Regional Small Towns Classification Assessment and Spatial Pattern Integration: A Case Study of the Yunnan Section of the China–Laos Economic Corridor

Abstract

The role of small towns in regional development is being emphasized, especially in developing countries, where small towns are driving regional spatial integration and optimization from the ‘bottom up’. In the context of further refinement of regional governance, it is important to identify the characteristics of regional small towns and explore the spatial pattern and structure of their development to achieve regional strategic goals. Taking the Yunnan section of the China–Laos Economic Corridor as an example, this study integrated small towns and regional high-quality development needs, constructed a regional small-town classification and evaluation index system, used various quantitative analysis methods to explore the spatial differentiation of regional small towns’ development levels, and constructed a spatial pattern of regional small towns. Our results reveal that: (1) Small towns in the Yunnan section of the China–Laos Economic Corridor showed large variations in the scores of the four indicator types, which were spatially distributed as ‘core-edge’ and ‘peripheral core’. (2) There was spatial autocorrelation in the classification assessment results of small towns, where small towns with similar levels of development were spatially adjacent and dominated by hot spot agglomerations, but with different agglomeration patterns and distribution locations. (3) The spatial pattern of regional small towns was composed of various elements such as points, lines, axes, rings, and clusters, which can meet the diversified development needs of the region. (4) Our study found that the horizontal transportation links of the Yunnan section require strengthening and suggested the construction of a ‘1 + 3’ regional transportation network.

1. Introduction

Small towns have played multiple roles in regional development as major nodes linking rural and urban areas [1,2]. They not only provide potential space for urbanization but also act as ‘growth poles’ to promote the revitalization of rural areas [3,4,5]. Therefore, they are perceived as ‘reservoirs’ to balance regional development [6]. Especially in developing countries, small towns have been regarded as a ‘bottom-up’ force for regional spatial integration and transformational development [7,8,9]. Achieving regional development goals will necessitate a focus on the quality of life and industry in small towns [10]. In China, along with high-quality development, the scale of regional governance will be further detailed. Small towns are the most fundamental administrative division and are receiving increasing attention as a weak link in the regional development system [11]. The Chinese government has introduced a series of policies to vigorously cultivate various ‘characteristic towns’ and ‘key towns’ and other small towns, and has proposed to promote urbanization with counties as significant carrier units [12,13,14]. Some small towns have the potential to turn into new districts or even small cities. It can be argued that small towns are the basic units for the optimal reorganization of regional resources and high-quality socioeconomic development.

Despite the demonstrated role of small towns, current research on regions has focused either on a ‘top-down’ macro policy approach or at the level of large cities or metropolitan areas [15,16,17,18], ignoring the physical and pragmatic nature of the bottom spatial unit [19]. Enhancing a comprehensive understanding of regional small towns and the study of their spatial patterns is essential for regional development.

Research on small towns is currently focused on social development [20,21,22], spatial planning [23,24], residents’ well-being [25,26,27], and industrial development [28,29,30]. Limited by the lack of microunit statistics, the assessment of small towns is still dominated by case surveys and qualitative analysis [31,32,33], and lacks multidimensional and comprehensive quantitative research. Despite the improvement of spatial information data and the development of analysis techniques in recent years, some scholars have used quantitative methods to analyze the current situation of small towns [34,35,36,37], and attempted to propose the direction of differentiated development of them [38,39]. Nonetheless, these studies still regarded small towns as independently developed units, and rarely explored their classification characteristics and distribution patterns, without combining the functions of regional small towns with the regional development requirements that match them. Several small towns in the region with high growth potential and development levels are not effectively identified, ignoring the current trend of small towns moving from ‘independent development’ to ‘regional synergy’.

The construction of the Belt and Road has now entered a phase of high-quality development [40], which means that the development of the regions along the route will also transfer from an emphasis on speed to a balance between quality and efficiency [41,42]. The Yunnan section is the beginning of the China–Laos Economic Corridor, and the development of this region is also pioneering and exemplary. With over four-fifths of the land area of the Yunnan section consisting of small towns, and about half of the population living in small towns, the spatial potential and influence of small towns cannot be underestimated. With the deepening of regional cooperation and the enhancement of development level, the small towns in the Yunnan section of the China–Laos Economic Corridor are facing further integration and undertaking the tasks of expanding space, carrying functions, and upgrading the level of development.

In general, this research aims to contribute to the following aspects: (1) Theoretically, this study contributes to a broad discussion on the relationship between small towns and regional development by constructing a classification assessment system for small towns and revealing the regional distribution pattern of small towns. (2) Methodologically, this study forms a combination of quantitative methods to scientifically and objectively integrate the characteristics between small towns, establish and optimize the spatial structure of regional small-town development, and provide a ‘bottom-up’ exploration model for achieving regional goals.

It is expected that the findings will serve to provide scientific guidance for the pioneering and exploratory spatial planning and construction of the Yunnan section of the China–Laos Economic Corridor and a methodology for the study of small-town development and spatial patterns in other regions.

2. Materials and Methods

2.1. Study Area

The Yunnan section of the China–Laos Economic Corridor is located in Yunnan Province in southwestern China, and represents one of China’s key gateways to the China–Indochina Peninsula, bordered to the south by the Luang Namtha in Laos and connected to the Mohan–Boten National Port. As the Yunnan section contains both the China–Laos Railway and Kunming–Bangkok Highway, the region is also an essential road and rail exit point to Southeast Asia in the Chinese ‘Belt and Road’ development strategy. Following the opening of the Kunming–Bangkok Highway at the end of 2013, the import and export of goods at the Mohan–Boten National Port increased by more than 36%, and the number of people crossing the border increased by more than 190% from 2014 to 2019. Since the opening of the China–Laos Railway in December 2021, cross-border freight volumes have exceeded one million tons in eight months, with 21 Chinese provinces and cities running cross-border freight trains on the railway, while the Lao section has also built a railway dressing yard based on the railway, truly interconnecting the railways of China, Laos, and Thailand. The Yunnan section has become an integral hub for the link between China and Southeast Asia. Despite the geographical importance of the Yunnan section and the positive impact of the two cross-border transportation arteries, it remains one of the less developed regions in China, and the level of social and spatial development still needs to be improved.

The Yunnan section consists of four prefecture-level cities, Kunming, Yuxi, Pu’er, and Sipsongpanna, of which Kunming is the provincial capital of Yunnan Province. According to the city size classification standards issued by the Chinese State Council, Kunming is a megacity with a resident population of over 5 million, Yuxi is a medium-sized city with a resident population of 500,000–1 million, and Pu’er and Sipsongpanna are small cities with a population of less than 500,000. These four cities contain a total of 241 town-level units, all of which combined account for approximately 86.8% of the total area of the region (Figure 1). Unlike many countries that define small towns based on population size, in China’s administrative division system a town is a definite administrative area, a grassroots unit within a region that is under the jurisdiction of a higher level of government. ‘Small town’, in this study, refers to this type of administrative area.

2.2. Small Town Classification Assessment System

Current research has revealed several factors affecting the development of small towns, including location, population, industry, policy, facilities, etc., involving diverse indicators (

Table 1. Relevant indicators from existing studies.

Source	Important Indicators
Alam et al., 2016 [33]	Road network density, distance from large cities, the number of public service facilities, the number of jobs offered, industrial scale, terrain, etc.
Han et al., 2020 [43]	The number of industrial enterprises, Industrial Aggregation Capability, rate of urbanization, capacity of population aggregation, distance to the prefecture-level city center, distance to transport facilities, the number of national distinctive designations received, terrain undulation, etc.
Ren et al., 2020 [44]	The number of public service facilities, road network construction, population size, number of employed, distance from the main road, distance from the main city, etc.
Tripathi et al., 2022 [45]	The number of public service facilities, distance from large cities, distance from railway stations, distance from the main road, etc.

).

In terms of the demand for high-quality regional development, China has attached great importance to the spatial expansion and layout of the Yunnan section since the completion of the China–Laos Railway, with construction focusing on industry, people’s livelihoods, transportation, and tourism. The goal of the construction is not only to promote the formation of industrial agglomerations and tourism cooperation zones but also to explore some driven service spaces and characteristic spaces. Given this, this study selected 16 indicators from four dimensions: industrial development, transport location, social livelihood, and resource potential to construct an evaluation system. The relevant indicators cover the relevant influencing factors and regional development needs of small towns. The entropy value method was used to assign weights to the indicators of small towns (

Table 2. Assessment indicators system for regional small towns.

Dimensions (Weight)	Indicators	Index Meaning/Measurement Method	Weight	Attribute
Industrial development (0.427)	Number of industrial enterprises	Statistics of towns	0.078	+
	Number of industrial enterprises above the scale	Statistics of towns	0.193	+
	Industrial Aggregation Capability	The number of industrial enterprises above the size of the town/the number of industrial enterprises above the size of the county	0.091	+
	Rate of enterprise employment	Number of employees in urban enterprises/resident population in urban areas	0.065	+
transport location (0.177)	Distance to the nearest highway entrance	ArcGIS proximity analysis	0.038	−
	Distance to the nearest high-speed rail station	ArcGIS proximity analysis	0.067	−
	Distance to the nearest border crossing	ArcGIS proximity analysis	0.014	−
	Distance to prefecture-level city center	ArcGIS proximity analysis	0.058	−
social livelihood (0.243)	Rate of urbanization	Urban built-up area resident population/urban resident population	0.108	+
	Capacity of Population Aggregation	Urban resident population/resident population in the county	0.033	+
	Density of the road network	ArcGIS density analysis	0.058	+
	Number of medical institutions	POI data	0.013	+
	Number of primary and secondary schools	POI data	0.031	+
resource potential (0.153)	Policy tilt strength	Number of honorable mentions in the national town competition	0.060	+
	Number of tourism resources	Statistics of related policy documents	0.059	+
	Degree of terrain undulation	ArcGIS raster analysis	0.034	−

).

(1) The dimension of industrial development contained four indicators: the number of industrial enterprises, the number of industrial enterprises above the scale, the employment rate of enterprises, and the industrial gathering capacity. The first three indicators reflected the scale and level of industrial development of small towns, while the latter reflected the level of the industrial gathering of small towns in the county.

(2) The dimension of transport location included four indicators: distance from the nearest highway entrance, distance from the nearest high-speed railway station, distance from the nearest border crossing, and distance from the prefecture-level city center, which reflected the competitiveness and superiority of small towns in terms of location and transportation.

(3) The dimension of social livelihood included five indicators: urbanization rate, population gathering capacity, road network density, number of medical institutions, and number of primary and secondary schools. The first two indicators represented the level of urbanization development and the ability of small towns to absorb population factors in the county, while the last three indicators considered the construction of livelihood infrastructure and service facilities such as education, medical and health care, and transportation.

(4) The dimension of resource potential consisted of three indicators: the strength of policy inclination, the number of tourism resources, and the topographic relief, reflecting the strength of policy support, tourism development potential, and spatial construction potential for small towns, respectively.

2.3. Research Methods

2.3.1. Assessment Methods

Due to the large sample size and the different attributes of the indicators in this study, standardization of the indicators was required. An indicator attribute of ‘+’ indicates that a larger indicator value is more advantageous, while an indicator attribute of ‘−’ indicates the opposite.

Indicators of ‘+’ attribute:

$$b_{ij}=\frac{a_{ij}-a_{ij}^{\min }}{a_{ij}^{\max }-a_{ij}^{\min }}$$

(1)

Indicators of ‘−’ attribute:

$$b_{ij}=\frac{a_{ij}^{\max }-a_{ij}}{a_{ij}^{\max }-a_{ij}^{\min }}$$

(2)

In the formula, i = 1, 2, 3…m, j = 1, 2, 3…n. The number of small towns is defined as m, the number of evaluation indicators is defined as n.

The improved entropy method with objective characteristics was used to assign weights to the indicators [46,47]. This method minimizes the errors caused by subjective arbitrariness. We used this method to calculate the weights of each category of indicators. According to the degree of variation of each indicator, the entropy weight of each indicator was calculated using information entropy, and the weights were modified. The indicator weight indicates the degree of influence of the indicator on the corresponding function. The calculation steps are shown below.

Firstly, we constructed the raw data matrix C:

$$C={\left\{C_{ij}\right\}}_{m\times n}$$

(3)

Secondly, we calculated the information entropy of the indicators:

$$e_j=-k{\displaystyle \sum _{i=1}^m{p}_{ij}}In(p_{ij})$$

(4)

$$k=1/In(m)$$

(5)

In this formula above, p_ij is the percentage of sample i in indicator j.

Thirdly, the variance degree of information entropy was calculated:

$$D_j=1-e_j$$

(6)

Fourthly, the weight values of the indicators were calculated:

$$W_j=\frac{D_j}{{\displaystyle \sum _{j=1}^n{D}_j}}$$

(7)

Finally, the total score for each type of indicators and the comprehensive score for all indicators were calculated using the following formula:

$$X_i={\displaystyle \sum _{j=1}^n{w}_j}{p}_{ij}$$

(8)

2.3.2. Analysis Method of Spatial Pattern

Exploratory spatial data analysis (ESDA) can determine the presence of agglomeration in spatial units and identify agglomeration characteristics. This study used this method to analyze the spatial autocorrelation of the development quality of small towns and to identify the specific locations of spatial agglomerations in various types of small towns.

Global Moran’s I index is used to describe the degree of association of spatial units with their surrounding areas throughout the study area, revealing whether there is a significant distribution pattern of spatial objects in the global context. Moran’s I indices are between [–1, 1]; if the index value is positive and the p-value < 0.05, it indicates a significant positive correlation, i.e., small towns with comparable levels of development are spatially significantly clustered; if the index value is negative and the p-value < 0.05, it indicates a significant negative correlation, i.e., small towns with comparable levels of development are spatially discrete; if the index value is 0 or tends to 0, this indicates that there is no spatial autocorrelation [48].

Its calculation formula is as follows:

$$I=\frac{k{\displaystyle \sum _{i=1}^k{\displaystyle \sum _{j=1}^k{{\displaystyle w}}_{ij}\left(x_i-\overline{x}\right)\left(x_j-\overline{x}\right)}}}{\left({\displaystyle \sum _{i=1}^k{\displaystyle \sum _{j=1}^k{w}_{ij}}}\right){\displaystyle \sum _{i=1}^n{\left(x_i-x\right)}^2}}$$

(9)

$$\overline{x}=\frac{1}{n}{\displaystyle \sum _{i=1}^n{x}_i}$$

(10)

In the formula above, k is the number of small towns in the study area; w_ij denotes the spatial weighting factor of small towns: if they are adjacent to each other, then w_ij = 1, if not adjacent, then w_ij = 0; x_i and x_j denote the development quality of small towns i and j.

Hot spot analysis is a cluster distribution mapping tool that can identify statistically significant spatial clusters by calculating the Getis–Ord G_i* statistic for each sample element, revealing the spatial clustering characteristics and distribution patterns of local areas. This study used this method to explore the clustering characteristics of small towns in the region. The method reflects the confidence level through confidence intervals, and confidence levels of 99%, 95%, and 90% indicate that the factor clustering is statistically significant, with higher confidence levels indicating a lower probability of randomness in factor creation.

Its calculation formula is as follows:

$$G_i\left(d\right)=\frac{{\displaystyle \sum _{j=1}^n{w}_{ij}\left(d\right)x_j}}{{\displaystyle \sum _{j=1}^n{x}_j}}$$

(11)

$$Z\left(G_i\right)=\frac{\left[G_i\left(d\right)\right]-E\left(G_i\right)}{\sqrt{VAR\left(G_i\right)}}$$

(12)

In the formula above, G_i(d) is the degree of correlation between i and j in small towns; w_ij is the spatial weight within the distance d; E(G_i) and VAR(G_i) are the mathematical expectation and variance, respectively. If Z > 0 and is significant, it indicates that a small town i is surrounded by high quality development areas and small town i is a hot spot, vice versa for a cold spot.

2.4. Data Sources

The data sources for this study are shown below (

Table 3. Data sources.

Data Types	Data Sources
Spatial data (including administrative boundaries, topography, roads, high speed rail stations, border crossings, etc.)	China National Geographic Information Resources Catalogue Service System
POI information	Baidu Maps open platform
industry statistics and population statistics	China County Statistical Yearbook
Tourism resources	China Association of National Parks and Scenic Sites, National Forestry Administration, Yunnan Provincial Department of Culture and Tourism

):

3. Results

3.1. Results of the Assessment

The study standardized the data of 16 indicators in four types using descriptive statistics, calculated the weights of the indicators using assessment indicators and methods, and assessed the development indicators of 241 small towns contained in four cities of the Yunnan section of the China–Laos Economic Corridor in each of the 4 aspects (Table 2); the indicators of the four dimensions were then weighted to obtain the comprehensive evaluation results of small towns. The assessment results were classified into four levels: high, moderately high, moderately low and low, using the natural breakpoint method, which can clarify the spatial differentiation of the development quality of regional small towns, and to visualize them in ArcGIS (Figure 2 and Figure 3).

3.1.1. Global Distribution Characteristics of Small Towns

As shown in Figure 2 and Figure 3, the comprehensive assessment results of small towns differ significantly from the results of different types of assessments, and the small towns with high scores showed a spatial distribution pattern of ‘core-edge’ in the region. In terms of the comprehensive assessment, the high scores of small towns around the urban area of Kunming contrast with the low scores of small towns in other areas. Small towns at high and moderately high levels were apparently around the capital of Yunnan Province—the urban area of Kunming, such as Yanglin Town, which has the highest quality of development, and Xiaojie, Beigucheng, Jiucun, Er’jie and Shangsuan, which are at moderately high level, and these small towns are clearly driven by the big cities. In addition, Ning’er Town in Pu’er City, which is not only the center of the county but also located near the urban area of the prefecture-level city, scored moderately high in the comprehensive assessment.

In terms of different types, the small towns with high and moderately high scores in the industrial development dimension were mainly located around the urban area of Kunming. In the transportation location type, the farther away from the main transport routes, the lower the score of small towns, and almost all small towns at the high level were located around urban areas. In the resource potential type, small towns near urban areas in Kunming, Pu’er and Sipsongpanna showed strong development potential, while small towns far from urban areas were dominated by low potential. The small towns with high and moderately high scores for social and livelihood development were generally scattered, whereas the small towns with high and higher quality in the northeast side of Kunming showed a phase gathering trend, and also reflected a ‘core-edge’ characteristic.

3.1.2. Local Distribution Characteristics of Small Towns

Although small towns with high development levels were mainly located around the urban areas of prefecture-level cities, there were also ‘special’ small towns with high scores in the peripheral local areas in the assessment of small towns by different types, and most of these small towns have one or more functional attributes. These small towns have certain core positions compared with the surrounding space, and are an important force in serving and driving the high-quality development of peripheral areas.

The county towns supported the high-quality development of people’s livelihood and industry in the local area. Towns such as Menglang, Menghai, Nayun, Weiyuan, Jinping and Mengla scored high and moderately high in the social livelihood aspects as county towns, and among these, the industrial development level in Weiyuan was prominent; thus, the influence of these county towns in the peripheral regions should be further emphasized and brought into effect. The border ports also enhance the transport location superiority of the surrounding small towns to a certain extent. For example, the towns of Mohan and Mengla, which are in the vicinity of the Mohan Port, are vital nodes on the Kunming–Bangkok Road and China–Laos railway line, and Mohan Town is at the same time an integral part of the China–Laos Economic Cooperation Zone. In terms of the score of the transport location aspect, the hierarchical differences between Mohan Town and Mengla Town and other small towns in the vicinity are obvious and show core characteristics. In addition, in terms of resource potential, the town of Gasa has won national business cards such as National Characteristic Town, National Key Town, and National Model Pilot Town by virtue of its resource endowment, and is a small town with high development potential in the peripheral region.

3.2. Spatial Agglomeration Pattern of Regional Small Towns

Based on the above evaluation results, there are more obvious geographical differences in the spatial distribution of small towns’ development quality in the Yunnan section of the China–Laos Economic Corridor. To further identify the spatial clustering of development quality of small towns and clarify the spatial pattern, Global Moran’s I index (

Table 4. Results of Global Moran’s I.

Dimensions	z-Score	p-Value	Moran’s I
Industrial Development	3.476	0.000	0.123
Transportation Location	21.606	0.000	0.846
Social Livelihood	7.124	0.000	0.207
Resource Potential	3.128	0.001	0.114
Integrated Evaluation	4.436	0.000	0.160

) was used to analyze the four dimensions of small towns and the global spatial autocorrelation characteristics of integrated evaluation, and the INVERSE_DISTANCE was chosen to conceptualize the spatial relationships.

The global Moran’s I indices of small towns in terms of industrial development, transport location, social livelihood, resource potential, and integrated evaluation were 0.123, 0.846, 0.207, 0.114, and 0.160; the z values were 3.476, 21.606, 7.124, 3.128, 4.436. All global Moran’s I passed the significance test at the 0.001 level. This indicates that there is a significant positive correlation in the spatial distribution of the development quality of small towns in the Yunnan section, and that small towns with comparable development levels were clustered in space. Further, on the basis of clarifying the global autocorrelation characteristics, local spatial autocorrelation analysis was conducted using hot spot analysis to explore the spatial clustering characteristics of regional small towns (Figure 4 and Figure 5).

3.2.1. Clustering Characteristics of Small Towns in Different Dimensions

(1)

Industrial development: undertaking the transfer of urban industrial functions

In the industrial development dimension, Figure 4a presents two larger hot spot clusters and two independent hot spot areas in the north and south, both of which are located at the edges of urban areas of prefecture-level cities. The northern hot spot group is located in the east of Kunming downtown and consists of four small towns, namely Yanglin, Niulanjiang, Majie and Beigucheng. The southern hot spot group is adjacent to the urban area of Pu’er and consists of four small towns, which included Ning’er, Mengxian, Mohei and Tongxin. The two hot spot groups featured a good industrial foundation and a strong industrial aggregation capacity. The two independent hot spot areas were both located at the edge of Kunming urban area, namely Erjie and Shangsuan in Jinning District of Kunming City, both of which had good foundations for industrial specialization and development (

Table 5. Cluster results of industrial development dimension.

Significance of Hot Spot	Small Towns	Numbers
99% Confidence	Yanglin, Niulanjiang, Majie, Beigucheng, Ning’er, Mengxian, Mohei, Tongxin	8
95% Confidence	Shangsuan, Erjie	2

).

(2)

Transport location: influenced by the China–Laos Railway and Kunming–Bangkok Road

In the transport location dimension, Figure 4b shows a hot spot zone with a continuous longitudinal direction ‘belt’ distribution of small towns mainly along the China–Laos Railway and the Kunming–Bangkok Road. The hot spot zone extended from Yangzong in the north to Mochan in the south, spanning three cities: Yuxi, Pu’er and Xishuangbanna. Among them, the small towns with the highest significance of hot spot basically contained high-speed railway stations or highway access, which assume certain transportation hub or channel functions and provide important spatial support for the flow of people, logistics, technology, and information in the China–Laos Economic Corridor (

Table 6. Cluster results of transport location dimension (hot spot).

Significance of Hot Spot	Small Towns	Numbers
99% Confidence	Shangsuan, Jiuxi, Yangwu, Lianzhu, Tongguan, Mohei, Jingxing, De’an, Yutang, Mengxian, Ni’er, Nanping, Puwen, Dadugang, Mengyang, Mengla, Mohan	17
95% Confidence	Luohe, Fuliangpeng, Dianzhong, Chahe, Tadian, Huanian, Pingdian, Manlai, Jianxing, Mengnong, Mili, Yinyuan, Yangjie, Na’nuo, Xin’an, Longba, Yayi, Naha, Longtan (in Pu’er), Puyi, Tongxin, Dehua, Zhengxing, Yunxian, Liushun, Jingne, Menglun, Ji’nuo, Menghan, Jingha, Liujie, Xiaoshiqiao, An’hua, Qianwei, Jiangcheng, He’xi, Si’jie, Nagu	38
90% Confidence	Yangzong, Mosha, Si’nanjiang, Qixiang, Longtan (in Yuxi), Wa’die	6

).

In addition, it is indicated that there are three cold spot clusters in the north, northwest, and southwest of the Yunnan section of the China–Laos Economic Corridor (Figure 4b). Small towns within these cold spot clusters need to improve the transportation support structure of the corridor by strengthening transportation infrastructure and linkages with the urban areas to the east, thus further expanding the radiation impact of the China–Laos Railway (

Table 7. Cluster results of transport location dimension (Cold spot).

Significance of Cold Spot	Small Towns	Numbers
99% Confidence	Jiaopingdu, Wudongde, Tanglang, Malutang, Shekuai, Yinmin, Sayingpan, Zehei, Jingpin, Jingfu, Linjie, Wendong, Xuelin, Muga, Laba, Zhongke, Zhutang, Mengsuo, Mengka, Li’suo, Menggake, Nayun, Mangxin, Donghui, Jingxin, Xinchang, Yuesong, Wumeng, Xueshan	29
95% Confidence	Yunlong, Tuanjie, Zhongpin, Zhuanlong, Tuobuka, Tangdan, Hongtudi, Dashanchaodong, Mandeng, Dajie, Taizhong, Longjie, Wenlong, Manwan, Fudong, Dashan, Ankang, Donghe, Nanling, Nuofu, Fuyan, Fubang, Shangyun	23
90% Confidence	Jiulong, Maoshan, Cuihua, Anding, Wenjing, Mengda, Huashan, Menglang, Jiujing, Huimin, Mengma	11

).

(3)

Social livelihood: the role of small towns’ hierarchy is obvious

In terms of the social livelihood dimension, Figure 4c indicates four hot spot clusters. Among them, one cluster was adjacent to downtown Kunming, while three clusters contained three county seats: Weiyuan, Menghai, and Mengla, respectively. These agglomerations were close to cities of high administrative level or contained small towns with administrative functions, which had a good foundation for social development and were attractive in terms of investment and construction of infrastructure and public service facilities to serve the surrounding areas. Compared with other small towns, such clusters can provide residents with a wider variety of public services and have a higher population carrying capacity (

Table 8. Cluster results of social livelihood dimension.

Significance of Hot Spot	Small Towns	Numbers
99% Confidence	Niulanjiang, Yanglin, Xiaojie, Menghai	4
95% Confidence	Mengla	1
90% Confidence	Yangjie, Weiyuan, Minle, Menghun, Mengzhe, Mohan	6

).

(4)

Resource potential: inclined to tourism resource aggregation areas

In the resource potential dimension, one hot spot cluster and two hot spot zones can be seen in Figure 4d; hot spots are generally distributed in areas with abundant tourism resources. The hot spot cluster consisted of 9 small towns, which were surrounded by the periphery of Sipsongpanna urban area. This cluster featured a large potential for tourism, which included not only natural resources such as Sipsongpanna National Nature Reserve, national forest parks and scenic spots, but also hundreds of human landscape resources in the form of Dai-style Buddhist temples, residential houses, traditional villages and other material cultural heritage. Two independent hot spot zones are Yixiang Town in Pu’er City and Gasa Town in Yuxi City, the former of which was located within the Sun River Provincial Nature Reserve and was the main location of the Sun River National Forest Park, with rich natural scenic tourism resources. The latter has been rated as a ‘National Characteristic Town’, which combined the natural scenery represented by the original forest of Ailao Mountain, the human landscape represented by the Huayao Dai ethnic cultural village, and the agricultural tourism resources represented by the Chu Orange Manor (

Table 9. Cluster results of resource potential dimension.

Significance of Hot Spot	Small Towns	Numbers
99% Confidence	Mengyang, Jinuo, Menghan, Jingha, Menglong, Gelanghe, Mengsong, Menghai, Mengzhe	9
95% Confidence	Yixiang	1
90% Confidence	Gasa, Bulangshan	2

).

3.2.2. Integrated Agglomeration Characteristics

From the perspective of integrated agglomeration characteristics, there were 11 small towns in the north that passed the significance test (

Table 10. Cluster results of integrated evaluation.

Significance of Hot Spot	Small Towns	Numbers
99% Confidence	Yanglin, Niulanjiang, Majie, Beigucheng, Ning’er, Mengxian	6
95% Confidence	Jiucun, Yousuo, Shangsuan, Er’jie, Tongxin	5
90% Confidence	Yangzong, Liujie, Jiangcheng, Mohei	4

). Figure 5 indicates that they are distributed around the urban area of Kunming in the form of ‘semi-surrounding’. Among them, Yanglin–Niulanjiang–Majie–Beigucheng on the east side of Kunming form the most significant hot spot group, which coincides with the northern hot spot group of the industrial dimension. The hot spot cluster located in the south is neighboring Pu’er, which also partially overlaps with the southern hot spot cluster in the industrial dimension. Consequently, the small towns located in these two hot spot clusters not only have outstanding industrial aggregation capacity, but also possess a high level of comprehensive development.

3.2.3. Construction of Spatial Pattern of Regional Small Towns

According to the agglomeration characteristics of regional small towns in the Yunnan section of the China–Laos Economic Corridor, it can be concluded that: (1) The hot spot clusters in the industrial development dimension were adjacent to urban areas and highly overlap with the integrated clustering results, and were integrated town clusters led by industries. (2) The traffic-dominated areas of the small towns were primarily distributed vertically along the China–Laos Railway and the Kunming–Bangkok Road, but the horizontal traffic in the region needs further improvement, which can form a traffic network of ‘1 + 3’. (3) Some small towns with ‘county’ attributes had higher social and people’s livelihood scores; among them, Weiyuan, Menghai, and Mengla have formed a hot spot group together with neighboring small towns, and the central function was more prominent than other counties. This hot spot group can become a regional service center. (4) The small towns around Sipsongpanna City form an ‘aggregation circle’ of natural and humanistic tourism resources. The town of Yixiang and Gashan are endowed with rich tourism resources, and were characteristic towns with the role of radiation.

Based on the identification of spatial features and agglomeration characteristics mentioned above, we coordinated the organization of points, lines, axes, rings, groups, and other elements to build an integrated development belt around Kunming, a tourism agglomeration circle around Sipsongpanna, three service-oriented town clusters, two industry-oriented town clusters, multiple core nodes, and a ‘1 + 3’ transportation support network, which jointly constitute a spatial pattern of regional small towns (Figure 6).

4. Discussion

The role of small towns in regional development is increasingly emphasized [49]; additionally, China has continued to focus on the positive role of small towns in regional development since the beginning of its reform and opening up [50,51]. At present, small towns have transformed from quantitative expansion to quality improvement, and balanced development to focused construction. Nevertheless, most of the current studies on regional spatial patterns focus on the macro and meso levels and pay insufficient attention to the hidden micro differences within regions [52]. In this regard, we explored the regional spatial pattern from the perspective of small towns to propose a more pragmatic path for intra-regional development.

The study found that the development of small towns showed a clear spatial differentiation in the region, with most of the small towns close to urban centers having a higher level of development in terms of industry, transportation, and living services, and higher potential, showing a ‘core-edge’ global distribution, which was similar to some other studies in developing countries [53,54,55]. Instead, some of these small towns stand out under their unique spatial resources or social characteristics, becoming peripheral areas of the local development plateau. It was pointed out that the ‘core groups’ also exist within the regional periphery, and the phenomenon was referred to as the ‘peripheral core’ [56]. This study confirmed this phenomenon from the perspective of small towns. Such small towns can assist in overcoming inequalities in the regional spatial structure by becoming service centers, transportation hubs, or specialized sectors within a certain geographical area.

Cluster analysis further concluded that small towns with comparable levels of development are spatially clustered; some adjacent small towns have formed functional clusters with regional development advantages, while others are at a disadvantage in some dimensions. By gaining a comprehensive perspective on the clustering characteristics and distribution patterns of small towns, the town-units can be synergistically integrated to better identify the strengths and weaknesses of small towns in regional development. This research identified small towns with high growth potential and was able to leverage the group effect of small towns to guide the planning and construction of the region more efficiently.

In comparison with other research [38,39,56], this study was able to better match the functions of small towns with regional development goals. On the one side, in the selection of assessment indicators, we not only learned from published studies on small towns but also took into account the regional development goals of the Yunnan area. On the other side, we explored the regional spatial distribution patterns and characteristics of small towns. As a result, the constructed spatial structure can also be integrated with the demands of regional development. For instance, small town clusters with high-quality industrial development are not only adjacent to urban centers, but also of high comprehensive quality, providing suitable spatial carriers for the demands of ‘developing industries’ along the corridors and promoting the construction of new cities in the implementation plan of the Yunnan section of the China–Laos Economic Corridor, which can be developed into new industry-led urban areas. The ‘agglomeration circle’ of tourism resources and the two characteristic small towns included in the resource potential dimension are also compatible with the requirements of ‘developing cultural tourism’ and ‘building tourism demonstration zones’. The three cold clusters identified in the transport location dimension can also help the corridor region to improve transportation and promote the corridor to enhance access capacity and build hubs. The various clusters and nuclei formed by the social livelihood type also contribute to the promotion of balanced regional development and equity in public services.

5. Conclusions

Based on quantitative methods, this study assessed small towns within the Yunnan section of the China–Laos Economic Corridor considering four aspects: industrial development, transport location, social livelihood, and resource potential, as well as determining the clustering characteristics and spatial patterns of small towns through spatial cluster analysis methods. The study found that small towns in the Yunnan section showed significant differences in development levels among the four types, and that these differences were generally characterized by a ‘core-edge’ distribution, but there were ‘peripheral cores’ in the local areas of the region. The spatial forms of regional small towns can be divided into four categories: group, line (belt), ring and point, with different spatial structure composition patterns and associated elements, resulting in one integrated development belt, one tourism-oriented development circle, two industry-oriented town clusters, three regional service-oriented town clusters, and several county services points. Furthermore, the study found that the Yunnan section should further improve the transportation links with small towns in the western part of the corridor on the basis of the China–Laos Railway and Kunming–Bangkok Road as the vertical transportation spine, forming a ‘1 + 3’ transportation network to support the construction of high-quality regional development.

This study was of reference value for regional resource integration and spatial pattern construction and provided a basis for the development and construction of China–Laos economic corridors. However, whether the proposed construction ideas and methods are universal in the typological sense remains to be verified by more cases of different geographical areas and different types of economic corridors. In addition, in the next phase of the study, we will analyze the characteristics of enterprises within small towns and study the networks of enterprises and industrial relations between small towns in the region, thus making additional recommendations for the sustainable development of small towns in economic terms.