Urbanization, Rural E-Commerce Villages, and Regional Solutions for Urban–Rural Coordinated Development in China

Abstract

With the rapid development of e-commerce, Taobao Villages have emerged as a representative form of rural e-commerce in China, exerting a profound influence on rural economic transformation and urban–rural integration. However, their spatiotemporal distribution is uneven and exhibits a complex interaction with urbanization. Drawing on data from 178 cities between 2017 and 2022, this study employs the Spatial Autoregressive Model (SAR), Spatial Error Model (SEM), and Spatial Durbin Model (SDM) to examine both the direct and spillover effects of urbanization on Taobao Villages. The results indicate that Taobao Villages display significant spatial clustering across China. While urbanization exerts a positive short-term effect on the number of local Taobao Villages, this effect weakens in the long term and under spatial interaction, and higher levels of urbanization in one region impose significant negative spillover effects on neighboring areas. These findings highlight the dual nature of urbanization in simultaneously promoting and constraining rural e-commerce development. Accordingly, policy efforts should focus on rational administrative spatial adjustment and boundary reorganization, optimizing the urban–rural spatial structure, supporting the development of peripheral and disadvantaged regions, enhancing the balanced and efficient flow of factors across regions, and fostering differentiated development pathways to strengthen the resilience of Taobao Villages and promote healthy and sustainable urban–rural integration.

1. Introduction

With the rapid rise in e-commerce, rural China has also witnessed vigorous development in this sector. Around 2011, the first “Taobao Villages” emerged, represented by Dongfeng Village in Suqian’s Shaji Town, Jiangsu Province, and Wantou Village in Boxing, Shandong Province [1]. Taobao Villages constitute a distinctive form of rural e-commerce development in China. Scholars generally define a Taobao Village as a village-level unit where e-commerce industries serve as the backbone and online retailing is the primary mode of transaction, forming a specialized rural e-commerce agglomeration. In essence, this organizational form reflects industrial clustering, platform-based operations, and networked logistics. In 2013, the AliResearch Institute released the first official list of Taobao Villages, and the number expanded rapidly thereafter, reaching 7780 by 2022 [2].

Taobao Villages have profoundly reshaped the rural economic structure, promoting income growth and employment for farmers while fostering urban–rural integration and rural revitalization. Through e-commerce platforms, they have stimulated the formation of new industrial clusters, driven the transformation and upgrading of traditional industries, expanded the market space for family farms, and strengthened both the branding and commercialization of agricultural products. At the same time, Taobao Villages have inspired entrepreneurial enthusiasm among rural youth and returning migrants. By combining online transactions with offline services such as warehousing, packaging, and logistics, they have created substantial employment opportunities, alleviated the outflow of surplus rural labor, and enabled farmers to earn higher incomes than in traditional agriculture. In terms of urban–rural coordinated development, Taobao Villages, leveraging e-commerce platforms and logistics networks, have facilitated the two-way flow of information, capital, talent, and goods between cities and rural areas, thereby accelerating the process of urban–rural integration.

However, alongside the rapid expansion of Taobao Villages, a number of pressing issues have also emerged. Their regional distribution is highly uneven, with a high density in the eastern coastal areas, while development in the central, western, and less developed regions has lagged significantly behind [3]. Moreover, the quality of development varies markedly across villages. Some Taobao Villages are excessively dependent on a single platform or industrial path, making them prone to “herd effects” and low-price competition, and thus lacking the momentum for sustainable growth [4]. Similar developmental dilemmas can also be observed in other countries. Based on a longitudinal analysis of e-commerce development across European regions from 2010 to 2019,it was found that the level of e-commerce penetration exhibited substantial regional disparities [5]. Northern and Western European countries showed relatively high penetration rates, whereas Southern and Eastern European countries lagged behind, and such differences were closely associated with income levels, broadband coverage, and the degree of urbanization. Their study further indicated that e-commerce development tends to concentrate initially in large cities and metropolitan areas, while rural and peripheral regions remain significantly behind, suggesting that the digital divide has not narrowed despite the overall increase in e-commerce adoption [5]. In recent years, facing intensified competition among e-commerce platforms, rural labor outmigration, and shifts in consumer preferences, the development of Taobao Villages has entered a bottleneck period [6]. The growth rate has slowed, and the phenomenon of disappearing Taobao Villages has become evident: as early as 2015, cases of disappearance were reported in Guangdong Province, and since 2016, the number of disappearing Taobao Villages has continued to grow with an expanding geographic scope. By 2020, this phenomenon reached its peak, with withdrawals accelerating in the eastern regions and similar trends beginning to emerge in the central, western, and northeastern regions.

The process of urbanization demonstrates a complex interaction with the development of Taobao Villages [7]. On the one hand, the advancement of urbanization provides solid infrastructure support and broad market opportunities, facilitates the flow of population, resources, and information between rural and urban areas, and effectively enhances the development level of rural e-commerce. On the other hand, while urbanization generates positive effects, it also imposes constraints on the growth of Taobao Villages through resource siphoning and market crowding-out effects. The Chinese government attaches great importance to advancing urbanization and improving the quality of urban–rural integration. On 31 July 2024, the State Council issued the Five-Year Action Plan for the In-Depth Implementation of the People-Centered New Urbanization Strategy, which emphasizes measures such as deepening the reform of the household registration system, enhancing the comprehensive carrying capacity of cities, and fostering modern metropolitan areas, with the aim of adjusting the complex interactions of coordinated urban–rural development from a policy perspective. Against this backdrop, exploring the impact of urbanization on the spatial patterns of Taobao Villages carries important significance.

Amidst the rapid development of urbanization in China, the spatiotemporal evolution of professional villages like Taobao Villages and their interaction with urbanization have attracted academic attention. Regarding the relationship between Taobao Villages and urbanization, scholars believe that the rapid development of rural economies and the modernization of social structures have been facilitated by the impetus of e-commerce platforms [8]. In European villages, village-level e-commerce platforms can reduce rural communities’ dependence on large multinational e-commerce platforms and enhance the autonomy and sustainability of local economies. Compared with cross-border platforms, localized platforms are better embedded in regional industrial chains, directly linking farmers with consumers, thereby preventing rural areas from being marginalized in the digital economy (Rundel et al., 2024) [9].

Luo Zhendong and He Heming [10] pointed out that rural urbanization driven by e-commerce represents a new bottom-up development process. Through the systematic reconstruction of the social, economic, and physical environment of rural areas, it has facilitated the non-agricultural transformation of employment, the modernization of rural lifestyles, and the intensive urbanization of spatial structures. Moreover, the development of Taobao Villages has broken through the constraints of traditional location factors, enabling participation in national and even global industrial divisions of labor, and thereby realizing a leapfrog transformation in the scale and functions of rural communities. Xu Chan et al. (2015) explored the impact of e-commerce on the urbanization pattern in China, arguing that in the internet era [11], the popularization of e-commerce has made opportunities for talent and economic entities in cities at all levels more equalized, bringing unprecedented opportunities to county-level areas and rural regions. The development of Taobao Villages not only challenges the centrality of large cities but also promotes a new situation of urbanization driven by informatization rather than industrialization. Fang Guanxin (2016), from the perspective of urban sociology, explained the emergence, evolution, characteristics, and significance of “Taobao Villages,” suggesting that Taobao Villages have effectively integrated traditional and modern factors in rural areas through the internet economy and technological innovation, achieving a “one-stop” modernization transformation of “agriculture, rural areas, and farmers [12].” Other scholars have also studied the role of digitization in promoting in situ urbanization in rural areas of China, analyzing the impact of e-commerce on local industrial development, social transformation, governance structure, and spatial layout [13,14].

From the perspective of spatial evolution, existing studies on the development and influencing factors of Taobao Villages have mainly focused on their spatial distribution patterns, directions of agglomeration, clustering structures, density changes, and regional linkages [15,16,17,18]. However, in reality, urbanization, as a core driving force of regional economic and social transformation, not only promotes the emergence and agglomeration of Taobao Villages through mechanisms such as infrastructure improvement, population mobility, and market expansion, but may also generate constraining effects through factor polarization, market crowding-out, and policy resource bias, thereby profoundly reshaping their spatial patterns.

The spatiotemporal evolution of urbanization and professional villages is a complex process involving multiple factors, including economic, social, and environmental aspects [12,19,20]. Existing research has yielded a relatively rich array of scholarly outcomes, yet certain deficiencies persist. Firstly, there is a limitation in the theoretical frameworks and perspectives. Current studies tend to rely on singular theoretical frameworks, such as economics or geography, lacking an interdisciplinary and comprehensive viewpoint. This singular perspective may overlook the multidimensional sociocultural factors, environmental impacts, and policy dynamics in the development of professional villages. Secondly, in terms of research methodology, there is a predominance of qualitative research with a deficiency in quantitative studies. Thirdly, there is insufficient consideration of dynamic changes and sustainable development. The development of professional villages is a dynamic process influenced by various factors; existing research often focuses on the state at a particular point in time, neglecting the trends and patterns that evolve. Additionally, current research on Taobao villages is primarily conducted within the framework of the urban-rural divide. Due to differences in economic development levels, resource endowment conditions, and characteristics of population migration and mobility across regions, the traditional assumption of spatial homogeneity cannot effectively explain the spatial relationship between urbanization and Taobao villages [21,22,23].

Most existing studies remain at the descriptive level of spatial patterns, lacking systematic exploration of how urbanization influences the spatial configuration of Taobao Villages, what spillover effects urbanization generates on Taobao Villages in surrounding regions, and how Taobao Villages interact with the urbanization process within the broader context of urban–rural coordinated development and its deeper implications. In addition, current research on Taobao Villages has largely been conducted within the framework of the urban–rural divide [24]. Given the differences in regional economic development levels, resource endowments, and population migration and mobility characteristics, the traditional assumption of spatial homogeneity cannot effectively explain the spatial relationship between urbanization and Taobao Villages [25].

Building on theoretical analysis, this study develops an analytical framework to examine the impact of urbanization on the spatial patterns of Taobao Villages. Using ArcGIS 10.5, we constructed a spatial dataset of Taobao Villages within Chinese city jurisdictions and employed panel data from 178 cities between 2017 and 2022. Spatial econometric models, including the Spatial Autoregressive Model (SAR), Spatial Error Model (SEM), and Spatial Durbin Model (SDM), were applied to empirically test the heterogeneous impacts of urbanization and its spatial spillover effects on Taobao Villages. The aim is to enrich the academic understanding of how urbanization influences the spatial dynamics of Taobao Villages and their underlying mechanisms, while also providing scientific evidence and practical references for promoting urban–rural integration, improving rural e-commerce policies, and refining the new urbanization strategy.

The main contributions of this study are as follows: (1) From the combined perspective of spatial evolution and urbanization, it investigates the impact of urbanization on the spatial patterns of Taobao Villages, addressing the limitations of existing studies that largely remain at the descriptive level and lack exploration of the underlying mechanisms. (2) By introducing urbanization as a key analytical variable into the study of rural e-commerce, this research extends the existing frameworks that have predominantly focused on industry, market, or individual levels, offering a new theoretical perspective for understanding the spatial distribution and evolution of rural e-commerce. (3) By employing spatial econometric models, the study identifies both the positive promotion and negative spillover effects of urbanization, thereby providing new empirical evidence for understanding the complex relationship between urbanization and rural e-commerce development.

2. Theoretical Analysis and Research Hypothesis

Regional economics has produced a rich body of theories on spatial agglomeration and regional disparities. As early as von Thünen’s location theory and A. Weber’s theory of industrial location, it was emphasized that factor endowments and transportation costs shape the geographical distribution of economic activities. Krugman (1991), within the framework of New Economic Geography, further argued that adjacent regions, by sharing transportation networks, labor markets, and industrial linkages, are more likely to exhibit similar development trajectories [26,27,28].

In the development of Taobao Villages, the joint influence of factor endowments and market accessibility has created similar conditions of transportation costs, delivery efficiency, and market reach in neighboring regions through transport, logistics, and urban cluster networks. This has fostered comparable conditions for e-commerce entry and growth, leading to the clustering of high-value regions and the contiguous distribution of low-value regions. Moreover, agglomeration economies and supply chain sharing reinforce such spatial linkages. The division of labor and supporting elements associated with e-commerce—such as raw materials and packaging, photography and design, warehousing and logistics, live-streaming services, financial payment systems, and training or intermediary services—tend to concentrate regionally and radiate outward, enabling neighboring areas to benefit from externalities and form localized “high–high” development clusters.

At the same time, geographic proximity reduces the costs of information diffusion and imitation, allowing product selection models, multi-store operations, live-streaming scripts, and platform management experience to spread quickly across adjacent areas, further strengthening positive spatial correlations. Labor and social network externalities also play a role: inter-village and inter-town labor markets, acquaintance networks, and industrial chains often transcend administrative boundaries, creating mutually supportive employment and collaboration relationships that promote the coordinated development of Taobao Villages in neighboring regions. Finally, early-developing areas often enjoy advantages in branding, credibility, and platform resources, while government policies and industrial support are typically implemented in a regional cluster pattern. Such contiguous driving effects further consolidate and amplify existing spatial agglomeration.

Based on the above analysis, we propose the following hypothesis:

Hypothesis 1.

Taobao Villages exhibit significant positive spatial correlation.

Jacobs (1969) argued that the economic growth of cities depends not only on the specialized clustering of industries but also on the coexistence of diverse industries [29,30]. Diversified economic activities facilitate cross-boundary knowledge exchange and the emergence of new industries, and such diversification externalities constitute an important source of sustained innovation and long-term growth for cities [31]. In other words, the process of urbanization does not merely imply the spatial concentration of population [32]. More importantly, it systematically alters local production and exchange conditions by expanding market capacity, improving infrastructure, strengthening the supply of labor and services, and enhancing institutional support [33].

With respect to the formation of Taobao Villages, the role of urbanization can be understood from several dimensions [34]. First, urbanization brings an increase in permanent population and consumption capacity, thereby significantly enhancing market potential. As consumer groups concentrate and consumption structures upgrade, local e-commerce can more easily form a stable demand base. In particular, the rising proportion of intra-city and nearby orders helps shorten delivery distances and reduce demand uncertainty, making it easier for e-commerce villages to achieve efficient operating scales. Second, urbanization is often accompanied by improvements in transportation, logistics, and digital networks. The development of roads, warehousing and distribution nodes, express delivery outlets, broadband, mobile communication, and third-party payment systems significantly lowers entry and transaction costs, improves delivery efficiency and reliability, and thus provides essential support for the operation of e-commerce villages. Third, urbanization enriches labor and service markets, not only attracting more human resources but also cultivating diversified professional services such as operations, customer service, live-streaming, design, outsourcing, and finance. These factors provide rural e-commerce with nearly “ready-made” supporting services, allowing individual villages to embed efficiently into the existing e-commerce ecosystem without bearing the entire cost of building a complete supply chain. Finally, as urbanization progresses, local governments’ governance capacity and institutional support are also enhanced. Measures such as quality inspection, brand and credit systems, SME financing, industrial parks, and live-streaming bases are continuously improved, reducing uncertainties in entrepreneurship and transactions and further creating a favorable institutional environment for the growth of e-commerce villages.

Based on the above analysis, we propose the following hypothesis:

Hypothesis 2.

A higher urbanization rate increases the number of local Taobao Villages.

The theory of unbalanced growth posits that economic development does not occur simultaneously across all regions, but rather first concentrates in certain “growth poles” or “growth centers” (Hirschman, 1958; Perroux, 1955) [35,36]. Growth centers may exert dual effects on surrounding areas: on the one hand, economic prosperity can generate spread effects through demand spillovers and technology diffusion; on the other hand, excessive concentration of factors and demand may result in a “polarization effect,” crowding out the development space of neighboring regions (Myrdal, 1957) [37]. When industries exhibit significant economies of scale and network externalities, polarization and backwash effects often dominate, thereby constraining the growth opportunities of adjacent areas. New Economic Geography provides a further explanation of this process. Under conditions of increasing returns to scale and declining trade costs, differences in market accessibility tend to push economic activities toward the “core,” resulting in a “core–periphery” structure (Krugman, 1991; Fujita & Krugman, 2004; Fujita, Krugman & Venables, 1999) [28,38,39]. For rural e-commerce, such “core formation” is reflected not only in the concentration of population and enterprises but also in the agglomeration of logistics nodes, platform traffic, and service elements. As a result, the marginal attractiveness of core areas continues to strengthen, while the entry and expansion conditions for neighboring regions become relatively unfavorable.

The inhibitory effects of urbanization on the development of Taobao Villages in neighboring regions are manifested in several ways. First, as local urbanization levels rise, consumer density and platform visibility increase substantially, leading to higher search rankings, greater exposure in live-streaming and short videos, and stronger customer conversion capacity. This more aggressively captures market demand, diverting potential customers away from neighboring areas. Second, capital, operational, and live-streaming talent, as well as third-party service providers, tend to cluster in highly urbanized areas with dense orders, causing surrounding regions to face service scarcity and rising costs, thereby raising entry barriers and operational expenses. Third, critical nodes such as express distribution centers, intra-city warehouses, and transportation trunk lines are often concentrated in core areas, creating economies of scale and timeliness advantages. Neighboring regions that lack similar conditions face fulfillment uncertainties and delivery disadvantages, which undermine their competitiveness. Finally, institutional resources such as industrial parks, live-streaming bases, brand and quality inspection systems, and financing support also tend to concentrate in core areas. This provides stronger credit endorsement and brand externalities for core regions, while weakening the institutional environment of neighboring areas and increasing their development uncertainty.

Based on the above analysis, we propose the following hypothesis:

Hypothesis 3.

A higher level of urbanization in one region suppresses the formation and development of Taobao Villages in neighboring regions.

3. Materials and Methods

3.1. Data Sources

Data on Taobao villages were obtained from the Ali Research Institute (http://www.aliresearch.com/cn/index accessed on 15 July 2024). Based on the public lists of Taobao villages from 2017 to 2022, effective vector data of cities where Taobao villages are located during the period from 2017 to 2022 were collected. Utilizing ArcGIS, a spatial dataset of Taobao villages within the jurisdiction of Chinese cities was constructed. This study designates the period from 2017 to 2022 as the observation window, based on the following considerations. First, during this period, the definitions and coverage of Taobao Village lists and related socioeconomic indicators remained relatively stable, which facilitates inter-annual comparability and spatial matching. Second, the sample spans the pre-pandemic, pandemic shock, and recovery phases, thereby providing a more comprehensive reflection of the staged characteristics of Taobao Village development and ensuring representativeness. Finally, 2022 is the most recent year with fully available public data, which enables the construction of a balanced panel and ensures the comparability of estimation results. Socio-economic data is primarily derived from the “China City Statistical Yearbook,” “China County Statistical Yearbook,” various provincial statistical yearbooks, local government work reports, and public reports on national economic and social development from 2017 to 2022. For missing data in certain years, interpolation calculations were applied to fill the gaps. The map of China’s administrative divisions was sourced from the National Geomatics Center (http://www.ngcc.cn accessed on 19 July 2024).

3.2. Variable Selection and Descriptive Statistics

• Dependent Variable

The dependent variable in this study is the number of Taobao villages. A Taobao village is defined as a natural or administrative village where there are more than 30 active online stores with an annual transaction volume of over CNY 10 million. Active online stores refer to those that have actual transaction records in the past year. Data on the number of Taobao villages come from the annual lists of Taobao villages published by the Ali Research Institute. This study uses data on the number of Taobao villages from 2017 to 2022, covering 178 cities across the country.

• Core Explanatory Variable

The core explanatory variable in this study is the level of urbanization (urban). Urbanization reflects the process of rural population migration and aggregation in urban areas. Following the approach of most scholars, the urbanization rate of permanent residents is used as a proxy variable, calculated by the formula:

$$urba{n}_i=P_{ui}/P_i$$

(1)

In the formula, $urba{n}_i$ represents the level of urbanization development in city i, $P_{ui}$ denotes the permanent urban population of the city i, and $P_i$ signifies the total permanent population of city i.

• Control Variables

To avoid bias in the estimation results due to the omission of important factors, this study, following prior research [24,25,40,41,42,43], incorporates five control variables covering three dimensions. The first dimension concerns supply conditions, including the value added of the primary industry. In the supply side of rural e-commerce, agriculture, forestry, animal husbandry, and fishery constitute the main sources of e-commerce products. The economic scale and value-added level of the primary industry not only affect the categories and quantities of e-commerce products but also reflect the degree of agricultural industrialization and commercialization, thereby influencing the conditions for the formation of Taobao Villages (Zhu et al., 2016) [44]. This dimension also includes employment in agriculture, forestry, animal husbandry, and fisheries. The development of e-commerce villages requires a large labor force to participate in production, processing, packaging, warehousing, and distribution. The scale of agricultural employment reflects not only the abundance of labor resources but also the sustainability of rural e-commerce in terms of organization and operation. The second dimension concerns development level, for which per capita GDP is selected as an indicator to represent the overall level of regional economic development and consumption capacity. A higher level of economic development typically indicates a better business environment, stronger purchasing power, and a more complete factor market, all of which may promote the growth and expansion of e-commerce villages. The third dimension concerns circulation and information conditions, primarily including total freight volume and total postal service volume. Total freight volume refers to the total weight of goods transported by various means over a given period, reflecting the regional conditions of transportation and logistics capacity. Total postal service volume refers to the total amount of business handled by the postal system within a given period, such as letters and parcels. Both freight and postal volumes reflect the activity level of express delivery and information circulation. To some extent, they determine the efficiency and reliability of e-commerce fulfillment and delivery, thereby exerting significant influence on the spatial agglomeration and diffusion of Taobao Villages. To ensure consistency of scale, all control variables are transformed into natural logarithms.

The types and definitions of variables are shown in

Table 1. Variable types and definitions.

Variable Type	Variable Name	Definition
Dependent Variable	Number of Taobao Villages	The number of villages within a natural or administrative village boundary that have ≥30 active online shops and annual transaction volume ≥ RMB 10 million.
Independent Variable	Urbanization Rate	The proportion of the permanent population residing in urban areas of the total permanent population.
Control Variables	Added Value of Primary Industry	The added value created during a certain period by agriculture, forestry, animal husbandry, fishery, and related services.
	Employment in Agriculture, Forestry, Animal Husbandry, and Fishery	The number of laborers engaged in agricultural, forestry, animal husbandry, and fishery activities during a certain period.
	Per Capita GDP	The ratio of regional gross domestic product to the permanent resident population during a certain period.
	Total Freight Volume	The total weight of goods transported through various transportation modes during a certain period.
	Total Postal Service Volume	The total amount of postal business handled by the postal system during a certain period, including letters, parcels, etc.

.

The relevant data mainly come from the “China Statistical Yearbook,” “China City Statistical Yearbook,” “China Regional Economic Statistical Yearbook,” and the statistical yearbooks and bulletins of cities where Taobao villages are located over the years. The sample consists of 1068 panel data from 178 prefecture-level and above cities where Taobao villages are located from 2017 to 2022. Linear interpolation was used to fill in missing values for some areas or years. Descriptive statistical characteristics of the above variables are shown in

Table 2. Descriptive statistics.

Variable Name	Variable Symbol	Mean	Std.Dev.	Min	Max
Number of Taobao Villages	TBC	1.757	1.684	0	6.295
Urbanization Rate	UR	0.631	0.128	0.344	1.178
Added Value of Primary Industry	VA1	14.571	0.725	12.378	16.818
Employment in Agriculture, Forestry, Animal Husbandry, and Fishery	AFF	0.39	0.934	0.001	11.98
Per Capita GDP	PGDP	11.079	0.483	9.367	12.198
Total Freight Volume	CT	9.657	0.892	7.014	12.844
Total Postal Service Volume	PTV	12.056	1.247	9.026	16.795

. The results show that from 2017 to 2022, the urbanization rate of cities where Taobao villages are located was 63.1%, with an average urbanization rate of 0.631 and a small standard deviation, indicating that there is not much difference in the urbanization rate among the regions in the sample, but there is still a certain range of variation. The average number of Taobao villages in the sample was 1.757, with a large standard deviation, indicating a significant difference in the number of Taobao villages between different regions.

3.3. Research Methods

3.3.1. Kernel Density Analysis

Kernel density analysis is a spatial algorithm based on the clustering of data density functions. It can reflect the characteristics of the aggregation or dispersion of observed elements in space. The formula is as follows:

$$f\left(x\right)={\displaystyle \frac{1}{n{h}^d}}{\displaystyle \sum _{i=1}^{n}K}\left({\displaystyle \frac{x-x_i}{h}}\right)$$

(2)

3.3.2. Spatial Autocorrelation Analysis

Spatial autocorrelation is a method of spatial analysis that reflects whether there is a correlation between spatial variables and their neighboring variables. It is divided into Global Moran’s I for global spatial autocorrelation and Local Moran’s I for local spatial autocorrelation.

• Global Spatial Autocorrelation

Using municipal administrative units as spatial units, global spatial autocorrelation analysis is employed to examine the spatial correlation of Taobao villages across the country. The formula is as follows:

$$\mathrm{Global} \mathrm{Moran}’\mathrm{s} \mathrm{I}={\displaystyle \frac{n{\displaystyle \sum _{i=1}^n{\displaystyle \sum _{j=1}^n{W}_{ij}\left(X_i-\overline{X}\right)\left(X_j-\overline{X}\right)}}}{{\displaystyle \sum _{i=1}^n{\displaystyle \sum _{j=1}^n{W}_{ij}{\displaystyle \sum _{i=1}^n\left(X_i-\overline{X}\right)}}}}}$$

(3)

In the formula, n is the number of municipal administrative units, $X_i$ and $X_j$ are the numbers of Taobao villages in cities i and j, respectively; $\overline{X}$ is the mean value; $W_{ij}$ is the spatial weight. The value of the global spatial autocorrelation coefficient ranges from [−1, 1]. When Global Moran’s I > 0, it indicates that elements are aggregated spatially; when Global Moran’s I < 0, they are dispersed; and when Global Moran’s I = 0, it indicates a random distribution of elements.

• Local Spatial Autocorrelation

Different from global spatial autocorrelation analysis, local spatial autocorrelation focuses on analyzing the relationship between each observation point or area in the spatial dataset and its surrounding neighboring areas, revealing local spatial correlation and heterogeneity. The formula for local spatial autocorrelation is as follows:

$$\mathrm{Local} \mathrm{Moran}’\mathrm{s} \mathrm{I}={\displaystyle \frac{\left(X_i-\overline{X}\right)}{S_i^2}}{\displaystyle \sum _{i=1,j\ne i}^n\left(X_j-\overline{X}\right)}$$

(4)

In the formula, Var($X_i$) is the variance of the number of Taobao villages in city ii. Local Moran’s I can reflect local spatial characteristics such as heterogeneity and instability in the distribution of spatial variables.

3.3.3. Spatial Econometrics

This paper utilizes spatial econometrics to further explore the impact of the level of urbanization on the spatial distribution of Taobao villages. Spatial econometrics is an important method that integrates spatial information into economic analysis, revealing the interrelationships and mechanisms of influence of economic variables in geographical space by considering spatial correlation and heterogeneity [45,46]. Commonly used models in spatial econometrics include the Spatial Lag Model (SAR), the Spatial Error Model (SEM), and the Spatial Durbin Model (SDM). The Spatial Lag Model (SAR) captures the mutual influence between spatially adjacent areas by introducing the spatial lag term of the dependent variable into the regression equation. The Spatial Error Model captures the mutual influence between spatially adjacent areas by introducing a spatial autoregressive structure into the error term. The Spatial Durbin Model combines the features of SAR and SEM, introducing not only the spatial lag term of the dependent variable but also the spatial lag term of the independent variables, including direct effects (the impact of local variables on the dependent variable), indirect effects (the influence of neighboring areas’ variables on the local dependent variable through spatial transmission mechanisms), and total effects (combining direct and indirect effects), revealing more complex spatial interaction relationships.

4. Evolutionary Characteristics of the Spatial Pattern of Taobao Villages

4.1. General Geographic Pattern of Taobao Villages

Observing the geographical distribution of Taobao villages from 2017 to 2022 (Figure 1), it is evident that the overall geographic distribution of Taobao villages exhibits a pattern of being denser in the east and sparser in the west, with uneven distribution between the north and south. The eastern coastal regions, particularly Zhejiang, Jiangsu, Shandong, and Guangdong provinces, have become the core areas of high-density agglomeration for Taobao villages. Although the number of Taobao villages in central and western provinces such as Henan, Hubei, and Sichuan has been increasing year by year, there is still a significant gap compared to the eastern coastal regions. Northern provinces like Hebei, Inner Mongolia, and Heilongjiang have relatively fewer Taobao villages, which are more dispersed in distribution.

4.2. Spatiotemporal Agglomeration Density and Evolution of Taobao Villages

The Kernel Density Plot is a spatial analysis tool that uses kernel functions to smooth data points, generating a continuous density distribution map. It can intuitively display the degree and trend of aggregation of data points in space, eliminating discreteness and noise, and providing a smoother and more readable distribution image, suitable for the visualization and analysis of spatial data.

From the Kernel Density Plot of Taobao villages from 2017 to 2022 (Figure 2), it can be seen that the geographical distribution of Taobao villages shows obvious evolutionary characteristics. Firstly, a high-density agglomeration effect has been formed in the eastern coastal regions. The density of Taobao villages in Zhejiang, Jiangsu, Shandong, and Guangdong provinces has been increasing year by year, forming a stable area of high-density agglomeration. Secondly, the expansion of Taobao villages in central provinces continues. The number of Taobao villages in central provinces such as Henan, Hubei, and Sichuan has been increasing year by year, showing the potential for e-commerce development and the trend of Taobao villages expanding inland. Thirdly, regional differences are gradually emerging. Although the number of Taobao villages in central and western regions and northern regions such as Fujian, Anhui, and Hebei has increased, the overall density is still lower than that of the eastern coastal regions, indicating an unbalanced development of Taobao villages across the country.

4.3. Spatial Autocorrelation Characteristics of Taobao Villages

4.3.1. Global Moran’s I

This paper selects the adjacency matrix to test the spatial correlation of the number of Taobao villages in 178 cities from 2017 to 2022. The I values of the number of Taobao villages for each year are all greater than 0 and pass the significance test of p < 0.001 (

Table 3. Global Moran’s I Index.

Year	Moran’s I Index	z	p
2017	0.324	5.885	<0.001
2018	0.328	5.960	<0.001
2019	0.35	6.340	<0.001
2020	0.364	6.751	<0.001
2021	0.355	6.430	<0.001
2022	0.369	6.670	<0.001

), indicating that the distribution of Taobao villages across the country has obvious spatial dependence, and there is a strong positive correlation between the number of Taobao villages in neighboring areas, showing significant spatial agglomeration characteristics, and Hypothesis 1 is thus supported. In particular, Moran’s I value in 2020 is 0.364, which is the highest among all years, indicating that the degree of spatial agglomeration of Taobao villages is the most significant in that year.

4.3.2. Local Moran’s I Analysis

Looking at the LISA cluster map from 2017 to 2022 (Figure 3), the spatial distribution of Taobao villages shows significant spatial heterogeneity. The eastern coastal areas, especially Zhejiang, Jiangsu, Shandong, and Guangdong provinces, have a larger number of Taobao villages and exhibit a high-high cluster characteristic. In provinces such as Fujian, Anhui, and Hebei, the number of Taobao villages is smaller, showing a low-low cluster characteristic, failing to form an agglomeration effect for the development of Taobao villages.

5. Empirical Analysis of the Impact of Urbanization on the Spatial Distribution of Taobao Villages

5.1. Model Construction

To analyze the impact of the level of urbanization on the spatiotemporal evolution of Taobao villages, this study constructs three spatial econometric models: the Spatial Lag Model (SAR), the Spatial Error Model (SEM), and the Spatial Durbin Model (SDM).

• SAR

$$Y_{it}=\rho W{Y}_{jt}+{\beta }_0+{\beta }_{1}U{R}_{it}+{\beta }_{2}VA{1}_{it}+{\beta }_{3}AF{F}_{it}+{\beta }_{4}PGD{P}_{it}+{\beta }_{5}C{T}_{it}+{\beta }_{6}PT{V}_{it}+{\alpha }_i+{\gamma }_t+{\epsilon }_{it}$$

(5)

• SEM

$$\begin{array}{l}{Y}_{it}={\beta }_0+{\beta }_{1}U{R}_{it}+{\beta }_{2}VA{1}_{it}+{\beta }_{3}AF{F}_{it}+{\beta }_{4}PGD{P}_{it}+{\beta }_{5}C{T}_{it}+{\beta }_{6}PT{V}_{it}+{\alpha }_i+{\gamma }_t+{\mu }_{it}\\ {\mu }_{it}=\lambda W{\mu }_{it}+{\epsilon }_{it}\end{array}$$

(6)

• SDM

$$Y_{it}=\rho W{Y}_{\mathrm{jt}}+{\beta }_0+{\beta }_{1}U{R}_{it}+{\beta }_{2}VA{1}_{it}+{\beta }_{3}AF{F}_{it}+{\beta }_{4}PGD{P}_{it}+{\beta }_{5}C{T}_{it}+{\beta }_{6}PT{V}_{it}+{\displaystyle {\sum }_{\mathrm{k}}{\theta }_k}W{X}_{kit}+{\alpha }_i+{\gamma }_t+{\epsilon }_{it}$$

(7)

In the aforementioned models, $Y_{it}$ represents the number of Taobao villages in region i during period t, $\rho$ is the spatial lag coefficient, $\rho W{Y}_{\mathrm{jt}}$ is the spatial lag term of the number of Taobao villages, W is the spatial weight matrix, ${\beta }_0$ is the constant term, ${\beta }_1$, ${\beta }_2$, …, ${\beta }_6$ represents the regression coefficients for the explanatory variables, ${\alpha }_i$ denotes the individual fixed effects, ${\gamma }_t$ signifies the time fixed effects, ${\epsilon }_{it}$ is the random error term, ${\theta }_k$ stands for the spatial lag coefficients of the explanatory variables, ${\mu }_{it}$ represents the error term, and $W{\mu }_{it}$ is the spatial lag term of the error.

5.2. Model Testing and Selection

To determine the presence of spatial autocorrelation in the model, this paper conducted the Lagrange Multiplier (LM) test. The results indicate significant spatial lag and spatial error effects. Subsequently, the Likelihood Ratio Test (LR Test) was used to compare the merits of different models [45,46]. When comparing the time effect model with the bidirectional effect model, the LR test showed a statistic of 92.65 with a p-value less than 0.01, indicating that the bidirectional effect model is superior. When comparing the individual effect model with the bidirectional effect model, the LR statistic was 2266.01 with a p-value less than 0.01, indicating that the bidirectional effect model is again superior. To determine which model, fixed effect or random effect, is more suitable for the data analysis in this paper, the Hausman test was conducted. The results showed significant differences in the coefficients of the urbanization rate (UR) and the added value of the primary industry (VA1) between the fixed effect and random effect models, with values of −1.218 and −1.104, respectively. According to the Hausman test results, the fixed effect model is more appropriate for explaining the impact of variables. In the LR test results, the p-values were significantly less than 0.01, rejecting the null hypothesis that the SDM cannot be simplified to SAR and SEM. Through the above model testing and selection process, this paper ultimately chose the Spatial Durbin Model (SDM) for empirical analysis to more comprehensively reveal the impact of urbanization and other control variables on the spatial distribution of Taobao villages (

Table 4. Test results of spatial econometric model.

Text	Statistics	p
Hausman Test	39.11	<0.01
Spatial Lag LM	344.712	<0.01
Spatial Error LM	107.081	<0.01
Spatial Lag Robust LM	240.23	<0.01
Spatial Error Robust LM	2.599	0.107
Bidirectional and Individual Effect LR Test	92.65	<0.01
Time and Bidirectional Effect LR Test	2266.01	<0.01
Spatial Lag Effect Wald Test	39.11	<0.01
Spatial Error Effect Wald Test	38.06	<0.01
Likelihood Ratio Test SDM vs. SAR	39.41	<0.01
Likelihood Ratio Test SDM vs. SEM	39.45	<0.01

).

5.3. Empirical Results

The model was tested for multicollinearity (

Table 5. Variance inflation factor.

Variable	VIF	1/VIF
UR	3.518	0.284
PGDP	2.524	0.396
PTV	2.064	0.484
VA1	1.869	0.535
CT	1.309	0.764
AFF	1.036	0.965
Mean-VIF	2.053	0.965

All variable abbreviations are defined in Table 2.

), and the results showed that all VIF values were significantly less than 10, indicating that there is no problem with multicollinearity among the variables.

5.3.1. Baseline Regression Analysis

As shown in

Table 6. Estimation results of the spatiotemporal impact of urbanization development on Taobao Villages.

Variable	(1)	(2)	(3)	Variable	(4)	(5)	(6)
	Ind	Time	Both		Ind	Time	Both
ur	1.332 ***	0.549	0.332	W×ur	−0.434	−3.671 ***	−2.062 **
	(3.62)	(1.05)	(0.90)		(−0.65)	(−4.71)	(−3.09)
lnva1	0.00234	0.107	−0.731 ***	W×lnva1	1.140 ***	−0.685 ***	0.530
	(0.01)	(1.61)	(−3.35)		(4.04)	(−6.80)	(1.84)
lnaff	0.0396	−0.202 ***	0.0793 **	W×aff	−0.0564	−0.0206	−0.0599
	(1.45)	(−5.54)	(2.96)		(−0.63)	(−0.23)	(−0.69)
lnpgdp	−0.00422	−0.236 *	−0.172	W×lnpgdp	−0.0501	0.620 ***	−0.304
	(−0.03)	(−2.09)	(−1.21)		(−0.24)	(4.61)	(−1.48)
lnct	0.0117	−0.206 ***	0.0109	W×lnct	0.621 *	−0.202 *	0.673 **
	(0.09)	(−4.83)	(0.08)		(2.32)	(−2.54)	(2.61)
lnptv	0.116 **	0.627 ***	−0.00784	W×lnptv	0.500 ***	0.660 ***	0.255 ***
	(3.16)	(14.71)	(−0.21)		(7.81)	(8.22)	(3.73)

The values in parentheses are t-statistics. ***, **, * denote significance at the 1%, 5%, and 10% levels, respectively. All variable abbreviations are defined in Table 2.

, Column (1) presents the baseline regression model, which only considers contemporaneous variables without including spatial or temporal lag terms. This reflects the impact of the current urbanization rate on the number of local Taobao Villages. The impact coefficient of the urbanization rate is 1.332 and has passed the significance test, indicating that the urbanization rate has a positive impact on the number of Taobao villages in the short term, thereby confirming Hypothesis H2. Under the spatial effect of the adjacency matrix, the time effect captures the dynamic changes in the impact of the urbanization rate on the development of Taobao villages in different years’ fixed effects. In column (5), the urbanization impact coefficient is −3.671 and has passed the significance test, indicating that an increase in the urbanization rate in neighboring areas will have a negative impact on the number of Taobao villages in this area. In the dual fixed-effects model, the impact of the urbanization rate on the number of Taobao villages is significantly negative (−2.062), indicating that even when considering both time and space effects, an increase in the urbanization rate in neighboring areas still has a negative impact on the number of Taobao villages in this area (Table 6).

5.3.2. Analysis of Direct Effects and Spatial Spillover Effects

This paper utilizes the “partial derivative method” to further decompose the spatial spillover effects of the explanatory variables into direct and indirect effects (

Table 7. Results of direct effect and spatial spillover effect tests.

	(1)	(2)	(3)	(4)	(5)
Variables	Main	Wx	Direct	Indirect	Total
UR	0.332	−2.062 ***	0.333	−1.976 ***	−1.643 **
	(0.367)	(0.668)	(0.375)	(0.657)	(0.735)
VAL	−0.731 ***	0.530 *	−0.737 ***	0.489 *	−0.248
	(0.218)	(0.289)	(0.210)	(0.267)	(0.269)
AFF	0.0793 ***	−0.0599	0.0818 ***	−0.0498	0.0320
	(0.0268)	(0.0871)	(0.0256)	(0.0833)	(0.0862)
PGDP	−0.172	−0.304	−0.172	−0.293	−0.465 **
	(0.142)	(0.206)	(0.139)	(0.199)	(0.198)
CT	0.0109	0.673 ***	0.0166	0.640 **	0.656 **
	(0.130)	(0.258)	(0.125)	(0.249)	(0.278)
PTV	−0.00784	0.255 ***	−0.00461	0.250 ***	0.245 ***
	(0.0380)	(0.0683)	(0.0378)	(0.0688)	(0.0813)

The values in parentheses are t-statistics. ***, **, * denote significance at the 1%, 5%, and 10% levels, respectively. All variable abbreviations are defined in Table 2.

) [47]. The results show that the direct effect of the urbanization rate (UR) is 0.333, indicating that an increase in the urbanization rate in the local area has a positive impact on the number of Taobao villages, but this effect is statistically insignificant (p > 0.1). This means that although the estimated value is positive, we do not have sufficient evidence to prove that the urbanization rate has a significant direct impact on the number of Taobao villages in the long term. As for the indirect effect, local urbanization development significantly reduces the number of Taobao Villages in neighboring regions at the 1% level. The indirect effect coefficient is −1.976, indicating that an increase in the local urbanization rate has a negative impact on the development of Taobao Villages in adjacent areas, further confirming Hypothesis H3.

The total effect, combining both direct and indirect components, shows that the long-term impact of the urbanization rate (UR) is −1.643, indicating a significant negative effect on the number of Taobao Villages (p < 0.05).

In terms of control variables, looking at the direct effects, the impact coefficient of the added value of the primary industry on the number of local Taobao villages is −0.737, indicating that the higher the added value of the primary industry, the fewer the number of Taobao villages. This may be because the primary industry mainly includes agriculture, forestry, animal husbandry, and fishery, which have a lower economic added value and are difficult to support in a complex e-commerce ecosystem. The impact coefficient of the number of people employed in agriculture, forestry, animal husbandry, and fishery on the number of local Taobao villages is 0.0818, indicating that a large number of employees can participate in the e-commerce industry while engaged in agricultural, forestry, animal husbandry, and fishery production, engaging in production, packaging, and distribution, and more labor resources help support the development of Taobao villages.

Looking at the spatial spillover effects, the impact coefficient of the added value of the primary industry on the number of Taobao villages is 0.489, and it passes the significant test, showing a significant spatial spillover effect, indicating that in the long run, an increase in the added value of the primary industry in neighboring areas will have a positive impact on the number of Taobao villages in this area. The impact coefficients of the total freight volume and the total postal service volume on the number of Taobao villages are 0.640 and 0.250, respectively, and both pass the significance test, showing a significant spatial spillover effect, indicating that the total freight volume and postal service volume in neighboring areas have a positive impact on the number of Taobao villages in this area.

6. Discussion

The relationship between urbanization and rural e-commerce development is an important issue in the context of China’s urban–rural coordinated development. As a typical form of rural e-commerce, the spatial distribution of Taobao Villages is not only influenced by local market and factor conditions but is also significantly constrained by the level of urbanization in surrounding regions. Existing research on Taobao Villages has largely focused on perspectives such as industrial agglomeration and regional disparities, while discussions on the mechanisms through which urbanization affects the number of Taobao Villages—via both direct effects and spatial spillovers—remain insufficient.

This study finds that Taobao Villages exhibit significant positive spatial correlation, with stable “high–high” clusters forming in the eastern coastal regions and “low–low” clusters observed in some parts of central and western China. Such evidence indicates that the spatial distribution of Taobao Villages is characterized by strong spatial dependence and regional clustering, a conclusion that is largely consistent with related research findings on rural e-commerce and industrial agglomeration [47,48,49,50]. The clustering peaked in 2020, which may reflect both the cumulative effect of Taobao Village development and the surge in online consumption demand during the COVID-19 pandemic, though the underlying mechanisms require further investigation.

This study further finds that an increase in the level of urbanization in one region has a significant negative spillover effect on the number of Taobao Villages in neighboring areas. This suggests that, during the process of urbanization, resources, traffic flows, and policy support tend to concentrate in regions with higher levels of urbanization, thereby weakening the conditions for Taobao Village development in surrounding areas. This conclusion is consistent with existing research arguing that urbanization and industrial agglomeration may generate “polarization effects,” and it also provides micro-level empirical evidence that reinforces and extends previous findings [51,52,53].

The results indicate that the impact of urbanization on rural e-commerce development is not unidirectionally positive; its inhibitory effects on neighboring regions also warrant attention. This provides new evidence for understanding urban–rural interactions and regional balanced development, and offers policymakers a scientific basis for mitigating regional disparities, alleviating the negative spillover effects of urbanization, and promoting the coordinated development of rural e-commerce and urban–rural integration.

However, this study is not without limitations. First, the factors influencing the development of Taobao Villages are diverse. This study mainly focuses on the impact of urbanization level on their spatial distribution, centering on a single core explanatory variable without considering additional influencing factors. Second, the agglomeration of Taobao Villages is a long-term evolutionary process, and data covering a longer time span may yield more accurate results. Finally, different spatial units vary significantly in terms of resource carrying capacity, market accessibility, and policy support. A further examination of distribution characteristics across multiple spatial scales—such as metropolitan peripheries, county-level areas, and small towns—would help to enrich the conclusions. Nevertheless, due to data availability constraints, the present analysis is primarily based on the prefecture-level scale. Future research will seek more detailed micro-level data to conduct a deeper exploration of the interactive mechanisms between Taobao Villages and urbanization in the process of coordinated development.

7. Conclusions

The findings of this study highlight several important insights. First, the number of Taobao Villages shows significant positive spatial autocorrelation, with a clear clustering pattern across the country. From 2017 to 2022, their development followed an evolutionary trajectory that began with initial clustering in the southeastern coastal regions, subsequently expanded to the north, south, and inland areas, and later experienced growth in the central regions, with notable regional differences in the degree of clustering. Second, the urbanization rate has a significant positive effect on the number of local Taobao Villages in the current period, indicating that urbanization directly contributes to the expansion of rural e-commerce at the local level. Third, when decomposing the effects, the average direct effect of urbanization on local Taobao Villages is found to be positive but not statistically significant, suggesting that the positive role of urbanization may be offset or mediated by other factors. Finally, urbanization demonstrates a significant negative spillover effect on neighboring regions, suppressing the development of Taobao Villages in adjacent areas. These results imply that while urbanization can promote local development, it may also generate adverse externalities that constrain the broader regional diffusion of rural e-commerce.

The policy implications of this study are multifaceted. First, rational administrative spatial adjustment should be promoted through scientific boundary reorganization in order to optimize administrative divisions and reduce the resource misallocation and spatial fragmentation caused by divided jurisdictions. Strengthening integrated regional and urban–rural planning can help guide cities and their surrounding rural areas toward more balanced polycentric structures, thereby avoiding excessive dependence on a single growth center and promoting the optimization of urban–rural spatial structures. Second, greater attention should be given to the development needs of peripheral, rural, and disadvantaged areas in the process of policy design and resource allocation, so as to avoid excessive concentration of resources in core urban regions. By optimizing institutional arrangements, policymakers can ensure a more balanced distribution of public services and infrastructure between urban and rural areas, narrowing the development gap and improving overall coordination. In addition, cities should be encouraged to follow differentiated development paths based on their specific resource endowments and industrial foundations. Leveraging the comparative advantages of rural areas in agricultural production, cultural resources, and ecological value, while maintaining regional coordination, can support both urban and rural areas in exploring development models suited to their conditions, thus enhancing the diversity and resilience of Taobao Villages as well as urban–rural development as a whole. Finally, the rational allocation of resources between regions and between urban and rural areas should be further promoted. Establishing cross-city and urban–rural collaborative development platforms will improve the efficiency of factor flows such as talent, capital, technology, and information. Strengthened inter-regional and urban–rural cooperation can reinforce division of labor and complementarity, thereby fostering a virtuous cycle of coordinated development within and across regions.