Does Carbon Emissions Trading Policy Improve Inclusive Green Resilience in Cities? Evidence from China

Abstract

With the impact of external globalization uncertainties and the pressure of internal national environmental policies and markets, maintaining inclusive green resilience while coordinating economic, environmental, and social systems is critical for achieving green and sustainable urban development. We define inclusive green resilience for cities in this research and build a system of inclusive green resilience indicators. The DID model and entropy approach were used to examine the impact of carbon trading policies on inclusive green resilience in 184 representative Chinese cities from 2008 to 2018, and PSM-DID was utilized for further validation. According to the findings of the study, carbon emissions pricing policies can considerably increase inclusive green resilience in cities. Mechanism verification demonstrates that carbon trading programs improve inclusive green resilience in cities through industrial restructuring, technical innovation capability, and employment benefits. According to the city heterogeneity study, the implementation of the carbon emissions trading system has a scale effect and significant urban functional differences, and its impact on inclusive green resilience of cities is greater in large and medium-sized cities and non- resource-based cities. This research offers a new way of thinking about inclusive green resilience as well as empirical data for future sustainable policy development.

1. Introduction

Under the background of global warming, the world economy is facing a contradiction between high energy consumption and low-carbon economic transition [1]. As the largest developing country, China is also under enormous pressure to achieve the goals of carbon emissions reduction, economization, and environmental protection in its efforts to promote rapid economic growth and improve people’s livelihoods. The outline of China’s 14th Five-Year Plan clearly states that it aims to achieve a “carbon peak” by 2030 and the “3060 Goal Vision” of carbon neutrality by 2060. The “3060 Goal Vision” emphasizes the importance of cities reducing carbon emissions, managing the relationship between environmental and economic systems while maintaining social equity and justice, improving the affordability of urban economic and social development, and ultimately promoting inclusive green resilience [2,3]. With the legislation and execution of a large number of environmental regulatory regulations, China’s carbon emissions intensity will have decreased by 48.4% from 2005 levels by the end of 2020. Nonetheless, despite domestic policy support, China confronts huge obstacles in terms of inclusive green resilience. The key barriers to China’s inclusive green resilience are a lack of “inclusive resilience” and the difficulties in harmonizing it with “green development resilience [4,5]. Therefore, how to promote inclusive green resilience in cities while improving low-carbon economic efficiency has become an issue demanding an urgent solution for China.

The notion of inclusive green resilience in cities is taken primarily from the concept of resilience in ecology, with a focus on dynamic adaptive adjustment pathways for the long-term sustained expansion of urban low-carbon economic and social systems [3,6,7]. Rizzo believes that so-called “green resilience” is the integration of technological solutions and resilience thinking that emerged to solve urban ecological problems [8]. Shokry et al. propose that green resilience aims to reduce the vulnerability of socially vulnerable groups to climate risks and impacts [9]. The article argues that urban inclusive green resilience refers to the fact that, in the face of current internal and external shocks, such as resource and environmental constraints, a city recovers by upgrading the level of economic and social inclusiveness, as well as environmental development, and then transitions to a new path of growth and a state of sustainable development. The academic community currently uses the core variable method and the indicator system method to measure inclusive green resilience capacity. When compared to the core variable method, the indicator system method can comprehensively and systematically measure inclusive green resilience capacity. The indicator system method separates green resilience into numerous aspects and selects indicators to quantify each dimension. Shokry et al. constructed an index system from the three dimensions of exposure, sensitivity, and adaptability to measure urban green resilience [10]. However, investigating the measurement of the level of inclusive green resilience by creating indicators based solely on the operational process of green resilience, neglecting the interaction between economic, social, and green development, still poses a problem for scholars. As a result, this essay develops an evaluation index system based on three factors: economic resilience, social inclusive resilience, and ecological environment resilience. The existing literature on the factors influencing urban inclusive green resilience focuses primarily on urban development activities [11,12], and there is still potential to investigate other aspects impacting inclusive green resilience capability. Furthermore, although a large number of scholars have confirmed that environmental policies have a positive influence on the level of urban inclusive green development through qualitative and quantitative analyses, there is still a research gap in exploring the impact of environmental policies on inclusive green resilience capacity.

Carbon emissions trading policy has garnered substantial attention from academics in recent years due to its operating mechanism and low-carbon development as a market incentive-based environmental strategy. Carbon emissions trading policies, according to relevant researchers, can greatly improve regional economic efficiency [13], improve energy efficiency [14], and play a vital role in stimulating low-carbon technology innovation in businesses [15]. In terms of research methods, some researchers have used qualitative and theoretical research approaches to examine the impact of carbon emissions trading policies, such as developing an impact evaluation index system and analyzing and evaluating the consequences of policy implementation [16]. Some researchers analyze the influence of carbon emissions trading programs using quantitative methods such as the instrumental variable method, comprehensive control method, dynamic computable general equilibrium model (CGE), DID model, PSM model, PSM-DID model, and so on [17,18,19]. The policy assessment impact is disturbed when the process of assessing policy outcomes is influenced by some observable or unobservable confounding factors. To address this type of issue, the DID model may reduce the influence of confounding factors by inserting dummy variables to increase policy assessment accuracy; hence, the DID model has been widely employed in the policy evaluation process. The PM-DID model is also valued in the policy analysis process because it combines the benefits of the DID and the PSM while effectively avoiding the difficulties of selection bias and endogeneity [20,21].

To sum up, although there is a considerable amount of literature on carbon emissions trading policy and inclusive green resilience, some limitations still exist. First of all, although previous studies on carbon emissions trading policy have discussed the impacts of carbon emissions trading policy on enterprises, economy, and ecological environment from different perspectives, few studies have discussed the impacts of carbon emissions trading policy on urban economic, social, and environmental coordination, especially on urban inclusiveness, from the perspective of urban integrity. Secondly, in terms of measuring inclusive green resilience, most scholars mainly focus on the level of green resilience, exploring the level of urban green development space and green infrastructure construction, and few studies have deeply captured and explored the development status of urban economic, social, and ecological comprehensive resilience capacity. In addition, there is an urgent need for empirical studies on whether urban environmental policies have a significant impact on improving inclusive green resilience.

In order to bridge the research gap, based on the current research content, this paper uses the PSM-DID method to thoroughly investigate the response of inclusive green resilience to the carbon emissions trading policy and conducts an in-depth analysis of its influence mechanism. Based on China’s actual national conditions, the heterogeneous impact of carbon emissions trading policy on cities in different categories is analyzed in depth.

2. Theoretical Analysis and Research Hypothesis

2.1. Mechanism of Analysis of Carbon Emissions Trading Rights Policy for Inclusive Green Resilience

In this section, we describe the theoretical mechanism by which emissions trading rights policy impacts inclusive green resilience. We extend the theoretical mechanism based on the previous carbon emissions trading rights policy, including three main aspects:

First is urban industrial structure adjustment. The secondary industry has long accounted for a relatively large part in China’s economic aggregate. Due to long-term “path dependence” and uncontrolled exploitation of natural resources, there are prominent urban environmental problems and human settlement problems, and it is difficult to achieve sustainable economic and social development. As the key to inclusive green urban growth, reasonable industrial structure adjustment can help reduce energy consumption, reduce the energy “dependence path,” promote the development of clean energy, and thus reduce carbon emissions. Through reasonable control of carbon emissions and strengthening of carbon emissions constraints on the energy industry, carbon emissions trading policy will help promote the transformation and upgrading of the original energy-intensive industries to the tertiary sector, forming an industrial restructuring and a change in the mode of economic development and ultimately achieving inclusive green resilience [17,22]. To this end, the following hypothesis is proposed:

Hypothesis 1. 

Carbon emissions trading policy promotes the adjustment of urban industrial structure and enables inclusive green resilience.

Second is urban technological innovation. On the one hand, under the impact of carbon emissions trading policy, technological innovation will accelerate R&D and the application of production technology and environmental protection technology of enterprises. By applying clean technology into energy and production systems, enterprises can fundamentally improve resource utilization efficiency, implement prevention from the front end, and reduce environmental pollution. On the other hand, the development of technological innovation meets the demands of production and life, improves the productivity of social labor, and provides more human, material, and financial resources for industrial enterprises. Hence, the implementation of carbon emissions trading policy can effectively stimulate the innovation capacity of enterprises and create more new jobs, provide more employment opportunities for urban labor, and promote inclusive green resilience. To this end, this paper proposes the following hypothesis:

Hypothesis 2. 

Carbon emissions trading policy stimulates the technological innovation capacity of cities and thus enables inclusive green resilience.

Third, the employment effect is remarkable. On the one hand, the carbon emissions trading policy encourages enterprises with high energy consumption and low efficiency to transform into enterprises with high efficiency and low energy consumption. For enterprises with successful transformation, they can take advantage of the opportunities brought by green transformation to reduce carbon emissions and save carbon quotas, trade and make profits in the carbon trading market, improve their capital stock, and thus promote the expansion of enterprise scale and technological improvement, enabling green and upgraded development. In this process, in order to meet the needs of expanding scale and improving technology, enterprises increase employment and employment income, which meets the public’s demand for employment and narrows the income distribution gap [23]. On the other hand, under the influence of carbon emissions trading policy, natural resources and environmental standards inevitably rise. In order to meet their own needs, enterprises reduce the use of natural resources and turn to labor with a relatively low cost instead of natural resources in development. Therefore, the substitution of resource elements by labor promotes the increasing labor demand of industrial enterprises, thus effectively adjusting the balance between social labor supply and demand, enabling coordination and unity between the economic system, social system, and environmental system and promoting inclusive green resilience. To this end, this paper proposes the following hypothesis:

Hypothesis 3. 

Carbon emissions trading policy affects urban employment and thus enables inclusive green resilience.

2.2. Research on the Heterogeneity of Carbon Emissions Trading Policy to Inclusive Green Resilience

The role of carbon emissions trading policy has different significance in cities of different development types. This paper mainly analyzes the difference from two parts: urban development scale and urban function.

First is urban development scale. Urban scale is a primary factor affecting the economic and social development of a city. The economic, social, and environmental benefits vary for cities of different scales, and a city of a certain scale will form economic aggregation and population aggregation to varying degrees. Economic agglomeration and population agglomeration promote the economic and social development of cities but also lead to the consumption of more energy and more carbon dioxide emissions, thus hindering the green development ability of cities. In addition, larger cities face more social problems and need to deal with more problems in their efforts to promote inclusive green resilience. Therefore, the impact of carbon emissions trading policy on the inclusive green resilience of cities is somewhat different due to the effect of urban development scale. Larger cities have higher economic agglomeration and population agglomeration, with great energy consumption and carbon emissions, but industrial enterprises have a stronger influence on the external environment and greater inter-industry competition reaction. When exposed to external conditions such as environmental policy constraints, they can make rapid adjustments to enhance the production and supply capacity of enterprises, thus improving the economic benefits of the whole city. However, due to the low degree of economic development and market activity in small-scale cities, carbon emissions trading policies finds it challenging to effectively foster inclusive green resilience. To this end, the following hypothesis is proposed:

Hypothesis 4. 

In the development of large cities, carbon emissions trading policy has a better impact on inclusive green resilience.

Second is urban function difference. One important reason for China’s rapid economic development since the reform and opening up is the driving role of large-scale resource-based cities. For cities that have relied on heavy industry for a long time and excessively exploit natural resources to develop their economy, there is serious environmental pollution, backward infrastructure, a large proportion of labor force emigration, and significant vulnerability in economic development. As a result, these cities have significantly higher energy consumption and carbon emissions intensity than other non-resource-based cities in China. Due to the emphasis on the development of the secondary industry, serious industrial structure imbalance occurs in resource-based cities, resulting in overcapacity, weak urban innovation and development capacity, and serious urban pollution. Carbon emissions trading policy can control the carbon emissions of resource-based cities, adjust the industrial structure, inject new vitality into the development of industrial enterprises, stimulate the internal production and technological innovation of industrial enterprises, reasonably adjust the allocation of the urban labor force, improve the green development ability of enterprises and the well-being of social residents, and ultimately improve the inclusive green resilience level of cities. To this end, the following hypothesis is proposed:

Hypothesis 5. 

In resource-based cities, carbon emissions trading policy has a more significant impact on inclusive green resilience.

3. Research Model and Variables

3.1. Study Area

In order to put China’s carbon trading system requirements into action, the National Development and Reform Commission (NDRC) designated Beijing, Tianjin, Shanghai, Chongqing, Hubei Province, Guangdong Province, and Shenzhen City as pilot cities for the carbon emissions trading policy in 2011. In 2013, Shenzhen City led the way in launching the nationwide carbon emissions trading market, and other cities and provinces followed suit. As a result, this paper views the 2013 pilot carbon emissions trading policy as a quasi-natural experiment. After deleting the cities with the most missing data on green inclusive resilience, the panel data of the remaining 184 cities from 2008 to 2018 are used as research samples, of which 139 are non-resource cities, 37 are mature resource cities, and 8 are declining resource cities. Given the policy implementation time lag, 2014–2018 is designated as the implementation year of the carbon emissions trading policy, and 2008–2013 is designated as the year preceding the system’s implementation in both the treatment and the control group. The 37 prefectural-level cities taking part in the pilot carbon emissions trading strategy are considered the treatment group, whereas the remaining 147 cities are considered the control group. A double-difference model is built to analyze the influence of carbon emissions trading policy on the level of green inclusion resilience.

3.2. Research Model

PSM-DID model. The DID was proposed by Ashenfelier. The model can deduce differences between the experimental group impacted by the policy and the control group not affected by the policy before and after the policy’s implementation, mostly through the use of dummy variables. The DID model, as one of the most widely used tools for assessing policy success, has also been used to investigate the process of the influence of carbon emissions trading policies on the level of economic, social, and other development [24,25]. As a result, this article employs the DID model to assess the impact of carbon emissions pricing policies on the level of green inclusive resilience. The model is as follows:

$${\mathrm{green}}_{\mathrm{it}}={\mathsf{\alpha }}_0+{\mathsf{\alpha }}_1\mathrm{du}\ast \mathrm{dt}+{\displaystyle \sum }_{\mathrm{i}=1}^{\mathrm{N}}{\mathrm{b}}_{\mathrm{j}}{\mathrm{X}}_{\mathrm{it}}+{\mathsf{\epsilon }}_{\mathrm{it}}$$

(1)

where i represents a city, t represents the year, ${\mathrm{green}}_{\mathrm{it}}$ represents the interpreted variable, du represents the city grouping dummy variable, and dt represents the time grouping dummy variable. ${\mathrm{X}}_{\mathrm{it}}$ represents the control variable, including the economic development level (sgdp), innovative research and development ability (echol), urbanization level (city), industrial structure (ind), and population density (pop). ${\mathsf{\epsilon }}_{\mathrm{it}}$ is the disturbance term. Since the policy is implemented after 2013 and there is a certain time lag effect, 2014–2018 is defined as the implementation year of the carbon emissions trading policy, which is defined as 1. The period 2007–2013 is defined as 0 as the early stage of system introduction. In the division of treatment group and control group, pilot cities in two provinces and five cities in 2014 are taken as the experimental group and defined as 1. Other cities serve as the control group and are defined as 0.

However, due to the great heterogeneity in the development of different cities in China, it is difficult for different cities to meet the condition of time effect consistency [26,27]. Thus, a control group with characteristics as similar as possible to the experimental group should be selected before adopting the DID method. The PSM-DID model is an estimation method that combines the PSM method and the DID method. The basic concept of propensity score matching is that for some control variables, the assumption that the mean value can be ignored must be met. Compared with the DID method, PSM-DID exhibits the advantage of controlling unobservable differences between groups that do not change over time, so this paper adopts the PSM-DID model for robustness analysis [27]. Specifically, PSM is used to reduce sample selection bias, and radius matching is conducted between the experimental group and the control group samples. On this basis, DID analysis is performed. The PSM-DID model is as follows:

$${\mathrm{green}}_{\mathrm{it}}{}^{\mathrm{PSM}}={\mathsf{\alpha }}_0+{\mathsf{\alpha }}_1\mathrm{du}\ast \mathrm{dt}+{\displaystyle \sum }_{\mathrm{i}=1}^{\mathrm{N}}{\mathrm{b}}_{\mathrm{j}}{\mathrm{X}}_{\mathrm{it}}+{\mathsf{\epsilon }}_{\mathrm{it}}$$

(2)

3.3. Measure the Level of Inclusive Green Resilience in Cities

Index system construction. This paper comprehensively refers to the inclusive growth evaluation index published by the Asian Development Bank and the description of inclusive green resilience index system by Li et al. and constructs three subsystems, including economic development, social inclusion, and environmental sustainable development [7,28]. Economic resilience is mainly measured by the basis of urban economic development and economic stability. In the selection of indexes, six indexes are used to describe the subsystem of economic development. The social inclusion resilience is mainly measured by eight indexes, including education, employment, health, and infrastructure. In terms of environmental resilience, the analysis mainly focuses on green production and consumption and environmental protection, including seven evaluation indexes. See

Table 1. Indicators of inclusive green resilience in cities.

Goal System	Subsystem	Name of Index	Unit	Character
Inclusive green resilience	Economic resilience	GDP growth rate	%	+
		Proportion of tertiary industry	%	+
		Proportion of fiscal revenue	%	+
		Per capita disposable income of urban residents	CNY/person	+
		Per capita disposable income of rural residents	CNY/person	+
		Urban and rural per capita income ratio	%	−
	Social inclusion resilience	Educational resources for every 10,000 people	%	−
		Educational fund input	%	+
		Urban registered unemployment rate	%	−
		Employment training rate	%	+
		Health professionals per 1000 population	Person/1000	+
		Number of beds in medical and health institutions per 1000 population	Person/1000	+
		Length of transportation lines per 10,000 people	Km/10,000 people	+
		The number of buses per 10,000 people	Car/1000 people	+
	Ecological environment resilience	Energy consumption per unit of gross regional product	Ton/1000 yuan	−
		Amount of discharge wastewater	Ton/1000 yuan	−
		Carbon dioxide emissions	Ton/1000 yuan	−
		Sulfur dioxide emissions	Ton/1000 yuan	−
		Urban per capita public green space	Person/square meter	+
		Comprehensive utilization rate of solid waste	%	+
		Intensity of investment in environmental pollution control	%	+

Note: “+”, “−” represent positive indicators and negative indicators, respectively.

for details.

Research methods. On the basis of constructing an evaluation index system, this paper refers to the research method of Wang et al. (2022) and uses the entropy method to analyze the level of inclusive green resilience [29]. The specific steps are as follows:

First, according to the index attributes, the evaluation data are standardized with the help of the entropy method:

Forward standardization formula:

$${\mathrm{X}}_{\mathrm{ij}}^{}=\left({\mathrm{X}}_{\mathrm{ij}}^{}-{\mathrm{X}}_{\mathrm{minij}}\right)/\left({\mathrm{X}}_{\mathrm{maxij}}-{\mathrm{X}}_{\mathrm{minij}}\right)\times 0.99+0.001$$

(3)

Negative normalization formula:

$${\mathrm{X}}_{\mathrm{ij}}^{}=\left({\mathrm{X}}_{\mathrm{maxij}}-{\mathrm{X}}_{\mathrm{ij}}^{}\right)/\left({\mathrm{X}}_{\mathrm{maxij}}-{\mathrm{X}}_{\mathrm{minij}}\right)\times 0.99+0.001$$

(4)

where ${\mathrm{X}}_{\mathrm{ij}}^{}$ represents the evaluation standard value, ${\mathrm{X}}_{\mathrm{minij}}$ represents the minimum value of the $\mathrm{j}$-th item in the $\mathrm{i}$-th city, and ${\mathrm{X}}_{\mathrm{maxij}}$ represents the maximum value of the j-th item in the i-th city.

Second, calculate the characteristic proportion of the j-th index:

$${\mathrm{P}}_{\mathrm{ij}}={\mathrm{X}}_{\mathrm{ij}}^{}/{\displaystyle \sum }_{\mathrm{i}=1}^{\mathrm{m}}{\mathrm{X}}_{\mathrm{ij}}^{}$$

(5)

Third, calculate the index of the information entropy:

The information entropy calculation formula of the j-th index of the i-th city can be expressed as:

$${\mathrm{e}}_{\mathrm{j}}=-\mathrm{k}{\displaystyle \sum }_{\mathrm{i}=1}^{\mathrm{mT}}\left({\mathrm{X}}_{\mathrm{ij}}^{}/{\displaystyle \sum }_{\mathrm{i}=1}^{\mathrm{mT}}{\mathrm{X}}_{\mathrm{ij}}^{}\right)\ln ({\mathrm{X}}_{\mathrm{ij}}^{}/{\displaystyle \sum }_{\mathrm{i}=1}^{\mathrm{mT}}{\mathrm{X}}_{\mathrm{ij}}^{})$$

(6)

where $\mathrm{k}=1/\ln \mathrm{mT}$, $1\le \mathrm{i}\le \mathrm{m},\mathrm{and}1\le \mathrm{j}\le 21$.

Fourth, calculate the weight of the evaluation index:

$${\mathrm{w}}_{\mathrm{j}}=\left(1-{\mathrm{e}}_{\mathrm{j}}\right)/{\displaystyle \sum }_{\mathrm{j}=1}^{21}\left(1-{\mathrm{e}}_{\mathrm{j}}\right)$$

(7)

Fifth, calculate the comprehensive score of the weight of the j-th index:

$${\mathrm{S}}_{\mathrm{j}}={{\displaystyle \sum }}^{}{\mathrm{w}}_{\mathrm{j}}{\mathrm{P}}_{\mathrm{ij}}$$

(8)

Sixth, according to the obtained comprehensive score of the index weight, the comprehensive index of inclusive green resilience is calculated:

$${\mathrm{green}}_{\mathrm{i}}={\displaystyle \sum }_{\mathrm{i}}^{\mathrm{N}}{\mathrm{S}}_{\mathrm{ij}}$$

(9)

where ${\mathrm{green}}_{\mathrm{i}}$ represents the comprehensive index, and N is the number of indexes under all levels in the evaluation system.

3.4. Variable Selection and Data Source

The DID model is used to analyze the impact of carbon emissions trading policy on the level of inclusive green resilience in cities, and its explained variable is the level of inclusive green resilience in cities (green). The data used to calculate the level of inclusive green resilience in cities are mainly collected from the “China Urban Statistical Yearbook,” the “China Environmental Statistical Yearbook,” and the statistical yearbook and bulletin of each province from 2007 to 2018.

In terms of control variables, urbanization(city) is mainly expressed by the proportion of non-agricultural population in the total population [30]. The level of economic development (sgdp) is expressed as a logarithm of real GDP per capita. Innovative research and development capability (echol) is expressed by the number of urban patent grants. Industrial structure (ind) is expressed as the proportion of added value of secondary industry in GDP. Population density (pop) is expressed by the number of people per unit administrative area.

4. Empirical Results and Robustness Test

4.1. Empirical Results

The descriptive statistical results of the main variables in this paper are shown in

Table 2. Descriptive statistics.

Variable	Mean	SD	Min.	Max.
Green	0.357	0.172	0.010	0.998
sgdp	4.563	0.282	2.670	5.280
echol	5.613	8.962	0.000	64.580
ind	0.497	0.120	0.000	3.739
pop	5.986	0.750	0.177	3.739
city	0.516	0.146	0.189	1.000

. Figure 1 shows the distribution of inclusive green resilience capacity across cities in 2008, 2013, and 2018. In

Table 3. Carbon emissions trading policies and inclusive green resilience in cities: DID regression results.

Variable	(1)	(2)	(3)	(4)
	Green	Green	Green	Green
DID	0.138 ***	0.029 ***	0.053 ***	0.022 ***
	(20.130)	(7.799)	(11.354)	(5.588)
city			0.358 ***	−0.111 ***
			(12.822)	(−4.181)
sgdp			0.230 ***	0.024 **
			(25.214)	(2.481)
echol			0.004 ***	0.002 ***
			(6.530)	(5.051)
ind			−0.050 ***	−0.004
			(−3.927)	(−0.411)
pop			0.042 ***	0.025 ***
			(6.621)	(4.888)
_cons	0.344 ***	−39.287 ***	−1.129 ***	−39.564 ***
	(205.542)	(−72.530)	(−22.535)	(−33.537)
N	2024	2024	2024	2024
R2	0.099	0.770	0.650	0.778

Note: ** and *** indicate significance at the levels of 10%. In parentheses is the robust standard error.

, model (1) is the model without adding control variables, and model (2) is based on model (1) with the addition of the time effect. Model (3) is the regression result after adding control variables on the basis of model (1), and model (4) is the regression result after adding control variables on the basis of model (2). All regression results show that carbon emissions trading policy has a significant positive impact on the inclusive green resilience level of cities with or without the inclusion of control variables, indicating that carbon emissions trading policy significantly improves the inclusive development level of cities. In terms of control variables, high population density and technical innovation level can significantly improve the level of inclusive green resilience in cities. The proportion of urbanization has a significant negative impact on the level of inclusive green resilience in cities, indicating that a higher proportion of urbanization reduces the level of inclusive green resilience in cities, whereas the coefficients of other control variables are insignificant, indicating that they are not the core factors affecting the level of inclusive green resilience in cities.

4.2. Robustness Test

Parallel trend test. For the DID, a quasi-natural experimental method, an important prerequisite for the establishment of its results is to meet the parallel trend hypothesis—that is, for the inclusive green resilience level of the experimental group and the control group to have the same time variation trend before the impact of carbon emissions trading policy [31]. The research completed a parallel trend test on the level of inclusive green resilience in cities, using 2014 as the base year and data from the first four years and the last four years as samples. The test results are shown in Figure 2. As can be seen from Figure 1, in the years before the pilot of the carbon emissions trading policy, the inclusive green resilience level (green) of cities fluctuates around 0 and does not reject 0 at the 95% confidence interval, which conforms to the common trend hypothesis.

PSM-DID analysis. In order to overcome the differences in the change trends between the pilot cities of the carbon emissions trading policy and other cities and to reduce the bias of the DID analysis, further robustness tests were conducted in this paper using PSM-DID’s nearest neighbor matching, kernel matching, and radius matching. As can be seen from

Table 4. The average treatment effect.

Stage	Type	Nearest Neighbor Matching	Kernel Matching	Radius Matching
		Green	Green	Green
Before	Treated	1.010	1.010	1.010
	Control	0.051	0.051	0.051
	T-C	0.959	0.960	0.960
After	Treated	0.706	0.917	0.706
	Control	0.050	0.050	0.050
	T-C	0.656	0.868	0.656
S.E. (ATT)		0.131	0.146	0.131
T (ATT)		5.01	5.95	5.01
Variable		Yes	Yes	Yes
N		2024	2024	2024

, carbon trading policies still significantly increase the level of inclusive green resilience of cities after using the three PSM-DID tests. The estimated results of the PSM-DID do not differ significantly from those of the above DIDs, further validating the empirical results of this paper. Carbon trading policies play an important role in inclusive green resilience.

5. Verification of Influence Mechanism

5.1. Verification of Influence Mechanism

Through DID analysis and the robustness test, it can be seen that the carbon emissions trading policy can promote the inclusive green resilience level of cities. In order to further explore the internal influence mechanism of carbon emissions trading policy on the inclusive green resilience level of cities, based on Formula (1), this paper sets the mechanism variables from the three parts of industrial structure adjustment, technological innovation ability, and employment effect, and, drawing on the study of Wen and Ye (2004), a three-step approach is used to test the model to further investigate the action path of carbon emissions trading policy on inclusive green resilience [31]. The model is as follows:

$${\mathrm{green}}_{\mathrm{it}}={\mathsf{\beta }}_0+{\mathsf{\beta }}_1\mathrm{du}\ast \mathrm{dt}+{\displaystyle \sum }_{\mathrm{i}=1}^{\mathrm{N}}{\mathrm{b}}_{\mathrm{j}}{\mathrm{X}}_{\mathrm{it}}+{\mathsf{\epsilon }}_{\mathrm{it}}$$

(10)

$${\mathrm{M}}_{\mathrm{it}}={\mathsf{\alpha }}_0+{\mathsf{\alpha }}_1\mathrm{du}\ast \mathrm{dt}+{\displaystyle \sum }_{\mathrm{i}=1}^{\mathrm{N}}{\mathrm{b}}_{\mathrm{j}}{\mathrm{X}}_{\mathrm{it}}+{\mathsf{\epsilon }}_{\mathrm{it}}$$

(11)

$${\mathrm{green}}_{\mathrm{it}}={\mathsf{\gamma }}_0+{\mathsf{\gamma }}_1\mathrm{du}\ast \mathrm{dt}+{\mathsf{\gamma }}_2{\mathrm{M}}_{\mathrm{it}}+{\displaystyle \sum }_{\mathrm{i}=1}^{\mathrm{N}}{\mathrm{b}}_{\mathrm{j}}{\mathrm{X}}_{\mathrm{it}}+{\mathsf{\epsilon }}_{\mathrm{it}}$$

(12)

where ${\mathrm{M}}_{\mathrm{it}}$ is the mechanism variable. If the coefficients ${\mathsf{\alpha }}_1$ and ${\mathsf{\gamma }}_1$ are significant, then there is an effect. In the index selection of variables, we set mechanism variables for industrial structure adjustment, technological innovation capacity, and employment effect in accordance with the requirements of scientificity, representativeness, and comprehensiveness. Industrial structure adjustment is measured by the proportion of added value of the secondary industry and tertiary industry (upgrade). Technological innovation capacity is measured by the number of effective invention patents of industrial enterprises (patents). Employment effect is measured by the average wage of working employees (wage). The data for each index derive from the statistical yearbook of each city.

5.2. Influence Mechanism Analysis

Table 5. A mechanistic test of the impact of carbon trading policies on green inclusive resilience.

Variable	Industrial Structure Adjustment	Technological Innovation Capacity	Employment Effect
upgrade	−0.023 ***
patents		0.157 ***
wage			0.028 ***
did	0.052 ***	0.048 **	0.046 ***
city	0.302 ***	0.288 ***	0.272 ***
sgdp	0.242 ***	0.237 ***	−0.219 ***
pop	0.019 ***	0.018 ***	0.018 ***
ind	0.059 **	−0.029 **	−0.037 **
echol	0.003 ***	0.003 ***	0.003
_cons	−1.037 ***	−1.046 ***	−1.047 ***
R2	0.685	0.689	0.701

Note: ** and *** indicate significance at the levels of 5%, and 10%, respectively. In parentheses is the robust standard error.

shows the analysis of the internal mechanism by which carbon emissions trading policy affects the level of inclusive green resilience in cities. As can be seen from the table, the three influencing mechanisms can significantly promote the impact of carbon emissions trading policies on inclusive green resilience. First of all, it can be seen from the table that, as the mechanism variable of industrial structure adjustment, the proportion of added value of the secondary and tertiary industries has a significant negative impact, indicating that the economic structure dominated by the secondary industry does not use carbon emissions trading policy to promote inclusive green resilience. This is because, compared with the tertiary industry, the secondary industry has a single industrial structure, high energy consumption, and carbon emission intensity, which leads to serious urban environmental pollution and low economic and social development, and finally results in the weak green sustainable development level of cities. As a result, if we can raise the entry threshold of the secondary industry, reduce the market share of the secondary industry and transform it in a green and clean direction, and increase the proportion of the tertiary industry, we can reduce the negative impact of industrial structure imbalance in the process of enhancing cities’ green inclusive resilience through the optimization and upgrading of the industrial structure so as to better play the impact of carbon emission trading policy on the urban inclusive green resilience capacity. Therefore, Hypothesis 1 is verified.

Secondly, in terms of the level of technological innovation, the number of patentable inventions of industrial enterprises has a significant impact on the impact of carbon trading policies on inclusive green resilience. Therefore, to enhance inclusive green resilience through technological innovation, more attention should be paid to innovation, invention, scientific research, and technology. Overall, carbon emissions trading policy can promote inclusive green resilience in cities with the help of technological innovation. Therefore, Hypothesis 2 is verified.

Finally, in terms of employment effect, the total number of jobs in a city has a significant impact on the inclusive green resilience of a city due to carbon emissions trading policies. This suggests that, in order to better improve the inclusive green resilience of a city, we should strengthen the improvement of urban employment capacity and employment opportunities. Overall, carbon emissions trading policy can promote inclusive green resilience in cities through employment effects. Therefore, Hypothesis 3 is verified.

6. Heterogeneity Analysis

6.1. Heterogeneity Analysis of City Development Scale

Large-scale cities can lead to economic aggregation and population aggregation, which in turn promotes regional economic development and technological innovation. However, excessive aggregation leads to high energy consumption, increased carbon emissions, and more complex social problems, which is unfavorable to inclusive green urban growth. Therefore, whether carbon emissions trading policy can play an equal role in cities of different scales needs to be further studied. Based on the “Notice on Adjusting the Standards of City Scale Division” issued by the State Council, the size of the resident population in an urban area is used as a criterion in this article, and cities with a resident population of 500,000 or fewer are classified as small cities. Cities with a population of more than 500,000 but fewer than 1 million people are classed as medium-sized. Large cities are classified as having a resident population of more than one million people [32]. The paper divides 184 representative cities into large and medium cities and small cities for the regression test.

Table 6. Heterogeneity analysis of urban development scale.

	Small Cities	Large and Medium Cities
	Green	Green
DID	0.001	0.029 ***
	(0.09)	(6.24)
Control variable	Yes	Yes
Year fixed effect	Yes	Yes
_cons	−0.756 ***	−0.383 ***
	(−4.78)	(−6.62)
N	451	1573
R2	0.806	0.816

Note: *** indicate significance at the levels of 10%. In parentheses is the robust standard error.

shows the impact of different types of cities on the level of inclusive green resilience. As can be seen from the table, the carbon emissions trading policy significantly improves the inclusive green resilience level in large and medium-sized cities, but there is an insignificant effect on small cities. This shows that the carbon emissions trading policy can better fit the economic aggregation and population aggregation of large and medium-sized cities to jointly achieve inclusive green resilience. As small cities have lagging development, inferior market factor allocation ability, and weak information response ability compared to large and medium-sized cities, carbon emissions trading policy exerts an insignificant effect on them. In addition, in the process of policy implementation, due to the insufficient advantages of such cities, a large amount of capital and labor need to be consumed after policy implementation, resulting in greater output than income, which is ultimately unfavorable to the sustainable development of cities. Thus, Hypothesis 4 is verified.

6.2. Heterogeneity Analysis of Urban Function

According to statistics, China has 262 resource-based cities that can be classified into four types: growing, mature, declining, and regenerative. Compared with growth- and regenerative-type resource-based cities, the mature and declining types have advantages in transformation and development and can quickly get rid of resource dependence. As a result of the examination of resource-based and non-resource-based cities, this article separates resource-based cities into typical mature resource-based cities and declining resource-based cities, taking into account the sample cities chosen previously in this research. As shown in

Table 7. Heterogeneity analysis of ecological resilience.

	Non-Resource-Based Cities	Resource-Based Cities Group		Resource-Based Cities
		Mature Resource-Based Cities	Declining Resource-Based Cities
DID	0.054 ***	0.053 ***	0.133 ***	0.046 ***
	(9.42)	(4.04)	(5.88)	(4.70)
_cons	−1.135 ***	−0.780 ***	−2.782 ***	−0.907 ***
	(−16.52)	(−7.52)	(−2.10)	(−10.85)
Control variable	YES	YES	YES	YES
N	1430	385	99	594
F	449.75	85.62	28.64	162.53
R2	0.685	0.708	0.930	0.705

Note: *** indicate significance at the levels of 10%. In parentheses is the robust standard error.

.

Carbon trading policies have a significant influence on the green inclusive resilience of both declining and mature resource-based cities at the 1% level, green inclusive resilience improves significantly after the implementation of carbon trading policies, and the increase is greater for declining resource-based cities. On the one hand, declining resource-based cities have more space for improvement in green inclusive resilience than mature resource-based cities, and they are more responsive to environmental policies aimed at carbon reduction and green development. On the other hand, declining resource-based cities experience more severe urban development concerns, such as slow economic growth and significant issues with means of subsistence. As a market-based policy, carbon emissions trading can effectively promote the sustainable development capacity of cities by taking into account both equity and efficiency. It is therefore easier to inject new impetus into the reform and development of cities and to achieve an increase in their green, inclusiveness, and resilience capacity. Thus, Hypothesis 5 is verified.

7. Conclusions and Policy Implications

Based on 184 representative cities from 2008 to 2018 as a study sample, this study analyzes the impact of carbon trading policies on the inclusive green resilience of pilot cities with the help of the DID model. The results show that carbon trading policies can significantly promote inclusive green resilience in cities, at about 2.2% and pass parallel trend and PSM-DID model tests, confirming the reliability of the findings. In order to gain insight into the mechanism of the impact of carbon trading policies on inclusive green resilience, this paper provides a further in-depth analysis from three components: industrial structure, technological innovation capacity, and employment effects. The results show that carbon trading policies can achieve inclusive green resilience by adjusting industrial structure, improving the capacity for technological innovation, and promoting employment effects. Carbon emissions trading programs, at 15.7%, have the largest influence on inclusive green resilience through technological innovation capacity. Carbon emissions pricing policies have a negative and large influence on inclusive green resilience through industrial restructuring, with a value of −2.3%. The results of the heterogeneous analysis show that, on the one hand, carbon trading policies affect cities of scale, and the policy can have a significant positive impact on the inclusive green resilience of large and medium-sized cities. On the other hand, there are significant differences in the impact of carbon emissions trading policies on inclusive green resilience in urban functions. For declining resource-based cities, carbon trading policies can most rapidly improve the inclusive green resilience of cities, at 13.3%, followed by non-resource-based cities, at about 5.4%. Based on the above findings, the paper proposes the following:

First, focus should be on the adjustment of the industrial structure of a city. The secondary industrial economy, which is based on high energy consumption and high emissions, still accounts for a large proportion of China’s economic development. In this regard, on the one hand, on the basis of carbon emissions trading policies, the coordination of carbon emission trading policies should be promoted with other mandatory policies the low-carbon incentive and constraint mechanism should be improved while raising the entry threshold of the secondary industry through carbon emissions trading policies to guide and encourage the development of emerging low-carbon industries and promote the development of modern service industries. On the other hand, urban environmental supervision should be strengthened to reduce the intensity of urban carbon emissions with the help of market guidance and government control. With regard to the secondary industry, which is high in energy consumption and pollution, the industrial energy consumption structure should be gradually adjusted, clean technology of resources should be vigorously developed, and the flow of industries with high energy consumption and pollution should be reasonably guided to those with low emissions and high efficiency.

Second, urban technological innovation and development should be promoted. From the analysis of the influence mechanism, we can see that the number of patents of industrial enterprises in cities has a significant positive effect on the development of technological innovation in cities. Therefore, while implementing the carbon emissions trading policy, attention should be paid to enhancing the technological input of cities, actively guiding enterprises to carry out green technological innovation, guaranteeing the technological innovation of non-high-tech enterprises and private enterprises from the policy level, and fundamentally reducing the carbon dioxide emissions intensity of cities.

Third, attention should be paid to the effect of urban employment. The influence mechanism shows that urban employment effects can improve inclusive green resilience. Therefore, the government should increase training in employment skills of the workforce and improve the level of employment skills and the quality of the workforce. In addition, the function of public employment services should be fully utilized to alleviate structural contradictions and employment inequalities in the urban workforce, and the social impact of the workforce should be recognized in enhancing the inclusive green resilience of a city.

Finally, measures should be adjusted to local conditions. Different city sizes have different impacts on carbon trading policies. On the one hand, large and medium-sized cities should implement appropriate tightening of carbon emissions quotas to stimulate market competition, improve industrial enterprises’ technological innovation ability, promote employment, and improve inclusive green resilience. Small cities, on the other hand, should establish appropriate relaxation policies, as well as matching auxiliary policies, to provide financial and technical support for technological innovation and green development to alleviate the negative effects of the lack of impetus for urban development and promote inclusive green resilience. In addition, for resource-based cities, government departments at all levels should avoid the “one-size-fits-all” policy. For mature resource-based cities, reasonable carbon emissions targets should be set to encourage cities to develop new pillar industries and green industries, strengthen workforce absorption, and achieve inclusive green urban growth. For declining resource-based cities, we should not only scientifically set reasonable total carbon emissions targets, relax carbon emissions indicators, and stimulate energy-intensive and high-emissions enterprises to reduce carbon emissions by means of technological innovation but also take into account the market economy vulnerability of declining resource-based cities and introduce corresponding carbon emissions-supporting measures in a timely manner. The relationship between industrial structure and the allocation of labor should be properly adjusted to reduce structural unemployment and help gradually achieve sustainable urban transformation.

Our research has several limitations, which may also be the future directions for our follow-up research. First, our analysis exclusively takes into account data from Chinese cities and excludes data from other countries. With the substantial trend of globalization, the network of economic and social links between countries is becoming increasingly intertwined. The study could be expanded in the future to investigate the dynamic impact of national interaction networks on the capacity for inclusive green resilience. Second, due to the scarcity of data on inclusive green resilience indicators for some provinces, the article’s study period is limited to 2008–2018. If the research data years could be expanded, the article could more precisely depict variations in impacts over time.