The Transformative Impacts of Green Finance Governance on Construction-Related CO2 Emissions

Abstract

In the context of the commitment to peak carbon emissions by 2030, specific sectors in China should take responsibility to change their energy consumption patterns. In China and across the globe, the construction sector is a major source of carbon dioxide emissions, as well as an indicator of economic growth and structural transformation. In this study, we examine panel data for 30 provinces or regions from 2008 to 2019 to dissect which macro-factors contribute to growth in carbon emissions, and which will lead to carbon emission reductions. Derived by the entropy method, the Green Finance Index is a comprehensive environmental regulation index related to reduction in emissions in each province. It presents an N shape for construction emissions, and provinces are currently striving to cross the first inflection point, which will help to curb emissions. Judging from the combined effects of this and other structural factors, the Green Finance Index can promote the decarbonization of production by playing the role of guiding and screening capital allocation. Population expansion, income levels, and financial development initially stimulate demand for construction, but their effects eventually level off. This paper can serve as a reference for developing countries that are experiencing industrialization and urbanization processes and handling gas discharge pressure at the same time.

1. Introduction: Broad Context

Emerging economies represented by China, India, Brazil, and Turkey are facing multiple development pressures of economic growth—carbon emission reduction and employment stability. In the process of industrialization and urbanization, the acceleration of energy consumption inevitably causes the continuation of a state of high carbon emissions [1]. How to balance multiple targets and reduce carbon emissions through administrative and financial means has become an urgent problem to be solved [2,3].

In 2006, China released about 6 gigatons (Gt) of carbon emissions, becoming the world’s largest carbon emitter. China’s carbon emissions will significantly affect the realization of the global goal of reducing carbon emissions, and play an important role in mitigating climate warming. In 2014, China pledged to peak carbon dioxide emissions by 2030 in the China–US Joint Statement Addressing the Climate Crisis. If China can successfully fulfill its promise, it will become a model for developing countries around the world to reduce emissions.

However, is the contradiction between income growth and emission reduction really irreconcilable? Are the measures taken effective? The macroeconomic structure, accompanied by the spontaneous constraints of various industries and the impact of government incentives, means that the mechanisms of carbon emissions in any industry are complicated. Fortunately, China is highly heterogeneous across provinces, showing different economic endowments and development stages. Therefore, the research on the provincial panel has reference significance for countries in different stages of industrialization.

The industrial structural transformation and population migration lead to dynamic adjustment of consumption demand structure which, in turn, leads to changes in the structure of energy consumption and sources of carbon emissions. Considering also the incentives and penalties of industry sectors and government, the mechanisms by which carbon emissions are affected are perplexing. Based on factor decomposition, most of the existing studies analyze the effect of each factor, but lack a means of sorting out the conduction mechanism.

This paper attempts to clarify the direct and indirect effects of macroeconomic environmental regulations on carbon emissions from construction through regression models. By sorting out mechanisms and estimating parameters, we explore the functional relationships and conduction paths of financial measures with respect to construction-related CO₂ emissions through industrial structures and real estate markets. This makes up for the deficiencies of the existing literature. Financial and fiscal policies themselves have the function of guiding funds and filtering enterprises. In the following paragraphs, special emphasis is placed on their nonlinear effects on the main source of carbon emissions, namely, the construction industry.

1.1. Realistic Features and the Main Aim of this Article

As a result of rapid industrialization and population concentration in urban areas, the total carbon emissions of China have increased in recent decades, accounting for about 30% of the world’s carbon dioxide emissions. Being the highest emitter in the world [4], China submitted the Chinese Intended Nationally Determined Contributions (INDCs) to the Paris Agreement, which commit to peaking its emissions by 2030, with a 60–65% reduction in intensity compared to 2005.

The specific sector that performs most prominently during economic booms is the construction industry, which is both the physical cornerstone of industrial investment and consumer demand expansion, and a barometer of their rise and fall. Construction can be classified into six categories, including productive construction and non-productive construction: residential, manufacturing, infrastructure, commercial, public, and other buildings. As a result of the traditional construction industry’s crude development methods, construction has become characterized by high energy consumption, low utilization, and high emissions. It was proposed at the 21st United Nations Climate Change Conference (COP21) that the energy consumed by buildings throughout their whole life cycle constitutes 30% of global energy consumption. The construction industry is one of the major sources of carbon emissions, particularly in China—an emerging economy—so this article focuses on this energy-intensive industry.

The Chinese economy has been recovering slowly from the global financial crisis, and has entered a “new normal”—a new phase of economic development accompanied by various structural transformations. In the process of China’s promotion of ecological civilization construction and high-quality socioeconomic development, green finance is an important driving force for the green development of the economy. In June 2017, the Standing Committee of the State Council decided to build pilot zones for green financial reform and innovation in five provinces and regions.

Green finance policies contain all aspects of government and financial sectors’ indicators related to emissions reduction and environmental protection, including governmental guiding regulations and incentive-based voluntary covenants. Assuredly, the green finance policies have the functions of resource allocation and project filtering of finance itself, restricting the high-emission development mode of the traditional construction industry.

According to the statistical data from the National Bureau of Statistics of China, the China Statistical Yearbook, and the China Energy Yearbook, the carbon emissions of the construction industry and the development level of green finance in 30 provinces from 2001 to 2019 are exhibited in Figure 1. The steady growth trend of green finance is roughly the same in the 30 provinces, among which Beijing, Shanghai, Jiangsu, and Zhejiang have the highest growth rates.

However, the changes in construction industry carbon emissions vary widely among the 30 provinces. Overall, construction industry emissions in all provinces have generally experienced a continuous rise, growing in tandem with green finance. However, as green finance exceeds a certain level, the growth rate of emissions in many provinces—such as Hunan, Anhui, Yunnan, Fujian, Gansu, and other provinces—has begun to slow down. There are even some provinces where the carbon emissions have reached an inflection point, showing a declining trend, such as Beijing, Hebei, Hubei, Liaoning, Shandong, Tianjin, Zhejiang, Shanghai, and other provinces or regions. We therefore expect high levels of green finance to help curb the construction-related carbon emissions.

1.2. Core Explanatory Variables: Government and Financial Sector Regulations

Calculated by the entropy method using the following sub-indices in Table 1, we derived the environmental regulation index—the Green Finance Index (GreenF)—at the provincial level. GreenF is a comprehensive index of government and financial sector indicators related to emissions reduction and environmental protection. The raw data of province-level sub-indices are derived from the China Statistical Yearbook, the Provincial Statistical Yearbook, and the China Insurance Yearbook.

Entropy is capable of providing an objective basis for the comprehensive evaluation of multiple indicators. As a measurement of uncertainty and dispersion, it conforms to mathematical laws, and has a strict mathematical connotation. Using the information entropy instrument, the entropy method calculates the weights of each indicator based on its degree of variation. This study adopts the entropy value method to weight the above positive and negative indicators, distilling and summarizing the degree of incentives and restrictions on emissions in each province. In short, the GreenF variable is comparable, and effectively reflects levels of green governance across provinces.

2. Review of the Literature: Theoretical Mechanisms and Empirical Analysis

The increasingly severe challenges that have accompanied emerging economies’ rapid economic growth in the past 30 years include the unsustainable economic development mode, the lag in industrial structural adjustment, and the destruction of the ecological environment. In the research and discussion of green development, population explosion, industrial structure, construction industry, and residents’ disposable income are indispensable influencing factors in energy saving and carbon reduction. Researchers in the field of cleaner production have thought about the mechanisms between construction-related carbon emissions and green development from different perspectives. Here, we sort them out from the macroscopic angles of economic growth, industrial structural optimization, and urbanization development.

The carbon emissions of the construction industry mainly come from the combustion of fossil raw materials, and their levels are closely related to the strength of human economic activities. Economic growth apparently has a boosting effect on carbon emissions. Meanwhile, economic structural transformation and green transition may have a nonlinear impact on carbon emissions. Moreover, the increase in income levels profoundly changes the consumer demand of the residents, and makes people pay more attention to environmental protection, thereby curbing carbon emissions. This section summarizes the impact of economic development on the carbon emissions of the construction industry from three perspectives: economic growth, environmental regulation, and structural transformation of society and demographics.

2.1. Economic Expansion

With the deepening of research on the relationship between the economy and the environment, a large number of empirical analyses have proven that economic growth and growth in carbon emissions or pollutant emissions are correlated, or exhibit co-integration relationships [5,6].

Grossman’s research on the relationship between the concentration of sulfur dioxide, “smoke” pollutants, and economic growth found that when the national income level is low, emission concentrations increase with the increase in per capita GDP. An inflection point occurs between USD 4000 and 5000. When the income level is higher than the inflection point, its concentration decreases with the growth of GDP per capita, indicating that there is an inverted U-shaped relationship between environmental pollution and economic growth [7].

Salih Ozturk [8] applied a nonlinear analysis to Turkey—an emerging economy. According to its carbon emissions data from 1897 to 2019, they found that there is an asymmetric co-integration relationship between CO₂ emissions and GDP, and argued that the economic growth in Turkey is not sustainable in terms of its negative effects on environmental quality. Hence, it is of vital importance that policymakers pave the way for sustainable economic development by specifying and pursuing sustainable development goals.

Taking cement—the main source of CO₂ emissions in the world—as a typical object, rapid economic growth is the main driving force for the growth of cement-related carbon emissions. Studies [9] have used LMDI (logarithmic mean Divisia index)—an exponential decomposition analysis method—to decompose the changes in China’s cement-related CO₂ emissions from 2002 to 2016 to explore their driving factors, and found that growth in emissions is correlated with China’s GDP growth trend. In 2007, in response to the global financial crisis, the Chinese government invested CNY 4 trillion (about USD 586 billion) to expand domestic demand. During the period 2008–2010, stimulated by related measures, this factor contributed 26.4% of the emissions growth, indicating that the expansion of economic scale is a key factor in the growth of carbon emissions.

2.2. Income Level’s Effect: The Environmental Kuznets Curve (EKC)

Panayotou [10] first called the relationship between environmental quality and per capita income the environmental Kuznets curve (EKC), based on Kuznets’ “inverted U-shaped curve” hypothesis. Regarding environmental quality, further research by Grossman et al. [7] showed that in most cases, the per capita income of USD 8000 is the turning point from the deterioration of environmental quality to the improvement of environmental quality.

Income growth’s effect on emissions reduction is linear, or takes place through a single channel. Studies have shown that the pollution emissions of high-income communities in the United States have significantly reduced. A possible reason for this is that income levels affect the public’s preference for environmental quality and the ability to pressure polluting companies. The higher the income level, the greater the lobbying power against the government or polluting companies [11].

Based on the Chinese urban household survey from 2002 to 2009, Zan et al. (2017)’s results [12] show that the higher the income and housing wealth of a household, the higher their carbon emissions. In particular, they found the increasing margin effect of income on emissions for households of different wealth.

Ozturk, S et al. (2021) [8] integrated economic growth and financial development with the CO₂ emissions specification over the period of 1987–2019 in Turkey, employing NARDL (nonlinear autoregressive distributed lag) model to explore the long-term nonlinear linkages between the series. The results indicate the long-run and short-run asymmetric relationships between income inequality and CO₂ emissions in the classical model. Positive and negative income inequality shocks are positively related to CO₂ emissions, implying that positive and negative shocks of income inequality enhance CO₂ emissions in the long run.

2.3. Industrial Structure Transformation and Consumption Change

At the same time, there is evidence that the optimization of the economic structure, including industrial and investment structure, can improve production efficiency and resource utilization [13,14], help increase the probability of peaking carbon dioxide emissions [15,16,17], and even accelerate the process of peaking emissions [18].

Changes brought by the proportional increase in tertiary industry with respect to GDP are considered to be important for reducing CO₂ emissions [19]. Accelerating the pace of economic transformation will be greatly beneficial to the early achievement of peak carbon dioxide emissions, because the inhibitory effect of structural changes on carbon emissions has shown an increasing marginal trend. To assess the effects of economic structural changes on the successful peaking of carbon dioxide emissions, scholars have constructed different models, including the use of optimization models, computable general equilibrium (CGE) models, and other simulation models.

Using a new economic–carbon emission–employment multi-objective optimization model, Yu et al. (2018) [20] found that if the industrial structure measured by the proportion of the added value of the tertiary industry is optimized from 2013 to 2030, China would be able to reach peak carbon emissions around 2023–2025, without compromising GDP. Zhang et al. (2017) [21] used an extended LMDI model incorporating investment and R&D activities, dynamic Monte Carlo simulation, and scenario analysis to test the impact of China’s Industrial Green Development Plan 2016–2020 Targets on future carbon emissions, finding that investment intensity is the primary driver for the increase in ICEI (industrial CO₂ emissions intensity), while R&D intensity and energy intensity are the leading contributors to the reduction in ICEI. Results suggest that under existing policies, ICEI has the maximum possibility to decline by 23.1–23.7%, 60.8–61.1%, and 75.8–76.1% during 2015–2020, 2005–2020, and 2005–2030, respectively. Based on the research results, it is suggested that the Chinese government should further strengthen and improve the policies and measures of efficiency improvement and structural adjustment in the industrial sector.

The industrial structure indirectly changes the emission source structure by changing the energy consumption structure. With the rise of the secondary and tertiary industries one after another and the advancement of urbanization, residential and commercial energy sources have increased accordingly, and the energy consumption structure has shifted from direct combustion of coal and organic matter to the use of cleaner energy sources, such as electricity, oil, and natural gas [1,22].

2.4. Environmental Governance and Regulation

Environmental regulations can be divided into the following two types according to the entities implementing environmental regulations: (1) formal environmental regulations refer to governmental regulations; (2) while informal regulation refers to market-based or incentive-based environmental regulations and voluntary agreements. When the formal environmental regulations implemented by the government are absent or weak, many groups will negotiate with local polluters to promote the realization of pollution reduction; that is, social groups pursue higher environmental quality based on their own interests (Yijun Yuan, Ronghui Xie, 2014) [3].

Scholars have conducted a lot of research on the relationship between environmental policies and corporate behavior. Zhu Dongbo, Ren Li, et al. (2017) [23] analyzed the heterogeneous impacts of environmental regulation on cleaner production and pollution-intensive enterprises from the angle of heterogeneous industries. They used Chinese provincial dynamic panel data and a systematic generalized moment estimation method, combined with measurement methods such as mixed OLS, individual fixed effects, random effects, and differential GMM, to explore the relationship between environmental regulation, FDI (foreign direct investment), and industrial green transformation from the perspective of an open economy.

Among them, the intensity of environmental regulation was represented by the punitive environmental tax, which can be afforded without implementing environmental regulation policies and the environmental governance costs caused by the implementation of environmental regulation. Environmental regulation is an important entry threshold for foreign investment, which can guide foreign investment to flow to cleaner production industries [23].

2.5. Urbanization and the Blooming of the Real Estate Industry

The acceleration of urbanization has led to the prosperity and development of the real estate industry, and the increase in construction area has also brought more carbon emissions. In order to satisfy local development, GDP increase, basic government operations, and social welfare, localities often gain benefits through the construction of infrastructure and real estate.

Based on data from Malaysia during 1971–2015, Bekhet and Othman (2017) [24] aimed to examine the relationship between CO₂ emissions and urbanization growth. They used F-bounds tests and VECM (vector error correction model) Granger causality, and examined the inverted U-shaped relationship between CO₂ emissions and urbanization in the long-term. The elasticity of CO₂ was found to be positively elastic in the early stage of urbanization, but turned to negatively inelastic at the later urbanization stage.

Ahmed, Z et al. (2019) [25] analyzed the nonlinear relationship between urbanization and CO₂ emissions over the period 1971 to 2014 in Indonesia. Their findings revealed an inverted U-shaped relationship between urbanization and CO₂ emissions. The elasticity of carbon emissions with respect to urbanization was 5.81, which implies that a 1% increase in urbanization would increase emissions by 5.81%. However, after achieving a threshold level, urbanization reduces CO₂ emissions, and a −1% increase in urbanization would reduce carbon emissions by 0.87%.

Since China embarked on the path of reform and opening up, its urbanization has shown a rapid development trend. The urbanization rate reached 53.7% in 2013, with an average annual growth rate of 3.10%—nearly twice the average annual growth rate (1.75%) of urbanization before the implementation of reform and opening-up policies. At the same time, total carbon emissions increased to 6.2 times their levels 35 years ago, and per capita carbon emissions increased to 4.4 times their level 35 years ago. The process of urbanization is accompanied by high carbon emissions. This indicates that future urbanization development in China will face huge pressure due to high carbon emissions. How to reduce carbon emissions in the process of urbanization has become an urgent problem to be solved [3]. The research of Lin Boqiang and Liu Xiying (2010) [26], Liu Mengqin et al. (2011) [27], He Jiduo (2010) [28], and Wang, SJ et al. (2018) [29] also proved that the process of urbanization directly aggravated CO₂ emissions, and there is a “U-shaped” relationship between urbanization and carbon emissions in the Pearl River Delta [30,31].

Some results present significant temporal and spatial differences in the effects of the urbanization quality on CO₂ emissions between provinces. Wang, YN et al. (2019) [32] established an evaluation system for urbanization quality to estimate the urbanization development level. The geographically weighted regression (GWR) model was employed to examine the impact of the urbanization quality on CO₂ emissions, and revealed the spatial differences in 30 provinces in 2000, 2005, 2010, and 2015. Improvements in the urbanization quality have contributed to cutting CO₂ emissions in most provinces. The impact of the urbanization quality on CO₂ emissions in the central and western regions is greater than that in the eastern region.

The review of the literature shows that the mainstream paradigm for studying the relationship between economic expansion and emissions is still based on the environmental Kuznets hypothesis [7]. However, most existing articles have paid little attention to providing a clear picture of intricate mechanisms, including direct and indirect pathways of action for the effects of economic output, industrial structure, and governance.

3. Methods and Data Sources

3.1. Panel Regression Analysis: Direct Effects and Combined Effects

We utilized panel data of 30 provinces from 2008 to 2019 to quantify and estimate the impacts of various factors on carbon emissions from the construction industry. Equation (1) is the baseline regression model, regressing the urbanization development level and other macroeconomic scale or structural factors to the construction-related carbon emissions in each province—the typical research object in this essay. Whether to use a fixed-effects or random-effects regression model was determined later, based on the Hausman test.

(1)$\begin{array}{l}ConsCar{E}_{it}={\alpha }_i+{\lambda }_t+{\beta }_1\cdot Green{F}_{it}+{\beta }_2\cdot Green{F}^2{}_{it}+{\beta }_3\cdot Green{F}^3{}_{it}\\ +\overrightarrow{\gamma }\cdot {\overrightarrow{X}}_{it}+{\beta }_4{\overrightarrow{X}}_{it}\cdot Green{F}_{it}+{\epsilon }_{it}\end{array}$

The explained variable $ConsCarE$ represents the carbon emissions of the provincial construction industry. Considering that there may be a potential nonlinear relationship between the carbon emissions and this core explanatory variable GreenF, its square term GreenF² and cubic term GreenF³ are introduced in the empirical analysis. Several $\beta$s are the corresponding regression coefficients to be estimated.

The vector $\overrightarrow{X}$ is a set of macroeconomic variables, including economic scale and economic structural factors that probably affect the carbon emissions of the construction industry, which are treated as the control variables; $\overrightarrow{\gamma }$ is the coefficient vector of the corresponding control variables; ${\alpha }_i$ and ${\lambda }_t$ represent the fixed effects of each province and year, respectively, given that emissions are heterogeneous across years and provinces; ${\overrightarrow{X}}_{it}\cdot Green{F}_{it}$ is the intersection of GreenF and the control variables, indicating combined effects on construction emissions; ${\epsilon }_{it}$ is the disturbance term, considering other unknown influencing factors.

The production activity of the construction industry is influenced by the demand side of economic growth, population size, and consumption level. The needs of different industries for infrastructure construction also vary greatly, and policy constraints and environmental governance also affect the efficiency of cleaner production. There is great heterogeneity in the interprovincial macroeconomic environment, and the effects of factors such as population size, industrial structure, and income level were taken into consideration to more accurately fit the mechanism and estimate their impact. The specific indicators and economic implications are described below.

Resip indicates the permanent residents of each province; UrbanR, representing the urbanization development, is calculated from the proportion of urban residents to all residents. SecIR represents the percentage of GDP added by secondary industry; TerIR represents the percentage of GDP added by tertiary industry; FinR represents ratio of added value of the financial industry to GDP; REInvR represents the ratio of real estate investment to GDP; and NewAreaR represents the new housing construction area of real estate development enterprises in a certain year. The statistical descriptions of these variables are presented in Table 2, and it is evident that the indicators reflect a wide heterogeneity of development environments.

3.2. Explained Variable: Carbon Emissions from the Construction Industry

Provincial construction carbon emissions are the research object in this paper. The emissions data were extracted from the CEADs (Carbon Emission Accounts and Datasets). CEADs is a common compilation of multiscale carbon accounting lists and social economy and trade databases covering China and other developing economies, which is led by China’s Science and Technology Department International Cooperation Project and Key R&D Plans, with the joint effort of the British Research Council, Cambridge University, Tsinghua University, and other institutions, clustering nearly 1000 Chinese scholars who collect and verify data by data crowdfunding.

This database has long been committed to creating an open, transparent, comparable, verifiable, and completely free carbon accounting database. The researchers at this institution calculate the carbon dioxide emissions inventory based on the latest energy consumption statistics published by the China Bureau of Statistics. This differs by 2.87% from the emission result calculated by using the emission factors obtained from the carbon-specific survey data.

In 2020, the CEADs’ database officially released the latest carbon dioxide emission inventories of China, including 30 provinces and cities, in “Scientific Data”, which is an authoritative publication of the Nature Publishing Group. Subsequently, the CEADs released the carbon emissions data updated for 2019, which have been used by a large number of mainstream researchers.

The emissions account includes CO₂ emissions from both fossil fuel combustion (i.e., energy-related emissions) and cement production (process-related emissions). Energy-related carbon dioxide emissions are converted from the carbon content of fossil fuels such as raw coal and gasoline when they are burned. We used mass balance to calculate emissions according to IPCC guidelines, as shown in Equation (2)

(2)$C{E}_{\mathrm{i}}=A{D}_{\mathrm{i}}\times NC{V}_{\mathrm{i}}\times C{C}_{\mathrm{i}}\times O$

where $C{E}_{\mathrm{i}}$ refers to carbon dioxide emissions from the combustion of fossil fuels. There are 26 types of fossil fuels in China’s energy statistics system for carbon dioxide emissions. $A{D}_{\mathrm{i}}$ refers to the “activity data” used to estimate emissions, referring to the burned amount of fossil fuel $i$; $NC{V}_{\mathrm{i}}$ represents the “Net Calorific Value”, which is the calorific value produced by the combustion of each physical unit of fossil fuel $i$; $C{C}_{\mathrm{i}}$ is the carbon content of the fuel, derived from the quantified carbon emissions/net calorific value; and $O$ represents “oxidation efficiency”, which represents the oxidation ratio during the combustion of fossil fuels.

4. Empirical Analysis

Before considering the constraints of environmental regulations, we first regressed some macro-variables such as population size, income level, and industrial structure with respect to emissions. In subsequent sections, in order to control the size of the construction industry and its carbon emissions more accurately, we replace the share of secondary industry value added to GDP (SecIR)for robustness analysis.

4.1. Baseline Regression Results

Utilizing a benchmark regression, we examined the effects of macroeconomic structures on the explained variable—carbon emissions from the construction sector. The regression results are shown in Table 3. According to the Hausman test, the fixed-effects model outperformed the random-effects model, so the provincial fixed effects should be included in the regression model.

We first put the secondary industry’s added value into the regression equation to control the impact of the secondary industry development requirements on construction carbon emissions. As is shown in column (1), the nonlinear effects of lnResiP, lnGDPpc, and SecIR are significant. The coefficients of SecIR and its squared term are positive and negative, respectively, indicating an inverted U-shaped relationship between construction carbon emissions and the contribution of secondary industry to GDP.

We added TerIR in column (2). With the rapid integration and development of tertiary industry, such as the information technology, service, and financial industries, technological innovation drives the scale of the tertiary industry, with relatively low construction carbon emissions. The regression results show that the effect of tertiary industry is significant and nonlinear, indicating that there is also an inverted U-shaped relationship between construction carbon emissions and the contribution of tertiary industry to GDP.

After entering the post-industrialization stage, the industrial structure is continuously optimized and upgraded, and the technology is improved to solve the pollution problem. The proportion of the tertiary industry in the national economy will progressively increase until it becomes the leading industry, and after the secondary industry exceeds the inflection point, it will gradually decline, both of which will curb the carbon emissions of the construction industry.

The coefficients of lnResip and its squared term in each province are negative and positive, respectively, indicating a U-shaped relationship between carbon emissions and total population. As calculated by the regression results in column (1), the inflection point is about 6.75, while the rising total population in most provinces has a facilitating effect on carbon emissions at the current stage.

The environmental Kuznets curve hypothesis also holds in China. The regression coefficients of lnGDPpc and its squared term are positively and negatively significant, respectively, at the 1% confidence interval, indicating an inverted U-shaped curve relationship between emissions and economic growth. From the regression results, we calculated that the inflection point is around 11, i.e., the per capita GDP is around RMB 59,871, which is about USD 8805. In 2021, there were still half of the provinces with lnGDPpc not reaching this critical value, showing that the increases in economic growth and income levels at this stage have a catalytic effect on carbon emissions.

We believe that the introduction of the indicator of real estate investment can more clearly and directly reflect the construction of non-productive fixed assets. In order to control the impact of the consumption of materials and energy in their construction, we included the share of real estate investment in GDP, and its squared term, in columns (2)–(5). The results show that the scale of real estate investment has a significant positive effect on the carbon emissions from the construction industry in the first order. As shown in columns (3) and (4), the effect pattern of real estate investment scale has an inverted U shape, with the inflection point around 0.32. This shows that real estate investment promotes carbon emissions from the construction industry in most provinces during this period.

Subsequently, for the sake of considering the influence of financial investment factors, we introduced the ratio of financial investment to GDP, and its squared term, in columns (4) and (5). The results show that the development of the financial sector has a suppressive effect on the carbon emissions of the construction industry, and that the goodness of fit is significantly improved and the explanatory power is effectively enhanced by adding the indicator of financial industry, compared with other regressions.

Mainly to perform the functions of financing and resource allocation, the initial demand for construction projects in the financial industry is relatively low. The financial support can be further strengthened in the development of the green building industry chain to assist in technological upgrading and, consequently, reduce carbon emissions.

4.2. Baseline Regression Results: Green Finance Indicators Added

In Table 4, we present the empirical results after adding the Green Financial Index and the urbanization rate with fixed panel effects. In columns (1), (3), and (4), the coefficients of green finance and its squared and cubic terms are positive, negative, and positive, respectively, and they are all significant at the 1% level. When controlling variables such as the population urbanization rate (PUrbanR), financial sector (FinR), and real estate investment (REinvR), the previous results are still robust. In column (2), the coefficients of GreenF and its multiple terms are not significant, probably owing to the inclusion of both secondary industries’ value added and the share of real estate investment, which has a multicollinearity problem. In columns (3) and (4), we consider replacing SecIR with FinR, which is not collinear with real estate investment, and the results show that the coefficient of GreenF turns out to be significant.

In addition, the coefficients of high-order terms of GreenF are significant, revealing an N-shaped curve relationship between GreenF and construction carbon emissions. This is calculated by the regression result in column (1), where the first inflection point is around 0.36, and the second inflection point is around 0.58. In other words, the impact path of GreenF on construction carbon emissions is to first promote, then to inhibit after exceeding the first inflection point, and finally turn to promote after reaching the second inflection point.

The increased level of green finance has successively produced an economic expansion effect and a structural transformation effect on carbon emissions. The initial increase in green finance has a catalytic effect on construction emissions, and the stage with a low Green Financial Index may be dominated by the economic expansion effect because the support for green industries expands the size of the economy, and the facilities involved in environmental protection and pollution control are under construction, thus raising carbon emissions. When the investment in green finance reaches a certain level, the function of screening green enterprises and promoting the green transformation that green finance itself carries will gradually appear. The proportion of support for green industries will keep increasing, and the means and facilities for emission reduction will be formed at large scale, at which time the total carbon emissions in the construction industry will instead decrease, and the effects of economic structural transformation will dominate.

Combined with the actual situation, the average value of the Green Finance Index in each province has gradually increased 0.07 in 2001, to 0.238 in 2020. China is still in the start-up phase of green finance, while moving along the left side of the inverted U-shaped curve, approaching the stage of crossing the inflection point and effectively curbing construction-related carbon emissions.

In columns (2), (3), and (4), the population urbanization rate is introduced with positive, negative, and positive coefficients for the primary, square, and cubic terms, respectively, revealing that the nonlinear effect of urbanization level on construction-related emissions also appears as N-shaped. According to the regression results of column (4), we calculated that the first inflection point is about 0.50, and the second inflection point is around 1.01. Considering the possible range in the actual situation, the impact of urbanization level on construction carbon emissions is at the inverted U-shaped stage, which means that the low level of urbanization promotes emissions and, after reaching the inflection point, it gradually shifts to suppression.

The root cause is that, in the middle and late stages of urbanization, people’s awareness of environmental protection is enhanced, and higher requirements for the living environment are put forward, which significantly reduce the construction industry’s carbon emissions. This is also verified by the changing trend in provinces and cities with high levels of urbanization, such as Beijing, Shanghai, and Liaoning.

The demand of the financial industry for construction is smaller than that of the secondary industry. In columns (1), (3), and (4), controlling the financial sector development variable FinR further, its coefficient is significantly negative, which indicates that financial sector development has a significant dampening effect on carbon emissions in the construction sector. In columns (2) and (4), the variable of real estate investment is introduced, and its coefficient is positive and significant, indicating that real estate investment significantly contributes to the increase in construction-related carbon emissions.

4.3. Robustness Tests

Comparing the regression results of each column in Table 5, it can be seen whether by means of indicator substitution, such as replacing REInvR with IncREE—the business income of real estate development enterprises—and replacing lnGDPpc with PerIncR—the per capita income of all residents—or by adding new control variables to the regression model, subterms of the level of green financial development and the population urbanization rate are still significant, indicating that the direct mechanisms of both on the carbon emissions remain robust.

In addition, we also explored whether there was any indirect influence path from green finance level to carbon emissions in the construction industry, by including various interaction terms. In column (1), the effect pattern of green financial development on carbon emissions is still N-shaped, the inflection points of 0.24 and 0.65, respectively; however, inter_PUrbanR × REInvR.—the interaction term between GreenF and REInvR—is not significant, probably owing to the insufficient precision of the real estate sector indicator REInvR.

Nevertheless, as expected, in columns (2) and (3), the coefficients of the interaction term between population urbanization rate and real estate industry indicators are significantly positive, which could indicate that the fundamental trigger for the real estate industry to influence the carbon emissions is the promotion of population urbanization. On one hand, the continuous promotion of urbanization provides fresh land supply for the real estate industry; On the other hand, it brings a constant demand for productive and non-productive construction for the real estate industry, thus leading to the intensification of carbon emissions. In column (2), the two inflection points are at 0.43 and 1.11 respectively, and the population urbanization rate specifically is shown as an inverted U-shaped effect on the carbon emissions.

The development of green finance is able to suppress carbon emissions from the construction industry indirectly by curbing the development of the real estate sector, considering that the coefficient of its interaction term with the business income of real estate development enterprises—Inter_GreenF × IncREE—is significantly negative, as shown in column (3). A probable explanation would be that green finance, as a specific initiative of China’s economic transformation, can not only play the function of screening green enterprises and promoting their green transformation by limiting investments in high-emission enterprises, but also provide financial support for their green technology upgrading and environmental transformation by issuing green financial products.

In the process of population urbanization, green finance is also capable of playing a constraining role in emissions reduction, as the combined effect of the Green Finance Index and population urbanization level—Inter_PUrbanR × GreenF—is significantly negative.

In the following section, we explore the magnitude of the total effect of green finance levels on carbon emissions in the construction industry. At the sample mean—that is, GreenF is 0.145, IncREE is 6.55, and PUrbanR is 0.477—holding the size of other variables constant, when the Green Finance Index increases by 0.01 units, and will act as a direct effect, increasing its emissions by 1.09 unit in the confidence interval. By limiting the high-emission scale expansion of the real estate industry, GreenF reduces the carbon emission of the construction industry by 0.47 units, and through the combined effect with the level of population urbanization, the carbon emissions are reduced by 0.53 units. In total, the level of green financial development has a more obvious restraining effect on the carbon emissions of the construction industry, leading to an emissions reduction of 0.09 units.

5. Conclusions

In this study, we examined panel data for 30 provinces or regions in China from 2008 to 2019 to evaluate the development of green finance and construction carbon emissions. In the context of promoting green finance, we sought to dissect which macro-factors contribute to growth in carbon emissions, and which can lead to reductions in carbon emissions. In this research, an N-shaped effect of the level of green finance development on the emissions was found. Today, the average level of green finance in China is approaching the first inflection point, which implies that the promotion effect of green finance may further increase carbon emissions in the construction industry in the short-term but, shortly, the promotion effect will gradually slow down, and turn into a negative suppression effect after crossing the first inflection point. Generally, green finance can play an important role in integrating and balancing economic growth and sustainable development. Relying on the government and financial sectors, implementing green finance is supported by national finances and guided by policy dividends, which stimulate more social capital to flow into low-carbon and environment-friendly industries. The paths of effects on carbon emissions in the construction industry can be divided into two categories: direct and indirect.

This research reveals that green finance enables an indirect effect by constraining the impacts of industrial structure and real estate market development on carbon emissions. With the advancement of urbanization, the restraining effect of green finance on carbon emissions in the construction industry shall be gradually amplified. Firstly, green finance can gather capital and channel it to sustainable green industries by issuing green bonds, green funds, green insurance, and other financial products, facilitating low-energy industries and raising the threshold for high-energy industries, thereby accelerating the optimization and upgrading of industrial structure and the green transformation of energy structure. Secondly, carbon emissions generated by the construction industry in the process of real estate construction can be reduced. As a capital-intensive and energy-intensive industry, the real estate industry is tightly connected with the financial industry and the construction industry. Green finance can guide the real estate industry to pay attention to green and sustainable development at the investment and financing end, to a large extent limiting its scale of disorderly expansion,

Moreover, the prosperity of tertiary industry and population urbanization will eventually continue to curb carbon emissions after the economic development enters the post-industrialization stage and a high level of urbanization. While it is an inevitable process for developing countries to experience rapid growth of carbon emissions during the period of industrialization and population migration, in contrast to the effect of green finance, this study concludes that the urbanization process boosts the development of the real estate sector and, consequently, raises carbon emissions. However, in terms of the total effect, the average level of population urbanization in China has already produced a disincentive impact on carbon emissions in the construction sector, which implies that the contribution of the traditional modes of economic and social development to carbon emissions has diminished.

The inhibiting influence of the level of green financial development on carbon emissions not only reflects the role of government management, but also reveals the spontaneous choice of the financial industry. Through policy guidance, government should stimulate the function of credit and investment funds to allocate resources by screening green projects. The preliminary results of these practices are not only a success of China, but also a solution for many developing countries to escape their current predicament. The experience of China shows the world that economic development does not necessarily require the sacrifice of the environment, because economic development itself can be green and sustainable.

Author Contributions

Conceptualization, Z.L. and L.W.; methodology, Z.L., L.W. and Y.J.; software, R.C. and Z.Z.; validation, W.J. and Y.J.; formal analysis, Z.L.; investigation, Z.L. and Z.Z.; resources, R.C. and Y.P.; data curation, Z.Z. and W.J.; writing—original draft preparation, Z.L. and L.W.; writing—review and editing, Z.L., Y.P. and K.Z.; visualization, Y.P. and K.Z.; supervision, R.C. and Z.L.; project administration, Z.L., Y.P. and K.Z.; funding acquisition, Z.L. and Y.J.; All authors have read and agreed to the published version of the manuscript.

Institutional Review Board Statement

Not applicable.

Informed Consent Statement

Not applicable.

Data Availability Statement

Publicly available datasets were analyzed in this study. This data can be found here: https://www.ceads.net.cn/ (accessed on 25 July 2022), https://www.yearbookchina.com/index.aspx (accessed on 25 July 2022).

Conflicts of Interest

The authors declare no conflict of interest.

Publisher’s Note: MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Figure and Tables

Figure 1. Trends in green finance and construction emissions in 30 provinces over the last decade. Notes: For the visibility of the graph, the unit of the corresponding carbon emissions indicator (construction CO2) in the graph is 10 million tons.

Table 1

Sub-indices that constitute the Green Finance Index.

First Level Indicators	Characterization	Index Description	Effect Symbols
Non-green credit	Proportion of interest expenditure in high-energy-consumption industries	Interest expenditures of the six energy-intensive industries/total industrial interest expenditures	—
Green investment	Investment in environmental pollution control as a percentage of GDP	Environmental pollution control investment/GDP	+
Green insurance	Agricultural insurance depth	Agricultural insurance income/gross agricultural output value	+
Governmental support	Proportion of fiscal environmental protection expenditure	Fiscal expenditure on environmental protection/general budget fiscal expenditure	+

Table 2

Descriptive statistics of variables.

Variables	Meanings	Obs	Mean	Std. Dev.	Min	Max
ConsCarE	Carbon emissions from the construction industry	360	1.992	1.438	0.0482	6.3263
GreenF	Green Finance Index	603	0.145	0.098	0.0418	0.839
lnGDPpc	Log value of GDP per capita	570	10.192	0.821	8.0064	12.009
PerIncR	Per capita income of all residents	473	9.3876	0.7203	7.7776	11.1247
Resip	Resident population of each province	620	4292.415	2725.643	258	11521
UrbanR	Proportion of urban residents to all residents	620	47.747	17.765	23.95	89.6000
SecIR	Ratio of the secondary industry’s added value to GDP	580	0.448	0.082	0.1615	0.5904
TerIR	Ratio of the tertiary industry’s added value to GDP	580	0.433	0.090	0.2830	0.8352
FinR	Ratio of added value of the financial industry to GDP	551	0.051	0.300	0.0063	0.1797
REInvR	The ratio of real estate investment to GDP	719	0.093	0.062	0.0024	0.4601
IncREE	Log value of real estate business income of real estate development enterprises	549	1631.124	2208.731	6.77	16160.51

Table 3

Baseline regression results.

Column	(1)	(2)	(3)	(4)	(5)	(6)
Explained variables	ConsCarE	ConsCarE	ConsCarE	ConsCarE	ConsCarE	ConsCarE
lnResiP	−20.7765 *	−15.2702	−9.7439	−21.2729 **	−19.5622 *	−4.5196
	(10.8164)	(10.9324)	(11.2987)	(10.7877)	(10.8720)	(11.3256)
sqlnResiP	1.5381 **	1.1538	0.9247	1.4856 **	1.4019 *	0.6066
	(0.7181)	(0.7250)	(0.7432)	(0.7180)	(0.7232)	(0.7466)
lnGDPpc	8.2828 ***	2.0998	7.9870 **	5.9178 *	5.4469 *	14.5094 ***
	(2.9558)	(3.4324)	(3.5274)	(3.1993)	(3.2003)	(3.5206)
sqlnGDPpc	−0.3621 **	−0.0757	−0.3446 **	−0.2607 *	−0.2322	−0.6423 ***
	(0.1430)	(0.1643)	(0.1687)	(0.1532)	(0.1536)	(0.1685)
SecIR	11.1627 **				8.4659	6.5792
	(5.5056)				(5.6892)	(6.1968)
sqSecIR	−10.8609 *				−7.9607	−10.8134
	(6.1031)				(6.3072)	(6.8928)
TecIR		13.7825 ***
		(5.1189)
sqTerIR		−16.2715 ***
		(5.6787)
FinR			−10.1234			−28.4805 ***
			(6.7992)			(9.2644)
sqFinR			−36.2049			33.1138
			(39.6242)			(46.6816)
REInvR		6.6515 **	7.3227 **	7.5189 **	6.3385 **
		(2.9876)	(2.8644)	(2.9991)	(3.0985)
sqREInvR		−9.5589	−11.6498 **	−11.4110 *	−9.2072
		(6.0503)	(5.7112)	(6.0771)	(6.2877)
_cons	18.3182	31.8291	−26.5727			−82.6221 *
	(42.7193)	(42.7467)	(45.0151)			(46.2622)
N	348	348	319	348	348	319
R2	0.191	0.216	0.268	0.194	0.204	0.269
R2_w	0.1906	0.2157	0.2678	0.1938	0.2038	0.2687

Notes: Standard errors in parentheses; * p < 0.1, ** p < 0.05, *** p < 0.01.

Table 4

Baseline regression results: green finance indicators added.

Column	(1)	(2)	(3)	(4)
Explained variables	ConsCarE	ConsCarE	ConsCarE	ConsCarE
GreenF	32.7958 ***	12.5505	26.0016 ***	30.3663 ***
	(−8.5384)	(9.2741)	(9.4567)	(9.4598)
sqGreenF	−73.6462 ***	−35.0863	−64.9587 ***	−76.6139 ***
	(20.6282)	(22.8065)	(23.1079)	(23.3546)
cubeGreenF	52.2290 ***	24.0348	45.7153 **	55.1206 ***
	(−15.8998)	(17.3085)	(17.7636)	(18.0698)
PUrbanR		70.1182 ***	66.3761 ***	64.2441 **
		(26.4498)	(24.9881)	(25.8576)
sqPUrbanR		−1.0 × 102 **	−1.0 × 102 **	−96.5335 **
		(44.2353)	(41.6430)	(42.8223)
cubePUrbanR		46.7329 *	46.1987 **	42.6659 *
		(24.1702)	(22.4688)	(22.9739)
lnResiP	−4.9782	−22.1085 *	−2.7415	−8.2520
	(−11.5205)	(11.7627)	(12.5694)	(12.7307)
sqlnResiP	−4.9782	1.6162 **	0.5457	0.8899
	(−11.5205)	(0.7882)	(0.8330)	(0.8413)
lnGDPpc	24.7925 ***	−11.4879 **	−1.7203	−7.2843
	(−4.312)	(5.7788)	(6.3177)	(6.2845)
sqlnGDPpc	−1.1906 ***	0.5520 **	0.1062	0.3489
	(−0.212)	(0.2755)	(0.3059)	(0.3054)
SecIR		6.8004	3.4564
		(6.5145)	(6.3497)
sqSecIR		−5.2656	−6.6487
		(7.4310)	(7.2063)
TerIR
sqTerIR
REInvR		5.0339		5.5674 *
		(3.2778)		(2.9472)
sqREInvR		−5.2658		−5.9281
		(6.5863)		(5.8577)
FinR	−35.3285 ***		−30.0492 ***	−18.6861 **
	(7.7403)		(9.5677)	(8.2858)
sqFinR	96.2805 **		23.8078	−13.2658
	(44.7647)		(51.1019)	(48.6126)
_cons	−1.3 × 102 **	115.0623 **	−19.9738	32.1781
	(−50.9514)	(51.9950)	(57.9111)	(58.9480)
N	319	346	319	319
R2	0.370	0.246	0.326	0.332
r2_w	0.3701	0.2455	0.3257	0.3320

Notes: Standard errors in parentheses; * p < 0.1, ** p < 0.05, *** p < 0.01.

Table 5

Robustness tests.

Column	(1)	(2)	(3)
Explained variables	ConCarE	ConCarE	ConCarE
GreenF	28.9121 ***	27.6396 *	71.9378 **
	(9.5575)	(15.2247)	(28.3891)
sqGreenF	−83.4013 ***	−76.5323 **	141.2046 **
	(25.184)	(34.488)	(69.2435)
cubeGreenF	62.8681 ***	58.7136 **	−79.8205 *
	(20.4791)	(23.2155)	(43.5438)
lnResiP	−4.9201	−10.3157	17.8925
	(12.5818)	(12.8975)	(15.1038)
sqlnResiP	0.6996	0.9915	−1.092
	−4.9201	(0.8462)	(0.9685)
ln GDPpc	(0.8347)	−8.9733
	−2.6885	(6.8283)
sqlnGDPpc	(5.5231)	0.4361
	0.1269	(0.3322)
PUrbanR	0.1983 **	80.1605 ***	107.3626 ***
	(0.0836)	(26.8994)	(31.4695)
sq PUrbanR	−0.0019 ***	−1.3 × 102 ***	−2.4 × 102 ***
	(0.0007)	(44.7441)	(68.057)
cube PUrbanR		56.3506 **	146.8401 ***
		(24.6034)	(49.6336)
FinR	−19.6173 **	−10.8666
	(−8.2971)	(19.737)
sq FinR	−4.4317	−43.5391
	(−48.4458)	(94.052)
REInvR	4.1487	−11.3815
	(−3.5524)	(7.7123)
sq REInvR	−11.3521 *	−8.6221
	(6.3063)	(6.1188)
inter_GreenF × REInvR	23.0255
	(20.1337)
inter_PUrbanR × FinR		−3.3105
		(46.2574)
inter_PUrbanR × REInvR		31.5141 **
		(13.0364)
inter_PUrbanR × GreenF		0.7208	−1.1 × 102 **
		(31.7759)	(50.8874)
PerlncR			−10.6762
			(7.5179)
sq PerlncR			0.5429
			(0.3902)
IncREE			0.5504
			(2.0671)
sqIncREE			−0.4191
			(0.3464)
cube IncREE			0.027
			(0.0191)
inter_PUrbanR × IncREE			5.4561 ***
			(1.8683)
inter_GreenF × IncREE			−7.2444 **
			(3.1828)
_cons	2.3611	48.3389	−39.9343
	(55.6086)	(59.4602)	(70.124)
N	319	319	317
R2	0.327	0.346	0.270
r2_w	0.3268	0.3463	0.2703

Notes: Standard errors in parentheses; * p < 0.1, ** p < 0.05, *** p < 0.01.

References

1. Sathaye, J.; Meyers, S. Energy use in cities of the developing countries. Annu. Rev. Energy; 1985; 10, pp. 109-133. [DOI: https://dx.doi.org/10.1146/annurev.eg.10.110185.000545]

2. Yuan, Y.; Xie, R. Does pollution reduction policy affect the threshold effect of industrial structure adjustment?. Econ. Rev.; 2014; 5, pp. 75-84.

3. Yuan, Y.; Xie, R. Research on the Effect of Environmental Regulation on Industrial Structure Adjustment—An Empirical Test Based on China’s Provincial Panel Data. China Ind. Econ.; 2014; 8, pp. 57-69.

4. Friedlingstein, P.; Andrew, R.M.; Rogelj, J.; Peters, G.P.; Canadell, J.G.; Knutti, R.; Luderer, G.; Raupach, M.R.; Schaeffer, M.; van Vuuren, D.P. et al. Persistent growth of CO₂ emissions and implications for reaching climate targets. Nat. Geosci.; 2014; 7, pp. 709-715. [DOI: https://dx.doi.org/10.1038/ngeo2248]

5. Byeon, D.-H.; Jung, J.-M.; Park, Y.; Lee, H.-S.; Lee, J.-H.; Jung, S.; Lee, W.-H. Model-based assessment of changes in the potential distribution of Solenopsis geminata (Hymenoptera: Formicidae) according to climate change scenarios. J. Asia-Pac. Biodivers.; 2020; 13, pp. 331-338. [DOI: https://dx.doi.org/10.1016/j.japb.2020.03.011]

6. Schlesinger, M.E.; Malyshev, S.; Rozanov, E.V.; Yang, F.; Andronova, N.G.; de Vries, B.; Grübler, A.; Jiang, K.; Masui, T.; Morita, T. et al. Geographical Distributions of Temperature Change for Scenarios of Greenhouse Gas and Sulfur Dioxide Emissions. Technol. Forecast. Soc. Chang.; 2000; 65, pp. 167-193. [DOI: https://dx.doi.org/10.1016/S0040-1625(99)00114-6]

7. Grossman, G.M.; Krueger, A.B. Environental Impacts of a North America Free Trade Agreement; NBER Working Paper 3914 NBER: Cambridge, MA, USA, 1991.

8. Ozturk, S.; Cetin, M.; Demir, H. Income inequality and CO₂ emissions: Nonlinear evidence from Turkey. Environ. Dev. Sustain.; 2021; pp. 1-18. [DOI: https://dx.doi.org/10.1007/s10668-021-01922-y]

9. Shan, Y.; Zhou, Y.; Meng, J.; Mi, Z.; Liu, J.; Guan, D. Peak Cement-related CO₂ Emissions and the Changes in Drivers in China. J. Ind. Ecol.; 2019; 23, pp. 959-971. [DOI: https://dx.doi.org/10.1111/jiec.12839]

10. Panayotou, T. Empirical Tests and Policy Analysis of Environmental Degradation at Different Stages of Economic Development. Pac. Asian J. Energy; 1993; 4, Available online: https://www.academia.edu/download/79436985/93B09_31_engl.pdf (accessed on 25 July 2022).

11. Wheeler, D.; Pargal, S. Informal Regulation of Industrial Pollution in Developing Countries: Evidence from Indonesia. J. Political Econ.; 1996; 104, pp. 1314-1327.

12. Yang, Z.; Wu, S.; Cheung, H.Y. From income and housing wealth inequalities to emissions inequality: Carbon emissions of households in China. Neth. J. Hous. Built Environ.; 2017; 32, pp. 231-252. [DOI: https://dx.doi.org/10.1007/s10901-016-9510-9]

13. Mi, Z.-F.; Pan, S.-Y.; Yu, H.; Wei, Y.-M. Potential impacts of industrial structure on energy consumption and CO₂ emission: A case study of Beijing. J. Clean. Prod.; 2015; 103, pp. 455-462. [DOI: https://dx.doi.org/10.1016/j.jclepro.2014.06.011]

14. Wang, Z.; Yang, L. Delinking indicators on regional industry development and carbon emissions: Beijing–Tianjin–Hebei economic band case. Ecol. Indic.; 2015; 48, pp. 41-48. [DOI: https://dx.doi.org/10.1016/j.ecolind.2014.07.035]

15. Duan, H.; Mo, J.; Fan, Y.; Wang, S. Achieving China’s energy and climate policy targets in 2030 under multiple uncertainties. Energy Econ.; 2018; 70, pp. 45-60. [DOI: https://dx.doi.org/10.1016/j.eneco.2017.12.022]

16. Li, F.; Xu, Z.; Ma, H. Can China achieve its CO₂ emissions peak by 2030?. Ecol. Indic.; 2018; 84, pp. 337-344. [DOI: https://dx.doi.org/10.1016/j.ecolind.2017.08.048]

17. Steckel, J.C.; Jakob, M.; Marschinski, R.; Luderer, G. From carbonization to decarbonization?—Past trends and future scenarios for China’s CO2 emissions. Energy Policy; 2011; 39, pp. 3443-3455. [DOI: https://dx.doi.org/10.1016/j.enpol.2011.03.042]

18. Guan, D.; Meng, J.; Reiner, D.M.; Zhang, N.; Shan, Y.; Mi, Z.; Shao, S.; Liu, Z.; Zhang, Q.; Davis, S.J. Structural decline in China’s CO₂ emissions through transitions in industry and energy systems. Nat. Geosci.; 2018; 11, pp. 551-555. [DOI: https://dx.doi.org/10.1038/s41561-018-0161-1]

19. British Petroleum. BP Statistical Review of World Energy; 68th ed. British Petroleum: Londen, UK, 2019; pp. 54-96.

20. Yu, S.; Zheng, S.; Li, X.; Li, L. China can peak its energy-related carbon emissions before 2025: Evidence from industry restructuring. Energy Econ.; 2018; 73, pp. 91-107. [DOI: https://dx.doi.org/10.1016/j.eneco.2018.05.012]

21. Zhang, X.; Zhao, X.; Jiang, Z.; Shao, S. How to achieve the 2030 CO₂ emission reduction targets for China’s industrial sector: Retrospective decomposition and prospective trajectories. Glob. Environ. Chang.; 2017; 44, pp. 83-97. [DOI: https://dx.doi.org/10.1016/j.gloenvcha.2017.03.003]

22. Gates, D.F.; Yin, J.Z. Urbanization and energy in China: Issues and implications. Urbanization and Social Welfare in China; Routledge: London, UK, 2018; pp. 351-371.

23. Zhu, D.; Ren, L. Environmental Regulation, Foreign Direct Investment and China’s Industrial Green Transformation. Int. Trade Issues; 2017; 11, pp. 70-81. [DOI: https://dx.doi.org/10.13510/j.cnki.jit.2017.11.007]

24. Bekhet, H.A.; Othman, N.S. Impact of urbanization growth on Malaysia CO₂ emissions: Evidence from the dynamic relationship. J. Clean. Prod.; 2017; 154, pp. 374-388. [DOI: https://dx.doi.org/10.1016/j.jclepro.2017.03.174]

25. Ahmed, Z.; Wang, Z.; Ali, S. Investigating the non-linear relationship between urbanization and CO₂ emissions: An empirical analysis. Air Qual. Atmos. Health; 2019; 12, pp. 945-953. [DOI: https://dx.doi.org/10.1007/s11869-019-00711-x]

26. Lin, B.; Liu, X. China’s Carbon Dioxide Emissions under the Urbanization Process: Influence Factors and Abatement Policies. Econ. Res. J.; 2010; 8, pp. 66-78.

27. Liu, M.; Liu, Y. The Relationship between the Development of China’s Urbanization and Carbon Emissions: An Empirical Study Based on 30 Provinces’s Data. Urban Stud.; 2011; 18, pp. 27-32.

28. He, J. Emipirical Analysis of Dynamic elationship between Urbanization and Carbon Emission Level in China. West Forum; 2010; 20, pp. 78-86.

29. Wang, S.; Zeng, J.; Huang, Y.; Shi, C.; Zhan, P. The effects of urbanization on CO₂ emissions in the Pearl River Delta: A comprehensive assessment and panel data analysis. Appl. Energy; 2018; 228, pp. 1693-1706. [DOI: https://dx.doi.org/10.1016/j.apenergy.2018.06.155]

30. Hu, J.; Jiang, X. Research on the effect of urbanization on carbon emissions from the perspective of urban agglomerations. J. China Univ. Geosci. (Soc. Sci. Ed.); 2015; 15, pp. 11-21. [DOI: https://dx.doi.org/10.16493/j.cnki.42-1627/c.2015.86.002]

31. Xu, S.-C.; He, Z.-X.; Long, R.-Y.; Shen, W.-X.; Ji, S.-B.; Chen, Q.-B. Impacts of economic growth and urbanization on CO₂ emissions: Regional differences in China based on panel estimation. Reg. Environ. Chang.; 2016; 16, pp. 777-787. [DOI: https://dx.doi.org/10.1007/s10113-015-0795-0]

32. Wang, Y.; Li, X.; Kang, Y.; Chen, W.; Zhao, M.; Li, W. Analyzing the impact of urbanization quality on CO2 emissions: What can geographically weighted regression tell us?. Renew. Sustain. Energy Rev.; 2019; 104, pp. 127-136. [DOI: https://dx.doi.org/10.1016/j.rser.2019.01.028]