Research on the Spatio-Temporal Characteristics and Influence Path of High-Quality Economic Development from the Perspective of Urban Land Transfer

Abstract

The transfer of urban land is an important means for the government to optimize the allocation of resources and promote economic development, and its impact on high-quality economic development (HQED) in the new era is worthy of attention. Based on the panel data of 108 cities in the Yangtze River Economic Belt from 2004 to 2017, the entropy method and panel regression model are used to analyze the direct, indirect, and non-linear effects of land transfer on HQED. The study found that: (1) The HQED level of the Yangtze River economic belt increased steadily from 2004 to 2017, but the overall level was low, showing the spatial characteristics of high downstream and low middle and upstream. (2) Urban land transfer (ULT) has a significant positive direct impact on HQED, especially in the upper and middle reaches, while the indirect impact is more beneficial to the downstream areas. The impact of industrial structure upgrading (uis) and urbanization (urb) on HQED is significantly negative in the upstream area, while significantly positive in the downstream area. (3) There is a significant threshold effect in ULT, and its positive impact on HQED decreases with the expansion of the threshold scale, but increases in the upstream and midstream areas. The overall impact of uis and urb shows inverted “U” characteristics and significant differences in different areas. The conclusion is that the government should first further reform the land resource management system, and then improve the efficiency of construction land use, to reduce the dependence on land resources, and finally promote the upgrading of industrial structure and improve the quality of urbanization.

1. Introduction

In recent years, the downward pressure on China’s economic operation has been increasing, the external environment has been deteriorating, and the traditional factor and investment-driven model has been weak [1]. For a long time, economic development has been carried out in a high-speed and extensive mode, resulting in the adverse situation of low quality of economic development and severe damage to the ecological environment such as “low efficiency, high pollution and uncoordinated” [2,3]. In this regard, China’s economy has shifted from the stage of high-speed growth to the stage of high-quality economic development, reflecting the new development concept of innovation, coordination, green, open, and sharing. According to the connotation, HQED is a development model that pursues good social and economic benefits and emphasizes excellent ecological and environmental benefits. As the space carrier of all activities, the contribution of land elements and land systems to China’s rapid economic development is self-evident [4,5]. In addition, compared with other developing countries, the unique land system adopted by China has significant advantages in promoting economic growth. The key lies in the government’s rapid industrialization by attracting investment from low-cost land and accelerating urbanization by land capitalization [6]. However, experience shows that in the past few decades, the government relies on urban land transfer (ULT) to achieve rapid land capitalization, industrialization, and urbanization, breaking the balance between the socio-economic and ecological environment systems [7,8]. With China pursuing high-quality economic development, the traditional unsustainable, uncoordinated, and unhealthy economic development model has been abandoned, and exploring a new land management model has become a trend. Meanwhile, the contribution of government-led ULT to economic growth has also been given a new evaluation criterion, which has become a trend. Re-evaluating the relationship between ULT and HQED is important for promoting the transformation of traditional development mode and land management mode.

From the perspective of the evaluation of HQED, some scholars use total factor productivity indicators to measure the level of HQED [9,10]. Many scholars believe that HQED is a systematic and comprehensive concept that should be evaluated by a comprehensive index system made up of different dimensions [11,12,13]. The comprehensive index system has been widely adopted because it can better meet the research needs and has the advantages of rich connotation, comprehensive indicators, and a straightforward process. As for the evaluation index system, it was constructed by different scholars for the research content is different. However, most scholars consider that the HQED index system should be constructed from the five dimensions of “innovation, coordination, green, openness and sharing” [13]. Although the existing index system can better reflect HQED, there are still some indexes that could be improved.

On the one hand, the social and economic indicators at the municipal level are more than enough, while the ecological environment indicators need to be more. In pursuing HQED and ecological civilization construction in China, the status of the natural ecological environment is becoming more and more prominent. The indicators reflecting “green” should be limited to treating industrial wastes and comprehensively consider the level of carbon emissions, air quality, urban greening, and other indicators. Secondly, there is little consideration of the “healthy” operation state between economy and ecology. In most studies, the “coordination” dimension indicators mainly consider regional and urban-rural coordination and lack of consideration of the coordination between man and nature [14]. Therefore, this paper further enriches the understanding of the connotation of “coordinated development” in high-quality economic development, and further introduces the coordinated development between economy and environment on the basis of the coordinated development of social economy. This can expand and enrich the evaluation system of HQED, so as to evaluate the HQED level of urban economy more scientifically and objectively.

Clarifying how ULT influences HQED is an essential prerequisite for promoting the transformation of the development and land management modes. From the perspective of the research on ULT and HQED, direct research on the two is relatively scarce. In contrast, the related research on ULT and economic growth, land factor allocation, land finance, and economic development are relatively prosperous. From limited studies, it is found that government-led ULT can promote industrialization and urbanization, thus promoting rapid economic growth [6]. Further study indicate that urban land still accounts for a large proportion in China due to the marketization of land factors and land finance. However, some scholars have found that ULT has a dynamic influence on HQED, which changes with the level of marketization [15]. Thus, the relationship between government-led ULT and HQED is complex. Current research on ULT and HQED needs further exploration. On the one hand, the evaluation of HQED needs to consider the coordination of social economy and ecological environment. On the other hand, the nonlinear impact of ULT and its influencing path on HQED should be considered. Thus, according to the understanding of the connotation of HQED of “innovation, coordination, green, open and sharing”, we construct the evaluation index system of HQED of the prefecture-level city by combining with the theory of system coupling and coordination. Then, the mediating effect model and the panel threshold model are further introduced to analyze the mechanism and nonlinear characteristics of ULT for HQED, which makes up for the limitation of the analytical perspective in the existing research. Finally, relevant policy recommendations are put forward to promote HQED and land management reform in the Yangtze River Economic belt.

2. Theoretical Analysis and Hypothesis

2.1. Literature Review

2.1.1. High-Quality Economic Development

The quality of economic development is a global issue, and the discussion about it has never stopped in the world. Soviet economist Kamaev first defined the concept of economic growth quality in 1977, stating that the quality of economic growth includes an increase in both living and production materials, an improvement in product quality, an increase in production efficiency, and an increase in consumer goods [16]. Since then, scholars around the world have been discussing and researching the content and evaluation of economic growth and development quality. In terms of content, Barro believes that economic growth can be divided into quantity and quality, and emphasizes that the quality of economic growth is a broad concept that is closely related to various aspects of economic development, political systems, and income distribution [17]. Schumpeter and Backhaus believe that economic growth is closely related to income growth, and to enrich the content of economic growth, micro factors such as capital accumulation, population, and resource allocation efficiency should be included in the scope of economic growth [18]. Mlachila et al. proposed an economic growth index for developing countries, believing that high-quality economic development is different from general economic development, and should not only be reflected in the improvement of the economic growth rate, but also in the improvement of the welfare level of all social residents [19]. For China, pursuing high-quality economic development has become a key focus of the Chinese government. The report of the 19th National People’s Congress of China publicly stated that high-quality development is a development that requires less input from production factors, high efficiency in resource allocation, low environmental costs, and good economic and social benefits. Some scholars have also put forward different opinions, such as Liu and Chen, who believe that the connotation of high-quality development includes demand, supply, economic cycle, input-output, resource allocation, and income distribution [20]. However, people believe that the concepts of “economic growth quality” and “high-quality economic development” are both related and different. The commonality between them is that both emphasize that economic growth cannot simply pursue expansion in terms of quantity and scale, but must also be fully reflected in quality aspects such as structural optimization, efficiency improvement, welfare distribution improvement, and environmental cost reduction [21]. Similarly, Yan et al. believe that high-quality economic development should be comprehensively considered from three aspects: ecological environment, economic benefits, and economic structure [22]. So far, for the quality of economic development, it is still at a single research level, mostly summarizing the inherent meaning of economic quality as the multi-dimensionality of economic development, including the quantity and quality connotation of products.

2.1.2. Urban Land Transfer

Land, as the spatial carrier of all production activities, makes an indelible contribution to social and economic development. Neoclassical economists believe that land is one of the core factors that promote economic growth [23]. Land transfer, as the main way of land factor flow, plays a fundamental role in the high-quality development of the economy. From the current research content, land transfer mainly includes the two major parts of urban construction land transfer and agricultural land transfer [24,25]. In most cases, urban construction land has a more significant and far-reaching impact on urban economic growth and high-quality development. Overall, some scholars directly studied the contribution of construction land to economic development. From a static perspective, Mo et al. found that in Chinese county-level cities from 2009 to 2014, when the revenue from urban land transfer was high, their economic growth rate was greater [26]. From a dynamic analysis perspective, Yan and Yang found a significant U-shaped non-linear relationship between land circulation marketization and urban high-quality economic development by using the region development and livelihood index published by China as a substitute for the level of economic high-quality development [15]. From a specific land type and spatial perspective, Zhang et al. found that the continuous increase of urban transportation land transfer is essential to ensure sustainable economic growth and adjacent transportation land transfer has a positive spatial spillover effect, from the perspective of urban transportation land transfer [24]. In addition, some scholars indirectly analyzed the impact of construction land transfer on other factors that affect economic growth through econometric models. For example, the dynamic impact of urban construction land circulation on new urbanization [27], the impact of land transfer marketization on provincial green total factor productivity in China [28], and the impact of construction land transfer methods and scales on local fiscal revenue [29]. In summary, high-quality economic development is a complex dynamic problem that includes elements of the social-economic-ecological tripartite system. Although a large amount of evidence shows that urban land transfer has a positive impact on high-quality economic development, this positive impact also has uncertainty in terms of temporal and spatial scales. A small amount of research has further analyzed the impact path of urban land transfer on high-quality economic development from different perspectives such as urbanization, land finance, and industrial structure, but a comprehensive analysis of the impact of urban land transfer on high-quality economic development from multiple dimensions is still relatively scarce.

2.2. Hypothesis

2.2.1. The Total Effect of ULT

In the stage of rapid economic development, the government obtains financial revenue through ULT. It invests land revenue in urban infrastructure construction to further improve the investment environment and attract foreign investment. Secondly, the government attracts foreign investment and undertakes industrial transfer in developed areas through cheap land to improve the level of industrialization and urbanization and ultimately achieve economic growth [7]. In this process, the large-scale “bidding, auction and listing” transfer meets the local government’s goal of “seeking development by cheap land”, and the government-led ULT significantly promotes rapid economic development by promoting industrial development and urbanization. Industrialization, population growth and agglomeration, and the improvement of infrastructure and urbanization have further expanded the demand for land and ultimately stimulated the expansion of the scale of ULT. However, the social, economic, and ecological problems caused by rapid economic development have forced local governments to adjust their industrial development strategies and control the scale of urban expansion, gradually shifting from focusing on economic growth to HQED. As the social economy enters a high-level stage of development, the upgrading of industrial structure (uis) and the development of urbanization (urb) gradually slow down, economic development increasingly relies on the promotion of technology and talents, the dependence on land elements gradually decreases, and the impact of the ULT on HQED weakens. Thus, we put forward Hypothesis 1:

 H1:

Government-led ULT directly promotes HQED but gradually weakens with the improvement of the economy.

2.2.2. Upgrading of the Industrial Structure (uis), Urbanization and HQED

The uis represent one of the directions of HQED, and observing the impact of the ULT on the transformation of local industrial structure is also the fundamental perspective to analyze the HQED of a regional economy. Under the dual pressure of finance and promotion, local governments first regard land as the primary source of extra-budgetary finance and then transfer many incredibly cheap lands to construct industrial parks to realize industrialization [30]. According to statistics, from 2005 to 2017, the average annual supply of industrial and mining storage land in China is as high as 140,000 hectares. This also proves the intention of local governments to promote industrial development through land supply and transfer. Under the government’s low-price land supply strategy and the policy of “building nests to attract phoenixes”, local governments have introduced a large number of “bees” (low-level manufacturing) rather than “phoenixes” (high-level manufacturing). However, compared with the traditional industrial structure, many low-level manufacturing industries in the short term has increased local fiscal revenue and led to economic growth. But the drawbacks of backward industries are very prominent, such as low technology level, high energy consumption, low output efficiency and severe environmental pollution, which seriously inhibit the HQED. With the progress of the social economy and science and technology, the development concept of local governments has also changed. ULT is increasingly inclined to more technological, environmentally friendly, and efficient enterprises.

Government-led ULT is one of the critical factors of urbanization. The improvement of the urbanization level profoundly impacts the urban economy, society, and environment in both tangible and intangible aspects. The government promotes the rapid development of urban industrialization by selling land at a low price during rapid urbanization [31,32]. On the one hand, the impact of the ULT on urbanization lies in the promotion of urban industrialization by low-price transfer, which attracts a large number of the agricultural labor force to “work” in cities through non-agricultural employment opportunities and indirectly promotes population urbanization. On the other hand, the construction of industrial parks and new urban areas directly promotes the urbanization of land, leading to the large-scale outward expansion of urban built-up areas. During economic growth, the urbanization level will also effectively promote improving innovation. High-level urbanization has a more substantial cumulative effect on innovation capability than low-level urbanization and has a more pronounced impact on HQED [2,33]. But the disorderly expansion has become a thorny social problem, especially in the cities with a low level of urbanization, economy, and technology. The ULT has attracted many low-level manufacturing industries, and the expansion of urban scale has directly promoted economic development, resulting in ecological and environmental problems. With the improvement of social and economic levels, urbanization has changed from low to high. High-quality urbanization has become more and more attractive to talents, technology, and capital, which is beneficial to HQED. Thus, we propose Hypothesis 2:

 H2:

ULT impacts HQED by affecting the uis and the urb, and this impact has significant nonlinear characteristics.

2.2.3. Regional Heterogeneity of the Influence of ULT on HQED

The scale of ULT is limited by the construction land index of the superior government. In different regions and stages of development in China, the central government’s allocation of urban construction land index has apparent regional heterogeneity [34]. Due to convenient transportation and foreign trade conditions, the downstream area of the Yangtze Economic Belt has obtained many construction land indicators in the early stage. The scale of ULT is significantly higher than that of the middle and upper reaches of the Yangtze Economic Belt, which leads to the early industrial upgrading and urbanization in the downstream area, resulting in the level of HQED being much higher than that of the middle and upper reaches. However, because the regional endowment of the middle and upper reaches lags that of the downstream areas, the construction land in the middle and upper reaches is low. The scale of ULT is small, and the role of the government in promoting HQED through ULT is far lower than that of the downstream areas, which indirectly leads to the long-term lower level of HQED in the middle and upper reaches. With the widening development gap and the change in development policies, the central government has gradually increased its support for the middle and upper reaches. The land use indicators have gradually tilted to the middle and upper reaches, which has gradually expanded the scale of ULT in the central and Western cities, and accelerated the government’s promotion of HQED through ULT. The dependence of economic development of the downstream area on land elements is low owing to the developed economy and the abundant elements of high-tech industries, technology, talent, and capital in the downstream area. Thus, the impact of the ULT on HQED is weak in the downstream area. Then, we propose Hypothesis 3:

 H3:

The impact of the ULT on HQED in the downstream, midstream, and upstream is significantly different, and this impact is more evident in the middle and upper reaches.

3. Study Area, Materials and Methods

3.1. Study Area

The Yangtze River Economic Belt in China, which covers 11 provincial administrative regions and spans the east, middle and west, is the golden economic belt with the longest depth, the widest coverage and the greatest influence in China. it is also the leading area to promote the formation of complementary and high-quality development in China (Figure 1). The Chinese president has focused on the future development direction of the Yangtze River Economic Belt three times, and proposed to build the Yangtze River Economic Belt into the main force leading high-quality economic development. According to the China Statistical Yearbook, the Yangtze River economic belt contributed more than 50% of the country’s economic growth in 2021, and its total population accounted for 42.9% of the country’s total population. From 2004 to 2018, the area of urban construction land in 11 provinces and cities of the Yangtze River Economic Belt continued to grow. In 2018, with the exception of Guizhou Province, the area of urban construction land in other provinces and cities exceeded 1000 square kilometers, and the area of construction land in Jiangsu Province even exceeded 4000 square kilometers. The synchronous characteristics of economic growth and the expansion of construction land are particularly significant in the Yangtze River economic belt. Land transfer, as the main way of construction land expansion, exploring its influence mechanism on economic growth is of great significance for the reform of the land system and reducing land financial dependence in the Yangtze River Economic Belt. It is also one of the urgent needs of the current region to promote high-quality economic development.

3.2. Model and Method

3.2.1. Entropy Method

According to the connotation of the concept of HQED and the availability of relevant data, and referring to the current research results, this paper focuses on the overall effectiveness [11,35], coordinated development [12,36], innovation promotion [36,37], achievement sharing [38] and sustainable development [22] to construct the evaluation index system of HQED of 108 cities in the Yangtze River Economic belt from 2004 to 2017. Details of the indicators are shown in

Table 1. Evaluation Index System of HQED.

Subsystem	Indicator Layer	Calculation Method	Attribute	Weight
Overall effectiveness	GDP per capita	GDP/total population	+	0.0932
	Per capita fiscal revenue	Fiscal revenue/total population	+	0.0376
	Coefficient of investment effect of fixed assets	Newly increased GDP/investment in fixed assets in the same period	+	0.0424
Coordinated development	Income ratio of urban and rural residents	Disposable income of urban residents/disposable income of rural residents	−	0.1792
	Level of foreign capital utilization	Foreign capital actually used in the year/GDP	+	0.0449
	Coordination of economy and environment	Total discharge of industrial wastewater, waste gas and solid waste/GDP	−	0.0520
Innovation promotion	R & D investment intensity	R & D expenditure/GDP	+	0.0915
	Intensity of investment in education	Education expenditure/fiscal expenditure	+	0.0641
	Innovation output level	Number of three types of patents granted/total population	+	0.0303
Sharing of results	Level of transport facilities	Total road area/total population	+	0.1072
	Medical and health level	Number of beds in medical and health institutions/total population	+	0.0316
	Internet penetration rate	Internet users per 100 population	+	0.0501
Sustainable development	Level of urban greening	Green area of built-up area/built-up area	+	0.0657
	Urban Air Quality Level	Daily Average Air Quality Index	−	0.0565
	Intensity of environmental regulation	Average treatment rate of industrial waste water, waste gas and solid wastes	+	0.0537

.

Referring to Wu et al., the entropy method with a time variable is used to measure the HQED level of the regional economy [39]. The specific steps are as follows:

x ₊ ’_θij = x_θij/x_max; x_-’_θij = x_min/x_θij

(1)

y_θij = x’_θij/ΣθΣ_ix’_θij

(2)

e_j = −kΣ_eΣ_iy_ijln(y_θij)

(3)

w_j = g_j/Σ_jg_j

(4)

HQED_θi = Σ_j(w_jx’_θij)

(5)

where, x_θij is the value of indicator “j” of city “i” in year ”θ”; x ₊ ’_θij is the standardized value of the positive indicator; x_-’_θij is the standardized value of the negative indicator; y_θij is the weight in local year; e_j is the entropy of index j, so g_j can be calculated using 1-e_j; w_j is the weight of index j; HQED_θi is the value of high quality economic development level (Figure 1).

3.2.2. Benchmark Regression Model

According to the characteristics of panel data and the needs of research content, a benchmark model is constructed to analyze the direct impact of the ULT on HQED. The model is as follows:

HQED_it= ∂ +β₁land_it +β₂uis_it + β₃urb_it + β₄Controls_it +μ_i +σ_t +ε_it

(6)

where i is the city, t is the time; HQED is the explained variable, representing the level of HQED; ∂ is the constant term. The coefficient of the explanatory variable is β; the land is the core explanatory variable, which represents the ULT; uis and urb are the two key explanatory variables, which represent the level of upgrading of industrial structure and urbanization respectively. μ_i represents the personal effect, σ_t is the time effect, and ε_it is the error term.

3.2.3. Mediating Effect Model

On the basis of theoretical analysis, this paper analyzes the indirect impact mechanism of ULT on HQED from two aspects of uis and urb. We build an intermediary effect model to verify it [27]. The mediating effect model is as follows:

HQEDit = ∂1 +α1landit +α2Controlsit +μi +σt +εit

(7)

Mit = ∂2 +β1landit +β2Controlsit +μi +σt +εit

(8)

HQEDit = ∂3 +γ1landit +γ2Mit +γ3Controlsit +μi +σt +εit

(9)

where M_it is the mediator variable: uis and urb; α, β, and γ represent the coefficients of the variables.

3.2.4. Panel Threshold Model

From the theoretical analysis, we can see that the impact of the ULT on HQED has nonlinear characteristics. Based on this, this paper further adopts the threshold regression model proposed by Hansen and constructs the panel single threshold model (Formula (10)) and multi-threshold model (Formula (11)) [40,41] based on the research of Chen et al. The model is as follows:

$$HQE{D}_{it}={\delta }_0+{\delta }_1{x}_{it}I(T_{it}\le {\lambda }_1)+{\delta }_2{x}_{it}I(T_{it}>\lambda )+{\beta }_{n}Control{s}_{it}+{\mu }_i+{\sigma }_t+{\epsilon }_{it}$$

(10)

$$\begin{array}{l}HQE{D}_{it}={\delta }_0+{\delta }_1{x}_{it}I(T_{it}\le {\lambda }_1)+{\delta }_2{x}_{it}I({\lambda }_1<T_{it}\le {\lambda }_2)\\ +{\delta }_3{x}_{it}I(T_{it}>{\lambda }_3)+{\beta }_{n}Control{s}_{it}+{\mu }_i+{\sigma }_t+{\epsilon }_{it}\end{array}$$

(11)

where δ is the coefficient of the explanatory variable X_it to the explained variable HQED_it when the threshold variable T_it is greater than or less than the threshold λ; I is the indicator function, and I = 1 when T_it ≤ λ. Otherwise, it is equal to 0. Among them, the threshold variable of this paper is the scale of ULT (land) which lags two periods. The variables involved in this paper are shown in

Table 2. Variable description.

Variable	Attribute	Explanation	Acquisition	Data Source
HQED	Explained variable	The high-quality economic development level of city.	Calculated by using the index in Table 1 and formula 1–5.	-
land	Core explanatory variable/threshold variable	Total scale of urban land transfer of city.	Sum of different forms of land transfer area, and with a lag of two years.	Statistical Yearbook of Land and Resources of China
uis	Mediating variable	Industrial structure upgrading index of city.	The ratio of the tertiary industry’s output value to the secondary industry’s output value	China Urban Statistics Yearbook; EPS global database
urb	Mediating variable	The level of urbanization of city.	The ratio of the urban built-up area to the total urban area	Statistical Yearbook of Chinese cities
gov	Control variable	The degree of government intervention of city.	The proportion of local fiscal expenditure to GDP	EPS global database
emp	Control variable	The employment level of city.	The proportion of employment to total population	EPS global database
open	Control variable	The level of the city’s opening to the outside world.	Foreign direct investment to GDP	EPS global database
dep	Control variable	The level of economic development of city.	Total retail sales of social consumer goods	EPS global database

.

3.3. Description of Variables

3.3.1. Explained Variable

HQED. Based on the concept of HQED and the existing research foundation, this paper expands and constructs a new evaluation index system to calculate the HQED of 108 cities in the Yangtze River Economic belt and takes it as the explanatory variable.

3.3.2. Explanatory Variables

• Core explanatory variable and threshold variable. Existing studies generally characterize the government-led ULT by the full scale of ULT, the scale of “bidding, auction and listing” transfer or the scale of agreement transfer [42]. With the development of the social economy, the behavior of urban ULT is becoming more and more common, and the way of ULT is mainly divided into “bidding, auction and listing” transfer and agreement transfer according to different uses. In order to comprehensively analyze the impact of government-led ULT on HQED, we take the full scale of ULT as the core explanatory variable (land). We select the land as the threshold variable to analyze the non-linear relationship between ULT and HQED. Due to the lag of social and economic construction after the ULT, all the analysis models in this paper treat the ULT variable with a lag of two years before operation.
• Mediating variables. industrial structure upgrading (uis) and urbanization (urb). Based on theoretical analysis, on the one hand, this paper analyzes the indirect impact of government ULT on the HQED from the perspective of uis and urb; On the other hand, this paper analyzes the nonlinear impact of ULT, uis and urb on HQED under different levels of ULT scale. Therefore, the intermediates variable of this paper are the uis and urb.

3.3.3. Control Variables

Based on the existing research [43,44], the degree of government intervention (gov), the employment level (emp), the degree of opening to the outside world (open) and the level of economic development (dep) are taken as the control variables.

3.4. Data Sources

The primary data are the socio-economic panel data and the vector administrative boundary data of 108 cities in the Yangtze River Economic belt from 2004 to 2017. Among them, socio-economic data mainly come from EPS global database, China Urban Statistics Yearbook, China Environmental Statistics Yearbook and China Land and Resources Statistics Yearbook. As China Land and Resources Statistics Yearbook is only updated to 2018, our research cannot explore more recent years. The vector boundary data are from the 1:4 million databases of the National Fundamental Geographic Information System of China.

4. Results

4.1. Spatio-Temporal Characteristics of HQED

According to Table 1 and the improved entropy method, the HQED levels of 108 cities (Figure 2a) were calculated, and the natural breakpoint method was used to divide them into six categories from low to high. On this basis, the coupling coordination degree model is further used to calculate the coupling coordination degree (D) of ULT and HQED levels in each city (Figure 2b). Referring to Wang et al., it is divided into five intervals, representing severe imbalance, imminent imbalance, prior coordination, intermediate coordination, and advanced coordination from low to high [44]. In order to facilitate comparison, this paper mainly shows the specific results of 2004, 2010 and 2017. It can be seen from Figure 1 that the city-level HQED is in a steady upward trend in terms of time, but the overall level is still low. The HQED of most cities was lower than 0.42 in 2010 and 0.55 in 2017. The difference is that the coupling coordination degree of ULT and HQED is in an overall upward trend from 2004 to 2010, reaching the primary coordination state, while it declines from 2010 to 2017, and some cities in the downstream region decline from the advanced coordination state to the intermediate coordination state. From the perspective of space, the level of HQED and the level of coupling coordination show a decreasing trend from east to west. The high-value areas are concentrated in the Yangtze River Delta region, consistent with the characteristics of the economic level.

4.2. Direct Impact Analysis

According to the steps of panel regression analysis and related tests, this paper finally uses the fixed effect model for benchmark regression analysis. The regression results are shown in

Table 3. Benchmark regression results.

Variables	Global	Upstream	Midstream	Downstream
land	0.111 ***(0.022)	0.219 ***(0.031)	0.115 ***(0.033)	0.009 ***(0.031)
uis	0.210 **(0.024)	−0.134 *(0.030)	0.242 **(0.036)	0.513 ***(0.176)
urb	−0.391 **(0.072)	−0.440 ***(0.108)	−0.114 **(0.118)	0.059 ***(0.023)
gov	0.141 ***(0.098)	0.183 ***(0.195)	−0.009(0.122)	1.005 ***(0.135)
emp	−0.076(0.062)	−0.095(0.092)	0.084 ***(0.082)	0.191(0.123)
open	0.173 ***(0.027)	0.087 *(0.048)	0.189 ***(0.038)	0.186 **(0.070)
dep	0.289 *(0.024)	0.092(0.072)	2.230 ***(0.602)	1.353 ***(0.383)
Constant term	−2.200 ***(0.420)	−2.27 ***(0.492)	−5.597 ***(1.457)	−4.740 ***(0.944)
Bidirectional fixation	YES	YES	YES	YES
R2	0.551	0.497	0.616	0.557

Note: ***, **, * indicate significant at 1%, 5%, 10% statistical level respectively.

. From the results of Figure 1, in order to intuitively compare the characteristics of regional differences, this paper further divides them into upstream, midstream and downstream for analysis. Due to the lack of a large amount of data in the upstream area, in order to ensure the accuracy of the conclusions, this paper mainly analyzes the results of the middle and downstream areas.

From the results of Table 3, the impact coefficient of ULT on HQED in all regions is positive and significant at the statistical level of 1%. However, the impact effect decreases from the middle reaches and downstream, which are 0.219, 0.115 and 0.009, respectively. This indicates that the ULT has a substantial impact on HQED in areas with relatively low development levels but has a weak impact on HQED in areas with relatively high development levels, and the results are consistent with Hypothesis 1 and Hypothesis 3. In addition, the coefficients of uis and urb on HQED are significant at the statistical level of 5%. The difference is that the overall impact of uis is positive, while the overall impact of urb is negative. The positive impact of uis is mainly in the middle reaches and downstream, while the negative impact of urb is mainly in the upper and middle reaches. This shows that the Yangtze River Economic belt should take complete account of regional differences in promoting HQED so as to “implement policies according to local conditions”, and the ULT in the middle reach and downstream area with relatively well economic development should focus on promoting the uis so as to improve the HQED level, while the cities in middle reaches should control the speed of land urbanization to avoid the adverse effects of extensive urbanization on HQED.

4.3. Indirect Impact Analysis

After the direct effect analysis, the indirect effect of ULT on HQED through uis and urb is further analyzed by using the intermediary effect model.

Table 4. Test results of mediating effect.

Mediating Effect		HQED	uis	HQED		HQED	urb	HQED
		(1)	(2)	(3)		(4)	(5)	(6)
Global	land	0.137 *** (0.028)	0.038 *** (0.014)	0.067 *** (0.012)	land	0.137 *** (0.028)	0.566 ** (0.236)	0.035 *** (0.008)
	uis			0.204 ** (0.040)	urb			0.028 * (0.06)
Upstream	land	0.184 ** (0.019)	0.002 (0.018)	0.182 ** (0.009)	land	0.184 ** (0.019)	0.145 ** (0.135)	0.018 ** (0.10)
	uis			−0.089 (0.051)	urb			−0.003 (0.005)
Midstream	land	0.215 ** (0.009)	0.049 * (0.028)	0.113 (0.008)	land	0.215 ** (0.009)	1.064 ** (0.624)	0.022 ** (0.008)
	uis			0.209 *** (0.062)	urb			−0.000 (0.001)
Downstream	land	0.082 ** (0.016)	0.149 ** (0.019)	0.047 ** (0.020)	land	0.082 ** (0.016)	0.449 * (0.445)	0.035 (0.017)
	uis			0.249 *** (0.075)	urb			0.074 ** (0.025)
Control variable		YES	YES	YES		YES	YES	YES
Bidirectional fixation		YES	YES	YES		YES	YES	YES

Note: ***, **, * indicate significant at 1%, 5%, 10% statistical level respectively.

shows the main results of the global and local mediating effects models. Other results of the models, such as control variables, have been omitted. Consistent with the analysis of direct effects, the indirect effect analysis of this paper mainly focuses on the results of the middle and lower reaches.

The regression results of the main effects of models (1) and (4) in Table 4 indicate the impact of the core explanatory variable (land) on the HQED, and the impact coefficient is significantly positive, indicating that the impact of the ULT on the HQED of each region is positive. Model (2) verifies whether the ULT impacts the uis. According to model (2), except for the upstream region, ULT in other regions has a significant positive impact on uis. The downstream region has the most significant impact, with a coefficient of 0.149, indicating that the scale of ULT in the downstream increases by 1%, and the level of uis increases by 0.149%. Model (3) is to verify the mediating effect of uis, in which the results of global scale and downstream regional model are similar, and the coefficients of uis are significant and are 0.204 and 0.249, respectively, indicating that the mediating effect of the global scale and downstream regional uis exists. The indirect promotion effects on HQED were 0.204% and 0.249%, respectively, and the coefficients of ULT were significant, indicating that the global scale and the uis in downstream areas played a partial intermediary role. The coefficient of uis in the middle reaches is 0.209, but the coefficient of ULT is not significant, which indicates that the uis in the middle reaches a full intermediary role between ULT and HQED.

Model (5) verifies whether the ULT has an impact on urb. The results show that both global scale and local scale ULT have a significant positive impact on urb, and the middle reaches have the most significant impact, with a coefficient of 1.064, indicating that the scale of ULT in the middle reaches increases by 1%, and the level of urb increases by 1.064%. Model (6) is the verification of the intermediary effect of urb. From the global scale, the coefficients of urb and ULT are significantly positive, indicating that urb plays a partial intermediary role in the positive impact of ULT on HQED at the global scale. In model (6), the urb coefficients of midstream regions are negative and insignificant, indicating that urbanization’s intermediary role in the midstream regions is not apparent. In model (6), the coefficient of urb in the downstream area is significant and 0.074, while the coefficient of ULT is not significant, indicating that urb in the downstream area plays a full intermediary role between ULT and HQED.

4.4. Threshold Effect Analysis

According to the panel threshold effect model, Bootstrap analysis is carried out 500 times by using Stata17 software, and the significance tests of single threshold, double threshold and three thresholds are carried out step by step for the global and local models of the Yangtze River Economic Belt. Through the tests, it is found that there is only a double threshold effect in the global and local areas, which is consistent with Hypothesis 2 in this paper. On the basis of passing the threshold effect significance test, the significance level of the threshold value is further analyzed. Finally, the parameter estimation results of the threshold regression are obtained (

Table 5. Threshold effect significance test and confidence interval.

Area	Threshold Number	F Value	10%	5%	1%	Threshold Value	95% Confidence Interval
Global	Single threshold	38.47 **	33.34	37.92	47.13	897.85	(862.64, 906.87)
	Double threshold	27.94 *	26.88	29.89	40.76	365.04	(354.25, 368.71)
Upstream	Single threshold	57.20 **	46.07	51.88	68.68	265.07	(259.82, 267.74)
	Double threshold	34.59 **	28.18	32.71	42.27	94.07	(89.74, 95.37)
Midstream	Single threshold	50.89 **	33.32	39.27	53.50	432.68	(419.89, 441.42)
	Double threshold	17.04	27.45	34.10	43.78	—	—
Downstream	Single threshold	46.44 ***	25.98	31.31	38.06	845.56	(828.82, 854.06)
	Double threshold	30.15 *	25.88	30.27	39.66	561.16	(544.57, 566.80)

Note: ***, **, * indicate significant at 1%, 5%, 10% statistical level respectively.

and

Table 6. Parameter estimation results of the panel threshold model.

	Global (7)	Upstream (8)	Midstream (9)	Downstream (10)
Tit : land	λ1 = 365.04 λ2 = 897.85	λ1 = 94.07 λ2 = 265.07	λ1 = 432.68	λ1 = 561.16 λ2 = 845.56
land (Tit ≤ λ1) land (λ1 < Tit ≤ λ2) land (Tit > λ2)	0.131 ***(0.014) 0.155 ***(0.013) 0.018 ***(0.012)	0.071 ***(0.020) 0.121 ***(0.017) 0.159 ***(0.015)	0.159 ***(0.019) 0.171 ***(0.017)	0.065 ***(0.019) 0.092 ***(0.182) 0.005 ***(0.017)
uis (Tit ≤ λ1) uis (λ1 < Tit ≤ λ2) uis (Tit > λ2)	−0.180 ***(0.043) 0.110 **(0.050) 0.186 ***(0.059)	−0.241 *(0.128) −0.026 (0.080) 0.070 (0.066)	−0.198 **(0.077) 0.312 ***(0.074)	0.152 **(0.068) 0.264 ***(0.090) 0.471 ***(0.072)
urb (Tit ≤ λ1) urb (λ1< Tit ≤ λ2) urb (Tit > λ2)	−0.252 (0.067) 0.064 ***(0.002) 0.113 *(0.004)	−0.017 *(0.010) 0.006 (0.012) 0.015 (0.010)	−0.008 ***(0.003) −0.001 (0.003)	0.005*(0.003) −0.011 ***(0.004) −0.002 (0.003)
Control variable	YES	YES	YES	YES
R2	0.6192	0.5104	0.5973	0.5688
N	1296	372	432	492

Note: ***, **, * indicate significant at 1%, 5%, 10% statistical level respectively.

). Similarly, the threshold effect analysis of this paper mainly focuses on the results of the middle and lower reaches.

Table 5 shows the significance test, threshold value and confidence interval of the threshold effect of ULT in the global and local areas. It can be seen from the results that there is only a single threshold effect in the midstream region. However, there is a double threshold effect on the global scale, upstream and downstream regions, and all of them pass the test at least at the 10% significance level. Table 6 shows the threshold regression parameter estimation results of the global scale (model 7) and local scale models (models 8~10) of the Yangtze River Economic belt.

From a global perspective, from model (7) in Table 5, we can see that the ULT scale threshold is 365.04 hectares and 897.85 hectares, respectively. From the coefficient estimation results, we can see that the impact of the ULT on high economic quality is significantly positive at different threshold stages. However, the impact size increases with the scale of ULT. The coefficients are 0.131, 0.155 and 0.018, respectively. When the threshold value is less than or equal to 365.04 hectares, it is significantly negative, −0.180 and −0.252, respectively, while when the threshold value is more excellent than 365.04 hectares and 897.85 hectares, it is significantly positive. 0.110, 0.186 and 0.064, 0.113, respectively, showing inverted “U” type characteristics. This shows that with the HQED, the Yangtze River Economic belt should gradually reduce its dependence on ULT, control the growth of the scale of construction land, promote the inclination of new construction land to high-tech industries using policy regulation and administrative intervention, promote the upgrading of urban industrial structure, orderly promote the process of urbanization, avoid blind expansion of construction land, and improve the quality of urbanization.

From the local perspective, it can be seen from the models (9) ~ (10) in Table 6 that there is only one threshold in the middle reaches, which is 432.68 hectares. There are two thresholds in the downstream area, which are 561.16 hectares and 845.56 hectares, respectively. From the coefficients of ULT in models (9) ~ (10), we can see that ULT in the midstream, and downstream significantly positively impacts HQED in different threshold ranges. However, the difference is that the impact of ULT in the middle reach increases with the expansion of the scale. In comparison, it has a more negligible impact in the lower reaches and shows the characteristics of an inverted U-shape. From the results of uis and urb in models (9) ~ (10), with the threshold value increase, the impact of uis in the middle reaches negative first and then positive, showing an inverted U-shape. In contrast, the impact of urb has been negative. The impact of uis in the downstream area is always optimistic in the two threshold ranges. In contrast, the impact of urb is positive when the ULT scale is less than or equal to 561.16 hectares, but harmful when it is more than 561.16 hectares.

To sum up, ULT in the Yangtze River Economic belt is good for HQED, but its marginal impact is declining, and it is ultimately difficult to rely on ULT to promote HQED. The impact of the ULT on HQED in the middle reach is more vital than in the downstream, which is consistent with Hypothesis 3. The uis in the midstream region has gradually shown a strong positive impact on HQED under large-scale ULT. The HQED of downstream areas has gradually eliminated the dependence on the index of construction land and the expansion of urban scale and has been more and more positively affected by uis. The reason may be that in recent years, the national construction land index has been inclined to the midstream and upstream areas, which enables the midstream areas to attract investment, expand cities and towns on a large scale, and develop land finance to promote economic growth. On the other hand, the poor endowment and foundation of the midstream region led to the relatively backward industries introduced, which play a feeble role in uis and improving the economic quality. In addition, due to poor endowment conditions, significant investments in the early stage of economic construction and high financial requirements for local governments. The relevant supporting infrastructure in the midstream areas cannot be synchronized after rapid land urbanization [16]. Thus, the attraction for industrial and population agglomeration is weak, and the short-term economic effect is low.

5. Discussions

In order to ensure the accuracy of the conclusions of the article, we discuss the robustness of the results from two aspects. First of all, we test the correlation between variables, and eliminate the highly related variables, and retain the variables with relatively small correlation. The correlation test results of the final reserved variables are shown in

Table 7. Test of correlation coefficient.

	land	uis	urb	gov	emp	open	dep
land	1
uis	0.060 **	1
urb	−0.026	−0.054 **	1
gov	0.203 ***	0.140 ***	0.106 ***	1
emp	0.251 ***	0.322 ***	−0.028	0.300 ***	1
open	0.108 ***	0.389 ***	0.026	0.200 ***	0.193 ***	1
dep	0.313 ***	0.142 ***	−0.004	0.376 ***	0.298 ***	0.311 ***	1

Note: ***, ** indicate significant at 1%, 5% statistical level respectively.

.

Secondly, we further use the variance expansion factor to test the multicollinearity of variables, and the test results are shown in

Table 8. Test of VIF.

	land	uis	urb	gov	emp	open	dep	mean
VIF	2.62	2.38	2.38	1.85	1.68	1.29	1.04	1.89
1/VIF	0.3813	0.4199	0.4203	0.5397	0.5964	0.7725	0.9571

. As can be seen from Table 8, all VIF values are much less than 10, indicating that the multicollinearity between variables can be ignored [41].

In addition, in this paper, we refer to the relevant authoritative literature, use the lag of two periods of ULT as the core explanatory variable for regression, the main regression results have passed the significance test. Furthermore, we use ULT without lag processing as an instrumental variable for regression analysis, and the regression results are consistent with the ULT regression results of two lagging periods, which shows that the regression results and conclusions of this paper are reliable. Due to the limitation of the length of the article, the direct, indirect and threshold effects of ULT without lag processing are not included in the text. If necessary, you can ask the author of this article.

Based on robust discussion, we further discuss some shortcomings of this paper. Firstly, the research period of this paper has limitations, and due to the limitation of access to core variable data, the research content of this paper is only up to 2017. It is a pity that the latest year has not been analyzed. Although the research conclusions can be compared with subsequent years to achieve the forecasting effect, due to the lack of support of the latest data, the conclusions of this paper have a weakening impact on the reference of relevant policy-making in the future. Secondly, the division of the study area according to the watershed may weaken the heterogeneity of the internal development of the region, so the conclusion has a certain deviation. Therefore, in the follow-up research, a variety of means and methods should be used to obtain the latest data to ensure the timeliness of the research. In the process of regional analysis, we should break the idea of dividing geographical boundaries and think about the division methods of research regions from a socio-economic point of view, to further promote the accuracy and scientificalness of the research conclusions.

6. Solutions and Recommendations

According to the results, the following suggestions can be put forward. The government should promote the reform of the land transfer mechanism and break the mechanism of “making money from land”. The government should introduce third-party evaluation agencies and market competition mechanisms, and learn from the auction strategy of commercial land to optimize the supply of industrial industry. Secondly, the government should streamline the examination and approval procedures and reduce excessive interference in the allocation of land resources. strengthen the supervision responsibility of the government.

The efficiency of construction land use and the quality of urbanization should be improved. Firstly, Government-led spatial planning must improve the binding force of the three “red lines”, strictly control the expansion of construction land, make use of urban renewal to vigorously tap the potential of existing construction land, and improve utilization efficiency. Secondly, the government can make use of urban development planning and urban renewal to promote the transformation of land use and function, adjust the structure of construction land, and increase the supply of residential land. Thirdly, the government can use administrative and economic means to reduce the supply proportion and preferential policies of industrial land, raise the barriers to scientific and technological and environmental access, and further optimize the financial services of science and technology enterprises, in order to promote the optimization and upgrading of industrial structure, improve the quality of urbanization.

Regional differences should be paid more attention to. Due to the high level of economic development and industrial upgrading in the lower reaches of the Yangtze River Economic belt, on the one hand, the government should transfer industrial land to high-level industries to enhance industrial economic competitiveness and land use efficiency; On the other hand, through ULT, the proportion of urban life and ecological space will be increased, and the quality of urbanization will be improved. Due to the poor endowment conditions, the midstream and upstream areas are competitive in undertaking the industrial transfer from the eastern coastal areas. The strategy of selling industrial land at a low price is still of practical significance. Meanwhile, ULT in the midstream and upstream areas should also focus on the layout of future industrial transformation and upgrading, pay attention to economic quality in formulating medium and long-term social and economic development goals, properly improve the entry threshold of enterprises and strengthen the level of environmental control, further improve infrastructure conditions, and enhance the quality of urbanization and the potential of HQED.

7. Conclusions

China’s rapid economic growth for a long-time past must be separated from the support of the unique land system, and local governments have also formed a high degree of dependence on land resources in promoting economic development. With China’s economic growth entering the “new normal”, changing the old mode of “seeking development by land” and promoting high-quality economic growth has become the future development trend. Based on this, this paper takes the Yangtze River Economic belt as the study area, uses an empirical model to evaluate the level of HQED from 2004 to 2017, and analyses the connection between ULT and HQED. The results show that the HQED level during the study period increased steadily. However, the overall level was low, and the coupling coordination with ULT showed a trend of first increasing and then decreasing and there were significant regional differences. With the improvement of the level of HQED, the direct impact of ULT is gradually weakened, and it is unsustainable to seek development by land. The ULT has a significant intermediary effect, which further stimulates HQED by affecting the uis and urb, which is more significant in the midstream and downstream. The ULT also has a threshold effect. With the expansion of the threshold scale of ULT, the overall impact of ULT on HQED is decreasing but rising in the upper and middle reaches; The overall impact of uis and urb shows an inverted “U” shape, which is first negative and then positive, but the impact varies significantly in different regions. The impact of uis and urb has the characteristics of alienation. Finally, according to the research results, we put forward targeted policy recommendations, which provide a reference for the high-quality economic development of the Yangtze River Economic Belt and even the whole of China.