Distribution Characteristics and Influencing Factors of Supply Chain Innovation Firms: A Case Study of Zhejiang Province

Abstract

The establishment of supply chain innovation enterprises is conducive to maximizing production efficiency, deepening the division of labor mechanism in all links, and promoting supply-side structural reform. In order to study the factors related to the distribution of supply chain innovation enterprises, this study was based on the pilot list of the first and second batches of supply chain innovation enterprises in Zhejiang Province from 2017 to 2019, and based on the geographical location points of 187 enterprises. The POI data were analyzed with the GeoDa 1.20 and ArcMap 10.8 systems, using tools such as the Moran index, nearest neighbor index, point density estimation, standard deviational ellipse, etc. The analysis results show that the distribution characteristics and influencing factors of supply chain innovation enterprises in Zhejiang Province demonstrate global autocorrelation and a high degree of local aggregation, forming a “one main multi-point” distribution with Hangzhou as the center. The influencing factors are mainly affected by employment density and local economy, and there are multi-factor interactions. Finally, this study puts forward suggestions for the improvement of supply chain innovation enterprises, hoping to promote the sustainable development of supply chain innovation enterprises.

1. Introduction

A supply chain is the organizational form of a series of activities, such as joint manufacturing, logistics and transportation, and customer demand, aiming to improve the production capacity level, promote the efficiency of logistics, meet customer demand, and realize the maximization of interests and optimal resource conditions. However, the traditional supply chain has many disadvantages which are not conducive to reducing the cost and increasing the efficiency of logistics links. With the development of information technology, intelligent logistics arises at the same historical moment. Supply chain enterprises refer to enterprises that coordinate logistics activities such as order management, warehousing, sorting, and transportation. The birth of supply chain enterprises symbolizes the regulation and informationization of logistics activities, and their competitiveness is remarkable in traditional logistics enterprises. At present, the integration of supply chain enterprises and Internet technology has spawned a new stage of development and has formed supply chain innovative enterprises. The motivation of supply chain innovation enterprises comes from two aspects: one is the industry environment, the other is the regional innovation environment. Firstly, the regional innovation system theory puts forward the view that the regional innovation environment will have a certain impact on the development of enterprises in the region. The industrial environment is divided into market policy, industrial innovation and development, and market demand. Market policies include laws and regulations, management systems, and government service systems. The promulgation of policies affects the overall environment of enterprises. Positive policies can boost the development of enterprises, stimulate the vitality of enterprises, and form a good “enterprise industry” double cycle; it is helpful to realize the sustainable development of supply chain innovation enterprises. Industry innovation and development is the embodiment of high industry concentration and the overall trend of enterprises. The constant change of market demand promotes the transformation and upgrading of enterprises, which is mainly reflected in the penetration and application of informatization. Regional environment covers resource environment and consumer environment. Resource environment refers to enterprise capital, information, and other elements. Capital assistance is the booster for enterprises to accelerate. A large amount of information collection is conducive to the improvement of the competitiveness of enterprises in market innovation. The consumer environment is the feedback performance of market demand. The improvement of market demand requires the consumer environment to be more immediate, transparent, and traceable. The conceptual diagram of the development motivation of supply chain innovation enterprises is shown in Figure 1.

Supply chain innovation is embodied in many aspects: (1) the various industries, such as the combination of rural industry and the supply chain; (2) as a service for an intelligent, modern, and global supply chain system [1]; (3) combined with finance, to prevent financial risks; (4) to establish the green supply chain [2] and the reverse logistics supply chain, and to explore the new consumer market. However, there are still many problems in the current supply chain innovation enterprises. Due to the high requirements of supply chain innovation for enterprises and the need for data system support, the imprecision and multiple sources of data lead to the reduction or even abnormality of system processing efficiency, and it is difficult for ordinary enterprises to maintain the operation of the system with sufficient funds. The integration and innovation of supply chains need the overall consideration of upstream and downstream, and to put forward higher requirements for information integration and symmetry. At the same time, the improvement of living standards makes consumers demand timeliness and effectiveness in supply chains, and improving customer satisfaction has become a major difficulty. There are loopholes in the existing supply chain innovation management system. Although the tax, quality inspection, business administration, and other departments have increased the supervision of supply chain innovation enterprises, the management function is relatively separate, and it is difficult for it to play a significant role. Since 2017, The General Office of the State Council has issued guidelines on promoting innovation and application of supply chains, which has gradually realized a deep integration between supply chains, production, and circulation, and has effectively realized supply-side structural reform. In 2018, the Smart Logistics and Supply Chain Innovation and Development Summit Forum was held in Yiwu to promote the construction of smart logistics supply chains and cross-border e-commerce supply chain layout. In 2019, Zhejiang Province put forward a series of implementation opinions and safeguard measures according to the guidance of The State Council. At the same time, the provincial supply chain innovation and application pilot cities and enterprises in Zhejiang Province have been released, and they are committed to giving full play both to the decisive role of the market and to the role of the government in promoting it, so as to make Zhejiang an important center for the innovation and the development of modern supply chains in China. In 2020, Zhejiang held the 12th of China’s logistics and supply chain informatization conferences, which further promoted supply chain innovation and blockchain technology [3], Internet of things, and big data. In April 2021, work on the promotion of collaborative innovation in Zhejiang Province’s supply chain was held in Ningbo for the release of “Zhejiang Province Modern Supply Chain Development”, “14th Five-year plan”, and “Zhejiang Province Supply Chain Innovation and application pilot enterprise typical case set”. At present, Zhejiang Province is a frontier province in the development of the supply chain industry, and supply chain innovative enterprises are booming. The distribution of supply chain innovation enterprises in Zhejiang Province has a certain research significance. This result can provide a representative reference for the distribution of supply chain innovation enterprises in China and carry out national expansion research. Therefore, this study will analyze the list of the first and second batches of supply chain innovation and application pilot enterprises in Zhejiang Province. A total of 189 pilot enterprises of supply chain innovation can be obtained through the list, among which 187 enterprises were selected as analysis samples due to inaccurate information on the geographic location of two enterprises. This study used GIS spatial methods such as standard deviational ellipse and point density analysis to conduct correlation analysis on the spatial distribution of pilot enterprises, in order to obtain the distribution structure and pattern characteristics and thereby provide feasible improvement policies and guidance for the relevant departments.

2. Literature Review

POI data represents the point feature data of real geographical entities, reflects the spatial location of geographical elements, and includes attributes such as name, address, and longitude and latitude coordinates [4,5]. POI data [6] can improve the accuracy of geographical location description, better reflect the spatial structure and distribution of research objects, and improve the accuracy of research [7,8]. Through the spatial distribution method of innovation-oriented enterprises in Zhejiang Province, Liu Kaikai, Feng Xiuli, Zhou Meiling et al. [9] explored the spatial distribution characteristics and influencing factors of innovation-oriented enterprises at the provincial scale by using the nearest proximity index, kernel density analysis, and other research methods, combined with the geographical factor detection method. The nearest proximity index [10] refers to the comparison of the average distance of point elements under the assumption of a random distribution by calculating the average distance of all point elements and their nearest neighbors. Kernel density analysis [11] refers to density analysis of point and line patterns and the use of core estimation to simulate the detailed distribution of variable data. Zhan Dongsheng and Zhang Qianyun et al. [12] analyzed the distribution of real estate enterprises in Hangzhou by using spatial point model methods such as nearest proximity index and Ripley’s K-function. Li Cong and Lu Minghua [13] took 361 prefecture-level administrative regions in China as spatial units and analyzed the spatio-temporal characteristics and influencing factors of China’s import e-commerce enterprises by using kernel density, spatial autocorrelation analysis, and spatial measurement methods. Spatial autocorrelation analysis [14] is a method of statistical testing which is divided into global and local hypotheses. It can measure the distribution of spatial things with autocorrelation. The higher the correlation, the stronger the spatial agglomeration. Yan Dongsheng and Wu Qiang [15] took the Yangtze River Delta as their research area and explored the spatio-temporal pattern evolution of high-tech enterprises based on the coefficient of variation and other methods. The coefficient of variation is used to analyze spatio-temporal evolution. The larger the coefficient of variation, the more concentrated the growth and the more dispersed the distribution.

Another kind of study in the literature addresses the relationship between distribution characteristics and variable factors. Pan Fangjie and Wan Qing et al. [16] analyzed the data of A-level logistics enterprises for 14 years, and concluded that A-level logistics enterprises showed a trend of “south (west)—north (east)” on the national scale. The development stage of enterprises was unbalanced and insufficient, and there were obvious step-type differences in cold or hot spots. The distribution of logistics enterprises is the comprehensive result of economic strength, industrial structure, and industrial level. Luo Xiang, Song Xin, and Zhu Lixia et al. [17] obtained the spillover situation and employment agglomeration degree of the Beijing–Tianjin–Hebei urban agglomeration by analyzing the panel data of 180 districts and counties, and concluded that there was a significant correlation between the two. Li Hui, Ren Qilong, Mao Guangxiong, and Liu Chuanming et al. [18] took 31 provinces in China as their research unit and analyzed the pattern of logistics enterprises from 2005 to 2018. They concluded that the overall distribution of enterprises showed a pattern of greater concentration in the south than in the north, with that distribution being smaller at both ends and larger in the middle. The distribution pattern of enterprises showed a trend of moving southwest; among the economic factors, government planning and traffic location are the important affecting the distribution of logistics enterprises. Huang Xiaoliang [19] studied the high-tech industry in the Pearl River Delta region to explore how the high-tech industry promotes regional economic growth. According to the relationship between industrial isomorphism and industrial elasticity, he found that high-tech industries have significant positive externalities.

Through a review of the relevant literature, it can be found that there is a large number of publications on spatial distribution, but there is still room for improvement. First, the existing research mainly focuses on the distribution of regional industries, supply chain enterprises, and innovative enterprises, but there is little research on the distribution of innovative enterprises in the supply chain. Second, under the same conditions, the choice of spatial scale affects the reliability and accuracy of the results. The existing studies are based on national samples, with a wide distribution range, resulting in significant regional differences which cannot accurately reveal the situation on a provincial scale. When judging the relationship between enterprise distribution pattern on the one hand and economic factors, educational factors, industrial structure factors, urbanization factors, etc. on the other, it is difficult to analyze the distribution aggregation relationship. In order to prevent the phenomenon of data deviation, we can better analyze the spatial distribution law by selecting an inter-provincial scale and an accurate location of enterprises. Third, most of the existing research methods limit themselves to description of the phenomenon; they lack scientific and systematic spatial econometric analysis as well as any comparative analysis of various influencing factors. Therefore, this paper takes the supply chain innovation enterprises in Zhejiang Province as an example. By visualizing the pattern distribution of supply chain innovation enterprises on the map and the relationship between the relevant factors, we can better explore the internal relationship and external influence relationship of enterprises. According to the analysis results and conclusions, this paper puts forward good countermeasures for the development of enterprises, realizes the sustainable development of supply chain innovation enterprises, and helps enterprises improve their competitiveness in the market and integrate themselves into the industry environment.

3. Data Sources and Research Methods

3.1. Data Sources

This study aims to analyze the spatial distribution of supply chain innovative enterprises in Zhejiang Province. Since Zhejiang Province implemented the policy of promoting supply chain innovation in 2019, the “14th Five-year plan” of developing the modern supply chain was put forward in 2021. Therefore, the enterprise sources selected were the first batch and the second batch of the supply chain innovation pilot enterprises. Through the coordinate picker system of the Tencent location service, the specific coordinates of 187 supply chain innovation enterprises in Zhejiang Province were obtained. Using the ArcMap 10.8 software, https://developers.arcgis.com (accessed on 17 October 2021), the vector map of Zhejiang Province was imported as the base map and the spatial coordinate points of the enterprises were located to obtain the distribution map of points. In addition, variable indexes obtained in this study include GDP, number of employed persons in the secondary industry, number of employed persons in the tertiary industry, employment density, education level, and urbanization rate in Zhejiang Province. Data of variable indexes are all from the Statistical yearbook of Zhejiang Province in 2019 and 2020. The setting and description of each variable are shown in

Table 1. Setting and description of each variable.

Variable Classification	The Variable Name	Variable Meaning	Unit
Explained variable	Employment density	The number of people employed per unit area	People/km2
Control variables	Municipalities’ GDP	Gross domestic product	One hundred million yuan
Control variables	Employment in the secondary and tertiary industries	Employment in secondary and tertiary industries	Ten thousand people
Control variables	Education level	Number of higher education institutions	Person
Control variables	Urbanization rate	Urban population to total population	%

Note: Employment density index refers to the number of employed persons in the secondary and tertiary industries/administrative area.

:

3.2. Research Methods

3.2.1. Moran Index

Spatial autocorrelation statistics is used to measure a basic property of geographic data, the interdependence between data in one location and data in other locations. Spatial autocorrelation statistics can be roughly divided into two categories: global autocorrelation and local autocorrelation. Global autocorrelation is to describe the overall distribution and judge whether this phenomenon has aggregation characteristics in space. Local autocorrelation can calculate the scope of the gathering place. The spatial distribution of supply chain innovation firms is correlated, and the correlation degree can be expressed by Moran index. The range of the spatial autocorrelation coefficient is [−1, 1], where −1 represents a negative correlation, 1 represents a positive correlation, and 0 represents no correlation. The correlation degree is proportional to the value. The GeoDa 1.20 software, http://geodacenter.github.io/, accessed on 11 October 2021, was used to analyze the correlation degree of supply chain innovation enterprises in Zhejiang Province, and the Z value and P value were obtained. The formula of the Moran index is:

$$I=\frac{{{\displaystyle \sum }}_{i=1}^n{{\displaystyle \sum }}_{j=1}^n{\omega }_{ij}\left(x_i-\overline{x}\right)\left(x_j-\overline{x}\right)}{S^2{{\displaystyle \sum }}_{i=1}^n{{\displaystyle \sum }}_{j=1}^n{\omega }_{ij}}=\frac{{{\displaystyle \sum }}_{i=1}^n{{\displaystyle \sum }}_{j=1}^n{\omega }_{ij}\left(x_i-\overline{x}\right)\left(x_j-\overline{x}\right)}{S_0}$$

(1)

where, n is the number of spatial point elements and is also the spatial weight matrix, reflecting the positional relation between spatial elements I and j. The calculation method of spatial weight in this paper was Queen, and the spatial coefficients were divided into two categories, represented by 0 and 1, respectively. A value of 0 means that two regions are not adjacent, and 1 means the reverse. x_i and x_j represent geometric attribute information and $\overline{x}$ represents the average of the set point density, respectively, and S² represents the variance of the point density.

3.2.2. Nearest Point Index

The spatial distribution of point elements can be divided into three types: aggregated, uniform, and random. The nearest distance of point elements with uniform distribution is the largest, followed by random distribution and the smallest condensed distribution. The nearest neighbor distance is a geographical index indicating the degree of mutual proximity of point things in geographical space. The nearest neighbor index can well reflect the spatial distribution characteristics of point elements.

The nearest neighbor point index compares the average distance of the point elements under the assumption of random distribution by calculating the average distance of all the enterprise point elements and its nearest enterprise spatial points. If the average distance of random distribution is larger than the actual average distance, there is a spatial agglomeration effect among the enterprises. If they are equal, however, they are evenly distributed. The formula for calculating the nearest proximity index is:

$$R=\frac{\overline{r_1}}{r_E}=2\surd D\times \overline{r_1}$$

(2)

$$\overline{r_E}=\frac{1}{2\sqrt{\raisebox{1ex}{$n$}\!\left/ \!\raisebox{-1ex}{$A$}\right.}}=\frac{1}{2\surd D}$$

(3)

where R is the nearest point index, $\overline{r_1}$ is the average value of $r_1$ between the nearest points, $\overline{r_E}$ is the nearest distance assumed, and D is the point density. When R = 1, that is, $\overline{r_1}$ = $\overline{r_E}$, enterprise point factors are randomly distributed. When R > 1, that is, $\overline{r_1}$ > $\overline{r_E}$, enterprise point factors are evenly distributed. When R < 1, that is, $\overline{r_1}$ < $\overline{r_E}$, the point elements have a condensed distribution. A is the area and n is the number of research objects. The nearest neighbor index analysis can be obtained by using Average Nearest Neighbor from the ArcMap 10.8 software.

3.2.3. Point Density Estimation

Density analysis is one of the analysis tools of a point factor set. This study uses point density analysis to identify high-density and low-density areas. This operation can be completed with the Spatial Analyst option of the ArcMap10.8 software. The point density analysis tool was used to calculate the density of point elements around each output grid pixel. Conceptually, a neighborhood was defined around the center of each grid pixel. The density of point elements can be obtained by adding the number of points in the neighborhood and then dividing by the neighborhood area. The formula of point density [20] is as follows:

$$D=\frac{n}{s}$$

(4)

where n is the number of points and s is the area of the field.

3.2.4. Kernel Density Estimation

Kernel density analysis is used to calculate the density of point features around each output grid pixel. Conceptually, each point is covered with a smooth surface. The surface value is the highest at the position where the point is located. As the distance from the point increases, the surface value gradually decreases, and the surface value is zero at the position where the distance from the point is equal to the search radius. Only circular neighborhoods are allowed. The function estimation formula of kernel density [21] is as follows:

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

(5)

$$K\left(\frac{x-x_i}{h}\right)=\frac{3}{\pi }{\left[1-\left(\frac{x-x_i}{h}\right)\right]}^2,{\left(\frac{x-x_i}{h}\right)}^2\le 1$$

(6)

where, $f\left(x\right)$ is the kernel density calculation function of space location in x, d denotes the dimension of space, and h is the distance attenuation threshold (bandwidth). N denotes the number of points whose distance from the location in x_i is less than or equal to h. The K-function represents the spatial weight function.

3.2.5. Standard Deviation Ellipse

Standard deviational ellipse, also known as Lefevre directional distribution, can be used to measure the direction and distribution of point elements, which can be intuitively reflected in the image. In this study, the ArcMap10.8 software is used to perform standard deviational ellipse analysis and visual processing on Zhejiang supply chain innovation enterprises so as to obtain the overall trend development of the enterprise point factor distribution pattern. The algorithm for the standard deviational ellipse is divided into three parts: ① determine the center of the circle; ② determine the rotation angle; ③ determine the length of the XY axis.

First determine the center of the circle. The center of the circle formula is as follows:

$$SD{E}_x=\sqrt{\frac{{{\displaystyle \sum }}_{i=1}^n{\left(x_i-\overline{X}\right)}^2}{n}}$$

(7)

$$SD{E}_y=\sqrt{\frac{{{\displaystyle \sum }}_{i=1}^n{\left(y_i-Y\right)}^2}{n}}$$

(8)

where $x_i$ and $y_i$ are the spatial coordinate positions of each point element, X and Y are the arithmetic mean, and SDE represents the center of the ellipse.

Next, determine the angle of rotation of the ellipse. Based on the X-axis, the 0° position is twelve o’clock, and clockwise rotation represents the angle. The calculation formula is as follows:

$$\tan \theta =\frac{A+B}{C}$$

(9)

$$A=\left({{\displaystyle \sum }}_{i=1}^n\widetilde{x_i}{ }^2-{{\displaystyle \sum }}_{i=1}^n\widetilde{y_i}{ }^2\right)$$

(10)

$$B=\sqrt{{({{\displaystyle \sum }}_{i=1}^n\widetilde{x_i}{ }^2-{{\displaystyle \sum }}_{i=1}^n\widetilde{y_i}{ }^2)}^2+4{\left({{\displaystyle \sum }}_{i=1}^n\widetilde{x_i} \widetilde{y_i}\right)}^2}$$

(11)

$$C=2{{\displaystyle \sum }}_{i=1}^n\widetilde{x_i} \widetilde{y_i}$$

(12)

$\widetilde{x_i}$ and $\widetilde{y_i}$ are the differences between the mean center and the XY coordinates. Finally, the length of the ${\sigma }_x$ and ${\sigma }_y$ are determined by the following formulas:

$${\sigma }_x =\sqrt{2}\sqrt{\frac{{{\displaystyle \sum }}_{i=1}^n{\left(\widetilde{x_i}-\cos \theta -\widetilde{y_i}-\sin \theta \right)}^2 }{n}}$$

(13)

$${\sigma }_y =\sqrt{2}\sqrt{\frac{{{\displaystyle \sum }}_{i=1}^n{\left(\widetilde{x_i}-\sin \theta -\widetilde{y_i}-\cos \theta \right)}^2}{n}}$$

(14)

4. New Industry Pattern of Zhejiang Supply Chain

4.1. Geographical Distribution

In this study, 187 supply chain innovative enterprises in Zhejiang Province were used to obtain the geographical spatial location of enterprises through the latitude and longitude positioning system. The specific operation process is as follows. ① The longitude and latitude of the enterprise point elements are listed in an excel table, and the table was converted into csv format. ② Open the ArcMap 10.8 software, add data to layer. ③ Display X and Y data, add geographic coordinate system to layer. ④ Export data point SHP file. Some examples of enterprise longitude and latitude are showed in

Table 2. Longitude and latitude of some supply chain innovation enterprises.

The Serial Number	Region	The Name of The Company	Longitude	Latitude	Main Business	Level of Enterprise Assets	Gradient Division
1	Hangzhou	Best Logistics Technology (China) Co., Ltd.	120.208683	30.207683	Road transport industry	1949.5 million	6
2	Hangzhou	Hangzhou debang Freight forwarding Co., Ltd.	120.374664	30.339747	Road transport industry	70.0 million	3
3	Hangzhou	Hangzhou Haicang Technology Co., Ltd.	120.171609	30.286448	Science and technology extension and application services	16.4 million	1
4	Ningbo	Ningbo International Logistics Development Co., Ltd.	121.602321	29.874359	Software and information technology services	30.0 million	2
5	Ningbo	Ningbo Hailide import and export Co., Ltd.	121.63492	30.05662	Wholesaling	66.6 million	3
6	Ningbo	Ningbo Shipping Booking platform Co., Ltd.	121.600299	29.862176	Multimodal transport and transportation agency business	32.0 million	2
7	Jiaxing	Haining Zhongyue Electronic Commerce Co., Ltd.	120.386616	30.441158	Retail	18.5 million	1
8	Shaoxing	Shaoxing Port Modern Logistics Group Co., Ltd.	120.637832	30.027286	Road transport industry	261.3 million	4
9	Shaoxing	Zhejiang Zhenong Maoyang Agricultural Products Distribution Co., Ltd.	120.851171	29.752658	Road transport industry	20.4 million	2
10	Jinhua	Hefeng Information Technology (Jinhua) Co., Ltd.	119.58337	29.096271	Information Technology services	7.0 million	2
11	Jinhua	Shangxiang Group Co., Ltd.	120.149396	29.398069	Business services	1000.0 million	6
12	Quzhou	Gallop Holding Group Co., Ltd.	118.613274	28.709293	Wholesaling	50.0 million	3
13	Zhoushan	Zhoushan Port Comprehensive Bonded Zone Commodity Exchange Clearing House Co., Ltd.	122.217977	30.098995	Wholesaling	12.5 million	1
14	Taizhou	Zhejiang Africa International Trade port Service Co., Ltd.	121.434177	28.644715	Retail	50.0 million	5

Note: Enterprise gradient is divided according to the level of enterprise assets.

:

According to the enterprise distribution map, the supply chain innovation enterprises in Zhejiang Province are not evenly distributed geographically, showing the phenomenon of local aggregation. The number of enterprises in different cities varies greatly, with 44 enterprises located in Hangzhou and only 7 in Taizhou. Enterprises are mainly distributed in the east of Hangzhou, Jiaxing, and Huzhou. Because Jiaxing and Huzhou are affected by Hangzhou’s economic radiation, there are many supply chain innovation enterprises there. This difference is affected by the economic level, regional policies, industrial structure, education conditions, and other factors. A further analysis was conducted on the situation affecting the distribution pattern of enterprises.

The enterprise distribution map produced is shown in Figure 2:

The enterprise density distribution is show in

Table 3. Enterprise density distribution.

	Hangzhou	Ningbo	Wenzhou	Jiaxing	Huzhou	Shaoxing	Jinhua	Quzhou	Zhoushan	Taizhou	Lishui
2018	2	15	1	7	18	2	12	5	1	2	4
2019	17	3	14	13	8	6	7	6	9	5	4

:

4.2. Moran Index

GeoDa 1.20 software was used to analyze the Moran index of the global autocorrelation of supply chain innovation enterprises among cities in Zhejiang Province. The index result was 0.341, in an interval of [0, 1]. This shows that there is a significant correlation between supply chain innovation enterprises in different cities, showing a kind of aggregation effect. The specific image is shown in Figure 3:

However, spatial global autocorrelation analysis can only reflect global characteristics; it ignores local relationships. Therefore, this study further conducted a local Moran index analysis. The results of the analysis form a LISA cluster graph, it can be shown in Figure 4.

The display results of the LISA cluster diagram are represented by different colors, in which “insignificant” means that the spatial autocorrelation is not obvious. “High-high” means that the supply chain innovation enterprises have a high development level, and the enterprise innovation degree in the surrounding areas is also high. The high-level areas have radiation and spillover effects on the surrounding areas, and the gap between the surrounding areas is narrowed. It is not difficult to find from the LISA cluster diagram that, from a local aspect, most supply chain innovative enterprises in Zhejiang Province are not significantly clustered, accounting for nearly 70%. The high-high cluster areas are mainly distributed in Quzhou, Wenzhou, and Wenzhou. The low-low concentration areas are mainly distributed in the northeast of Hangzhou and Ningbo. The high- and low-anomaly accumulation areas are mainly in the eastern part of Hangzhou. Low-high anomaly areas are scattered in the south of Zhejiang Province. To sum up, the significance of supply chain innovation enterprises in Zhejiang Province is weak and relatively scattered in the global autocorrelation distribution, but there is a certain significance in the local distribution.

4.3. Nearest Neighbor Index

The ArcMap10.8 software was used to analyze the nearest neighbor point index of point elements, and the statistical data of the nearest neighbor point index were obtained as shown in

Table 4. Nearest neighbor index.

Statistics	Min	Max	Sum	Average	Standard Deviation
Nearest proximity index	0	0.393771	9.762546	0.052206	0.068789

:

The spatial distribution of point elements can be divided into three types: aggregated, uniform, and random. The nearest distance of point elements with uniform distribution is the largest, followed by random distribution and the smallest condensed distribution. Since the interval of R is [0–0.393711] and the maximum proximity index of R is <1, it can be judged that the point factors of innovative supply chain enterprises in Zhejiang Province tend to be clustered rather than discrete. The frequency distribution of the nearest point index is showed in

Table 5. Nearest neighbor exponential frequency distribution.

Index	0.0–0.05	0.05–0.1	0.1–0.15	0.15–0.2	0.2–0.25	0.25–0.3	0.3–0.4
frequency	125	37	10	5	3	3	4

:

4.4. Point Density Estimation

In order to further observe the spatial pattern characteristics of supply chain innovation enterprises, the Spatial Analyst option of ArcMap10.8 software allows for the point density estimation of 187 point factors. The color concentration of the point density is proportional to the aggregation situation, and the aggregation of the decimal point forms the nucleus point, which represents the location of the density center. The estimated figures of point density and kernel density obtained through analysis are shown in Figure 5 and Figure 6. Through the point density and core density estimation map of supply chain innovation enterprises, it can be seen that the supply chain innovation enterprises in Zhejiang Province are dominated by a distribution pattern in which Hangzhou is the strong center and Hangzhou Hujia the multi-center. Secondly, they also show the distribution characteristics of “one main point and multiple points”, and the multi-intensive distribution of the main points in the east of Hangzhou. The distribution of enterprises is mainly concentrated in the northern part of Zhejiang Province. The density of enterprises in Ningbo, Zhoushan, Jinhua, and Wentai is more concentrated, while the density of enterprises in Lishui is more dispersed, which may be caused by the level of economic development and the radiation influence of neighboring cities.

4.5. Standard Deviation Ellipse

In this study, the ArcMap 10.8, was used to draw the standard deviational ellipse map of supply chain innovative enterprises in Zhejiang Province, as shown in Figure 7:

According to the inclining direction of the ellipse, the central axis distribution trend of the enterprise spatial pattern is Jinhua—Hangzhou—Huzhou—Jiaxing, which is generally consistent with the employment density and economic level of each city. The long axis of the ellipse is similar to the Y axis, indicating that the centripetal force is weak. The difference between the short axes is not significant, and the degree of bias is low. The regional point elements in the ellipse are mainly in the north of Zhejiang, especially in the east of Hangzhou. On the whole, the spatial distribution pattern of supply chain innovation enterprises in Zhejiang Province shows a “ladder” trend, and there is a significant gap between Hangzhou, the leading city, and Taizhou, the last city.

5. Results

Combined with the literature review, it is not difficult to find that the clustering of enterprises in spatial pattern distribution characteristics is mainly related to regional GDP. Regional economic level is the foundation of enterprise development. Moreover, economically developed regions can attract more industrial investment and form a good double-cycle structure. The number of people employed in the secondary and tertiary industries can reflect the emerging industrial structure from the side. If more people are employed in the secondary and tertiary industries, employment density increases accordingly. Education level and urbanization rate are also important factors affecting the distribution of innovative enterprises in supply chains, and they are mutually attractive.

Therefore, this study selects the above aspects as influencing factors to analyze the distribution of supply chain innovation enterprises in Zhejiang Province. The table after the data processing is shown in

Table 6. Data on influencing factors of cities in Zhejiang Province.

Region	City GDP in Zhejiang Province	The Number of Secondary Industries	The Number of Tertiary Industries	Employment Density	Education Level	Urbanization Rate
Hangzhou	15,373.05	256.6	406.7	0.04	431,965	0.68
Ningbo	11,985.12	319.1	254.11	0.06	149,804	0.70
Wenzhou	6606.11	269.94	244.61	0.04	88,810	0.56
Jiaxing	5370.32	176.47	134.21	0.07	70,711	0.66
Huzhou	3122.43	94.6	78.74	0.03	26,941	0.57
Shaoxing	5780.74	173.01	129.9	0.04	99,270	0.60
Jinhua	4559.91	162.37	123.56	0.03	77,480	0.51
Quzhou	1573.52	43.44	44.57	0.01	14,413	0.40
Zhoushan	1371.6	28.77	36.27	0.04	23,032	0.55
Taizhou	5134.05	184.22	157.56	0.03	35,216	0.56
Lishui	1476.61	44.67	53.56	0.01	21,008	0.45

Note: Table 1 for specific units.

:

Local univariate Moran index and bivariate Moran index were used to analyze the influencing factors. Through the univariate Moran index analysis of each influencing factor, the index results are shown in

Table 7. Local univariate Moran index.

Moran Index	City GDP in Zhejiang Province	The Number of Secondary Industries	The Number of Tertiary Industries	Employment Density	Education Level	Urbanization Rate
2018	−0.201	−0.118	−0.321	0.299	−0.32	0.130
2019	−0.208	−0.101	−0.317	0.191	−0.226	0.126

:

It can be seen from the precipitation results that among the single factors, only the employment density and urbanization rate are positively related to the distribution of enterprise factor points; that is, the greater the employment density, the wider the distribution of supply chain innovation enterprises, and the higher the urbanization rate, the higher the number of supply chain innovation enterprises. The other four factors show spatial differences with the distribution pattern of enterprises. Due to the correlation between factors, univariate factors may not have a strong relationship; bivariate Moran index was therefore used for further analysis. The first variable and the second variable were selected as GDP and employment density, respectively; others include employment in the secondary industry and employment in the tertiary industry, as well as GDP and the urbanization rate. The results analyzed by the GeoDa 1.20 software were: −0.128, −0.144, −0.136, 0.072, −0.238, −0.047. This shows that the distribution pattern of supply chain innovation enterprises is affected by many factors, especially GDP and employment density.

5.1. GDP of Each City

GDP reflects the economic strength of cities in Zhejiang Province, which can be measured to a certain extent. It directly affects the local productivity level and indirectly affects the spatial layout of local listed companies. Regions with a high economic level can promote the selection and development of enterprises. The level of economic development is also one of the indicators for measuring the market environment for the comprehensive development of enterprises. Cities with a high level of economic development have sufficient government revenue, with more funds being invested in urban infrastructure construction, resulting in better urban infrastructure and public services, which can provide more convenience and greater guarantees for the normal operation of enterprises. Overall, the economic level is positively correlated with the number and agglomeration degree of supply chain innovation enterprises.

5.2. Number of People in Secondary and Tertiary Industries

The number of people in secondary and tertiary industries can represent the proportion of the labor force in the city to a certain extent. At present, the level of economic development is constantly improving, and demand for emerging industries and non-material aspects is increasing. These demands are mainly met by secondary and tertiary industries, and the development of the market promotes the employment of these industries and reflects the comprehensive development ability and level of the region.

5.3. Employment Density

The employment density of a city reflects the employment space and economic development level. A city is a gathering of a large number of employed people. To a certain extent, the employment density can measure the size of a city, and it directly affects the local productivity level, but also indirectly affects the spatial layout of local listed companies. Higher employment density can provide a greater labor force, a good employment situation can attract more employees, and the overall development of cities and markets can increase.

5.4. Educational Level

The level of education reflects the level of talents. The higher the level of education, the greater the value, quantity, and selectivity of regional talents. Talents play a vital role in today’s social development and promote social and economic progress. In particular, supply chain innovation enterprises need a large number of high-level scientific and technological talents. Because of rapid technological change in the era of science and technology, enterprises are required to innovate continuously to meet the requirements of technological renewal in order to remain competitive in the market. High-level talent innovation can bring new technology, new production modes, and new management modes, and can thereby inject new vitality into the growth of enterprises.

5.5. Urbanization Rate

The urbanization rate reflects the objective level elements of the city. The higher the urbanization rate, the stronger the regional inclusiveness and the stronger the infrastructure and government, which is more beneficial to the development of enterprises and future activities. The urbanization rate accelerates the construction of smart cities, promotes high-quality and stable development of industry, brings about the continuous increase in investment, and forms an environment conducive to the survival of enterprises.

6. Conclusions

6.1. Conclusions

Starting from pilot enterprises of supply chain innovation in Zhejiang Province, this study conducted a series of analyses on point elements by collecting point coordinate systems and importing them into the GeoDa 1.20 and ArcMap 10.8 softwares. We reached the following three conclusions.

Firstly, the spatial pattern of supply chain innovation enterprises in Zhejiang Province mainly presents an agglomeration phenomenon, with this agglomeration effect being especially strong in some regions, mainly in the west of Hangzhou and the north of Zhejiang.

Secondly, the overall distribution of innovative enterprises in supply chains in Zhejiang Province is uneven. There are many influencing factors.

Thirdly, the distribution trend of supply chain innovation enterprises in Zhejiang Province is north—south, with the west of Hangzhou and Huzhou serving as the core radiating and dispersing to the surrounding areas. The gap between different cities is large, showing a ladder trend.

6.2. Outlook

This outlook for this study on supply chain innovation enterprises in Zhejiang Province outlook is as follows. Due to operating differences in the innovation of supply chain enterprises in Zhejiang Province, we should adopt the policy of seeking common ground while setting aside differences, fostering strengths, and circumventing weaknesses, combined with taking advantage of features of municipal enterprises, strengthening enterprises’ specialization degrees, forming key enterprises, and constantly improving the market competitiveness of enterprises. According to the evaluation results of the pilot list of the first and second batches of enterprises, it can be seen that the proportion of enterprises with good or better supply chain innovation is not large, mainly because of insufficient investment in Internet intelligence and the lack of a long-term vision from management. Innovation is wisdom. Supply chain enterprises can drive their development to the greatest extent only by combining with the Internet and cloud platforms, thereby realizing the dual cycle of economy and resources [22,23].

Different from the traditional supply chain, the innovative supply chain places significant demands on the leadership of enterprise managers and the informatization ability of internal talents [15]. Therefore, enterprises must attract and gather talents to form internal think tanks through all-round, multi-level, professional scientific and technological personnel, to build a supply chain innovation system. Enterprises should combine with each other to learn from the many excellent supply chain innovation enterprises and achieve mutual benefit and win-win results. It is essential to actively respond to government policies and form a supply chain network system covering the whole province and radiating throughout the whole country. This means integrating outstanding enterprises, creating diversified collaborative and efficient operations of modern supply chain platforms, and being committed to leading the innovative supply chain platform in China [16].

6.3. Limitations

The limitations of this study are mainly reflected in three aspects.

Firstly, the enterprise data from only 2017 to 2019 were used. Because the time span is small, there may be a problem with the generalizability of the results.

Secondly, relevant government departments have evaluated the first and second batches of pilot enterprises in Zhejiang Province and divided them into four levels, but the enterprise distribution pattern after this division was not reflected in this study, which may have affected the enterprise distribution pattern to a certain extent.

Thirdly, the scope of writing is limited to Zhejiang Province.