The Drag Effect of Land Resources on New-Type Urbanization: Evidence from China’s Top 10 City Clusters

Abstract

Land resources are the basis of human production and life, and they face many challenges in the process of urbanization, such as the prominent contradiction between land supply and demand and the inefficient use of land, which in turn restricts the socio-economic development and the promotion of urbanization. This paper takes China’s ten largest urban agglomerations as its research object and constructs a land resource drag effect model based on the C-D production function. The geographical weighted regression method is used to calculate the coefficient of the land drag effect. Combining kernel density analysis and spatial autocorrelation analysis, the paper reveals the temporal and spatial evolution patterns of the drag effect and discusses the impact of land resources on new urbanization and its temporal and spatial differentiation characteristics. The study shows that during the period of 2006–2022, China’s new-type urbanization as a whole rises, but the development of each urban agglomeration has significant differences, showing a spatial pattern of “east high, west low”; the drag effect of land resources shows a decreasing trend, but regional differences are obvious, showing a distribution of “east strong, west weak”; the kernel density curve of drag effect of land shows a “right-skewed-left-skewed” change, with the overall level weakening and the degree of concentration increasing; the drag effect of land resources shows significant positive global autocorrelation, and there are spatial proximity effect and spillover effect in space. The findings provide a theoretical basis for land resource utilization and spatial development in China’s new-type urbanization process. Therefore, it is necessary to implement differentiated land resource allocation and urban planning policies according to different types of urban spatial agglomeration and to give full play to the cooperative linkage effect of urban agglomerations in reducing the drag effect of land resources.

1. Introduction

With the advancement of industrialization, urbanization has become an important path for countries to promote economic growth and improve social structure [1,2]. Since the reform and opening up, China’s urbanization process has accelerated significantly, with the proportion of permanent urban residents increasing from 17.9% in 1978 to 67% in 2024 [3]. In 2024, the permanent urban population increased by 10.83 million. The labor and consumption potential unleashed in this process has contributed significantly to China’s rapid economic development. However, the rough urban expansion model has also brought about multiple problems, such as sloppy land use [4], resource constraints [5], energy inefficiency [6], ecological degradation [7], and shortage of public services [8]. The traditional speed- and land-expansion-oriented urban development model is difficult to adapt to the current demand for high-quality development transformation, which has led to the birth of a new-type urbanization strategy that is “people-centered, resource-intensive, and eco-friendly”. In 2014, China officially released the National New-Type Urbanization Plan (2014–2020), which proposed adhering to the fundamental reality of the primary stage of socialism, following the laws of urbanization development, and pursuing a path of new-type urbanization with Chinese characteristics. Subsequently, the Chinese government successively issued the Outline of the 14th Five-Year Plan for National Economic and Social Development and the Long-Range Objectives Through the Year 2035 and the National New-Type Urbanization Plan (2021–2035), further emphasizing the need to uphold a people-oriented approach in advancing new-type urbanization with Chinese characteristics. Essentially, new-type urbanization refers to urbanization centered on people. Against this policy backdrop, promoting high-quality, sustainable, and coordinated urbanization has become the central theme of China’s urban development [9].

Land resources refer to the comprehensive manifestation of land space that supports production, living, and ecological functions, as well as its inherent natural attributes. They are an important foundation for promoting sustainable economic and social development. Land resource allocation encompasses the process of reasonably distributing and efficiently utilizing land resources among different uses and regions through market regulation and policy guidance. The aim is to improve land use efficiency, achieve optimal resource allocation, and promote coordinated regional development [10]. As a core production factor in economic and social development, the efficiency of land resource allocation is directly related to the sustainable advancement of urbanization. Traditional economic theories emphasize the role of land as a driving force in promoting production and growth, but the limited and non-renewable nature of land resources makes them a constraint to development under conditions of high-intensity development. Especially given China’s basic national conditions of a large population and limited land resources, issues such as mismatches between land supply and demand, excessive development, and low utilization efficiency are becoming increasingly prominent. Land resource constraints are becoming increasingly stringent, and the trend of land resource shortages is becoming increasingly evident, posing a key bottleneck constraining the high-quality advancement of new-type urbanization. As the positive supportive effect of land resources on urban development weakens, its negative constraining effect gradually emerges, a phenomenon that can be summarized as the “drag effect” caused by land resources [11,12]. Identifying and measuring the drag effect caused by land resources is of great practical significance in clarifying the resource constraint mechanism behind urban expansion [13].

The conceptual basis of the resource drag effect can be traced back to The Limits to Growth [14], which highlighted how environmental and resource constraints can create feedback pressures on long-term economic development. Based on this premise, Nordhaus used the production function framework to assess the limiting effect of natural resources on US economic growth [14]. Based on this concept, Bruvoll measured environmental resistance in Norway using a computable general equilibrium model [15]. Noel calculated the impact of energy on economic growth in the U.S. [16]. These studies introduced resource and environmental factors into endogenous growth models, revealing that technological progress can overcome resource constraints. Romer furthered this line of research by incorporating resource constraints into the classical Solow growth model, leading to the concept of the growth drag effect—a measure of the diminishing contribution of natural resources to sustained output expansion [17]. In recent years, this theoretical perspective has gradually gained prominence in studies exploring the nonlinear effects of resource and environmental factors on economic systems. Pan et al. used the growth damping coefficient to quantitatively assess the degree of constraints imposed by water and land resources on economic development in the Chengdu-Chongqing Economic Zone [13]. Zhao et al. used cities as the unit of analysis to determine the direct and indirect drag effects of land and energy resources on economic growth [18]. On this basis, many scholars have further explored the impact of resource and environmental constraints on urbanization. An et al. quantitatively analyzed the drag effect of water consumption on urbanization using panel data from 11 provinces in China’s Yangtze River Economic Belt [19]. Li et al. measured the drag effect of carbon emissions on the urbanization process of 30 provinces in mainland China based on the intrinsic relationship between economic growth and urban development [20]. Although existing research has provided a basic understanding of the drag effect of land resources on economic growth and urbanization, research specifically addressing the land drag effect in the context of new-type urbanization remains relatively limited. In particular, systematic analyses of the mechanisms and pathways through which the “drag effect of land” affects the process of new-type urbanization are notably lacking. Secondly, the widespread use of traditional econometric methods (especially linear regression) tends to ignore the spatial spillover effects and interactive dynamics of land use in neighboring regions. This methodological limitation increases the risk of biased parameter estimates and misinterpretation of causal mechanisms. In addition, existing research on the relationship between new-type urbanization and land resources is mostly based on the national or provincial level. This makes it difficult to reflect the characteristics of inter-regional factor flows and cross-regional linkages. In recent years, although some scholars have used the urban agglomeration scale for analysis, most of the relevant research has focused on a single urban agglomeration. It is difficult to represent the differences in the development stages, spatial structures, and policy contexts of different types of urban agglomerations, and it is also insufficient to reveal the overall pattern and regional differences of the drag effect of land resources nationwide. It is worth noting that urban agglomerations have become the main spatial carriers of China’s new-type urbanization, with intensive flows of factors such as land, population, and industry within them, and significant spatial spillover effects. Therefore, it is necessary to use urban agglomerations as the basic unit of analysis and carry out systematic comparative research on multiple urban agglomerations.

This study aims to uncover the mechanisms through which land resources constrain the development of new-type urbanization across different regions and time periods. It seeks to address core scientific questions, including whether the drag effect of land resources is widespread, how it evolves over time, and what spatial differences it exhibits. By doing so, the paper endeavors to fill the gap in existing research that lacks a systematic analysis of the land resource drag effect, thereby deepening our understanding of the interactions between resource constraints and urbanization. In view of the above reasons, this paper selects the top ten representative urban agglomerations with high development level and wide spatial distribution in China as the study area, covering the eastern coast, central and western inland and other types of development zones, which can comprehensively reflect the overall level of the drag effect of land resources in China, and also help to reveal the heterogeneity of the characteristics of the development of different regions [21,22]. In terms of methodology, we first construct a measurement expression for the drag effect of land based on the Cobb–Douglas extended production function that includes land factors. We then use geographic weighted regression (GWR) to estimate the spatial heterogeneity parameters required for the model and calculate the drag effect values of land resources for each city (or urban agglomeration) at different points in time. Subsequently, this paper uses kernel density estimation to reveal the temporal distribution evolution of deadweight effects. It also employs the global Moran index and local clustering analysis to uncover the spatial aggregation characteristics of deadweight effects. This enables the identification of the heterogeneous manifestations of deadweight effects across different types of urban agglomerations, as well as their temporal and spatial distribution characteristics.

The core contributions of this paper are as follows: First, it concretizes the resource drag theory into a multidimensional measurement framework for new-type urbanization, clarifying the transmission channels through which the drag effect of land affects the quality of urbanization. Second, it uses China’s ten representative urban agglomerations as units of comparison, filling the gap in previous analyses that tended to focus on single cities or the country as a whole. Third, by combining spatial heterogeneity parameter estimation with temporal and spatial distribution analysis, it provides systematic empirical evidence to understand the variability of land resource constraints across different regions and development stages. In addition, the research results can provide a basis for identifying constraints in urban land resource allocation, optimizing land use structures, and formulating differentiated resource control policies. It is of positive significance for improving the development strategy of new urbanization and promoting regional synergy.

2. Theoretical Framework

New growth theories generally agree that natural resources are indispensable elements to support economic growth and urbanization, and that the quantity, type and allocation efficiency of resources directly affect the evolution of economic and urban systems [13]. Under conditions of limited resources, when supply is insufficient to meet the expanding demands of production and living, regional development will be significantly constrained [14]. This view breaks through the limitations of the traditional Cobb–Douglas production function, which only considers capital and labor. Incorporating natural resources such as land into the analytical framework reveals the “resource marginal constraint” effect in the process of economic growth. In the improved production function model, economic output (Y) is jointly determined by capital (K), labor (L), and land (T):

$$Y=A{K}^{\alpha }{L}^{\theta }{T}^{\beta }$$

(1)

Among them, β > 0 reflects the positive effect of land on output formation. When capital and labor continue to accumulate but the supply of land cannot increase at the same pace [13], marginal output will decline, and economic growth will be systematically suppressed. This mechanism can be defined as the “drag effect of land” [12].

Based on this, this study regards the drag effect of land as a dynamic path dependence mechanism in the process of new-type urbanization: on the one hand, insufficient land supply and declining utilization efficiency directly inhibit the productivity of capital and labor [23]; on the other hand, it indirectly affects population agglomeration and urban functional layout through intermediary links such as industrial restructuring, public service supply [24], and spatial governance efficiency. At the same time, factors such as land finance dependence and institutional fragmentation will also form a feedback effect, further solidifying resource misallocation [25] and constructing a closed cycle of the economy–resources–city system (Figure 1).

Within this theoretical framework, the tail effect of land resources is not merely a static consumption outcome, but rather a constraint mechanism that accumulates over the long term and evolves dynamically within the economic–resource–urban system. This mechanism helps to systematically explain phenomena such as declining urbanization quality, obstacles to industrial upgrading, and intensifying ecological pressures [26,27], and lays the theoretical foundation for constructing a new sustainable urbanization theory model that incorporates land factors. Therefore, to decipher the deep impact of land resources on new-type urbanization, we should start from improving the efficiency of land use, promoting the reform of system supply, optimizing the structure of factor allocation, and so on, so as to realize the coordinated development between the economic driving function of land resources and the sustainable goal of cities.

3. Materials and Methods

3.1. Study Area

The ten major urban agglomerations are the earliest national urban agglomerations constructed in China, which are the most developed regions in China at present, and also the regions with the most prominent contradiction between people and land. The development level of their new-type urbanization is more obviously constrained by the degree of land resources. The ten major city clusters are distributed in various regions of the country, and their macro can reflect the general law of various regions in China [28]. Based on this, the ten major city clusters are selected as the research object (Figure 2). In recent years, the increase in the level of urbanization has led to the further expansion of the number and scale of cities, the driving effect of city clusters on China’s economy has increased significantly, and the influence of regional linkage development has become increasingly prominent. The top ten city clusters are areas with high population and economic density, and currently, the top ten city clusters are in the leading position in various aspects such as economy, society and culture, and the level of urbanization is more prominent.

3.2. Selection, Processing, and Data Sources of Specific Variables

The data for this study primarily comes from the China City Statistical Yearbook, China Urban Construction Statistical Yearbook, provincial statistical yearbooks, the National Economic Research Network database, and municipal statistical bulletins. Additionally, to ensure representativeness and readability, this paper uses node years for visualization in charts and spatial analysis: 2006, 2011, 2016, and 2022. On the one hand, the four years are evenly spaced in time, covering the starting point, transition, and convergence stages within the research period (2006–2022). On the other hand, they match the phased characteristics of China’s 10th to 14th Five-Year Plans, making it easier to identify the phased changes in new-type urbanization and the drag effect of land. It is important to note that all calculations and statistical tests are based on annual data from 2006 to 2022; the node years are used solely for visualization and narrative purposes to avoid cluttering the charts and reducing readability. Our tests on intermediate years indicate that their trends align with the conclusions from the node years, and do not affect the overall judgment of this paper. Special data processing methods are described as follows [29]:

(1)

Gross Domestic Product (GDP) (Y): The GDP of each city is converted into real GDP using the GDP deflator index for comparability across years.

(2)

Capital Stock (K): The fixed asset investment from 2006 to 2019 for each city is used to calculate the capital stock, employing the perpetual inventory method and adjusting for the provincial fixed-asset investment price index. The formula is as follows:

$$K_{it }=K_{it-1} \left(1-{\delta }_{it}\right)+I_{it}\leftarrow$$

(2)

K_it represents the capital stock, K_it−1 is the capital stock of the previous year; i represents the prefecture-level city; t represents the year; δ is the depreciation rate.

(3)

Total Land Resources (T): Due to the lack of comprehensive data on land resources at the city level, we use the sum of cultivated land area and urban construction land area as a proxy for total land resources used for economic output. Since data on agricultural land is incomplete, we approximate it using the cultivated land area, which reflects primary sector land resources, while the urban construction land area represents secondary and tertiary sector land resources.

(4)

Labor Force (L): Given that the drag effect model considers all variables at the city level, the total labor force is represented by the sum of rural and urban workers in each city.

(5)

Per Capita Output (y): Per capita GDP is used to represent per capita output.

(6)

New-type Urbanization Level (U): The index is composed of six dimensions: population, economy, spatial development, society, ecology, and urban–rural integration. The final urbanization level is calculated using the entropy method combined with a comprehensive indexing approach.

3.3. Calculation of the Development Level of New-Type Urbanization

3.3.1. Establishing Evaluation System

The development of new-type urbanization is a dynamic and systematic process, and accurately assessing the level of new-type urbanization is of critical importance. Based on the concept of new-type urbanization, which emphasizes “people-oriented, coordinated, green, intensive, and shared development,” existing research has mainly constructed evaluation systems from five dimensions: population, economy, land, society, and environment [9]. However, some studies have also pointed out that urban–rural integration is an important mechanism driving new-type urbanization [30]. Therefore, this paper integrates and refines previous research, draws on the comprehensive measurement index system for urbanization, follows the basic principles of comprehensiveness, operability, accessibility, quantifiability, and comparability in the selection of indicators, and constructs an indicator system based on policy guidance from six aspects: population, economy, society, space, ecology, and urban–rural integration (

Table 1. Evaluation index system and weight of new-type urbanization.

System	Subsystem	Indicators	Unit	Attribute	Weight
New-Type Urbanization	Economic Urbanization	Per capita GDP	CNY	+	0.0758
		Share of output value of secondary and tertiary industries	%	+	0.0789
		Per capita retail sales of consumer goods	CNY	+	0.0741
	Population urbanization	Proportion of urban population	%	+	0.0785
		Number of university students per 10,000 persons	pers	+	0.0722
		Urban population density	Pers/m2	+	0.0745
	Spatial urbanization	Per capita built-up area	m2	+	0.0780
		Per capita urban road area	m2	+	0.0778
	Social urbanization	Public vehicles per 10,000 inhabitants	veh	+	0.0749
		Health technicians per 10,000 people	pers	+	0.0773
	Ecological urbanization	Greening coverage rate of built-up area	%	+	0.0795
		Per capita green park area	m2	+	0.0790
	Urban-rural integration	Ratio of disposable income per capita of urban and rural residents	%	−	0.0795

).

Economic urbanization: New-type urbanization emphasizes a shift from “scale and speed” to “quality and efficiency.” The economic dimension is used to describe how industrial structure upgrading and productivity support the quality of urbanization [31]. The relationship between urbanization and economic development is highly interactive and synergistic and is a key driver of coordinated regional development. In the selection of specific indicators, in addition to the GDP per capita to measure the level of economic growth, two dimensions have been introduced, namely, the degree of optimization of industrial structure and retail sales of consumer goods per capita, in order to comprehensively reflect the vitality of the regional economy and the consumption capacity of the residents. The former reflects the direction of economic transformation and high-quality development, while the latter reflects the extent to which the urban economy responds to the population concentration effect, and together they constitute an important supporting indicator for measuring the level of economic urbanization.

Population urbanization: The core of new-type urbanization is “people-oriented” urbanization and orderly population aggregation, involving the urbanization of migrant workers. It emphasizes the central role of people in urbanization [32]. In the context of new-type urbanization, population urbanization is not only reflected in the spatial reorganization of population quantity but also in the continuous improvement of population quality. As the rural population is gradually exposed to and integrated into the urban education system, the quality of the population continues to improve. Based on this, this paper introduces the number of students enrolled in higher education per 10,000 people as a key indicator of the level of population quality, so as to comprehensively reflect the breadth and depth of population urbanization.

Spatial urbanization: Spatial dimensions are used to characterize land urbanization, urban expansion, density, and form optimization, and serve as the physical spatial foundation for supporting population and industry. Recent studies on China’s urbanization trajectory and “land urbanization” have emphasized the importance of spatial restructuring and expansion governance based on land [33]. New-type urbanization emphasizes the construction of a scientific, orderly, compact and efficient spatial development pattern and requires that the process of urban spatial expansion and population agglomeration be coordinated so as to promote the balanced evolution of the urban and rural spatial structure. In the process of spatial urbanization, the efficient allocation and utilization of land resources become the key factors for consideration. For this reason, this paper selects the built-up area and the per capita urban road area as the core indicators, reflecting the degree of expansion of urban construction land and the rationality of the infrastructure layout, respectively, so as to measure the level of urban spatial utilization intensification and the quality of development.

Social urbanization: The social dimension complements the people-oriented system and welfare provision, including healthcare and public services. The accessibility of public services and healthcare provision determines the welfare foundation for sustained population concentration [34]. Therefore, it is necessary to measure new-type urbanization from the perspective of social service provision. Continuous improvement of people’s quality of life is the core embodiment of the “people-oriented” development concept of new-type urbanization, the essence of which is to improve the regional public service system and enhance the residents’ sense of access and happiness, especially the acceptance and integration of the rural migrant population. Among the many elements of social security, the convenience of transportation and accessibility of basic medical services are the key factors affecting the quality of life of urban residents. Based on this, this paper selects the number of public transportation vehicles and the number of health technicians per 10,000 people as representative indicators to measure the level of social urbanization, reflecting the ability of urban public service provision and the degree of improvement of the social security system.

Ecological urbanization: New-type urbanization elevates “ecological civilization” to one of its core tasks, emphasizing harmony between humans and nature and green transformation. Recent studies have shown that the ecological dimension is both a goal and a constraint [35]. Therefore, the evaluation system must incorporate an independent ecological dimension. The construction of ecological civilization is an important support for the sustainable development of new-type urbanization, emphasizing the simultaneous promotion of ecological environmental protection in the process of urban expansion and economic growth. Ecological civilization requires that the quality of the living environment be improved in tandem with expansion and growth. In order to scientifically assess the quality of the urban ecological environment, this paper selects two indicators, park green space area and green coverage rate of built-up area, which represent the absolute supply of ecological space and the relative density of green coverage, respectively, and can comprehensively reflect the level of urban ecological livability and the effectiveness of environmentally friendly development.

Urban-rural integration: This refers to breaking down the barriers between urban and rural areas, achieving the free flow of factors and optimal allocation of resources, thereby promoting coordinated development and integration of the economy, society, and space. This process is not only reflected in the reshaping of spatial patterns but also in the sharing of economic and social resources and the balanced distribution of development results. According to China’s New Urbanization Plan (2014–2020) and the State Council’s Opinions on Promoting Integrated Urban–Rural Development, narrowing the urban-rural income gap is considered a core economic dimension of the integration process. Therefore, this study adopts the ratio of per capita disposable income between urban and rural residents as the primary indicator, where a higher ratio indicates a larger income gap and lower integration levels. This indicator is treated as a non-positive variable during standardization, ensuring that higher values correspond to lower integration levels. Although urban–rural integration is inherently multidimensional, due to limitations in the availability and consistency of multidimensional data, this study selected the ratio of per capita disposable income between urban and rural residents as the core representative indicator to highlight the degree of integration at the economic level.

The above six dimensions are consistent with China’s new urbanization policy objectives and can characterize the “new” features from multiple levels, including elements, structure, function, and performance. The specific indicators under each dimension are selected based on availability, operability, and quantifiability, and objective weights are determined using the entropy method to reduce subjective weighting bias.

3.3.2. Data Processing

To ensure the objectivity of the evaluation results, the panel data entropy method was used to process the indicator weights, and the comprehensive indicator evaluation method was used to calculate the comprehensive indicators. The attributes of the multi-dimensional evaluation indicators are significantly different. In order to unify the comparison scale and improve the accuracy of the evaluation, a dimensionless method was used to standardize the positive and negative indicator data:

$$x_{ij}^{\prime }=\left\{\begin{array}{c}{\displaystyle \frac{x_{ij}-\min \left(x_j\right)}{\max \left(x_j\right)-\min \left(x_j\right)}},\\ {\displaystyle \frac{\max \left(x_j\right)-x_{ij}}{\max \left(x_j\right)-\min \left(x_j\right)}}, \end{array}\right.$$

(3)

where i is the year, j is the indicator, $x_{ij}^{\prime }$ is the standardized value, and $x_{ij}$ is the original value. After standardization, the entropy method is used to calculate the information entropy and information utility value of the indicator, and then the weight of the indicator is calculated [36]. The calculation formula is as follows:

$$p_{ij}={\displaystyle \frac{x_{ij}^{\prime }}{{\sum }_{i=1}^N x_{ij}^{\prime }}}$$

(4)

$$e_j=-{\displaystyle \frac{1}{\ln N}}\sum _{i=1}^N p_{ij}\ln p_{ij}$$

(5)

$$d_j=1-e_j$$

(6)

$$w_j={\displaystyle \frac{d_j}{{\sum }_{j=1}^M d_j}}$$

(7)

Among these, $p_{ij}$ is the proportion of the jth indicator in the ith year, $e_j$ represents the information entropy of the jth evaluation indicator, $d_j$ denotes the redundancy of the entropy value of the jth indicator, $w_j$ represents the weight of the jth evaluation indicator (0 ≤ $w_j$ 1), and n represents the number of research units. The weighting results for various indicators are shown in Table 1.

Based on the values and corresponding weights of each indicator within the assessment framework, the comprehensive index weighting method is used to calculate the level of new urbanization development $U_i$. The calculation formula is as follows:

$$U_i=\sum _{j=1}^M w_j{x}_{ij}^{\prime }$$

(8)

3.4. Constructing the Growth Drag of the Land Model

In this study, the land element is added to the production function and simplified to derive a model of the drag effect of land resources in economic growth, and then, based on the semi-logarithmic relationship between the economy and urbanization, the model is made to be connected to the first and the last, and a model of the drag effect of land resources in the process of new-type urbanization is derived [12]. (New-type urbanization is the same as urbanization in essence, so this study follows the formula of the function of the semi-logarithmic relationship between urbanization and the economy.)

$$\left\{\begin{array}{c}Y\left(t\right)=K(t{)}^{\alpha } T(t{)}^{\beta } \left[A\right(t) L(t){]}^{1-\alpha -\beta }\\ A>0, \alpha >0, \beta >0, \alpha +\beta <1 \end{array}\right.$$

(9)

Specifically, let Y(t) represent the actual output, K(t) the capital stock, T(t) the input of land resources, L(t) the employed labor force, and A(t) the effectiveness of labor. The term AL thus denotes effective labor. The parameters α and β represent the output elasticities of capital and land, respectively. Given the difficulty of accurately quantifying technological progress, this study excludes the impact of technical change from the model. Furthermore, since land is a fixed and non-increasing resource, we assume that $\dot{T}\left(t\right)=0$. According to the Solow model, it also follows that:

$$\begin{array}{c}\dot{K}\left(t\right) =sY\left(t\right) -\delta K\left(t\right), \dot{ L}\left(t\right) = nL\left(t\right),\\ \dot{A}\left(t\right) = gA\left(t\right)\end{array}$$

(10)

Let s denote the marginal savings rate, n the population growth rate, and δ the depreciation rate of capital. Based on the above formulation, we observe that Y(t) increases with K(t), implying the existence of a balanced growth path in the process of economic development. Taking the natural logarithm of both sides of Equation (9) and differentiating with respect to time, we obtain that the time derivative of the logarithm of each variable corresponds to its growth rate. According to the model assumptions and Equation (10), the growth rates of T, A, and L are 0, g, and n, respectively. Under the balanced growth path assumption, the growth rate of economic output equals the growth rate of capital, $g_Y\left(t\right) = g_K\left(t\right)$. Based on this, the growth trajectory of Y(t) can be derived. Accordingly, the growth rate of output per unit of labor under the balanced growth path is given by:

$$g_Y^{bgp} = {\displaystyle \frac{\left(1-\alpha -\beta \right) \left(n + g\right)}{1-\alpha }}$$

(11)

The analysis indicates that economic growth is inherently constrained by the non-increasing nature of land resources. To quantitatively assess the extent of this limitation, we introduce a modified assumption in which the growth rate of land resources is represented by n, reflecting the degree to which land expands in proportion to population growth. Specifically, the original assumption is adjusted to $T\left(t\right) = nT\left(t\right).$ Under this condition, the model can be reformulated, yielding the following expression for the growth rate of output per unit of labor along the balanced growth path:

$$g_{Y/L}^{-bgp}\left(t\right) = {\displaystyle \frac{\left(1-\alpha -\beta \right) g}{1-\alpha }}$$

(12)

The growth rate of output per unit of labor along the balanced growth path is calculated under two different assumptions. The difference between the two rates represents the drag effect of land resources on economic growth:

$${drag}^y = g_{Y/L}^{\sim bgp}(t) -{ g}_{Y/L}^{bgp}(t) = {\displaystyle \frac{\beta n}{1-\alpha }}$$

(13)

According to previous studies, there exists a clear semi-logarithmic relationship between the level of urbanization and economic growth in a given country or region [37]:

$$u = a + b\ln y + \epsilon (a<0, b<0)$$

(14)

where u denotes the level of urbanization and y represents per capita output. To align this formulation with the extended production model described in Equation (13), we define $a =-{\displaystyle \frac{\ln q}{\lambda }}, b = {\displaystyle \frac{1}{\lambda }}$. Substituting these into Equation (14) and simplifying yields:

$$\dot{u }= {\displaystyle \frac{1}{\lambda }}\dot{y}$$

(15)

Let $\dot{u}$ denote the growth rate of urbanization, $\dot{y}$ the growth rate of per capita output, and $\dot{\lambda }$ the elasticity of urbanization with respect to per capita output. By incorporating Equation (15) into the balanced growth path framework, the drag effect of land resources in the process of new-type urbanization can be expressed as:

$$dra{g}^u = {\displaystyle \frac{\beta n}{\left(1-\alpha \right)\lambda }}$$

(16)

From the above equation, it can be seen that the degree of constraint of land resources on new-type urbanization is directly proportional to the elasticity of land resources $\beta$, the population growth rate n and the elasticity of capital $\alpha$, and inversely proportional to the elasticity of the per capita output of new-type urbanization $\lambda$, so that when the process of new-type urbanization relies too much on the inputs from the land instead of the technological progress, the growth rate of the new-type urbanization will be reduced.

3.5. Calculation of Growth Drag of Land Model Coefficients

The Geographically Weighted Regression (GWR) model, an important tool for ad-dressing spatial heterogeneity, represents a significant improvement and extension of the traditional Ordinary Least Squares (OLS) model. By embedding spatial location in-formation into the regression parameters, GWR employs a locally weighted least squares method to estimate parameters at each observation point. This enables the relationships between explanatory and dependent variables to vary across geographic space, thereby capturing spatial non-stationarity more effectively. The model allows for a detailed exploration of the spatial variation and differentiation of regression coefficients, providing a highly intuitive means of identifying spatially varying relationships within the data.

In this study, all variables are logarithmically transformed to mitigate the effects of heteroscedasticity. First, we perform regression analysis on the extended production function to estimate the parameters α and β in the land resource drag effect equation. Subsequently, we regress the semi-logarithmic model of new-type urbanization and economic growth on the elasticity coefficient λ. Based on these estimates, the final value of the drag effect of land resources in the new-type urbanization process is calculated.

(1)

Regression analysis of the land resource drag effect model in economic growth.

$$\left\{\begin{array}{c}\mathrm{l}\mathrm{n}{Y}_{it}=\vartheta \left({\mu }_i,{\nu }_i\right)+\alpha \left({\mu }_i,{\nu }_i\right)\mathrm{l}\mathrm{n}{K}_{\mathrm{it}}+n\left({\mu }_i,{\nu }_i\right)\mathrm{l}\mathrm{n}{L}_{\mathrm{it}}+\beta \left({\mu }_i,{\nu }_i\right)\mathrm{l}\mathrm{n}{T}_{it}+{ \epsilon }_{it}\\ i=1,2,\dots , 137; t=2006, 2007, \dots , 2022\hspace{0.25em}\end{array}\right.$$

(17)

where $Y_{it}$ represents the economic output of a city; α is the constant term; $\left({\mu }_i,{\nu }_i\right)$ denote the geographical coordinates (longitude and latitude) of city i; α, β, and γ are the geographically varying elasticity coefficients corresponding to capital, labor, and land inputs, respectively; $K_{it}$ is the capital input; $L_{it}$ is the labor input; $T_{it}$ is the land input; i indexes each prefecture-level city; ${ \epsilon }_{it}$ is the stochastic error term; and $t$ refers to the time period under analysis.

(2)

Regression analysis of urbanization and economic growth model.

According to the drag effect model, in order to reflect the sensitivity of changes in the level of new-type urbanization to changes in per capita economic output, this study uses the GWR model to estimate the elasticity coefficient λ of the level of new-type urbanization and per capita output value for each city and each year. The formula of this model is as follows:

$$\left\{\begin{array}{c}\mathrm{l}\mathrm{n}{y}_{it} = {\beta }_0^{\prime }\left({\mu }_i,{\nu }_i\right) + \lambda \left({\mu }_i,{\nu }_i\right)\mathrm{l}\mathrm{n}{U}_{it} + {\epsilon }_{it}\\ i=1, 2, \dots , 137; t = 2006, 2007, \dots , 2022 \hspace{0.25em}\end{array}\right.$$

(18)

3.6. Time-Series Dynamic Evolution Estimation of Land Drag Effect

Kernel density estimation is a nonparametric method for estimating the probability density function, which has the advantage of not requiring any parametric modeling assumptions and being able to characterize the distribution pattern and evolution of random variables with continuous density curves. In this study, the nonparametric kernel density estimation method was used to analyze the time-series evolution characteristics of the study area in terms of the intensity, variability, and dispersion of the drag effect of land resources. The method uses the probability density function to represent the data to be analyzed by continuous density curves, which can describe the distributional and evolutionary characteristics of the variables without constructing any parametric model.

3.7. Analysis of Spatial Evolution Trend of Land Drag Effect

This study utilized ArcGIS 10.8 software for spatial analysis and map creation. The global Moran’s I index was calculated using the Spatial Autocorrelation (Global Moran’s I) tool, while the local Moran’s I index and spatial clustering types (HH, LL, HL, LH) were identified using the Cluster and Outlier Analysis (Anselin Local Moran’s I) tool. The spatial distribution of urban tail effects was visualized using ArcMap layer symbolization and layout functions.

This paper adopts the spatial autocorrelation method in exploratory spatial data analysis, which is mainly through the setting of the spatial weight matrix and embedding the geographic location information relationship into the data analysis. The method can identify the spatial correlation characteristics of the research object from the global and local perspectives. Moran’s I index is used to identify the global spatial pattern characteristics of drag effect of land in new-type urbanization, and its value ranges from [−1, 1]; if it is greater than 0, it means positive spatial correlation, if it is less than 0, it means negative spatial correlation, and if it is equal to 0, it means that the attribute data are randomly distributed in space. The local Moran’s I index was used to identify the characteristics of the local spatial pattern of the drag effect of land resources in new-type urbanization, which was specifically classified into four types of agglomeration, namely, HH (High-High), LH (Low-High), LL (Low-Low) and HL (High-Low).

4. Results

4.1. Analysis on the Level of New-Type Urbanization

This study measures the development level of new-type urbanization from six dimensions: economic, demographic, social, spatial, ecological and urban–rural integration. According to the results, the new-type urbanization development level is divided into five levels: Low level (u ≤ 28.63%), Medium-low level (28.63% < u ≤ 38.4%), Medium level (38.4% < u ≤ 43.43%), Medium-high level (43.43% < u ≤ 50.82%), and High level (u >50.82%) by adopting the natural breakpoint method (Figure 3). The overall results of the study show a continuous upward trend, and the spatial distribution pattern is characterized by a significant “east-high, west-low” pattern.

During the study period, the Pearl River Delta city cluster has consistently maintained a high level and ranked among the top in the country, while Guanzhong and Chengdu-Chongqing have long been in a relatively backward position. In addition, in the Pearl River Delta and Yangtze River Delta city clusters, apart from high-level cities such as Shenzhen and Shanghai, the rest of the cities also generally have a high quality of development, forming a relatively balanced pattern of high-quality development. In contrast, the development within the Shandong Peninsula City Cluster is more coordinated, and although it lacks high-level cities with great advantages, the overall comprehensive level is higher. The Beijing-Tianjin-Hebei City Cluster, driven by the two high-quality cities of Beijing and Tianjin, has realized rapid development in some areas, but most cities in Hebei Province are still at a lower level, which restricts the overall synergistic effect. In the Chengdu-Chongqing and Guanzhong city clusters, except for the core cities of Chengdu, Chongqing and Xi’an, the level of new-type urbanization in the rest of the cities is generally low, and the problem of uneven development is prominent.

Specifically, during the period 2006–2011, low-level cities dominated, mainly in the central and western regions; while the top-ranked cities were mostly super first-tier cities with strong economic foundations, such as Beijing, Shanghai, Guangzhou and Shenzhen, and most of them were concentrated in the eastern coastal region. With the continuous development of the regional economy, the number of low-level cities decreased significantly during the period of 2011–2016, and the proportion of medium-low and medium-level cities increased gradually. By 2022, a number of central and western cities jump from low level to low or medium level, such as Baoji City, Xianyang City, Ya’an City and Leshan City, etc., and the number of cities at low level further decreases. At the same time, some high-level cities are gradually expanding from the eastern coast to the central and western regions, especially capital cities such as Zhengzhou, Changsha and Wuhan, whose levels have increased significantly, indicating that the center of gravity of the development of high-level urbanization is showing a trend of spreading from the east to the central and western regions.

4.2. Analysis on the Level of Drag Effect

In this study, the drag effect of land resources in the process of new-type urbanization is measured using the drag effect model. From 2006 to 2022, the spatial distribution pattern of the drag effect of land resources in China has gone through an evolutionary process from “high drag effect agglomeration” to “a general weakening of drag effect”. From 2006 to 2022, the spatial distribution pattern of land resources drag efficiency in China has experienced an evolution from “high drag efficiency agglomeration” to “generally weakening drag efficiency”.

In order to further observe the spatial and temporal changes in the drag effect of land in each city of the ten major urban agglomerations, the natural breakpoint method was used to classify the cities into Low level (drag ≤ 0.002), Medium-low level (0.002 < drag ≤ 0.022), Medium level (0.022 < drag ≤ 0.060), Medium -high level (0.022 < drag ≤ 0.060) and high level (0.060 < drag ≤ 0.108) (Figure 4).

From the perspective of the overall evolution trend, the spatial distribution of the drag effect of land resources on new-type urbanization is characterized by a strong distribution in the east and a weak distribution in the west. Specifically, the drag effect of land resources in 2006 is dominated by medium-high and high-grade, concentrated in the developed eastern coastal areas such as Beijing-Tianjin-Hebei, Yangtze River Delta and Pearl River Delta, with high drag effect cities such as Beijing, Shanghai and Guangzhou as the core, showing the spatial characteristics of the centralized and contiguous distribution of medium-high and medium level cities; in the central and western regions, due to the generally low elasticity of land resources and the inconspicuous changes in labor resources, the Central Plains City Cluster and the city cluster in the middle reaches of the Yangtze River, on the other hand, are dominated by medium- and low-ranking drag-efficiency type cities. After entering 2011, there is a small increase in the number of high-level and medium–high-level cities on the east coast, and low-level cities in the central and western regions, such as Changsha and Chongqing, are shifting to medium–low grades, with an increase in the effect of the constraints on land resources. By 2016, there was a shift in this trend, with the number of high-level regions declining nationwide, and drag-efficiency grades generally evolving to medium–low grades. In 2022, the drag effect of land resources will further decline, with the vast majority of prefecture-level cities entering the medium–low grade and below, and the proportion of high-level effect areas significantly decreasing, with the spatial pattern further equalizing.

4.3. Characteristics of the Dynamic Evolution of the Land Drag Effect in Time

In this study, the kernel density estimation method is used to obtain the kernel density plots of the drag effect of land in the new-type urbanization process of China’s ten major urban agglomerations in 2006, 2011, 2016 and 2022 (Figure 5), which portrays the dynamic evolution characteristics of the drag effect of land in the ten major urban agglomerations in terms of the center of gravity of the curves, the peaks and tails, respectively.

Changes in the center of gravity of the kernel density curve clearly reflect the overall dynamic characteristics of the drag effect of land resources. From 2006 to 2011, the position of the center of gravity of the curve shifted slightly to the right, indicating a slight upward trend in the drag effect of land resources. In 2016, the center of the curve began to shift to the left, suggesting that the overall level of the drag effect began to decline, demonstrating a gradual trend of optimizing the efficiency of the use of land resources. The main reason is that China’s economic structure is constantly adjusting and improving as well as the promotion of supply-side structural reform in the land sector, and urban economic growth gradually and consciously reduces its dependence on land resources, which makes the intensity of the drag effect of land resources on new-type urbanization show a weakening trend. By 2022, the center of the curve shifts further to the left and the level of drag effect tends to lower values, while the distribution of kernel density concentrates between −0.05 and 0.2, indicating that the overall level of the drag effect of land resources in the region has weakened significantly.

The changes in the peaks of the kernel density curves reflect the dynamics of the variability of the drag effects in the study area in different years. In terms of the height of the main peak of the curve, there was a significant decrease from 2006 to 2011, indicating that the differences in the drag effect of land resources in urban areas have shown a continuing trend of expansion, which may be attributed to the implementation of the “four trillion” investment program during the global financial crisis, which required land resources as a carrier to support large-scale investment and construction, resulting in a spatial imbalance in the allocation of resources in different regions, which further contributed to an increase in the intensity of differences in the drag effect of land resources between different regions. The difference in the intensity of the role of land resources between different regions has increased. While the main peak rises in 2011–2016, constrained by the Chinese government’s intensive land use policy, the difference in the drag effect of land resources has converged. In 2016–2022, the peak of the main peak again shows a significant decline, marking the deepening of the spatial differentiation of the drag effect of land resources. This may be due to the dual effects of the deepening of the new-type urbanization strategy and the dramatic changes in the external environment: under the constraints of the full implementation of the land spatial planning system, the core cities have formed drag-effect depressions by virtue of their institutional privileges and technological innovations, while the peripheral cities are caught in structural dilemmas under the multiple pressures, and the technological differences have been transformed into the degree of drag-effect divergence, which has resulted in the widening of the drag-effect differences in land resources.

In terms of the trailing characteristics of the tail of the curve, the right-hand side of the tail shows a lengthening and elevated trend from 2006 to 2016, indicating that the proportion of cities in the high-value zone increased during this period. The trailing phenomenon weakens in 2022, indicating that the proportion of cities in the high-value drag effect decreases sharply, and that the dispersion of the drag effect is significantly weakened and centralization is enhanced.

4.4. Characteristics of the Evolution of the Spatial Pattern of the Land Drag Effect

In order to further explore the spatial clustering characteristics and degree of clustering of the drag effect of land resources, this study demonstrates the spatial distribution pattern of the drag effect of land resources from both global and local aspects.

Observing the four periods of 2006, 2011, 2016 and 2022, Moran’s I index is positive in the range of 0.1998–0.402638, and the Z-Value is greater than 1.96, which all passed the significance test (

Table 2. Changes in value of the Moran Index.

Time	Moran’s I	Z Value	p Value
2006	0.402638	7.5765	0.001
2011	0.491087	9.4151	0.001
2016	0.1998	4.1711	0.002
2022	0.2669	5.2017	0.001

). It shows that there is a significant positive global spatial autocorrelation of the drag effect of land resources in the study area over the four periods of time, i.e., the drag effect of local land resources not only affects the neighboring cities but also is affected by the neighboring cities.

The spatial clustering changes in the drag effect of land resources in the new-type urbanization process of China’s ten major urban agglomerations were analyzed using ArcGIS 10.8 during the period from 2006 to 2022, and the results were classified into four clustering types, namely, high-value agglomeration (HH), low-value collapse (LH), low-value agglomeration (LL), and high-value bulge (HL) (Figure 6).

As a whole, during the period of 2006–2022, the characteristics of the drag effect of land resources show the distribution characteristics of “big agglomeration and small agglomeration” and “east high and west low”, with HH and LL being distributed in succession, while LH and HL are scattered and sporadic, and the distribution pattern of the four types is generally LL > HHH > HL > LH in terms of quantity. HH and LL are continuously distributed, while LH and HL are scattered and sporadically distributed, and the distribution trend of LL > HHH > HL > LH is presented in terms of the number of the four types. Specifically, in 2006, HH was mainly concentrated in the eastern region, especially the Yangtze River Delta, the Pearl River Delta and the West Coast of the Taiwan Strait, showing that these regions had a high drag effect of land at that time, while LL appeared in the central and western regions that were relatively lagging behind in development. In 2011, the HH region in general converged, but there was an expansion of HH cities in the YRD region, indicating that the level of drag effect of land resources in these regions had improved somewhat. In addition, the LL region has expanded compared to 2006, mainly in the southeastern part of the middle reaches of the Yangtze River and the northwestern cities of the city cluster on the west coast of the Taiwan Strait, indicating that the level of the drag effect of land resources in these cities is low. In 2016, although HH in the east still existed, its scope was significantly reduced, and HL was sporadically distributed in the vicinity of HH. This indicates that the economic and urbanization levels in these regions are balancing. By 2022, new HH regions are added in the Beijing-Tianjin-Hebei city cluster, which may be related to the establishment of Xiongan New Area, where large-scale infrastructure and industrial undertaking lead to the squeezing of construction land indexes in the central and southern Hebei region, and the gap between the supply and demand of land expands regionally, leading to an increase in the drag effect of land resources [5]. This change reflects the gradual concentration of the drag effect of land resources in the more economically developed regions and new large-scale construction areas. In addition, the number of cities in the LL distribution area has decreased, mainly in the Guanzhong urban agglomeration area.

5. Discussion

Taking China’s ten largest urban agglomerations as the research object, this study systematically analyzes the spatial and temporal dynamic characteristics of the development level of new-type urbanization and the drag effect of land resources with the data from 2006 to 2022. The study shows that with the continuous promotion of new-type urbanization, the drag effect of land resources shows a decreasing trend in general, but there are still significant differences between regions, reflecting the complex relationship between new-type urbanization and land resource utilization.

5.1. Comparison with Previous Studies

Based on a comprehensive literature review, this study explores the drag effect of land in the ten major urban agglomerations in the context of China’s new-type urbanization by applying the land resource drag effect model, geographically weighted regression and spatial autocorrelation. The results of the study are analyzed in the following sections.

(1)

After measuring and analyzing the overall characteristics of the new-type urbanization level of cities, it is found that in recent years, the new-type urbanization level of Chinese cities has continued to rise, but the growth rate has shown a trend of slowing down gradually, and the spatial distribution pattern has been clearly shown to be “high in the east and low in the west”. This trend is consistent with the findings of Zhang et al., who also observed that eastern coastal regions have maintained relatively high levels of urbanization due to their geographical advantages and economic foundations, while central, western, and northeastern regions lag behind in development [38]. This further validates the robustness and representativeness of the conclusions drawn in this study. The difference is that this paper introduces multi-dimensional indicators such as ecology, spatial coordination, and urban–rural integration into the measurement. This makes the urbanization level assessment values of some resource-based cities lower than the results under a single population indicator, more accurately reflecting the differences in the quality of urbanization. This is also an important reason for the differences between some of the results and previous studies. In order to further reveal the temporal evolution and spatial characteristics behind these differences, this study conducted a decomposition analysis of the new-type urbanization levels at different stages and in different regions.

Specifically, during the period 2006–2011, driven by the policy support of the Eleventh Five-Year Plan and the rapid growth of the macroeconomy, the development of new-type urbanization showed a rapid upward trend. In 2016–2022, as regional development gradually enters a mature stage, the growth rate of urbanization development tends to slow down, but its connotative growth is increasingly significant. In terms of spatial distribution, the eastern coastal region, relying on its superior location and strong economic foundation, has maintained a high level of new-type urbanization for a long time, and the development of cities is becoming more and more coordinated, especially in the Yangtze River Delta, the Pearl River Delta and the Shandong Peninsula, etc., where the core cities, through the agglomeration of resources and spillover of functions, have led to synergistic enhancement of the neighboring cities to form a more significant spatial agglomeration effect. In contrast, although Wuhan, Zhengzhou and other core cities are the engine of development, the central region has shown a faster momentum of improvement, but there are still structural problems within the region such as the lagging development of marginal cities and the overall development level is not as high as that in the east. The western and northeastern regions show a typical pattern of “core-driven-peripheral lag”, except for Xi’an, Chengdu, Shenyang and a few other core cities that have achieved more obvious urbanization, some resource cities (such as Fushun) are constrained by a single industrial structure and intensified pressure of transformation, and their urbanization has stagnated at a certain stage. The development of urbanization in some resource cities (e.g., Fushun) has been stagnant due to the single industry structure and the increasing pressure of transformation.

In view of the persistence of the pattern of “high in the east and low in the west” in China’s new-type urbanization, the uneven development among regions and the overall slowdown in growth rate, it is suggested that differentiated urbanization strategies be promoted according to local conditions, the radiation-driven role of the core cities in central and western China be strengthened, the policy support and industrial transformation support for the peripheral and resource cities be increased at the same time, and a transformation from external expansion to internal upgrading be promoted. It is also recommended to promote the transformation from external expansion to internal enhancement, focusing on improving the equalization of public services, the quality of the ecological environment and the governance capacity of urban and rural areas; optimize the internal synergy mechanism of the urban agglomerations; and strengthen the complementary functions and linkage of factors between the core cities and the peripheral cities, so as to enhance the level of regional integrated development and help China’s high-quality new-type urbanization development.

(2)

From the overall measurement results of land resource deadweight levels, a general trend of weakening is evident, which is consistent with the findings of Zhao et al. [12]. Additionally, in terms of spatial patterns, there is a “strong east, weak west” characteristic, with some urban agglomerations in central and western regions still showing an upward trend in deadweight levels. This conclusion aligns with Zhao’s view that “land constraints in eastern regions are higher than in western regions [12]”. However, at a more refined spatial scale, specific cities in the eastern coastal regions with high tail effect levels have been identified, and the spatial connectivity between these cities and their surrounding areas has been revealed. For example, despite a general decline in the overall tail effect levels of eastern cities in recent years, certain core functional spillover zones—such as Suzhou, Wuxi, and Hangzhou surrounding Shanghai—still exhibit a pronounced concentration of high tail effects. This finding not only addresses the spatial resolution limitations of existing research but also partially corroborates Chu et al.’s assessment, from a fiscal perspective, that “land pressure in coastal megacities remains prominent [39].” At the same time, the distribution pattern of high-efficiency cities along the eastern coast is consistent with Zhou et al.’s conclusion that “political factors such as land reserves, planned floor area ratios, and benchmark land prices, as well as transportation accessibility and natural resource endowments” have a significant driving effect on land development [40]. This enhances the robustness and policy orientation of the research conclusions.

It is worth noting that this spatial distribution of “strong east and weak west” is not static but rather exhibits obvious temporal evolutionary characteristics at different stages of development. Further analysis of time series data reveals the changing trajectories of tail effect levels in different regions. In fact, this may reflect the differentiated mechanisms of land use intensity, stage of development, and policy environment. Specifically, in 2006, due to the rapid urban expansion and high intensity of land development in the Yangtze River Delta and the Pearl River Delta and other developed eastern coastal regions, the marginal output effect of land resources was gradually diminishing, thus forming a certain degree of constraint on economic growth; whereas in the central and western regions, due to factors such as a lower stage of development, less mobility of labor and low land use density, land resources have not yet become a core bottleneck constraining regional development. After entering 2011, the constraint effect of land resources has increased significantly. This trend of change may be due to the superimposed effect of multiple mechanisms, on the one hand, with the rapid advancement of new-type urbanization, the population density of cities and the economic volume of synchronous growth, so that the rigid demand for new land for construction continues to intensify; on the other hand, in the context of the international financial crisis, China accelerated the pace of investment in infrastructure and real estate through a large-scale stimulus plan, and this has led to an increase in land use density. On the other hand, against the backdrop of the international financial crisis, China has accelerated the pace of infrastructure and real estate investment through a large-scale economic stimulus program, which is highly reliant on land inputs and has significantly increased the pressure on marginal output of land resources in the short term. At the same time, land resources available for development are subject to the double constraints of the red line of arable land and ecological protection policies, with limited space for new additions, further magnifying the contradiction between supply and demand. These factors have synergistically exacerbated the resource bottleneck effect of land elements on economic development and new-type urbanization, reflecting the structural challenge of land resources facing normalized constraints as China enters the middle and late stages of urbanization. Such changes highlight the practical urgency of optimizing the land use structure and enhancing the land intensification efficiency to promote high-quality urbanization. During the 2016–2022 period, the overall drag effect of land resources shows a declining trend, and regional spatial differences have converged, indicating that the constraint of land resources on new-type urbanization is slowing down. This shift may be attributed to the synergistic driving effect of institutional innovation and technological progress: at the institutional level, the structural reform of the land supply side effectively breaks down the institutional barriers in the process of land factor allocation and improves the spatial and temporal efficiencies of land utilization; at the technological level, developed cities such as Beijing and Shanghai gradually weaken the dependence on traditional land factors for economic growth by promoting the advancement of the industrial structure and reinforcing the path of innovation and intensification. At the technical level, Beijing, Shanghai and other developed cities have gradually weakened the dependence of economic growth on traditional land elements by promoting the advanced industrial structure and strengthening the path of innovation intensification. This transformation path essentially reflects the paradigm shift of the power mechanism of new-type urbanization from crude factor-led to institutional dividend-led—when institutional reforms ease the constraints of land resources, the drag effect of land resource constraints on the quality improvement of urbanization is also weakened. Nevertheless, the drag effect of land resources remains high in some cities along the eastern seaboard, especially in Suzhou, Wuxi, Nanjing and Hangzhou around Shanghai. This phenomenon may be due to the fact that with the spillover of Shanghai’s core functions, these node cities have rapidly taken over the transfer of industries and population, creating a regional agglomeration effect. However, under the dual pressure of the red line of arable land protection and ecological space control, it is difficult for the supply of construction land to expand in tandem, resulting in a further contradiction between land supply and demand, thus reinforcing the restrictive effect of land elements on the process of new-type urbanization.

In response to the phenomenon of the drag effect of land resources in China being “stronger in the east and weaker in the west”, and of uneven regional development, it is recommended that the construction of a new factor-driven system centered on the accumulation of knowledge, technological progress and institutional innovation be accelerated to break through the reliance on traditional factors of production such as land, capital and labor. In particular, in the eastern urban agglomeration, where the drag effect of land is high and the pressure on the carrying capacity of resources is prominent, the leading role of scientific and technological innovation in the efficiency of land utilization should be strengthened, and the development of new industries such as high-tech, green and low-carbon, and the digital economy should be promoted, so as to gradually transform the city into an efficiency-led city by improving total factor productivity.

(3)

In terms of spatio-temporal analysis, this study addresses the shortcomings of previous literature. Existing research has primarily focused on the numerical estimation and temporal changes of resource deadweight effects, with insufficient attention paid to their spatial dependence and the dynamic evolution of local aggregation patterns. Under a unified framework of a long time series and multiple urban agglomerations, this study reveals for the first time that the overall spatial dependence of the drag effect of land resources between 2006 and 2022 shows a fluctuating characteristic of “first weakening, then rebounding.” The Moran index was relatively high in 2006 and 2011, reflecting the strong spatial agglomeration of the drag effect, which was concentrated in the eastern coastal regions with a strong economic foundation and high level of urbanization. The index dropped significantly in 2016, indicating a notable weakening of spatial clustering, closely related to the relative rise of central and western regions. Although it rebounded slightly in 2022, it remained below the peak level. This suggests that the spatial distribution of land resources is evolving from a unipolar concentration toward a multipolar equilibrium.

Local spatial autocorrelation analysis further reveals the spatial differentiation characteristics of different regions. High-value clusters are primarily concentrated in eastern coastal developed cities and some provincial capitals (such as Shanghai, Hangzhou, and Suzhou). These regions have experienced rapid urbanization, high land development intensity, and spatial saturation, resulting in high costs for new construction land. Urban expansion in these areas focuses on optimizing existing land use. Therefore, they maintain a long-term “high-high” clustering pattern. In contrast, western regions lag behind in urbanization, with less pronounced land-population conflicts and relatively abundant construction land. Urban expansion in these areas primarily relies on incremental supply, resulting in a “low-low” pattern. Notably, in 2022, some central and western regions showed an upward trend toward a “high-low” pattern, indicating that certain inland cities have achieved breakthroughs in land use efficiency, with regional economic development potential gradually being unlocked, and are beginning to break away from the traditional coastal-inland dichotomy.

The research findings not only enrich the empirical evidence on the spatial evolution of land drag effects but also provide targeted insights for policy-making: in regions where the drag effect is significant, particularly in rapidly urbanizing urban agglomerations, high drag effects indicate that these areas still face significant pressure in the allocation of non-productive land uses such as public facilities and ecological spaces. Therefore, efforts should be made to strengthen spatial control and land-use regulation, optimize the layout structure of non-productive land, and enhance the comprehensive benefits of public spaces and ecological land. Priority should be given to redeveloping idle and underutilized land, strictly controlling the expansion of new land use, and encouraging compact urban development. In urban agglomerations in central and western China, infrastructure is inadequate, and industrial structures are unbalanced. Among these, urban agglomerations with significant drag effects should increase support for industrial transformation and infrastructure development to enhance infrastructure connectivity. Reasonably increase the supply of construction land to alleviate development constraints. Prioritize promoting mixed-use land development, guiding industrial upgrading, and improving land use efficiency. For regions with minimal resistance effects, integrate land allocation with industrial transformation, population mobility, and ecological protection to drive a shift in land use patterns from extensive to intensive. Maintaining a balanced land use structure and preventing disorderly expansion are key to avoiding the recurrence of land resource constraints. Overall, differentiated land allocation policies and urban planning strategies tailored to the developmental stages and spatial characteristics of each urban agglomeration are crucial for transforming non-productive land into productive assets and supporting sustainable new-type urbanization.

5.2. Limitations and Future Research

First, due to data availability constraints, this study only selected two types of land, namely agricultural land and construction land, and did not include forest land, grassland, water areas, and other land types. This may have affected the accuracy of the drag effect of land to some extent and limited the comprehensive analysis of the role of different land types in the process of new-type urbanization. In response to this limitation, future research can combine remote sensing images, land use/land cover data, and big data technology to construct a high-precision land resource data system covering multiple types of land, thereby improving the accuracy of drag effect assessment and further revealing the differentiated contributions of different land types in the process of new-type urbanization, providing a more scientific basis for policy-making and land management. Second, the measurement model in this study mainly refers to the production function model, which is more mature at home and abroad, but the parameter estimation and measurement method chooses the geographically weighted regression model, which still needs to be further verified and improved in future research, so as to prepare the foundation for scientifically calculating the “drag effect”. In addition to land resource constraints, the development of new-type urbanization may also be affected by a variety of exogenous factors, for example, regional development policies, urban planning strategies, and fiscal incentives. Depending on their specific design and implementation, these factors may weaken or strengthen the effect of land resource constraints. Due to data availability and the core objectives of this study, such policy factors were not directly included in the model. However, it is undeniable that exogenous variables play an important role in the process of new-type urbanization. Future research can introduce relevant policy variables based on the existing framework to more comprehensively reveal the impact mechanism and dynamic evolution characteristics of new-type urbanization.

6. Conclusions

Based on the statistical data of China’s ten largest urban agglomerations from 2006 to 2022, this study constructs a model for measuring the drag effect of land resources based on the C-D production function, and combines it with geographically weighted regression methods to systematically analyze the characteristics of the drag effect of land resources in the process of new-type urbanization in the ten largest urban agglomerations in China, as well as their spatial evolution patterns. The results of the study show that:

(1)

During the period of 2006–2022, the development level of new-type urbanization shows a general upward trend, but there are obvious differences in the development of each urban agglomeration, showing a spatial distribution pattern of “high in the east and low in the west”. In this process, the urbanization development level of the Pearl River Delta city cluster is in a leading position in the country, while that of the Guanzhong city cluster is relatively low, reflecting the unevenness of regional development.

(2)

During the study period, the drag effect of land resources generally showed a declining trend, but with significant regional differences. Spatially, the constraint of land resources on new-type urbanization development has been eased, and the intensity of the drag effect of land resources shows a distribution characteristic of ‘strong in the east and weak in the west’, a spatial distribution that is similar to the spatial pattern of new-type urbanization development. The centre of gravity of the growth of the drag effect of land resources is moving from south to north and from east to west. Specifically, the Beijing-Tianjin-Hebei urban agglomeration and the urban agglomeration in the middle reaches of the Yangtze River show an increase in the drag effect of land, while the Pearl River Delta and Yangtze River Delta urban agglomerations show a significant decrease in the drag effect of land.

(3)

From 2006 to 2022, the centre of gravity of the kernel density curve of the drag effect of land resources has experienced a trend of shifting to the right and then to the left, indicating that the drag effect of land resources has begun to show a downward trend as a whole after experiencing a slight increase. Most of the values of the drag effect of land in each city are concentrated in the range of −0.02 to 0.03, indicating that the overall level of drag effect of land resources in the region has significantly weakened. At the same time, the peak of the kernel density curve shows a fluctuating trend of ‘decreasing-rising-decreasing’, reflecting that the differences in the level of drag effect of land between the urban agglomerations are gradually narrowing. Observing the trailing characteristics at the end of the kernel density curve, the right-hand side of the trailing shows a tendency of lengthening and elevation in the early stage, while the trailing phenomenon is weakened in the later stage, and the dispersion of drag decreases and the concentration increases.

(4)

During the study period, the drag effect of land resources showed a spatial pattern of “high in the east, low in the west, and obvious agglomeration”, with the HH and LL types of areas distributed in a centralized and continuous manner, while the HL and LH types of areas were distributed in a fragmented manner, and the overall spatial structure remained relatively stable. The Yangtze River Delta region continues to be the core agglomeration of HH-type areas, while the central and western regions are mainly dominated by LL-type areas, and there is still much room for improvement in the utilization efficiency of land resources. Overall, the spatial scope of low drag effect areas has been reduced, and the drag effect of land resources has gradually gathered in economically active and key development areas.