Low-Carbon Transformation and Common Prosperity: An Analysis of the “Inverted U-Shaped” Relationship

Abstract

Low-carbon transformation and common prosperity are critical pillars of China’s economic growth. To explore the mechanism relating the two, this paper analyzes how carbon efficiency influences the urban–rural income gap, including its transmission mechanism and heterogeneity, and uses panel data from 240 Chinese prefectural cities (2006–2019). The results reveal an “inverted U-shaped” relationship between the low-carbon transition and urban–rural income gap. Specifically, as the carbon emission efficiency improves, the impact of the low-carbon transition on the urban–rural income gap changes from positive to negative. This finding remains robust under robustness tests. The heterogeneity test indicates that the “inverted U-shaped” relationship exhibits regional heterogeneity, resource endowment heterogeneity, economic development stage heterogeneity, and urban–rural income gap level heterogeneity. Furthermore, urban low-carbon transition influences the urban–rural income gap through industrial structure, employment structure, and human capital. This paper discusses the combination of low-carbon transformation and common prosperity, and takes into account both ecological sustainability and social sustainability. The findings of this paper offer policy proposals for advancing the achievement of dual-carbon goals and common prosperity, and provide references for developing countries.

1. Introduction

In September 2020, General Secretary Xi Jinping announced China’s development goals of achieving “carbon neutrality” and “peak carbon” during the 75th United Nations General Assembly. China has proactively taken on international responsibilities, exemplifying its role as a global leader. China is in a pivotal stage of socioeconomic transition, with accelerating low-carbon initiatives forming a cornerstone of its high-quality economic development strategy. According to the 2023 Annual Report on China’s Policies and Actions to Address Climate Change released by the Ministry of Ecology and Environment, China’s CO₂ emissions per GDP unit dropped by more than 51% by 2022 compared with the 2005 levels, underscoring steady progress in the nation’s low-carbon economic transformation. This transition significantly influences various dimensions of China’s socio-economic progress, including the redistribution of industries, resources, and labor, which inevitably influences the urban–rural income gap.

General Secretary Xi Jinping has stressed that “common prosperity is an essential requirement of socialism and a defining feature of Chinese-style modernization.” He has advocated for “unswervingly following the path of common prosperity” and stated that “people’s happiness and well-being is the ultimate goal of promoting high-quality development.” In recent years, China’s urbanization rate has risen steadily, accompanied by a significant improvement in living standards. The National Bureau of Statistics claims that the Gini coefficient for per capita disposable income in 2019 was 0.465, a decline from 0.474 in 2012, indicating a reduction in the income gap among residents. However, with the Gini coefficient remaining in the range of 0.4–0.5, income inequality persists, and the urban–rural income gap remains a critical component of this disparity. Addressing the pronounced urban–rural income gap continues to be a pressing challenge for China’s socioeconomic progress. Against the backdrop of China’s ongoing economic and social transformation, exploring the effects of low-carbon transformation on the urban–rural income gap holds significant theoretical and practical value.

From a global perspective, the low-carbon transition serves not only as a critical strategy for mitigating climate change, but also as a pivotal pathway for developing nations to optimize economic structures, enhance social equity, and achieve sustainable development. However, the urban–rural income gap remains a persistent challenge in these countries, and the low-carbon transition may either exacerbate or alleviate this disparity, with profound global implications.

Several successful initiatives demonstrate how low-carbon policies can reduce urban–rural inequality. For instance, India’s Solar Water Pump Program has provided irrigation solutions for rural areas while stimulating local photovoltaic industry growth. Similarly, Ethiopia’s Green Legacy Program has enhanced carbon sequestration while improving rural livelihoods through large-scale afforestation. These cases illustrate how low-carbon transitions can narrow the urban–rural income gap when properly implemented.

Conversely, in many developing nations, low-carbon investments remain disproportionately concentrated in urban areas, leaving rural regions underserved and exacerbating disparities in energy transition benefits. Furthermore, advanced low-carbon technologies promoted by developed countries—such as hydrogen energy and carbon capture, utilization, and storage—often entail high costs, rendering them unsuitable for rural applications in developing economies. This mismatch may deepen technological dependency and widen income disparities.

Existing research highlights the potential adverse effects of low-carbon policies on vulnerable populations. Jakob et al. (2014) found that carbon pricing mechanisms (e.g., carbon taxes) could disproportionately burden low-income households by raising energy costs [1]. Bowen et al. (2015) analyzed the structural employment shifts induced by green growth, noting that, while low-carbon transitions may displace low-skilled workers in carbon-intensive sectors (e.g., coal mining), they also generate new green employment opportunities [2]. Carley et al. (2020) further emphasized the need for targeted workforce retraining and regional transition support in fossil fuel-dependent communities to mitigate job losses [3]. These factors collectively suggest that poorly designed low-carbon policies may inadvertently widen the urban–rural income gap.

The interplay between low-carbon transition and urban–rural inequality underscores a critical intersection of global climate governance and domestic equity. Ignoring regional disparities risks transforming climate policies into new drivers of inequality. Conversely, inclusive strategies—such as universal access to green energy and rural green job creation—can serve as powerful tools for reducing income gaps. Only by integrating low-carbon development with rural economic advancement can nations achieve truly sustainable and inclusive growth, ensuring that climate action does not exacerbate social stratification.

The current literature lacks a comprehensive analysis of how urban low-carbon transitions influence urban–rural income disparities. While some studies examine environmental regulations’ impacts on income inequality, few directly quantify the effects of low-carbon transformation. Moreover, existing research often focuses narrowly on either economic or ecological sustainability, neglecting social sustainability—particularly human well-being during transitional development. Addressing these gaps, this study adopts a novel, holistic framework that integrates economic, ecological, and social sustainability dimensions. By examining the synergy between the low-carbon transition and urban–rural common prosperity, this paper aims to contribute to policy designs that simultaneously achieve decarbonization and equitable development.

2. Literature Review

The urban–rural income gap has long been a subject of widespread concern, with much focus placed on identifying the key elements influencing urban–rural income. Numerous scholars have explored these factors from various perspectives. Specifically, their research primarily focuses on aspects such as urbanization, digitalization, and energy transition.

2.1. Urbanization and Urban-Rural Income Gap

Yuan et al. (2020) [4] discovered that, when measuring the urban–rural income gap (URIG) through urban–rural income or consumption metrics, urbanization tends to exacerbate this gap. Conversely, when URIG is assessed using inequality indices—such as the Thiel index or Gini coefficient—urbanization is correlated with urban–rural income convergence [4]. In contrast, Su et al. (2015) [5] adopt a heterogeneity perspective in their analysis and argue that, while urbanization may lead to an increase in the urban–rural income gap in certain provinces of China—including Anhui, Fujian, and Sichuan—it appears to reduce this disparity in other regions. This can largely be explained by how urbanization enhances individual resource endowment in rural settings by facilitating the transfer of surplus rural labor into cities, consequently raising rural incomes [5]. Additionally, Sulemana et al. (2019) identified a positive correlation between urbanization and income inequality within cities across Sub-Saharan Africa [6].

2.2. Digitalization and Urban–Rural Income Gap

Xia et al. (2024) proposed that the influence of the digital economy on the income gap between urban and rural areas in China exhibits spatial heterogeneity [7]. In addition, Fu et al. (2024) [8] proposed that a comprehensive dimension of digitalization should be considered when assessing the impact of digitalization, rather than relying solely on a single dimension or a composite index. In regions with significant urban–rural disparities, digital financial services and digital governance play a more crucial role in bridging the gap. Conversely, in areas where the disparity is less pronounced, factors such as digital infrastructure, production digitization, and supply chain digitalization are more influential. However, without appropriate policy guidance, the process of digitization risks aggravating the current digital divide, potentially deepening the development gap between urban and rural areas [8]. Some scholars argue that the digital economy’s impact on the urban–rural income gap may follow a non-linear pattern. Peng et al. (2023) [9] contend that the relationship between the digital economy and the urban–rural income gap is not merely linear; rather, it follows a U-shaped trajectory. In the early stages of rapid digital economy development, policymakers should prioritize harnessing the digital dividend effect to mitigate the urban–rural income disparity. Therefore, in the initial stage of the rapid development of the digital economy, the primary goal of policy-making should be to realize the potential of digital dividends to mitigate income inequality between urban and rural areas. In the later stage of the development of the digital economy, with the emergence of the digital divide, close supervision of policy makers is needed [9]. Li et al. (2024) [10] used a sample of prefecture-level cities in China and found that the progression of network infrastructure significantly diminishes urban–rural income inequality. Furthermore, this reduction is more marked in eastern and central cities, those with well-developed traditional infrastructure, and areas characterized by elevated administrative and management capacities [10].

2.3. Energy Transition and Urban–Rural Income Gap

Gao et al. (2023) [11] argue that the energy transition is instrumental in narrowing the urban–rural income gap in China. They find that, for every 1% increase in energy transition, the Tel index decreases by 0.24%. This effect exhibits significant regional heterogeneity; specifically, while the energy transition contributes to reducing urban–rural income disparity in western, eastern, and southern regions, its impact is not statistically significant in northern areas. Furthermore, their research indicates that the energy transition reduces the urban–rural income gap to a comparable degree across both more and less economically developed regions. However, these effects are not pronounced in areas with constrained clean energy capacity [11]. Zhang et al. (2024) [12] present a similar perspective, positing that China’s energy transition policy can facilitate rural revitalization through industrial upgrading as a mediating factor. Additionally, they assert that advancements in the digital economy enhance this effect. Nevertheless, this influence varies between developed and less developed regions; developed areas tend to reap more substantial benefits, whereas less developed regions encounter more significant challenges [12].

It is important to note that other scholars have examined factors influencing the urban–rural income gap from different angles, such as financial development [13,14], industrial growth [15], human capital accumulation [16], fiscal expenditure policies [17,18], and international openness [19].

2.4. Environmental Factors and Urban–Rural Income Gap

Scholars have extensively studied the interplay between environmental factors and the urban–rural income gap, but a large proportion of existing studies primarily focuses on how the income gap affects environmental outcomes, with limited exploration of how environmental changes impact the urban–rural income gap. For instance, Jing et al. (2021) employed the spatial Durbin model to reveal that a widening urban–rural income gap exacerbates air pollution [20]. Yan et al. (2023) found that an increasing income gap contributes to higher carbon emission intensity [21]. While some studies have examined the effects of environmental factors on the urban–rural income gap, they predominantly focus on the effects of environmental regulations, with mixed findings. For example, Lu et al. (2022) demonstrated that ecological compensation through horizontal transfer payments significantly reduces income inequality [22]. In contrast, Wang et al. (2022) noted that the establishment of National Ecological Civilization Construction Demonstration Counties has led to a widening of the urban–rural income gap [23].

2.5. Summary

In summary, the research results on urban–rural income gap are relatively abundant, but specific studies on the impact of the low-carbon transition on the urban–rural income gap is still scarce. Thus, the marginal contributions of this study are as follows: first, this study begins from a completely new perspective on the theoretical and empirical analysis of the impact effect of the low-carbon transition on the urban–rural income gap; to some extent, this is to supplement the existing research theories and research frameworks. Second, this study contributes to comprehensively exploring the linear and non-linear relationship between China’s low-carbon transition and the urban–rural income gap, and ensuring that the conclusions of the study are comprehensive and accurate. Third, it explores the transmission mechanism and heterogeneity of the low-carbon transition affecting the urban–rural income gap from a multi-dimensional perspective, which provides policy insights for achieving dual-carbon goals and common prosperity.

3. Theoretical Mechanism Analysis and Research Hypotheses

The investigation of the relationship between economic growth and income inequality originates from Kuznets’ (1955) seminal work on the inverted-U curve hypothesis, which argues that, along with the evolution of industrialization and urbanization, the income distribution gradually improves from the initial state of inequality, and ultimately reaches a more equal pattern of income distribution [24]. The Kuznets curve provides an important inspiration for studying economic transformation and income distribution. Based on this hypothesis, this study speculates that there may be an “inverted U-shaped” relationship between the low-carbon transition and urban–rural income gap. At the initial stage of the urban low-carbon transition, all regions actively formulate energy-saving and carbon reduction-related policies, and high-carbon enterprises, especially industrial enterprises, face the important task of transforming and upgrading, but completing the industrial upgrading and transformation is a long process; within a short period of time, if high-carbon enterprises cannot immediately adapt to the changes in the economic environment, it will lead to a decline in production efficiency, which will lead to a declining wage income for enterprise employees, especially low-skilled laborers, who may even face the risk of unemployment, thereby widening the urban–rural income gap. The de-industrialized economy will rely on the development of highly skill-intensive service sectors, such as finance and IT, but the absorptive capacity of these sectors for low- and middle-income workers is much lower than that of the manufacturing sector [25]. Therefore, de-industrialization will lead to the inflow of unskilled labor into the low-paying service sector [26]. These service sectors, which lack technological and transactional attributes, will not be able to develop in the long run, but will be able to absorb low-wage labor. The development of these service sectors is limited due to the lack of technical and transactional attributes [25]; “de-industrialization” also hinders service sector expansion due to the lack of support from the real economy [27], thus further exacerbating the urban–rural income problem. Along with the low-carbon transformation of the city to promote in-depth, pollution-intensive and energy-intensive manufacturing sectors have successively completed transformation and upgrading, production efficiency has been gradually improved, and the improvement in technological innovation will further improve the production efficiency of the enterprise, thereby reducing the production cost; workers’ wages and income are then increased, and the expansion of the enterprise scale provides more jobs for the unemployed, so the urban–rural income gap gradually converges. Next, this study will discuss the theoretical mechanism in detail from three aspects: industrial structure, employment structure and human capital. This selection establishes a comprehensive “macro–meso–micro” analytical framework. At the macro level, industrial structure transformation reveals how the low-carbon transition reshapes urban–rural economic patterns through industrial upgrading and resource reallocation. The meso-level employment structure mechanism captures labor market adjustments and differential adaptations to emerging job opportunities. At the micro level, human capital accumulation determines individuals’ capacity to benefit from transition opportunities.

This tripartite framework offers several analytical advantages. First, it provides a complete causal chain from macroeconomic restructuring to micro-level outcomes. Second, in China’s dual urban–rural economy, industrial and employment transformations disproportionately affect rural areas, while human capital serves as the critical factor in overcoming the “low-carbon poverty trap.” Third, these mechanisms are particularly suitable for China’s context, with readily quantifiable indicators and high data availability.

3.1. Industrial Structure

Industrial structure transformation encompasses two fundamental dimensions: industrial structure rationalization and industrial structure upgrading. Industrial structure rationalization reflects the degree of inter-sectoral coordination and optimal resource allocation, characterized by balanced industrial proportions, efficient factor mobility, and supply–demand matching. This dimension is typically quantified using metrics such as the Theil index and structural deviation degree. In contrast, industrial structure upgrading represents the evolutionary process from low-value-added to high-value-added production, from labor-intensive to technology-intensive sectors, and from traditional to emerging industries, as evidenced by the growing share of tertiary industries and technology-dominated sectors. This dimension is usually quantified using a weighted sum of the output of the primary, secondary, and tertiary sectors or by applying the ratio of the tertiary sector to the secondary sector. Together, these dimensions constitute the core aspects of industrial structure optimization, with rationalization emphasizing horizontal coordination and upgrading focusing on vertical progression.

The low-carbon transition drives industrial restructuring through three primary pathways: (1) environmental regulation, (2) technological innovation, and (3) market mechanisms. Carbon pricing policies and environmental regulations compel the phase-out or modernization of high-carbon industries while facilitating factor reallocation to more efficient sectors, thereby enhancing structural rationalization. Concurrently, breakthroughs in green technologies foster emerging industries, and digitalization coupled with service-oriented transformation accelerates knowledge-intensive sector development, promoting both tertiary industry expansion and high-tech industry growth—key manifestations of industrial structure upgrading. Furthermore, the low-carbon transition facilitates resource reallocation through green finance mechanisms and carbon market operations.

China’s economic development is currently undergoing significant structural transformation, with many cities experiencing distinct processes of deindustrialization. Among them, cities such as Shenzhen exemplify proactive industrial restructuring, emphasizing high-end manufacturing and service industries to mitigate the risks of economic hollowing-out resulting from the phase-out of low-end production capacity.

Shenzhen represents one of China’s most prominent cases of managed deindustrialization. The city has witnessed a steady decline in its industrial sector’s share of GDP, accompanied by a corresponding rise in the tertiary sector. Guided by policy interventions, low-end manufacturing has been systematically relocated to neighboring regions such as Dongguan and Huizhou, while domestic technology leaders like Huawei and DJI have driven the growth of high-value-added manufacturing. Despite maintaining the highest R&D investment intensity nationwide, Shenzhen has encountered challenges related to manufacturing relocation in recent years. In response, the municipal government has adopted an “industrial anchoring” strategy, prioritizing advanced manufacturing sectors such as new energy vehicles and semiconductors to ensure a balanced economic structure that retains a robust industrial base alongside service sector expansion.

In contrast, deindustrialization in northeastern Chinese cities reflects passive industrial decline, rather than strategic upgrading. Changchun, for instance, remains heavily reliant on traditional automotive manufacturing, which still accounts for over 50% of its economy, yet its transition to new energy vehicle production has been sluggish. State-owned enterprises (e.g., FAW Group) exhibit limited innovation capacity, while the private sector remains underdeveloped. Persistent outmigration has further exacerbated industrial contraction, resulting in a “low-end lock-in” effect. Unlike Shenzhen, the northeastern region lacks emerging industries to compensate for the decline in traditional sectors, highlighting divergent trajectories in China’s urban deindustrialization.

Some scholars have proposed that there is a problem of premature and rapid “de-industrialization” in China [28,29,30]. Cai (2021) [29] suggests that per capita GDP, the share of agriculture, and the development of service industries are measures of premature and rapid “de-industrialization”; when the per capita GDP enters ranks of high-income countries, the proportion of agriculture declines to a sufficiently low level, i.e., there is no longer any pressure on rural surplus labor transfer in a country and the service sector is at a superior position in the value chain. In the higher value chain position, the decline in the proportion of manufacturing will not lead to a decline in labor productivity [29]. Taking a comprehensive view, China’s current premature, too rapid “de-industrialization” arises from industrial structure upgrading, saving energy and reducing emissions emerging from the excessive pursuit of industrialization. Industry is, to some extent, considered to be highly polluting and energy-consuming, which leads to a decline in the momentum of industrial growth and expansion of the scale of the service industry [31]. China still relies on a high proportion of rural surplus labor, so it has not yet reached the mature conditions of “de-industrialization,” and due to the transformation and upgrading of the manufacturing sector requires a certain amount of time for the transition; premature “de-industrialization” impedes innovation in the manufacturing industry. The manufacturing industry is considered to be highly polluting and energy-consuming, despite R&D and productivity enhancement. As the transformation and upgrading of the manufacturing sector requires some time for transition, premature “de-industrialization” hinders innovation, R&D, and productivity improvement in various manufacturing sectors, and the decline in production efficiency in the manufacturing industry will further aggravate the problem of unequal urban and rural income. However, the service sector’s progress, which lacks the support of industrial development, will also be inhibited [27], leading to a further decline in the wage income of the labor force that flows to low-paying service sector jobs. Huang et al. (2022) suggest that the decarbonization of industrial processes and the circular economy, which involves the transformation and upgrading of equipment and the development and production of new materials, cannot be separated from the support of manufacturing enterprises [30]. Therefore, premature and rapid “de-industrialization” is not conducive to the low-carbon economy’s progress. Empirical evidence from Mehmet Akif Destek et al. (2024) confirms that premature deindustrialization exerts persistent adverse effects on environmental quality [32]. Conversely, during advanced stages of urban low-carbon transitions, manufacturing enterprises typically complete their transformation and upgrading processes. Subsequent technological innovation enhances production efficiency while reducing operational costs, as demonstrated in recent findings. Wang et al. (2025) further substantiate that industrial structure transformation and upgrading contribute significantly to reducing urban–rural income disparities [33]. Industrial restructuring and upgrading are also conducive to agricultural development, which promotes the modernization of agriculture through the “trickle-down effect,” consequently decreasing the urban–rural income inequality [34].

3.2. Employment Structure

First, at the initial phase of the low-carbon transition, many enterprises, especially those which are high-carbon, accelerate the pace of transformation and upgrading, and tend to change the production mode through enhanced technological innovation; this change will substantially boost the need for skilled workforce, forcing unskilled labor into mobility or face unemployment. Regarding the relationship between technology and employment, the British classical economist Ricardo put forward the idea that machines expel workers, resulting in a decline in the living conditions of the working class, and then to overpopulation [35]. Acemoglu (2002) argues that, after entering the twentieth century, technological progress has been transformed from skill substitution to skill bias, and that skill bias is more favorable for the employment of high-skilled labor, which consequently replaces the employment of unskilled labor [36]. However, there is still a significant volume of rural surplus labor in China, and the labor market is unable to match the demand for industrial structural transformation, thus triggering structural unemployment. The promotion of industrial intelligence and the popularization and application of industrial robots have accelerated the manufacturing greening process, which also has a “crowding out” effect on the employment of low-skilled labor. Moreover, along with the mobility and transfer of labor, urban–rural income disparity has gradually narrowed [37,38], which is also confirmed by Stark et al. (2009) [39], who concluded that the Gini coefficient is positively correlated with labor mobility. The intelligent transformation of the Wuhan Iron and Steel Company has significantly reduced high-risk occupational positions, particularly front-line steelmaking furnace operators. Concurrently, government–university partnerships have established “digital technician” training programs to facilitate workforce transition. Nevertheless, substantial numbers of displaced workers face structural unemployment due to age-related and educational barriers that limit their adaptability to emerging job requirements.

Similar patterns are observed in the automotive sector. Leading manufacturers, including Chongqing Changan Automobile, have implemented smart manufacturing systems in Liangjiang New Area, where robotic automation now handles welding and painting processes with over 40% workforce reduction per facility. While this technological shift has created new occupational categories, such as industrial robot operation and maintenance, many former production-line workers experience significant skill mismatches. Consequently, a portion of this workforce has been compelled to accept lower-paying alternative employment. On the other hand, Stark et al. (2009) demonstrated a positive correlation between the Gini coefficient and labor mobility, suggesting that workforce transitions can contribute to urban–rural income convergence over time [39]. In the late stage of urban low-carbon transition, when industries have basically completed the transformation and upgrading, the employment structure also tends to stabilize; all aspects of society have gradually adapted to the needs of low-carbon economy progress, the measures and infrastructure have gradually improved, the state attaches more importance to the development of rural education and skills training for rural labor force, which improves the suitability for the labor market, there is a gradual increase in the employment rate, and urban and rural income problems have been mitigated. The problem of urban and rural income has been alleviated.

3.3. Human Capital

Along with the advancement of industrial intelligence and industrial robotics technology, if the level of human capital is not able to keep pace with its development, a low-skilled labor force will develop that cannot be competent to operate industrial robots. Moreover, the problem of human–machine mismatch will lead to reduced productivity in industrial enterprises, which will aggravate the problem of urban and rural income. Wang et al. (2020) found that artificial intelligence through the job turnover effect and the productivity effect will exacerbate the income inequality between high and low-skilled sectors, and low-skilled sector income inequality [40]. Dauth et al. (2017) found that robotics applications benefit high-skilled managers and negatively affect the personal income of middle-skilled laborers who operate machines [41]. Building upon these findings, Zhang et al. (2025) further empirically validated the displacement effect of automation technologies on medium and low-skilled labor employment through the construction of a general equilibrium model [42]. Rodrik (2014) also emphasized the importance of human capital to achieve economic growth in developing countries, as it constitutes a key determinant of the urban–rural income gap [43]. However, an increase in human capital investment in farm households, especially education investment, can effectively reduce the income gap [44]. Notably, Suzhou pioneered China’s first “Professional Farmers College,” specializing in modern agricultural technology education. Similarly, Qingyuan district in Guangzhou has implemented poverty alleviation initiatives through skill development programs, including “Cantonese Cuisine Culinary Arts” and “South China Home Services” training, offered at no cost to underprivileged households in mountainous regions. These programs incorporate strategic partnerships with Guangzhou-based catering enterprises to ensure employment opportunities. These interventions have effectively enhanced farmers’ earnings. In the late stage of the low-carbon transition, the education level and skill training of the rural labor force develops and matures, and employment security and life security are improved. The improvement in rural human capital levels has narrowed the income gap between urban and rural areas. The improvement in human capital improves the corresponding degree of labor market and industrial intelligent development. As the amount of rural surplus labor decreases, transfers to high-tech enterprises and high-skill-intensive service industry, and the improvement in human–machine matching greatly advance the productive efficiency of industrial enterprises; subsequently, the expansion of the production scale absorbs more labor force employment, and the urban–rural income gap gradually converges.

Following the above analysis of theoretical mechanisms, we put forward the following hypotheses:

Hypothesis 1.

The urban low-carbon transition and the urban–rural income gap exhibit an “inverted U-shaped relationship.”

Hypothesis 2.

The urban low-carbon transition affects the urban–rural income gap through industrial structure, employment structure, and human capital mechanism.

4. Research Design

4.1. Model Setting

Given the panel data structure employed in this study, the fixed-effects model offers several methodological advantages. First, it effectively controls for time-invariant regional heterogeneity through city-level fixed effects. This is particularly relevant as urban low-carbon transition processes exhibit strong path dependence, being significantly influenced by initial conditions, including industrial composition and energy infrastructure. The within-group transformation characteristic of fixed effects models helps to mitigate potential selection bias arising from these pre-existing factors.

Second, in examining the direct causal mechanisms through which the low-carbon transition affects urban–rural income inequality, the fixed-effects model provides superior identification of structural economic effects. By incorporating city-level fixed effects, this approach effectively isolates the net impact of economic restructuring while controlling for time-invariant urban characteristics. Empirically, the fixed-effects model does not depend on the assumption of a spatial weight matrix. This approach can circumvent the misconfiguration of the spatial matrix that may arise from the “administrative region economy” phenomenon in China. Additionally, it eliminates the need to address common issues related to weak instrumental variables found in GMM models. Consequently, this model is considered more reliable in scenarios involving non-policy evaluations. Therefore, in order to investigate the impact of urban low-carbon transformation on the urban–rural income gap in China, this paper selects the fixed-effects model and introduces carbon emission efficiency and its quadratic term to test whether there is a nonlinear relationship between urban low-carbon transformation and the urban–rural income gap. The benchmark regression model is as follows:

$${urid}_{it}={\alpha }_0+{\alpha }_1{carbon}_{it}+{\alpha }_2{carbon}_{it}^2+{\alpha }_3{control\text{\_}var}_{it}+v_i+u_t+{\epsilon }_{it}$$

(1)

where urid is the urban–rural income gap, carbon is the urban low-carbon transition, control_var is a set of control variables, v is a city fixed effect, u is a year fixed effect, and ε is a random error term.

4.2. Variable Setting

Explained variable: income gap between urban and rural residents. In this study, it is expressed as the ratio of the disposable income of urban residents to the disposable income of rural residents. Several important limitations should be noted regarding the use of the urban–rural income ratio as a singular metric: First, this indicator solely captures mean value comparisons, consequently obscuring intra-regional income distribution disparities (e.g., urban high-income groups inflating averages whereas rural poverty becomes statistically diluted) and overlapping income ranges (where certain agricultural households may surpass urban low-income earners). Second, the metric fails to account for cost-of-living differentials between urban and rural areas, particularly implicit expenditures on housing, education, and healthcare, leading to systematic underestimation of real purchasing power disparities. Third, methodological inconsistencies in data collection introduce measurement bias, as urban income calculations incorporate hidden benefits like social security provisions, whereas rural datasets remain incomplete and inadequately address income attribution challenges for migrant workers.

Core explanatory variable: urban low-carbon transition. The low-carbon transition has a narrow and broad sense, the narrow aspect of the low-carbon transition refers to the economy from relying on high-carbon development to the low-carbon development transition, while the broad sense of the low-carbon transition refers to the economy from high- to low-pollution development transition; at this time, pollution not only refers to the carbon dioxide emissions, but also includes all kinds of pollutants in addition to carbon dioxide [45]. To comprehensively examine the urban low-carbon transition, this study measures the urban low-carbon transition with carbon emission efficiency, referring to the study by Guo et al. (2022) [46], which takes the real urban GDP (2006 = 100) as the desired output and the urban CO₂ emissions as the non-desired output. The carbon emission data were from the official website of CEADs, the China Carbon Accounting Database. Capital stock, labor input, and energy input are taken as input indicators, where capital stock is expressed as fixed capital stock calculated by the perpetual inventory method with 2006 as the base period, and the formula is $k_{i,t}=k_{i,t-1}(1-\delta )+I_{i,t}$, where k_i,t denotes the fixed capital stock of city i in year t, $\delta$ denotes the capital depreciation rate, the depreciation rate $\delta$ is calculated using 9.6% with reference to Zhang et al. (2004) [47], I_i,t denotes the real fixed asset investment of city i in year t (deflated using the fixed asset investment price index (2006 = 100)), and the fixed capital stock in the base period is calculated as $k_{i,2006}=I_{i,2006}/\rho$. With reference to Zhang et al. (2004) [47], ρ is calculated by using 10%; labor input is expressed by the number of employees per unit in the city in the calendar year; energy input is expressed by the total energy consumption, which specifically includes the total electricity consumption of the whole society, the total amount of artificial gas and natural gas supply, and the total amount of liquefied petroleum gas (LPG) supply. Here, we refer to the “General Principles for Calculation of Comprehensive Energy Consumption,” which are calculated according to the formulae of 0.1229 kg tec/(kW-h), 1.33 kg tec/m³, and 1.7143 kg tec/kg, respectively, to convert energy consumption to tons of standard coal. As the carbon emission efficiency measured using the non-expected output SBM model will appear in the case of multiple cities with a value of one to more accurately estimate the impact of the urban low-carbon transition, this study adopts the non-expected output super-efficiency SBM model to measure carbon emission efficiency. The non-expected output super-efficiency SBM model is commonly employed to assess environmental efficiency. This model can simultaneously address both desired and non-desired outputs, thereby considering economic benefits alongside negative environmental impacts. It overcomes the limitations of traditional DEA models that cannot distinguish between efficient decision-making units with an efficiency score of one, allowing for the precise measurement of efficiency values under environmental constraints. This model is particularly suitable for evaluating low-carbon transitions, green development, and resource–environmental efficiency. It provides policymakers with a scientific basis for making decisions that balance economic growth and environmental protection.

Control variables: Considering the influence of other factors on the income gap between urban and rural residents, and referring to the existing literature, this study selects the following control variables: gross domestic product, expressed as the logarithm of real GDP (2006 = 100); government behavior, expressed as the proportion of the general government budget expenditures to the GDP; foreign direct investment, expressed as the proportion of the amount of foreign investment actually used to the GDP; and transportation infrastructure, expressed as the proportion of the road passenger traffic as a share of total population.

The variable definitions are shown in

Table 1. Variable definitions.

Variable	Variable Symbol	Variable Name	Variable Definition
Dependent variable	urid	Urban–rural income gap	Ratio of disposable income of urban residents to disposable income of rural residents
Independent variable	carbon	Urban low-carbon transition	Carbon emission efficiency
	carbon2	Urban low-carbon transition secondary term	Carbon emission efficiency secondary term
	lngdp	Gross domestic product	Logarithm of real GDP (2006 = 100)
Control variables	gova	Government behavior	Ratio of general government budget expenditures to GDP
	fdi	Foreign direct investment	Ratio of foreign capital actually utilized to GDP
	transport	Transportation infrastructure	Ratio of road passenger traffic to total population

.

4.3. Sample Selection and Data Sources

This paper analyzes how carbon efficiency influences the urban–rural income gap, including its transmission mechanism and heterogeneity, and uses panel data from 240 Chinese prefectural cities (2006–2019); the data mainly come from China’s city statistical yearbooks in previous years, the official website of the National Bureau of Statistics, the wind database, and the EPS data platform, while the data on carbon emissions come from the official website of China Carbon Accounting Database (CEADs).

Linear interpolation methods for filling in missing data can effectively preserve the trends of time series, such as the smooth changes observed in economic indicators. This approach is particularly suitable for small-scale, randomly missing continuous variables, as it avoids sample size loss and maintains the balance of panel data. The calculations involved are straightforward and easy to implement. However, it is important to recognize that this method may obscure fluctuations present in real data, such as sudden change points resulting from policy shocks or economic crises. Consequently, this could artificially reduce variance and lead to inflated statistical significance. For some of the missing data, this study uses linear interpolation to fill in the blanks, and to eliminate the influence of outliers and extreme values, this study shrinks the data at the 1% level. The descriptive statistics of main variables are shown in

Table 2. Descriptive statistics.

VarName	Obs	Mean	SD	Min	Max
urid	3360	2.4694	0.5156	1.5787	4.2159
carbon	3360	0.5119	0.2323	0.1933	1.3202
carbon2	3360	0.3159	0.3251	0.0373	1.7428
lngdp	3360	16.1892	0.9423	14.2736	18.7512
gova	3360	0.1677	0.0761	0.0634	0.4697
fdi	3360	0.0195	0.0191	0.0000	0.0926
transport	3360	18.5957	20.5032	2.3676	142.8395

.

5. Empirical Results and Analysis

5.1. Benchmark Regression

Table 3. Baseline regression results.

VARIABLES	(1)	(2)	(3)
	Urid	Urid	Urid
carbon	0.0300	0.4995 **	0.6098 ***
	(0.5393)	(2.2842)	(3.0468)
carbon2		−0.3192 **	−0.4177 ***
		(−2.2085)	(−3.3229)
lngdp	−0.7524 ***		−0.7872 ***
	(−5.1989)		(−5.6158)
gova	−0.3973 **		−0.4349 **
	(−2.0899)		(−2.3355)
fdi	−1.3815 **		−1.3540 **
	(−2.0243)		(−2.0033)
transport	−0.0011 **		−0.0010 **
	(−2.1065)		(−2.0508)
Constant	14.4509 ***	2.5451 ***	14.8256 ***
	(6.4157)	(35.7818)	(6.8136)
id fe	yes	yes	yes
year fe	yes	yes	yes
Observations	3360	3360	3360
R-squared	0.5877	0.5599	0.5921
U-Test inflection point value		0.7824473	0.7298599
U-Test t-value		1.93	2.93
U-Test p-value		0.0274	0.00188

The symbols *** and ** denote significance at the 1% and 5% levels, respectively.

shows the results of the benchmark regression; column (1) is the linear effect of the urban low-carbon transition on the urban–rural income gap. The results show that the estimated coefficient of carbon is not significant, which indicates that the linear relationship between the urban low-carbon transition and urban–rural income gap does not hold, thus this study introduces carbon² into the benchmark model to test whether there is a non-linear relationship between the urban low-carbon transition and urban–rural income gap. Column (2) is the estimated results without adding control variables and column (3) is the estimated results after adding control variables. The results show that the estimated coefficient of carbon, the primary term of the urban low-carbon transition, is significantly positive, and the estimated coefficient of carbon², the secondary term, is significantly negative, which indicates that there is an “inverted U-shape” relationship between the urban low-carbon transition and urban–rural income disparity; the results of the Utest test show that the inflection point value is 0.7298599, that is to say, when the inflection point value is 0.7298599, i.e., when the carbon emission efficiency is lower than 0.7298599, the urban low-carbon transition will expand the urban–rural income gap; when the carbon emission efficiency crosses the inflection point of 0.7298599, the low-carbon transition will continue to promote the urban–rural income gap convergence, and the assumption expressed in H1 has been proven. Figure 1 illustrates an inverted U-shaped relationship between the urban low-carbon transformation and the income gap between urban and rural areas. Notably, the inflection point value is 0.72985986, which aligns with the benchmark regression results discussed earlier. First of all, in the initial stages of low-carbon technology adoption, there is a significant reliance on substantial investments and high technical thresholds. Urban areas, characterized by their well-developed infrastructure, capital-intensive nature, and concentration of talent, are more likely to attract green investments and thereby create advantages for industrial upgrading. In contrast, rural regions often face challenges such as outdated technologies and difficulties in securing financing, which hinder their participation in high value-added green industries.

Secondly, the contraction of traditional high-carbon industries may result in unemployment among the rural labor force. Meanwhile, emerging green industries located in urban centers typically demand higher skill levels from workers. This disparity can exacerbate inequalities in employment opportunities between urban and rural populations in the short term. Furthermore, early low-carbon policies frequently prioritized cities or industrial zones at the expense of rural areas. These factors contribute to an expanding income gap between urban and rural areas during the early phases of low-carbon transformation.

As the costs associated with low-carbon technologies decrease over time, the inherent natural resource advantages found in rural areas become increasingly evident. Income growth can be achieved through initiatives such as distributed energy systems and ecological agriculture. Moreover, the government plays a crucial role by addressing the urban–rural gap through fiscal transfer payments, investment in green infrastructure within rural settings, and carbon sink trading mechanisms.

In the long run, training programs focused on green skills along with enhanced mobility of resources between urban and rural areas will facilitate greater integration of rural populations into the green industrial chain. Therefore, this ongoing low-carbon transformation is expected to further narrow the income gap between urban and rural regions.

5.2. Robustness Tests

5.2.1. Replacement of Explanatory and Interpreted Variables

This study replaces the explanatory and interpreted variables, using the difference between the disposable income of urban residents and rural residents diff_urid to indicate the level of the urban–rural income gap, and using carbon emissions carbon_emi and its quadratic term carbon_emi² as explanatory variables.

Table 4. Robustness test results I.

VARIABLES	Diff_urid
carbon_emi	−0.3147 **
	(−2.4517)
carbon_emi2	0.2074 ***
	(3.2758)
Constant	−1.2631
	(−0.4049)
Control vars	yes
id fe	yes
year fe	yes
Observations	3360
R-squared	0.8808
U-Test inflection point value	0.7585915
U-Test t-value	2.43
U-Test p-value	0.00795

The symbols *** and ** denote significance at the 1% and 5% levels, respectively.

shows that the estimated coefficients of carbon_emi are significantly negative at the 5% level, and the estimated coefficients of carbon_emi² are significantly positive at the 1% level, which initially indicates the existence of a “positive U-shaped” relationship between carbon emissions and urban–rural income difference; the results of the Utest test further verify the “positive U-shaped” relationship between the two. It can be observed from the results presented in Table 4 that the inflection point value of this positive U-shaped relationship is 0.7585915. This suggests that, when carbon emissions are below 0.7585915, the income gap will continuously decrease as carbon emissions increase. However, once this inflection point is surpassed, further increases in carbon emissions will result in a continuous widening of the income gap between urban and rural regions. From a horizontal perspective regarding the income gap, it indicates that, while an increase in carbon emissions may lead to a temporary contraction of this income gap in the short term, it ultimately contributes to a greater long-term disparity. Carbon emissions serve as a direct quantitative indicator for low-carbon transformation, thereby elucidating the necessity for such transformations from another angle and affirming the nonlinear relationship between the low-carbon transformation and the urban–rural income gap. To some extent, this finding reinforces the robustness of our benchmark regression results.

5.2.2. Adding the Control Variable

Liao et al. (2024) affirmed the important role of new urbanization in improving unbalanced development [48]. To circumvent the impact of the omitted variable problem, this study introduces the urbanization rate urban as a control variable into the benchmark regression model, which is shown by the results in Column (1) of

Table 5. Robustness test results II.

VARIABLES	(1)	(2)	(3)	(4)	(5)
	Adding the Control Variable	Reduction of Sample Period	2SLS with the First-Order Lag Term as the Instrumental Variable	2SLS with the Second-Order Lag Term as the Instrumental Variable	Excluding the Central City Sample
carbon	0.6114 ***	0.4418 **	0.834 ***	1.192 ***	0.7093 ***
	(3.0964)	(1.9836)	(0.170)	(0.308)	(3.4170)
carbon2	−0.4187 ***	−0.2976 *	−0.549 ***	−0.796 ***	−0.5066 ***
	(−3.3605)	(−1.9534)	(0.122)	(0.240)	(−3.9728)
Constant	14.8580 ***	12.3307 ***			12.0143 ***
	(6.5880)	(4.5878)			(4.1765)
Control vars	yes	yes	yes	yes	yes
id fe	yes	yes	yes	yes	yes
year fe	yes	yes	yes	yes	yes
Observations	3360	2640	3120	2880	3108
R-squared	0.5921	0.5338	0.605	0.6	0.6126
Stock-Yogo 10%			7.03	7.03
Cragg-Donald Wald F statistic			833.475	119.928
U-Test inflection point value	0.7302339	0.7423745	0.7597419	0.748943	0.7000589
U-Test t-value	2.98	1.76	3.8	2.67	3.22
U-Test p-value	0.0016	0.0395	0.0000727	0.00379	0.000748

The symbols ***, **, and * denote significance at the 1%, 5%, and 10% levels, respectively.

. The estimated coefficient of carbon is significantly positive at the 1% level, and the estimated coefficient of carbon² is significantly negative at the 1% level, which suggests that there is a possible “inverted U-shape” between the urban low-carbon transition and common prosperity. For the “inverted U-shaped” relationship, Utest test further verifies the non-linear relationship between the two. Upon incorporating urbanization indicators as control variables, we find that the inflection point value of the inverted U-shaped curve relating low-carbon transformation to urban–rural income disparity shifts to 0.7302339—an increase compared with its counterpart in our benchmark regression analysis. This change implies that after accounting for urbanization factors, this curve’s inflection point occurs at a later stage. Furthermore, it highlights from another perspective that advancements in urbanization levels can expedite the transition of income disparities between cities from widening to narrowing due to low-carbon transformation efforts.

5.2.3. Reduction of Sample Period

The rural revitalization strategy is the top priority for solving the three rural issues, considering the positive impact of the rural revitalization strategy on the income level of rural residents; this study shortens the sample period to 2006–2016 and re-estimates the baseline model, which is shown by the results in Column (2) of Table 5. The estimated coefficients of carbon are significantly positive at the 5% level, and the estimated coefficient of carbon² is significantly negative at the 5% level, indicating that there is an “inverted U-shaped” relationship between the urban low-carbon transition and the urban–rural income gap, which is verified by the results of the Utest test. It can be observed from the results that the inflection point value of this inverted U-shaped curve is 0.7423745, which exceeds the inflection point value obtained from the benchmark regression analysis. This finding suggests, from an alternative perspective, that the rural revitalization strategy has expedited the transition of the urban–rural income gap’s low-carbon transformation from a positive impact to a negative one, thereby demonstrating the effectiveness of this policy.

5.2.4. Dealing with Endogeneity

To avoid the impact of endogeneity problems on the model, this study takes the first- and second-order lag of urban low-carbon transition and its quadratic term as instrumental variables and conducts two-stage least squares estimation (2SLS) for the relationship between the urban low-carbon transition and urban–rural income gap. Here, the weak instrumental variables test is first conducted to verify the validity of the instrumental variables. Table 5 column (3) shows the estimation results of 2SLS with the first-order lag term as the instrumental variable, and Column (4) shows the estimation results of 2SLS with the second-order lag term as the instrumental variable. By comparing the size of the critical values of the Cragg–Donald Wald F-statistic and the Stock–Yogo weak ID test at the 10% level in columns (3) and (4) of Table 5, it can be seen that the former is much larger than the latter; this difference indicates that the instrumental variables selected in this study are valid instrumental variables with a strong correlation with endogenous variables. The results of both columns show that the estimated coefficients of carbon are significantly positive at the 1% level, and the estimated coefficients of carbon² are significantly negative at that level, which indicates that, after addressing the endogenous problems, the “inverted U-shaped” relationship between the urban low-carbon transition and urban–rural income gap is still valid. This study further applies the Utest to test the relationship between the urban low-carbon transition and the urban–rural income gap; the test results again verify the “inverted U-shaped” relationship between the two.

5.2.5. Excluding the Central City Sample

To test the robustness of the results, this study excludes the central cities (municipalities and sub-provincial cities) from the sample, and regresses the benchmark model again. The results in column (5) of Table 5 show that there is still an “inverted U-shaped” relationship between the low-carbon transition and common prosperity. In contrast, the inflection point value in this analysis is 0.7000589, which is lower than that found in the benchmark regression result. This indicates that, after excluding samples from central cities, there has been an acceleration in reaching the inflection point of the inverted U-shaped curve concerning the low-carbon transformation and income disparity between urban and rural areas. Consequently, it appears to have advanced to a stage where common prosperity is promoted through low-carbon transformation earlier than anticipated. The underlying reason for this phenomenon may lie in the relatively developed economic conditions, resource availability, and market dynamics within central cities. As such, during early stages of transformation, initiatives related to low-carbon policies implementation, dissemination of low-carbon technologies, and attraction of highly skilled talent tend to favor these central urban areas more significantly. This trend could further exacerbate income disparities between urban and rural regions; thus, at a national level perspective, this inadvertently hampers progress toward achieving common prosperity via low-carbon transformation.

5.2.6. Establish the Evaluation Index System for Low-Carbon Transformation

This paper identifies energy structure, industrial structure, green technology level, green attention, and carbon emissions as a comprehensive index system for measuring low-carbon transformation. The entropy method has been employed to construct this index system for low-carbon transformation, and finally calculate the level of low-carbon transformation. The composition is detailed in

Table 6. Evaluation index system for low-carbon transformation.

Dimension	Indicator Selection	Indicator Description	Indicator Polarity
Foundation	Energy structure	Proportion of coal in total energy consumption	Negative (−)
Carrier	Industrial structure	Percentage of secondary industry in GDP	Negative (−)
Driving force Social pressure	Green technology level Green attention	Share of green patents in regional annual patent applications Measured through word frequency analysis in government work reports	Positive (+) Positive (+)
Ultimate state	Carbon emissions	Data sourced from China Carbon Accounting Database (CEAD)	Negative (−)

. These five dimensions collectively form a complete logical chain characterized by “foundation–carrier–driving force–pressure–state”: the energy structure (the foundation of transformation) directly influences carbon emission levels; the industrial structure (the carrier of transformation) reflects the extent of decarbonization within the economic system; green technology (the driving force of transformation) provides essential technical support; green attention (social pressure) encourages shifts toward low-carbon behaviors; and carbon emissions (the ultimate state) quantify the effectiveness of these transformations. This framework encompasses both supply-side factors (energy and industry) and demand-side elements (social participation), integrating hard power (technology) with soft environmental aspects (awareness), thereby enabling a systematic assessment of low-carbon transition efforts.

After establishing an evaluation index system for low-carbon transformation and adding the calculated level of low-carbon transformation to the benchmark regression model for re-estimation, new estimation results were obtained, as shown in

Table 7. Robustness test results III.

VARIABLES	Urid
Comcarbon	0.6787 **
	(2.0325)
Comcarbon2	−0.8448 **
	(−2.0564)
Constant	14.4318 ***
	(6.3905)
Control vars	yes
id fe	yes
year fe	yes
Observations	3360
R-squared	0.5891
U-Test inflection point value	0.4017026
U-Test t-value	1.86
U-Test p-value	0.0318

The symbols *** and ** denote significance at the 1% and 5% levels, respectively.

. The first term of the core explanatory variable comcarbon is significantly positive, and the second term is significantly negative. Moreover, the estimation results have passed the Utest. This indicates that the “inverted U-shaped” relationship between the low-carbon transformation and the urban–rural income gap still holds true, verifying the robustness of the benchmark regression results.

5.3. Mechanism Analysis

This study starts from the industrial and employment structures and human capital to study the mechanism of their role in the process of urban low-carbon transformation affecting the urban–rural income gap. This study specifically analyzes the mechanism of industrial structure from the two dimensions of industrial structure advanced and industrial structure rationalization. Referring to the research of Xu (2008) [49], we calculate the advanced industrial structure and assign values to the proportion of one, two, and three industries, respectively. The formula is ${industry}_{it}={\sum }_{n=1}^3{indu\text{\_}gdp}_{i,t,n}\times n, n=1,2,3,$ where indu_gdp represents the proportion of one, two, and three industries in GDP. Referring to the study of Gan et al. (2011) [50], the rationalization of industrial structure is measured by using Theil’s index, which is calculated as ${theil}_{it}={\sum }_{n=1}^3{indu\text{\_}gdp}_{i,t,n}ln({indu\text{\_}gdp}_{i,t,n}/{labor}_{i,t,n}), n=1,2,3,$ where labor is the share of employees in primary, secondary, and tertiary industries in the total employed population. The employment structure is expressed by the number of employees in one, two, or three industries as a proportion of the total employed population respectively, and the level of human capital is expressed by the number of students enrolled in general higher education schools as a proportion of the total population.

Table 8. Mechanism test results.

VARIABLES	(1)	(2)	(3)	(4)	(5)	(6)
	Advancement of Industrial Structure	Rationalization of Industrial Structure	The Proportion of Employed People in Primary Industry	The Proportion of Employed People in Secondary Industry	The Proportion of Employed People in Tertiary Industry	Human Capital
carbon	−0.0749 *	0.1875 **	0.0030	−0.1324 ***	0.1320 ***	0.0058 **
	(−1.9695)	(2.2742)	(0.6135)	(−6.9889)	(7.0927)	(2.3542)
carbon2	0.0557 **	−0.1153 **
	(2.0705)	(−2.2605)
Constant	1.7096 ***	1.2278	0.4383 ***	−1.2668 ***	1.7754 ***	0.0511
	(3.2224)	(1.6443)	(2.9391)	(−2.6104)	(3.7145)	(0.3876)
Control vars	yes	yes	yes	yes	yes	yes
id fe	yes	yes	yes	yes	yes	yes
year fe	yes	yes	yes	yes	yes	yes
Observations	3360	3360	3360	3360	3360	3360
R-squared	0.7317	0.0452	0.1569	0.2966	0.3033	0.1754
U-Test inflection point value	0.6730085	0.8129219
U-Test t-value	1.88	2.00
U-Test p-value	0.031	0.0235

The symbols ***, **, and * denote significance at the 1%, 5%, and 10% levels, respectively.

(1) shows the estimation results of the relationship between urban low-carbon transformation and industrial structure advanced, the estimated coefficient of carbon is significantly negative, and the estimated coefficient of carbon² is significantly positive, so it can be initially judged that there exists a “U-shaped” relationship between the two, and the Utest further verifies this conclusion. Column (2) shows the estimation results of the relationship between urban low-carbon transformation and Theil’s index (i.e., rationalization of industrial structure). The smaller Theil’s index is, the more rational the industrial structure is; here, the estimated coefficient of carbon is significantly positive, the estimated coefficient of carbon² is significantly negative, and it can be preliminarily judged that there is an “inverted U-shape” relationship between them, which the Utest further verifies. Columns (3), (4), and (5) are the estimation results of the relationship between the urban low-carbon transformation and the proportion of employed people in primary, secondary, and tertiary industries, among which the urban low-carbon transformation has no significant impact on primary industry employment, but has a significant negative impact on secondary industry employment, and has a significant positive impact on tertiary industry employment; these results suggest that China’s low-carbon transformation has generally prompted the transfer of labor from the secondary to the tertiary industry. Column (6) shows the estimation results of the relationship between the urban low-carbon transformation and human capital, and the results show that the urban low-carbon transformation has a significant positive impact on the level of human capital. The low-carbon transformation of cities serves a dual purpose. On one hand, it fosters the development of high-tech and green industries, which typically necessitate a workforce equipped with advanced skills. This demand consequently elevates the overall level of human capital within these urban environments. On the other hand, the decline in high-carbon industries resulting from this transformation has led to increased unemployment among low-skilled workers. In response to this challenge, state-sponsored re-employment skills training for the unemployed aims to enhance their employability and simultaneously contributes to an improvement in human capital levels. In summary, urban low-carbon transformation has a significant impact on China’s industrial structure, employment structure, and human capital level, and the previous three mechanisms have been analyzed from the theoretical perspective of the urban–rural income gap mechanism in China: therefore, hypothesis H2 has been proven.

5.4. Tests for Heterogeneity

5.4.1. Regional Heterogeneity

The research sample was divided into two groups of samples according to China’s coastal and inland areas for group regression, and the results in columns (1) and (2) of

Table 9. Heterogeneity test results.

VARIABLES	(1)	(2)	(3)	(4)	(5)	(6)	(7)	(8)
	Coastal Areas	Inland Areas	Resource Cities	Non-Resource Cities	Industrialization Stage	Service Stage	Larger Urban–Rural Income Gap	Smaller Urban–Rural Income Gap
carbon	0.1607	0.9211 ***	0.6716 **	0.5482 **	0.5061 **	0.8001 **	0.9220 ***	0.2901 *
	(0.5254)	(3.7526)	(2.1815)	(2.1636)	(2.2472)	(2.3594)	(2.9244)	(1.8378)
carbon2	−0.0590	−0.6620 ***	−0.5392 ***	−0.3491 **	−0.3606 ***	−0.5035 **	−0.5911 ***	−0.1825 *
	(−0.3177)	(−4.3710)	(−3.0541)	(−2.1619)	(−2.6545)	(−2.3686)	(−3.1317)	(−1.7004)
Constant	11.1398 ***	14.2149 ***	14.3382 ***	15.0080 ***	9.5276 ***	18.3061 ***	16.8370 ***	8.0424 ***
	(4.3690)	(4.8233)	(3.4695)	(5.8887)	(2.9695)	(5.5057)	(5.3967)	(4.2849)
Control vars	yes	yes	yes	yes	yes	yes	yes	yes
id fe	yes	yes	yes	yes	yes	yes	yes	yes
year fe	yes	yes	yes	yes	yes	yes	yes	yes
Observations	1246	2114	1274	2086	2401	959	1680	1680
R-squared	0.5200	0.6345	0.5644	0.6142	0.5851	0.5822	0.6088	0.4476
U-Test inflection point value		0.6957244	0.6227813	0.7852671	0.7016496	0.7945524	0.7798417	0.7948833
U-Test t-value		3.53	1.93	1.98	2.11	2.05	2.84	1.36
U-Test p-value		0.000273	0.0286	0.0248	0.018	0.0207	0.00253	0.0871

The symbols ***, **, and * denote significance at the 1%, 5%, and 10% levels, respectively.

show that an inverted U-curve correlation exists between China’s inland low-carbon transition and the urban–rural income gap, whereas there is no such relationship in China’s coastal areas. Its extreme value is 1.360519, which is located on the right side of the research interval, indicating that the urban low-carbon transition and urban–rural income gap in coastal areas have entered a negative correlation stage; the low-carbon transition promotes the urban–rural income gap to continue to converge. This may be because of the relative advantages of China’s coastal areas in economic development, industrial structure, social security, and resource concentration, and the soundness of the income distribution mechanism, while in these same aspects of the inland areas relative to the coastal areas, the gap still exists, and along with the promotion of the low-carbon transition, some of the high-energy-consuming and high-polluting industries are gradually being transferred inland. As the industrial takeover area, the inland cities are facing more severe economic transformation problems, so the regression results have regional heterogeneity.

5.4.2. Resource Endowment Heterogeneity

According to the list of resource cities in the National Sustainable Development Plan for Resource Cities (2013–2020), this study divides the research sample into two groups: resource and non-resource cities. The results in Columns (3) and (4) of Table 9 show that, whether it is a resource or a non-resource city, the urban low-carbon transition and the urban–rural income gap show an “inverted U-shape” between the urban low-carbon transition and common prosperity. The inflection point value of resource cities is 0.6227813, which is smaller than the benchmark regression inflection point value of 0.7298599, while the inflection point value of non-resource cities is 0.7852671, which indicates that, in comparison, the low-carbon transformation of resource cities ushered in the role of the inflection point earlier, and entered the stage of narrowing the income gap between urban and rural areas earlier. The main reason is that resource cities have gathered a large number of industrial enterprises, are the main supply areas of energy resources, have more prominent resource and environmental problems, and have a single direction of labor force employment, so the low-carbon transition has a greater impact on the economic development of resource cities.

5.4.3. Heterogeneity of Economic Development Stages

This paper refers to the study of Yuan et al. (2022) [51], considers the impact of different stages of economic growth, takes the ratio of the tertiary to the secondary sector as the basis for the division of the stage of economic growth, and divides the samples with a ratio equal to or less than one into the industrialization stage, and the samples with a ratio greater than one into the service stage. From the results of columns (5) and (6) of Table 9, it can be seen that, whether it is the industrialization or the service stage, the relationship between the urban low-carbon transformation and urban–rural income gap is an “inverted U”, but the inflection point value of the industrialization stage is 0.7016496, which is significantly smaller than the benchmark regression inflection point value of 0.7298599. While entering the service stage, the inflection point of urban low-carbon transformation affecting the urban–rural income gap arrives later, with an inflection point value of 0.7945524, which demonstrates that, in the industrialization stage, urban low-carbon transformation will promote the decline of the urban–rural income gap earlier. The main reason for this is the more homogeneous structure of industry and employment during the industrialization phase, an imperfect income distribution system, and high energy consumption and high-pollution characteristics, which make the urban–rural income gap more sensitive to the impacts of the urban low-carbon transition.

5.4.4. Heterogeneity of Income Gap

This paper calculates the median in accordance with the ratio of urban–rural income gap in each city, and divides the samples into two groups for group regression with the median as the cut-off point. The results in columns (7) and (8) of Table 9 show that there is an inverted U-curve correlation existing between the low-carbon transition and the urban–rural income gap in two groups of samples, but the inflection value of the group with a larger urban–rural income gap is 0.7798417, which is smaller than that of the group with a smaller urban–rural income gap, indicating that cities with a larger income gap are more sensitive to the impact of low-carbon transition. This also validates the effectiveness of the low-carbon transformation process in achieving the goal of common prosperity, particularly for cities characterized by a significant urban–rural gap. Consequently, through well-designed policy measures, the potential of low-carbon transformation to contribute to poverty alleviation can be actively harnessed.

6. Conclusions and Recommendations

This paper analyzes how carbon efficiency influences the urban–rural income gap, including its transmission mechanism and heterogeneity, and uses panel data from 240 Chinese prefectural cities (2006–2019). We found an inverted U-curve correlation existing between the low-carbon transition and the urban–rural income gap. The low-carbon transition influences the urban–rural income gap through industrial structure advancement, industrial structure rationalization, employment structure, and human capital. After the heterogeneity test, it was found that the “inverted U-shaped” relationship between the urban low-carbon transition and urban–rural income gap is characterized by regional heterogeneity: in resource endowment, in the economic development stage, and in the level of the urban–rural income gap. On the basis of these findings, this study makes the following policy recommendations:

(1) We must fully recognize the critical role of manufacturing in the development of the real economy. In the context of low-carbon economic development, it is essential not only to promote the growth of the service sector, but also to emphasize the transformation and upgrading of the manufacturing industry. We should support the gradual transformation and upgrading of manufacturing by formulating relevant policies and contingency plans to effectively manage associated risks, ensuring both economic growth and stable employment. Efforts should be made to minimize the unemployment risk for low-skilled labor resulting from the transformation and upgrading of the manufacturing sector. Existing research indicates that deindustrialization in developing countries often represents premature deindustrialization, characterized by the decline or absence of mid-to-high-end manufacturing and the loss of low-to-mid-end manufacturing, leading to a trap of mid-to-high-end manufacturing stagnation. Therefore, maintaining an appropriate proportion of manufacturing requires increasing support for mid-to-high-end manufacturing sectors, particularly in integrated circuits, automotive manufacturing, display panels, and internet industries. By fostering collaboration among industry, academia, and research, we can enhance the technological sophistication and vitality of mid-to-high-end manufacturing, avoiding the trap of stagnation and preserving a balanced domestic manufacturing structure. This approach also mitigates the risk of labor unemployment caused by premature and rapid industrialization. Additionally, accelerating financial reforms to provide adequate and accessible financial support for the manufacturing industry, while continuously improving the business environment through initiatives such as innovation fund support, encourages manufacturing enterprises to innovate and develop, thereby reducing workforce losses in the manufacturing sector. Finally, it is crucial to prioritize the coordinated development of the service and manufacturing industries. The two sectors are interdependent, rather than isolated. Encouraging the growth of complementary service industries, such as industrial design, business consulting, and inspection and testing, not only provides supportive services for the manufacturing industry, but also creates employment opportunities for those displaced by manufacturing upgrades.

(2) Strengthening the skills training of the rural labor force and enhancing their competencies through increased investment in human capital is essential for better alignment with job market demands. It is imperative to establish academic programs that reflect the needs of modern agriculture and related industries. Vocational colleges should be encouraged to expand enrollment, lower age limits for admission, and implement preferential policies regarding tuition reduction and exemptions. In particular, in response to the challenge posed by industrial robots “squeezing out” low-skilled labor amid advancements in industrial intelligence, vocational institutions must focus on training individuals in operating industrial robots to continuously enhance the professional skills of the workforce. Concurrently, targeted on-the-job skills training should be reinforced to improve compatibility between humans and machines while effectively mitigating unemployment risks.

The inadequate infrastructure prevalent in rural areas, coupled with traditional mindsets, contributes significantly to a low willingness among rural migrant workers to return home following unemployment. This phenomenon represents a critical factor behind severe brain drain and stagnation within these regions, further exacerbating disparities in development and income between urban and rural settings. Therefore, accelerating improvements in the rural living environment emerges as a pivotal strategy for fostering development. Additionally, formulating favorable employment policies aimed at encouraging those displaced by manufacturing industry upgrades to seek opportunities back home can play a vital role in narrowing the income gap between urban and rural populations.

(3) Based on the heterogeneous characteristics of the impact of urban low-carbon transformation on the income gap between urban and rural areas, this paper proposes that a low-carbon economy should be developed in accordance with local conditions, tailored to both the coastal and inland regions of China. Coastal areas, characterized by high-density industrial clusters, relatively complete green financial systems, extensive application of renewable energy such as offshore wind power and photovoltaic power, as well as advanced digital technologies, often exhibit these features predominantly in urban settings. However, rural areas still lag behind in terms of development conditions. Research findings indicate that the low-carbon transformation of cities and the urban–rural income gap have entered a negative correlation stage, where promoting urban low-carbon transformation can continuously narrow the income gap between urban and rural areas. For coastal areas, it is recommended to fully leverage their advantages in talent, technology, and geographical location to sustain the momentum of low-carbon economic development. This can be achieved through the establishment of a carbon compensation fund, “cities giving back to rural areas,” by allocating part of the revenue from urban carbon trading to rural photovoltaic poverty alleviation and biomass energy development projects. Additionally, the active promotion of transferring low-carbon industrial chains from urban enterprises to surrounding rural areas is encouraged. Inland areas, which often rely on energy-intensive industries for development and face a high rural energy poverty rate, experience significant transformation pressure and a pronounced urban–rural income gap. To address this, a “tiered carbon pricing” system can be implemented for high energy-consuming cities, applying a low tax rate within the emission threshold and using the revenue exceeding the threshold for rural biogas projects. Furthermore, exemptions in carbon taxes for domestic energy use in rural areas are suggested. Coastal cities can also provide technology transfer to inland areas, while inland areas offer green electricity quotas to enhance cross-regional collaboration.

(4) As resource-based cities, industrialized cities, and regions with a significant urban–rural income gap have entered the phase of low-carbon transformation earlier to narrow the urban–rural income disparity, they should capitalize on the economic transformation period and expedite the development of a low-carbon economy. However, given their heavy reliance on high-carbon industries and the dependence of rural labor forces on traditional energy supply chains for employment, these cities face concentrated unemployment risks during the transition. To mitigate this, exit subsidies for high-carbon industries can be implemented in a phased manner based on enterprises’ carbon emissions levels. By fostering integrated industries such as “mine restoration + renewable energy” and establishing regional green metallurgy/chemical clusters, additional job opportunities for rural laborers can be created. Furthermore, specialized training subsidies for rural workers can enhance the level of rural human capital.

For non-resource-based cities, service-oriented cities, and regions with a smaller urban–rural income gap, the development of low-carbon industries may trigger a siphon effect. This could attract young and middle-aged rural laborers from surrounding areas, leading to a shortage of rural construction personnel, thereby adversely affecting agricultural development. To address this, urban universities and enterprises can collaborate to support rural areas through targeted assistance programs and formulate rural employment subsidy policies to provide talent and bolster rural development.

(5) To establish a comprehensive monitoring indicator system for the low-carbon transition in economic and social contexts, it is recommended to develop a multidimensional and dynamic monitoring framework, which should encompass core indicators such as carbon emission intensity, the proportion of renewable energy, added value from green industries, and changes in the urban–rural income gap. Furthermore, differentiated monitoring standards should be established for coastal cities versus inland cities, resource-based versus non-resource-based cities, and industrialized versus service-oriented cities.

7. Limitations of This Paper

This paper may have the following limitations: firstly, due to the differences in the caliber of income statistics between urban and rural areas in China (e.g., the problem of attributing the income of rural migrant workers and the difficulty of capturing hidden income), the use of the urban–rural income ratio as an indicator of common prosperity may lead to the underestimation or overestimation of the income gap; secondly, as this study relies on the efficiency of carbon emissions as an indicator of the low-carbon transition, this study focuses on the macro level and ignores the low-carbon transition’s micro impact. Finally, the data span of this study is relatively short because the low-carbon transition is a long-term process.