Impacts of Low-Carbon City Pilot Policy on Urban Land Green Use Efficiency: Evidence from 283 Cities in China

Abstract

On the global scale, the low-carbon city pilot policy (LCCPP) has important significance for and influence on the study of urban land green use efficiency (ULGUE). Based on the panel data of 283 cities in China from 2007 to 2019, this study uses the super-SBM model, multi-period DID model, spatial econometric model, intermediary effect model, and heterogeneity analysis methods to deeply explore the specific impact mechanism of LCCPP on ULGUE. The results show the following: (1) During the study period, the average ULGUE of the selected samples increased by 11.71 percentage points overall and showed a certain spatial agglomeration effect. (2) LCCPP has a significant promoting effect on the improvement of ULGUE, and there is a positive spatial spillover effect. (3) The impact of LCCPP on ULGUE is mainly achieved through two paths: reducing energy utilization intensity and improving urban innovation level. (4) In cities with different levels of land green use efficiency, geographical location, and resource endowment, there are significant differences in policy effects. This paper puts forward countermeasures and suggestions to comprehensively promote the sustainable development of global cities and the improvement of land green use efficiency.

1. Introduction

With the acceleration of the industrialization process, the greenhouse gases emitted by human activities continue to rise, leading to the rise in the earth’s temperature and the intensification of climate change. On a global scale, climate change is already causing melting glaciers, rising sea levels, loss of biodiversity, and more. And from mountain peaks to ocean depths, climate change continues, extreme weather and its devastating effects persist, and the environmental and socio-economic costs are rising [1]. Climate impacts on people and ecosystems are far greater than expected, and risks will escalate rapidly as the climate warms, requiring urgent action to minimize and avoid loss and damage [2]. Carbon dioxide, as the first greenhouse gas emission, has an important impact on global climate change. China is the country with the largest total carbon emission in the world and, also, the country with the largest carbon emission reduction efforts [3]. In 2022, China’s CO₂ emissions reached 11.47 billion tons, accounting for about 31% of the global total, and decreased by 23 million tons compared with 2021 [4]. Therefore, China’s carbon emission reduction work has attracted the attention of all countries in the world and has important reference value for developing countries that are still in the stage of industrialization.

In order to effectively control greenhouse gas emissions and promote the green and low-carbon transformation of the economy and society, China has implemented a range of policy measures [5]. These measures include the low-carbon city pilot policy, carbon emission rights trading pilot policy, and green finance policy, etc. Among these, the low-carbon city pilot policy has been implemented earliest and has the widest coverage [6], making it highly valued by the pilot governments.

Urban land is a crucial strategic resource for the development of a country or region. The efficient utilization of urban land resources not only fosters economic growth but also yields significant positive impacts [7]. Moreover, land use serves as a pivotal factor in depicting the spatial distribution of carbon emissions and carbon sinks, acting as a spatial conduit for carbon emissions originating from terrestrial ecosystems and human activities [8]. A large number of studies have shown that the development and utilization of urban land is the main source of carbon emissions [9,10,11]. Traditional land use methods mainly emphasize the advantages of speed and scale, while ignoring the benefits of green and low-carbon development. Therefore, integrating the concept of green and low-carbon development into the urban land use process will help to grasp urban land use efficiency more comprehensively [12]. It can be seen that urban land green use efficiency (ULGUE) refers to maximizing the economic, social, and ecological benefits of land through scientific planning, effective management, and rational use of land resources in the process of urbanization. However, while gradually changing the natural ecosystem and affecting the environmental sustainability [13], urbanization has also caused great harm to human health [14]. Based on this, how to improve ULGUE to achieve the coordination and unity of economy and environment and the sustainable development of human society is an issue worthy of close attention.

Low-carbon city pilot policy (LCCPP) refers to the policy measures implemented by the government in specific cities to reduce carbon emissions and promote sustainable development. The specific cities selected by the government are known as low-carbon pilot cities. In the global context, this policy has become one of the main ways for countries to promote sustainable urban development [15,16]. The policy includes but is not limited to reducing carbon emissions, improving energy efficiency, optimizing transportation systems, promoting renewable energy, etc. Through the implementation of these measures, cities can effectively reduce carbon emissions, improve resource utilization efficiency, and provide a more livable environment for residents [17,18]. At present, research on LCCPP in China mainly focuses on urban green efficiency [6], green technology innovation level [19], industrial structure [20], carbon emission reduction effect [21], pollution emission reduction efficiency [22], etc. Previous studies have shown a close relationship between the implementation of LCCPP and ULGUE. The policy promotes urban development in a low-carbon, sustainable, and efficient direction by optimizing land use structure, changing land development methods, and improving land use efficiency, which leads to the coordinated development of the economy, society, and environment [23]. However, the existing literature lacks a specific exploration of the impact of LCCPP on ULGUE, and the results from the few similar studies conducted are inconsistent [24,25]. The reason for this inconsistency may be due to different scholars choosing different indicators and methods to measure urban land green use efficiency. Therefore, this study aims to further optimize the measurement method of ULGUE and build a reasonable theoretical analysis framework, in order to expand and improve the existing research content on LCCPP and ULGUE. Notably, China holds the distinction of being the world’s largest developing country with the third largest land area globally. Consequently, conducting research on the impact of LCCPP on ULGUE in China carries significant reference value for other developing nations and even countries worldwide, offering insights into effective strategies for reducing carbon emissions and promoting environment-friendly land use practices.

The marginal contribution of this research mainly lies in two aspects: research content and research method. Specifically, in terms of research content, this study puts LCCPP and ULGUE in the same theoretical framework, explores the spatial spillover effect of LCCPP based on spatial perspective, and tests the influence path of the policy on ULGUE from the perspectives of energy utilization intensity and urban innovation level. In addition, LCCPP is a macro-policy at the central level, and its impact on ULGUE in prefecture-level cities cannot be generalized. In this regard, this study adopts quantile regression, sub-region, and sub-city type to conduct empirical analysis, so as to explore the effect of implementing LCCPP in different cities in a more detailed manner. Therefore, this study contributes to the existing research on LCCPP and ULGUE. In terms of research methodology, this study utilizes the Super-Efficiency Slack-Based Measure model with undesirable outputs to gauge ULGUE. However, it is worth noting that previous studies primarily focus on indicators such as industrial soot discharge, industrial sulfur dioxide emissions, and industrial wastewater [26,27]. Recognizing that LCCPP has a direct impact on carbon emission reduction, this study incorporates not only industrial soot, industrial sulfur dioxide, and industrial wastewater emissions but also carbon emissions as undesirable outputs. This comprehensive approach allows for a more precise exploration of the collaborative emission reduction effect of LCCPP in influencing ULGUE.

Next, the rest of the research is arranged as follows. Section 2 analyzes the mechanism of LCCPP on ULGUE and puts forward the research hypotheses. Section 3 describes the research area, measurement of ULGUE, model building, variable setting, and data resource. Section 4 reports the empirical results. Section 5 summarizes the research conclusions and provides policy implications.

2. Theoretical Analysis and Research Hypotheses

2.1. Direct Impact of LCCPP on ULGUE

The promotion effect of LCCPP on ULGUE is reflected in three dimensions: economy, society, and ecology. From an economic perspective, LCCPP can facilitate the transformation and upgrading of industrial structures in pilot cities [20], thereby not only extending the industrial value chain and enhancing product value-added but also driving out low-end industries through competition [28]. Consequently, this contributes to the intensive land use and improves the economic output efficiency of land [29]. From a societal perspective, although LCCPP is a result of government behavior and decision-making, its effective implementation requires full cooperation among enterprises, society, and the government [30,31]. Research has demonstrated that this policy has significant positive effects on promoting social environmental protection concepts, optimizing labor structure, rationalizing enterprise fund allocation, and enhancing regional innovation capacity [25,32]. These positive effects further incentivize the government to innovate policies aimed at improving green land use efficiency. From an ecological standpoint, while the original intention of LCCPP is primarily to reduce greenhouse gas emissions—which it has indeed achieved by significantly improving carbon emission efficiency [33]—there has also been continuous improvement in energy utilization efficiency due to a decrease in energy-consuming industries’ proportion and advancements in green technology innovation within enterprises. Consequently, this generates a “joint governance” effect for air pollution and carbon emission reduction [22,34], ultimately leading to improved ULGUE. Based on the above analysis, this paper proposes research hypothesis 1.

Hypothesis 1.

The implementation of LCCPP contributes to enhancing the efficiency of land green use in pilot cities.

2.2. The Spatial Spillover Effect of LCCPP

The free flow of factors constitutes the fundamental competition mechanism among local governments. In order to enhance their own advantages, local governments engage in competition centered around attracting liquidity factors, which significantly influences policy formulation and environmental governance by the government [35]. Within the framework of tournament theory [36], superior officials establish performance appraisal mechanisms, motivating lower-level officials to drive economic development for political advancement. This tendency towards an “economic man” [37,38] mindset among local officials leads to strategic games and imitation behaviors in policy formulation and implementation [39], including those related to LCCPP. However, in China’s current context, increasing attention is being paid to the effectiveness of environmental governance within the promotion and evaluation mechanisms for government officials. Environmental performance evaluations have been explicitly incorporated into key indicators for assessing local government performance [25]. This indicates a shift away from previous “race-to-the-bottom” behavior that prioritized economic development at the expense of the environment [40]. As an environmental regulation measure, LCCPP exhibits spatial spillover effects based on geographical distance and economic proximity. Additionally, occasional instances of air pollution spillover occur [41]. Under the influence of LCCPP, if pollution control measures are significantly improved within these pilot cities themselves, it will also mitigate spatial concentration characteristics of pollution in surrounding areas to some extent while enhancing ULGUE. Based on the above analysis, this paper proposes research hypothesis 2.

Hypothesis 2.

The implementation of LCCPP has a positive spatial spillover effect on the ULGUE.

2.3. Influence Mechanism of Energy Utilization Intensity and Urban Innovation Level

As previously mentioned, the impact of LCCPP and factors influencing ULGUE are multidimensional, encompassing economic, social, and ecological benefits. However, their core lies in sustainable development: minimizing environmental harm while maximizing economic output during the development process [42]. Environmental pollution is directly linked to energy usage. Currently, traditional energy sources such as oil, natural gas, and coal dominate global energy consumption with clean energies like nuclear power, wind power, and biomass being supplementary. The promotion of new energies is still accelerating. LCCPP will directly affect urban energy utilization intensity by developing low-carbon green industries and advocating for a low-carbon green lifestyle [23,43], thereby reducing environmental pollution while improving ULGUE. Additionally, besides economic output being related to resource input amounts and allocation methods [44], the technological innovation level also plays an important role in this regard. According to “Porter’s hypothesis” [45], the impact of LCCPP on economic output is to promote the technological innovation of enterprises. Although short-term costs may increase for enterprises due to environmental protection measures taken during production processes, in the long run, the production efficiency of enterprises will be improved to make up for the cost incurred in the process of environmental protection, thus promoting economic growth. Furthermore, implementing this policy has promoted innovation in low-carbon green technology which has had significant positive impacts on pollution reduction [46] leading ultimately towards continuous improvement in urban land’s green use efficiency. To sum up, LCCPP can affect ULGUE through two paths: energy utilization intensity and urban innovation level. This paper proposes research hypothesis 3 and research hypothesis 4 in order to examine the rationality of its influence mechanism.

Hypothesis 3.

LCCPP improves ULGUE by reducing energy utilization intensity.

Hypothesis 4.

LCCPP improves ULGUE by improving urban innovation level.

Through empirical testing of the four research hypotheses mentioned above, we can gain a better understanding of whether LCCPP will lead to an improvement in ULGUE. Additionally, we can analyze the impact of these policies on economic, social, and ecological benefits based on relevant studies. This will provide a solid foundation for drawing final policy recommendations.

3. Methodology and Data

3.1. Research Area

In terms of individuals, considering the availability and validity of data, 283 cities including Beijing, Tianjin, Shanghai, Chongqing, Shijiazhuang, and Tangshan were selected as research samples. In terms of the research period, on one hand, LCCPP was implemented in 2010. In order to better evaluate the policy’s effectiveness, the starting year of the research was set as 2007. On the other hand, to mitigate any potential impact from the global novel coronavirus outbreak in 2020 on land’s green use efficiency, the end year of the research was set as 2019. The specific distribution of low-carbon pilot cities and research area can be seen in Figure 1.

3.2. Measurement of ULGUE

3.2.1. Super-Efficiency Slack-Based Measure Model with Undesirable Outputs

Currently, the primary methods utilized for frontier efficiency evaluation are Stochastic Frontier Analysis (SFA) [47] and Data Envelopment Analysis (DEA) [48]. However, in contrast to the DEA method, the SFA method is more complex when dealing with multi-output scenarios as it requires combining multiple outputs into a comprehensive output. Moreover, if there are numerous input indicators, the reliability of SFA results may be affected due to indicator correlation. On the other hand, DEA does not necessitate a specific production function. Instead, it cleverly constructs an objective function and transforms fractional programming problems into linear programming problems through Charnes–Cooper transformation. It eliminates the need for standardizing indicator dimensions or setting input–output weights by determining them through optimization processes. Consequently, Decision Making Unit (DMU) evaluation becomes more objective [49]. Traditional DEA models include Charnes–Cooper–Rhodes (CCR) and Bank–Charnes–Cooper (BCC), both of which are radial models assuming proportional changes in inputs and outputs. Subsequently, scholars have proposed non-radial models such as the Slack-Based Measure (SBM) model [50], which incorporates slack variables into the objective function to eliminate result deviations caused by radial and angle selection biases. However, this model cannot further analyze cases where efficiency values remain at 1 or have identical values. Additionally, traditional DEA methods do not incorporate negative environmental externalities generated during production processes into their objective functions. To overcome these limitations, Tone [51] also proposed a Super-Efficiency Slack-Based Measure (super-SBM) model that includes undesirable outputs. Based on the above analysis and taking into account the undesirable outputs such as carbon dioxide and industrial pollution generated in the process of land use, this paper adopts this model to measure the ULGUE. The model is specifically expressed as follows:

$${\theta }^{*}=\underset{\lambda ,s^{-},s^{+}}{\min }\frac{1+\frac{1}{m}\sum _{i=1}^m\frac{s_i^{-}}{x_{io}^t}}{1-\frac{1}{q+h}(\sum _{r=1}^q\frac{s_r^{+}}{y_{ro}^t}+\sum _{k=1}^h\frac{s_k^{-}}{b_{ko}^t})}$$

(1)

$$s.t.\left\{\begin{array}{l}{x}_{io}^t\ge {\sum }_{t=1}^T{\sum }_{j=1,j\ne o}^n{{\lambda }_j^{t}x}_{ij}^t-s_i^{-},\hspace{1em}\hspace{1em}i=\mathrm{1,2}, \dots ,m;\\ y_{ro}^t\le {\sum }_{t=1}^T{\sum }_{j=1,j\ne o}^n{{\lambda }_j^{t}y}_{rj}^t+s_r^{+},\hspace{1em}\hspace{1em}r=\mathrm{1,2},\dots ,q;\\ b_{ko}^t\ge {\sum }_{t=1}^T{\sum }_{j=1,j\ne o}^n{{\lambda }_j^{t}b}_{kj}^t-s_k^{-},\hspace{1em}\hspace{1em}k=\mathrm{1,2},\dots ,h;\\ \hfill {\sum }_{t=1}^T{\sum }_{j=1,j\ne o}^n{\lambda }_j^t=1\hfill \\ \hfill {\lambda }_j^t\ge 0(\forall j), s_i^{-}\ge 0(\forall i), s_r^{+}\ge 0(\forall r), s_k^{-}\ge 0(\forall k)\hfill \end{array}\right.$$

(2)

Equations (1) and (2) together form a global non-angular super-SBM model with variable returns to scale including undesirable outputs. In the model, ${\theta }^{*}$ represents ULGUE; T is the research period span; n is the number of decision-making units; m, q, and h are the number of variables of input, desirable output, and undesirable output, respectively; $x_{io}^t$, $y_{ro}^t$, and $b_{ko}^t$ are slacked variables corresponding to input, desirable output, and undesirable output, respectively. Due to the adoption of panel data encompassing multiple years at the prefecture level in this study, a global estimation method is employed for specific calculations, which facilitates the comparison of observed objects across different years. The model is also closer to China’s actual development situation by using variable returns to scale and non-angle methods to measure ULGUE. In this paper, the efficiency value is calculated by using MAXDEA 8.22 Ultra software.

3.2.2. Selection of Measurement Indicators

The objective of promoting sustainable land use in urban areas is to achieve the harmonization and integration of economic, social, and ecological benefits [52]. Given the influence of LCCPP, ensuring high-quality sustainable development of the social economy becomes particularly crucial. Therefore, based on relevant literature research, this paper establishes an indicator system for assessing the efficiency of ULGUE from three perspectives: input, desirable output, and undesirable output. The specific information regarding indicator variables is presented in

Table 1. Evaluation index system of ULGUE.

Index	Specific Index	Index Composition	References
Input	Capital	Fixed capital stock of municipal district (unit: 10,000 yuan)	Chang et al. [26]
	Labor	Number of employees of secondary and tertiary industries in municipal districts (unit: 10,000 people)	Feng et al. [53]
	Land	Built-up area of municipal district (unit: km2)	Xie et al. [52]
Desirable output	Economy	Added value of secondary and tertiary industries in municipal districts (unit: 10,000 yuan)	Wang and Han. [54]
	Society	Average wages of urban workers in municipal districts (unit: yuan)	Xie et al. [52]
	Ecology	Total carbon sink of urban green space in municipal district (unit: 10,000 tons)	Dou et al. [55]
Undesirable output	Industrial pollution	Industrial sulfur dioxide emissions (unit: 10,000 tons)	Wang and Han [54]
		Industrial soot emission (unit: 10,000 tons)
		Industrial wastewater discharge (unit: 10,000 tons)
	Carbon emission	Municipal district carbon emissions (unit: 10,000 tons)	Liu et al. [56]

.

The variables in this paper are generally utilized at the municipal district level to mitigate the potential inconsistency of data statistical caliber resulting from significant administrative division changes. In terms of input, this paper selects the factors of capital, labor, and land. For estimating the capital stock, this paper adopts the “perpetual inventory method” proposed by Zhang et al. [57]. Firstly, 2007 is designated as the base period, and ten times of the actual fixed asset investment in that year is considered as the base period capital stock. Then, the annual nominal fixed asset investment is converted to constant prices in 2007 by using the fixed asset investment price index. Finally, the annual fixed capital stock is calculated based on a depreciation rate of 9.6%. The utilization of urban land primarily focuses on developing secondary and tertiary industries, so total employment in these sectors is chosen to represent labor input. Compared with urban construction land area, urban built-up area better reflects actual land input due to its concentration of human activities. Henceforth, urban built-up area serves as an indicator for measuring land factor input.

In terms of desirable outcomes, this paper selects indicators from three dimensions: economy, society, and ecology. Also taking into account the urban land development utilization, this paper takes the added value of secondary and tertiary industries and the average wage of urban workers as the output indicators of economic and social levels, respectively. In order to conduct a more comprehensive assessment of the effectiveness of LCCPP, this paper quantifies the ecological output by measuring the total carbon absorption capacity of urban green spaces [58]. The specific calculation formula is shown in Equation (3):

$$C_i=A_i\times f_i$$

(3)

where $C_i$ represents the total amount of urban green space carbon sink each year; $A_i$ represents the urban green space area; $f_i$ represents the urban green space carbon sink coefficient, which is taken as 1.66 C/ha/year [59].

In terms of undesirable outcomes, considering the LCCPP will yield a collaborative reduction in emissions; it encompasses industrial sulfur dioxide, soot, wastewater, and carbon emissions simultaneously. Among these, carbon emissions originate from direct sources within the municipal district, primarily including greenhouse gas emissions resulting from transportation and construction activities, industrial production processes, agricultural practices, forestry activities, and land use changes, as well as waste treatment operations [60].

3.3. Model Building

3.3.1. Multi-Period Difference-in-Differences Model

At present, the prevailing approach for policy evaluation is the difference-in-differences (DID) model, which effectively addresses endogeneity issues associated with unobservable time-invariant individual characteristics [61]. Given that the implementation of LCCPP occurs in three distinct phases, this study employs a multi-period DID model to examine the impact of the policy. The concrete model is set to Equation (4):

$${ULGUE}_{it}={\alpha }_0+{\alpha }_1{did}_{it}+X_{it}\gamma +{\mu }_i+{\delta }_t+{\epsilon }_{it}$$

(4)

where i and t represent individual and year, respectively; ${ULGUE}_{it}$ represents urban land green use efficiency; ${did}_{it}$ is the product of ${treated}_i$ and ${time}_t$, where ${treated}_i$ and ${time}_t$ represent the dummy variables of processing group individual and policy time, respectively; $X_{it}$ represents a set of control variables; vector $\gamma$ represents the corresponding regression coefficients of each control variable; ${\mu }_i$ represents the fixed effect of the controlled individual; ${\delta }_t$ represents the fixed effect of the controlled time; ${\epsilon }_{it}$ represents the random interference term; ${\alpha }_0$ represents the intercept term of the result of the regression estimate and; ${\alpha }_1$ is the regression coefficient of the core explanatory variable we need to pay attention to, which reflects the difference in ULGUE changes between low-carbon pilot cities and non-low-carbon pilot cities before and after the implementation of the policy.

3.3.2. Spatial Difference-in-Differences Model

Considering the competition and imitative behavior exhibited by local government officials during the decision-making process, as well as the negative externality of environmental pollution, this paper adopts the spatial Durbin DID model to examine the spatial spillover effect of LCCPP, drawing reference from Chen and Wang’s method [6]. Therefore, we constructed Equation (5):

$$\begin{gathered} {ULGUE}_{it}={\alpha }_0+{\rho }_1{\sum }_{j=1,j\ne i}^n{W}_{ij}{ULGUE}_{jt}+{\alpha }_1{did}_{it}+{\rho }_2{\sum }_{j=1,j\ne i}^n{W}_{ij}{did}_{jt} \\ +X_{it}\gamma +{\sum }_{j=1,j\ne i}^n{W}_{ij}{X}_{jt}\theta +{\mu }_i+{\delta }_t+{\epsilon }_{it} \end{gathered}$$

(5)

where W is the normalized spatial weight matrix, $\sum _{j=1,j\ne i}^n{W}_{ij}{ULGUE}_{jt}$ is the spatial lag term of ULGUE, and ${\rho }_1$ is the corresponding spatial autoregressive coefficient; $\sum _{j=1,j\ne i}^n{W}_{ij}{did}_{jt}$ is the spatial lag term of LCCPP, and the corresponding regression coefficient ${\rho }_2$ represents the spatial spillover effect of the policy. Individual fixed effect, time fixed effect, and random interference terms are denoted by ${\mu }_i$, ${\delta }_t,$ and ${\epsilon }_{it}$, respectively.

Drawing on theoretical analysis, this study selects a spatial weight matrix based on economic geographical distance nested [62] to capture the potential spatial spillover effects associated with both geographic and economic factors. The specific construction process of this matrix is outlined in Equation (6):

$$W_{ij}^{de}=\left\{\begin{array}{c} W_{ij}^d\times W_{ij}^e i\ne j\\ 0 i=j\end{array}\right.$$

(6)

where $W_{ij}^d$ and $W_{ij}^e$ are the inverse geographical distance matrix and the economic distance matrix, respectively.

3.3.3. Intermediary Effects Model

Based on theoretical analysis, LCCPP has the potential to enhance ULGUE by reducing energy utilization intensity and improving urban innovation level. Therefore, drawing inspiration from the studies conducted by MacKinnon et al. [63] and Edwards and Lambert [64], this paper constructs a mediation effect model to empirically examine these two influential pathways:

$$M_{it}={\beta }_0+{\beta }_1{did}_{it}+X_{it}\gamma +{\mu }_i+{\delta }_t+{\epsilon }_{it}$$

(7)

$${ULGUE}_{it}={\alpha }_0+{\alpha }_1{did}_{it}+{\alpha }_2{M}_{it}+X_{it}\gamma +{\mu }_i+{\delta }_t+{\epsilon }_{it}$$

(8)

where $M_{it}$ represents the intermediary variable, which is substituted into the variables of energy utilization intensity and urban innovation level, respectively, in the concrete empirical process. In this paper, Equation (7) is firstly used to test the impact of LCCPP on the intermediary variables, and then, the intermediary variables are included in Equation (4), that is, Equation (8) is used to test the rationality of the impact path.

3.4. Variable Setting and Data Resource

Dependent variable: Urban land green use efficiency (ULGUE). As mentioned above, the super-SBM model with undesirable outputs is used for calculation.

Core independent variable: Virtual indicator for LCCPP (did). During the research period, the did value for non-low-carbon pilot cities is 0. For low-carbon pilot cities, the did value is 0 before the year of policy implementation and 1 after it. Regarding the selection of the year of policy implementation, since both of the first and second batches of low-carbon city pilot took place in the latter half of their respective years, considering that there is a certain time lag in policy effects, we set 2011 and 2013 as the years for policy implementation for the first and second batches, respectively.

Control variables: In order to alleviate the endogenous problem caused by the impact of missing variables on ULGUE, this paper selects the following control variables: (1) Economic level (agdp). Cities with a high level of economic development usually have more resources and technology and are better able to support the implementation of green land use [56], so the natural logarithm of per capita urban gross regional product is utilized. (2) Population size (density). Cities with high population density often experience greater land pressure, which can limit land use efficiency. Additionally, population density has a direct impact on the severity of environmental pollution. Higher population density increases environmental pollution and can negatively affect ULGUE [56]. Therefore, the natural logarithm of urban population density is employed as a measure. (3) Government financial support (government). China’s fiscal decentralization has resulted in an increase in local fiscal deficits. To address the gap between fiscal revenue and expenditure, as well as intense regional competition, local governments have resorted to expanding the scale of industrial land to boost fiscal revenue and achieve rapid economic development. However, this practice may have a negative impact on ULGUE [65]. Therefore, it is assessed by the ratio of general financial revenue to general financial expenditure for each local government. (4) Level of opening up (fdi). The introduction of foreign capital can contribute to the development of urban industrial systems and improve the overall economic level. However, without proper government monitoring, it can also lead to the proliferation of low-end industries [20], which can harm ULGUE. And the proportion of actual foreign capital utilization in GDP denominated in RMB is calculated by using the exchange rate for each city in the current year. (5) Traffic condition (road). A well-developed public transport system and a convenient transportation network can effectively reduce reliance on private vehicles, alleviating road congestion and reducing traffic emissions. This not only contributes to the conservation of land resources but also enhances the overall efficiency of land use [7]. It is measured by the natural logarithm of per capita urban road area owned. (6) Science education level (technology). Investment in science and education is essential for fostering research and innovation in environmental protection and sustainable development, as well as enhancing people’s environmental awareness and skills [66]. And the natural logarithm of total science and education expenditure for each city is employed to measure this. (7) Industrial structure (industry). The type of industrial structure plays a crucial role in determining the mode of land use development. Excessive expansion of the secondary industry can lead to increased pollution emissions, while the development of the tertiary industry promotes the efficient use of land [67]. Hence, this is measured by the proportion of value added from tertiary industry and secondary industry within each city’s economy. (8) Human capital level (hcl). To improve the level of human capital, strengthen education and training, and improve people’s environmental awareness and skills are crucial to improve ULGUE. The measurement is determined by taking the natural logarithm of the student population in regular urban higher education institutions.

Intermediary variables: Energy utilization intensity (energy) and urban innovation level (innovation). Regarding the data on energy utilization intensity, we initially convert the total social electricity consumption, total supply of artificial gas and natural gas, as well as total supply of liquefied petroleum gas based on their respective standard coal coefficients. We then aggregate them to obtain the overall tons of standard coal consumption and apply natural logarithm to characterize urban energy utilization intensity. Concerning the data on urban innovation level, this study adopts Kou and Liu’s approach [68] to comprehensively calculate the urban innovation index using patent data from China’s State Intellectual Property Office and newly established enterprise data from various cities. Given that original urban innovation power indices vary significantly across different cities, we also treat it with the natural logarithm in order to avoid heteroscedasticity during analysis.

The index data utilized in this paper primarily originate from the China Urban Statistical Yearbook, China Urban Construction Statistical Yearbook, China Urban and Rural Construction Statistical Yearbook, China Energy Statistical Yearbook, China Environmental Statistical Yearbook, as well as provincial and municipal statistical yearbooks and government statistical bulletins. In cases of missing data, this study employs adjacent year substitution or linear interpolation methods for imputation purposes. All price data are adjusted to the constant price of 2007 based on the corresponding price index of the respective year. Following a series of calculations and processing steps, we ultimately obtain a balanced panel dataset comprising 283 cities spanning from 2007 to 2019.

Table 2. Descriptive statistics of variables.

Variable	Symbol	Observations	Mean	Std. Dev.	Min	Max
Urban land green use efficiency	ULGUE	3679	0.2823	0.1888	0.0039	1.9215
Virtual indicator for LCCPP	did	3679	0.0960	0.2946	0.0000	1.0000
Economic level	agdp	3679	10.5876	0.6187	8.0868	15.4191
Population size	density	3679	6.4587	0.9342	2.5649	9.3457
Government financial support	government	3679	0.5667	0.2767	0.0228	8.3902
Level of opening up	fdi	3679	0.0216	0.0235	0.0000	0.2265
Traffic condition	road	3679	2.4363	0.5347	0.0182	4.6943
Science education level	technology	3679	11.6701	1.1422	8.2933	16.2634
Industrial structure	industry	3679	1.0762	0.6619	0.0943	6.5326
Human capital level	hcl	3679	10.3659	1.4181	2.3026	13.9169
Energy utilization intensity	energy	3679	13.6816	1.2344	9.3537	17.5211
Urban innovation level	innovation	3679	1.3345	1.2907	0.0000	7.5828

presents descriptive statistics for each variable, which includes the symbol representing each variable, the number of observations in the sample, the mean value, standard deviation, minimum value, and maximum value. The software Stata 17 is used to validate the aforementioned model and conduct related tests. By calculating the variance inflation factor (VIF) for each variable, it is determined that all VIF values are significantly below 5, indicating an absence of severe multicollinearity among the variables.

4. Empirical Results

4.1. Spatial Evolutionary Characteristics of ULGUE

In order to facilitate the observation and comparison of land green use efficiency across different cities and years, we organized the 3679 data on ULGUE in ascending order and classified them using a quantile classification method with a classification interval of 20%. Based on this analysis, this study utilized ArcGIS 10.8 software to depict spatial pattern characteristics of ULGUE in four selected years, namely, 2007, 2011, 2015, and 2019 (see Figure 2). From the perspective of distribution convergence, ULGUE exhibits distinct spatial agglomeration characteristics at various levels, indicating both intra-city variations in land use and inter-city interactions during the process. This phenomenon may be attributed to the spatial spillover effect of environmental pollution and the mutual imitation behavior of local government decision-making, among other factors. Specifically, the manner in which land is utilized within urban areas can exert a substantial influence on the environment. For instance, high-density construction can result in air and water pollution, among other problems. These pollutants can not only spread and circulate within the city but can also extend to neighboring areas through atmospheric wind patterns and water flow. As a result, the spatial spillover effect of environmental pollution can cause a decline in ULGUE in the surrounding areas of the city, leading to a spatial agglomeration of lower levels of ULGUE. Additionally, when formulating land use policies, local governments may look to and imitate the successful experiences of neighboring cities. This imitation behavior can lead to a convergence between cities in terms of land green use efficiency, with more and more cities adopting similar strategies and measures. Consequently, this further reinforces the spatial agglomeration characteristics of higher levels of ULGUE. In terms of distribution, the ULGUE was relatively high in the central and western regions in 2007, while it generally remained low or below average in the eastern region. However, based on distribution maps from 2011 to 2015, there was a gradual inward spread of higher green use efficiency from central and coastal areas. By 2019, except for certain cities such as Weinan, Shangluo, and Xiaogan that still exhibited low or below-average levels, most cities demonstrated improved overall green use efficiency with reduced inter-city disparities. According to statistics, the average ULGUE of the selected sample of prefecture-level cities increased by 11.71 percentage points from 2007 to 2019. Additionally, the average ULGUE of non-low-carbon pilot cities is 0.2704, while the average ULGUE of low-carbon pilot cities is 0.3940. This represents a 12.36 percentage point increase compared to the former. Therefore, this paper conducted an empirical test to examine the impact of LCCPP on ULGUE.

4.2. Direct Effect Analysis

4.2.1. Parallel Trend Test

Before conducting regression analysis of multi-period DID model, it is essential to perform parallel trend testing (see Figure 3).

In this study, an event analysis was employed to generate fixed-period dummy variables based on the year of occurrence for each LCCPP. By comparing the dynamic effects of LCCPP before and after its implementation, we can assess whether the policy has a significant impact on land green use efficiency in the pilot city. As depicted in Figure 3, values ranging from −9 to −1 represent 9 years prior to policy implementation, 0 represents the year when the policy was implemented, and values from 1 to 8 indicate 8 years following policy implementation. The dynamic effect value of the policy exhibits slight fluctuations around the zero line before its implementation but demonstrates a noticeable upward trend afterward. This suggests that there exists a substantial disparity in policy effectiveness between pilot cities and non-pilot cities, thereby satisfying the parallel trend hypothesis.

4.2.2. Basic Regression

In order to examine the direct impact of LCCPP on ULGUE, this study conducted an empirical analysis based on model (4) and obtained the subsequent regression results (see

Table 3. The results of multi-period DID regression.

Variable	ULGUE
	(1) DID	(2) DID	(3) PSM-DID	(4) PSM-DID
did	0.0628 ***	0.0589 ***	0.0397 ***	0.0381 ***
	(6.5098)	(6.2279)	(3.9436)	(3.8542)
agdp		0.0417 ***		0.0415 ***
		(3.3714)		(3.1424)
density		−0.0206 **		−0.0235 ***
		(−2.4268)		(−2.8436)
government		0.0058		0.0323 *
		(0.5950)		(1.7258)
fdi		−0.4439 ***		−0.4942 ***
		(−3.2416)		(−3.4873)
road		−0.0159 *		−0.0105
		(−1.8262)		(−1.2457)
technology		0.0219 ***		0.0219 ***
		(2.9523)		(2.8776)
industry		0.0393 ***		0.0345 ***
		(4.3359)		(3.7238)
hcl		−0.0040		−0.0029
		(−0.6870)		(−0.3691)
Year fixed	Yes	Yes	Yes	Yes
City fixed	Yes	Yes	Yes	Yes
Observations	3679	3679	3509	3509
R-squared	0.6840	0.6903	0.6920	0.6980

Note: T statistics are in parentheses. *** p < 0.01, ** p < 0.05, * p < 0.1.

). Column (1) presents the regression results without incorporating any additional control variables except for the control year and city fixed effect, while column (2) represents the regression results after including a series of control variables. The findings reveal that regardless of whether control variables are added or not, the regression coefficients of policy dummy variables exhibit significant positive effects at a 1% significance level. This suggests that the implementation of LCCPP significantly promotes improvements in land green use efficiency within pilot cities. Specifically, after controlling for various variables, it is observed that low-carbon pilot cities experienced a 5.89% increase in land green use efficiency compared to non-pilot cities. Therefore, hypothesis 1 is confirmed.

Based on the results of the control variables, it can be observed that the coefficient of the variable agdp is significantly positive. This implies that as the level of economic development of a city increases, the ULGUE also increases. This can be attributed to the fact that as the economy develops, people become more aware of environmental protection. Consequently, both the government and enterprises focus more on green development and implement measures to reduce land damage and pollution [69]. Furthermore, the coefficient of the variable density is significantly negative. This suggests that a higher population density, which is often associated with urbanization, leads to a decrease in green land use efficiency. The reason behind this is that the construction of numerous residential, commercial, and transportation facilities in urban areas occupies a substantial amount of land, limiting the available space for green land utilization. Moreover, the coefficient of the variable fdi is significantly negative. This indicates that foreign-funded enterprises from regions with lower environmental standards may not prioritize environmental protection to the same extent [70]. Consequently, they tend to adopt cheaper production methods that exert greater pressure on the environment, resulting in lower ULGUE. Additionally, there is a significant positive correlation between industrial structure (industry) and ULGUE. Traditional heavy industries and resource-intensive industries often exert high pressure on land resources, leading to lower ULGUE. However, with the optimization and upgrading of industrial structure, the transition to technology-intensive, knowledge-intensive, and service industries occurs. This shift changes the demand for land resources and the methods of utilization, ultimately improving ULGUE. The impact of science and education investment (technology) on land green use efficiency is multifaceted. On one hand, investing in science and education promotes scientific and technological innovation, enhances production efficiency, and improves resource utilization efficiency in various sectors such as agriculture, industry, and services. On the other hand, science and education investment may also lead to changes in industrial structure. For instance, the development of technology-intensive industries may present new challenges in land resource utilization, necessitating corresponding green development measures to enhance ULGUE. Relatively speaking, the variables government, road, and hcl have no significant influence on ULGUE.

4.2.3. Placebo Test

The placebo test in the DID model aims to mitigate the influence of non-policy factors on research outcomes, thereby avoiding subjective changes resulting from subjects being informed about the policy signal in advance and potential errors in assessing the “policy effect” [71]. To address this concern, this study re-estimates the model by randomly advancing the implementation timing of policy for the treatment group and employing new samples. This process is repeated 1000 times to obtain a probability distribution as depicted below (see Figure 4). As illustrated in the figure, the estimated coefficient conforms largely to a normal distribution with an average value significantly lower than that of the true coefficient estimate (0.0589). Consequently, it can be inferred from a counterfactual perspective that advancing policy establishment time at random leads to a substantial decrease in land green use efficiency driven by LCCPP. This finding further confirms that low-carbon cities indeed enhance land green use efficiency within their respective locations.

In order to avoid the possible individual bias in the original selection of the treatment group and the control group [72], this study also randomizes the policy time and the samples of the treatment group at the same time for re-estimation, and the process is repeated 1000 times to draw the following coefficient kernel density plot and p-value scatter plot (see Figure 5). The randomized coefficients for the did term predominantly cluster around zero, with most of the p-values exceeding 0.1. Moreover, the distribution of random coefficients is primarily skewed towards lower values compared to the true value of 0.0589, indicating that double randomization significantly weakens both the statistical significance and effect intensity of the policy impact, thus indirectly confirming the robustness of our original conclusion.

4.2.4. PSM-DID

To mitigate the endogeneity problem caused by sample selection bias and further enhance the randomness of treatment group and control group selection, this study employed Propensity Score Matching (PSM) method to match original samples year by year with a 1 to 2 caliper radius. Figure 6 displays kernel density estimation results of propensity score values for both groups before and after matching. The figure demonstrates that, following the matching process, the nuclear density curves of the treated group and control group exhibit a closer alignment compared to their pre-matching state. And the mean line distance between the two groups has been reduced. Hence, it can be inferred that year-by-year matching results has proven effective to a certain extent.

The new sample, obtained after matching, is utilized for regression estimation, and the corresponding results are presented above (see Table 3). Columns (3) and (4) represent the regression outcomes without and with control variables, respectively. Irrespective of whether control variables are incorporated, the results consistently demonstrate a significantly positive coefficient for did at a 1% significance level. However, this coefficient has decreased compared to its value prior to matching. This indicates that the implementation effect of the LCCPP is overestimated to some extent due to the existence of sample selection bias, but the test results for hypothesis 1 prove to be robust.

According to the regression results of the control variables, the positive and negative signs of the coefficients for economic level (agdp), population size (density), level of opening up (fdi), science education level (technology), and industrial structure (industry) remain unchanged and are all significant at the 1% level. This reaffirms the robustness of the potential influencing factors of ULGUE mentioned above. However, the impact of government financial support (government), traffic condition (road), and human capital level (hcl) remains relatively insignificant.

4.3. Spatial Spillover Effect Analysis

4.3.1. Spatial Correlation Test

To examine the spatial spillover effects of LCCPP, we initially analyze the global Moran index of ULGUE in 283 cities across China from 2007 to 2019. The values of the Moran index from 2007 to 2019, as shown in

Table 4. The global Moran index of ULGUE in China from 2007 to 2019.

Year	I	Z	p-Value
2007	0.033	1.159	0.246
2008	0.052	1.726	0.084
2009	0.035	1.215	0.224
2010	0.020	0.738	0.461
2011	0.049	1.630	0.103
2012	0.066	2.170	0.030
2013	0.074	2.429	0.015
2014	0.058	1.915	0.056
2015	0.128	4.133	0.000
2016	0.116	3.746	0.000
2017	0.216	6.851	0.000
2018	0.132	4.220	0.000
2019	0.149	4.719	0.000

, are all greater than zero, indicating a positive spatial autocorrelation in ULGUE. Although the p-value results are not significant from 2007 to 2011, they are significantly positive from 2012 to 2019. This consistency with the starting year of low-carbon city pilot policy implementation suggests that the policy had a spatial spillover effect and met the prerequisite for subsequent spatial DID regression.

4.3.2. Spatial DID

Before conducting the specific regression analysis, this paper employed LM test, LR test, Hausman test, and Wald test to determine the appropriate model for spatial DID. The results show that SDM model cannot be degraded to SEM or SAR model, and two-way fixed effects are selected, that is, model (5). Subsequently, model (5) was used for regression analysis and yielded the following results (see

Table 5. The results of spatial DID regression.

Variable	ULGUE
	(1) Main	(2) Wx	(3) LR_Direct	(4) LR_Indirect	(5) LR_Total
did	0.0468 ***	0.1412 ***	0.0495 ***	0.1607 ***	0.2101 ***
	(2.9106)	(4.8576)	(2.9789)	(5.0698)	(5.8304)
agdp	0.0497 ***	−0.0400	0.0485 ***	−0.0405	0.0081
	(2.9147)	(−1.4823)	(2.9455)	(−1.4170)	(0.2366)
density	−0.0238	0.0115	−0.0218	0.0143	−0.0075
	(−1.4953)	(0.3177)	(−1.4151)	(0.3360)	(−0.1541)
government	0.0122	−0.0165	0.0120	−0.0176	−0.0056
	(1.0516)	(−1.6245)	(1.0558)	(−1.6192)	(−0.3378)
fdi	−0.3368 **	−1.0310 ***	−0.3489 **	−1.1675 ***	−1.5165 ***
	(−2.0691)	(−2.6418)	(−2.1772)	(−2.7544)	(−3.2025)
road	−0.0181	−0.0258	−0.0180 *	−0.0311	−0.0492 *
	(−1.7661)	(−1.2093)	(−1.8113)	(−1.3970)	(−1.9150)
technology	0.0238 **	0.0209	0.0242 **	0.0248	0.0490 *
	(2.1132)	(0.9049)	(2.1504)	(0.9549)	(1.7500)
industry	0.0424 ***	0.0088	0.0418 ***	0.0168	0.0585 **
	(3.1028)	(0.3840)	(3.1289)	(0.6815)	(2.1271)
hcl	−0.0009	0.0248	0.0004	0.0271	0.0275
	(−0.0985)	(1.2258)	(0.0411)	(1.2101)	(1.0800)
rho	0.0936 ***
	(2.7262)
sigma2_e	0.0098 ***
	(7.9595)
Year fixed	Yes	Yes	Yes	Yes	Yes
City fixed	Yes	Yes	Yes	Yes	Yes
Observations	3679	3679	3679	3679	3679
R-squared	0.1352	0.1352	0.1352	0.1352	0.1352

Note: Z statistics are in parentheses. *** p < 0.01, ** p < 0.05, * p < 0.1.

). In column (1), the regression coefficient of did is significantly positive, indicating that LCCPP contributes to enhancing land green use efficiency in pilot cities themselves. The regression coefficient of did in column (2) also demonstrates a significant and positive spatial spillover effect on the enhancement of ULGUE, thereby highlighting the effectiveness of this policy. The main reason is that LCCPP not only enhances ULGUE in the target city but also serves as a role model for neighboring cities. As a result, other city governments are more likely to adopt the practices of low-carbon cities, such as bolstering the construction of ecological and green infrastructure, optimizing the industrial layout of low-carbon economy, and establishing mechanisms for ecological compensation, which can enhance the effectiveness of green land use and facilitate sustainable urban development. Similarly, the inclusion of did terms in columns (3) and (4) further substantiates these findings. Consequently, it can be concluded that hypothesis 2 is supported by these results.

4.4. Influence Mechanism Test

In order to examine the impact pathway of LCCPP on ULGUE, this study empirically investigates the effects of energy utilization intensity and urban innovation level using models (6) and (7), yielding the following results (see

Table 6. The results of the influence mechanism test.

Variable	Energy	ULGUE	Innovation	ULGUE
	(1)	(2)	(3)	(4)
energy		−0.0276 ***
		(−4.6607)
innovation				0.0635 ***
				(8.8407)
did	−0.1141 ***	0.0557 ***	0.3600 ***	0.0361 ***
	(−3.8535)	(6.0124)	(12.7223)	(3.9364)
agdp	0.2129 ***	0.0476 ***	−0.0934 ***	0.0476 ***
	(3.8308)	(3.5951)	(−3.5058)	(3.8135)
density	−0.0221	−0.0212 **	−0.0834 **	−0.0153 *
	(−0.4970)	(−2.4030)	(−2.3229)	(−1.8135)
government	0.0361	0.0068	−0.0003	0.0058
	(1.1712)	(0.6721)	(−0.0142)	(0.6104)
fdi	0.1475	−0.4398 ***	−1.8696 ***	−0.3253 **
	(0.3515)	(−3.2568)	(−3.9707)	(−2.3465)
road	−0.0502	−0.0173 *	−0.0099	−0.0153 *
	(−1.1372)	(−1.9515)	(−0.3784)	(−1.7904)
technology	0.1855 ***	0.0270 ***	0.2012 ***	0.0092
	(7.0601)	(3.5453)	(9.0105)	(1.2392)
industry	0.0356	0.0403 ***	−0.0712 ***	0.0439 ***
	(1.1182)	(4.4258)	(−2.9523)	(4.8815)
hcl	0.0299	−0.0032	−0.0507 ***	−0.0008
	(1.4029)	(−0.5523)	(−2.9577)	(−0.1405)
Year fixed	Yes	Yes	Yes	Yes
City fixed	Yes	Yes	Yes	Yes
Observations	3679	3679	3679	3679
R-squared	0.9181	0.6928	0.9390	0.7017

Note: T statistics are in parentheses. *** p < 0.01, ** p < 0.05, * p < 0.1.

). Columns (1) and (2) present the test outcomes for energy utilization intensity, while columns (3) and (4) represent the test results for urban innovation level. From an energy utilization intensity perspective, column (1)’s regression coefficient for did is significantly negative, indicating that implementing LCCPP can stimulate a reduction in energy consumption. The regression coefficient for energy in column (2) is also significantly negative, while that of did is significantly positive, suggesting that decreasing energy utilization intensity in low-carbon pilot cities contributes to enhancing ULGUE—thus confirming hypothesis 3. This shows that the government will encourage low-carbon pilot cities to adopt clean energy and efficient energy technologies to promote the reduction in energy consumption, which, on the one hand, is conducive to reducing the dependence on traditional energy and the pressure of energy consumption on land resources but, on the other hand, reduces the impact of energy pollution on the land environment and, thus, improves ULGUE. Similarly, both did and innovation items’ regression coefficients in columns (3) and (4) are significantly positive, implying that implementing LCCPP facilitates improvements in urban innovation levels with a positive effect on land green use efficiency—confirming hypothesis 4. As a comprehensive environmental regulatory policy, the government aims to encourage cities to engage in technological innovation and industrial upgrading. Simultaneously, it imposes higher costs on enterprises for pollutant discharge, thereby compelling them to pursue technological advancements and strive for a balance between economic maximization and pollution reduction. Therefore, the green land use efficiency of low-carbon pilot cities can be improved.

4.5. Heterogeneity Analysis

4.5.1. Heterogeneity in Region

Chinese cities exhibit significant disparities in economic development across different regions and geographical locations. Hence, based on China’s western development policy, this paper divides cities into three regions: east, middle, and west, respectively, to test the differences in the impact effects of LCCPP. The results are shown in

Table 7. The results of regional heterogeneity regression.

Variable	ULGUE
	(1) East	(2) Mid	(3) West
did	0.0812 ***	0.0275	0.0513 **
	(6.2092)	(1.5121)	(2.5457)
agdp	0.0630 ***	0.0780 ***	0.0316
	(3.6773)	(4.2684)	(1.5349)
density	−0.0218 *	0.0014	−0.0285
	(−1.7589)	(0.1241)	(−1.1087)
government	−0.0089 *	0.0078	0.1072 *
	(−1.8871)	(0.3105)	(1.7914)
fdi	−0.2657	−0.3286	−0.6313
	(−1.4612)	(−1.6414)	(−0.8957)
road	−0.0121	−0.0159	−0.0052
	(−0.6457)	(−1.1782)	(−0.3065)
technology	0.0102	0.0067	0.0588 ***
	(0.8725)	(0.6844)	(3.2285)
industry	0.0357 **	0.0362 **	0.0499 ***
	(2.0641)	(2.4986)	(2.9978)
hcl	−0.0555 ***	−0.0064	0.0161 *
	(−3.3518)	(−1.0465)	(1.7659)
Year fixed	Yes	Yes	Yes
City fixed	Yes	Yes	Yes
Observations	1300	1300	1079
R-squared	0.7624	0.5874	0.6705

Note: T statistics are in parentheses. *** p < 0.01, ** p < 0.05, * p < 0.1.

. The findings are as follows: The influence of the policy on ULGUE is most pronounced and robust in the eastern region, relatively more substantial and stronger in the western region, but not statistically significant in the middle region. This discrepancy can be attributed to the relatively advanced state of development in the eastern region, characterized by a high level of economic progress and a strong commitment to environmental protection and sustainable growth. Consequently, under the promotion of LCCPP, improvements in land green use efficiency are most prominent within this region. In contrast, although both of the middle and western regions lag behind economically, the latter possesses abundant land resources and mineral reserves. While this may result in greater environmental harm during land development and economic production processes within the western region, it also generates higher economic output from land utilization compared to that observed within its middle counterpart. It can be seen that the implementation effectiveness of LCCPP closely correlates with regional economic development levels, which suggests that future improvements and innovations regarding low-carbon policies should be formulated based on each city’s individual developmental status.

4.5.2. Heterogeneity in Urban Types

Moreover, considering the disparities in resource endowments among Chinese cities, this study categorizes 283 cities into resource-based and non-resource-based cities based on the National Sustainable Development Plan for Resource-based Cities (2013–2020) issued by The State Council. Subsequently, a multi-period DID model is employed to conduct regression analysis separately (see

Table 8. The regression results of heterogeneity in urban types.

Variable	ULGUE
	(1) Non-Resource-Based	(2) Resource-Based
did	0.0774 ***	−0.0341 ***
	(6.6931)	(−2.9591)
agdp	0.0543 ***	0.0274 *
	(3.5898)	(1.7498)
density	−0.0043	−0.0672 ***
	(−0.4600)	(−4.5285)
government	0.0022	−0.0032
	(0.2532)	(−0.1157)
fdi	−0.1732	−0.6782 ***
	(−1.1172)	(−3.2167)
road	−0.0217 *	0.0050
	(−1.9154)	(0.3806)
technology	0.0199 **	0.0093
	(2.1813)	(0.7711)
industry	0.0429 ***	0.0286 **
	(3.6057)	(2.1638)
hcl	−0.0103	0.0031
	(−1.4859)	(0.3125)
Year fixed	Yes	Yes
City fixed	Yes	Yes
Observations	2210	1469
R-squared	0.7101	0.6697

Note: T statistics are in parentheses. *** p < 0.01, ** p < 0.05, * p < 0.1.

). The findings reveal that the coefficient of the did term in column (1) is significantly positive at a 1% level, while it is significantly negative at a 1% level in column (2). This suggests that LCCPP can enhance ULGUE in non-resource cities but hinder it in resource-based ones. This may be attributed to the fact that economic growth in resource-based cities heavily relies on exploiting and processing resources, and implementing LCCPP could have significant impacts on their industrial structure and economic development mode, leading to certain hindrances to their development. Additionally, since resource-rich cities tend to prioritize natural resource utilization over improving green utilization efficiency when it comes to land use planning, they may have lower adaptability towards LCCPP, which would further impede improvements of their land green use efficiency.

4.5.3. Quantile Regression

In the aforementioned empirical process, we conducted an analysis on the impact of LCCPP on the conditional expectation of ULGUE. However, this analysis overlooked the potential for structural changes in such impact across different levels of ULGUE. Therefore, to address this limitation, this paper adopts the quantile regression method [73] to explore variations in the effects of LCCPP based on different levels of ULGUE. Starting from the 5% sub-point and progressing at every 5% quantile interval up to 95%, we estimate the impact of LCCPP on ULGUE. The estimated results for each subsite are presented below (see Figure 7). As shown in the figure, the horizontal axis represents the different loci of ULGUE, and the vertical axis represents the regression coefficient value of did term in quantile regression results. The solid line depicting quantile regression results exhibits an upward trend, indicating that as ULGUE improves, the LCCPP has a stronger promoting effect on enhancing ULGUE. This observation also suggests that the LCCPP is independently implemented by local governments, and its effectiveness may be closely linked to their own implementation efforts. If a local government demonstrates strong positive initiative towards urban economic development and environmental governance, it will also exhibit greater proactiveness in embracing pilot policy for low-carbon cities.

5. Conclusions and Policy Implications

5.1. Conclusion Summary

5.1.1. Discussion

The increasingly pressing issues of global climate change and environmental pollution, combined with population growth and the escalating strain on land resources, have posed a formidable challenge for all nations in achieving low-carbon development within limited land availability. Therefore, conducting research on the impact of LCCPP on ULGUE can offer valuable insights and guidance for global urban planning and development.

By reviewing the studies on the impact of existing LCCPP on ULGUE, we have found that different scholars have reached different conclusions. Therefore, this study aims to contribute to and enhance the research content and theoretical framework of this topic. The main differences between this study and the other similar literature are as follows: (1) Connotation of ULGUE: Most scholars only consider industrial waste as a measure of undesirable output when assessing the connotation of ULGUE. However, this study takes into account the original intention of LCCPP, which is to reduce carbon emissions. Additionally, this study evaluates the sustainable use of land resources by comprehensively considering different factors. Therefore, this study includes carbon emissions and industrial waste as indicators of undesirable output to measure ULGUE. Furthermore, this study converts the coverage of green space into a carbon sink index in the ecological dimension of desirable output, providing a more objective and accurate assessment of the effects of LCCPP. (2) Measurement methodology of ULGUE: This study utilizes the global super-efficiency SBM model, which includes undesirable output, to measure ULGUE. Unlike the stochastic frontier analysis method, this measurement approach can identify decision-making units with high efficiency in both desirable and undesirable output and facilitates inter-individual comparisons and rankings, offering a valuable tool for sharing experiences and providing references for cities. Additionally, this method does not require the establishment of specific models during the calculation process, thereby enhancing objectivity. (3) Research methods: In addition to employing a multi-period DID approach combined with PSM to examine the direct effects of LCCPP, this study also employs a spatial Durbin DID model to explore the spatial spillover effects generated by the policy. Furthermore, a heterogeneity analysis is conducted through quantile regression, regional differentiation, and city type to provide a more detailed understanding of the impact of implementing LCCPP in different cities. (4) Impact mechanisms: This study examines the impact of LCCPP on ULGUE from the perspectives of energy utilization intensity and urban innovation level. By investigating these impact paths, the theoretical framework is further improved.

In summary, based on panel data from 2007 to 2019, encompassing 283 cities in China, this study employs the super-SBM model to measure ULGUE and analyzes its spatial-temporal pattern. The model incorporates the undesirable outputs of industrial pollution and carbon emissions. Subsequently, the quasi-natural experiment of LCCPP is utilized to examine its impact on land green use efficiency in both pilot and non-pilot cities through a multi-period DID model and spatial econometric model. Furthermore, employing an intermediary effect model, this study explores the specific influence mechanism of LCCPP on ULGUE via two pathways: energy utilization intensity and urban innovation level. Finally, a heterogeneity analysis is conducted based on different levels of ULGUE, regions, and urban types to elucidate how LCCPP affects such efficiency.

5.1.2. Conclusions

The key findings are as follows: (1) Overall, there has been an upward trend in ULGUE from 2007 to 2019, with an average increase of 11.71 percentage points. Moreover, there is a discernible spatial agglomeration feature, and the disparity between cities has gradually diminished. (2) The implementation of LCCPP enhances land green use efficiency within pilot cities themselves while also generating positive spillover effects on surrounding cities in terms of geographical distance and economic proximity, by improving their own land green use efficiency. (3) LCCPP incentivizes low-carbon pilot cities to reduce energy utilization intensity and enhance their level of urban innovation, thereby enhancing overall ULGUE. (4) Cities with higher land green use efficiency demonstrate greater initiative and enthusiasm in implementing the specific measures outlined in the LCCPP, thereby enhancing the effectiveness of policy implementation. The impact of the LCCPP on ULGUE is most pronounced in the eastern region, followed by the western region, while it exhibits no significant effect in the middle region. The pilot policy has played a substantial role in promoting green land use efficiency in non-resource-based cities, but it exerts an inhibitory influence on resource-based cities.

5.2. Policy Recommendations

It should be emphasized that, despite China’s top-down and multi-level administrative system, the implementation of LCCPP is reported from the grassroots level by local governments and through a series of independent initiatives. Therefore, while various countries around the world may have different administrative systems, this does not hinder the adoption of LCCPP to guide and inspire global efforts in sustainable construction. Consequently, based on these findings, this paper proposes the following countermeasures and suggestions for reference by regions and countries worldwide:

• Pay greater attention to the ULGUE. Various countries should establish a comprehensive land use planning system in urban areas, clearly define the functional positioning and rational utilization objectives of land, strengthen land management and supervision, and enhance the system of land use rights and market. This will effectively improve the ecological efficiency of land usage. Moreover, cities can be encouraged to engage in ecological restoration and greening initiatives, introduce sustainable urban agriculture and ecological farming models, as well as achieve multi-functional land use and resource recycling.
• Promote the implementation of LCCPP in a similar manner. Local governments can develop specific guidelines and policy measures for constructing low-carbon cities based on local resource conditions, infrastructure, economic development, and other factors. They should also clarify the tasks and objectives of pilot cities. For instance, preferential fiscal and tax policies can be introduced to incentivize enterprises and residents to adopt energy-saving and emission reduction measures. Additionally, special funds dedicated to low-carbon cities can be established to support the implementation of innovative low-carbon technologies and demonstration projects. Furthermore, it is crucial to enhance publicity and education efforts in order to raise public awareness about and engagement in low-carbon city construction.
• Strengthen the research and development and innovation of low-carbon technologies. During the development of low-carbon cities, it is essential for the government to increase investments in the research and development of low-carbon technologies. This will encourage technological innovation and application, ultimately leading to a reduction in energy consumption and an improvement in the efficient use of urban land. Additionally, it is crucial to foster collaboration between enterprises, universities, and scientific research institutions. This collaborative effort will drive the development and application of low-carbon technologies and help establish low-carbon innovative industries that can compete internationally.
• Enhance international cooperation and exchanges. All regions and countries should enhance international cooperation and exchanges to facilitate the sharing of successful experiences and technologies in green land use. By establishing an international cooperation platform, cities can promote collaboration and exchanges and jointly research and address global challenges related to green land use, thereby advancing sustainable development. Additionally, collaborative research projects should be conducted and data and information resources shared, while also emphasizing technological innovation and application.

5.3. Limitations and Prospects

Nevertheless, this study has some limitations that require further exploration. Primarily, due to the reliability and availability of data, this study solely examined Chinese urban panel data from 2007 to 2019 without taking into account the longer-term temporal trends. Secondly, when it comes to gauging ULGUE, the measurement outcomes may vary depending on the different methodologies and indicators employed. It is also worthwhile to consider whether disparities will arise in assessing the implementation effectiveness of LCCPP. In future research, we shall endeavor to delve into additional potential data that can be subjected to analysis, thus enhancing the measurement of land green utilization efficiency in scholarly investigations. Furthermore, with respect to the global COVID-19 outbreak in 2020, we shall strive to examine whether public health events such as COVID-19 may alter the impact of LCCPP on ULGUE, thereby facilitating a more comprehensive comprehension of the effects of said policies. This will ensure that the research findings possess heightened practical significance and serve as a reliable reference for the promotion of sustainable development in global cities and the enhancement of land green utilization efficiency.