Low-carbon transformation of China’s cities: evaluation and spatiotemporal pattern evolution

Abstract

The challenges of global warming and the rapid and considerable growth of carbon emissions have attracted enormous attention from the government and society, highlighting the urgency of promoting regional low-carbon transformation, particularly in developing countries. This study focuses on China, the largest developing country, constructing an index system to evaluate cities’ low-carbon transformation and quantifying the index for 209 Chinese prefecture-level cities. This study then examines the spatiotemporal pattern evolution of Chinese cities’ low-carbon transformation and uses geographically and temporally weighted regression to investigate the influencing factors. The relevant results are threefold. (1) From 2011 to 2019, the index rose rapidly, and the eastern region experienced the strongest effect. (2) Chinese cities’ transformation exhibits obvious regional agglomeration characteristics. (3) Population density, per capita foreign direct investment (FDI), the low-carbon pilot policy, and per capita fixed asset investment generally have positive impacts, while resource-based cities, infrastructure status, and GDP target growth have overall negative effects; however, the regression coefficients of the influencing factors exhibit obvious spatial imbalance, and the dominant influencing factor of low-carbon transformation in 2011 was per capita fixed asset investment. In 2015, per capita FDI and population density were the dominant factors of central and eastern regions’ low-carbon transformation, and per capita FDI and per capita fixed asset investment were the dominant factors of low-carbon transformation in the western region. In 2019, per capita FDI and GDP target growth rate were the dominant factors in all regions.

Introduction

Global warming has emerged as a significant international concern, so the urgent task for all nations is to advocate for the reduction of greenhouse gases, such as carbon dioxide, in order to achieve low-carbon development. For developing countries in particular, the balance of economic development and mitigating carbon dioxide (CO₂) emissions is difficult to achieve (Bai et al. 2019). Previous studies have noted that developing countries have experienced rapid CO₂ emissions growth in recent years, exhibiting a continuously increasing proportion, with a significant impact on global CO₂ emissions (Jiang et al. 2019). Consequently, the facilitation of low-carbon transformation in developing nations is imperative for sustainable global economic progress and the preservation of human health. Overall, developing countries produce more carbon emissions, have low economic development levels, outdated industrialization and urbanization processes, and limited technological development (Balsa-Barreiro et al. 2019). As the world’s largest developing country, China is experiencing an ongoing industrialization and urbanization process, and its low-carbon transformation is representative and typical. Therefore, investigating the current circumstances and driving factors of China’s low-carbon transformation will offer a valuable reference for other developing nations.

According to the International Energy Agency’s (IEA) CO₂ Emissions in 2022 report, global CO₂ emissions exceeded 36.8 billion tons in 2022, of which China’s CO₂ emissions reached 11.477 billion tons (IEA, 2023). As the largest developing country, following decades of rapid growth, China has enjoyed remarkable economic achievements and attracted worldwide attention. However, the early extensive development model has also incurred challenges with rapid and considerable CO₂ emissions growth. China is currently the largest energy-related carbon emissions country, which diminishes sustainable development and residents’ health (Bai et al. 2019). In response to this pressing challenge, the Chinese government’s commitment to achieving peak carbon emissions, attaining carbon neutrality, and accelerating regional low-carbon transformation has become increasingly vital.

For academic research, scientifically quantifying regional low-carbon transformations is the prerequisite for accelerating the implementation of effectively targeted policies in various regions. Previous studies have conducted research measuring low-carbon development in China. Some studies have employed single indices such as carbon emissions intensity (Sun, 2005; Zhou et al. 2019; Zeng et al. 2023) and total CO₂ emissions (Wang and Zhang, 2022; Lee et al. 2022; Dong et al. 2023) to quantify low-carbon development. This method is simple to calculate and relatively intuitive; however, it only considers carbon emissions as a function of a single factor. In addition, based on full factor indicators, many studies have adopted nonparametric methods such as the data envelopment analysis approach (Meng et al. 2016; Cheng et al. 2018) to measure carbon emissions efficiency (Lin and Chen, 2019; Yu and Zhang, 2021; Li et al. 2023; Kou and Xu, 2022). While these studies have enriched the body of research considerably, the low-carbon city is a system including economy, society, energy, and environment for more comprehensive insights (Wang et al. 2021). Carbon emissions intensity and efficiency cannot measure cities’ economic development, the resource environment, or urban construction (Zhou et al. 2015), which makes it difficult to capture the circumstances of low-carbon development and transformation comprehensively and accurately. Therefore, to quantify regional low-carbon development, some studies have constructed indicator systems that include multilevel variables such as low-carbon economy, environment, society, urban planning, and resource utilization (Wang et al. 2021).

Prior studies have offered extensive insights into China’s low-carbon development, serving as a valuable reference for assessing regional low-carbon transformation. However, there are several areas that warrant further exploration. First, existing research has been centered on evaluating China’s low-carbon development. However, there is a scarcity of studies that have measured and analyzed the transformation towards low-carbon across various regions in China. Previous literature has noted that low-carbon transformation should include two dimensions of low-carbon development and economic transformation (Wang et al. 2021). Although some studies have measured low-carbon development by constructing an indicator system (Wang et al. 2021), these are only one dimensional, and few studies have scientifically and comprehensively quantified China’s low-carbon transformation from these two dimensions. Second, previous research has predominantly focused on provinces in China (Price et al. 2013; Xie et al. 2017; Song et al. 2018) and a limited number of cities or individual cities (Tan et al. 2017; Zhou et al. 2015); however, studies have not measured or analyzed the low-carbon development of all Chinese cities, which cannot reflect the spatiotemporal evolution of prefecture-level cities’ low-carbon transformation comprehensively. Although Wang et al. (2021) measured the low-carbon development indicators of some cities in China, due to data limitations, the study lacked a longer time dimension for measurement, not examining low-carbon development changes in terms of the time dimension. Third, the existing literature has often used ordinary least squares, panel fixed effect models, quantile regression, threshold effect models, and other approaches to reflect the average impact of various factors on regional low-carbon development, neglecting the spatial heterogeneity of various economic and social factors’ impact.

In response, an evaluation index system is constructed to quantify regional low-carbon transformation from the perspectives of low-carbon development and economic transition. The low-carbon transformation index of China’s 209 cities is calculated, and its spatiotemporal pattern evolution characteristics are examined. The study also explores the spatial heterogeneity of the impact of various social and economic factors, and analyzes the dominant influencing factors in China’s regions in different periods. In addition, this paper explores how to extend the research findings to other developing countries. Due to the significant development heterogeneity in China’s different regions, the nation can be considered a microcosm of developing countries worldwide. Therefore, this study proposes targeted policy recommendations based on the effects of various influencing factors in different Chinese regions, and analogically applies them to developing countries at different stages of industrialization and urbanization.

Compared with existing research, the contribution of this study is reflected in the following three aspects. First, from the research perspective, the majority of the existing literature has only measured and investigated regional low-carbon development (Zeng et al. 2023; Lin and Chen, 2019; Wang et al. 2021); however, limited literature has quantified and analyzed regional low-carbon transformation. This study provides a comprehensive evaluation of the low-carbon transformation of Chinese cities from two dimensions, thereby calculating the index and filling a gap in existing research on measuring low-carbon transformation. Second, this study clearly reveals cities’ spatiotemporal pattern evolution and agglomeration characteristics, contributing valuable insights to understanding the characteristics of low-carbon transformation and providing empirical evidence to further advance cities’ low-carbon transformation. Third, in contrast to existing research that has examined the average impact of various influencing factors on regional low-carbon development, this study examines the spatial heterogeneity of various social and economic factors in China’s cities and the dominant influencing factors in different periods, which is beneficial for the government to develop strategically targeted policies based on local conditions. The study delivers a substantial policy reference for China to achieve its dual carbon goals. Furthermore, it offers practical solutions and pathways for other developing nations at diverse stages of industrialization and urbanization to accomplish low-carbon transformation.

The remainder of this paper consists of six parts. Section “Model design, index system construction, and data sources” presents our index construction. Section “Spatiotemporal pattern evolution of Chinese cities’ low-carbon transformation” examines spatiotemporal evolutionary characteristics of Chinese cities’ low-carbon transformation. Section “Spatial heterogeneity analysis of factors influencing Chinese cities’ low-carbon transformation” analyzes the heterogeneous spatial impacts of social and economic factors and the dominant influencing factors in different periods and regions. Section “Discussion” discusses the reference significance for other developing countries. Section “Conclusion and policy implications” details the research conclusions and policy implications.

Model design, index system construction, and data sources

This study’s research path is as follows. First, 9 third-level indicators and 15 fourth-level indicators are selected from the dimensions of low-carbon development and economic transformation, and the entropy method is applied to construct a system for evaluating the low-carbon transformation of Chinese cities. Based on this, the low-carbon transformation index of Chinese cities is calculated. Second, this paper reveals the spatiotemporal pattern evolution and agglomeration characteristics of low-carbon transformation in Chinese cities using this index. Third, using geographically and temporally weighted regression (GTWR) analysis, the spatial heterogeneity in the impact of various economic and social variables on urban low-carbon transformation is examined. Fourth, this study identifies the dominant influencing factors of urban low-carbon transformation in different periods. Finally, this paper examines how to extend the research findings to other developing countries. The technical roadmap of this article is illustrated in Fig. 1.

Fig. 1: Technology roadmap.

It shows the overall research approach and methodology of this article.

Model design

Employing the entropy method, this study constructs a novel low-carbon transformation index system for Chinese cities from low-carbon development and economic transformation perspectives. The values of the indicators and their spatiotemporal evolution can be revealed based on this index, filling the gap in existing research on low-carbon transformation measurement. The spatial heterogeneity in the impact of various socioeconomic factors on low-carbon transformation across Chinese cities is examined, and the dominant influencing factors in different periods are identified using GTWR. These findings provide valuable insights for different regions in China and other developing countries to formulate targeted policies based on local conditions and promote urban low-carbon transformation.

Entropy method

Based on previous research, this study employs the entropy method to construct the evaluation system (Dai et al. 2019; Wang et al. 2021). First, the positive indices are processed according to Eq. (1), and the negative indices are processed according to Eq. (2), where i represents the city, j represents the index, t represents the year, ${x}_{{ijt}}$ represents the value of the j index for year t in city i, and ${y}_{{ijt}}$ represents the value of ${x}_{{ijt}}$ after standardization. Second, ${p}_{{ijt}}$ is calculated according to Eq. (3). Third, the entropy value (${e}_{j}$) of index j is obtained according to Eq. (4), then determine the weight of each index according to Eq. (5). Finally, according to Eq. (6), the standardized index values are multiplied by their corresponding weights and then summed to obtain the composite low-carbon transformation index for each city in each year.

$${y}_{ijt}=\frac{{x}_{ijt}-\,\min {x}_{j}}{\max {x}_{j}-\,\min {x}_{j}},1\le i\le n,1\le j\le m,1\le t\le T$$

(1)

$${y}_{ijt}=\frac{\max {x}_{j}-{x}_{ijt}}{\max {x}_{j}-\,\min {x}_{j}},1\le i\le n,1\le j\le m,1\le t\le T$$

(2)

$${p}_{ijt}=\frac{{y}_{ijt}}{{\sum }_{i=1}^{n}{\sum }_{t=1}^{T}{y}_{ijt}}$$

(3)

$${e}_{j}=-\frac{{\sum }_{i=1}^{n}{\sum }_{t=1}^{T}{p}_{ijt}{\mathrm{ln}}\,{p}_{ijt}}{{\mathrm{ln}}\,nT}$$

(4)

$${W}_{j}=\frac{1-{e}_{j}}{{\sum }_{j=1}^{m}(1-{e}_{j})}$$

(5)

$${S}_{ijt}=\mathop{\sum }\limits_{j=1}^{m}{W}_{j}\times {y}_{ijt}$$

(6)

GTWR model

Due to the significant differences in social and economic conditions across China’s cities, spatial heterogeneity in their impacts on regional low-carbon transformation may exist. Notably, simple multiple linear regression models cannot reflect this feature, while geographically weighted regression (GWR) can capture the spatial heterogeneity of various low-carbon transformation influencing factors in different cities. Furthermore, the GTWR model incorporates the time dimension into the GWR model to simultaneously handle nonstationary spatiotemporal data (Huang et al. 2010). Therefore, the GTWR model is employed to examine the spatial differences in the impact of various socioeconomic factors. Based on Wu et al. (2014), the GTWR model is constructed as follows:

$$Inde{x}_{i}={\beta }_{0}({\mu }_{i},{v}_{i},{t}_{i})+\mathop{\sum}\limits_{k}{\beta }_{k}({\mu }_{i},{v}_{i},{t}_{i}){x}_{ik}+{\varepsilon }_{i}$$

(7)

where ${Inde}{x}_{i}$ is the low-carbon transformation index of city i; β₀ is the intercept term; ${x}_{{ik}}$ is the explanatory variable, including population density, per capita foreign direct investment (FDI), resource-based cities, low-carbon pilot cities, per capita fixed asset investment, infrastructure status, and GDP target growth rate; ${\mu }_{i}$ and ${v}_{i}$ are the latitude and longitude of city i, respectively; ${t}_{i}$ is the year; therefore, $\left({\mu }_{i},{v}_{i},{t}_{i}\right)$ represents the coordinates of city i; ${\beta }_{k}\left({\mu }_{i},{v}_{i},{t}_{i}\right)$ is the coefficient of various influencing factors in city i; and ${\varepsilon }_{i}$ is an error term.

Index system construction principle and method

Index system construction principle

Table 1 details the composition of the index system and the corresponding weights. At present, some research has measured regional low-carbon development using indicators such as carbon emissions intensity (Wang et al. 2021), industrial wastewater (Wang et al. 2021), and government low-carbon attention (Bao and Liu, 2022); however, economic transformation is also a significant dimension of low-carbon transformation. Notably, existing literature has not yet included economic transformation in the comprehensive measurement of low-carbon transformation. Therefore, this study references previous studies to measure economic transformation from the perspectives of green innovation (Magat, 1978) and financial transformation (Pan and Dong, 2023; Sun and Tang, 2022). By incorporating both low-carbon development and economic transformation dimensions, this study provides a comprehensive assessment of Chinese cities’ low-carbon transformation. Specifically, based on previous literature, 15 indicators from these two dimensions are selected to extend the existing research, forming an evaluation system for calculating the composite index.

Table 1 Low-carbon transformation index system.

Carbon emissions intensity can directly reflect regions’ low-carbon development circumstances, and is a negative indicator (Wang et al. 2021). Environmental pollution includes the average concentration of fine particulate matter (PM_2.5), the volume of industrial wastewater discharged, and the volume of industrial soot (dust) emissions. PM_2.5 includes a large quantity of toxic and harmful substances that can have a serious negative impact on air quality and visibility. Similarly, industrial wastewater and industrial soot (dust) are harmful to the environment, with negative attributes (Wang et al. 2021). Finally, lower energy consumption efficiency has a greater negative impact on the environment and diminishes low-carbon development; therefore, it has a negative direction.

This study employs two indices to measure government low-carbon attention and public environmental attention. First, government low-carbon attention can offer guidance and policies, contributing to cities’ low-carbon transformation (Bao and Liu, 2022). Second, in addition to government attention, low-carbon transformation also requires the joint effort of the public. Internet development has established a new path for China’s environmental governance (Mol and Carter, 2006). Using online platforms, the public can effectively follow, supervise, provide feedback, and discuss environmental pollution and governance issues. These indicators’ attributes are positive.

Economic growth can be measured using the per capita GDP and GDP growth rate, with positive indicator attributes. For industrial transformation, the growing dominance of the tertiary sector can reflect industrial structure upgrading and generate less pollution and carbon emissions overall; therefore, the indicator’s direction is positive (Wang et al. 2021).

Green innovation can not only improve enterprises’ production efficiency, but also reduces environmental pollution and energy use; therefore, its indicator direction is positive.

To comprehensively measure financial transformation, green finance development and financial inclusion are used. First, the development of green finance introduces additional financing opportunities for low-carbon industries and technology development (Pan and Dong, 2023; Du, 2023). Second, digital financial inclusion development is beneficial for sustainable economic development (Sun and Tang, 2022). Therefore, the directions of these indicators are positive.

The process of economic transformation must also improve human capital, so its attribute is positive.

Index system construction method

This study uses four features based on previous research to measure Chinese cities’ low-carbon development, encompassing carbon emissions intensity, environmental pollution, energy consumption, and low-carbon attention. Table 1 presented the definitions of each measure.

First, following Shan et al. (2017, 2018, 2019, 2022), carbon emissions data are obtained from the Carbon Emission Accounts and Datasets (CEADs). These data are then divided by regional GDP to calculate each city’s carbon emissions intensity, and missing values are supplemented using interpolation.

To comprehensively measure environmental pollution, this study uses the average PM_2.5 concentration, the volume of industrial wastewater discharge, and volume of industrial soot (dust) emissions. For energy consumption, energy consumption intensity, calculated by dividing each city’s total energy consumption by its GDP, is employed.

To quantify government attention to low-carbon development, keywords related to low-carbon topics are extracted from prefecture-level cities’ government work reports during the sample period, and Python is used to calculate their frequency. The frequency of these keywords, expressed as a proportion of the total number of words in the reports, is then used to represent the government’s attention to low-carbon development. The keywords include “air pollution,” “smog,” “ammonia nitrogen,” “nitrogen oxides,” “sulfur dioxide,” “carbon dioxide,” “low-carbon,” “greenhouse gases,” “waste material,” “environment,” “environmental quality,” “environmental protection,” “environmental supervision,” “environmental collaboration,” “pollution prevention and control,” “pollution treatment,” “treat pollution,” “air,” “air quality,” “garbage,” “green,” “green development,” “green water and mountains,” “energy consumption,” “new energy,” “clean energy,” “forest,” “ecology,” “ancient woodland,” “return more marginal farmland to forest,” “emissions reduction,” “dust fall,” “particulate matter,” “sustainable,” “PM_2.5,” and “PM₁₀.” To measure public environmental attention, this study uses the Baidu Index, which reflects public interest in specific topics. A set of environment-related keywords, such as “low-carbon,” “sulfur dioxide,” “carbon dioxide,” “environmental protection,” “environmental pollution,” “emissions reduction,” “water conservation,” “sustainable,” “air quality,” “green land,” “green,” “pollution discharge,” “clean energy,” “decontamination,” “global warming,” “ecology,” “acid rain,” “greenhouse effect,” “pollution,” “sewage,” “smog,” and “PM2.5”, is selected from the Baidu platform. The daily average frequency of these keywords is calculated and used as a proxy for public environmental attention.

As previously described, this study measures the economic transformation dimension from five aspects, encompassing economic growth, industrial transformation, green innovation, financial transformation, and human capital.

Per capita GDP and GDP growth rate are used to measure economic growth. Industrial transformation is measured using the ratio of the tertiary industry to regional GDP. Green innovation is measured by the per capita number of authorized green invention patents, as invention patents typically embody higher technological content and innovation, and are subject to stricter examination standards. This ensures greater quality and better reflects cities’ substantive innovation capabilities. This approach is also supported by previous research. (Hashmi, 2013). In addition, compared with the application form, authorized green invention patent data can better reflect the essence of regional innovation capabilities (Hsu et al. 2014).

Financial transformation is comprehensively measured using green finance development and financial inclusion. First, following Huang et al. (2022) and Zhang et al. (2023), green finance is incorporated into the index system, which consists of seven sub-indices, and the entropy method is applied for evaluation. Second, the Digital Inclusive Finance Index is used to measure financial inclusion.

Finally, the number of students in regular higher education institutions per 10,000 residents and the proportion of employees in computer services, information transmission, and software industries are used to measure human capital.

Data sources

This study obtains total carbon emissions data from CEADs¹ and GDP data from the China City Statistical Yearbook to calculate carbon emissions intensity. Average of PM_2.5 concentration is obtained from the Atmospheric Composition Analysis Group.² The government’s low-carbon attention is from prefecture-level cities’ government work reports. Public environmental attention is quantified by crawling keywords on the Baidu website. The number of green patent authorizations is obtained from the Chinese Research Data Services Platform³ (CNRDS). The green finance index data are obtained from the Ministry of Science and Technology, the Bureau of Statistics, and People’s Bank of China websites and various authoritative statistical yearbooks. The Digital Inclusive Finance Index is obtained from the Institute of Digital Finance at Peking University.⁴ The data sources of each influencing factor are identified in corresponding sections. All other variables are obtained from the China City Statistical Yearbook. By consulting provincial and city statistical yearbooks, some missing data are supplemented, and the interpolation method is used to supplement the remaining missing data.

Notably, the entropy method and GTWR require all indicators to not have missing values, whereas many annual indicators in autonomous prefectures and leagues have missing values and are excluded. According to our statistics, from 2011 to 2019, the sum of regional GDP of the selected sample cities accounts for an average of 88.55% of China’s GDP, the population accounts for an average of 72.81% of China’s total population, and the sum of carbon emissions accounts for an average of 95.48% of China’s total carbon emissions. Therefore, the research sample is clearly representative, and the impact of missing regions is relatively small, meaning that the approach taken in this study is reasonable.

Spatiotemporal pattern evolution of Chinese cities’ low-carbon transformation

Temporal change characteristics

This paper next calculates the mean index value for each year by region, illustrating the temporal evolution characteristics for China and three regions in Fig. 2. The low-carbon transformation index continuously increased at national and three major regions’ levels from 2011 to 2019. Specifically, the mean value of the index continuously increased from 0.0538 to 0.0966, reflecting an overall acceleration in Chinese cities’ low-carbon transformation. Furthermore, the eastern region exhibits the fastest low-carbon transformation, while the western region is better than the central region. This can be attributed to China’s concerted efforts in recent years to promote the construction of an ecological civilization, which have effectively strengthened resource conservation and environmental protection. These efforts also include exploring market-based mechanisms, such as carbon emissions trading pilot programs, and implementing policy initiatives like low-carbon pilot zones. These measures have reduced energy consumption and total carbon emissions, improved the ecological environment, and promoted low-carbon development (Lee et al. 2022). At the same time, following the 2008 global financial crisis, the Chinese economy gradually shifted from an extensive economic development model to a model that is predominantly driven by innovation and human capital (Xie et al. 2023). Significant improvements in industrial structure optimization, technological innovation, financial transformation, and human capital have occurred that have promoted economic transformation and accelerated urban low-carbon transformation (Zeng et al. 2023).

Fig. 2: Temporal evolution of Chinese cities’ low-carbon transformation index.

It shows the low-carbon transformation situation of China as a whole, eastern region, central region, and western region from 2011 to 2019.

Spatial evolution characteristics

This study selects cross-sectional data from 2011, 2015, and 2019. Figure 3 illustrates the spatial evolution characteristics of Chinese cities’ low-carbon transformation, revealing significant spatial agglomeration characteristics. Specifically, Fig. 3 shows that high-value areas are primarily concentrated in the Beijing–Tianjin region, the Yangtze River Delta, and the Pearl River Delta, which all have the strongest economic vitality in China. The low-carbon economy continues to develop as energy efficiency continuously improves (Long and Shao, 2015), resulting in rapid progress in low-carbon transformation. Spatial differences in the low-carbon transformation process are also evident at the city level, with Beijing, Shanghai, Guangzhou, and Shenzhen (the four first tier cities) consistently leading the low-carbon transformation process. Furthermore, agglomeration characteristics also emerged in some high-level administrative cities, with Nanjing, Hangzhou, Jinan, Wuhan, Xi’an, Taiyuan, Lanzhou, Chengdu, Changsha and other cities at the forefront. These cities have stronger economic foundations, including more favorable energy infrastructure, rational industrial structure, and relatively high green innovation and human capital levels. In addition, these cities’ administrative level is higher than that of ordinary prefecture-level city; therefore, they are more likely to attract national or provincial governments’ attention and can obtain more resources and policy support for low-carbon transformation.

Fig. 3: Spatial pattern of Chinese cities’ low-carbon transformation of in 2011, 2015, and 2019.

It displays the distribution of China’s low-carbon transformation index in 2011, 2015, and 2019.

Different low-carbon transformation dimensions

To determine the temporal evolution of China’s low-carbon development and economic transformation, the mean value of low-carbon development and economic transformation in each year by region is calculated, as shown in Fig. 4. The low-carbon development indices of various regions have consistently increased, suggesting that China has made significant progress in this area. In 2011 and 2012, the western region lagged behind the eastern region; however, since 2013, it surpassed the eastern region and significantly surpassed the central region and the national average, becoming the region with the fastest low-carbon development. The steady rise in the economic transformation index suggests that China has achieved substantial advancements in this area, marked by an enhanced quality of economic development. The economic transformation in the eastern region is significantly faster. The economic transformation in the western region is similar to the national average and better than that in the central region, which also implies that the economic transformation in the western and central regions is not ideal and depends heavily on resources and energy. The eastern region has a higher demand for low-carbon development and economic development level, with significantly better energy conservation and emissions reduction, energy structure adjustment, industrial structure transformation, green innovation, financial transformation, and human capital. China’s central and western regions are heavily dependent on resources and energy, predominantly relying on traditional industries to drive economic growth, which hinders economic transformation (Cheng et al. 2019). As a result, it is essential to promote environmental governance and energy transformation, accelerate the optimization of economic transformation methods, and further promote low-carbon transformation in these regions.

Fig. 4: Temporal evolution characteristics of Chinese cities’ low-carbon development and economic transformation.

It displays the changes in the average values of low-carbon development and economic transformation in various regions of China from 2011 to 2019.

Spatial heterogeneity analysis of factors influencing Chinese cities’ low-carbon transformation

Variable description

Referencing previous research (Rahman, 2017; Song et al. 2018; Yu and Zhang, 2021) and avoiding the duplication of the low-carbon transformation index system and selected influencing factor indices, the following indices are chosen to assess their impact on low-carbon transformation and their spatial heterogeneity.⁵

Population density (Density)

Increased population density in a region often leads to industrial agglomeration, which can increase the scale of resource development and energy use and raise carbon emissions (Rahman, 2017), hindering low-carbon development and regional low-carbon transformation. In contrast, increased population density may also suppress regional carbon emissions (Wang and Li, 2021) as it can promote the concentration of production factors and improve utilization efficiency, enhancing energy use and resource development efficiency, promoting low-carbon development and economic transformation. Furthermore, individuals tend to migrate to regions with favorable ecological conditions and more advanced economic development. An increase in population density may also indicate a higher level of low-carbon transformation, thus exerting a significant influence on the regional low-carbon transition.

Per capita FDI (Perfdi)

FDI may hinder regional low-carbon transformation, leading to the “pollution paradise” effect (Copeland and Taylor, 1994). Compared with developed countries, developing countries have lower environmental standards and less strict environmental regulations. Therefore, some developed countries’ polluting industries will shift to developing countries, leading to the “pollution havens” effect. Therefore, FDI may increase regional carbon emissions and hinder regional economic transformation. Conversely, FDI can introduce advanced equipment and technology from abroad (Frankel, 2003) and serve as a demonstration for host country enterprises through technology spillovers (Letchumanan and Kodama, 2000), with the “pollution halo” effect that can reduce environmental pollution and promote low-carbon development and economic transformation.

Resource-based cities (Resource)

In the past few decades, resource-based cities have had the significant influence on driving China’s economic growth; however, the continuous promotion of resource exploitation has resulted in resource depletion. In addition, energy supply infrastructures that are dominated by resource-based industries, heavy industries, and fossil fuels make urban development heavily dependent on these industries, resulting in large-scale energy consumption and decreased energy use efficiency (Yan et al. 2019; Li and Dewan, 2017), which hinders low-carbon development and economic transformation. The definition of resource-based cities is based on the National Sustainable Development Plan for Resource-based Cities (2013–2020).⁶

Low-carbon pilot cities (Low-carbon)

The Notice on Carrying out Pilot Work in Low-Carbon Provinces and Cities,⁷ issued in 2010, initiated efforts for low-carbon transformation in five provinces and eight cities. The second and third batches were determined in 2012 and 2017, respectively. This policy can accelerate energy transitions (Lee et al. 2022) and significantly reduce regional carbon emissions intensity by improving urban environmental management, optimizing industrial structure, and increasing urban innovation (Zeng et al. 2023). Therefore, the low-carbon pilot policy should have a significant impact on low-carbon development and economic transformation.

Per capita fixed asset investment (Perfixed)

Increased fixed asset investment may directly lead to higher carbon emissions due to the expansion of production activities. However, it may also generate scale effects, which can help improve energy efficiency and reduce carbon emissions, thereby contributing to regional low-carbon transformation.

Infrastructure status (Infrastructure)

Urban road infrastructure construction is inevitably accompanied by increased carbon emissions, and increased per capita road area also promotes energy consumption and transportation development (Lin and Chen, 2019), resulting in more carbon emissions (Xie et al. 2017). In contrast, infrastructure investment can promote the agglomeration of production factors and the economy, improve resource utilization and production efficiency, which are conducive to reducing carbon emissions (Aschauer, 1989).

GDP target growth rate (Target)

This section extracts specific numerical targets for GDP growth from government work reports, which are often described using modifiers such as “approximately,” “about,” “higher than,” “above,” “minimum,” and “not below.” For expressions given as intervals, the average value of the range is used as a representative figure. The central government’s growth targets are broken down through administrative hierarchy to local governments at all levels, becoming public targets (Li et al. 2019). Increased target growth can motivate the government to support economic development and enhance regional economic growth. However, local governments will relax environmental regulations and implement short-term economic development strategies that support the expansion of high pollution industries to achieve the expected growth goals (Du and Yi, 2022), which can increase carbon emissions and environmental pollution, and inhibit industrial upgrading, thereby impeding regional low-carbon transformation. However, some scholars have argued that economic growth targets do not necessarily have negative impacts, and the relationship is U-shaped (Chai et al. 2021).

Model verification and analysis

In this section, the dependent variable is the low-carbon transformation index, and population density, per capita FDI, resource-based cities, low-carbon pilot cities, per capita fixed asset investment, infrastructure status, and GDP target growth rate are independent variables. Then the GTWR model is used for empirical analysis. The results reveal that the R² and adjusted R² are both higher than 0.5,⁸ confirming that the GTWR model can be effectively used in this study.

Estimation results

After obtaining the estimation results of the GTWR model, ArcGIS software is employed to visualize the spatial distribution of the effects of influencing factors on low-carbon transformation for the years 2011, 2015, and 2019, using the natural breaks classification method. This approach directly reflects the spatial heterogeneity characteristics of the impact of population density, per capita FDI, resource-based cities, low-carbon pilot cities, per capita fixed asset investment, infrastructure status, and GDP target growth ratio, as shown in Figs. 5–11.⁹

Fig. 5: Regression coefficients of population density.

It shows the impact of population density on low-carbon transformation in 2011, 2015 and 2019.

Fig. 6: Regression coefficients of per capita foreign direct investment.

It shows the impact of per capita forign direct investment on low-carbon transformation in 2011, 2015 and 2019.

Fig. 7: Regression coefficients of resource-based cities.

It shows the low-carbon transformation of resource-based cities in 2011, 2015 and 2019.

Fig. 8: Regression coefficients of low-carbon pilot cities.

It shows the low-carbon transformation of low-carbon pilot cities in 2011, 2015 and 2019.

Fig. 9: Regression coefficient of per capita fixed asset investment.

It shows the impact of per capita fixed asset investment on low-carbon transformation in 2011, 2015 and 2019.

Fig. 10: Regression coefficients of infrastructure status.

It shows the impact of infrastructure status on low-carbon transformation in 2011, 2015 and 2019.

Fig. 11: Regression coefficients of GDP target growth rate.

It shows the impact of GDP target growth rate on low-carbon transformation in 2011, 2015 and 2019.

(1) Spatiotemporal differentiation is evident in the impact of population density. The regression coefficients remained positive in 2011, 2015, and 2019, indicating that the impact of population density on cities’ low-carbon transformation was positive during the sample period. In addition, the mean of its regression coefficient first increased and then decreased, exhibiting a significant overall increase, indicating that the promotional effect of population density on cities’ low-carbon transformation is generally increasing. In Fig. 5, the coefficients of population density are all positive and generally increased from 2011 to 2019.

As China’s economic development quality improves, population agglomeration has promoted the aggregation of physical and human capital factors, resulting in clear agglomeration effects. Population agglomeration promotes scientific and technological progress, enhances resource use efficiency, and contributes to regional low-carbon development and economic transformation (Wang and Li, 2021). In the early stages, positive and high-value areas were primarily located in regions such as the northeast, Xinjiang, Shandong, and Jiangsu; however, over time, they gradually shifted to areas with high population density such as Shandong, Jiangsu, and Shanghai, indicating that the effect of factor aggregation was more obvious in these regions.

(2) Spatiotemporal differentiation is also revealed in the impact of per capita FDI. In 2011, 2015, and 2019, the lower quantile and median of the regression coefficients of per capita FDI were positive, indicating that it was mainly positively correlated with cities’ low-carbon transformation. Furthermore, the mean values of regression coefficients were all positive and continued to increase significantly, indicating that the promotional effect of per capita FDI on cities’ low-carbon transformation continued to increase. Figure 6 reveals that mostly positive coefficients of per capita FDI, indicating that FDI predominantly had “technology spillover effects.” Advanced technology and management experience from abroad are beneficial for improving enterprises’ production and management efficiency, improving energy and resource utilization efficiency, and promoting low-carbon development and economic transformation (Frankel, 2003; Letchumanan and Kodama, 2000). In the early stages, positive and high-value areas were mostly located in the underdeveloped regions of China’s western and central regions, which could be attributable to the marginal diminishing effect of FDI, which has a more obvious impact in underdeveloped areas. The negative areas of per capita FDI were concentrated in the eastern region in 2011, which may be because these regions have taken over more polluting industries from developed countries, resulting in the “pollution paradise” effect of FDI (Copeland and Taylor, 1994), which hindered the regional low-carbon transformation. In 2015 and 2019, the positive high-value areas gradually shifted from the western and central regions to the eastern coastal areas of the northern region such as Shandong and Jiangsu. The regression coefficients were only negative in a few cities and increased overall, indicating a gradual shift from the “pollution paradise” effect to the “technology spillover effect” of FDI.

(3) Spatiotemporal differentiation is also demonstrated in the impact of resource-based cites on low-carbon transformation. The regression coefficients were negative in 2011, 2015, and 2019, when the low-carbon transformation index was relatively low. In addition, the mean values of their regression coefficients were all negative, and the numerical changes were minimal, indicating that resource-based policy had a relatively stable inhibitory influence. Figure 7 shows that the coefficients were almost all negative, and negative and high-value areas were generally focused in the western and central regions, indicating that low-carbon transformation in these cities remained a significant challenge. The industries of resource-based cities predominantly include heavy industries, and energy supply structures primarily rely on fossil fuels. This high pollution, high energy consumption extensive urban development model exacerbates environmental pollution and constrains high-quality economic development, hindering low-carbon transformation.

(4) Spatiotemporal differentiation is also evident in the effect of the low-carbon pilot policy. Although the minimum regression coefficient was negative in 2011, 2015, and 2019, the lower quantile and median were both positive, suggesting that the indices of low-carbon pilot cities in the sample period were high. In addition, the mean values of regression coefficients were all positive and continued to increase during the sample period, indicating that the positive influence of this policy on cities’ low-carbon transformation was consistently improving. In the early stage, negative values were mainly located in the northeast region, but in the later stage, only a few cities had negative regression coefficients. Figure 8 demonstrates that the influence of the low-carbon pilot city policy was mostly positive, and the intensity of the effect significantly increased over time. The low-carbon policy can accelerate energy transformation, optimize industrial structure, and improve innovation (Lee et al. 2022; Zeng et al. 2023), with increasingly significant impact on promoting low-carbon development and transforming economic development patterns.

(5) Spatiotemporal differentiation is also revealed in the impact of per capita fixed asset investment. Although the minimum regression coefficient of per capita fixed asset investment was negative in 2011, 2015, and 2019, most of the lower quantile and median values were positive, indicating that per capita fixed asset investment in the sample period primarily had a positive impact. Furthermore, the mean values were all positive and continued to decrease, indicating that the promotional effect of per capita fixed asset investment is weakening. Figure 9 also reveals that the regression coefficients of most regions were positive in 2011 and 2015; however, in 2019, the regions with negative coefficients increased significantly, indicating that the positive impact of fixed asset investment is weakening or even becoming negative.

The initial fixed asset investment had a scale effect and improved the energy utilization rate, reducing the CO₂ emissions produced per unit of investment. However, when the scale of fixed assets reaches a certain threshold, further investment will increase carbon emissions and pollution and hinder the low-carbon transformation. The negative and high-value regions in 2019 were primarily located in the northeast region, Shandong, and other areas, which may be attributable to the delayed transformation of industrial structure and the high fixed asset investments in heavy and polluting industries.

(6) Spatiotemporal differentiation is also demonstrated in the impact of infrastructure status. The regression coefficients of infrastructure status were almost all negative in 2011, 2015, and 2019, indicating a negative correlation between infrastructure status and low-carbon transformation. In addition, the mean values were all negative, and the absolute values continued to increase, indicating that the inhibitory influence of infrastructure status on cities’ low-carbon transformation was increasing. Figure 10 shows that the coefficients of infrastructure status were negative in most regions, and the absolute values continued to increase over time. The rationale for this may be that road infrastructure construction directly generates carbon emissions, and road infrastructure can indirectly stimulate energy demand (Lin and Chen, 2019) and increase carbon emissions (Xie et al. 2017), hindering cities’ low-carbon transformation of.

(7) Spatiotemporal differentiation is also revealed in the impact of GDP target growth rate. The regression coefficients of GDP target growth rate were almost all negative in 2011, 2015, and 2019, indicating that GDP target growth rate during the sample period predominantly had an inhibitory influence. Furthermore, the mean values were all negative, and the absolute values continued to increase notably during the sample period, indicating that the inhibitory effect of GDP target growth rate on cities’ low-carbon transformation has been significantly strengthened. Figure 11 reveals that GDP target growth rates generally hindered low-carbon transformation and this inhibitory effect intensified over time. In 2015 and 2019, the regression coefficients of all cities’ GDP target growth rates were negative. Local governments will relax environmental regulations, support polluting industries, and postpone industrial transformation to meet predetermined economic growth goals (Du and Yi, 2022). This can result in increased carbon emissions, exacerbate environmental pollution, and hinder cities’ low-carbon development, which becomes more evident with improved economic development quality. Negative and high-value regions were primarily located in the Beijing–Tianjin–Hebei region, Inner Mongolia, and Shandong, indicating that the GDP target growth rates set by the government in these areas has had a significant hindering influence.

Chinese cities’ low-carbon transformation and the distribution patterns of its influencing factors

Based on the regression coefficients from the GTWR model, this paper identifies the dominant influencing factors across different periods and regions by comparing the absolute values of the coefficients to determine their relative impact on regional low-carbon transformation. Figure 12 presents the low-carbon transformation level of cities in three regions and the dominant influencing factors in 2011, 2015, and 2019.

Fig. 12: Low-carbon transformation in 2011, 2015, and 2019 and the structural pattern of dominant influencing factors.

Based on the the absolute values of the coefficients across different periods and regions, it displays the low-carbon transformation level of cities in three regions and the dominant influencing factors in 2011, 2015, and 2019.

Notably, in 2011, 2015, and 2019, the eastern region was at a high level of low-carbon transformation, the western region was at a medium level, and the central region was at a low level. Benefiting from its latecomer advantages and strong policy support, the western region has achieved energy transformation and industrial upgrading, resulting in a higher level of low-carbon transformation than the central region. The dominant influencing factor is per capita fixed asset investment, which may be explained by the significant role of such investment in driving economic growth in 2011.

As shown in Fig. 12, in 2015, the dominant factor in the three major regions was per capita FDI, indicating that with China’s continuous promotion of opening up, the role of FDI in economic development became increasingly prominent as the new dominant factor in cities’ low-carbon transformation. Population density was also a dominant factor in eastern and central regions because senior talent tends to migrate to regions with superior ecological environments and economic foundations. Human capital becomes increasingly important with the continuous advancement of economic transformation. Increased population density suggests the agglomeration of human capital elements, which significantly impacts the promotion of cities’ low-carbon development and economic transformation. Furthermore, fixed asset investment remained a dominant factor in the western region, suggesting that the western region still heavily relied on such investments.

In 2019, per capita FDI was the dominant factor in the three major regions, indicating that FDI was the most important factor and steadily became the dominant factor of cities’ low-carbon transformation. The GDP target growth rate emerged as a dominant factor across the three major regions, indicating that higher economic growth targets increasingly impeded cities’ low-carbon transformation.

Discussion

This study explores the current circumstances of Chinese cities’ low-carbon transformation and spatial heterogeneity in impact of economic and social factors. It is inadequate to only investigate the low-carbon transformation in some regions of China. Therefore, this section also explores how to extend the research conclusions to other regions and developing countries, and considers how different industrialization and urbanization processes affect the extrapolation of our conclusions (Balsa-Barreiro et al. 2019).

First, in terms of other regions in China, as the research sample in this article has covered most parts of the country, their total economic output, population size, and carbon emissions have a pivotal position in the country and have strong representativeness. Therefore, the impact of the economic and social factors examined in this study on urban low-carbon transformation can be extended to other Chinese regions.

Second, a more important concern is proposing beneficial practical solutions for other developing countries based on the research conclusions of this study to promote their low-carbon transformation. Previous studies have suggested that industrialization and urbanization are significant factors contributing to increased carbon emissions. Developing countries often pursue rapid economic growth accompanied by rapid industrialization and urbanization processes. At this time, the demand for energy consumption is high and energy utilization efficiency is low, resulting in a rapid increase in carbon emissions (Balsa-Barreiro et al. 2019). Some scholars have also argued that the most important facilitators of environmental degradation are rapid economic growth and early industrialization (Toth and Szigeti, 2016). For example, in India, rapid GDP growth and industrialization have resulted in a rapid increase in CO₂ emissions (Balsa-Barreiro et al. 2019).

Although a positive correlation between economic growth, energy consumption, and environmental pollution is typically observed in developing countries, differing industrialization processes can also result in different energy utilization efficiency, resulting in different carbon emissions characteristics. For example, China is in a later stage of industrialization and urbanization, and improved electricity utilization efficiency and ecological benefit policies have increased the trend in decoupling between China’s economic growth and emissions in recent years (Balsa-Barreiro et al. 2019; Wang et al. 2019; Juaidi et al. 2019). However, developing countries such as India and African nations are in early stages of industrialization, resulting in lower energy efficiency and economic growth that is often accompanied by continuous carbon emissions increase (Wang et al. 2019).

These studies provide us with a good foundation to tentatively extend our conclusions to other developing countries based on China’s national condition. China also exhibits significant heterogeneity in various regions’ development circumstances, presenting a microcosm of developing countries around the world. The eastern region is in the late stage of industrialization, with high energy utilization efficiency, relatively complete urbanization, and a high level of low-carbon transformation. In contrast, the central and western regions are in the middle stage of industrialization, with imbalanced industrial structure, lagging innovation and high-end industry development capabilities, imbalanced industrialization speed and resource and environmental carrying capacity, and insufficient low-carbon economy development (Huang, 2013; Wen, 2021). Therefore, it can be inferred that relevant policies could be applied to developing countries in the early stages of industrialization and urbanization by drawing on the experience of western China.

For example, our research results employing GTWR demonstrate that fixed asset investment in central and western China has generally inhibited low-carbon transformation, which may be attributable to the relatively lagging industrialization and urbanization process in central and western China and relatively high fixed asset investment in heavy and polluting industries. In recent years, the Chinese government has enacted tax incentives, such as reductions in corporate income tax for high-tech industries in the western region. These measures have fostered the development of advantageous industries and facilitated industrial transformation, thereby mitigating this phenomenon. Therefore, developing countries in the early stage of industrialization and urbanization should also appropriately control the scale of fixed asset investment, rather than indiscriminately stimulating investment. Such countries can learn from China’s practices of implementing tax incentives for emerging industries, improving investment and energy efficiency to advance low-carbon transformation. The eastern region of China has a superior economic foundation and higher scientific and technological development. The Chinese government has also established many high-tech development zones in the eastern region; therefore, fixed asset investment in the eastern region has had a scale effect of investment and improved energy utilization, reducing CO₂ emissions generated per unit investment, generally promoting low-carbon transformation. This indicates that countries in the later industrialization and urbanization stages are able to use energy more reasonably and can leverage the investment scale effect and improve investment efficiency by establishing development zones.

In addition, developing countries at different stages of development can learn from some other relevant conclusions. For example, China’s resource-based cities have clearly hindered the low-carbon transformation process. Therefore, resource-based countries should also consider industrial structure adjustment and energy structure optimization, improving energy utilization efficiency, reducing the proportion of resource-based and heavy industries, and increasing the integration of clean and renewable energy. Progressive reforms, such as China’s implementation of low-carbon pilot city policies, which promote overall low-carbon transformation, serve as a strategy that developing countries can refer to. The government can provide additional financial support to pilot areas and encourage increased green investments. This study shows that the GDP growth targets set by Chinese cities have generally impeded their low-carbon transformation. Similarly, developing countries pursuing catch-up strategies to achieve rapid economic growth may increase carbon emissions and worsen environmental pollution. Therefore, it is crucial for these countries to set realistic growth targets and place greater emphasis on the quality of economic development and environmental sustainability.

This study did not use cross-border data because it is difficult to reveal the heterogeneity between different regions within a country from national level data, which can make the research appear vague and imprecise. Research based on data from Chinese cities can reveal the circumstances of low-carbon transformation in regions with different levels of development and explore the heterogeneous impact of various influencing factors on low-carbon transformation. However, incorporating an international perspective is also essential. Future studies should enhance the use of cross-border data to support more robust international analyses.

In summary, the research findings of this study can promote China’s low-carbon transformation, achieve the dual carbon goal as quickly as possible, and provide a reference for other developing countries with similar development levels as China. Moreover, the findings also provide useful solutions and practical paths for developing countries with industrialization and urbanization processes lagging behind China to achieve low-carbon transformation.

Conclusion and policy implications

Conclusion

This study constructs a low-carbon transformation index to calculate Chinese cities’ low-carbon development and economic transformation, and also examines the spatiotemporal evolution characteristics of this index, subsequently explores the spatiotemporal heterogeneity of the impact of various economic and social factors. Based on the analysis, four key conclusions are drawn.

First, the low-carbon transformation index continued to rise rapidly from 2011 to 2019, and the process of low-carbon transformation has continued to advance rapidly. From a national perspective, the average value increased from 0.053 to 0.091 from 2011 to 2019. The eastern region has the best low-carbon transformation situation, and the western region is better than the central region.

Second, Chinese cites’ low-carbon transformation exhibits obvious regional agglomeration characteristics. High indices in China were mostly concentrated in the regions with the strongest economic vitality and strength. At the city level, agglomeration characteristics also appeared in high administrative level cities such as Beijing, Shanghai, Shenzhen, Nanjing, Guangzhou, Hangzhou, Jinan, Wuhan, Xi’an, Taiyuan, Lanzhou, Chengdu, and Changsha that have better economic foundations and higher government levels, and their low-carbon transformation index remained stable at the forefront in 2011, 2015, and 2019.

Third, the GTWR results reveal obvious spatiotemporal heterogeneity across the proposed influencing factors. Population density, per capita FDI, the low-carbon pilot policy, and per capita fixed asset investment had an overall positive influence. In recent years, population density and per capita FDI have increased the low-carbon transformation index in the eastern coastal regions of the northern region, and per capita fixed asset investment hindered the low-carbon transformation in the northeast region and Shandong. Resource-based cities, infrastructure status, and GDP target growth rate all had negative impacts. The inhibitory effect of resource-based cities and infrastructure status was stronger in western and central regions, and GDP target growth rate had a significant hindering effect on low-carbon transformation in the Beijing–Tianjin-Hebei region, Inner Mongolia, and Shandong.

Finally, per capita fixed asset investment was the dominant influencing factor in all three regions in 2011. The effect of per capita FDI on regional low-carbon transformation strengthened in 2015, basically replacing per capita fixed asset investment as the dominant factor in the three regions. However, per capita fixed asset investment continued to play a significant role in the western region, highlighting the region’s ongoing reliance on fixed asset investments. In addition, with the human capital agglomeration in the eastern and central regions, population density has become the dominant factor. In 2019, per capita FDI remained a stable dominant factor in the three major regions, and GDP target growth rate also became the dominant factor. This indicates that in recent years, the rising emphasis on economic growth targets has increasingly posed challenges to Chinese cities’ low-carbon transformation.

Policy suggestions

Our findings regarding China’s low-carbon transformation can also provide significant insights for other developing countries that are also pursuing low-carbon transformation.

(1) Other developing countries and China exhibit similar economic and environmental characteristics, with industrialization and urbanization still in progress. Therefore, in the future process of industrialization and urbanization, developing countries’ governments should continue to enhance energy structures, reduce the use of polluting energy, support clean energy applications and the growth of clean energy industries, and strengthen environmental regulation by comprehensively using market-oriented and legal approaches. In addition, developing nations should reduce reliance on heavy industries, focus on developing high-end manufacturing and service industries, and improve production efficiency and energy utilization efficiency; support green innovation and promote the transformation of green patent technology achievements; promote the transformation of the financial system toward green and inclusive development; and increase investment in human capital, cultivate “dual carbon” professional talents, and promote economic transformation and upgrading.

(2) Developing countries should implement strategically targeted initiatives for low-carbon transformation and narrow the gaps between different regions. The low-carbon transformation in the different regions varies due to economic foundations, resource endowments, path dependence, and other reasons, indicating that cities in different regions should adopt different modes. For China, since cities in the eastern region have stronger economic foundations, they should focus on fostering green innovation, financial transformation, and enhancing human capital. The western and central regions lag behind, with economic growth that is heavily reliant on traditional industries and resource exploitation. Therefore, it is more important to accelerate the cultivation of new driving forces, support manufacturing industry optimization, cultivate emerging industries, and promote energy utilization efficiency. Based on this, it can be inferred that countries in the early stages of industrialization and urbanization should focus on improving energy efficiency and optimizing industrial structures to decouple economic growth from carbon emissions. In contrast, countries in the later stages should prioritize enhancing innovation capacity and developing human capital.

(3) The impact of various influencing factors in cities’ low-carbon transformation exhibits spatial heterogeneity. Therefore, when developing low-carbon transformation policies, the government should adapt to local conditions to make more strategically targeted decisions. For example, regions with underdeveloped economies should introduce policies to attract high-quality talent and foreign investment. With policy support, China’s central and western regions should transform the driving force of economic development and reduce reliance on fixed asset investment. Resource-based cities should optimize the energy and industrial structures. The government should also optimize investment structure, appropriately reduce reliance on infrastructure construction investment, and reasonably set GDP target growth rates. Furthermore, in recent years, per capita FDI has become a stable and dominant factor; therefore, the government should focus on optimizing the business environment, strengthening intellectual property protection, and generating technology spillover effects from FDI to promote cities’ low-carbon transformation.

These insights may also be relevant for other developing nations. For instance, countries rich in resources should prioritize the enhancement of energy efficiency, decrease the reliance on resource-based and heavy industries, and augment the utilization of clean energy. Resource-based countries can promote reform pilot policies to advance low-carbon development and environmental protection, extend successful experiences to other regions, and promote low-carbon transformation at the national level. Developing countries should set reasonable economic growth targets, emphasize the quality of economic growth and environmental benefits, and avoid exacerbating carbon emissions and environmental pollution by implementing catch-up strategies. At the same time, countries in the early stages of industrialization and urbanization should avoid indiscriminately stimulating investment and population aggregation. Instead, they should focus on enhancing investment and energy efficiency while leveraging the influence of high-quality talent aggregation to promote low-carbon transformation.