Spatial Pattern Evolution Characteristics and Influencing Factors in County Economic Resilience in China

Abstract

In the face of shocks, a region’s economic resilience decides whether it can quickly recover or slip into long-term economic stagnation. This study took 2801 counties in China as the research object and distinguished them into long-term and short-term economic resilience by taking 2007–2020 as the research time period, and used spatial autocorrelation, the semi-variance function, and the geodetector method to analyze the spatial evolution pattern and driving mechanism of economic resilience of China’s counties in different time periods. The research found that: (1) From a long-term perspective, the economic resilience of China’s counties was dominated by the moderate level of resilience, and although its characteristics varied slightly over time, the overall performance showed that the level of resilience was increasing. Over time, the number of counties with very high levels of resilience has been increasing, and the number of counties with very low levels has been gradually decreasing. (2) In terms of spatial layout, China’s county economic resilience exhibited spatial autocorrelation, with similar areas clustered and distributed spatially, with high-high concentration (H-H) and hot spot (99% confidence) areas distributed in the eastern coast and its hinterland, and low-low concentration (L-L) and cold spot (99% confidence) areas distributed in Inner Mongolia and the northeast. The evolution of its spatial pattern was influenced by both stochastic and structural factors, and the spatial divergence was mainly reflected in the northeast–southwest direction, while the northwest–southeast direction was more balanced. (3) Long-term economic resilience and short-term economic resilience had different influencing factors. The industrial structure diversification index, which characterized economic factors, could significantly improve the long-term economic resilience of cities, while the influencing factors of short-term economic resilience differed from period to period.

1. Introduction

China’s economy has continued to grow rapidly since the reform and opening up, creating the “Chinese miracle” in the world’s economic history. However, there was an objective law of development for economic growth, and China inevitably encountered bottlenecks and risks, as well as sudden crises and challenges, in order to achieve the transition from the low-income to the high-income stage and to cross the “middle-income trap”. For example, a series of shocks such as the global financial crisis in 2008, the US–China trade war in 2018, and the COVID-19 outbreak so far in 2019 have had a huge impact on China’s sound economic development. Therefore, how to drive China’s economy steadily forward in a complex environment has become the focus of scholarly research.

As China’s economy gradually transforms and upgrades, counties, as the basis of regional economic development, have an increasingly prominent role in driving the regional economy. Due to the vast size of China, different regions have certain differences in resource endowment, industrial structure, and social and cultural backgrounds. After suffering from external shocks, some counties have successfully recovered and regained steady economic growth, while others have been in decline ever since. The differences exhibited by different counties are precisely the role of economic resilience, which has become the key for counties to achieve both quality and efficiency of economic growth, and therefore, it is necessary to conduct an in-depth study on the economic resilience of China’s counties. Unfortunately, however, the existing literature contains few studies on the economic resilience of China’s counties, and even fewer results analyze the spatial characteristics of economic resilience based on a national perspective. As the research object at the middle-level of China, county-level regions can reflect the spatial distribution pattern of economic variables in China in more detail and reflect the development differences existing among different regions. In this context, the central questions of interest in this paper are: How should China’s county economic resilience be measured? What is the magnitude of economic resilience across counties? What are the factors that lead to differences in county economic resilience? In-depth research and discussion of the above issues will not only help explore the path to enhancing the resilience of county economies, but also provide new ideas and perspectives to solve the problem of unbalanced regional development and promote coordinated regional development.

2. Literature Review

The word resilience comes from the Latin root “resilire” and describes the ability of a system to recover in response to shocks and perturbations. Regional economic resilience, which is an important application of the concept of resilience in the socioeconomic field, has become an important theoretical tool for economic geographers to study economic recovery and sustainable development [1,2,3]. A more accurate and accepted definition of regional economic resilience among existing studies comes from Martin et al. [4], which argues that the resilience of regional economies to recessionary shocks should include four dimensions. The first is resistance, which emphasizes the degree of response to shocks, determined not only by endogenous factors in the economy, but also by the size, nature, and duration of the shock. The second is recovery, the ability to recover from a shock and return to its original growth path. The third is reorientation, the ability of a region to adapt to shocks by restructuring industries, technology, labor, and so on. The fourth is renewal, the ability of the region to step into a new growth path through innovation and transformation.

In recent years, empirical studies on regional economic resilience have been enriched, and the research results mainly focus on resilience measurement and exploration of influential factors. Scholars have two main methods for measuring regional economic resilience: the first method is the indicator system measurement method. The method measures the economic resilience of the region under study by constructing a system of assessment indicators through the selection of several indicators and synthesizing a regional economic resilience index. For example, Briguglioet et al. [5] constructed a system of economic resilience indicators in four dimensions: macroeconomic stability, micro market efficiency, economic governance, and social development and thus measured the economic resilience of 86 countries. Using Austrian agriculture as a case study, Quendler et al. [6] measured the economic resilience of Austria over the period of 1995–2019 in terms of six dimensions: market repositioning, value chain optimization, strategic cooperation, innovation upgrading, production migration, and corporate culture. Zhang Mingdou et al. [7] measured the economic resilience of China’s provinces using four indicators, including fiscal revenue and the amount of actual foreign capital used and so on. Indicator system measures have not been recognized as uniform indicators until now, which has led to widely varying results from different studies. Another approach is the unidimensional indicator method, in which scholars usually chose core indicators such as employment rate and GDP to explore regional economic resilience through changes in the core indicators, taking a particular economic shock as an example. Martin et al. [4] measured economic resilience during the UK recession using one indicator, the number of unemployed people. Tan Juntao et al. [8] used GDP indicators and regional sensitivity index methods to measure the economic resilience of each of China’s provinces. Wang Xiaowen et al. [9] applied the employment growth rate to measure the regional economic resilience of 284 cities in China and analyzed the factors influencing regional economic resilience. The second method is now widely used [10,11,12,13,14].

Scholars have explored different aspects of the factors influencing regional economic resilience. Martin [15] proposed a more comprehensive framework for analyzing the factors influencing regional economic resilience in terms of industrial structure, labor force, finance, and systems, and scholars have since conducted empirical studies from resource endowment, the level of specialization in manufacturing, services, and other industries, and policy systems [16,17,18,19,20]. For instance, Davies et al. [21] studied the impact of sectoral structure on the economic resilience of European countries and showed that regions with a high share of the financial sector had significantly more resilience and those with a predominantly manufacturing and construction sector had less resilience. Boschma [22] argued that institutional factors affect regional resilience in two ways: first, technological-industrial diversification, and second, the ability of regions to nurture new growth paths after being exposed to external shocks. The impact of enterprises, as subjects of microeconomic activities, on regional economic resilience has begun to attract the attention of scholars. Wang Chen et al. [23] argued that firm heterogeneity and entrepreneurship were conducive to enhancing regional economic resilience; a more open and diversified environment is more conducive to the recovery of productive activities. In conclusion, scholars have conducted theoretical and empirical studies on regional economic resilience, concentrating on the ability of regions to cope with crises, and have studied the factors influencing regional economic resilience, but these studies have mainly quantified the impact of a particular influencing factor on regional economic resilience, lacking a systematic perspective to analyze the combined impact of multiple factors.

Throughout the existing literature, although there is a wealth of regional economic resilience research results, there are a few studies that address the economic resilience of Chinese counties. In the small literature on China’s regional economic resilience, studies are mainly focused on the national provinces, individual provinces, and regions. These include, for instance, the old industrial base of Liaoning (Li Lianggang et al. [24]), resource-based cities in the northeast (Liu Shiwei et al. [25]), and the Guangdong-Hong Kong-Macao Greater Bay Area (Liu Yi et al. [26]). The results of the regional economic resilience variance analysis differ according to the methodology used, the scale of the study, and the perspective from which the problem is analyzed. In particular, the existing empirical analysis on regional economic resilience mostly focuses on the economic resilience of a certain region, ignoring the spatial impact, especially the comparative analysis of smaller-scale regions, which makes it difficult to accurately reflect the spatial pattern differences, evolutionary characteristics and influencing factors of regional economic resilience. The county is the basic unit of the country to promote high-quality economic development and regional coordination and plays an important role in the government’s macro-control and policy formulation. Therefore, the article selected 2801 county-level regions in China as the main research objects, measured the economic resilience intensity of different regions, analyzed their spatial evolution characteristics, and further explored their main influencing factors, which not only can analyze the ability of each county-level region to cope with potential crises and risks and provide policy suggestions to enhance their regional economic resilience, but can also provide new perspectives and ideas to further solve the problem of unbalanced regional development in China.

3. Research Data and Methods

3.1. Research Area and Data Sources

The research object was all county-level administrative units in mainland China (excluding Hong Kong, Macao, and Taiwan). Since the national administrative divisions have been adjusted and changed several times, to ensure the comparability of the study results, the 2020 county-level administrative divisions published by the Ministry of Civil Affairs of the People’s Republic of China (http://xzqh.mca.gov.cn/map (accessed on 4 February 2022)) were used uniformly. As of December 2020, there were 2844 county-level administrative units in mainland China. We excluded 43 county-level administrative units for which data were not available, leaving 2801 county-level administrative units as the subject of this study. This study has covered the majority of county-level administrative units across the country, with diverse geographical types and differences in the stages and characteristics of county economic development. Influenced by data availability, the study unit we chose for the impact factor analysis was 1906 counties (cities and districts) with relatively complete data. The statistical data were mainly obtained from “China statistical yearbook for regional economy”, “China city Statistical Yearbook”, “China County Statistical Yearbook”, provincial statistical yearbooks, yearbooks of some provinces and prefecture-level cities, as well as statistical bulletins on national economic and social development of prefectures, counties, and regions in each year. Meanwhile, we processed some of the data: (1) In the case of the unit of the removal of counties and the establishment of districts during the study period, we have included the original jurisdictions in the newly established urban areas for statistics, and in addition, the areas where the administrative divisions have changed are counted according to the adjusted divisions. (2) Some county-level data, which we obtained based on growth rate projections. The study period was from 2007 to 2020, totaling 14 years.

3.2. Research Methodology

In the regional economic resilience measurement method, this paper used the maximum and minimum values of GDP growth rate for calculation and reflected the economic resilience through the fluctuation of the GDP growth rate. Spatial autocorrelation can reveal the phenomenon of spatial dependence and spatial heterogeneity of data. This paper used this methodology to analyze whether there was spatial aggregation and dispersion in the differences of economic resilience of China’s counties in space. The semi-variance function is a basic method to describe the stochasticity and structure of regional variables. In this paper, we used this method to analyze the status of economic resilience differences in China’s counties, including from the east, south, west, and north in each direction. The geodetector method as an influence factor analysis method is widely applied to urbanization, economic growth, and other socioeconomic fields, and the geodetector model is less constrained in terms of assumptions than traditional statistical methods are. For this paper, the geodetector method was applied to analyze the factors influencing the differences in economic resilience of China’s counties.

(1)

Regional economic resilience measurement methods

In terms of short-term economic resilience measures, unlike many developed Western cities that experienced absolute quantitative declines in GDP or employment when faced with shocks, the economy and employment across China have rarely experienced negative growth in the 21st century, thus requiring an adjustment in the measurement of short-term economic resilience. Most of the existing studies on China’s economic resilience have borrowed from measures proposed by foreign scholars [4,15,27], and there are some shortcomings: first, studies using panel data, which are calculated annually for each region to obtain an economic resilience index, are not consistent with the connotation of economic resilience (since shocks cannot exist every year), and therefore these formulas are only applicable to the period before and after a shock. Second, this calculation reflected the economic resilience before and after a particular shock, ignoring the long-term economic “slow burn”. In order to address the above issues, we combined the existing scholars’ discussion of China’s economic development stages [28,29] with the research results of Peng Rongxi et al. [30] and divided the period of 2007–2020 into three stages (short-term economic resilience). The first period was from 2007 to 2009. China’s economic development was also affected by the financial crisis in 2008, and the GDP growth rate dropped. Therefore, the period of 2007–2009 was classified as a resistance period. The second period was 2009–2017. We classified 2009–2017 as the recovery and adjustment period, mainly considering the “the 4-trillion-yuan stimulus package” adopted by the Chinese government to promote economic recovery in the face of the financial crisis, as well as the “new normal” proposed in 2014 and the “high-quality” development proposed in 2017 to promote the optimization and upgrading of industrial structure and economic transformation and adjustment. The third period was 2017–2020, which was also a period of resistance. The main consideration was the impact of the US–China trade war in 2018 and the COIVD-19 outbreak in China so far in 2019. Long-term economic resilience refers to the resilience of the economy for 14 years from 2007 to 2020. It is important to note that, unlike economics, which distinguishes between the long-run and the short-run in terms of factor variability, this paper focused on the shock cycle perspective, calling periods within a shock cycle “short-run” and periods spanning multiple shock cycles “long-run” [30].

The equation to calculate the short-term toughness is as follows:

$$\begin{array}{l}sRE{S}_{iT}=\left\{\begin{array}{l}{\mathrm{G}}_{{}_{\mathrm{i}},\min ,T}-{\mathrm{G}}_{{}_{\mathrm{i}},\max ,T}\hspace{1em}\hspace{1em}\hspace{1em}\hspace{1em}\mathrm{the} \mathrm{esistance} \mathrm{period}\\ {\mathrm{G}}_{{}_{\mathrm{i}},\max ,T}-{\mathrm{G}}_{{}_{\mathrm{i}},\min ,T}\hspace{1em}\hspace{1em}\hspace{1em}\hspace{1em}\mathrm{the} \mathrm{recovery} \mathrm{and} \mathrm{adjustment} \mathrm{period}\end{array}\right.\\ \end{array}$$

(1)

where sRES_iT refers to the short-term economic resilience of region i in time period T. The larger the value, the stronger the resilience. G_i,max,T refers to the maximum GDP growth rate of region i in time period T. G_i,min,T refers to the minimum value of the GDP growth rate in region i in time period T. For each county, due to factors such as geographical distance and economic upstream and downstream links, the year when the maximum shock arrives may be different, and there were differences in the speed of economic recovery. In order to avoid the trouble caused by using the same year growth rate, this article used the maximum and minimum values of the GDP growth rate in a certain period of time for calculations. It should be noted that in Equation (1), both the resistance period and the recovery period were calculated as the GDP growth rate in the last year of the resistance period or the recovery period minus the GDP growth rate in the first year. The principle of the calculation was to reflect economic resilience by calculating the fluctuation of the GDP growth rate during the resistance and recovery periods. With a view to facilitate inclusion in the same model for calculation: in the resistance period, the economy grew at a decreased rate, and resistance resilience was expressed as a negative value, and in the recovery period, the economy grew at an accelerated rate, and recovery resilience was expressed as a positive value.

The formula for calculating long-term economic resilience is as follows:

$${IRES}_{iT}=\sqrt{\frac{t-1}{{\displaystyle {\sum }_{\tau }(G_{i\tau }-{\bar{G}}_{i\tau }}{)}^2}}$$

(2)

where IRES_iT refers to the long-term economic resilience of region i at time period T; t refers to the total number of years in time period T, t > 1. Long-term economic resilience is the economic resilience that includes multiple years, so the long-term economic resilience measure is meaningful when t > 1. G_iτ refers to the GDP growth rate of region i in year τ. G_iT refers to the average of the economic growth rate of each year in time period T in region i. The GDP growth rate of a given year is used, minus the average of the GDP growth rates of all years, to judge how the GDP growth rate fluctuates in that year. A higher value of IRES_iT indicates more economic resilience. This method did not require the identification of the year in which the shock occurred and took into account the “slow burn”factor in the economic development process.

(2)

Spatial correlation and heterogeneity analysis methods

Spatial autocorrelation. Spatial autocorrelation is a measurement of the degree of clustering of attribute values of spatial units, and there are mainly global spatial autocorrelation and local spatial autocorrelation. The former is generally measured using Moran’s I index (Equation (3)), and the latter is often measured using hotspot analysis (Equation (4)) [31].

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

(3)

$$s^2={\displaystyle \sum _{i=1}^n{\left(x_i-\overline{x}\right)}^2}/n$$

(4)

I is the Moran index, W is the spatial matrix, n is the number of regions, and x_i and x_j are the observations in regions i and j, respectively. x is the mean value of the observations, and w_ij is the corresponding spatial weight value. The value of I lies between [−1,1], I < 0 indicates a negative correlation, I = 0 indicates no correlation, and I > 0 indicates a positive correlation. Moran’s I was calculated, and the results were subjected to statistical tests, generally using the Z test.

Semi-variance function. The semi-variance function captures the state of variability and distance variation within the region [32]. The formula is:

$$\gamma (h)=\frac{1}{2N(h)}{\displaystyle \sum _{i=1}^{N(h)}{[Z(x_i)-Z(x_i+h)]}^2}$$

(5)

N (h) is the sample amount of segmentation distance h; Z (x_i) and Z (x_i+h) are the county economic resilience values of Z (x) on spatial cells xi and x_i+ h, respectively. The semi-variance function is mainly composed of five parameters: nugget value (C₀), sill value (C₀ +C), partial base stations ©, variation (a), and nugget coefficients (C₀/(C₀ + C)). With h as the horizontal coordinate and γ(h) as the vertical coordinate, the semi-variance function can be plotted, and these five main parameters can be characterized by spatial variation in the plot. The nugget value (C₀) indicates the discontinuous variation when the regionalized variable is smaller than the observed scale, the sill value (C₀ + C) indicates the smooth value of the semi-variance function variable as the spacing increases to a certain scale, and the variation range (a) is the spacing when the semi-variance function reaches the sill value. The nugget coefficient (C₀/(C₀ + C)) can be used to characterize the spatial variation, and the larger the value is, the more spatial variation is caused by random components, otherwise, it is caused by specific geographic processes or multiple processes combined. The theoretical semi-variance functions are unknown, and it is possible to fit them by computing the values. The commonly used semi-variational fitting models are Gaussian, Exponential, Spherical, Linear, etc.

The fractal dimension (D) is another important parameter of the semi-variance function, whose value is determined by the relationship between y (h) and the interval distance h.

$$2\gamma (h)=h^{(4-2D)}$$

(6)

The fractal dimension (D) reflects the intensity of spatial variability, and the larger its value, the higher the spatial heterogeneity caused by structural factors, and conversely, the greater the spatial heterogeneity caused by random-type factors. The closer its value is to 2, the more balanced the spatial distribution.

(3)

Influence factor analysis method

The core idea of the factor detector of the geodetector method is to explain whether an influencing factor has a significant impact on the change of a geographical thing by comparing whether there is a significant consistency between the change of a certain influencing factor and the geographical thing in space. Therefore, this study used the geographical detector method to analyze the influence of the internal driver X of county economic resilience on Y, the county’s economic resilience. The calculation equation is as follows [33]:

$$q=1-\frac{1}{N{\sigma }^2}{\displaystyle \sum _{h=1}^L{N}_h{\sigma }_h^2}$$

(7)

q is the influence of influence factor X on county economic resilience, N and σ² are the sample number and total variance, respectively, and N_h and σ²_h are the type h (h =1, 2, …, L) of factor X within the sample numbers and variances. The value of q ranges between [0, 1], and the closer the value of q is to 1, the higher the degree of influence of factor X on the county’s economic resilience.

4. Result

4.1. Spatial Evolution Pattern of Economic Resilience in Chinese Counties

4.1.1. Analysis of Spatial Overall Characteristics

(1)

The number of high resilience counties was increasing and the number of low resilience counties was decreasing.

Based on the economic resilience measurement method above and the division of the different stages of the economy from 2007 to 2020, the short- and long-term economic resilience values for China were calculated. Using the natural breakpoint method of ArcGIS software, the economic resilience of counties was divided into five levels (very low, low, moderate, high, and very high level), and the number of counties at different levels was calculated for each time period (

Table 1. Number and proportion of counties with five levels of economic resilience.

Time Period	Very Low Level	Low Level	Moderate Level	High Level	Very High Level
2007–2009	869 (31.0%)	1029 (36.7%)	583 (20.8%)	288 (10.3%)	32 (1.1%)
2009–2017	400 (14.3%)	1179 (42.1%)	736 (26.3%)	205 (7.3%)	281 (10.0%)
2017–2020	352 (12.6%)	400 (14.3%)	991 (35.4%)	723 (25.8%)	335 (12.0%)
2007–2020	525 (18.7%)	652 (23.3%)	1078 (38.5%)	459 (16.4%)	87 (3.1%)

). Looking at the time periods, China’s overall economic resilience was not strong in 2007–2009, with low resilience counties dominating, numbering 1029, or 36.7% of all counties, while only 32 counties had very high resilience. It indicated that with the deepening of economic globalization, China’s economy was deeply embedded in the globalization network and external economic fluctuations, especially under the impact of the global financial crisis, have affected China’s economy to a certain extent. From 2009 to 2017, China’s economic resilience has been increasing, as shown by a decrease in the number of very-low-resilience counties and an increase in the number of very-high-resilience counties. In 2017–2020, despite the impact of the US–China trade war and the COVID-19 epidemic, China’s economy remained strongly resilient, with the number of very-high-resilience counties further increasing to 335, accounting for 12%, and the number of very-low-resilience counties decreasing to 352, accounting for 12.6%. This was due to the fact that since the 19th National Congress of the CPC, China’s economy has gradually shifted from a high-speed growth stage to a higher-quality development stage, with continuous innovation in macroeconomic control policies and continuous optimization of industrial structure, and China’s ability to cope with fluctuations in economic growth has been improving. In terms of long-term economic resilience from 2007 to 2020, China’s county economies were dominated by moderate resilience, with the number of 1078 and the proportion of 38.5%.

(2)

High resilience counties were concentrated on the east coast, and low resilience counties were contracting in spatial distribution.

Viewed from the spatial pattern, from 2007 to 2009, the distribution of very-high-resilience counties was relatively scattered, while the distribution of very-low-resilience counties was relatively concentrated, mainly in northwest and southwest China and in the middle reaches of the Yellow River (Figure 1a). During 2009–2017, spatially very-high-resilience counties showed a concentration of very-high-resilience counties to the Central Plains urban agglomeration and the eastern coast, while very-low-resilience areas in the northwest showed a continuous contraction, and Liaoning, which was dominated by heavy industry, generally showed low economic resilience (Figure 1b). As a result of the structural adjustment period, the Supply-side Structural Reform continues to deepen, cutting a part of the excess capacity in the old industrial base of northeastern China, which in turn has led to a lack of development momentum and therefore relatively low economic resilience. The spatial divergence characteristics were further revealed in 2017–2020 (Figure 1c), and very high resilience value areas were concentrated in Yangtze River Delta, Pearl River Delta, and Yangtze River Economic Belt regions. This was related to the implementation of national development strategies in recent years, such as the coordinated development of the Beijing-Tianjin-Hebei region, the strategy for integrated development of the Yangtze River Delta, the Yangtze River Economic Belt, the Guangdong-Hong Kong-Macao Greater Bay Area, and other national regional development strategies. It was because of the significant favorable national policies, superior location conditions, and good economic foundation that made these regions more capable of coping with external shocks and disturbances. The very-low-resilience-value area further contracted, but still formed three major agglomerations, including the northeast region, northern Inner Mongolia, and the border area between Tibet and Qinghai. In terms of long-term economic resilience, moderately resilient counties were widely distributed (Figure 1d), and very-low-resilience counties were still concentrated in the northeast, northern Inner Mongolia and Qinghai, Tibet, and other regions. This was due to the economic development model in northern China (especially in the northeast), which relied mainly on resource factors and policy-driven economic development after the reform and opening up, as well as the long-term planned economy, resulting in insufficient endogenous impetus for market-oriented reforms, a single industrial structure and difficulties in economic transformation and upgrading. Moreover, the institutional environment in these regions was rigid, the innovation consciousness of enterprises was weak, and historical institutional factors led to the tendency of enterprises to obtain support from national policies and financial transfers rather than actively exploring new growth paths to obtain development, so in the face of external shocks and disturbances, the adaptive capacity of these regions and enterprises was relatively poor and the economic resilience was low. The very-high-resilience and high-resilience counties were concentrated in the southeastern coastal region and its hinterland, as well as in and around the provincial capital cities, the Central Plains urban agglomeration, the Changsha-zhuzhou-xiangtan Urban agglomeration, and other regions. The siphon effect, together with the administrative rank of the city and its associated resource allocation capacity and fiscal revenues and expenditures, may also have some impact on economic resilience.

4.1.2. Characteristics of Spatial Association Pattern and Spatial Divergence Pattern

(1)

The economic resilience of China’s counties exhibits spatial autocorrelation, with similar regions clustered and distributed in space.

Using ArcGIS software, the global Moran I index of China’s county economic resilience from 2007 to 2020 was calculated, focusing on the long-term resilience of county economies. China’s economic resilience Moran I for 2007–2020 was 0.1186 and passed the Z-statistic test, indicating that the spatial distribution of China’s county economic resilience was not completely random. Counties with close economic resilience levels had obvious spatial clustering phenomenon in the region. That is, counties with high levels of economic resilience were close to other counties with high levels of economic resilience, and counties with low levels of economic resilience tended to be close to each other. Judging from the results of local autocorrelation and hot and cold spots analysis, the economic resilience of China’s counties showed significant east–west zonal variability in space (Figure 2). The LISA map (Figure 2a) revealed that the H-H agglomerations of county economic resilience were distributed in the eastern coastal region and its hinterland, and the L-L agglomerations were distributed in northeast China and Inner Mongolia. H-L agglomerations and L-H agglomerations presented a scattered distribution, with scattered distribution in northern China, Tibet, and eastern Xinjiang, indicating a differentiated distribution pattern of high and low agglomeration areas of county economies. Hot spot (99% confidence) areas were distributed in the eastern coast and its hinterland areas, concentrated in Zhejiang, Jiangsu, Fujian, Jiangxi, Hunan, and Guangdong. The hot spot (95% confidence) area was mainly distributed in Anhui, Henan, Hubei, and other places. The cities in the cold spot (95% confidence) zone were mainly located in Inner Mongolia, Qinghai, Xinjiang, and Tibet. The cold spot (99% confidence) area was concentrated in Inner Mongolia and Northeast China, which was consistent with the LISA map results. The reason for this was that the counties in the eastern region had the advantages of location and policy on the one hand, and on the other hand, they continued to improve infrastructure, upgrade public services, and vigorously promote industrial transformation and upgrading, and their economic resilience continued to improve. The counties in Northeast China and Inner Mongolia have strong resource dependence, heavy industrial structure, with difficulties in transforming the county economy and many bottlenecks in sustainable urban development. With vigorous national financial investment and policy support over the years, although the economic resilience value of counties in provinces such as Heilongjiang, Liaoning, Jilin, and Inner Mongolia has increased significantly, it was still low compared to that of the eastern coastal region.

(2)

Spatial differences in the economic resilience narrowed overall, but the variability increased in the south–north and northeast–southwest directions.

A semi-variance function was used to examine the evolution mechanism of the spatial pattern of economic resilience in Chinese counties at different time periods. We selected the county economic resilience for each time period as a spatial geographic variable and calculated the semi-variance function values for each time period separately with the help of GS+9 software. A model with a good fit was used to fit the sample data points for estimation, and the fractal dimension in different directions were calculated separately for each time period and interpolated by Kriging.

As can be seen from

Table 2. Fitting parameters of the spatial variation function of economic resilience of Chinese counties in different time periods.

Time Period	Variation α	Nugget Co	Sill C+Co	Nugget Coefficient Co/C+Co	Fitting Simulation	Fitting Coefficient R2
2007–2009	13.340	0.318	0.183	1.738	Gaussian	0.647
2009–2017	28.971	0.758	0.838	0.905	Gaussian	0.686
2017–2020	29.465	0.783	1.468	0.533	Gaussian	0.742
2007–2020	56.233	0.869	1.522	0.571	Gaussian	0.919

, the nugget and sill values for 2007–2020 were overall seen to be higher than those for other time periods, indicating that the degree of spatial divergence in the economic resilience of Chinese counties was more pronounced from a long time scale. In the short-term, the nugget and sill values of the other three time periods kept increasing, while the nugget coefficient kept decreasing, showing that the structural divergence caused by spatial correlation ketp increasing in the gradually increasing spatial differences of China’s regional economic resilience. The increasing parameter of the variation α reflected that the scope of the spatial effect of economic resilience of Chinese counties was gradually expanded, the scope of the spatial correlation effect caused by structuring was gradually expanded, the intraregional economic resilience was gradually influenced by the core region, the spillover effect was significant, and the strength of the correlation effect between counties was continuously increased. The fitted optimal model for each time period was a Gaussian model, suggesting that the spatially divergent evolution of China’s economic resilience exhibited good continuity and stability. The fit coefficients of the model were generally on an upward trend, which indicated that the spatial self-organization of China’s economic resilience was increasing year by year.

As shown by the fractional dimension of the spatial variance function (

Table 3. Fractal dimension of the spatial variation function of economic resilience of Chinese counties in different directions.

Time Period	Omnidirectional		South–North (0°)		Northeast–Southwest (45°)		East–West (90°)		Southeast–Northwest (135°)
	D	R2	D	R2	D	R2	D	R2	D	R2
2007–2009	1.854	0.508	1.989	0.001	1.801	0.672	1.818	0.161	1.813	0.398
2009–2017	1.863	0.671	1.883	0.401	1.759	0.639	1.778	0.597	1.926	0.398
2017–2020	1.882	0.711	1.815	0.022	1.619	0.607	1.867	0.323	1.962	0.213
2007–2020	1.893	0.804	1.820	0.678	1.729	0.734	1.957	0.590	1.978	0.044

), the omnidirectional fractional dimension kept increasing and fitting better, but there was a certain deviation from the ideal fractional dimension 2, and the spatial pattern of county economic resilience was out of the ideal pattern of mean distribution. When refined to different directions, the degree of fit of the north–south fractal dimension first increased and then decreased, but the fractal dimension roughly tended to decrease, indicating that the variability of the spatial pattern of economic resilience increased. In the northeast–southwest direction, the fractal dimension fitted relatively well, but the value of fractal dimension was lower relative to those of other dimensions, meaning that the economic resilience of the northeast–southwest direction counties was the weakest in homogeneity and had significant spatial variation. The fractal dimension of the east–west direction presented slight fluctuations in the fractal dimension values at each time period, reflecting a small but fluctuating spatial pattern of county economic resilience variation. The northwest–southeast directional fractal dimension value kept increasing and gradually approached 2, and the county economic development tended to be balanced. The reasons for this spatial deviation were, on the one hand, the vast territory of China and the great differences between counties in terms of location conditions, historical reasons, resource endowments, economic development bases, industrial structures, policies and cultures, and so on. Due to its absolute leading position within China in terms of economic strength, industrial structure and innovation effectiveness, the resilience value of the eastern region was higher than that of other regions, indicating that the eastern region was better able to cope with uncertain risks. The central and western regions were actively undertaking the industrial transfer from the eastern regions and continuously improving the industrial structure. Together with the “The Belt and Road” initiative, “Yangtze River Economic Belt”, “The Development of the Western Region in China“, and other related strategies, the regional economic resilience was also enhanced. The old industrial areas in Northeast China were affected by the history and the single industrial structure, which prolonged the time of transferring existing resources to new industries and further restricted the transformation and upgrading of industrial structure, and fell into a “locked” state without finding a new economic growth path, so the economic resilience was low and the gap with other regions gradually widened.

Figure 3, based on the results of isotropic variance fitting and anisotropic kriging interpolation, showed that the spatial pattern of economic resilience of China’s counties had a certain pattern during the period of 2007–2020. It exhibited a clear concave area without a clear peak during the shock periods of 2007–2009 and 2017–2020, indicating that the overall absolute values did not differ significantly during the shock periods, although there was spatial heterogeneity. The spatial pattern of county economies from 2009 to 2017 was largely a plain structure, with spatial variation highlighted, and the faster-growing Central Plains urban agglomeration as well as the middle reaches of the Yangtze River urban agglomeration began to rise, with most counties still on the same plane. Long time scales revealed that the spatial pattern of economic resilience evolved into a high-low structure, with deepening variation in the spatial structure of economic resilience and multiple peaks. These results suggested that China formed a pattern of multiple core cities to drive the common development of surrounding counties, and the spatial structural differences between counties were still very significant.

4.2. Analysis of Influencing Factors

4.2.1. Selection of Influencing Factors

The formation and evolution of the pattern of spatial differentiation of economic resilience differences in counties was driven by a combination of factors within the economic system and factors outside the system. In this thesis, through a systematic review of the existing literature [34,35,36,37,38,39,40], while following the principles of scientific and operability of data, five variables were selected as indicators affecting the economic resilience of Chinese counties. (1) The strength of government support. A healthy economic development cannot be achieved without the government’s support in economy, policies, and other aspects [41]. The financial solvency of local governments is an important basis for promoting economic recovery from external shocks, and good financial solvency is conducive to pulling the economy to a full recovery, as well as generating new growth poles by increasing investment in research and technology and stimulating new economies. We chose the share of public expenditure in GDP (X1) to indicate the strength of government support. (2) Economic base. The better the economic base, the more factors available for the region to call upon in the face of external shocks, and the stronger the resistance, resilience, and adaptability of the region as a whole. The economic base was expressed by the GDP per capita of each county (X2). (3) Industrial structure. The impact of industrial structure on economic resilience may vary at different stages of development [42]. In the early stage of economic development, the moderate concentration of specialized industrial structure is conducive to the rapid accumulation of factors, which in turn strengthens the economic foundation and enhances economic resilience. However, in the middle and later stages of economic development, it will inhibit the improvement of economic resilience. Diversified industrial structures are effective in spreading risks, and such regions can demonstrate better resilience in the face of shocks. The industrial structure was characterized by the industrial structure diversification index (X3), where the industrial structure diversification index was calculated using the entropy method proposed by Frenken et al. [43]. The formula is $VAR={\displaystyle {\sum }_{i=1}^n{p}_{i}ln(1/p_i)}$, where VAR is the entropy index of diversification, and a larger value represents a higher degree of diversification; n is the number of industrial sectors in the economic system; p_i is the proportion of employees in sector i in the total number of employees in the region, and this study used data on the proportion of employees in the three industrial sectors to measure VAR. (4) Social factors. Differences in population number and structure also affect economic growth and consumption levels, thus causing fluctuations in economic growth. An abundant labor force, especially high-quality human resources, can provide talent support and intellectual support for county economic innovation and industrial restructuring. In addition, a certain size of population implies high consumption potential and market potential, which can stimulate economic growth, recovery, and adjustment to some extent. Social factors were reflected by population density (X4). (5) Financial environment. Finance plays a key role in stabilizing market expectations, enhancing the resilience of economic development and keeping the economy running in a reasonable range. The recovery and adjustment of the economy, the reallocation of production resources, and the transformation and upgrading of the industrial structure are inseparable from the support of the financial sector. The financial environment was indicated by the proportion of loan balance of financial institutions to the GDP of the county in which they are located (X5).

4.2.2. Analysis of the Main Driving Factors

Factor detection is mainly used to analyze the extent to which different influencing factors explain differences in county economic resilience. In this paper, the mean values of the indicators for 2007–2009, 2009–2017, 2017–2020, and 2007–2020 were used as explanatory variables for different time periods, and the county economic resilience values (Y) for each time period were used as explanatory variables to analyze the driving factors of the differences in the economic resilience of Chinese counties. We first classified the continuum detection factors using ArcGIS natural interruption point hierarchy, and then calculated the influence of each detection factor on the economic resilience of China’s counties in each time period separately (

Table 4. Results of probing the resilience level factor of Chinese county economies.

Variable	2007–2009		2009–2017		2017–2020		2007–2020
	q-Value	p-Value	q-Value	p-Value	q-Value	p-Value	q-Value	p-Value
X1	0.644	0.086	0.502	0.124	0.482	0.443	0.258	0.602
X2	0.631	0.000	0.621	0.035	0.618	0.073	0.633	0.008
X3	0.620	0.000	0.616	0.004	0.701	0.004	0.785	0.005
X4	0.575	0.046	0.624	0.022	0.634	0.040	0.626	0.019
X5	0.481	0.308	0.474	0.699	0.351	0.780	0.572	0.177

) to study the strength of the influence of different factors on the economic resilience of counties. A larger q value indicates the stronger explanatory power of the independent variable X on the dependent variable Y.

In general, each factor had a certain influence on the economic resilience of China’s counties, and the validity and intensity of the effect of these factors varied at different stages. Looking at the long time period (2007–2020), the q-value of the industrial structure diversification index (X3) was in the first place and was the strongest factor influencing the resilience of the county economy, which confirmed the view of Xu Yuan et al. [12] and Lin Geng et al. [44]. Industrial diversification can promote rapid economic development, provide diversified driving force for economic growth, and effectively disperse or absorb crisis risks after suffering from external shocks. Therefore, future economic development should focus on the development of diversified industrial structure, especially to improve the level of related diversification, and reduce the economic volatility brought by uncertain spillover factors through technology linkage and knowledge spillover. X2 (per capita GDP) ranked second, indicating that the economic base has a greater impact on county economic resilience, especially as China’s economy shifts from high growth to high quality development, the development of county economies should focus more on the quality, structure and efficiency of economic development rather than economic scale. Population density (X4) ranked third, suggesting that the richer the labor force and the higher the population density, the more it helps to withstand external shocks and disturbances, and the more it helps the regional economy to recover. The financial environment (X5) ranked fourth and government support (X1) had the smallest q-value and ranked last, showing that the driving role of government support in the long run in the differences of county economic resilience all declined.

In the short term, the main factors affecting economic resilience differ from period to period. The impact factor of government support (X1) was the largest in 2007–2009, while the q-values of government support were relatively low in the other two periods. The loan balance of financial institutions (X5) ranked relatively low, while compared to other time periods, this period of 2007–2009 has the largest q-value, which is due to the global financial crisis, with a certain impact on the China’s economy, and the financial sector was the first to be hit. From 2009 to 2017, the population density (X4) ranked first and the q-value showed an overall upward trend, reflecting from the side that strengthening labor concentration is conducive to enhancing economic resilience. Labor concentration can promote the sharing of market, talent, and technology resources, which is conducive to the improvement of labor productivity and enhance economic resilience. In 2017–2020, industrial structural diversity had the deepest impact on economic resilience, and its role in economic stabilization and recovery has become more evident after the impact of the US–China trade friction and the COVID-19 epidemic.

5. Discussion and Suggestions

5.1. Discussion

Spatial differences in the economic resilience of China’s counties showed an overall decreasing trend, but the differences varied in different directions. The spatial pattern of economic resilience in the county has less variation in the east–west direction, but more variation in the north–south direction. For one thing, this may be due to the long-standing coordinated regional development strategy of the China’s government, which has implemented policies to take care of the western region and invested a lot of human and material resources, resulting in the continuous improvement of infrastructure, rapid economic development and increased resistance to risks in the western region. For another, as the western region is less market-oriented than the eastern region and less exposed to external shocks, these may contribute to the relatively balanced economic resilience of the western counties. In the Northeast, the economy is less resilient, probably caused by the long-term planned economy resulting in insufficient endogenous impetus for market-oriented reforms, a homogeneous industrial structure and difficulties in economic transformation and upgrading, and therefore its relatively poor adaptive capacity and low economic resilience when faced with external shocks and disturbances. Although the Northeast region has achieved some success in economic development with the promotion of the Northeast Revitalization Strategy, the level of economic resilience of the county remains low over time. In contrast, the eastern and southern regions, especially the eastern coastal region, have vigorously developed an export-oriented economy through market-oriented reforms, with high market dynamism, a diversified industrial structure and a strong ability to adapt in the face of a complex external environment.

The causes of the differences in the resilience of China’s counties at different stages of development are also different. Comparing the influencing factors at different stages can be seen: from a long time period, the industrial structure diversification index was the strongest factor influencing the resilience of counties, which may be due to the fact that external economic shocks generally affect one or a few industries directly, while a diversified industrial structure can effectively spread the risk, so such areas can reflect better resilience in the face of shocks. The effect of location is the weakest, probably attributed to the regional coordination policy implemented by the government of China in recent years, with the construction of new transport infrastructure such as railways and roads, and the increasing connectivity of the regions, so that the influence of location in economic resilience is becoming less pronounced. For example, since the 18th National Congress of the Communist Party of China, the Chinese government has placed greater importance on promoting coordinated regional development, and has proposed and implemented many major strategies such as the construction of “The Belt and Road”, the coordinated development of Beijing, Tianjin and Hebei, the development of the Yangtze River Economic Belt, the construction of the Guangdong-Hong Kong-Macao Bay Area, the integrated development of the Yangtze River Delta, and the ecological protection and high-quality development of the Yellow River Basin. At different stages, the dominant factors affecting the economic resilience of China’s counties were different. The period of 2007–2009 was when the location factor had the greatest impact, this was as a result of the fact that during this period, government infrastructure investment was not yet strong, and the infrastructure-driven effect was not yet visible. This period was also heavily influenced by the financial environment factor, probably influenced by the international financial crisis. The period of 2009–2017 was the most influenced by population density, probably caused by the Chinese government’s investment of four trillion to rescue the market after the financial crisis, implementing measures including the construction of infrastructure such as railroads and roads, and ecological investment as a way to boost economic growth and increase the demand for labor As China proposed the “new normal” in 2014, the industrial structure has been continuously adjusted and the demand for labor force structure has also changed accordingly, and the impact on economic resilience has also changed. In the period of 2017–2020, it was the diversification of the industrial structure that had the greatest impact, especially after the impact of the “new crown epidemic”. This was because COVID-19, as a public safety emergency, had a significant impact on economic development, and regions with a diversified industrial structure are better able to share the risk and recover quickly from the shock.

It should be noted that there are some limitations in the analysis and research of this study.In terms of research methodology, this study measured the economic resilience of China’s counties using the “short-term economic resilience” and “long-term economic resilience” measures. The study method is simple to operate and easy to generalize. However, it has certain shortcomings: firstly, the method can only perform comparative analysis before and after the shock, and is not very predictive of the future. As a method to cope with future changes, improving the effectiveness of forecasting is something that should be considered in future research. Secondly, the calculation method only focused on the core indicator of GDP, and reflected the county’s economic resilience by measuring the fluctuation of GDP growth rate in a certain period. There are limitations in the choice of variables and insufficient examination of the heterogeneity of economic resilience, while it is easy to ignore the multifaceted shocks to the economy after the occurrence of shocks. In the future, the differences in economic systems should be considered in an integrated regional economic-social-ecological system. Thirdly, we have classified “short-term economic resilience” and “long-term economic resilience” by the timing of shocks, major events and policies, and the state of China’s economic growth, but we have not explored the correlation between the two in depth due to the limitation of data acquisition. Thus, in future research, the measurement method of economic resilience should be further explored, and the relevance of “short-term economic resilience” and “long-term economic resilience” should be further verified by monitoring several key indicators with the help of big data. Fourthly, regional economic resilience is a dynamic evolutionary process, and the method can achieve horizontal comparison of different objects at the same stage, but there are still shortcomings in the comparative analysis of time series vertically. In addition, economic resilience may vary across shocks and stresses and across county types, and specific shock types and county types are not considered in this study. As for the future research, acute shocks and chronic stresses should be taken into account, while combining different counties and exploring their economic resilience to further expand the analytical methods and empirical scope of county economic resilience.

5.2. Suggestion

China is a vast country, and different counties have different resource endowments and development bases. Therefore, the adoption and implementation of development measures should not be a one-size-fits-all approach, but should be based on the actual situation, taking into account their own characteristics to enhance the economic resilience of counties.

At first, promote coordinated regional development. The government should strengthen economic exchanges between counties, coordinate cooperation between regions across administrative units based on the level of economic resilience and spatial differences characteristic of different counties, improve inter-regional infrastructure construction, and promote the free flow of factors. Second, improve the economic resilience of counties according to local conditions. The very high resilience of the eastern coastal region should actively promote the advanced adjustment of the industrial structure and improve the quality of the economy, while beware of the rigidity of the economic structure, maintain the diversity of the regional economic structure and enhance the risk resistance of the economy. The very low resilience of Northeast China and Inner Mongolia should break the inertia of resource dependence, accelerate the pace of upgrading traditional industries, and cultivate the formation of a matching total factor productivity structure, so as to cut into the domestic industrial and value chain division of labor, in order to achieve a smooth transition between resource-dependent cities and industrial monocities. Central and western regions should make full use of the country’s central and western tilted policies to lay a good foundation for undertaking related industries in the eastern coast, give full play to their own resource advantages, promote the development of special industry clusters, and find the optimal development path. Third, take targeted measures according to the factors influencing regional economic resilience differences. In the above analysis of the factors influencing regional economic resilience, we found that industrial structure diversification had a significant impact on the long-term economic resilience differences of counties, and location conditions and population density influenced the economic resilience differences of counties in different periods. Therefore, to improve economic resilience, on the one hand, we should promote the optimization and upgrading of industrial structure, establish a diversified power mechanism for economic development, improve the diversification of the regional economy, and thus improve the region’s ability to cope with external economic shocks. On the other hand, it is also necessary to maintain economic and social stability, create a strong macroeconomic base, continuously optimize the regional development environment and enhance the attractiveness of talents. Fourth, strengthen the county’s economic resilience monitoring, good planning and coordination. It should be combined with big data technology to continuously monitor the development of county economic resilience in both time and space dimensions, including the types of pressures and shocks affecting county economic resilience, to provide a basis for county economic resilience enhancement and governance. Improving the resilience of county economies and narrowing regional gaps requires strong leadership from the government and management, the formulation of relevant supporting policies, and the creation of an institutional environment conducive to promoting the resilient development of county economies. Specifically, the policy level, the introduction of a series of county economic development support policies to break the county economic development bottlenecks and barriers. At the planning level, strengthen the planning and coordination, do a good top-level design, governments at all levels should take into account their own reality, build planning and policies oriented to resilient development, and put forward suitable paths and development goals for county economic resilience enhancement.

6. Conclusions

This study took China’s counties as the research object, which has implications for the theory and practice of regional economic resilience research. First, we took counties as the research unit to explore the spatial characteristics of economic resilience in China’s counties, expanding the previous studies on the macro level such as national and provincial areas from the research object to make up for the lack of research on the county scale. Second, we distinguished economic resilience into short-term and long-term economic resilience, and included the “slow burn” factor, which was often neglected in previous studies, into the consideration of long-term economic resilience, which can reflect the pattern of economic resilience of China’s counties in a more comprehensive way. Finally, the factors were examined in the framework of short- and long-term economic resilience, providing a new perspective for the study of factors influencing the economic resilience of counties. The main conclusions obtained from the above analysis were as follows:

(1)

A comparative analysis of the spatial differentiation pattern of economic resilience of China’s counties to provide a theoretical basis for the improvement of China’s regional economic resilience and spatial pattern optimization. The study found that the pattern of distribution of short-run and long-run economic resilience varies considerably across time. In terms of long-term economic resilience, 2007–2020, China’s county economic resilience was dominated by moderate resilience, and the number of very high resilient counties was low and mostly distributed in the southeast coastal region and its hinterland or around the provincial capitals and urban clusters. Very low resilient counties were concentrated in the northeast region, northern Inner Mongolia and Qinghai and Tibet. Looking at short-term resilience, counties with low resilience dominated in 2007–2009, concentrated in northwest and southwest China and the middle reaches of the Yellow River. From 2009 to 2017, the overall level of economic resilience of Chinese counties has been increasing, spatially showing a concentration of very high resilience counties to the Central Plains urban agglomeration and the eastern coast, and very low resilience regions showing a contraction in the western region and an expansion in the northeast. This phenomenon was pronounced even more in 2017–2020. The results of the empirical study can better meet the actual situation of China’s development, and the findings of the study can provide a basis for the formulation of China’s regional development policies.

(2)

Compared the spatial correlation characteristics and evolutionary mechanisms of county economic disparities in China. The economic resilience of China’s counties was found to be significantly spatially correlated. The pattern of economic resilience development among counties reflected the spatial clustering of similar values, and the local spatial pattern and the distribution of cold and hot spots showed H-H and L-L clustering and a polarization pattern of cold spots (99% confidence) and hot spots (99% confidence). The H-H zone and the hot spot (99% confidence) zone were mainly distributed in the southeast coast and its hinterland, and gradually extended to the central region, while the L-L zone and the cold-spot (99% confidence) zone were mainly distributed in Inner Mongolia and northeast China. The results of the spatial variation function indicated the gradual increase of structural factors of the variation of the spatial pattern of economic resilience in China’s counties. Among them, the economic resilience differences in the whole direction were narrowing trends. In the east–west direction of the county economic resilience spatial pattern variation was relatively small, south–north direction increased, northwest–southeast upward spatial differences are more balanced, northeast–southwest direction of the county economic resilience in the homogeneity of the weakest, spatial variation is significant. The overall reduction in spatial differences in economic resilience of China’s counties is related to the promotion of regional coordination strategies such as the “The Belt and Road” initiative, the “Yangtze River Economic Belt”, and the “Western Development”, as well as the continuous improvement of infrastructure and the further strengthening of inter-regional factor flows, which have improved regional economic resilience. In particular, the “The Belt and Road” initiative and the continuous promotion of this makes the economic resilience of the counties along the route, especially in the western region, significantly improved. With the advantage of its own factor endowment, the Southwest region has explored a high-quality coordinated development path suitable for itself, with a diversified industrial structure, strong economic growth momentum and rising economic resilience. The accelerated transformation of China’s industrial structure and supply-side reforms in recent years have cut some of the excess capacity, resulting in the old industrial bases in the northeast generating insufficient development momentum, and therefore their economic resilience is lower, causing the economic resilience gap between counties in the northeast–southwest and south–north directions to expand.

(3)

Previous discussions on the influencing factors were limited to a certain time period or point in time, and there was a lack of research on the spatial mechanism of multi-factor effects in different time periods. This study analyzed the multivariate driving mechanisms of economic resilience differences in China’s counties from different time periods, bridging the gap of previous studies and better explaining the causes of economic resilience differences in China’s counties. The spatial differentiation of economic resilience in Chinese counties is subject to many factors, and the intensity of their influences varies at different stages. The study showed that the enhancement of industrial structure diversification had a significant positive effect on the enhancement of long-term economic resilience in the county, and the factors influencing short-term economic resilience varied depending on the type of shock. In 2007–2009, government support was a more significant influencing factor, while in the recovery and restructuring period of 2009–2017 it was the social factor of population density that had a greater impact on economic resilience, and in 2017–2020, industrial structural diversity played a dominant role in counties’ economic resilience.