Green Fiscal Policy and Urban Land Green Utilization Efficiency

Abstract

China is at a critical stage in addressing environmental issues and achieving sustainable development. To effectively promote environmental protection and the construction of an ecological civilization, government policy support has become essential. Accurately assessing the impact of green fiscal policy (GFP) on urban land green utilization efficiency (LGUE) is significant for achieving a “win–win” situation between economic development and environmental protection and for implementing a green and low-carbon path. This study evaluates the LGUE improvement effect of GFP using the DID method with the panel data of 270 cities in China from 2008 to 2019. We find that the GFP significantly improves the urban LGUE level. The mechanism analysis shows that the GFP improves LGUE through green technological innovation, industrial structure upgrading, and environmental regulation intensity. The heterogeneity analysis shows that the LGUE improvement effect of GFP is stronger for eastern cities and non-resource-based cities. This study has important theoretical and practical significance for coordinating green fiscal policies and deepening the promotion of the sustainable use of land resources.

1. Introduction

Urban land is the most primary foundation that sustains urban production, daily life, development, public services, and ecological protection. The green and low-carbon utilization of urban land resources and assets can induce fundamental transformations in the operation and development of urban systems, thereby providing robust support for the achievement of high-quality, sustainable development within regional urban socio-economic contexts [1,2]. With the continuous progress of China’s green development goals, government investment in pollution control and the intensity of regulations have been increasing annually [3,4]. As a result, the green utilization of urban land has become a key focus in the daily lives of residents, business management practices, government policy formulation, and the optimization of related systems [1]. The urban land area in China has grown from 6720 square kilometers in 1980 to 58,355 square kilometers in 2020. However, the supply–demand imbalance between urban spatial expansion and the limited availability of land resources constrains the potential for urban development [1]. The elasticity coefficient of urban expansion has reached 2.17, surpassing the ideal value of 1.12 [5]. The large-scale investment and high uncertainty of returns in the early stages of urban economic development have resulted in a relatively low level, small scale, and weak competitive strength of regional green and low-carbon industries [6]. This series of issues has been accumulating over time, resulting in urban land and soil pollution, resource depletion, and ecological damage, ultimately reflecting the low level of urban green utilization [1]. These factors have become a significant obstacle to the sustainable socio-economic development. A model of urban land green and efficient utilization, focused on material input savings, a reduction in environmental pollution, and enhanced comprehensive benefits, has become a crucial approach to alleviating the complex contradictions between humans and nature and accelerating the transformation of socio-economic structures [2].

China’s rapid industrialization over recent decades has contributed to increased air pollution emissions, particularly sulfur dioxide, leading to its status as one of the world’s largest emitters during certain periods [7]. As environmental pressures and resource constraints continue to intensify, improving resource utilization efficiency and unlocking the benefits of transformation through green development have become an inevitable choice for achieving the carbon neutrality goal [8]. This approach also serves as an important pathway for promoting high-quality development. As a foundational and crucial pillar of national governance, the fiscal sector plays a vital role in effectively reducing pollution emissions and addressing market failures through fiscal measures [3]. Green fiscal policy (GFP) can adjust fiscal funds and resource allocation to alleviate local governments’ financial pressures, incentivizing them to actively participate in green construction and achieve the low-carbon and green transformation of the economy and society [4]. Furthermore, GFP can promote regional green transformation by establishing environmental performance evaluation mechanisms and strengthening local governments’ efforts in ecological and environmental governance [5]. GFP can effectively solve the problem of resource dependence in resource-based cities. GFP can encourage enterprises to invest in new energy and energy-saving and environmental protection industries through tax incentives, promoting the diversified development of industries in resource-based cities. For the destruction of the ecological environment, financial subsidies can support pollution control and ecological restoration projects, reducing the environmental pressure on resource-based cities. Meanwhile, government procurement of green products and services can stimulate market demand and promote the development of green industries in resource-based cities [3,4,5]. Launched in 2011 by the Ministry of Finance and other relevant departments, the National Comprehensive Demonstration City of Energy Saving and Emission Reduction Fiscal Policy (NESERFP) is a typical green fiscal policy that combines fiscal incentives with target constraints. The selected demonstration cities not only receive central government reward funds but also benefit from provincial-level matching funds. Additionally, existing emission-reduction policies from the central government are prioritized for these demonstration cities. Therefore, NESERFP represents an important practice by the central government to encourage local governments’ proactive engagement in environmental governance. The policy effects of NESERFP have already attracted attention from academia. Scholars have found that NESERFP has facilitated the transformation of enterprises [6], and promoted urban pollution reduction and carbon reduction [5]. Moreover, in the context of environmental governance, NESERFP is beneficial for promoting urban sustainable development, although its effects only became apparent after three years of implementation [7].

It is crucial for fiscal policy to better leverage its role in guiding pollution reduction and emission control to improve the land green utilization efficiency (LGUE). Coordinating and aligning GFP with the efficient and green utilization of urban land has become a key measure in achieving China’s modernization goals [9]. As the main drivers of regional environmental governance, local governments play a central role in improving urban LGUE through the design of appropriate fiscal incentive policies, making the optimization of fiscal incentive systems a key direction for fiscal policy reform. However, existing literature mainly focuses on the environmental effects of fiscal incentives from the perspectives of fiscal decentralization and fiscal pressure, with relatively few studies analyzing the role of direct fiscal rewards from the central government to local governments. Additionally, there is a lack of research evaluating the impact of such policies on urban LGUE. The specific pathways through which green fiscal policies influence urban LGUE and the heterogeneity of these effects remain areas that warrant further exploration. Currently, China’s economy is undergoing a period of structural adjustment and transformation, while the construction of an ecological civilization faces mounting pressure. Economic development continuously impacts land resource allocation and utilization, and, conversely, land resource allocation and utilization play a critical role in the context of social development. Therefore, conducting research on the impact of GFP on urban LGUE not only enables land managers to adjust and refine land allocation strategies and methods in a timely manner, but also supports the implementation of green development principles in natural resource management.

This paper takes China’s NESERFP, implemented in 2011, as a case study and employs a multi-period difference-in-differences (DID) model to systematically explore the impact of GFP on urban LGUE. The contributions of this research are as follows: (1) Research Perspective: This study contributes to the existing literature on the environmental effects of GFP. While some studies have analyzed the policy effects of NESERFP, none have evaluated its impact on urban LGUE. By integrating GFP and urban LGUE into a unified research framework, this paper enriches the body of knowledge in this field. (2) Identification Strategy: Using NESERFP as an exogenous policy shock, this study systematically identifies the causal effects of GFP on the improvement of urban LGUE based on the DID model. The DID method effectively addresses variable errors in the explanatory variables by comparing the differences between the experimental and control groups before and after the policy implementation. This method effectively addresses the endogeneity issues highlighted in the existing literature and offers new policy insights for improving urban LGUE in China. (3) Research Content: This paper examines three key pathways through which GFP influences urban LGUE: environmental regulation (ER), industrial structure upgrading (ISU), and green technology innovation (GTI). Additionally, it conducts a heterogeneity analysis based on urban resource type and geographic location. This analysis enables governments to summarize in a timely manner the positive experiences from demonstration cities, maximizes the incentive and constraint effects of green fiscal policies, and develop scalable, replicable models that can inform future green policy formulation.

2. Materials and Methods

2.1. Literature Review

2.1.1. Research on Urban Land Green Utilization Efficiency

The LGUE is a key indicator for evaluating whether the input–output relationship of urban land use is effectively optimized and aligned with the principles of green development. It serves as a crucial reference for land management policies and has become a significant topic in academic discussions. This paper primarily reviews the concept and influencing factors of urban land LGUE. Scholars have introduced the concept of land use efficiency to assess the input–output structure of land use, comprehensively considering the costs and trade-offs involved to achieve land use benefits. Land use efficiency can be divided into structural efficiency and economic efficiency. The former refers to the benefits derived from the current land use structure and spatial layout [8], while the latter concerns the returns from various inputs in land use activities [9]. Given the growing demands for social welfare improvement and ecosystem protection, research on land use efficiency has expanded in both depth and scope, extending the narrowly defined economic efficiency of land use to encompass social and ecological outputs, forming the concept of LGUE. LGUE integrates the idea of green development into the notion of land use efficiency, representing a product of the social and economic transformation as well as conceptual advancements in the new era [10,11,12,13].

In evaluating LGUE, scholars typically employ comprehensive indicator-based evaluation methods to assess both the ecological benefits and costs of land use. The indicator system often includes desired ecological outputs, such as green coverage rate of built-up areas, as well as undesired pollution outputs, such as industrial pollution emissions and carbon emissions from construction land [11,12,13]. Based on the results of urban LGUE evaluations, scholars further explore the dynamic evolution patterns of urban LGUE. Methods such as kernel density estimation and descriptive statistics are employed to analyze the temporal characteristics of urban LGUE. For instance, at the national level, Zhou and Lu (2023) examined the temporal characteristics of urban LGUE in China, concluding that urban LGUE in China follows a fluctuating upward trend, with considerable room for improvement at the overall level [14]. At the regional level, Li et al. (2023) and Tan et al. (2021), respectively, examined the evolutionary patterns of urban LGUE in the Yellow River Basin and the Yangtze River Delta [15,16]. Koroso et al. (2021) [17] used ArcGIS to assess the urban LGUE of 16 cities in Ethiopia based on remote sensing data. The results indicated that all cities generally face the issue of low LGUE. Koroso and Zevenbergen (2024) found that the inefficient land transfer market, insecure land tenure, and land hoarding in Ethiopia have contributed to lower LGUE in cities, with gaps in policy formulation and implementation further exacerbating the issue [18].

Many scholars have further explored the driving factors and underlying mechanisms behind the evolution of LGUE, which can generally be categorized into natural, economic, social, and governmental factors. From a natural perspective, Chen et al. (2023) argue that urban ecological resources, such as urban landscapes, green space areas, and forest cover, contribute to improving LGUE [12]. Li et al. (2023) suggest that pollution emissions negatively affect urban LGUE, with their impact intensifying over time [19]. From an economic perspective, existing research extensively explores the complex impacts and mechanisms of such factors on urban LGUE. For instance, Zhou and Lu (2023) contend that economic development, urban construction, industrial transformation, market openness, and foreign investment all play significant roles in the evolution of urban LGUE [14]. Furthermore, the influence of economic factors often exhibits spatial spillover effects and regional heterogeneity. Zhao et al. (2024) highlight that industrial intelligence helps reduce disparities in urban LGUE, emphasizing the nonlinear characteristics of this innovation’s role [20]. From a social perspective, Lu et al. (2023) [21] have deeply examined the impact of compact urban transportation development on LGUE. They found that factors such as the construction of transportation infrastructure, accessibility, and the efficiency of public transportation systems all have positive effects on urban LGUE. From a governmental perspective, some scholars emphasize the significant influence of government actions on urban LGUE, particularly the roles of government constraints and regulations [22]. Zhou et al. (2024) found that China’s energy rights trading policy could enhance urban LGUE by optimizing resource allocation [23]. Additionally, Wang et al. (2025) observed that China’s green finance policies, through financial incentives, strengthen local governments’ environmental governance actions, thus affecting urban land LGUE [24].

2.1.2. Research on Green Fiscal Policy

Theoretically, GFP can address the pollution costs of enterprises through punitive environmental taxes and fees, thereby internalizing the external environmental costs of pollutant-emitting entities. Additionally, the government can constrain enterprises’ pollution behavior through direct investments in environmental governance, encouraging companies to improve green production technologies. Research by Gramkow and Anger (2018) demonstrated that, from 2001 to 2008, green fiscal measures were highly successful in Brazil’s manufacturing sector, effectively promoting green innovation. Furthermore, green innovation technologies played a crucial role in facilitating the transition of enterprises toward a green economy [25]. Gramkow (2020) [26] argued that one of the root causes of environmental and resource use issues is the lack of proper pricing and improper subsidies. The adjustment process of taxes and subsidies is slow, while, at the same time, if the pricing is correct, numerous green economic policies would be more effective. O’Callaghan et al. (2022) developed a machine-learning algorithm to qualitatively assess the literature and found that, compared to non-green investments, green investments can create more jobs and bring higher fiscal multipliers [27]. For policymakers, increasing green spending during recovery periods can foster economic revival. Fang et al. (2022) [28] explored the mechanisms and feasibility of green fiscal expenditure in addressing fiscal imbalances and promoting green economic growth. Their results indicated that green fiscal expenditure contributed to the expansion of green economic growth in BRICS countries. Yuan et al. (2022) [29] found that green fiscal expenditure enhanced the effectiveness of emission reduction and produced spatial–temporal effects. In the short term, technological spending is more effective in reducing carbon emissions than environmental protection spending; in the long term, environmental protection expenditure consistently suppresses carbon emissions, and its emission reduction effects are sustainable.

In recent years, an increasing body of literature has examined the environmental effects of China’s GFP using causal inference methods, particularly focusing on the NESERFP introduced in China. For example, Lin and Zhu (2019) evaluated the pollution reduction and sustainable development effects of the NESERFP, revealing that the implementation of NESERFP not only reduced pollution emissions in demonstration cities but also enhanced regional ecological efficiency [7,30]. However, some scholars hold opposing views, arguing that environmental tax policies have not been effective in providing adequate incentives in practice [31,32]. Specifically, in some developing countries, issues such as regional economic imbalances, local protectionism, and the imperfections in environmental fiscal systems have led to widespread skepticism about the practical effectiveness of GFP [33,34,35,36]. Currently, there is no literature that directly examines the relationship between GFP and green land use, which serves as the foundation of this study.

In summary, the existing literature offers a solid foundation for studying the relationship between GFP and LGUE, as shown in

Table 1. Literature review.

Research Topic	Main Research Content	Main Literature
Research on urban LGUE	The connotation and definition of LGUE	[8,9,10,11,12,13]
	The measurement indicators and methods for urban LGUE	[10,11,14,15,16,17]
	The factors influencing urban LGUE	[12,14,20,21,22,23,24]
Research on GFP	The economic effects of GFP	[17,18,19,20,21,22,23,24,25,26]
	The environmental effects of GFP	[25,28,29]
	The policy effect evaluation of NESERFP	[7,30]

below. The existing methods for measuring LGUE and GFP are well-established and provide a solid reference for this study [11,12,13,30]. Additionally, previous research on the effects of GFP and the influencing factors of LGUE offers valuable literature support for integrating GFP and LGUE into a unified analytical framework [22,23,24,28,29]. However, existing literature has several limitations. In terms of research content, studies mainly focus on policies like carbon trading and low-carbon cities, while research on the impact of GFP on urban LGUE remains scarce. Additionally, while some studies evaluate the effectiveness of NESERFP implementation, none have directly examined its impact on urban LGUE. From a methodological perspective, due to the reverse causality between environmental fiscal expenditure and environmental quality, existing literature generally faces endogeneity problems [26,27,28,29], underscoring the need for further investigation into the effectiveness of GFP. We employ the DID method to examine the relationship between GFP and LGUE, which effectively addresses endogeneity issues during the empirical analysis and ensures the robustness of the results. Our research further enriches the study of the factors influencing LGUE and the environmental benefits of GFP.

2.2. Research Hypothesis

2.2.1. Direct Impact

The demonstration period for the NESERFP in pilot cities lasts three years. During this period, the central government provides financial incentives for typical demonstration projects submitted by the pilot cities. The reward funding is tiered based on the nature of the demonstration city, with direct-controlled municipalities and urban clusters, separately planned cities, and provincial capitals receiving annual rewards of 600 million, 500 million, and 400 million yuan, respectively. The pilot cities have discretion over fund allocation. Therefore, the NESERFP is expected to effectively utilize fiscal incentives to optimize local governments’ environmental governance behaviors, encouraging them to increase environmental management efforts and green investment, thereby improving urban LGUE.

Firstly, the NESERFP ensures sufficient and targeted fiscal resources, improving the efficiency of government fiscal allocation and providing strong public goods support for regional innovation activities. Secondly, through demonstration projects, the NESERFP can leverage substantial local matching funds, encouraging local governments to increase their investments in environmental protection. This, in turn, attracts capital from society and enterprises, creating a strong momentum for energy conservation and emission reduction, thereby reducing undesirable land use outcomes. Moreover, the NESERFP enhances clean investment and strengthens the penalties for corporate pollution, directing fiscal resources towards energy conservation, emission reduction, and technological innovation [37]. Incentive funds not only motivate existing enterprises to engage in environmental governance but also send a strong policy signal to new enterprises, encouraging them to adopt clean and green technologies and projects [38]. This strengthens market competition in heavily polluting industries, and, through competition, accelerates the green transformation of enterprises, ultimately reducing land pollution. In conclusion, GFP is expected to significantly improve urban LGUE.

Therefore, this paper proposes the following hypothesis:

H1: 

Green fiscal policy has a significant impact on urban land green utilization efficiency.

2.2.2. Impact Mechanisms

(1)

Industrial Structure Upgrading Effect

One of the key objectives at the outset of the NESERFP is to achieve the ISU. The demonstration cities set specific emission-reduction targets, which were allocated to high-energy-consuming and high-emission industries, with mandatory interventions. This significantly raised their environmental compliance costs and survival thresholds, forcing high-pollution industries to eliminate outdated capacity or relocate [39]. Additionally, the central government’s comprehensive reward funds and provincial-level special funds directly increased the financial resources of demonstration cities, providing essential support for the growth and development of green industries. This, in turn, facilitated the development of high-tech industries focused on low-carbon environmental protection, ultimately driving regional industrial transformation.

Optimizing the industrial structure is a key pathway for improving urban LGUE. Existing research consistently demonstrates the positive impact of ISU on urban LGUE [23,24]. During the process of industrial structure upgrading, enterprises typically implement technological upgrades and industrial transformation to improve production efficiency and resource utilization. This involves adopting advanced production processes and equipment, increasing resource recycling rates, and reducing waste emissions. These measures not only enhance economic benefits for enterprises but also reduce land pollution and degradation, leading to greener and more efficient land use. Therefore, the GFP is expected to stimulate industrial upgrading in cities, ultimately improving urban LGUE.

(2)

Green Technology Innovation Effect

GTI refers to innovative activities aimed at promoting clean production, energy conservation, emission reduction, and the use of renewable energy, as well as reducing the use of raw materials and energy, and lowering pollution emissions from the outset. GTI combines both environmental governance and technological spillover effects and serving as a crucial long-term strategy for energy conservation and reduced resource consumption across various production sectors. From the perspective of public finance, green technologies often have long development cycles, high risks, and uncertainties, requiring substantial research and development investments at the initial stages. As a result, enterprises are often reluctant to engage in GTI, and market mechanisms alone are insufficient to incentivize such activities. Appropriate policy guidance and financial support are essential [40]. Stable fiscal policies play a key role in reducing business operating risks and innovation investment risks. Continuous environmental fiscal expenditure can significantly offset the costs of enterprise technology research and development, lower the risks associated with green innovation, and provide targeted improvements for environmental issues through strictly supervised funds. This creates a stronger incentive for green innovation activities. The NESERFP, which combines fiscal incentives with project governance, can effectively enhance local governments’ incentive and constraint behaviors, ensuring rigorous policy implementation and fostering improvements in GTI [41].

Green technologies, such as clean production technologies and resource recycling technologies, can reduce pollution generation and emissions at their source while improving resource utilization. The application of these technologies significantly lowers energy consumption and emissions in the production process, thus enhancing LGUE [20]. Moreover, GTI enables traditional enterprises to adopt more environmentally friendly and efficient production methods, reducing land pollution and degradation. The promotion and application of green technologies also foster the development of new green industries, such as green manufacturing and green services, which typically offer higher output efficiency and lower environmental impacts. As a result, green technologies increase output efficiency per unit of land and improve the level of LGUE. Consequently, the GFP will drive the improvement of urban GTI and, ultimately, enhance urban LGUE.

(3)

Environmental Regulation Effect

GFP plays a guiding regulatory role by compelling local governments to strengthen environmental regulations, which, in turn, improves urban land green use efficiency. Unlike general fiscal policies, GFP possesses a green regulatory attribute that directs the allocation of energy-saving and emission-reduction funds, thereby exerting governance effects on environmental regulation [42]. The NESERFP stipulates that, if the performance assessment determines that a demonstration city has failed to meet the overall emission-reduction targets set in the implementation plan, the city’s demonstration qualification will be revoked. This provision strengthens the environmental regulatory efforts of local governments. In the process of implementing the NESERFP, ER serves as a crucial tool to intervene in environmental pollution behavior and mitigate environmental externalities. Under the influence of ER, enterprises are required to increase their investment in environmental management equipment and environmental protection technology talents to control pollution, thereby internalizing the external environmental costs. Stricter ER raises the environmental costs and financing barriers for enterprises, particularly for high-pollution and high-energy-consuming industries. This forces companies to intensify their efforts in introducing technology, upgrading environmental technologies, and investing in green research and development, thereby mitigating environmental risks and accelerating green transformation. When stringent ER becomes a regular constraint at the local level, enterprises will prioritize environmental protection, actively invest in pollution control technologies, and reduce the production of high-pollution products [24,43]. This contributes to improving the expected benefits of land use and reducing undesirable outputs. Therefore, the GFP can enhance urban LGUE by strengthening regional ER.

Therefore, this paper proposes the following hypothesis:

H2: 

Green fiscal policy can significantly enhance urban land green utilization efficiency by promoting industrial structure upgrading, improving green technology innovation, and strengthening environmental regulation intensity.

2.3. Research Design

2.3.1. Model

The NESERFP, which began in 2011, is an exogenous policy of LGUE in cities, so we regard it as a quasi-natural experiment. Considering that the policy is implemented in batches and drawing on existing research practices, we use a multi-period DID model for causal inference. This study uses the city-wide caliber data from the China Urban Statistical Yearbook. For missing values in individual samples, we use local statistical yearbooks or apply linear interpolation to estimate them. Due to a significant number of missing values for some core variables in certain cities, we have excluded these cities, following the approach used in existing research [1,20]. We ultimately selected 270 prefecture-level cities from 2008 to 2019 as the research sample. The benchmark model is set as follows:

$${LGUE}_{it}={\alpha }_0+{\alpha }_1{DID}_{it}+{\alpha }_2{Conl}_{it}+{City}_i+{Year}_t+{\epsilon }_{it}$$

(1)

Among them, LGUE is the dependent variable of the city’s LGUE level, and DID_it represents the virtual variable of the NESERFP. If city i is selected in year t, then DID_it is 1; otherwise, it is 0. Conl is the set of control variables (CVs), City_i is city fixed effect, and Year_t is year fixed effect. The value of α₁ reflects the impact of NESERFP on the urban LGUE. This demonstration policy is not directly aimed at the urban LGUE, but rather at promoting regional green development. The DID can largely eliminate selective biases at the policy level and the interference of corporate strategic behavior, thus identifying relatively clean policy effects. We first empirically test the direct impact of GFP on LGUE based on model (1). To ensure robustness, we then perform a series of robustness checks. This primarily includes parallel trend testing, placebo testing, addressing non-random factors in pilot cities, and excluding the influence of other competing policies. After verifying hypothesis H1, we then adjusted model (1) to test the three impact channels through which GFP affects LGUE, in order to verify hypothesis H2. Finally, we test the heterogeneous impact of GFP on LGUE by adding interaction terms based on city type and geographic location, aiming to derive competitive policy recommendations.

2.3.2. Variable

LGUE: The comprehensive evaluation of urban LGUE is the use of certain measurement method to effectively quantify the structured relationship between traditional factor inputs, expected benefit outputs, and unexpected outputs in the process of urban land use, forming a scientific indicator system that reflects the degree of greening and ecologicalization of urban land use. This study employs the unexpected output super-efficient slack-based measure (SBM) model to assess urban LGUE. This model can not only distinguish between expected output and unexpected output, but also decompose effective decision-making units (DMUs) for comprehensive comparison, thereby preventing the loss of information from effective DMUs. In addition, the super-efficient SBM model allows the efficiency value of a DMU to exceed 1, which enables further differentiation of high-efficiency DMUs. The efficiency values of DMUs in traditional DEA model can only be between 0 and 1. When the efficiency values of multiple DMUs are all 1, effective sorting and differentiation cannot be achieved. The super-efficient SBM model can clearly distinguish these high-efficiency DMUs by allowing efficiency values to exceed 1. Many studies have used the super-efficient SBM model to measure and analyze urban LGUE [13,22,23,24], providing a methodological foundation and reference for this research.

In the selection of indicators, the input indicator system should include various elements necessary for economic production activities on a unit area of land. According to the Cobb–Douglas production function and existing literature [22,23,24], this study determines land, capital, and labor as input indicators. Land input is represented by the total area of urban construction land, capital input is represented by the amount of fixed assets investment, and labor input is represented by the number of employees in the secondary and tertiary industries. In terms of expected output, this study divides indicators into three categories: economic benefits, social benefits, and ecological benefits. We use the added value of the secondary and tertiary industries to represent economic benefits, the average salary of urban employees to represent social benefits, and the green coverage rate of built-up areas to represent environmental benefits. The unexpected output is reflected by the industrial wastewater discharge, sulfur dioxide, and dust.

GFP: We use NESERFP as proxy variables for GFP [4]. We use the interaction term (Treat*Post) between the virtual variable of pilot cities and the virtual variable of policy implementation time to characterize the treatment effect of NESERFP. In the following text, we directly use DID to represent the interaction term. Specifically, Treat is the grouping variable for whether city i is selected as a pilot city for NESERFP, and Post is the processing variable for whether NESERFP is implemented during period t. According to the release and approval time of the list of comprehensive demonstration cities for NESERFP, if city i is selected as a pilot city for NESERFP during period t, the DID value is 1; otherwise, it is 0.

CV: Considering that other urban characteristic factors may have an impact on urban LGUE, drawing on relevant research [22,23,24], this study controls for the following variables: economic development (EC), opening up (OU), population density (PD), foreign direct investment (FDI), government fiscal expenditure (GFE), and financial development (FD). Their calculation methods are shown in

Table 2. Descriptive statistics.

Variable	Calculation	Mean Value	Standard Deviation	Observation
LGUE	LGUE index based on super-efficient SBM model	0.627	0.465	3240
EC	The logarithm of GDP per capita	10.201	0.597	3240
OU	The ratio of total imports and exports to GDP	0.173	0.295	3240
PD	The logarithm of the population per unit area	5.745	0.962	3240
FDI	The ratio of actual utilization of foreign capital to GDP	0.016	0.019	3240
GFE	The ratio of government fiscal expenditure to GDP	0.191	0.141	3240
FD	The ratio of total financial deposits and loans to GDP	2.313	1.011	3240

below.

3. Results

3.1. Benchmark Results

Table 3. Benchmark results.

	(1)	(2)	(3)
	LGUE	LGUE	LGUE
DID	0.167 ***	0.273 ***	0.136 ***
	(0.0521)	(0.0379)	(0.0526)
EC		0.143 ***	0.300 ***
		(0.0278)	(0.0647)
FDI		−0.0225 **	0.0291 *
		(0.0106)	(0.0174)
POP		0.105 ***	0.432 **
		(0.0153)	(0.170)
FDI		0.223 ***	0.612 ***
		(0.065)	(0.1924)
GOV		0.292 *	0.120
		(0.171)	(0.229)
FIN		0.469	0.476
		(1.164)	(1.251)
City	N	N	Y
Year	N	Y	Y
R2	0.556	0.189	0.563

Note: *, **, *** represent significance levels at 10%, 5%, 1%, respectively, with robust stand errors in parentheses.

shows the effect of GFP on urban LGUE. Among them, column (1) does not include CVs, and columns (2) and (3) introduce CV sets and fixed effects based on the previous column. The regression results show that the estimated coefficient of DID is positive in all three columns, indicating that GFP can promote the improvement of LGUE, which is consistent with the research hypothesis. Local governments have significant information advantages in regional development and governance, as well as in environmental governance. The NESERFP increases local government environmental governance funds, which is essentially a special transfer payment. Correspondingly, more supporting funds will encourage local governments to increase their efforts in environmental governance to achieve the environmental protection goals set by the central government. The incentive mechanism of NESERFP can promote local governments to actively promote environmental governance, reduce negative impacts on land, and, ultimately, improve the LGUE in cities. Our results are similar to the conclusions obtained from existing research. Existing research has found that GFP promotes urban green development and environmental governance [4,37,38]. Moreover, environmental regulations, such as green finance and energy trading systems, can contribute to the improvement in urban LGUE [23,24]. This further strengthens the robustness of the results presented in this paper and expands the scope of existing literature on GFP and LGUE.

3.2. Robust Test

3.2.1. Parallel Trend Test

If there is a significant difference in the trend of LGUE changes between pilot and non-pilot areas before policy intervention, we cannot determine whether the observed policy treatment effect is actually caused by the policy or simply due to the different trends of LGUE changes in pilot and non-pilot areas themselves. The parallel trend assumption can ensure that LGUE in pilot and non-pilot areas have the same time trend before policy intervention, thereby ensuring the true effect of the policy. We conduct a parallel trend test using the event study method [44], and the results are shown in Figure 1. The coefficient values before the policy shock are in a fluctuating state, and there is no significant pre trend. However, after the policy shock, there is a clear upward trend in the urban LGUE, indicating that the policy shock had a significant effect. Therefore, the DID model set in this paper satisfies the parallel trend test.

3.2.2. Placebo Test

Although this study has controlled for some characteristic variables of each city in the quasi-natural experiment, there may still be some urban characteristic factors that may affect the regression results. In order to further eliminate the interference of other random factors and omitted variables, this study draws on the approach of previous scholars and uses a random selection of experimental groups for placebo testing [45]. We randomly select some cities as the experimental group, and then conducted regression analysis using generated false policy treatment variable. To ensure the unbiasedness of the estimation results, 500 random samples are taken from the sample cities, and the kernel density and p-value distribution are shown in Figure 2. During the random processing, the p-values are mostly above 0.1, and the estimated coefficients are concentrated around 0. Moreover, the estimated coefficients of the real policy dummy variable in this study are significantly different from the results of the placebo test. This also indicates to some extent that the policy’s promotion of urban LGUE has not been affected by random factors.

3.2.3. Discussion on Non-Random Selection

The ideal scenario for evaluating the effect of GFP on LGUE using the DID method is that the pilot cities are randomly selected. But, in reality, the selection of pilot cities is often based on factors such as the city’s economic development, government financial capacity, and fiscal expenditure. The inherent differences in these cities will have varying impacts on the LGUE over time, leading to estimation errors. To control the impact of the above factors on the construction of demonstration cities, this study draws on the approach of Li et al. (2016) [46] and adds an interaction term between control variables and time trends on the basis of benchmark regression to alleviate the non-random selection problem of NESERFP.

Table 4. Non-random selection of demonstration cities.

	(1)	(2)
	LGUE	LGUE
DID	0.0907 *	0.0995 **
	(0.0474)	(0.0475)
CV	N	Y
City	Y	Y
Year	Y	Y
R2	0.586	0.592

Note: *, ** represent significance levels at 10%, 5%, respectively, with robust stand errors in parentheses.

reports the results of the non-random selection of demonstration cities. From the results in column (1), after adding the interaction term between control variables and time trends, the regression coefficient of DID remains positive. At the same time, compared with the benchmark results, the magnitude of the estimated coefficients is not significantly different, which undoubtedly confirms the reliability of the benchmark results.

3.2.4. Excluding the Impact of Other Policies

In addition to the NESERFP, other emission reduction policies may also have an impact on the LGUE in cities, thereby affecting the estimation results. During the sample period of this paper, in order to control and reduce carbon emissions, the central government launched a pilot program for carbon emission trading at the provincial level in 2013 and a green finance innovation pilot policy in 2017. In addition, considering that the improvement effect of LGUE through GFP may be influenced by some regional economic and social environmental factors that change over time, and these factors cannot be controlled by the fixed effects of time and city. Therefore, following the approach of Zhou and Qi (2022) [47], we incorporated the interaction term between province and year into the benchmark model to exclude other provincial policies and capture the interference that may be caused by policy factors changes between different periods and regions. After incorporating the above interaction terms, the coefficients of DID in

Table 5. Exclude interference from other policies.

	(1)	(2)
	LGUE	LGUE
DID	0.198 ***	0.174 ***
	(0.066)	(0.0543)
CV	N	Y
City	Y	Y
Year	Y	Y
R2	0.626	0.635

Note: *** represent significance levels at 1%, with robust stand errors in parentheses.

are still significant. This indicates that the LGUE promotion effect brought by the GFP has not been interfered by other policies. It also further indicates that the pilot project of the comprehensive demonstration city selected in this study is unique and relatively exogenous, and the conclusions obtained are not affected by other policy factors.

3.2.5. PSM-DID

Although the DID method separates the average treatment effect of the pilot policy, there may still be selective bias in observing and researching data due to the fact that GFP is not a strictly natural experiment. To address this issue, we employ the PSM-DID model for robustness testing. We matched samples between the experimental group and the control group based on CV. We use model (1) for DID estimation based on caliper matching and kernel matching. As shown in

Table 6. PSM-DID.

	(1)	(2)
	Kernel	Radius
	LGUE	LGUE
DID	0.101 **	0.0835 *
	(0.0515)	(0.0463)
CV	Y	Y
City	Y	Y
Year	Y	Y
R2	0.557	0.565

Note: *, ** represent significance levels at 10%, 5%, respectively, with robust stand errors in parentheses.

, the coefficients of the DID under the matching of the two methods are still significantly positive, with no substantial difference from the benchmark results. This to some extent indicates that the promotion effect of the GFP on urban LGUE is robust.

3.2.6. Other Robustness Tests

To prevent a causal relationship between urban LGUE and CVs, which may affect the estimation results, we lagged all CVs by one period. The DID result shown in the first column of

Table 7. Other robustness.

	(1)	(2)
	LGUE	LGUE
DID	0.133 **	0.186 ***
	(0.0524)	(0.0348)
CV	Y	Y
City	Y	Y
Year	Y	Y
R2	0.554	0.562

Note: **, *** represent significance levels at 5%, 1%, respectively, with robust stand errors in parentheses.

is positive. The research period of this paper is relatively long, and it may contain many potential interfering factors, which is not conducive to the accuracy of the estimation results. Based on existing research, we further reduce the sample interval and retain the samples from 2008 to 2015 as our empirical sample. Then, we estimate the impact of GFP on urban LGUE using a benchmark model, and the DID result shown in the second column is still positive.

3.3. Mechanism

The above empirical results indicate that GFP can effectively promote the LGUE in cities. Thus, what is the underlying transmission mechanism? The previous theoretical analysis shows that GFP can improve the level of GTI, promote ISU, and strengthen local ER to affect the urban LGUE. Next, we refer to existing practices to construct the following model:

$${Med}_{it}={\beta }_0+{\beta }_1{DID}_{it }+{\beta }_2{Conl}_{it}+{City}_i+{Year}_t+{\epsilon }_{it}$$

(2)

M is the mechanism variable, representing GTI, ISU, and ER, respectively. Firstly, we test whether the GFP effectively improves the level of GTI in cities based on model (2). We use the number of green innovation patent applications to measure GTI. The results in column (1) of

Table 8. Mechanism analysis.

	(1)	(2)	(3)
	GTI	ISU	ER
DID	0.12 **	0.0288 *	0.155 ***
	(0.0468)	(0.0155)	(0.0403)
CV	Y	Y	Y
City	Y	Y	Y
Year	Y	Y	Y
R2	0.961	0.884	0.368

Note: *, **, *** represent significance levels at 10%, 5%, 1%, respectively, with robust stand errors in parentheses.

indicate that the GFP has improved the level of GTI in cities at the 5% statistical level. Therefore, cities selected as comprehensive demonstration cities for GFP can significantly increase local government environmental attention, which is conducive to promoting GTI. Subsequently, we examined the impact of GFP on the ISU. We measure ISU using the ratio of the output value of the tertiary industry to that of the secondary industry, as defined in the existing literature [7,8,23,24,25]. The results of column (2) indicate that the GFP significantly promotes the urban ISU. Demonstration cities can limit the pollution behavior of high-energy-consuming industries and promote the development of strategic emerging industries, achieving low-carbon and green industrial development. Finally, we examine the impact of GFP on the intensity of ER based on model (2). We use the logarithm of the frequency of environmental protection words in the city government work report to measure the ER. The results of column (3) indicate that the GFP significantly promotes the intensity of ER in cities. The regulatory constraint effect of GFP can further strengthen the intensity of local ER and force urban enterprises to undergo green transformation. The existing research has demonstrated that GTI, ISU, and ER can significantly enhance urban LGUE [19,20]. Therefore, GFP can improve urban LGUE through three channels: GTI, ISU, and ER.

3.4. Heterogeneity

3.4.1. Resource Type

Resource-based cities are typically endowed with abundant energy and resources. During the process of industrialization and urbanization, it is common for them to develop a model reliant on regional resource endowments, which can trigger the ’resource curse’ phenomenon in the green transformation of these cities. To further explore the heterogeneous effect of GFP under different resource endowments, we categorize the sample cities into resource-based and non-resource-based cities, following the approach outlined in the existing literature [21]. We use the dummy variable Type to represent city type, assigning a value of 1 to non-resource-based cities and 0 to resource-based cities. Heterogeneity testing is performed by including the interaction term between Type and the DID variable. The result is presented in column (1) of

Table 9. Heterogeneity analysis.

	(1)	(2)
	LGUE	LGUE
DID*Type	0.0755 ***
	(0.0261)
DID*Eastern		0.0253 *
		(0.0138)
City	Y	Y
Year	Y	Y
R2	0.569	0.567

Note: *, *** represent significance levels at 10%, 1%, respectively, with robust stand errors in parentheses.

. It shows that the coefficient of the interaction term between the Type and DID variables is significantly positive, suggesting that the pilot policy has a stronger impact on improving LGUE in non-resource-based cities compared to resource-based cities. From the perspective of resource dependence theory, resource-based cities are heavily reliant on traditional resource industries and heavy chemical industries due to the influence of resource endowments and industrial foundations. This dependence leads to path dependency and lock-in effects on the region’s resources and industrial structure. Meanwhile, previous studies have found that resource-based industries, characterized by pollution and resource dependence, tend to favor non-clean technologies, which reinforces the preference for such technologies [21,22]. Excessive fiscal bias may hinder technological innovation and economic efficiency in resource-based cities, leading to a relatively smaller improvement in LGUE in these cities.

3.4.2. Geographical Position

There are varying degrees of differences in policy support and basic conditions for development among cities in different locations, and the effects of GFP may also vary from city to city. We divide the sample into eastern and central western regions based on geographical location. We use the dummy variable East to represent the geographical location of cities, with cities in the eastern region set to 1 and cities in the central western region set to 0. We conducted heterogeneity testing by adding the interaction term between East and DID variable. As shown in column (2) of Table 9, the GFP has a more significant effect on improving the LGUE in eastern cities. The possible reason is that cities located in the eastern region have more higher education and financial resources, and their resource advantages lead to the agglomeration of high-tech industries and modern service industries, promoting the process of industrial cleanliness. Additionally, the economic development level of cities in the eastern region is typically higher than that of cities in the central and western regions. This suggests that eastern cities began prioritizing quality over quantity growth earlier and may have implemented more energy-saving and emission-reduction measures. Therefore, GFP has a positive impact on improving LGUE in eastern cities. Additionally, for cities in the central and western regions, to better improve LGUE and achieve sustainable urban development, local governments must further optimize the top-level design of green fiscal policies. This will help direct government and social funds toward emerging technology industries, such as new energy and artificial intelligence. For certain traditional polluting industries, local governments can allocate dedicated financial funds to support their upgrading and improvement.

4. Discussion

Accelerating the transformation of urban land green utilization and improving the urban LGUE is not only an important measure to meet the needs of social development, but also an important means to break the constraints of resource and environmental reality. Implementing effective green fiscal policies is an important means by which to achieve carbon neutrality. This study regards the implementation of the NESERFP as a quasi-natural experiment, and uses panel data at the Chinese city level from 2008 to 2019 to investigate the impact and mechanism of GFP on LGUE. The main research conclusion is as follows: (1) GFP has significantly improved the urban LGUE. (2) Heterogeneity analysis shows that the construction of demonstration cities will exhibit differences in promoting LGUE due to differences in urban geographical location, resource types, and urban size. The policy effects are more pronounced in eastern cities and non-resource-based cities. (3) A mechanism analysis shows that the GFP affects the LGUE by enhancing GTI, accelerating ISU, and promoting ER intensity. This study offers multi-dimensional and valuable empirical evidence for the practice and research of green environmental policies in other developing countries. Developing countries can implement a series of green fiscal policies to direct funds into new energy and artificial intelligence technologies, thereby promoting the research and application of green technologies. This will help developing countries reduce their dependence on traditional high-pollution industries, optimize land resource allocation, and achieve sustainable urban development.

As a fundamental pillar of national governance, fiscal policy plays a crucial role in helping China reduce pollution. The NESERFP is of significant importance in the current context of pollution and carbon reduction, as well as improving urban LGUE. We propose the following policy recommendations:

• Continue to improve the fiscal and tax system reform and design mechanisms to encourage compatibility. On the one hand, the government should further optimize the tax-sharing arrangements between central and local governments and enhance both the general and special transfer payment systems. This will ensure that local governments are provided with sufficient financial resources to invest in environmental governance. The central government should implement a variety of incentive policies that target both local economic development and environmental governance. These policies may include fiscal rewards, subsidies, and the establishment of green environmental demonstration zones. Such measures will grant local governments greater flexibility in terms of policy implementation and financial support for environmental governance, thereby maximizing the effectiveness of environmental pollution control at the local level. On the other hand, the government should further improve the multi-dimensional performance evaluation system for officials, strengthening incentives for those involved in urban public service construction. Historically, China’s relatively lenient environmental regulatory framework has been closely tied to the design of local officials’ performance evaluation systems, where economic performance was the primary focus. However, in the context of green development, the performance evaluation criteria for local officials must be diversified. The system should place greater emphasis on areas such as environmental protection, healthcare, social security, and other public service areas. The incorporation of environmental protection as a non-negotiable criterion in the performance assessments of local officials should be reinforced to ensure that both economic output (e.g., production and tax revenue) and environmental protection are given equal importance.
• Choose industry development paths that align with local green fiscal progress and economic development. While GFP plays a crucial role in enhancing urban LGUE, the efficiency of their impact varies considerably depending on the level of green fiscal development and the stage of economic development. Therefore, local governments should not only fully leverage the opportunities provided by green fiscal policies to upgrade industrial structure, but also take into account their own green fiscal conditions and economic development stages. They should opt for industrial development paths that are tailored to local conditions. For resource-based cities and western cities, the focus of demonstration projects should be placed on the low-carbon transformation of traditional industries, rather than blindly following trends or excessively promoting emerging strategic industries.
• Enhance environmental awareness, promote ISU, and strengthen GTI. The government should avoid short-termism in the economic development process and place equal emphasis on both economic growth and pollution reduction efforts. It should effectively utilize special fiscal funds for emission-reduction initiatives to enhance local governments’ environmental awareness and regulatory strength. This approach will foster the development of high-tech industries focused on low-carbon technologies and environmental protection, thereby optimizing and upgrading the industrial structure. Moreover, the government should encourage and guide enterprises in the research and development of advanced low-carbon technologies, breaking through technological lock-in, and promoting regional green innovation.

There are certain limitations to this study, as we did not thoroughly consider the potential impact of the pre trend on the parallel trend test results when conducting the parallel trend test on the DID model. We will overcome this issue in the future through parallel trend sensitivity testing. In addition, when analyzing the three impact channels of GFP on LGUE, we did not consider the potential connections and synergistic effects between the mechanisms, which may result in our theoretical analysis framework being insufficiently rich. We will also conduct an in-depth analysis of the potential synergistic effects between the influencing mechanisms in the future.