Wanping Zhang1，Zhibin Zhang1*，Zhuandi Shi2
1College of Geography and Environmental Science, Northwest Normal University, Gansu Lanzhou 730070, China
2Natural Resources Bureau of Jingning County, Gansu Pingliang 743400，China
Corresponding author:：Zhibin Zhang
email：zbzhang@nwnu.edu.cn
Wanping Zhang：zhangwanping@aliyun.com
Zhuandi Shi:dd07222025@163.com
Abstract：The Chengdu-Chongqing Economic Circle (CCEC) is the most densely populated region in Western China. Against the backdrop of China’s increasingly complex population dynamics, studying county-level population agglomeration within this region holds significant importance. Based on county-level population data from 2010, 2015, and 2020, this study employs methods such as population geographic concentration, spatial autocorrelation, and spatial econometric models to analyze the spatial-temporal distribution and driving factors of population agglomeration in the CCEC. The results indicate that: ① From 2010 to 2020, the degree of county-level population agglomeration exhibited an increasing trend, with particularly notable growth after 2015. The process of population agglomeration evolved from “slow” to “rapid,” primarily concentrated in the smaller core counties, revealing an uneven population distribution. ② During 2010–2020, the Moran’s I value of population agglomeration steadily increased, indicating an overall enhancement in spatial correlation. The agglomeration structure was characterized by a global spatial pattern of “point-to-surface, multi-center” agglomeration, with the Chengdu-Chongqing “dual-core” as the center, radiating along two north-south “belts,” further extending parallel to the main east-west axis, and eventually dispersing into the three subregions (“three zones”). ③ High-high (HH) and low-high (LH) agglomeration zones were distributed in the “dual-core” region and its peripheral areas, while high-low (HL) agglomeration zones were scattered in the densely populated urbanized areas of southern Sichuan, and low-low (LL) zones were primarily located in the peripheral areas of southern Sichuan and northeastern Chongqing, reflecting a local agglomeration pattern of “high core, low middle and periphery.” ④ From 2010 to 2020, urbanization rate, secondary industry proportion, and total retail sales of social consumer goods were the main drivers of population agglomeration. During the first five years, per capita GDP and the proportion of the tertiary industry also promoted population agglomeration but became insignificant by 2020. Geographical location and road density had no significant effect on population agglomeration.
Keywords：Chengdu-Chongqing Economic Circle (CCEC); population agglomeration; spatial-temporal distribution; driving factors
Introduction
On January 3, 2020, the development of the Chengdu-Chongqing Economic Circle (CCEC) was elevated to a major national strategy, alongside the Beijing-Tianjin-Hebei Region, the Yangtze River Delta, and the Guangdong-Hong Kong-Macao Greater Bay Area [1]. This strategic upgrade has attracted widespread attention [2]. As urban agglomeration integration has become an irreversible global trend in the urbanization process, population has begun to cluster in metropolitan regions and urban agglomerations at an unprecedented rate [3]. Population agglomeration not only impacts the future trajectory of urbanization within a region [4] but is also influenced by a combination of factors, including social development, economic conditions, industrial structure, and geographical location, all of which play a crucial role in regional development. Population agglomeration drives rapid socioeconomic development and urbanization, while also fostering regional social prosperity, cultural exchange, and intellectual innovation [5]. Currently, China is undergoing profound demographic changes, entering a critical transitional period in population studies in the context of a “new era” [6]. Against this backdrop, investigating the spatial-temporal distribution patterns and driving factors of population agglomeration in the CCEC is of significant practical importance.
In recent years, international studies on population in metropolitan regions have primarily focused on topics such as population and employment [8], population pressure [9], residential segregation of immigrant populations [10], population projection [11], and population growth and urban governance [12]. Beyond these topics, international research on population distribution has increasingly employed innovative models and methodologies. For example, Rahimi et al. [13] used taxi pickup and drop-off data to calculate trip volumes, allocate population data to individual communities, and develop models to estimate population distribution. Similarly, Burgi et al. [14] introduced a novel method for measuring spatial population concentration (SPC) to more accurately assess population density. Gibbs et al. [15] combined detailed location data from Facebook with census-derived population estimates to measure population redistribution and mobility. Domestically, the body of research on the spatial-temporal distribution of population and its influencing factors is extensive [16–22]. Findings from the Seventh National Population Census have revealed new trends in China’s population dynamics, directly advancing population research [23–26]. However, studies on the CCEC, as one of China’s key metropolitan regions, remain relatively limited. Existing studies predominantly adopt analytical methods such as spatial autocorrelation models [34], social network analysis [35], spatial simultaneous equation models [37], and coupled coordination models [28]. In terms of research scope, these studies have mainly focused on cities [27, 28], urban agglomerations [29–31], and the CCEC as a whole [35–39]. Regarding content and scale, research has addressed topics such as the spatial relationship between population and economic development [27, 36, 37], the coupled and coordinated development of population, land urbanization, and industries [28, 33, 39], urbanization [32], and population migration and mobility [29, 30, 35]. Overall, existing research predominantly focuses on population mobility, shrinkage, and aging. However, studies on population agglomeration patterns and their driving factors remain insufficient. Furthermore, most analyses of the CCEC emphasize cities, urban agglomerations, or metropolitan areas as the study unit, without examining the characteristics and influencing factors of population agglomeration at the county level.
Research on the spatial-temporal distribution and driving factors of population agglomeration sheds light on the dynamics and patterns of population clustering in the CCEC over different periods, while also delving into the underlying causes of population distribution imbalances. This paper aligns with the current context of population development in China and the strategic opportunities associated with the CCEC. The findings aim to provide the most accurate county-level population change data and serve as a critical basis for future urban population governance. From a developmental perspective, this research is of great significance for promoting high-quality development in the CCEC under the framework of Western China Development and offers valuable insights for government planning and decision-making regarding the construction of the CCEC.
- Study Area, Data Sources, and Research Methods
1.1 Study Area
The Chengdu-Chongqing Economic Circle is one of the most urbanized regions with significant development potential in Western China. It is a key component of the Yangtze River Economic Belt and the Belt and Road Initiative, serving as a pivotal element in the new development framework that emphasizes domestic circulation while promoting mutual reinforcement between domestic and international circulation. The CCEC holds a unique and critical position in this strategy. The CCEC’s geographic scope is roughly equivalent to the Chengdu-Chongqing Economic Zone and the Chengdu-Chongqing Urban Agglomeration. According to the “Outline of the Chengdu-Chongqing Economic Circle Development Plan,” issued by the Central Committee of the Communist Party of China and the State Council on October 20, 2021, the planning area includes 27 districts and counties in Chongqing, as well as parts of Kaizhou and Yunyang, and 15 cities in Sichuan Province, covering a total area of 185,000 square kilometers. To ensure the completeness of the study area and facilitate statistical analysis, this paper includes the entirety of Kaizhou District and Yunyang County in Chongqing within the study area. Additionally, Wulong District and Pengshui Miao and Tujia Autonomous County in Chongqing are incorporated into the study. In 2010, Wansheng District and Shuangqiao District in Chongqing were merged into Qijiang District and Dazu District, respectively, following administrative adjustments in October 2011. These adjustments are reflected in this study’s data processing.
According to the Seventh National Population Census, as of the end of 2020, the CCEC had a permanent population of nearly 98 million, comparable to the combined population of other urban agglomerations in Western China. The CCEC is now the region with the strongest comprehensive competitiveness in the west. Chengdu’s permanent population reached 20.938 million, an increase of 5.819 million compared to 2010, making it the fourth Chinese city to exceed 20 million residents after Beijing, Chongqing, and Shanghai. The population gap between Chengdu and Beijing has narrowed to less than 1 million. Meanwhile, Chongqing’s permanent population reached 32.0542 million, marking a significant increase compared to 2010 [7].
1.2 Data Sources
The socioeconomic statistical data related to the year-end permanent population, administrative area, and influencing factors of population agglomeration in the study area were obtained from the Sichuan Statistical Yearbook, Chongqing Statistical Yearbook, Chengdu Statistical Yearbook, statistical bureaus of different cities, the China Economic and Social Big Data Research Platform, China County Statistical Yearbook, and various city yearbooks. The population data for 2010 and 2020 were derived from the Sixth and Seventh National Population Census, respectively.
1.3 Research Methods
(1) Geographic Concentration: Geographic concentration is used to describe the spatial distribution of geographic elements in cities at a specific time. The population geographic concentration index is introduced to quantitatively measure the spatial concentration of the population within the study area:
(1)
Where: Rpopi: Population geographic concentration of region 𝑖. POPi: Total population of sub-region 𝑖. Si: Land area of sub-region. and : Total population and total land area of region 𝑖, respectively.
(2) Theil Index
The Theil Index is commonly used to determine whether a region exhibits a multi-center population distribution [3]. The formula is as follows:
(2)
Where: 𝑇: Theil Index. Pi : Population of city. : Average population of the Greater Bay Area (or the entire region).
(3) Exploratory Spatial Data Analysis (ESDA)
Global Spatial Autocorrelation Analysis: Moran’s I is used to reflect the similarity of attribute values between spatially adjacent units. Its value ranges from -1 to 1, with 𝐼 < 0 indicating negative correlation, 𝐼 >0 indicating positive correlation, and 𝐼 = 0 indicating no correlation. To study the overall spatial correlation characteristics of population geographic concentration and economic geographic concentration in the study area, Moran’s I is calculated as follows:
(3)
Where: 𝐼:Moran’s I index. : Number of regions. xi and xj : Attribute values of regions 𝑖 and 𝑗, respectively. : Mean attribute value. Wij : Spatial weight matrix.
Local Spatial Autocorrelation Analysis: Local spatial autocorrelation measures the spatial correlation degree of population geographic concentration in adjacent regions. Four types of clusters can be identified: high-high (HH), low-low (LL), high-low (HL), and low-high (LH). HH and LL indicate positive spatial autocorrelation with clustering of high (low) attribute values, while HL and LH indicate negative spatial autocorrelation where regions with high (low) values are surrounded by those with low (high) values. The formula is expressed as:
(4)
(4) Spatial Regression Models
The Spatial Lag Model (SLM) and Spatial Error Model (SEM) are special forms of spatial autoregressive models used to address spatial interaction effects among variables. SLM primarily examines whether variables in one region exhibit diffusion or agglomeration effects, analyzing the impact of population agglomeration in neighboring regions on the local region. SEM measures the extent to which errors in neighboring regions influence the observed values in the local region. Given the spatial autocorrelation in population agglomeration distribution, this study employs both SLM and SEM for comparative analysis to identify the optimal model for analyzing the influencing factors of population agglomeration.
Model Expressions:
SLM： (5)
SEM: (6)
(7)
Where: Y : Dependent variable. X: Independent variable. W1、W : Spatial weight matrices for the dependent variable. β: Coefficient of independent variables. ρ: Spatial correlation coefficient. ε、μ: Random errors, assumed to follow a normal distribution.
- Temporal Evolution Patterns of Population Agglomeration
2.1 Classification of Population Agglomeration Types
Population agglomeration is calculated using the geographic concentration index (Equation 1). To facilitate analysis, this study adopts the classification standards for county-level population agglomeration in China [40], tailored to the actual distribution of population agglomeration in the CCEC. The population agglomeration values, after data stretching (due to small values), primarily fall within the range of 2–4 (Fig. 1). Accordingly, the population agglomeration is divided into five categories based on the adjusted values: Highly Dense Population Areas(1.03≤R＜13.46).Low-Density Population Areas(0.3≤R＜1.06).Upper Density Balance Areas (0.2≤R＜0.3).Lower Density Balance Areas (0.08≤R＜0.2).Sparse Population Areas (0.02≤R＜0.08).
图1 2010-2020年人口集聚度分布图
Fig.1 Distribution map of population concentration from 2010 to 2020
2.2 Analysis of Population Agglomeration Types (2010–2020)
Based on the calculated population agglomeration values for the CCEC, this study obtained statistical data for different population agglomeration categories in 2010, 2015, and 2020 (Tab. 1) and visualized the data for each year (Fig. 2). The findings are as follows:
1)Uneven Population Distribution Concentrated in Highly Dense Areas: From 2010 to 2020, highly dense population areas accounted for just 1.67% of the total land area, significantly less than the 12.23% occupied by sparse population areas. However, the population density in highly dense areas exceeded that of sparse areas by 4,281 people/km² and was 3,847 people/km² higher than the largest land area category (lower density balance areas). Additionally, low-density population areas accounted for 11.05% of the land area, with an average population density of 930 people/km², indicating that the population predominantly clustered in small, economically developed core urban areas and central counties.
2)Faster Population Agglomeration (2015–2020):Although fewer counties experienced increased population agglomeration from 2015 to 2020 (13 counties) compared to 2010–2015 (21 counties), the total increase in agglomeration across all categories during 2015–2020 (17.67) far exceeded the increase during 2010–2015 (13.78). This indicates that population agglomeration accelerated between 2015 and 2020, particularly in highly dense population areas, whereas the earlier period saw slower agglomeration across various categories.
Tab.1 The classified statistics of population agglomeration about districts and counties in Chengdu-Chongqing Economic Circle
|
Population Agglomeration Type |
2010–2015 |
2015–2020 |
2010–2020 |
|||||
|
Change in County Number |
Change in Total Agglomeration Value |
Change in County Number |
Change in Total Agglomeration Value |
Average Area Proportion (%) |
Average Density (people/km²) |
|||
|
Highly Dense Population Areas |
0 |
+6.88 |
+6 |
+17.67 |
1.67 |
4465 |
||
|
Low-Density Population Areas |
+13 |
+4.77 |
-11 |
-4.42 |
11.05 |
930 |
||
|
Upper Density Balance Areas |
+8 |
+2.13 |
-1 |
-0.03 |
29.89 |
503 |
||
|
Lower Density Balance Areas |
-18 |
-2.54 |
+7 |
+1.08 |
45.16 |
618 |
||
|
Sparse Population Areas |
-4 |
-0.15 |
0 |
-0.02 |
12.23 |
184 |
||
Note:“+” indicates an increase; “-” indicates a decrease.
3) Evolution of Population Agglomeration Types (Fig. 2).The spatial distribution of population-dense areas evolved as follows: 2010–2015: Population agglomeration radiated outward from the core urban areas of Chengdu and Chongqing, forming concentric zones. The main radiation directions extended north-south along the Chengdu-Deyang-Mianyang-Leshan corridor and the riverine urban belt, covering key counties in these areas. Post-2015: Population agglomeration intensified within the metropolitan regions centered around the Chengdu-Chongqing “dual-core,” spreading parallel to the Chengdu-Chongqing axis along an east-west direction. This trend subsequently dispersed into the “three zones” surrounding the metropolitan core. 2010–2020: The upper-density balance areas were primarily distributed around population-dense zones, connecting to the southern Sichuan urban cluster, the Nanchong-Suining-Guang’an urban cluster, and the Dazhou-Wanzhou urban cluster, which collectively formed a roughly “V-shaped” region. Meanwhile, the lower-density balance areas and sparse population areas were concentrated in northeastern and southeastern Chongqing, as well as peripheral areas in southern Sichuan, Ya’an, and Leshan.
From the changes in the number of counties and the variation in population agglomeration values (Tab. 1), the following observations can be made: 2010–2015: During this period, population agglomerated toward highly dense, low-density, and upper-density balance areas. Among these, the low-density areas saw the largest increase in the number of counties, while the number of counties in the highly dense areas remained unchanged. However, the highly dense areas experienced the largest increase in population agglomeration values, indicating that population primarily concentrated in these highly dense areas.2015–2020: In the latter five years, the number of counties in highly dense areas increased less than that in the lower-density balance areas, but the growth in population agglomeration values was significantly higher. This suggests that population agglomeration toward highly dense areas became more pronounced during this period. Overall, the Chengdu-Chongqing Economic Circle exhibited a hierarchical diffusion pattern over the decade, characterized by a “dual-core—two-belts—three-zones” structure. Population distribution gradually dispersed outward in a stepwise manner, with the population center shifting northwestward, indicating an increasing concentration of population toward Chengdu. This trend highlights a “strong west, weak east” distribution pattern within the region.
Fig.2 Spatial distribution of population concentration degree from 2010 to 2020
- Spatial Evolution Pattern of Population Agglomeration
3.1 Multi-Center Distribution Pattern and Population Gravity Shift
This study calculates the Theil Index of the Chengdu-Chongqing Economic Circle population based on data from 2010, 2015, and 2020 (Table 2). The Theil Index is used to determine whether a region exhibits a multi-center population distribution and to analyze the evolution of the spatial layout of the population. A higher Theil Index indicates a more pronounced trend of multi-center population distribution [3].
Tab. 2 Theil Index of Population in Economic Circle Chengdu-Chongqing Region
from 2010 to 2020
|
Year |
2010 |
2015 |
2020 |
|
Theil Index |
0.1264 |
0.1259 |
0.1726 |
From Table 2, it can be observed that the Theil Index of the Chengdu-Chongqing Economic Circle population slightly decreased in 2015, but the change was insignificant, indicating that the multi-center distribution of the population was not prominent in the first five years. However, the Theil Index rose rapidly between 2015 and 2020, increasing from 0.1259 to 0.1726. This suggests that population dispersion accelerated during the latter five years, forming new agglomeration points and transitioning toward a multi-center population distribution trend. For instance, the establishment of Chengdu Tianfu New Area in 2014 and Chengdu Eastern New Area in 2020 created new strategic hubs for national westward and southward opening. These regions emerged as new growth poles driving the integrated development of Chengdu and Chongqing, resulting in a noticeable westward population flow within Chongqing during this period.
3.2 Spatial Correlation Analysis of Population Agglomeration
3.2.1 Global Spatial Autocorrelation Analysis
Using Global Spatial Autocorrelation analysis, the Moran’s I Index of county-level population concentration in 2010, 2015, and 2020 was calculated (Table 3). The results show that the Moran’s I values of the Chengdu-Chongqing Economic Circle were positive in all three years and exhibited an increasing trend. The Z-values were all greater than 1.96, indicating that population agglomeration was positively correlated and that the spatial autocorrelation of county-level population distribution strengthened over time. This suggests that the spatial dependency of population agglomeration within counties (districts) became increasingly significant.
Tab. 3 Moran＇s I value of population concentration degree of districts and counties in Chengdu-Chongqing Economic Circle
|
Year |
Moran’s I |
P-Value |
Z-Value |
|
2010年 |
0.45867 |
0.001 |
11.1811 |
|
2015年 |
0.503049 |
0.001 |
12.0166 |
|
2020年 |
0.540387 |
0.001 |
12.7644 |
3.2.2 Local Spatial Autocorrelation Analysis
To explore the local spatial relationships of county-level population distribution in the Chengdu-Chongqing Economic Circle, a Local Indicators of Spatial Association (LISA) analysis was conducted on the population geographic concentration index (Fig. 4). Due to significant population distribution differences, a distance-based spatial weight matrix with a threshold of 62,768.826 was used. LISA cluster maps for 2010, 2015, and 2020 were generated. The results reveal the following:
1)Distribution Characteristics of Different Agglomeration Types: 2010: High-High (HH) agglomeration areas were mainly concentrated in the core urban areas of Chengdu and Chongqing. These areas had strong development foundations, conditions, and trends, forming a clear “core-periphery” structure. This created a “magnetic effect” and competitive advantage, with significantly stronger population attraction compared to other counties. High-Low (HL) agglomeration areas were sporadically distributed along the “Chengdu-Deyang-Mianyang-Leshan Urban Belt” and the “Southern Sichuan Urban Cluster,” such as Leshan’s Central District and Naxi District. Low-High (LH) agglomeration areas were distributed in the outer periphery of the Chengdu-Chongqing “dual-core,” while Low-Low (LL) agglomeration areas were clustered in less developed regions in southwestern Sichuan (e.g., parts of Leshan and Ya’an) and northeastern Chongqing (e.g., Wulong and Pengshui).
2015: The area of HH agglomeration decreased. Counties like Pengzhou and Zhongjiang shifted from HH to LH, Naxi District transitioned from HL to insignificant, and Qionglai City shifted from HH to LH. Qianjiang District changed from insignificant to LL, indicating that HH agglomeration was primarily concentrated in Chengdu’s core area during this period. 2020: New HH agglomeration areas emerged in Chongqing, such as Banan District and Dadukou District, which transitioned from LH to HH. Yongchuan District changed from insignificant to HH, while Chengdu saw no significant changes. Counties like Wutongqiao transitioned from HL to insignificant, and Hechuan District shifted from LH to insignificant, while Qianjiang District moved from LL to insignificant. These changes suggest that HH agglomeration areas in this period were increasingly concentrated around Chongqing. The primary spatial correlation patterns in the study area were HH and LL, with HL and LH patterns showing relatively minor changes. This further confirms the uneven population agglomeration characteristics in the Chengdu-Chongqing region.
2)Spatial Distribution Pattern of Population Agglomeration: Overall, the population agglomeration pattern in 2010, 2015, and 2020 exhibited a trend of radiating outward from the Chengdu-Chongqing dual-core along “two belts,” spreading parallel to the “main axis,” and eventually dispersing into “three zones.” This formed a spatial pattern characterized by “multi-center distribution, central axis depression, and cluster-based development.” The spatial structure of “one axis, two belts, dual-core, and three zones” remained relatively stable, with the Chengdu-Chongqing dual-core acting as the primary drivers. However, there was insufficient agglomeration along the central axis of counties, highlighting the need to weaken this fixed development pattern and enhance coordinated development between Chengdu and Chongqing.
- Influencing Factors of Population Agglomeration in Chengdu-Chongqing Economic Circle
The positive spatial autocorrelation of population agglomeration in the Chengdu-Chongqing Economic Circle (Table 2) indicates that population distribution is not random but rather reflects strong spatial interactions. Therefore, Spatial Lag Model (SLM) and Spatial Error Model (SEM) analyses were used to examine the influencing factors of population agglomeration.
Fig. 3 Local spatial autocorrelation of population geographic concentration of the districts and counties in 2010、2015 and 2020
4.1 Selection of Indicators
Population agglomeration is primarily influenced by a combination of socioeconomic and geographic factors. Based on the spatiotemporal analysis of population agglomeration in the Chengdu-Chongqing Economic Circle and existing literature, eight explanatory variables were selected, covering social, economic, and geographic aspects (Table 3).
Tab. 3 The units and Settings of variables
|
Variable |
Variable Type |
Unit and Settings |
|
|
Dependent Variable |
Y |
Population Agglomeration (LnY) |
Natural logarithm of population agglomeration |
|
Social Factors |
Social development |
Urbanization rate (%) ln Medical personnel |
﹪ 10,000 persons |
|
Economic Factors |
Economic level Industry proportion Consumption demand |
GDP per capita Proportion of the secondary industry Proportion of the tertiary industry Per capita retail sales of consumer goods |
Thousand RMB/person Proportion of the secondary industry in GDP (%) Proportion of the tertiary industry in GDP (%) Thousand RMB/person |
|
Geographic Factors |
Transportation factors Geographic location |
Road length/area in urban districts, county-level cities, and counties |
Road length/area Urban district/county-level city = 1, county = 0 |
4.2 Regression Results and Analysis of Influencing Factors
4.2.1 Regression Results
Using Geoda 1.1.2, a spatial contiguity weight matrix based on Rook’s principle was constructed to estimate and test the spatial econometric model of factors influencing population agglomeration (Table 4). The results show that in 2010 and 2015, the Spatial Error Model (SEM) had higher log-likelihood values (Log-L), lower AIC values, and larger R² values. A higher Log-L and R² indicate a better fit and greater explanatory power of the model for the dependent variables. This suggests that in 2010 and 2015, the spatial effects of population agglomeration were reflected in spatial error terms, making the SEM more suitable. In contrast, in 2020, the Spatial Lag Model (SLM) exhibited a higher Log-L, lower AIC, and larger R², indicating that the SLM provided a better fit for this period. Therefore, this study analyzes the influencing factors of population agglomeration using the SEM for 2010 and 2015 and the SLM for 2020.
Tab. 4 Results of model estimation of influencing factors of population agglomeration
|
2010 |
2015 |
2020 |
||||||||||
|
Variable |
SLM |
SEM |
SLM |
SEM |
SLM |
SEM |
||||||
|
Coef. |
S.E. |
Coef. |
S.E. |
Coef. |
S.E. |
Coef. |
S.E. |
Coef. |
S.E. |
Coef. |
S.E. |
|
|
Intercept |
0.002 |
0.058 |
2.315 |
0.415 |
-0.582 |
0.465 |
1.665 |
0.458 |
0.324 |
0.701 |
2.398 |
0.692 |
|
Urbanization |
0.009** |
0.004 |
0.012*** |
0.003 |
0.012*** |
0.004 |
0.017*** |
0.004 |
0.031*** |
0.005 |
0.036*** |
0.005 |
|
GDPper |
0.005 |
0.007 |
0.017*** |
0.006 |
-0.002 |
0.003 |
0.006** |
0.003 |
0.006* |
0.003 |
-0.003 |
0.003 |
|
Sec |
0.011*** |
0.004 |
0.002 |
0.004 |
-0.003 |
0.004 |
0.001 |
0.004 |
0.001* |
0.001 |
0.001** |
0.001 |
|
Ter |
0.023*** |
0.006 |
0.032*** |
0.005 |
0.001 |
0.006 |
0.024*** |
0.006 |
0.003 |
0.006 |
0.008 |
0.006 |
|
Trscg |
-0.008 |
0.008 |
0.013* |
0.007 |
0.011** |
0.005 |
0.001 |
0.004 |
0.012*** |
0.004 |
0.014*** |
0.004 |
|
Ln medical |
0.218* |
0.113 |
0.282*** |
0.100 |
0.213 |
0.124 |
-0.156 |
0.134 |
-0.077 |
0.184 |
0.381** |
0.191 |
|
Density |
0.040 |
0.029 |
0.016 |
0.026 |
0.189*** |
0.056 |
0.140** |
0.058 |
-0.002 |
0.028 |
-0.011 |
0.028 |
|
District |
0.251** |
0.103 |
0.181** |
0.092 |
0.084 |
0.085 |
0.073 |
0.082 |
0.020 |
0.113 |
0.037 |
0.112 |
|
Spatial lag term |
0.676* |
0.058 |
0.644* |
0.054 |
0.464* |
0.070 |
||||||
|
Spatial error term |
0.838** |
0.047 |
0.802* |
0.054 |
0.557* |
0.088 |
||||||
|
R2 |
0.788 |
0.827 |
0.866 |
0.868 |
0.780 |
0.771 |
||||||
|
Log-L |
-105.884 |
-97.907 |
-67.399 |
-72.209 |
-112.481 |
-117.058 |
||||||
|
AIC |
231.769 |
213.813 |
154.797 |
162.417 |
244.962 |
252.116 |
||||||
|
N |
144 |
144 |
144 |
144 |
144 |
144 |
Note: ***, **, * indicate significance levels at 1%, 5%, and 10%, respectively.
4.2.2 Analysis of Factors Influencing Population Agglomeration Distribution
Social Factors：Urbanization determines the future flow of population and is the fundamental driver of population agglomeration. From 2010 to 2020, the urbanization rate had a significant impact on the population distribution pattern in the Chengdu-Chongqing Economic Circle. It was consistently significant over the three years, with its contribution increasing over time. During this decade, with the continuous economic and social development of the Chengdu-Chongqing Economic Circle, the rapid flow of production factors such as population, the improvement of the urban system, and the urbanization of central urban areas and major counties significantly enhanced their ability to absorb surrounding populations. However, urbanization rates within the region showed significant disparities. Between 2010 and 2018, the urbanization rate in the Chengdu-Chongqing Economic Circle increased slowly by an average of 1.5% annually. By 2018, Chengdu’s urbanization rate had reached 73.12%, and Chongqing’s main urban area had an even higher rate of 90.51%, holding a significant advantage[32]. This aligns with the observed concentration of population in major districts and counties of Chengdu and Chongqing. High urbanization has brought about more employment opportunities, attracting an increasing number of migrants and gradually expanding the population scale. The influence of medical personnel on population agglomeration was only significant in 2010 but was not significant in 2015 or 2020. This indicates that with urbanization, population agglomeration is influenced by factors beyond basic medical services. However, in theory, the development of healthcare plays a direct role in influencing birth and death rates. By reducing mortality caused by various diseases, it lowers the overall death rate and extends average life expectancy.
Economic Development Factors：Economic development serves as the foundation for population existence and growth and is a critical driver of population agglomeration. GDP per capita (GDPper) was highly significant in 2010 and 2015 but showed a weakened significance in 2020. Generally, GDP per capita is positively correlated with population agglomeration. However, after 2015, as the population in the Chengdu-Chongqing Economic Circle grew at a pace far exceeding economic growth, the influence of GDP per capita on population agglomeration diminished. The proportion of the secondary industry (Sec) showed an increasing trend in significance, which is closely related to national policies. The promotion of the “Chengdu-Chongqing Urban Agglomeration” in 2016 and the “Chengdu-Chongqing Economic Circle” in 2020 attracted high-end resources and industries, strengthening population agglomeration, increasing the region’s population-carrying capacity, and creating a more inclusive space for urbanization. The increase in population aligns with the logic of “people following industries.” The significance of total retail sales of consumer goods (Trscg) increased over time, with the strongest significance observed in 2020. This indicator has a strong regional characteristic and remains relatively stable, making it one of the key metrics for measuring population agglomeration. As the Chengdu-Chongqing Economic Circle continues to develop positively, its influence on population agglomeration will become increasingly prominent. The proportion of the tertiary industry (Ter) was highly significant in 2010 and 2015 but showed reduced significance in 2020. This indicates that the more developed the county economy and the more rational the industrial structure, the higher the proportion of people employed in the tertiary industry, particularly in service-related sectors, which promotes population agglomeration. However, the COVID-19 pandemic that began in late 2019 severely impacted the tertiary industry, hindering population mobility and reducing its role in population agglomeration.
Geographical Factors：The geographic location variable was significant only in 2010 but not in 2015 or 2020, suggesting that geographic location played a diminishing role in population agglomeration over time. Its influence shifted from strong to weak before and after 2015. The influence of road density on population agglomeration was significant at the 5% level in 2015 but not in 2010 or 2020. In the study area, the population is mainly concentrated in smaller core districts and counties, while peripheral counties with mountainous terrain and less efficient transportation systems often have higher road density without a corresponding population concentration. This aligns with the reality of the transportation conditions in the region, where road density does not significantly influence population agglomeration.
- Conclusions and Discussion
5.1 Conclusions
(1) Temporal Perspective: Distribution of Different Population Agglomeration Types：Between 2010 and 2020, the Chengdu-Chongqing Economic Circle experienced a transition from “slow” to “rapid” population agglomeration, with a significantly uneven population distribution. Population-dense areas were mainly located in smaller but economically developed core urban areas of Chengdu and Chongqing, extending outward to surrounding zones. Moderate-density population areas were distributed around these dense areas and connected with city clusters in southern Sichuan, southern Suining-Guangan, and Dazhou-Wanzhou, forming a “V”-shaped region. Low-density and sparse population areas were mainly located in the underdeveloped regions of northeastern and southeastern Chongqing and the southern fringes of Sichuan.
(2) Spatial Perspective: Structure of Different Spatial Agglomeration Types：From 2010 to 2020, the population center shifted slightly from the southeast to the northwest, with an increase in the Moran’s I index of population concentration, indicating stronger spatial correlation of population agglomeration. High-High (HH) and Low-High (LH) agglomeration areas exhibited a clear “core-periphery” structure. High-Low (HL) agglomeration areas were scattered in Leshan and Luzhou, while Low-Low (LL) agglomeration areas formed contiguous zones. Between 2015 and 2020, the Theil index rose rapidly, indicating a clear trend toward multi-center population agglomeration.
(3) Drivers of Population Agglomeration： Between 2010 and 2020, urbanization rate, the proportion of the secondary industry (Sec), and total retail sales of consumer goods (Trscg) were the most significant factors positively influencing population agglomeration. In 2010 and 2015, the main drivers were GDP per capita (GDP per) and the proportion of the tertiary industry (Ter). The influence of road density was significant only in 2015 at the 5% level. The impact of geographic location was significant only in 2010 but not in 2015 or 2020.
5.2 Discussion
(1) The highly concentrated population and economic strength of the “dual cores” of Chengdu and Chongqing, underpinned by strong industrial bases, abundant high-end resources, and exceptional innovation capabilities, exert a powerful siphoning effect on population agglomeration. However, the intermediate areas between the two metropolitan circles have significantly lower levels of population agglomeration. For the Chengdu-Chongqing Economic Circle to become a key driver of high-quality development in western China, it is crucial to address the “central collapse” issue by exploring the development potential of counties along the Chengdu-Chongqing axis and promoting better regional coordination. Since 2015, the Chengdu-Chongqing Economic Circle has been elevated to a national strategic priority. Policy support has accelerated regional economic development, with population-dense areas attracting more people compared to the previous five years. Regional initiatives, such as the “westward expansion” and the construction of a science city in Chongqing’s western region, as well as the establishment of the Chengdu Eastern New Area, facilitate the flow of population, economy, and information between the two cities. These efforts aim to break the “layered development” model of the Chengdu-Chongqing dual cores and guide the formation of a multi-polar, group-based development pattern.
(2) The HH agglomeration areas in the Chengdu-Chongqing Economic Circle have formed two stable core growth poles. The LH agglomeration areas are mainly located within parts of the Chengdu and Chongqing metropolitan circles and overlap with urban belts, featuring developed industries and convenient transportation, yet they lag behind the core urban areas of Chengdu and Chongqing. HL agglomeration areas, serving as new secondary growth poles, are scattered in cities like Leshan and Luzhou, where industrial and transportation advantages drive higher population agglomeration compared to surrounding regions. LL areas, characterized by harsh natural environments, lagging economies, and sparse populations, remain underdeveloped.
(3) Population agglomeration is the result of multiple factors working in tandem. Due to limitations in data availability, there are some gaps in the selection of driving factors, particularly in areas such as healthcare, transportation, and geographic location. Geographical location showed a weakening influence on population agglomeration from 2015 onward, likely due to the concentration of population growth in the core urban areas of the Chengdu-Chongqing “dual cores” with fewer counties (Table 1). Overall, urbanization rate, the proportion of the secondary industry, and total retail sales of consumer goods were the most significant factors influencing population agglomeration. This is closely tied to urbanization efforts in Chengdu and Chongqing and their active role in accommodating the westward relocation of industries from eastern China.
- References
[1] Zhou Chunshan. Utilizing Population Agglomeration Effects to Build Modernized Urban Agglomerations and Metropolitan Areas[J]. National Governance, 2021, 000(031): 19-24.
[2] Sheng Yinan, Yang Xuyu. Spatial Patterns and Mechanisms of Migrant Population Agglomeration in China’s Three Major Urban Agglomerations[J]. Population and Economics, 2021(6): 20.
[3] Chen Wensheng. Fundamental Characteristics and Historical Evolution of County-Level Development in China[J]. China Development Observation, 2014(6): 2. DOI: CNKI:SUN:FZGC.0.2014-06-012.
[4] Tang Chunyan. Spatiotemporal Differentiation and Influencing Factors of County-Level High-Quality Development in the Chengdu-Chongqing Twin-City Economic Circle[D]. Chongqing Technology and Business University, 2023. DOI: 10.27713/d.cnki.gcqgs.2022.000214.
[5] Zhai Zhenwu, Jin Guangzhao. China’s Negative Population Growth: Characteristics, Challenges, and Countermeasures[J]. Population Research, 2023, 47(2): 10.
[6] He Dan, Liu Zhongyi. Rethinking China’s Population Development Strategy from the Perspective of Population Security[J]. Population Research, 2023, 47(2): 9.
[7] Sarkar R, Hasnine M. Population Pressure and Urban Sprawl in Kolkata Metropolitan Area[J]. 2020.
[8] Benassi F, Bonifazi C, Heins F, et al. Comparing Residential Segregation of Migrant Populations in Selected European Urban and Metropolitan Areas[J]. Spatial Demography, 2020, 8(3): 269-290.
[9] Thurmer J, Taylor AJ. Internal Return Migration and the Northern Territory: New Migration Analysis for Understanding Population Prospects for Sparsely Populated Areas[J]. Population Research and Policy Review, 2020.
[10] Vinci S, Egidi G, Gay AL, et al. Population Growth and Urban Management in Metropolitan Regions: The Contribution of Natural Balance and Migration to Polycentric Development in Barcelona[J]. Applied Spatial Analysis and Policy, 2021: 1-24.
[11] Rahimi F, Sadeghi-Niaraki A, Ghodousi M, et al. Modeling Population Spatial-Temporal Distribution Using Taxis Origin and Destination Data[J]. Sustainability, 2021, 13.
[12] Burgi C, Gorgulu N. The Impact of the Spatial Population Distribution on Economic Growth[J]. Working Papers, 2021.
[13] Gibbs H, Waterlow NR, Cheshire J, et al. Population Disruption: Estimating Changes in Population Distribution in the UK During the COVID-19 Pandemic[J]. 2021.
[14] Fountoulakis KN, Karakatsoulis G, Abraham S, et al. Results of the COVID-19 Mental Health International for the General Population (COMET-G) Study[J]. 2021.
[15] Kwon HS, Jeong YW, Kim SH, et al. Annual Incidence of Dementia from 2003 to 2018 in Metropolitan Seoul, Korea: A Population-Based Study[J]. Journal of Clinical Medicine, 2022, 11(3). DOI: 10.3390/jcm11030819.
[16] Monazzam RM, Shamsipour R, et al. Evaluation of the Relationship Between Psychological Distress and Sleep Problems with Annoyance Caused by Exposure to Environmental Noise in the Adult Population of Tehran Metropolitan City, Iran[J]. Journal of Environmental Health Science and Engineering, 2022, 20(1).
[17] Wonjin S, Suyeon P, Jihye H, et al. Evaluation of Population Estimation Methods for Wastewater-Based Epidemiology in a Metropolitan City[J]. The Science of the Total Environment, 2022, 857(P1).
[18] Han Jiafu, Zhang Zhong, Qi Qingwen. Analysis and Visualization of Population Spatial Distribution Inequality in China[J]. Journal of Geo-Information Science, 2007(06): 14-19+2.
[19] Lü Chen, Fan Jie, Sun Wei. Spatial Pattern and Influencing Factors of China’s Population Based on ESDA[J]. Economic Geography, 2009, 29(11): 1797-1802.
[20] Yang Qiang, Li Li, Wang Yundong, et al. Spatial Patterns and Evolution Characteristics of China’s Population Distribution from 1935 to 2010[J]. Geographical Research, 2016, 35(08): 1547-1560.
[21] Liu Tao, Peng Rongxi, Zhu Yunxia, et al. Evolution and Influencing Factors of China’s Population Distribution Pattern from 2000 to 2020[J]. Acta Geographica Sinica, 2022, 77(02): 381-394.
[22] Liu Jie, Yang Qingshan, Xu Yiming, Tian Yufei. Spatial Differentiation Patterns of Population in Southeastern and Northern China and Their Influencing Factors[J]. Economic Geography, 2022, 42(01): 12-20.
[23] Liu Ruiwen, Feng Zhiming, You Zhen. Research on the Patterns and Formation Mechanisms of China’s Population Aggregation and Dispersion[J]. China Population, Resources and Environment, 2010, 20(03): 89-94.
[24] Bai Zhongqiang, Wang Juanle, Yang Yaping, et al. Characteristics and Influencing Factors of Population Distribution Across 25 Provinces in China Based on Township-Level Data[J]. Acta Geographica Sinica, 2015, 70(08): 1229-1242.
[25] Yin Xu, Wang Jing, Li Yurui, et al. Spatiotemporal Changes and Influencing Factors of Township-Level Population Distribution in China[J]. Geographical Research, 2022, 41(05): 1245-1261.
[26] Sun Tieshan, Li Guoping, Lu Minghua. Population Agglomeration and Dispersion in the Beijing-Tianjin-Hebei Metropolitan Area and Its Influencing Factors: Empirical Research Based on Regional Density Functions[J]. Acta Geographica Sinica, 2009, 64(08): 956-966.
[27] Yan Dongsheng, Sun Wei, Sun Xiaolu. Evolution and Driving Factors of Population Spatiotemporal Patterns in the Yangtze River Delta[J]. Scientia Geographica Sinica, 2020, 40(08): 1285-1292.
[28] Luo Jun, Shi Peiji, Zhang Xuebin. Characteristics of Population Distribution in the Lanzhou-Xining Urban Agglomeration Based on Township-Level Data[J]. Arid Zone Resources and Environment, 2020, 34(07): 104-111.
[29] Zhao Manru, Yan Dongsheng. Population Distribution Patterns, Spatial Stability, and Driving Factors in the Yangtze River Delta[J]. Journal of Xi’an University of Technology, 2022, 42(02): 204-216. DOI: 10.16185/j.jxatu.edu.cn.2022.02.501.
[30] Liu Haolan, Chen Wanxu, Zeng Jie. Characteristics and Influencing Factors of Population Distribution in the Middle Reaches of the Yangtze River[J]. Resource Development and Market, 2023, 39(04): 444-452+505.
[31] He Yanhua, Gong Zhenjie, Lin Kairong. Simulation of Fine-Scale Population Distribution in the Pearl River Delta Region Based on Geographic Big Data and Multi-Source Information Integration[J]. Scientia Geographica Sinica, 2022, 42(03): 426-435. DOI: 10.13249/j.cnki.sgs.2022.03.007.
[32] Li Shanshan, Lin Wentan. Characteristics of Population Spatial Distribution in the Guangdong-Hong Kong-Macao Greater Bay Area Based on Nighttime Light Remote Sensing[J]. Tropical Geography, 2023, 43(03): 384-394. DOI: 10.13284/j.cnki.rddl.003640.
[33] Yingna Z, Wang Yue, Hu Haoyu, et al. Spatiotemporal Distribution and Mobility Characteristics of Population in the Beijing-Tianjin-Hebei Urban Agglomeration Based on Mobile Signaling Big Data[J]. Territory Research and Development, 2023, 42(03): 161-167+180.
[34] Tang Shuangshuang, Zhou Jing, Deng Yinghui, et al. Spatiotemporal Differentiation and Influencing Factors of Urban-Township-Village Distribution of Migrant Population in Jiangsu Province[J]. Acta Geographica Sinica, 2022, 77(12): 3055-3071.
[35] Zhang Zhibin, Pan Jing, Li Xiaohu. Spatial Evolution of Population Density and Formation Mechanisms in Lanzhou City over the Past 30 Years[J]. Scientia Geographica Sinica, 2013, 33(01): 36-44.
[36] Zhao Kaixu, Zhang Shuaibing, Huang Xiaojun, et al. Evolution of Population Distribution and Influencing Factors in Xi’an City During COVID-19 Pandemic Control: A Perspective Based on Multi-Source Spatiotemporal Big Data[J]. Population and Development, 2022, 28(01): 140-150.
[37] Xu Lanfang, Wu Jilei, Pang Lihua. Spatiotemporal Evolution and Influencing Factors of Migrant Population in Prefecture-Level Cities in China[J]. Population and Development, 2023, 29(04): 89-99.
[38] Yang Huimin, Xu Jiawei, Li Xiaojian. Characteristics and Influencing Factors of Population Distribution in Plain Farming Areas During Urbanization: A Case Study of Zhecheng County in the Eastern Henan Plain[J]. Geographical Research, 2023, 42(06): 1528-1544.
[39] Huang Yanping, Zhai Youlong, Pan Juan, et al. Comparative Analysis of the Coupled Coordinated Development Level of Population, Land, and Urbanization in Chengdu and Chongqing[J]. Journal of China West Normal University (Natural Science Edition), 2018, 39(03): 306-313.
[40] Zhao Zhi. Coupling Analysis of the Population, Economy, and Resource-Environment Systems in the Chengdu-Chongqing Twin-City Economic Circle[J]. Journal of Chongqing University of Technology (Social Science Edition), 2022, 36(01): 108-118.
[41] Zhang Pengfei, Liu Xinzhi. Spatiotemporal Patterns and Evolution of “Industry-Transportation-Population” Coordinated Development: A Case Study of the Chengdu-Chongqing Twin-City Economic Circle[J]. Urban Problems, 2021(09): 60-74.
[42] Sun Jiping, Hou Langong. Network Structure Characteristics of the Chengdu-Chongqing Urban Agglomeration Based on Tencent Migration Data[J]. Modern Urban Research, 2020(09): 78-85.
[43] Pan Bilin, Wang Jianghao, Ge Yong, et al. Spatial Structure and Intercity Population Flow of the Chengdu-Chongqing Urban Agglomeration Based on Weibo Check-In Data[J]. Journal of Geo-Information Science, 2019, 21(01): 68-76.
[44] Deng Chengchao, Quan Nian, Chen Qijun. The Impact of the Chengdu-Chongqing Twin-City Economic Circle Construction on Regional Population Mobility: A Study Based on Panel Data of 16 Prefecture-Level Cities in the Chengdu-Chongqing Region[J]. Journal of Chongqing University of Arts and Sciences (Social Science Edition), 2022, 41(6): 12.
[45] Sun Pingjun, Liu Ju, Luo Ning, Peng Yali. Spatial Patterns and Influencing Factors of Shrinking Cities in the Chengdu-Chongqing Twin-City Economic Circle: An Analysis Based on the Fifth, Sixth, and Seventh National Population Census Data[J]. Journal of Southwest University (Natural Science Edition), 2022, 44(01): 46-56.
[46] Deng Jiaxu, Pan Yuhong. Characteristics and Influencing Factors of Population Shrinkage in the Chengdu-Chongqing Region[J]. Urban and Rural Planning, 2022(6): 10-19.
[47] Chu Wei. Spatial Relationship Between Population Mobility and Economic Growth in the Chengdu-Chongqing Twin-City Economic Circle[J]. China-Arab Technology Forum (Chinese and English Edition), 2023(1): 37-41.
[48] Feng Yue. Spatial Effects of Population and Economic Agglomeration: A Case Study of the Chengdu-Chongqing Twin-City Economic Circle[J]. Journal of Southwest Minzu University (Humanities and Social Sciences Edition), 2022, 43(02): 127-135.
[49] Lu Zuliang, Cai Fei, Huang Fei, et al. Research on the Coordinated Development Relationship Between Population and Economy in the Chengdu-Chongqing Twin-City Economic Circle[J]. Western Forum of Economic Management, 2021, 32(2): 1-10+43.
[50] Li Jia, Zhao Wei, Luo Jialing. Spatiotemporal Evolution of the Coordinated Development of the Population-Economy-Environment System in the Chengdu-Chongqing Twin-City Economic Circle[J]. Journal of Environmental Sciences, 2023, 43(2): 13.
[51] Zhou Yan, Tu Jianjun, Lu Debin, et al. The Relationship Between Population and Economic Spatial Distribution and Its Changes in Chongqing[J]. Economic Geography, 2011, 31(11): 1781-1785.
[52] Wang Hui, Yan Junping, Song Yongyong. Measurement and Spatial Pattern Changes of Economic Pressure Caused by County-Level Population Aging: A Case Study of the Chengdu-Chongqing Region[J]. Geography and Geo-Information Science, 2021, 37(1): 66-73.
[53] Li Cong, Zong Huiming, Xiao Lei. Population Flow Patterns in Typical Population Outflow Areas in China: A Case Study of the Sichuan-Chongqing Region[J]. Tropical Geography, 2021, 41(3): 516-527.
[54] Qin Xuehua, Wu Hua’an. Characteristics and Influencing Factors of Population Spatial Distribution in the Chengdu-Chongqing Twin-City Economic Circle[J]. Yangtze Forum, 2023(2): 8.
[55] Cheng Huan, Xiao Qin. Promoting the Integration of Elderly Care Services in the Chengdu-Chongqing Urban Agglomeration Under the Background of Population Aging[J]. Journal of China West Normal University (Philosophy and Social Sciences Edition), 2020, 39(06): 10-21.