emission trading scheme achieve
energy conservation and emission reduction? Evidence from the industrial sector in
Software. Qunyang Du: Writing – review & editing, Language. China. Energy Econ. 85, 104590.
Huang, W., Wang, Q., Li, H., Fan, H., Qian, Y., Klemeš, J.J., 2022. Review of recent
Declaration of competing interest progress of emission trading policy in China. J. Clean. Prod. 349, 131480.
Jacobson, M.Z., 2002. Atmospheric Pollution: History, Science, and Regulation.
Cambridge University Press.
We declare that we have no financial and personal relationships with Keller, W., Levinson, A., 2002. Pollution abatement costs and foreign direct investment
other people or organizations that can inappropriately influence our inflows to US states. Rev. Econ. Stat. 84 (4), 691–703.
work, there is no professional or other personal interest of any nature or Konisky, D.M., 2007. Regulatory competition and environmental enforcement: is there a
race to the bottom? Am. J. Polit. Sci. 51 (4), 853–872.
kind in any product, service and company. Li, Z., Wang, J., 2021. Spatial emission reduction effects of China’s carbon emissions
trading: quasi-natural experiments and policy spillovers. Chinese J. Population,
Data availability Resources Environ. 19 (3), 246–255.
Li, Z., Wang, J., 2022. Spatial spillover effect of carbon emission trading on carbon
emission reduction: empirical data from pilot regions in China. Energy 251, 123906.
Data will be made available on request. Li, H., Zhou, L.A., 2005. Political turnover and economic performance: the incentive role
of personnel control in China. J. Publ. Econ. 89 (9), 1743–1762.
Li, S., Liu, J., Wu, J., Hu, X., 2022. Spatial spillover effect of carbon emission trading
Acknowledgements policy on carbon emission reduction: empirical data from transport industry in
China. J. Clean. Prod. 371, 133529.
We sincerely thank the journal reviewers for their valuable com­ Liu, L., Chen, C., Zhao, Y., Zhao, E., 2015. China’s carbon-emissions trading: overview,
challenges and future. Renew. Sustain. Energy Rev. 49, 254–266.
ments on the manuscript. This work is supported by the Project of China Mo, J.L., Agnolucci, P., Jiang, M.R., Fan, Y., 2016. The impact of Chinese carbon
National Social Science Foundation (grant number 20BJY189). emission trading scheme (ETS) on low carbon energy (LCE) investment. Energy Pol.
89, 271–283.
Nie, P., Sousa-Poza, A., He, X., 2015. Peer effects on childhood and adolescent obesity in
Appendix A. Supplementary data
China. China Econ. Rev. 35, 47–69.
Sandoff, A., Schaad, G., 2009. Does EU ETS lead to emission reductions through trade?
Supplementary data to this article can be found online at https://doi. The case of the Swedish emissions trading sector participants. Energy Pol. 37 (10),
org/10.1016/j.enpol.2023.113803. 3967–3977.
Shipan, C.R., Volden, C., 2012. Policy diffusion: seven lessons for scholars and
practitioners. Publ. Adm. Rev. 72 (6), 788–796.

9
X. Yu et al. Energy Policy 183 (2023) 113803

Streimikiene, D., Roos, I., 2009. GHG emission trading implications on energy sector in Yu, D.J., Li, J., 2021. Evaluating the employment effect of China’s carbon emission
Baltic States. Renew. Sustain. Energy Rev. 13 (4), 854–862. trading policy: based on the perspective of spatial spillover. J. Clean. Prod. 292,
Van der Ploeg, F., 2011. Natural resources: curse or blessing? J. Econ. Lit. 49 (2), 126052.
366–420. Yu, Y., Zhang, N., 2021. Low-carbon city pilot and carbon emission efficiency: quasi-
Vogel, D., 1997. Trading up and governing across: transnational governance and experimental evidence from China. Energy Econ. 96, 105125.
environmental protection. J. Eur. Publ. Pol. 4 (4), 556–571. Yu, X., Shi, J., Wan, K., Chang, T., 2022. Carbon trading market policies and corporate
Volden, C., 2006. States as policy laboratories: emulating success in the children’s health environmental performance in China. J. Clean. Prod. 371, 133683.
insurance program. Am. J. Polit. Sci. 50 (2), 294–312. Zhang, Y.J., Peng, Y.L., Ma, C.Q., Shen, B., 2017. Can environmental innovation facilitate
Wang, P., Dai, H.C., Ren, S.Y., Zhao, D.Q., Masui, T., 2015. Achieving Copenhagen target carbon emissions reduction? Evidence from China. Energy Pol. 100, 18–28.
through carbon emission trading: economic impacts assessment in Guangdong Zhang, H., Duan, M., Deng, Z., 2019. Have China’s pilot emissions trading schemes
Province of China. Energy 79, 212–227. promoted carbon emission reductions?–the evidence from industrial sub-sectors at
Wei, Y.M., Fan, Y., Han, Z.Y., Zou, L., 2010. Energy Economics: Modeling and Empirical the provincial level. J. Clean. Prod. 234, 912–924.
Analysis in China. Center for Energy and Environmental Policy Research (CEEP), Zhang, S., Jiang, K., Wang, L., Bongers, G., Hu, G., Li, J., 2020. Do the performance and
Beijing Institute of Technology. efficiency of China’s carbon emission trading market change over time? Environ. Sci.
Wu, H., Guo, H., Zhang, B., Bu, M., 2017a. Westward movement of new polluting firms in Pollut. Control Ser. 27 (26), 33140–33160.
China: pollution reduction mandates and location choice. J. Comp. Econ. 45 (1), Zhao, X.G., Jiang, G.W., Nie, D., Chen, H., 2016. How to improve the market efficiency of
119–138. carbon trading: a perspective of China. Renew. Sustain. Energy Rev. 59, 1229–1245.
Wu, J., Kang, Z.Y., Zhang, N., 2017b. Carbon emission reduction potentials under Zhou, B., Zhang, C., Song, H., Wang, Q., 2019. How does emission trading reduce China’s
different polices in Chinese cities: a scenario-based analysis. J. Clean. Prod. 161, carbon intensity? An exploration using a decomposition and difference-in-
1226–1236. differences approach. Sci. Total Environ. 676, 514–523.
Yang, L., Li, F., Zhang, X., 2016. Chinese companies’ awareness and perceptions of the Zhu, X., Zhang, Y., 2016. Political mobility and dynamic diffusion of innovation: the
Emissions Trading Scheme (ETS): evidence from a national survey in China. Energy spread of municipal pro-business administrative reform in China. J. Publ. Adm. Res.
Pol. 98, 254–265. Theor. 26 (3), 535–551.
Yang, L., Li, Y., Liu, H., 2021. Did carbon trade improve green production performance?
Evidence from China. Energy Econ. 96, 105185.

10