Factors Influencing the Coordinated Development of Urbanization and Its Spatial Effects: A Case Study of Beijing-Tianjin-Hebei Region

Abstract

The purpose of the coordinated development of urbanization is to achieve the coordination of the internal subsystems of urbanization and the spatial coordination within the region. The coordinated development level of urbanization and its spatial effect are affected by many factors. Based on the influence mechanism of coordinated development of urbanization, in this study, the evaluation index system of coordinated development of urbanization and the analysis of influencing factors are constructed respectively. The coupling coordination model, fixed effect model and spatial lag model are used to analyze the coordinated development level and influencing factors of urbanization in Beijing-Tianjin-Hebei region, and the changes of influencing factors before and after the spatial effects are compared. The results are as follows: although the coordinated development level of urbanization in Beijing-Tianjin-Hebei region shows a trend of increasing year by year, the spatial differentiation phenomenon is becoming more and more obvious. In this process, the internal and external influencing factors of coordinated development of urbanization play a decisive role. In addition, after the spatial effect is included, the significance level of the influencing factors has changed, and the promotion effect on the coordinated development of urbanization is reduced. The performance is as follows: there is insufficient population transfer and employment among regions, low spatial spillover effect of economic development, strong dependence on land finance, expansion of habitat destruction and low degree of governance coordination, imbalance between public resource allocation and public service supply, lack of unified market-oriented environment and strong government intervention. The key points of future policies and reforms are as follows: under the guidance of the national strategic thinking, we should deepen the reform of the administrative system of the government, actively break through the institutional obstacles restricting the coordinated development of urbanization, continuously optimize the state of the urbanization system and form a joint force of coordinated development.

1. Introduction

The connotation of the new urbanization strategy is that under the guidance of the “people-oriented” ideology, the development mode of urbanization has gradually changed from “extension type”, “speed type” to “connotation type” and “quality type” [1]. The people-centered new urbanization can give full play to the role of improving the quality of the population, improving labor productivity and enhancing the vitality of innovation. It plays a key role in further tapping into the growth potential, promoting coordinated development and improving the quality of development. The coordinated development of urbanization conforms to the requirements of the new urbanization strategy. The coordinated development of urbanization focuses on the coordination of internal factors and the evolution of spatial pattern [2]. Firstly, the coordinated development of population, economy and other internal elements of urbanization can further optimize the allocation efficiency of various resources and stimulate the combined benefits of various resources; secondly, the coordination of urbanization in the spatial pattern can narrow the regional development gap and help the regional coordinated development strategy. In this process, the factors that influence the coordinated development of urbanization play a crucial role, affecting the level of coordinated development of urbanization and the spatial pattern of urbanization. Therefore, study on its influencing factors and spatial effects can provide decision-making basis for correctly guiding the coordinated development of urbanization within the regional scope [3].

The coordinated development of urbanization aims to pay attention to the coordination of the internal elements of urbanization and the spatial pattern of urbanization. Although there is no concept and special study on “coordinated development of urbanization” in foreign countries, from the perspective of complexity and diversity of urbanization process, certain study results have been formed in the aspects of population transfer and employment [4,5,6], economic agglomeration and spillover [7,8], land development and utilization [9,10], ecological damage and governance [11,12], infrastructure improvement [13] etc. To a certain extent, it confirms the importance of coordination of internal elements of urbanization. In China, there are abundant study results on “coordinated development of urbanization”, including the following aspects: measurement of coordinated development of urbanization (internal element level), spatial and temporal differentiation (spatial pattern level), and influencing factors [14,15]. As for the influencing factors, both macro and micro levels are involved. The macro level includes external power, endogenous power, government power and market power [16,17], while the micro level includes labor division, economic development level, industrial structure, degree of opening up, ecological environment resources, infrastructure construction etc. [18]. These factors not only affect the level of coordinated development of urbanization, but also affect the change of spatial pattern of coordinated development of urbanization [19,20]. Therefore, the study on the influencing factors of the coordinated development of urbanization should take into account the direction, degree and spatial effect.

It can be seen from the connotation of coordinated development of urbanization that it requires not only the coupling and coordination between internal subsystems of urbanization, but also the spatial coordination of urbanization development within the regional scope [20,21]. Additionally, although the existing achievements have laid a solid theoretical foundation for this study, existing research needs to be improved in the following two aspects: The first is that the existing study on the measurement and analysis of the coordinated development level of urbanization often uses population and land indicators as the standard, which can-not reflect the diversity and complexity of the urbanization system. In this study, starting from the connotation of coordinated development of urbanization, the economic, ecological, social and other factors are included in the evaluation index system, and the analysis results will be more stable after striving. The second is that the existing studies only focus on the direction and extent of their effects, and lack overall planning of spatial effects. In this study, combined with the ordinary panel and spatial panel econometric model, the influencing factors and spatial effects of coordinated development of urbanization are analyzed comprehensively.

2. Influence Mechanism of Coordinated Development of Urbanization

Urbanization is a comprehensive concept: its connotation is gradually changing from “population urbanization” to “human urbanization”, and the study on urbanization is also extending from demography to other disciplines. From different dimensions, urbanization is endowed with different connotations. For example, in demography, it is considered that urbanization is the process of transforming rural population into urban population (the transformation of population identity) [4]; in economics, it focuses on the economic benefits in the process of industrial structure transformation (transformation of economic structure) [7]; in geography, it focuses on the spatial reconstruction of urbanization (transformation of spatial structure) [9]; in ecology, attention is paid to the ecological environment state, pressure and response (transformation of ecological environment) [22]; in sociology, it is emphasized that urbanization is the whole process from rural lifestyle to urban lifestyle (transformation of lifestyle) [23]. In this study, it is considered that the urbanization system includes the following five dimensions: population, economy, land, ecology and society. There are different degrees of promoting or restricting relationships among the five dimensions. The coordinated development of urbanization shows the coupling and coordination among the above five dimensions. Coupling and coordination degrees are related but also different. The connection is that they both reflect the dynamic influence process between systems. The difference is that coupling degree only reflects the strength of the interaction between systems, but it is difficult to characterize whether there is a benign interaction, while coordination degree only reflects the benign interaction process between systems. The degree of coupling coordination refers to the dynamic change process between systems or subsystems, which gradually changes from disorder to order, reflecting the synergy effect between systems (Figure 1). Analyzing the coupling and coordination relationship can solve the problem of unbalanced development among systems, which is beneficial to the coordinated development of various systems or elements. Due to the strong coupling and coordination relationship among urbanization factors such as population, land, economy, ecology and society, the model is widely used in the study of coordinated development of urbanization, which is also an important reason why the model was selected for this study.

Population, economy, land, ecology and society, the components of urbanization system, directly affect the level of coordinated development of urbanization [16]. In addition, in the process of coordinated development of urbanization, the core position of population, the dynamic role of economy, the carrier function of land, and ecological protection effect all play corresponding roles [24,25,26,27]. (1) Population urbanization is the core of urbanization development. The traditional population urbanization pays one-sided attention to the process of rural population transferring to cities and towns, but ignores the demands of people. In essence, population urbanization is the whole process of changing the way of life and production, as well as the idea. The transfer of agricultural population to cities and towns is the essence of population urbanization, and research in this field has always been the focus of academic attention. (2) Economic urbanization is the driving force of urbanization and the process of human economic activities to cities and towns. The rapid development of economic society has strongly changed the natural structure and socio-economic structure of our country, from focusing on environmental changes dominated by nature to environmental changes dominated by human beings. (3) Land urbanization is the carrier of urbanization development, and the advancement of urbanization level will inevitably be reflected in the land space. In recent years, land urbanization has become a research hotspot in the field of urbanization, especially the process of land urbanization from the perspective of urban land use. Land provides spatial support for the transfer of agricultural population and the development of industrial economy. The most intuitive manifestation is the continuous expansion of urban space. The rapid development of urbanization highlights the value of land, and the urbanization process of many countries and regions shows the extension expansion of land area. (4) The research on the relationship between urbanization and ecological environment has attracted the attention of academic circles. There is an interactive and coupling relationship between urbanization and ecological environment. On the premise of ensuring the benign cycle of urban ecosystem, realizing the rapid and stable development of urbanization is the requirement of sustainable development and the guarantee of urbanization development. The realization of ecological urbanization should face up to the state, pressure and response of the ecological environment, that is, while assessing the quality and pressure of the current urban ecological environment, strengthen the governance of the ecological environment. (5) Social urbanization is reflected in the improvement of urban public facilities, public services and social security level, and is the process of providing services for the development of urbanization. The essence of social urbanization is to respond to the “human demand” in the process of urbanization, that is, the demand for basic life services with non-competitive and non-exclusive characteristics in the process of agricultural population transfer. The coordinated development of urbanization is a systematic project, which needs to give full play to the decisive role of the market in the allocation of resources, and to give full play to the role of the government in macro regulation and control. Hence, the market environment and government regulation can be used as external influencing factors of coordinated development of urbanization [16]. The internal and external factors influencing the coordinated development of urbanization are not completely isolated. External factors, often through economic and administrative means, act on internal factors, and then have an indirect impact on the coordinated development of urbanization.

The spatial effect also exists in the influencing factors of coordinated development of urbanization, as follows: the migration and employment of population in different places, the agglomeration and diffusion effect of economic development, the synergy between the expansion of habitat destruction and local governance, the equalization of basic public services and the degree of market integration. The spatial effects of the above factors will have an impact on the coordinated development of urbanization in the region [15]. The coordinated development of urbanization, all in all, is a process in which the internal and external elements flow and interact in time and space. Additionally, the combination of different types and regional spatial elements, the change of the intensity and scope of action will bring about the continuous evolution of the coordinated development level and spatial pattern of urbanization within the regional scope (Figure 2).

3. Study Methods and Data Sources

3.1. Index System Construction

By referring to the achievements of the evaluation index system of coordinated development of urbanization [2,28,29,30], following the principles of systematicness, scientificalness and operability, starting from the connotation of coordinated development of urbanization, it can be summarized into the following five aspects: the indicators of population mainly reflect the process of population concentration in cities and towns, including population structure, population employment, living standard and population quality; economic indicators mainly reflect the development of non-agricultural economy, including economic structure and development of export-oriented economy; land indicators mainly reflect urban land development and utilization, including land structure and land input-output benefits; ecological indicators mainly reflect habitat destruction and governance, including ecological quality, ecological pressure and ecological governance; social indicators mainly reflect the change of lifestyle, including public facilities, social services, social security. These five aspects are interrelated and different, and strive to comprehensively and accurately reflect the coordinated development process of urbanization (

Table 1. Measurement index system of coordinated development of urbanization.

First Level Index (A)	Second Level Index (B)	Third Level Index (C)
Population urbanization (X1)	Population structure (x11)	Population urbanization rate (%) (x111)
	Employment level (x12)	The registered urban unemployment rate (%) (x121)
	Living standard (x13)	Per capita disposable income of urban residents (yuan) (x131)
		Engel coefficient of urban residents (%) (x132)
	Population quality (x14)	Proportion of education expenditure in general public budget expenditure (%) (x141)
		Number of college students in 10,000 persons (person) (x142)
Economic urbanization (X2)	Economic development (x21)	GDP growth rate (%) (x211)
	Economic structure (x22)	Proportion of secondary and tertiary output value in GDP (%) (x221)
		General public budget income per capita (10,000 yuan) (x222)
		General public budget expenditure per capita (10,000 yuan) (x223)
	Export oriented economy (x23)	Proportion of total export to regional GDP (%) (x231)
Land urbanization (X3)	Land structure (x31)	Proportion of urban construction land to urban area (%) (x311)
		Proportion of urban built up area to urban area (%) (x312)
	Land input (x32)	Land-average fixed assets investment (10,000 yuan/km2) (x321)
		Land-average general public budget expenditures (10,000 yuan/ km2) (x322)
	Land output (x33)	Land-average second and third output value (10,000 yuan/ km2) (x331)
		Land-average general public budget income (10,000 yuan/ km2) (x332)
Ecological urbanization (X4)	Ecological quality (x41)	Green coverage rate of built-up areas (%) (x411)
	Ecological pressure (x42)	Industrial sulfur dioxide emissions per capita (t) (x421)
		Industrial dust emission per capita (t) (x422)
		Wastewater discharge per capita (t) (x423)
	Ecological management (x43)	Comprehensive utilization rate of solid waste (%) (x431)
		Harmless garbage treatment rate (%) (x432)
		Sewage treatment rate (%) (x433)
Social urbanization (X5)	Common facilities (x51)	Urban road area per capita (m2) (x511)
		Length of urban drainage pipeline per capita (m) (x512)
	Social services (x52)	The number of taxi operators owned by 10,000 persons (vehicle) (x521)
		The number of bus and tram operation owned by 10,000 persons (vehicle) (x522)
	Social security (x53)	Proportion of basic endowment insurance for urban employees (%) (x531)
		Proportion of basic medical insurance for urban employees (%) (x532)
		Proportion of urban unemployment insurance (%) (x533)

).

3.2. Study Method

3.2.1. Coupling Coordination Degree Model

In this study, subjective weighting method (G1 method) and objective weighting method (entropy weight method) are used to combine the weights of indicators. The two weighting methods are detailed in references [31,32]. After weighting the indicators, the coordinated development level of urbanization is measured, and the formula is as follows:

$$U_i={\displaystyle \sum _{j=1}^m{w}_j}\times Y_{ij}$$

(1)

$$T_i=a{U}_{ia}+b{U}_{ib}+c{U}_{ic}+d{U}_{id}+e{U}_{ie}$$

(2)

In the formula, U_i is the development index of urbanization subsystem in city i, Y_ij is the standardized value of index j of the city i, w_j is the entropy weight of the index j, m is the index number, T_i is the comprehensive development index of urbanization, and a, b, c, d, e are the contribution shares of population, economy, land, ecology and society to the urbanization system. As for the measurement of contribution share, the existing research generally states that the importance of various factors is the same [21], and the five dimensions of population, economy, land, ecology and society work together in the coordinated development of urbanization. At the same time, there is a strong coupling relationship between them. The lack of any kind of factors will lead to the imbalance of urbanization development. Therefore, this research refers to the existing research results and believes that a = b = c = d = e = 1/5.

The coupling coordination model is introduced to better reflect the coupling coordination level between subsystems [21]. The formula is as follows:

$$C_i=(U_{ia}+U_{ib}+U_{ic}+U_{id}+U_{ie})/\sqrt{{{\displaystyle U}}_{ia}^2+{{\displaystyle U}}_{ib}^2+{{\displaystyle U}}_{ic}^2+{{\displaystyle U}}_{id}^2+{{\displaystyle U}}_{ie}^2}$$

(3)

$${\mathrm{D}}_i=\sqrt{C_i\times T_i}$$

(4)

In the formula, C_i is the coupling degree of population, economy, land, ecology and society of city i, and D_i is the coupling coordination of urbanization development of city i.

3.2.2. Common Panel Data Model

The panel data model can be divided into fixed effect model (FE), random effect model (RE) and mixed effect model (ME) according to different assumptions of random error term μ_it [33,34,35]. The formula is as follows:

$${\mathrm{y}}_{it}=\alpha +{{\displaystyle x}}_{it}^{\prime }\beta +{\epsilon }_{it},i=1,2,...,N;t=1,2,...,T\hspace{1em}\hspace{1em}\left(\mathrm{ME}\right)$$

(5)

$$y_{it}={\alpha }_i+{{\displaystyle x}}_{it}^{\prime }\beta +{\epsilon }_{it},i=1,2,...,N;t=1,2,...,T\hspace{1em}\hspace{1em}\left(\mathrm{FE}\right)$$

(6)

$$y_{it}=\alpha it+{{\displaystyle x}}_{it}^{\prime }\beta +{\epsilon }_{it},i=1,2,...,N;t=1,2,...,T\hspace{1em}\hspace{1em}\left(\mathrm{RE}\right)$$

(7)

In the formula, y_it is the explained variable; α is the intercept term; α_i, α_it is the random variable; x_it’ is the K × 1 order regression variable sequence vector (including K regression variables); β is the K × 1 order regression system sequence vector; ε_it is the error term.

Generally speaking, the choice of the three models needs to be determined by the test method. In practical application, the specific steps are as follows: the fixed effect model or mixed effect model is decided by judgment. Through F-test, if the test result is significant, it belongs to fixed effect model, and otherwise it belongs to mixed effect model. On the basis of the above, the fixed effect model or random effect model is judged. Through Hausman test, if the result is significant, it belongs to the fixed effect model, otherwise it belongs to a random effect model.

3.2.3. Spatial Panel Data Model

Due to the different sources of spatial correlation effect factors, spatial econometric models can be divided into two types, namely spatial lag model (SLM) and spatial error model (SEM). The former mainly studies whether the variables have diffusion phenomenon (spillover effect) in a region. The latter mainly reflects the impact of error shocks in neighboring areas on the region [36]. The formula is as follows:

$$Y=\rho W_y+X\beta +\epsilon ,\epsilon ~N(0,{{\displaystyle \delta }}^2{{\displaystyle I}}_n)\hspace{1em}\hspace{1em}\hspace{1em}\left(\mathrm{SLM}\right)$$

(8)

$$Y=X\beta +\mu ,\mu =\lambda W\mu +\epsilon ,\epsilon ~N(0,{{\displaystyle \delta }}^2{{\displaystyle I}}_n)\hspace{1em}\hspace{1em}\hspace{1em}\left(\mathrm{SEM}\right)$$

(9)

In the formula, Y is the explained variable; X is the data matrix of the exogenous explanatory variable; ρ is the influence coefficient of the spatial lag effect; Wy is the spatial lag value of the explained variable; β is the coefficient vector to reveal the explanatory variable; ε is the random error vector of the spatial lag model; λ is the influence coefficient of the spatial error effect of the dependent variable and μ is the random error vector of the spatial error model.

In this study, the 0-1 weight matrix of car adjacency rule is used to select the weight, that is, if region i and region j have a common boundary, then W (i, j) = 1; otherwise, W (i, j) = 0, and the city without adjacent area and its nearest area are regarded as adjacent areas. The least squares (OLS) and maximum likelihood estimation (ML) are mainly used in the model estimation.

3.3. Data Sources

In this study, 13 cities of Beijing-Tianjin-Hebei urban agglomeration are taken as the basic research units. Considering the availability of data, 2009–2017 (five inspection periods: 2009, 2011, 2013, 2015 and 2017) are selected as continuous time series to measure and analyze the influencing factors of coordinated development of urbanization. Data sources are as follows: China Statistical Yearbook 2010–2018; China Urban Statistical Yearbook 2009–2017; China Population and Employment Statistical Yearbook 2010–2018; National Economic and Social Development Statistical Bulletin 2009–2017; China Land and Resources Statistical Yearbook 2010–2018; China Environment Statistical Yearbook 2010–2018; China Social Statistics Yearbook 2010–2018; 2010–2018 Beijing-Tianjin-Hebei related statistical yearbook and China Economic Information Network Statistical Database. As for some missing data, the moving average method is used for interpolation.

4. Measurement of Coordinated Development of Urbanization and Analysis of Influencing Factors

4.1. Measurement and Analysis of Coordinated Development of Urbanization in Beijing-Tianjin-Hebei Region

This paper uses the formula (1)–(4) to measure the coordinated development level of urbanization of 13 cities in Beijing, Tianjin and Hebei in five inspection periods (

Table 2. Coordinated development level of urbanization in Beijing-Tianjin-Hebei region (2009–2017).

	2009	2011	2013	2015	2017
Beijing	0.4645	0.4831	0.5033	0.5490	0.6039
Tianjin	0.3984	0.4143	0.4316	0.4708	0.5179
Shijiazhuang	0.3180	0.3307	0.3445	0.3758	0.4134
Tangshan	0.3101	0.3225	0.3359	0.3665	0.4031
Qinhuangdao	0.2907	0.3023	0.3149	0.3435	0.3779
Handan	0.2712	0.2821	0.2938	0.3205	0.3526
Xingtai	0.2735	0.2844	0.2963	0.3232	0.3555
Baoding	0.2575	0.2686	0.2799	0.3147	0.3349
Zhangjiakou	0.2576	0.2679	0.2791	0.3188	0.3349
Chengde	0.2506	0.2606	0.2715	0.2962	0.3258
Cangzhou	0.2886	0.3002	0.3127	0.3411	0.3752
Langfang	0.3048	0.3170	0.3303	0.3603	0.3963
Hengshui	0.2584	0.2687	0.2799	0.3054	0.3359

). From the view of time series, the coordinated development level of urbanization in Beijing-Tianjin-Hebei region shows an upward trend year by year. The fastest growing cities are Beijing and Tianjin, and the growth rate of Hebei is relatively slow. As far as Hebei is concerned, Shijiazhuang, the capital of Hebei, and Tangshan, Qinhuangdao and Cangzhou in the eastern coastal areas have a rapid growth rate, while other regions have a slower growth rate.

According to the classification and statistics of coordination types in different studies, it is basically concentrated on the classification of four levels, five levels and ten levels. This paper takes the extreme value of the coordinated development level of urbanization as the interval, and divides the coordinated development level of urbanization into the following five levels by using “equal interval method” of Arcgis10.1 software: [0.5333–0.6039] belongs to the high coordination level; [0.4627–0.5332] belongs to the coordination level; [0.3920–0.4626] belongs to the barely coordinated level; [0.3214–0.3919] belongs to the near maladjustment level; [0.2506–0.3213] belongs to the maladjustment level. The high coordination level means that, in the process of urbanization development, population, economy, land, ecology, society and other internal elements of urbanization achieve a coupling and coordination state of mutual promotion and mutual progress, and their combination efficiency will be continuously played. The coordination level is also a good state for the coordinated development of urbanization, but to jump to the highly coordinated level, it is necessary to further stimulate the cooperation and integration among the internal elements of urbanization. The barely coordinated level is the lowest level of coordinated development of urbanization. In this state, there is a trend of coupling and coordination among the internal elements of urbanization, but the trend is not obvious. The near maladjustment level means that there is disorder in the development of the internal elements of urbanization. If the disorder is stopped and solved, it may jump to the barely coordinated level, and if the disorder continues to deteriorate, it will fall to the maladjustment level. The maladjustment level means that the development of the internal factors of urbanization is chaotic, and various factors inhibit each other and have a vicious circle. The problem of urban disease is prominent, and the urbanization system is facing great risks.

From the perspective of spatial distribution, the spatial differentiation of coordinated development of urbanization in Beijing-Tianjin-Hebei region is serious. In 2009, all regions of Hebei were in a state of maladjustment, Tianjin was at the barely coordinated level, and Beijing was at the coordination level. In 2011, Shijiazhuang and Tangshan in Hebei Province were promoted from maladjustment level to on-the-spot maladjustment level. In 2013, Langfang, Hebei Province, was promoted from maladjustment to on the verge of maladjustment. In 2015, Xingtai, Cangzhou and Qinhuangdao in Hebei Province were promoted from maladjustment level to nearly maladjustment level, Tianjin from barely coordinated level to coordination level, and Beijing from coordination level to high coordination level. In 2017, there was no maladjustment level in all regions of Hebei Province. Shijiazhuang, Tangshan and Langfang were promoted to the barely coordinated level, while Tianjin and Beijing were still at the coordination level and high coordination level (Figure 3).

4.2. Analysis on Influencing Factors of Coordinated Development of Urbanization in Beijing-Tianjin-Hebei Region

4.2.1. Variable Selection and Model Establishment

The factors influencing the coordinated development of urbanization are multifaceted, and the factors considered from different research perspectives are also different, which can be roughly summarized as: labor division, economic development level, industrial structure, degree of openness, ecological environment resources [37,38,39,40,41,42], infrastructure construction, marketization level, government behavior etc. Sudden changes in the external environment may have a certain, or even greater, impact on the coordinated development of urbanization, such as sudden changes in national politics, major social emergencies etc. Such factors cannot be predicted and measured, so this study is not considered temporarily. The policy environment, market environment and other external factors that have a great impact on the coordinated development of urbanization are variable, and the specific quantitative work is difficult to cover comprehensively and calculate accurately. Therefore, this study adopts alternative indicators to represent. It should be noted that the internal and external aspects of the urbanization system do not represent the separation of the two systems. For example, although the economic factors in this study belong to the internal driving force of the coordinated development of urbanization, they also represent a certain external market environment to some extent. At the same time, the external environment and government actions often have to be effective through economic means. Based on previous studies and the actual situation of coordinated development of Beijing-Tianjin-Hebei, this study comprehensively considered the role of internal and external factors in the coordinated development of urbanization from the aspects of population, economy, land, ecology, society, market and government, and thus proposed a framework model of the impact mechanism of coordinated development of urbanization (Figure 4). The coupling coordination degree is selected as the explained variable, and the above aspects are selected as the explanatory variables for quantitative analysis (

Table 3. Variable setting table of influencing factors for coordinated development of urbanization.

	Variable Name	Abbreviation	Definition and Unit	Effect
Explained variables	Coordination degree	Cdu	It is calculated by Formula (1)–(4)
Explanatory variables	Economic development level	Edl	Per capita GDP (yuan)	+
	Industrial structure	Is	Contribution rate of secondary and tertiary industries (%)	+
	Degree of opening up	Od	Growth rate of actual direct utilization of foreign capital (%)	+
	Population transfer and employment	Le	Proportion of employed personnel in secondary and tertiary industries (%)	+
	Land revenue	Lpi	Proportion of land transfer revenue in government fund budget revenue (%)	-
	State of ecological environment	Es	Per capita park green space area (m2)	+
	Pressure of ecological environment	Ep	Energy consumption per (CNY) 10,000 GDP (tons of standard coal)	-
	Response of ecological environment	Er	Proportion of investment in environmental pollution control in GDP (%)	+
	Public resource allocation	Pra	Number of beds in medical and health institutions in 10,000 persons (number)	+
	Marketization level	Ml	Total retail sales of consumer goods per capita (CNY 10,000)	+
	Government regulation	Gr	Proportion of local fiscal expenditure in GDP (%)	+

). The setting of the explanatory variables is different from the variables in the evaluation index system of urbanization coordinated development in the previous article. Moreover, based on the panel data of 13 cities in Beijing-Tianjin-Hebei region from 2009 to 2017, this study establishes the econometric regression model of influencing factors of coordinated urbanization development [43].

$$\begin{array}{c}lnCd{u}_{it}={\beta }_0+{\beta }_{1}lnEd{l}_{it}+{\beta }_2\mathit{ln}I{s}_{it}+{\beta }_3\mathit{ln}O{d}_{it}+{\beta }_4\mathit{ln}L{e}_{it}+{\beta }_5\mathit{ln}Lp{i}_{it}+{\beta }_6\mathit{ln}E{\mathrm{s}}_{it}\\ +{\beta }_7\mathit{ln}E{p}_{it}+{\beta }_8\mathit{ln}E{r}_{it}+{\beta }_{9}lnPr{a}_{it}+{\beta }_{10}\mathit{ln}M{l}_{it}+{\beta }_{11}\mathit{ln}G{r}_{it}+{\epsilon }_{it}\end{array}$$

(10)

In the formula, the subscripts i and t represent the city and time, respectively; β₀ is the intercept term; β_i is the estimation coefficient, where i = 1, 2, 3,…, 11; ε_it is a random error term.

4.2.2. Stationarity Test and Cointegration Test

In order to avoid “pseudo regression”, LLC test and Fisher-ADF test are used to test the unit root of the panel data series’ stationarity [44]. Each variable has a unit root, and then the first-order difference value of the variable is tested. The results are as follows: there is no unit root for each variable at the significance level of 1% (

Table 4. Panel unit root test.

Variables	Unit Root Test Method				Conclusion
	LLC Test	p Value	ADF Test	p Value
Cdu	−3.6890	0.0000	25.8469	0.2570	Unstable
ΔCdu	−8.0315	0.0000	61.5357	0.0000	Stable
Edl	1.7596	0.9640	1.2875	1.0000	Unstable
ΔEdl	−7.8762	0.0000	56.0965	0.0000	Stable
Is	2.8562	0.9980	3.6359	1.0000	Unstable
ΔIs	−7.8770	0.0000	57.8961	0.0000	Stable
Od	−2.1088	0.0180	21.9523	0.4652	Unstable
ΔOd	−7.9785	0.0000	56.8962	0.0000	Stable
Le	−5.6987	0.5785	35.3587	0.1524	Unstable
ΔLpe	−7.5624	0.0000	58.3642	0.0000	Stable
Lpi	1.9856	0.8962	2.8752	1.0000	Unstable
ΔLpi	−8.0231	0.0000	55.2478	0.0000	Stable
Es	−9.2015	0.0000	62.0123	0.1238	Unstable
ΔEs	−10.9856	0.0000	65.3248	0.0000	Stable
EP	2.8544	0.9880	3.6527	1.0000	Unstable
ΔEP	−6.9854	0.0000	54.2178	0.0000	Stable
Er	2.5698	0.8996	2.9865	1.0000	Unstable
ΔEr	−8.7586	0.0000	59.3654	0.0000	Stable
Pra	−4.8790	0.0000	30.0563	0.1524	Unstable
ΔPra	−7.3215	0.0000	57.3220	0.0000	Stable
Ml	−0.9360	0.1860	15.2647	0.9117	Unstable
ΔMl	−20.0452	0.0000	185.0153	0.0000	Stable
Gr	−4.3289	0.0000	22.9632	0.5007	Unstable
ΔGr	−108.6352	0.0000	84.6385	0.0000	Stable

Note: Δ represents the first-order difference of each variable, and the calculation of each statistic value comes from Eviews 8.0 software.

), that is, all variables are first-order single integration, which can be tested for cointegration relationship.

Based on robustness, Pedroni test and Kao test are used to test the cointegration between variables [45]. The results are as follows: both tests reject the original hypothesis of “no cointegration relationship” at the 5% significance level (

Table 5. Panel data cointegration test.

Test Method	Name of Statistics	Statistical Value	p Value	Test Results
Pedroni test	PanelPP-Statistic	−30.3865	0.0000	Rejected
	PanelADF-Statistic	−7.8841	0.0000	Rejected
	GroupPP-Statistic	−30.9789	0.0000	Rejected
	GroupADF-Statistic	−7.6952	0.0000	Rejected
Kao test	ADF	−2.2152	0.0110	Rejected

Note: In the Pedroni test, only panel ADF and group ADF statistics are used in this paper, because they have better small sample properties than other statistics and conform to the data characteristics of this study.

), that is, there is a long-term stable equilibrium relationship among the variables, which can be used for regression analysis of panel data.

4.2.3. Regression Analysis of Influencing Factors

In this study, the common panel econometric model and spatial panel econometric model will be used to highlight the spatial effect of the influencing factors of coordinated development of urbanization in Beijing-Tianjin-Hebei region. In order to avoid multicollinearity among variables in ordinary panel econometric model, variance inflation factor (VIF) test is carried out. The results are as follows: the VIF values of each explanatory variable are less than 5, indicating that there is no serious multicollinearity among the variables. In model (1), model (4) and model (5), mixed effect, fixed effect and random effect are used to test the influencing factors (

Table 6. Estimation of common panel model.

Explanatory Variables	Mixed Effect	Fixed Effect			Random Effect	VIF
	Model (1)	Model (2)	Model (3)	Model (4)	Model (5)
c	−2.8129 ***				−2.8158 ***
	(−11.3536)				(−10.3368)
Edl	0.1650 ***		0.1686 **	0.1693 ***	0.1658 ***	4.68
	(9.0099)		(11.5126)	(9.2576)	(8.5163)
Is	0.0249		0.0627 *	0.0569 **	0.0071	1.59
	(1.2007)		(3.5465)	(2.3319)	(0.2947)
Od	0.1368 ***		0.1368 **	0.1033 ***	0.1349 ***	1.83
	(8.5572)		(9.0673)	(8.7461)	(6.6735)
Le	0.0346 **		0.2145 *	0.0238 *	0.0352 **	2.98
	(2.1888)		(1.5647)	(1.4734)	(2.1745)
Lpi	−0.0135 *		−0.0056	−0.0048	−0.0232 *	4.21
	(−2.7504)		(−1.6984)	(−1.9725)	(−3.9124)
Es	0.0056 *	0.0096		0.0082	0.0042 *	3.59
	(1.9177)	(4.2450)		(1.7113)	(2.0543)
EP	−0.0068 *	−0.0095 **		−0.0092 **	−0.0037 **	2.65
	(−1.9188)	(−1.2459)		(−1.7123)	(−2.0653)
Er	0.0359 ***	0.0421 *		0.0369 **	0.0563 **	2.33
	(8.6695)	(1.8652)		(1.6385)	(2.3654)
Pra	0.0569 ***	0.069 **		0.0246 ***	0.0482 ***	4.21
	(3.2568)	(2.3647)		(1.3625)	(2.3485)
Ml	0.0637 **	0.0466 **		0.1036 **	0.0730 **	3.76
	(8.9652)	(1.9865)		(1.6311)	(8.2823)
Gr	0.0853 ***	0.0754 ***		0.0654 ***	0.0534 ***	2.11
	(9.6354)	(2.3654)		(1.8652)	(7.2458)
F-sata	363.9354	479.6523	515.0635	427.9685	453.0214
Test	F = 3.8289. p = 0.0000			Chisq = 87.8597. p = 0.0000
R2	0.6835	0.7952	0.7042	0.7152	0.7075

Note: *, ** and *** are significant at 10%, 5% and 1% levels, respectively, and the values in brackets in the table are t values.

). F test rejected the hypothesis of mixed effect model estimation, and Hausman test rejected the hypothesis of random effect model estimation. Therefore, the fixed effect model is the best choice in this study. Model (2) and model (3), as robustness tests, test the impact of other variables on the coordinated development of urbanization after eliminating some variables, respectively. The results are as follows: there is no significant change in the direction and significance level of the parameters to be estimated, indicating that model (4) has strong explanatory power.

After estimating the ordinary panel model, taking the whole regional districts and counties of Beijing, Tianjin and Hebei as the basic units, arcgis10.1 software is used to test the spatial autocorrelation of the residuals of the spatial regression model based on OLS estimation to identify whether it is necessary to further incorporate the spatial effects. Additionally, this paper uses Lagrange multiplier (LM) to further test the spatial dependence of coordinated development of urbanization, and makes comparison and selection of spatial lag model and spatial error model (

Table 7. Spatial dependence test.

Test Method	Test Value	df	Small Probability p
Moran’s I	0.4378	-	0.0000
LMerr	201.4641	1	0.0000
LMlag	254.3258	1	0.0000
RLMerr	21.9842	1	0.0370
RLMlag	70.4257	1	0.0000
SARMA	261.7384	2	0.0000

).

Based on the above test results, the following aspects can be found: the Moran’s I value of regression error term is 0.4378, p value is 0.0000 and the spatial autocorrelation of residual term shows a significant trend. Moreover, the test value of spatial error term LMerr and spatial lag test value LMlag obtained by Lagrange multiplier (LM) test are also significant at the level of 1%. Therefore, the regression model rejects the hypothesis that there is no spatial dependence, and further verifies the necessity of incorporating spatial effects. From the test results, LMlag is greater than LMerr, and its robustness test value RLMlag is also greater than RLMerr, and the corresponding significance level RLMlag is higher than RLMerr. As a result, for the data of this study, the spatial lag model is better than the spatial error model. As for the choice of models, the two models have passed the significance level test of 5%. Finally, four results of the two models are listed for comparative analysis.

From the view of regression results, the following aspects can be obtained (

Table 8. Estimation of spatial panel model.

Explanatory Variables	Spatial Error Model (SEM)		Spatial Lag Model (SLM)
	Fixed Effect	Random Effect	Fixed Effect	Random Effect
	Model (6)	Model (7)	Model (8)	Model (9)
c		−3.0568 ***		−2.2774 ***
		(−10.0375)		(−9.3752)
Edl	0.1158 ***	0.1573 ***	0.1298 ***	0.1327 ***
	(7.8112)	(7.6957)	(7.2354)	(7.4121)
Is	0.0785 **	0.0306	0.0506 **	0.0273
	(2.1053)	(0.8964)	(2.0307)	(0.9445)
Od	0.0899 ***	0.0853 ***	0.0974 ***	0.0732 ***
	(4.0005)	(2.6587)	(4.2581)	(2.0655)
Le	0.0287	0.0299 *	0.0235	0.0374
	(1.4059)	(1.2345)	(1.3962)	(1.5471)
Lpi	−0.0034	−0.0193	−0.0043	−0.0202
	(−1.0765)	(−1.9822)	(−1.5347)	(−2.0954)
Es	0.0093	0.0069	0.0103	0.0073
	(2.0052)	(0.9658)	(2.3674)	(2.0098)
EP	−0.0206 **	−0.0098 **	−0.0165 ***	−0.0037 **
	(−4.3323)	(−3.1358)	(−4.4285)	(−2.1985)
Er	0.0456 **	0.0396 **	0.0473 *	0.0396 **
	(5.9868)	(4.1387)	(5.6417)	(3.9674)
Pra	0.0266 *	0.0387 *	0.0372 *	0.0488 *
	(1.3944)	(1.9685)	(1.5843)	(2.0124)
Ml	0.2249 *	0.0467 **	0.2158 *	0.0576 **
	(1.8783)	(2.3654)	(1.7564)	(3.8574)
Gr	0.0543 **	0.0634 **	0.0772 **	0.0556 **
	(1.8736)	(1.8654)	(1.9690)	(2.1637)
λ	0.4332 **	0.5006 **
ρ			0.3359 **	0.4153 **
sphtest	chisq = 20.1105.	p = 0.0650	chisq = 37.0534	p = 0.0003
R2	0.8009	0.7593	0.9396	0.8662

Note: *, ** and *** are significant at 10%, 5% and 1% levels, respectively, and the values in brackets in the table are t values.

): among the four models, the spatial error regression coefficient λ and the spatial lag regression coefficient ρ are all positive, and they all pass the significance level test of 5%. From the perspective of spatial lag regression coefficient, the coordinated development of urbanization in Beijing-Tianjin-Hebei region has spillover effect in geographical space. In terms of the spatial error coefficient, the random impact of various unobserved factors also plays a certain role in the transfer between regions. Model (6) and model (7) are spatial error models under fixed effect and random effect, respectively. The results of Hausman test are as follows: The chisq value did not pass the 5% significance test, so the random effect model was slightly better than the fixed effect model, but its fitting degree was relatively low. Therefore, these two models are only for reference. Model (8) and model (9) are spatial lag models with fixed effect and random effect respectively. The chisq value of Hausman test is very significant, so the fixed effect model is obviously better than the random effect model in the spatial lag model. Through comprehensive comparison, the fixed effect spatial lag model (8) is adopted.

Comparing the regression results of model (4) and model (8), the following aspects can be obtained: after considering the spatial effect, the significance level of some factors on the coordinated development of urbanization in Beijing-Tianjin-Hebei region has not changed. Among them, in the economic aspect, the influencing factors all show a significant positive promoting effect. The level of economic development and the degree of opening to the outside world are significant at the level of 1%, and the industrial structure is also significant at the level of 5%; in terms of land, the influencing factors of land financial revenue are negative and not significant; in terms of ecological environment, the influencing factors of ecological environment are all positive and not significant. Therefore, the spatial effect of the above factors on the coordinated development of urbanization in Beijing-Tianjin-Hebei region is not very obvious.

Considering the spatial effect, the significance level of some factors on the coordinated development of urbanization in Beijing-Tianjin-Hebei region has changed. Among them, in terms of population, the significant level of population transfer and employment of influencing factors has decreased, from the original 10% of the positive significant decline to not significant; in the aspect of ecological environment, the significant level of ecological environmental pollution of influencing factors has increased from the original negative significant 5% to 1%; on the contrary, the significance level of ecological and environmental governance has decreased, that is, from the original positive significant 5% to 10%; in the social aspect, the significant level of the allocation of public resources of influencing factors has decreased, that is, from the original positive significant 1% to 10%; in terms of the market, the significant level of the marketization of the influencing factors has decreased, that is, from the original positive significant 5% to 10%; and in terms of government, the significant level of government regulation of influencing factors has decreased, that is, from original positive significant 1% to 5%.

5. Results Analysis

Although the coordinated development level of urbanization in Beijing-Tianjin-Hebei region is increasing year by year, the growth rate of each region is not the same, and the spatial differentiation phenomenon is obvious. Additionally, the influencing factors of coordinated development of urbanization are the fundamental reasons for the above phenomenon. Therefore, the influencing factors play an important role in action direction, significance level and spatial effect.

As for economy, it is the sustainable power of coordinated development of urbanization in Beijing-Tianjin-Hebei region. The level of economic development and the degree of opening to the outside world have been gradually improved, the industrial structure has been constantly adjusted and optimized, and the development trend of new economic momentum is good. After the spatial effect was included, although the significance of economic factors did not change, the β coefficient was slightly smaller. The results show that the promoting effect of economy on the coordinated development of urbanization has a certain spatial spillover effect, but the effect is not strong. As for the regional economic development of Beijing-Tianjin-Hebei region, it is still in the typical “core–periphery” mode. The imbalance of economic development level, the difference of industrial structure and the different degree of opening to the outside world restrict the coordinated development of regional urbanization to a certain extent. In the future, Beijing-Tianjin-Hebei region should not only maintain the dynamic role of economy on the coordinated development of urbanization, but also improve the economic spillover effect of core cities [7].

The significant level of the influence of population factors on the coordinated development of urbanization in Beijing-Tianjin-Hebei region is low; the essence of “people-oriented” urbanization development has not been highlighted; after the spatial effect is included, the significant decline shows that the population transfer and employment are not sufficient, and the spatial spillover effect should not be strong. The significant level of influencing factors in land has not changed, and has been negative and not significant. After the spatial effect is included, the β coefficient decreases, which indicates that there is still high dependence on land finance and competition of urban construction land expansion in Beijing, Tianjin and Hebei, and the negative spillover effect of land is still continuing. Furthermore, the further emphasizing the core position of population and reversing the negative impact of land will become the key points of future attention in order to ensure the steady progress of coordinated development process of urbanization in Beijing-Tianjin-Hebei region [14].

Although the negative ecological pressure significantly affects the coordinated development of urbanization, ecological governance plays a positive and significant role, which indicates that the regional ecological governance in Beijing-Tianjin-Hebei region has achieved certain results in recent years. After the spatial effects were included, the significant level of ecological pressure increased and the significant level of ecological governance decreased. It shows that the ecological governance has achieved initial results, and the contradiction between the “diffusion effect” of ecological pressure and the low degree of coordination among the governance areas still restricts the coordinated development process of urbanization in Beijing-Tianjin-Hebei region. Because the ecological destruction is not restricted by administrative boundaries [11] while the overall development requirements of the region have been ignored, the governance is still fragmented among regions [12], so the task of ecological governance in the process of coordinated development of urbanization in Beijing-Tianjin-Hebei region is still very arduous.

In terms of social factors, it has a positive and significant impact on the coordinated development of urbanization in Beijing-Tianjin-Hebei region. The expansion and allocation efficiency of public resources, to a large extent, determine the quality of urbanization development, meet the requirements of “people-oriented” urbanization development, and provide guarantee for the coordinated development of various elements of urbanization [13]. After taking into account the spatial effect, the significant level of the impact of public resource allocation on the coordinated development of urbanization is significantly reduced. This shows that there is a serious imbalance in the allocation of public resources in Beijing-Tianjin-Hebei region, which, to a large extent, limits the coordinated development of regional urbanization. As for the unbalanced allocation of public resources, it is also an important reason for the population expansion of large cities in Beijing-Tianjin-Hebei region, the lack of absorption capacity of small and medium-sized cities, and the difficulty in forming a multicenter urbanization pattern [27].

As for the level of marketization and government regulation factors, they have a significant positive impact on the coordinated development of urbanization in Beijing-Tianjin-Hebei region, which is inseparable from the joint role of the government and the market. Additionally, the external environment of coordinated development of urbanization is composed of government policy, financial support and market maturity [46]. After taking into account the spatial effect, the significant level of the impact of marketization level and government regulation factors on the coordinated development of urbanization has decreased, which indicates that there is no unified market-oriented environment in Beijing-Tianjin-Hebei region. Meanwhile, due to the strong intervention of local governments, it is difficult to achieve the coordinated development of internal factors of urbanization but also the coordinated development of urbanization among regions. By properly handling the relationship between the government and the market, it is the key point to optimize the external environment for the coordinated development of urbanization in Beijing-Tianjin-Hebei region.

6. Conclusions and Discussion

In this study, starting from the connotation of coordinated development of urbanization, based on the analysis of the mechanism of coordinated development of urbanization, with the help of coupling coordination model, fixed effect model and spatial lag model, the level and influencing factors of coordinated development of urbanization in Beijing-Tianjin-Hebei region are analyzed. Furthermore, the change of the effect of influencing factors before and after the spatial effect is compared. The results are as follows: although the coordinated development level of urbanization in Beijing-Tianjin-Hebei region is increasing year by year, the regional gap is increasing and the spatial differentiation is serious. The coordinated development of urbanization is still in the typical “core-periphery” mode. After the spatial effect is included, the significance level of the influencing factors changes, and the promotion effect on the coordinated development of urbanization is reduced. The performance is as follows: the core of “people-oriented” urbanization development has not been highlighted; the spatial spillover effect of economic development is low; the regional economic development is unbalanced; the dependence of each region on land finance is too strong; the habitat destruction and expansion are serious; the coordination degree of habitat governance is low; the imbalance between public resource allocation and public service supply; the local government intervention is too strong, and there is a lack of a unified market-oriented environment. To sum up, the above phenomena seriously restrict the coordinated development of urbanization in Beijing-Tianjin-Hebei region.

The problems exposed in the hot spots, with certain representativeness, provide some important points for the coordinated development of urbanization. Therefore, the future policies and reforms should have a tendency. The first is to accurately grasp and deeply understand the coordinated development thought in the content of the national strategy, guided by the national strategies of new urbanization, regional coordinated development and Beijing-Tianjin-Hebei coordinated development. While paying attention to the coordinated development of various “human centered” elements in the process of urbanization, we should give full play to the driving role of advantageous areas to realize the overall efficiency improvement of urbanization development. Secondly, we should actively break through the institutional barriers that restrict the coordinated development of urbanization, strengthen the construction of regional cooperation mechanism and build a platform for coordinated development. In the process of coordinated development of urbanization, we should not only clarify the overall common interests, but also clarify the responsibilities shared by different regions. The third is to deepen the reform of the government administrative system, guide the coordinated development of urbanization, correctly handle the relationship between the government and the market [46], and give full play to the role of regional market-oriented mechanism and the government’s macro-control, so as to optimize the external environment for the coordinated development of urbanization. Finally, we should continue to promote the reform of internal factors of coordinated development of urbanization, further deepen the reform of household registration system and land system, promote regional economic integration and form a comprehensive opening pattern. Furthermore, we should control the expansion of habitat destruction and strengthen collaborative governance, promote the allocation of public resources and the equalization of public service areas, optimize the internal system development state of urbanization and form the resultant force of coordinated development.

In this study, there are some limitations, as follows: the measurement index system of coordinated development of urbanization involves a wide range of aspects, but the management system of urbanization is still not considered. In the setting of influencing factors variables, the use of alternative indicators to characterize the level of marketization and government regulation may lead to certain errors in the regression results. It is considered that the regions with higher urbanization quality have better fitting with urban agglomerations, so the study object is selected in Beijing-Tianjin-Hebei urban agglomeration, and the study scope is not extended to a more macroscopic geographical scale due to the limited space. Therefore, the study will strive to overcome the above limitations to enhance the theoretical and practical nature of the study content.