Evaluation and Influential Factors of Urban Land Use Efficiency in Yangtze River Economic Belt

Abstract

The study of urban land use efficiency is of great significance for optimizing the spatial allocation of urban land, thereby promoting the intensive use of urban land and the transformation of economic development modes. Taking the Yangtze River Economic Belt (YREB) as the study object, we chose the undesirable Slacks-Based Measure (SBM) model to calculate the urban land use efficiency (ULUE). Then, we utilized the spatial correlation analysis and econometric methods to discuss its spatio-temporal features and influential factors. The results show the following: (1) The urban land use efficiency in the YREB steadily improved from 2010 to 2022, but the inter-regional efficiency gap evidently increased. (2) There is an efficiency value to be found in a multi-center network structure, and it forms a “core-periphery” distribution pattern. The high-efficiency areas in the downstream and upstream regions of the YREB are gradually increasing, while the efficiency value in the midstream area remains low. (3) The urban efficiency values have strong correlation, and they are mainly “High-High agglomeration” and “Low-Low agglomeration”, and they show significant regional characteristics. (4) The economic level, industrial structure, and urbanization have obvious motivating effects on ULUE, and the positive spatial spillover effect is clear. The foreign direct investment and land finance hinder the boost of efficiency, and the latter has a negative spatial spillover role on the ULUE in the downstream cities.

1. Introduction

Land is not only space for human survival, but it is also the basic guarantee for promoting the urbanization process [1]. China is experiencing a quick urbanization, and the problems of a quick growth of construction land and imbalance between supply and demand in land use structure have led to increasingly salient conflicts among urban development, cultivated land protection [2], and environmental governance [3]. Enhancing ULUE is a vital choice to reduce land planning inefficiency and promote the rational allocation of land. Solving the land use contradiction in urbanization has become a crucial strategic combination of high-quality economic growth and land development.

The YREB has always been the most actively developed areas in China, as well as a giant industrial economic belt and an ecological civilization pioneer zone that spans the east, middle, and west of China [4,5]. The rational use of land is a crucial means to the realization of regional coordinated development and modern city construction. At present, the rapid and disorderly expansion of land in the YREB has become an important features of regional urban land use. It brings about the consequences of land abuse, and the restriction of insufficient exploitable land on city development have begun to appear. Now, the most effective solution is to improve ULUE. ULUE not only determines the performance level of land development, utilization, and protection policies, but it also impacts the layout of city multi-functional coordinated development [6]. The YREB is a vast area with rich resources and huge internal differences. Systematically studying the evolution characteristics, regional differences, and influential factors of ULUE in the YREB has major practical and long-range strategic significance for narrowing the development gap of regions and improving ULUE. It is hoped that it will offer a comprehensive decision-making reference to the planning and coordinated development of national and regional land resources.

To solve the contradiction between social development and land utilization, many scholars have attached greater importance to the study of ULUE. From the perspective of research categories, the studies involve both micro and macro aspects. This study analyzed the land use of different industrial types from the micro level and the focus on the rational land utilization [7]. However, an analysis from the micro perspective cannot effectively solve the external impact, so the attention of scholars to land studies has increasingly shifted to the macro level, thus tending toward provincial level [8], urban agglomeration, [9] and prefecture-level city studies [10], where the focus is on spatio-temporal features and influential factors [11,12]. Lin and Ling (2021) found that the ULUE in the Yangtze River Delta has steadily improved, and the spatial difference has gradually expanded [13]. Zhao et al. (2019) suggested that urban land use in Hangzhou shows an expanding tendency, and its spatial characteristics have become more complex [14]. Zhu et al. (2019) proved the heterogeneity between cities in terms of ULUE and found that economic level and infrastructure promote efficiency improvement [15]. Industrial structure has a promoting effect on ULUE [16], while the influence of industrial clusters is diversified, and its promoting or inhibiting effect depends on the form of agglomeration [17]. Urbanization and government actions can help to improve ULUE, while population density inhibits efficiency improvement [18]. Koroso (2023) analyzed the relation between lease policy and ULUE in Ethiopia, and believed that the lease policy was not well implemented, thus leading to more chaotic land use [19]. From the perspective of an evaluation system, the research on efficiency measures has realized a comprehensive development from single indicator to a multi-indicator approaches. Many scholars have characterized regional ULUE by the industrial added value per unit area [20] and land development intensity [21] metrics. Most scholars believe that the land production efficiency measure should consider a comprehensive index framework, including land elements, such as the “economic-social-ecological” framework [22] and the “input-expected output” framework [23], so as to evaluate ULUE from multiple angles. However, urban land use not only brings economic and social benefits, but it also carries pollutant discharges [24,25]. An increasing number of scholars have noticed the negative influence of land utilization on sustainability. Based on an understanding of this objective fact, researchers have included pollutant discharges as undesired outputs in the study framework of ULUE measurement [26,27], and they have also comprehensively evaluated ULUE from the perspectives of environmentally friendly and green development. In terms of evaluation methods, the DEA model [28], SFA model [29], and SBM model [30] are mainly used to calculate ULUE. The traditional DEA model is greatly affected by abnormal values when estimating efficiency, and it only calculates desired outputs, thereby ignoring the impact of undesired outputs on efficiency. The SFA model needs to establish a relationship equation and can only calculate one output when measuring efficiency, which is contrary to production practice. In fact, as the carrier of social development, land produces desired benefits, and it brings undesired responses. The SBM model not only fully considers the undesired outputs in economic activities, but it also eliminates the overestimation of land use efficiency caused by ignoring slack variables, thus making the measurement value of ULUE more accurate and widely used in land use efficiency research. The major studies are summarized in

Table 1. Summary of the main studies.

Aspects	Perspectives	Description	Literature Sources
Research scopes	Micro-view	High potential for business land use efficiency improvements	Yang et al. (2022) [7]
	Macro-view	The scholars explored the spatial and temporal features and influential factors of ULUE in provinces, urban agglomerations, and prefecture-level cities	Cao et al. (2024) [8], Lu et al. (2022) [9], Qian and Luo (2024) [10], Wang et al. (2022) [11], Ren et al. (2023) [12]
Influential factors	Economic level, infrastructural, industrial structure, urbanization, and government action	Significantly improved ULUE	Zhu et al. (2019) [15], Han et al. (2020) [16]
	Population density	Inhibited ULUE	Zhang et al. (2023) [18]
	Leasing policy	The influence on ULUE is uncertain	Koroso (2023) [19]
Indicator construction	Single indicator	Industrial added value per unit area and land development intensity	Li et al. (2014) [20], Li et al. (2024) [21]
	Indicator system	Comprehensive evaluation of urban land use efficiency	Wang et al. (2021) [22], Guo et al. (2024) [23], Habib et al. (2024) [24]
Evaluation methods	DEA model	Ignored non-expected outputs	Zhong et al. (2020) [28]
	SFA model	Single output	Meng et al. (2024) [29]
	SBM model	Considered expected output and unexpected output comprehensively, and the problem of more input and more output was solved	Lin et al. (2024) [30]

.

To summarize, these previous works conducted many in-depth studies on ULUE from multiple perspectives. The related research tends toward empirical analysis, i.e., the choice of evaluation indices is increasingly comprehensive and scientific, and the application of methods and models is rich and diverse. The research scale is comprehensive, and the study of ULUE has its own system. The existing research has a strong guiding significance for this paper to carry out research on ULUE in the YREB. But there are still many weaknesses: First, land use-related studies have mainly focused on provinces and city clusters; thus, there are fewer studies on the YREB as a complete region from the city level, and comparative analyses of regions with different levels of development are also lacking. Second, the selection of indicators is limited to the socio-economic perspective; thus, environmental factors carried by the land have been ignored, and a single indicator alone cannot comprehensively and accurately measure land use efficiency. Third, the traditional model does not consider the effect of the input–output slack variables on the results in measuring ULUE, thus resulting in the efficiency measurements being inconsistent with the reality.

The YREB is a dense economic zone in China, and it is a pilot demonstration belt and pilot area for new-type urbanization construction in the new era. However, in the past expression of a long development process, as an area with fast urbanization development, the conflict among resources, environment, and the economy became increasingly prominent. Blind expansion and irrational use have caused low land utilization efficiency, which directly affects regional green development. As a demonstration zone of ecological civilization construction, and due to its unique location and field, it is necessary to analyze its internal land use level and influential factors. Such an observation provides a theoretical basis for further improving ULUE in the future. Thus, we take 111 cities in the YREB as the study object and construct an ULUE index system. This index helps with scientifically measuring ULUE with the undesirable SBM model, comparative analyses of the heterogeneity of efficiency in different regions, and influential factors by the application of the spatial econometric model. Through the above study, we aim to explore the spatio-temporal features of ULUE in the YREB, identify the main influential factors, and provide suggestions for improving ULUE from the perspective of regional synergy.

The marginal contributions of this paper are as follows. First, the YREB is the pilot zone and demonstration base for China’s green development. The research on ULUE in the YREB has important theoretical and practical significance, which not only provides theoretical support for the decision making on improving efficiency in the region, but it also provides relevant experience for other basins. Second, in terms of research scale, considering that the provincial level is too broad and there are big differences among provinces, the research object of the article is the cities in the YREB as it will be more accurate to measure ULUE at the municipal level. Third, compared with previous works, which have paid attention to the effect analysis of a single factor, we utilize a spatial econometric model to analyze the influence of economic development, industrial structure, urbanization, foreign direct investment, and land financing on ULUE and the space spillover effect, thereby identifying the key influential factors, which will help the YREB become a pioneer in the efficient use of land.

2. Research Methodology and Data

2.1. Study Area

The YREB consists of 11 provinces, covering a total area of about 2.05 million km² (Figure 1). It is the location of China’s economic center and a strategic region for high-quality development, with 21.4% of the land area carrying more than 40% of the population and the total economic volume. The region spans the three levels of terrain in China, with large topographic fluctuations and complex landforms, and it mainly includes mountains, hills, and plains. The upper reaches of the YREB include Chongqing, Sichuan, Guizhou, and Yunnan; the middle reaches include Jiangxi, Hunan, and Hubei; and the lower reaches cover Shanghai, Jiangsu, Zhejiang, and Anhui.

2.2. SBM Model

Efficiency measurement methods mainly include parameter methods, such as the SFA model, and non-parameter methods, such as the DEA model. The SFA model has many parameters and assumes economic isomorphism, so it is not suitable for regional difference analysis. The DEA model can avoid the deviation caused by the preset production function form and the distribution characteristics of error terms when using the parameter method, so it has been extensively applied in efficiency calculation. In their article, the relaxation-measure-based SBM model proposed by Tone (2001) [31] demonstrated a better correction effect with the DEA model. It solved the non-zero slack problem and fully considered the undesirable outputs, and the measured efficiency values were more realistic [32]. The formula is as follows:

$$\begin{gathered} \theta =\min \frac{1-\frac{1}{N}{\displaystyle \sum _{n=1}^N{s}_n^x/x_{kn}^t}}{1+\frac{1}{M+I}\left({\displaystyle \sum _{m=1}^M{s}_m^y/y_{km}^t+{\displaystyle \sum _{i=1}^I{s}_i^b/b_{ki}^t}}\right)}, \\ s.t.\left\{\begin{array}{l}{x}_{kn}^t={\displaystyle \sum _{t=1}^T{\displaystyle \sum _{k=1}^K{\lambda }_k^t{x}_{kn}^t+s_n^x,n=1,\cdot \cdot \cdot ,N}}\\ y_{km}^t={\displaystyle \sum _{t=1}^T{\displaystyle \sum _{k=1}^K{\lambda }_k^t{y}_{km}^t-s_m^y,m=1,\cdot \cdot \cdot M}}\\ b_{ki}^t={\displaystyle \sum _{t=1}^T{\displaystyle \sum _{k=1}^K{\lambda }_k^t{b}_{ki}^t+}}{s}_i^b,i=1,\cdot \cdot \cdot ,I\\ {\lambda }_k^t\ge 0,s_n^x\ge 0,s_m^y\ge 0,s_i^b\ge 0,k=1,\cdot \cdot \cdot ,K\end{array}\right., \end{gathered}$$

(1)

where θ denotes ULUE; N, M, and I denote the number of input samples, desired output samples, and undesired output samples, respectively; x, y, and b denote the vector of inputs, desired outputs, and undesired outputs, respectively; k refers to the decision-making unit; s_n^x, s_m^y, and s_i^b are the slack variables for the inputs, the desired outputs, and the undesired outputs, respectively; and λ denotes the weighting coefficient.

2.3. Spatial Autocorrelation

For the purpose of testing the spatial pattern of the sample area, it was investigated whether there was a spatial correlation between the data. It is common to formulate a global Moran’s I to study, and the formula is as follows:

$${Moran}^{\prime }sI=\frac{{\displaystyle \sum _{i=1}^n{\displaystyle \sum _{j=1}^n{W}_{ij}\left(Y_i-\overline{Y}\right)\left(Y_j-\overline{Y}\right)}}}{S^2{\displaystyle \sum _{i=1}^n{\displaystyle \sum _{j=1}^n{W}_{ij}}}},$$

(2)

$$S^2=\frac{1}{n}{\displaystyle \sum _{i=1}^n\left(Y_i-\overline{Y}\right)},\overline{Y}=\frac{1}{n}{\displaystyle \sum _{i=1}^n{Y}_i},$$

(3)

where ${Moran}^{\prime }sI$ is the global spatial autocorrelation index; $W_{ij}$ is the spatial weight matrix; $Y_i$ and $Y_j$ represent the ULUE of the $i$ and $j$ cities, respectively; and $n$ is the number of cities in the study area.

The above formulas only measure the spatial correlation, and the identification of high-value and low-value regions cannot be completely determined. Therefore, on the basis of testing the global spatial correlation, a local correlation test was performed to measure the local correlation characteristics between the ULUE of the region and the surrounding areas. The formula is as follows:

$${Moran}^{\prime }sI=\frac{\left(x_i-\overline{x}\right)}{{\displaystyle \sum _i{\left(x_i-\overline{x}\right)}^2}}{\displaystyle \sum _j{w}_{ij}\left(x_i-\overline{x}\right)}.$$

(4)

2.4. Empirical Model

Owing to the transfer and flow of factors between regions, the change in ULUE in neighboring regions may have spillover effects on local efficiency. If there is a spatial relationship between things, the traditional linear regression model will not be able to accurately express the effect, so the spatial econometric models are usually chosen to research the relation between variables [33]. Common specific model forms are as follows.

(1) SAR model:

$${\theta }_{i,t}=\lambda W{\theta }_{j,t}+{\beta }_i{X}_i+{\mu }_i+{\eta }_t+{\epsilon }_{i,t},$$

(5)

where $\lambda$ is the correlation coefficient; $W$ is the spatial matrix; ${\mu }_i$ and ${\eta }_t$ are the individual and the time fixed effect, respectively; $X_i$ is the dependent variable; and ε_i,t is the interference term.

(2) SEM model:

$$\begin{array}{l}{\theta }_{i,t}={\beta }_i{X}_i+{\mu }_i+{\eta }_t+{\epsilon }_{i,t}\\ {\epsilon }_{i,t}=\rho W{\epsilon }_{i,t}+{\upsilon }_{i,t}\end{array},$$

(6)

where $\rho$ denotes the spatial autocorrelation coefficient.

(3) SDM model:

$${\theta }_{i,t}=\lambda W{\theta }_{j,t}+{\beta }_i{X}_i+W{X}_{i,t}\gamma +{\mu }_i+{\eta }_t+{\epsilon }_{i,t},$$

(7)

where $\gamma$ denotes the spatial autoregressive coefficient of the independent variable and $\lambda$ denotes the spatial autoregressive coefficient of the dependent variable.

2.5. Indicator System

Referring to the existing research results, the evaluation indicators were selected from the input and output aspects [34,35]. In terms of the inputs, basic elements such as the resources, capital, and manpower that reflect land use are usually selected. The urban land area provides space and potential for the adjustment of land structure. Capital and labor are essential elements for economic development. We chose three indicators to represent the inputs.

In terms of the outputs, the desired outputs were considered to be mainly the economic, social, and environmental aspects. Economic benefits reflect the economic level of land bearing. Social benefits are mainly reflected in the change in people’s income. Environmental benefit is the green level of urbanized development. In this study, three indicators were chosen to represent the desired outputs. In addition to being a carrier for social activities, urban construction land carries undesired outputs such as pollution discharges. We chose four indicators to represent the undesired outputs. Specific indicators are shown in

Table 2. Index system construction.

Target Level	Evaluation Dimension	Indicator Level	Units
Inputs	Land	Urban construction land area [36]	Square kilometers
	Capital	Fixed-asset investment [15]	CNY Ten thousand
	Labor	Number of employees in the secondary and tertiary industries [37]	Ten thousand people
Outputs	Desired outputs	Added value of the secondary and tertiary industries [38]	CNY Ten thousand
		Average wages of urban workers [39]	CNY
		Green coverage area of a built-up area [40]	Hectares
	Undesired outputs [41]	Industrial wastewater discharges	Ten thousand tons
		Industrial carbon dioxide discharges	Tons
		Industrial sulfur dioxide discharges	Tons
		industrial soot and dust discharges	Tons

.

2.6. Data Sources

Since 2010, the Chinese Government has repeatedly emphasized the importance of the YREB. During this period, President Xi Jinping has presided over three symposiums on the development of the YREB. From promotion, in-depth promotion, to comprehensive promotion, the construction of the YREB has been related to the national development plan. The data of the indicators involved in the study mainly come from the 2010–2022 China Urban Statistical Yearbook, as well as the yearbooks and statistical bulletins of provinces and cities in the YREB.

3. Results

3.1. Time Evolution Analysis

The calculation results of ULUE in the YREB are shown in Figure 2. The regional ULUE mainly has three characteristics. First, the efficiency has been significantly improved. The efficiency has risen from 0.391 in 2010 to 0.726 in 2022, thus showing a steady improvement of the YREB as a whole. The temporal evolution features of ULUE in the upstream region are roughly in line with the overall tendency; although the economic basis is relatively weak, the development speed is faster. With increased investment in environmental governance, pollution emissions have been curbed and ULUE has gradually improved. The growth of ULUE in the middle reaches is relatively slow even though some its cities are China’s manufacturing and energy supply base; in addition, the industrial structure is not reasonable, and the proportion of polluting industries is too high. Moreover, the transfer of industrial enterprises has raised local environmental pressures and inhibited the improvement of ULUE. The downstream areas are relatively economically developed and densely populated, and they are more in line with the rational planning of urban land. The production modes are mature, thus reducing environmental cost, and ULUE is growing faster.

Second, the regional efficiency gap is obvious, and the gap is gradually increasing. In 2010, the ULUE in the upper reaches, middle reaches, and lower reaches was 0.325, 0.357, and 0.472, respectively, with a difference of 0.147 between the highest and lowest values. In 2022, the efficiency in the upper reaches, middle reaches, and lower reaches increased to 0.622, 0.488, and 0.813, respectively, with a difference of 0.325 between the highest and lowest values. The inter-regional efficiency gap is thus becoming increasingly evident.

Third, the growth of the efficiency value has been found to be more tortuous. The growth tendency of the upper reaches of the YREB shows a ladder-like character, and ULUE is growing faster. As the geomorphology in the upstream areas is complex, and due to the fact that the city development and land are limited, fully utilizing land resources is a prerequisite for development. The efficiency disparity between the midstream and the upstream areas is progressively widening, thus indicating that, although ULUE has improved slightly, the total efficiency of the midstream area is low and the future growth space is large. The proportion of polluting industries in the midstream areas is high, and the pollution discharges carried by the land have reduced ULUE. The efficiency in the lower reaches increased first, then decreased and then increased, thus indicating that the early urbanization expansion makes full use of the construction land, which makes the efficiency value continue to rise. However, due to the serious environmental pollution caused by economic development, the efficiency has reduced. As economic growth has shifted from extensive to green, the government has increased its efforts to control pollution, thus effectively improving land use efficiency.

3.2. Spatial Feature Analysis

3.2.1. Spatial Distribution Pattern

To visually show the space features, we used the natural discontinuity method to render the ULUE of four time nodes, and the results are shown in Figure 3. The space characteristics can be summarized as follows.

First, the spatial non-equilibrium characteristics were found to be significant. The number of cities with high ULUE in the lower reaches of the YREB was more than that in the middle and upper reaches. In 2010, there was a significant step change in the efficiency from east to west, thus showing that the efficiency in lower reaches were higher than in the middle reaches and upper reaches. In 2022, the urbanization of the western region developed rapidly and ULUE significantly improved, but regional differences still existed. ULUE is not only relevant to the economy, but it is also affected by urban expansion. There are differences in the economic basis and available land resources in different regions, and ULUE also has regional heterogeneity.

Second, the “core-periphery” spatial distribution pattern was found to be significant. The ULUE gradually presented itself as a type of multi-center network structure. The lower reaches have basically formed a high-efficiency agglomeration area with Shanghai as the leader, and they then diffused outwardly layer by layer. The middle reaches were mainly of medium and low efficiency, and the efficiency values of Wuhan and Changsha were significantly higher than those of surrounding areas. But, as it the area has been affected by urbanization and the economic level, the overall efficiency of the middle reaches was found to be low; thus, the driving effect of key cities needs to be enhanced in the future. The upper reaches have formed a core area based on Chengdu and Chongqing. The two cities have developed rapidly, and the economic basis and urban construction have promoted the improvement of ULUE.

Third, the “Matthew effect” is becoming increasingly obvious. The high-efficiency areas in the YREB are evidently increasing, and the high-efficiency cities are gradually spreading from the east to west. The high-efficiency areas in the downstream region are clearly concentrated, and the efficiency of the midstream area is still relatively low, thus indicating that the ULUE in this region has fallen into the trap of the “Matthew effect”. Thus, it is difficult to effectively solve the contradictions between land allocation and social progress with its own efforts. The possible reasons for this are that the large-scale investment and construction in various cities leads to overcapacity, and traditional production patterns do not conform to the green development, thus resulting in insufficient endogenous motivation for urban land use.

3.2.2. Spatial Correlation Analysis

The correlation attributes of things are affected by spatial distance and may have spatial dependence. If the autocorrelation of ULUE in space is not considered, the results will be biased. We calculated the global Moran’s I index and analyzed its spatial correlation. The change trends of the index are shown in

Table 3. The global Moran’s I index in 2010–2022.

Years	Moran’s I	Z	P
2010	0.223	6.642	0.000
2011	0.169	6.358	0.000
2012	0.203	7.284	0.000
2013	0.234	7.937	0.000
2014	0.217	8.327	0.000
2015	0.228	9.434	0.000
2016	0.206	8.162	0.000
2017	0.185	7.623	0.000
2018	0.197	6.382	0.000
2019	0.188	5.293	0.000
2020	0.204	6.315	0.000
2021	0.216	5.746	0.000
2022	0.232	6.632	0.000

, which all passed the significance test, thus indicating that the spatial influence persists.

The global spatial correlation cannot reflect the local correlation intensity, so the local Moran’s I index was used to analyze the local correlation features of ULUE. We measured the local Moran’s I index of each city and drew a LISA (Local Indicators of Spatial Association) spatial correlation map. LISA cluster analysis is based on the data similarity in a geographic space, which is used to analyze the degree of data aggregation, as shown in Figure 4.

High–high agglomeration was mainly distributed in the downstream areas. With time, high-value agglomeration areas gradually appeared in the upstream regions. The downstream areas have significant location advantages and a developed economy, which is conducive to the internal and external circulation of production factors, and the inter-city link effect was found to be significant. The radiation range is gradually expanding, thus driving the raise of efficiency in the adjacent cities and showing high-value agglomeration. With the rise of western urban areas, the diffusion effect of cities is constantly emerging, and the land use is more reasonable. Many cities have also shown high-value agglomeration.

Low–low agglomeration was mainly distributed in the midstream regions. These cities have relatively weak economic conditions and infrastructure. To speed up the pace of urbanization, the land has been blindly expanded, and the urban construction land has not been fully and effectively used. Moreover, with the gradual deepening of trans-regional industrial restructuring, these regions have undertaken high-polluting industries, and the local environmental pollution has intensified, thus reducing the land use efficiency. The core cities have not formed regional links, and the high-efficiency cities cannot effectively improve efficiency in neighboring areas.

4. Analysis of the Influential Factors

4.1. Influential Factor Indices

The existing studies have indicated that economic level, government behavior, and industrial structure have an obvious role in ULUE [42,43]. Chen et al. (2022) found that population density contributes to improving ULUE, while the transportation level inhibits its improvement [44]. Huang and Xue (2019) believed that technological progress is a key factor to improve ULUE [45]. This previous work examined the factors that influence ULUE. According to the existing literature, we will further discuss the influential factors, and we will use the per capita GDP, industrial structure, urbanization, foreign direct investment, and land finance parameters to establish an indicator system of the influential factors (Table 3). The influence process of each factor is as follows.

The economic level is a vital factor for measuring the comprehensive strength of a city, and economic growth helps to increase the input of land unit area factors, thus affecting ULUE. We used per capita GDP to represent the economy level.

The advanced industrial structure contributes to the efficient allocation of land by raising the utilization rate of resources, reducing the pollution emission, and lowering the pressure on land carrying capacity. We measured the industrial structure by the ratio of the added value of the tertiary industry to that of the secondary industry.

Urban development not only helps to improve city planning, but it also boosts infrastructure, which helps with the rationalization of land resource allocation. Urbanization is expressed in terms of the proportion of the urban population to the total population.

Foreign direct investment has an influence on land use; for a long time, the government, in their pursuit of economic growth through formulating land policies to attract foreign capital, used the injection of different types of capital to trigger the land use structure to change. Foreign direct investment is represented by the actual amount of foreign capital utilized.

The decision-making authorities have the right to plan and manage urban land within their jurisdiction. Increased land financing may result in a waste of land resources. Land finance activities are reflected by the annual income from land transfer payments.

The description of specific factors is shown in

Table 4. Description of the variables.

Variables	Variable Codes	Mean	SD	Min	Max
Economic level	PGDP	4.243	2.972	1.221	17.949
Industrial structure	STR	0.795	0.466	0.252	1.845
Urbanization rate	URB	53.24	42.562	44.274	89.312
Foreign direct investment	FDI	0.018	0.013	0.005	0.024
Land finance	FIN	0.085	0.031	0.024	0.572

. We found that the mean values of PGDP, STR, URB, FDI, and FIN are 4.243, 0.795, 53.24, 0.018, and 0.085, respectively. Their standard deviations are 2.972, 0.466, 42.562, 0.013, and 0.031, respectively, thus indicating that the degree of data dispersion is small.

4.2. Analysis of the Spatial Regression Results

We processed each variable logarithmically, and we then carried out spatial econometric regression using Stata 14. According to the R², we chose the SDM model for further analysis.

From the regression results in

Table 5. The spatial regression results.

Variables	OLS	SAR	SEM	SDM
lnPGDP	0.123 ***	0.212 ***	0.114 ***	0.235 ***
lnSTR	0.423 **	0.348 *	0.396 *	0.401 **
lnURB	0.014 ***	0.013 *	0.024 *	0.016 *
lnFDI	−0.023 *	−0.028 **	−0.035 *	−0.031 **
lnFIN	0.008 **	0.017 *	0.014 **	0.013 *
lnPGDP×W	-	-	-	0.147 **
lnSTR×W	-	-	-	0.236 **
lnURB×W	-	-	-	0.007 *
lnFDI×W	-	-	-	−0.018
lnFIN×W	-	-	-	0.003
Spatial error term	-	0.166 ***	0.214 **	0.232 ***
Space effect	Control	Control	Control	Control
Time effect	Control	Control	Control	Control
R2-ad	0.721	0.697	0.752	0.773
Log-likelihood	-	209.342	203.925	228.649

Note: OLS is short for ordinary least squares, and it is compared here with the spatial measurement results. ***, **, and * are significant at the 1%, 5%, and 10% confidence levels, respectively (and this is the same for the below).

, the per capita GDP, industrial structure, urbanization, foreign direct investment, and land finance all passed the significance test. Then, we decomposed each indicator, as shown in

Table 6. Decomposition of the spatial effects of influential factors.

Variables	Direct Effect	Indirect Effect	Total Effect
lnPGDP	0.129 ***	0.075 ***	0.204 ***
lnSTR	0.413 **	0.248 **	0.661 **
lnURB	0.023 ***	0.017 **	0.040 **
lnFDI	−0.034 *	−0.023	−0.057 *
lnFIN	−0.012 **	−0.003	−0.015 *

Note: ***, **, and * are significant at the 1%, 5%, and 10% confidence levels.

.

The direct and indirect effects of economic development level on ULUE are both positive and pass the 1% significance test, thus implying that the economic level significantly improves ULUE and has a spatial spillover effect. Economic growth determines the number of resource inputs on the urban land per unit area. With economic development, the environmental protection technology and management level in the production process are improved, which brings significant positive effects to ULUE. Rapid economic growth drives the economic level of adjacent cities to improve, thus improving the ULUE of other cities.

The direct and indirect effects of industrial structure on ULUE are positive and significant at the level of 5%. This, thus, declares that the industrial structure positively affects ULUE, and the spatial spillover effect is significant. Industrial upgrading contributes to driving the shift of production factors to tertiary industries, and it also plays a role in adjusting land use structure. Clean industries promote the more reasonable utilization of urban land, thereby indirectly reducing environmental costs and improving land use efficiency. Industrial transformation will strengthen the technical exchanges between cities, establish a good industrial correlation, and better promote regional cooperation, thus enhancing the industrial upgrading of surrounding areas and improving ULUE.

The direct effect of urbanization on ULUE is positive and significant at the 1% level. Urban infrastructure has increasingly improved, and many rural people have flowed into the city; thus, the size of cities has gradually expanded. To improve the quality of urbanization, the management of construction land has been strengthened, the urban layout has been planned with greater rationality, and ULUE has improved. The indirect effect of urbanization is positive and significant at the 5% level, thus proving that urbanization can drive the improvement of ULUE in adjacent cities. The main reason for this is that there are links between cities and the relatively dense urban spatial network, which affects land use efficiency by driving the surrounding urbanization process.

The direct impact of foreign direct investment is negative and significant at the 10% level. Under the mode of long-term investment driven economic growth, foreign capital is a crucial way through which to guide economic development. In introducing foreign capital, local governments attract investors to enclosure land in the form of land price reduction and preferential treatment, and the quick expansion of industrial land causes the inefficient use of construction land. Although foreign direct investment drives economic development, the negative effects will reduce ULUE at present.

The direct effect of land finance on ULUE is negative and significant at the level of 5%. With an overall tightening of urban construction land, rational land planning is the basis for development. Land finance is an important means for local governments to develop their economies, but arbitrary regulation of land prices by the government can affect market operations and the allocation of construction land, thus leading to inefficient land utilization.

4.3. Analysis of Regional Heterogeneity

To explore the effect of numerous variables in diverse regions with respect to ULUE, urban samples were grouped and compared according to the upstream, midstream, and downstream areas, the results of which are shown in

Table 7. Effects of spatial decomposition by region.

Variables	Upstream			Midstream			Downstream
	Direct Effect	Indirect Effect	Total Effect	Direct Effect	Indirect Effect	Total Effect	Direct Effect	Indirect Effect	Total Effect
lnPGDP	0.158 ***	0.123 ***	0.281 ***	0.134 ***	0.115 ***	0.249 ***	0.122 ***	0.104 ***	0.226 ***
lnSTR	0.311 ***	0.223 ***	0.534 ***	0.426 ***	0.231 **	0.657 ***	0.332 **	0.265 **	0.597 ***
lnURB	0.031 *	0.013 **	0.044 *	0.106 ***	0.026 *	0.132 *	0.036 **	0.015 **	0.051 **
lnFDI	−0.041	−0.022	−0.063	−0.052 **	−0.027	−0.079	−0.021 *	−0.028	−0.049 *
lnFIN	−0.102 *	−0.124	−0.226	−0.024 *	−0.008	−0.032 *	−0.083 **	−0.032 *	−0.115 *

Note: ***, **, and * are significant at the 1%, 5%, and 10% confidence levels.

.

The economic development level has significantly positive direct and indirect impacts on the ULUE of the upstream, midstream, and downstream areas; however, the promotion effect does gradually weaken from the upstream regions to the downstream areas. The reason for this may be that as the economies of downstream areas develop, many enterprises and high-tech talents will be attracted the region. Under the existing infrastructure level, a congestion effect has been shown, and the limited land resources have made the marginal impact of economic growth on urban land use slow down. Comparatively, the upstream economy has a huge potential for economic growth, and it is more likely to drive the reasonable utilization of urban land.

Industrial structure has a significantly positive direct and indirect effect on the ULUE of the upstream, midstream, and downstream regions. The downstream region consists of mainly clean industries such as high-tech industry and service industry. The midstream area is dominated by industry and is relatively well developed. The upstream region’s industry is in the catch-up stage, and the agglomeration scale is small. Although the industrial structure varies from region to region, they are all influenced by land use structure. The transformation of industry to an advanced level is conducive to improving economic output and alleviating the environmental pollution caused by industrial land usage, which has a promoting effect on improving ULUE.

The existing urbanization has had significantly positive direct and indirect influence on the ULUE of the upstream, midstream, and downstream regions, but the effect is the most obvious in the midstream area. The urbanization in the midstream region is growing at a faster rate, and the urban land has been rationalized and reconfigured to meet urbanization. The city cluster effect in the downstream area is obvious, and the internal exchanges between cities are more frequent, but the urban development is approaching saturation, and the change in available land is not large. The urbanization level in the upstream region is relatively low, and the land has not been fully utilized.

Foreign direct investment has a significantly negative direct influence on the ULUE of the midstream and downstream regions. It shows that foreign direct investment is not conducive to improving the ULUE of local cities. Its direct and indirect influence on the ULUE of the upstream area was not found to be significant. The reasons for this are that the location advantages of the upstream region of the YREB are not obvious, and foreign investment is limited. Though this region has also been unable to obtain advanced technical support and management experience, it has little impact on ULUE.

The direct influence of land finance on the ULUE of the upstream, midstream, and downstream regions is significantly negative, and the indirect influence on the downstream area is significantly negative. Local governments manage urban land development and spatial expansion through land finance. In addition, they also regulate land prices, but this leads to the emergence of urbanization land imbalance, and irrational land use reduces land use efficiency.

5. Discussion

In this study, we aimed to discuss the spatio-temporal features of ULUE and its influential factors in the YREB. We found that the ULUE in the YREB has steadily improved. This conclusion is consistent with existing studies that have confirmed that ULUE in the YREB is in the process of growth [18,46]. However, our findings also show that the gaps of efficiency among the upstream, midstream, and downstream regions are gradually increasing. According to the above analysis, the ULUE in the upstream, midstream, and downstream regions have increased by 0.297, 0.131, and 0.360, respectively. The upstream region is growing steadily, the midstream is growing slowly, and the downstream is experiencing large fluctuations. As a result, the midstream region’s slow efficiency growth has led to widening regional disparities.

There are obvious spatial features of ULUE in the YREB [47]. The spatial distribution pattern of “core- periphery” is clear. There are three urban agglomerations in the upstream, midstream, and downstream regions of the YREB, which are the national urban agglomerations centered on the core cities. The core cities have kept a prominent status in terms of both economic and social influence [48,49]. The space distribution characteristics of ULUE are in line with the layout law of urban agglomeration, and the efficiency of the core cities is significantly higher than that of others. Therefore, in the future, the core cities should play a pioneering role and drive the surrounding cities to improve their ULUE. From the LISA map, H-H agglomeration and L-L agglomeration are the major space cluster patterns of ULUE in the YREB. The H-H agglomeration pattern is mostly in the downstream area, and the L-L agglomeration pattern is mostly in the midstream region. The results are in accordance with the spatial distribution characteristics of ULUE.

There are many factors affecting ULUE. We verified that the economic level, industrial structure, and urbanization significantly promoted the improvement of ULUE, which is supported by many studies [50,51,52]. This study found that the foreign direct investment and land finance inhibited ULUE. This finding is contrary to the conclusions of existing research [53]. They used a single economic indicator to measure ULUE. The results obtained do not include the impact of unexpected output, such as environmental pollution. In fact, it has been proven that foreign direct investment contributes to promoting economic development [54]. However, most foreign capital is invested in high-pollution and energy-consuming industries, and there is a risk of pollution transfer [55], so pollution factors should be considered. This paper fully considered the pollution discharge carried by land, such as wastewater, carbon dioxide, sulfur dioxide, etc., and the conclusion obtained was closer to the actual development situation. As for land finance, its negative effect on urban land has been supported by many scholars [56,57]. However, Zhong et al. (2022) explored the nonlinear effect of land finance, and they believed that its effect on ULUE presents an “inverted U-shape” [58]. In other words, when land finance exceeds a certain threshold, it will inhibit the improvement of ULUE. This conclusion also provides ideas for follow-up research, where we could analyze the effect from a non-linear perspective.

6. Conclusions, Implications, and Limitations

6.1. Conclusions

This paper adopted the undesirable SBM model to measure the ULUE of 111 cities in the YREB from 2010 to 2022. Spatial correlation analysis and the spatial econometric model were applied to analyze the spatio-temporal features of ULUE and its influential elements. We were thus able to draw the following conclusions:

(1) The ULUE in the YREB showed a significant upward tendency from 2010 to 2022. In 2010, the difference between the highest and lowest values of the ULUE in the upstream, midstream, and downstream areas was 0.147, and 0.325 in 2022. The disparity between the regional efficiency values is becoming increasingly obvious.

(2) There is a significant regional heterogeneity in the ULUE in the YREB, with more high-efficiency cities in the downstream than in the midstream and upstream regions. The regions gradually form a “core-periphery” spatial differentiation pattern dominated by core cities. The number of high-efficiency cities in the downstream region gradually increase, while the efficiency values of midstream cities remain low.

(3) ULUE is dominated by “High-High agglomeration” and “Low-Low agglomeration”, with significant regional features. The high-value agglomeration area was mainly concentrated in the downstream area of the YREB, and the low-value agglomeration area was mainly in the midstream region.

(4) The existing level of economic development, industrial structure, and urbanization are having a positive impact on ULUE, and the positive spatial spillover effect is obvious. However, foreign direct investment and land finance have inhibited the improvement of ULUE.

6.2. Implications

In response to the findings of the study, the implications are summarized as follows. In terms of the YREB as a whole, the government should unswervingly develop the economy, rely on the major strategic advantages of the YREB, and make it a smooth domestic and international double circulation aorta. Advanced industrial structure is a crucial way through which to improve ULUE. Local governments need to set strict industrial access standards, inspire the development of clean industry, and strengthen the role of innovation to accelerate the green transition of industries and boost the efficiency of resource utilization. Urban decision-making departments should formulate differentiated land-use planning and policies according to different influential factors. It is a requisite to improve the investment quality and prevent the transfer of pollution. Under the conditions of market economy, the government should reduce unreasonable intervention and respect the law of market operation.

From the perspective of the three major regions, the ULUE in the lower reaches area is relatively high. It is necessary to strengthen the radiation role of the downstream cities, realize regional coordination and links, and drive the efficiency improvement of other regions. The middle reaches area should carry out reasonable planning land utilization, make full use of the stock of land, effectively control urban disorderly expansion, and relieve land pressure. Cities with low land use efficiency should gradually reduce the proportion of polluting industries, absorb technology and knowledge spillover from downstream areas, and guide industries toward advanced development. The upper reaches region should increase vitality, improve the absorption capacity of production factors through preferential policies, optimize the investment environment, upgrade the industrial structure and other measures, as well as enhance the industrial layout while achieving economic catch up, a reasonable planning of land utilization, and improving ULUE.

6.3. Limitations

Although we have made efforts to improve the analysis in this paper, there are still some limitations. First, the study on the formation mechanism of the spatial differentiation of ULUE is not deep enough, and the indicators of influential factors need to be improved, which will be explored in the subsequent research. Second, there are uncovered areas in the study of ULUE at the municipal level, which will have a certain effect on the overall evaluation of the ULUE of the YREB. As such, a refinement of this study at smaller scales, such as at the county level, could enhance the coverage of the samples, and the results obtained would be more scientific. Therefore, an in-depth discussion of small-scale ULUE and its influential factors will be a crucial part of our future research. Third, taking the YREB as an example, we analyzed the influential factors of ULUE. Owing to different economic systems and development degrees, whether the influential factors of ULUE are consistent in other different river basins in the world was not analyzed. Future studies could explore the ULUE in other river basins and their influential factors.