Road to Resilient Cities: The Power of Education Investment from China’s Cities

Abstract

Educational investment is important for resilient city shaping. Based on the perspective of education resource input, this paper empirically examines the effect and mechanism of urban education investment on urban resilience construction with a sample of 280 prefecture-level cities in China from 2011 to 2023. The study finds that urban education investment can significantly promote urban resilience governance. In terms of the effect mechanism, urban education investment mainly enhances urban resilience through two paths: science and technology innovation and industry upgrading. The heterogeneity test reveals that the differences in economic level, administrative level, education input preferences, and geographic location of cities lead to the heterogeneous performance of the incentive effect of education investment on resilience shaping. Based on this, policy recommendations are put forward in terms of strengthening the stability and continuity of urban education investment; improving the level of innovation and industrial structure; and emphasizing the problem of unbalanced education development.

1. Introduction

Against the backdrop of intertwined crises, such as global climate change, economic downturn, and social volatility, scholars have viewed urban resilience as a common choice to address such challenges [1,2]. Urban resilience is defined as the ability of complex urban social-ecological systems to recover, change, adapt and transform in response to pressures and constraints [3,4,5], and its construction depends not only on the strengthening of physical infrastructure, but also on the long-term accumulation of human and social capital. In this context, education investment, as a fundamental condition for urban development, can provide important support for urban resilience. For example, Singapore provides universities with special research funds for urban resilience and encourages the introduction of smart technologies and data analysis into the curriculum to create an urban emergency-response training system. The Nordic countries, on the other hand, take the community as the driving body, guide public participation in disaster-management training through school workshops, optimize the fairness of emergency resource allocation, and promote the construction of resilient cities. However, for developing countries like China, education investment and the building of urban resilience are still at a relatively backward stage, and a large number of low-education groups tend to become vulnerable in disasters, exacerbating urban inequality. In contrast, the successful experiences of developed countries emphasize the fundamental role of education investment and form a significant urban resilience enhancement effect through education guidance, which provides inspiration for China’s urban resilience shaping. Especially in recent years, China’s education investment has continued to increase, which not only affects the knowledge accumulation and skill development of individuals, but also influences the innovation capacity, economic growth, and social coordination of the whole city. A growing number of studies show that education investment has a significant impact on urban resilience [6,7]. A high-quality education system not only improves the skill level of the workforce, but also enhances social resilience and innovation, making cities more resilient and capable of coping with the challenges they face.

Based on the above analysis, this paper aims to address the following research questions: ① Determine whether educational investment can play a substantial role in enhancing urban resilience? ② What kind of mechanisms and channels does education investment work through? ③ What kind of differentiation is there in the relationship between the impact of educational investment on urban resilience? Specifically, this paper empirically examines the direction and intensity of the impact of education investment on urban resilience using Chinese prefecture-level city-level panel data from 2011–2023 and explains the underlying mechanisms from the perspectives of both quality and efficiency.

The marginal contributions of this paper are as follows. First, combined with the current background of high-quality economic development and urban transformation, this paper deconstructs urban resilience into three dimensions: resistance and recovery ability; adaptation and adjustment ability; and transformation and development ability, so as to construct a comprehensive evaluation index, and to provide a new methodology for urban resilience evaluation. Secondly, this paper distinguishes from the above paths and breaks down the logic of education investment into two aspects, namely “quality improvement” and “efficiency enhancement”, which provides a new path reference for further releasing the education dividend. Third, for the research sample, economic data of Chinese cities at the prefecture level and above were used to construct a measure of urban resilience and matched with education investment, extending the data dimension of the existing empirical sample and providing more detailed research evidence.

2. Literature Review

2.1. Research of Urban Resilience

Resilience is related to the future direction of urban transformation, and has become a core topic of regional research in recent years [5,8,9] and scholars have richly discussed the connotation of urban resilience, measurement, and construction strategies.

2.1.1. Connotation of Urban Resilience

In terms of connotation, the established literature has redefined the connotation of urban resilience based on different perspectives with the transformation of resilience development history as the main line. The connotation of urban resilience in the early stage focuses on the rapid resilience of cities [10,11,12]. Since then, the diversified urban development has injected new content into the connotation of resilience. Urban resilience should not be limited to the restoration of the original state, it is a combination of short-term resilience and long-term adaptive capacity, and the current connotation of resilience places more emphasis on the balanced development of various capacities within urban resilience relative to the earlier stages [13,14,15].

2.1.2. Measurement of Urban Resilience

In terms of measurement, one category of approach is to understand resilience as disaster coping capacity, and to measure resilience through the quantification of social, natural, and physical polymorphic capital [16]. Another more common approach is to decompose urban resilience into social, economic, infrastructure, institutional, and environmental dimensions, and to evaluate urban resilience using microdata [17,18,19]. In addition, some scholars have adopted different approaches to measure urban resilience in terms of the systemic relevance of the city, the degree of external threats, resilience, and coping measures [1]; or evaluating urban resilience from the perspectives of both adaptive capacity and planning capacity [20], which also provides new ideas for urban resilience measurement.

2.1.3. Enhancement Strategies of Urban Resilience

In terms of construction strategies, established studies have proposed urban resilience construction measures from a multidimensional perspective. In the face of the requirements of high-quality economic transformation, improving economic diversity and flexibility, promoting infrastructure modernization and renewal, and encouraging technological innovation are important measures to enhance economic resilience [21,22]. From a social development perspective, urban resilience can be enhanced through social capital, public participation [23,24]. The role of institutions and governance capacity for urban resilience should not be neglected, and efficient urban governance and strong public policy support can better enhance urban resilience [25].

2.2. Effects of Educational Investment

In addition, the public has fully recognized the significant value-creating effects of education investment [26,27] and analyzed the multiple economic impacts of education investment. At the macro level, endogenous growth theory regards human capital accumulation as an endogenous factor for economic development, emphasizing that educational inputs not only increase labor productivity through long-term educational accumulation [28,29], but also stimulate technological innovation and diffusion [30,31], which in turn bringing about a double improvement in the quality and efficiency of economic development. Other studies have also demonstrated that education investment contributes to foreign capital inflows and alleviates regional poverty [32,33]. At the micro level, education investment enhances workers’ human capital [34], improves skill matching [35], increases employment competitiveness, and promotes social mobility by improving the quality and accessibility of education [36,37], creating positive conditions for individual employment, with significant improvements in employment opportunities and income levels.

2.3. Connection of Educational Investment and Urban Resilience

Based on the diversified economic effects of education investment, some scholars have refined their research perspectives, focusing on the impact of educational inputs on the sustainable development of cities. First, urban entities can form a “talent concentration effect” [38,39] by increasing education investment to enhance their innovative capacity and economic competitiveness. Second, education investment in disadvantaged groups and low-income areas can reduce social inequality and enhance urban inclusiveness [40,41], thus compensating for the unevenness of urban development. Third, the adaptive capacity and governance capacity enhancement generated by education investment can help cities adopt more rational coping strategies when they are hit by epidemics, disasters and economic crises, enhance crisis management resilience, and provide support for sustainable development [42,43].

2.4. Research Gap

The review of this paper shows that established studies have already conducted rich and valuable discussions on urban resilience, especially with the continuous deepening of the research on the concept of urban resilience. Its evaluation methods are also being updated, and the corresponding urban resilience construction strategies are more diversified, which lays a potential theoretical foundation for the research of this paper. However, there are also some shortcomings that limit the research outreach of urban resilience: Firstly, at this stage, domestic and foreign research on urban toughness mainly focuses on its connotation, evaluation measurement, enhancement strategies, etc. Less of the literature starts from the perspective of urban education investment, incorporates education factors into the research framework of urban toughness, and lacks the characterization based on the conditions of education investment. Secondly, in terms of the impact effect of education investment, more of the literature discusses the economic effect of education investment based on the traditional path, and research perspectives are mostly focused on the level of high-quality development of the economy, and there is a lack of sufficient attention to the impact of education investment in the future to cope with the challenges of the city; there is also a lack of theoretical logic in terms of how education investment affects urban resilience and the examination of the mechanism of the role.

3. Hypothesis Development

3.1. The Direct Impact of Educational Investment on Urban Resilience

Education, as one of the important influences on productivity, has a subtle driving effect on urban resilience. From the perspective of the practical process of urban resilience construction, the ability to cope with crises and adaptive capacity, and the ability of sustainable development are inseparable from the support of education investment. As a basic public service for urban development, education investment directly promotes the formation of a benign development pattern of human capital quality, production factor allocation, scientific and technological innovation, and social collaboration in cities. On the one hand, education investment cultivates high-quality human capital, enhances the adaptability and learning ability of the public, and strengthens the public’s sense of social participation in a subtle way, so as to cope with crisis impacts through social cooperation, human resource allocation adjustment, and other ways to quickly adapt to the needs of urban transformation, thus enhancing the stability of urban resilience [44,45]. On the other hand, education investment provides innovative technological solutions for crisis management and sustainable transformation by promoting scientific and technological innovation, laying a long-term foundation for sustainable resilience enhancement [46,47], successful experiences from countries such as Singapore, Japan, and the Netherlands also verify this. This paper proposes Hypothesis 1.

Hypothesis 1:

Education investment contributes significantly to urban resilience.

3.2. The Intermediary Mechanisms

3.2.1. Human Capital

First, the scaling up of education investment has produced a large number of highly qualified workers [48] with not only improved labor skills, but also stronger problem-solving and innovative thinking skills [49]. Supported by these capabilities, a high-quality labor force endows cities with more significant resilience, making them more capable of maintaining vitality and recovery speed amidst environmental fluctuations and crisis shocks [50,51]. Second, education investment promotes human capital diversity. Differentiated education investment and training models provide the market with a diverse pool of human capital, including traditional skilled workers and modern skilled labor, such as in data analysis, thus laying a solid human foundation for cities to adapt to technological change and meet the demands of future transformations. It also improves workforce adaptability and resilience in response to economic change, further enhancing the city’s crisis resilience. Third, education investment not only empowers individuals with knowledge and skills, but also enhances their social responsibility, sense of community, and ability to collaborate. This social capital increases the level of trust and willingness to collaborate among communities, making cities more capable of recovering quickly in times of crisis through resource sharing and collective action, providing a stable social foundation for urban transformation [52]. This paper proposes Hypothesis 2.

Hypothesis 2:

Education investment can enhance urban resilience through the human capital pathway.

3.2.2. Science and Technology Innovation

On the one hand, education investment can play a positive role in shaping urban resilience by financing technological research and development and innovation activities. In terms of education spending, higher education has always had a high share, for example, at least 24 universities in China will invest more than RMB 10 billion in education in 2024. The value of these large educational inputs is not only reflected in the ability to enhance the scientific research capacity of teaching institutions, but also in the greater value, which is to lead to technological innovation breakthroughs, providing good technical support for solving urban crises and challenges. As such, the social service capacity has become increasingly prominent, which promotes the integration of scientific and technological elements with the urban resilience of the confluence. On the other hand, the construction of education informatization has become a key area of current education investment and is catalyzed by this education tilt strategy. A large number of emerging technologies represented by digital technology can be rapidly popularized and applied to improve the urban scientific and technological environment, effectively enhance the foundation of urban resilience, and provide effective technological support for the construction of resilient cities [53,54]. Some studies confirm the application of new generation information technology, such as digital platforms, big data, and artificial intelligence, at the current stage to help cities realize key technological breakthroughs around risk assessment, monitoring and early warning, rescue and disposal, and comprehensive safeguards, which empowers the construction of resilient cities [55]. Therefore, education investment can promote the extensive penetration of digital technology into the urban governance system, which can not only enhance the city’s ability to resist, but also help the city to recover from shocks as soon as possible and explore new development paths through adaptive adjustments. This paper proposes Hypothesis 3.

Hypothesis 3:

Education investment can enhance urban resilience through the Innovation pathway.

3.2.3. Industry Upgrading

New structural economics fully affirms the close connection between education investment and industry upgrading and believes that education investment provides the factor endowment structure power for Upgrade. From the supply side, the expansion of educational inputs brings direct effects, such as technological progress and efficiency improvement, thus changing the comparative advantage of the original factor endowment structure, stimulating the upgrading and adjustment of industrial structure, and realizing the dynamic matching effect between education investment and industry upgrading [47,56]. From the demand side, the new structural economic theory believes that the essence of Upgrade is the continuous growth per capita income, and the basic logic lies in the fact that education investment creates micro output benefits for a large number of labor, improves the consumption ability and level of labor, renews the concept of consumption, and drives the development of the urban service industry, which promotes the upgrading of the industrial level from the demand side. Then, the coupling of education investment and industry upgrading has a positive effect on urban resilience. Industrial structure upgrading strengthens the coordination of inter-industry operation, promotes technology and knowledge spillover, enhances the adaptive capacity and learning ability of the urban system, disperses the risk of external shocks, and strengthens the resilience of the urban economy. At the same time, with the upgrading and transformation of industrial structure, different production sectors can effectively utilize all kinds of production resources to break through bottlenecks at the high end of industrial development and connect the middle and low end of industry, so as to realize the interconnection of the high, middle, and low end of the industrial chain, as well as the effective integration of resource elements, eliminating the factors of industrial instability, thus laying a solid foundation for the enhancement of the city’s economic resilience [57]. Accordingly, this paper proposes Hypothesis 4.

Hypothesis 4:

Education investment can enhance urban resilience through industry upgrading pathways.

3.2.4. Labor Efficiency

Based on the perspective of efficiency, this paper argues that education investment can play a positive role in shaping urban resilience through the mediating path of labor efficiency enhancement. Established studies have shown that labor efficiency is an important factor in strengthening urban resilience [58]. On the one hand, improved labor efficiency will increase the scale efficiency of the city’s overall economic output and form a spillover effect that leads to the simultaneous improvement of other aspects of social development, with more advanced production capacity and adjustment capacity, thus forming the systemic advantages of the city and helping to reduce the vulnerability of the city to crises. On the other hand, higher labor efficiency also puts higher demands on urban governance. When labor efficiency rises to a higher level, it must be matched by high-quality urban resilience in order to achieve a Pareto-optimal allocation of labor resources and other factors. At this time, labor efficiency forces the city to improve the carrying capacity of resources, enhance the environmental capacity, and optimize the public service system, so as to reduce the adverse impact of various crisis shocks on the city. From the source, education investment, as a fundamental function of urban development, can play a contributing role to labor efficiency in the upstream link of urban resilience shaping [59]. At the individual level, education investment significantly raises the level of professional skills of the labor force, which can help improve production processes and shorten production cycles, leading to greater marginal output benefits. At the overall level, educational input helps to improve the high-quality education system, thereby significantly increasing labor efficiency at the level of the entire industry through the universality and penetration of education, enabling rapid adaptation to technological change, and guaranteeing the continuity and adaptability of urban activities. In summary, labor efficiency assumes the role of a mediating mechanism for education investment for urban resilience management, connecting education investment and urban resilience as a unified system. Accordingly, this paper proposes Hypothesis 5.

Hypothesis 5:

Education investment can enhance urban resilience through the labor efficiency pathway.

And the theoretical framework of this paper is shown in Figure 1.

4. Research Design

4.1. Model Setting

This study focuses on exploring the impact of education investment on urban resilience, so a two-way fixed-effects model is selected for empirical analysis [45,47] as follows:

$$RE{S}_{it}={\alpha }_0+{\alpha }_{1}ED{U}_{it}+{\alpha }_{2}Control{s}_{it}+{\mu }_t+{\theta }_i+\epsilon$$

(1)

where i represents individual city, t represents year, $RE{S}_{it}$ represents the urban resilience of city i in year t, and $ED{U}_{it}$ is the core explanatory variable representing the urban education investment of city i in year t. $Control{s}_{it}$ is a collection of control variables that may affect the urban resilience. ${\mu }_t, {\theta }_i$ denote time and city fixed effects, respectively, and ε is a random error term. It should be noted here that the two-way fixed-effects model was selected for this study because a series of control variables were selected in the early stage to deal with the omissions or biases in the model, but there are still unobservable factors in the city and time dimensions, which interfere with urban resilience; in order to ensure that the conclusions are more accurate, the two-way fixed-effects model was finally established as the baseline regression model.

4.2. Variable Selection

Explained variable: urban resilience. Existing studies have provided different solution ideas for economic resilience measurement; based on different research focuses, single sensitivity index, comprehensive evaluation index system, the GEE model, and other methods have been proposed, respectively. Combined with data availability and the aforementioned connotation definition of urban resilience, this paper decomposes urban resilience into resistance and resilience, adaptation and adjustment, transformation and development, and selects the corresponding subdivided indicators for entropy value method measurement, and finally constructs the evaluation index system of urban resilience [60]. And the urban resilience evaluation indicator system is shown in

Table 1. Urban resilience evaluation indicator system.

Variable Name	Competency Decomposition	Breakdown of Indicators
RES	Resistance and Resilience	Per capita GDP (RMB)
		Disposable income of urban residents (RMB) Savings balance of urban and rural residents (RMB)
		Number of urban registered unemployed (persons)
		Total import and export as a share of GDP (%)
		Ratio of local fiscal revenue to expenditure (%)
	Adaptation and Adjustment	Total retail sales of social consumption (RMB)
		Share of tertiary industry in GDP (%)
		Deposit and loan ratio of financial institutions at the end of the year (%)
		Investment in fixed assets (RMB)
		Number of patents granted (pieces)
	Transformation and development	Number of students in general higher education schools per 10,000 Persons (persons)
		Fiscal expenditure on science (RMB)
		Fiscal expenditure on education (RMB)

.

Explanatory variable: education investment. In this paper, we use the city’s annual public budget expenditure on education to measure the city’s education investment [61]. In order to avoid the problem of heteroskedasticity, the indicator is logarithmized. At the same time, considering the impact lag of education investment, this paper lags one period of the education investment indicator.

Mediating variables: Based on the previous theoretical mechanism, this paper selects human capital, science and technology innovation, labor efficiency and industry upgrading as mediating variables. Among them, human capital is expressed by the natural logarithm of the number of students enrolled in general higher education institutions [62], science and technology innovation is measured by the number of invention patents applied by the city in the year [63], industry upgrading is expressed by the proportion of the output value of the tertiary industry and the output value of the secondary industry in the city [64], and labor efficiency is expressed by the total factor productivity of the city [65].

Control variables: in order to control the potential impact of other related factors on urban resilience, this paper introduces a series of control variables in the benchmark regression model. Economic level (Eco): cities with high economic levels form a pool of crisis response funds through tax revenue accumulation and fiscal surpluses, and they are more likely to attract innovative resources to promote resilient technologies [66]. Openness (Open) cities that are more open to international trade can reduce the impact of a single shock by diversifying their markets [67]. Financial efficiency (Fin), mature financial markets hedge resilience risk through insurance, derivatives instruments [68]. Market size (Cap), large consumer markets buffer external shocks by enhancing demand stability [69]. Social investment (Inv), public services investments, such as community centers and public health care, can strengthen residents’ mutual aid networks and enhance their ability to help themselves after a disaster [70]. Infrastructure construction (Infras): a balanced layout of infrastructure reduces the vulnerability of disadvantaged groups in crises and avoids the widening of the resilience divide [71]. Government intervention (Gov): excessive administrative interventions can lead to duplication of infrastructure, such as “performance projects” which, in turn, undermines long-term resilience [72]. Urbanization level (Urban): high urbanization rates require refined resilience management, accelerating the promotion and application of smart city platforms [73]. All the variable settings are shown in

Table 2. Variable definition.

	Variable	Definition	Calculation Methods
Dependent variable	RES	Urban resilience	Entropy method
Independent variable	EDU	Educaiton investment	Ln (Urban education investment in the previous period)
Mediating variable	Human	Human capital	Ln (Number of students enrolled in general higher education institutions)
	Innovation	Science and technology innovation	Ln (Number of patent applications in the city)
	Upgrade	Industry upgrading	The output value of the tertiary industry/The output value of the secondary industry in the city
	Efficiency	Labor efficiency	Calculated using the SBM model
Control variable	Eco	Economic level	Ln (GDP)
	Open	Openness	Total exports and imports/GDP
	Fin	Financial efficiency	Loan balance of financial institutions/GDP
	Cap	Market size	Ln (Total retail sales of consumer goods)
	Inv	Social Investment	Total investment in social fixed assets/GDP
	Infras	Infrastructure construction	Ln (Road freight per capita)
	Gov	Government intervention	General fiscal expenditure/GDP
	Urb	Urbanization level	Urban population/Total population

.

4.3. Data Sources

This paper is based on the panel data of 280 prefecture-level cities in China from 2011 to 2021, in which the data of urban education investment are from 2010 to 2020. The relevant data come from China Urban Statistical Yearbook, CSMAR database, and so on. Continuous variables are preprocessed with upper and lower 1% shrinkage in the data collation process, so as to alleviate the interference of results caused by outliers. The final sample of 2832 observations from 280 prefecture-level cities is obtained.

4.4. Descriptive Statistics

The descriptive statistics of the relevant variables are shown in

Table 3. Descriptive statistics.

VarName	Obs	Mean	Median	Max	Min	SD
RES	3374	0.073	0.053	0.364	0.025	0.059
EDU	3374	13.217	13.216	15.472	11.293	0.755
Eco	3374	16.731	16.621	19.186	14.764	0.915
Open	3374	0.168	0.076	1.390	0.002	0.250
Fin	3374	1.494	1.437	2.812	0.777	0.398
Cap	3374	15.771	15.703	18.243	13.610	0.992
Inv	3374	0.909	0.883	2.228	0.179	0.380
Infras	3352	3.212	3.232	4.955	1.587	0.634
Gov	3374	0.203	0.181	0.508	0.081	0.089
Urb	3374	0.578	0.557	0.949	0.296	0.141

, which reveals that the maximum value of urban resilience in the study period is about 15 times that of the minimum value, reflecting the basic fact that there is a significant level of resilience in Chinese cities.

At the same time, the median urban resilience is smaller than the mean, indicating that the data distribution pattern of urban resilience is to the right, which also means that the resilience level of more than half of the cities in the overall observation sample is located below the mean. This also reflects the great variability of urban resilience. In terms of education investment, the difference in education investment is similarly large, and the data distribution tends to be left-skewed. Although there are still unbalanced differences in education investment across Chinese cities, along with the strengthening of the fundamental role of education investment in urban development, the importance attached to education investment by local governments has also risen, narrowing the education investment gap between regions.

In addition, this paper uses ArcGIS to map the trends in urban resilience and education investment, and the results are shown in Figure 2. First, both urban resilience and education investment in 2021 show a significant increase compared to 2011; this improvement is more significant in the eastern coastal region. At the same time, the results also confirm the significant differences in resilience and education investment between cities in descriptive statistics: cities with higher resilience levels and larger education investment are generally concentrated in the eastern and southern regions, while cities in the western and northern regions lag behind in resilience levels and education investment. The above results reflect the dynamic changes of urban resilience and education investment; on the one hand, the synchronous change of the two proves the correlation between urban resilience and education investment. On the other hand, the basic facts of urban resilience and education investment also have obvious differentiation, which together provide a good realistic situation for this paper to explore the relationship between education investment on the impact of urban resilience.

5. Empirical Analysis

5.1. Baseline Regression

This paper examines the base impact effect of urban education investment on urban resilience using a benchmark regression model, and the results are presented in

Table 4. Benchmark regression.

	(1)	(2)	(3)	(4)	(5)	(6)
	RES	RES	RES	RES	2011–2019	2020–2023
EDU	0.051 ***	0.012 ***	0.011 ***	0.012 ***	0.005 **	0.019
	(40.57)	(5.10)	(4.43)	(4.64)	(2.57)	(0.99)
Eco		0.038 ***	0.022 ***	0.023 ***	0.023 ***	0.006
		(12.04)	(6.01)	(6.06)	(7.27)	(0.22)
Open		0.013 ***	−0.014 ***	−0.014 ***	−0.022 ***	0.042
		(3.40)	(−2.80)	(−2.84)	(−5.79)	(1.39)
Fin		−0.003	0.000	−0.000	0.002	−0.005
		(−1.44)	(0.15)	(−0.19)	(1.22)	(−0.40)
Cap		0.001	0.002	0.003	−0.001	−0.015
		(0.62)	(0.99)	(1.56)	(−0.34)	(−0.76)
Inv		−0.003 *	−0.002	−0.001	−0.002	−0.008
		(−1.86)	(−1.34)	(−0.74)	(−1.21)	(−0.46)
Infras		0.000	0.001	0.000	−0.000	−0.003
		(0.24)	(0.53)	(0.47)	(−0.06)	(−0.84)
Gov		0.063 ***	0.036 **	0.032 **	−0.005	0.253 ***
		(5.73)	(2.56)	(2.25)	(−0.39)	(5.70)
Urb		0.070 ***	0.002	−0.001	−0.003	0.018
		(10.01)	(0.17)	(−0.07)	(−0.36)	(0.20)
Const	−0.600 ***	−0.780 ***	−0.486 ***	−0.525 ***	−0.371 ***	−0.081
	(−35.83)	(−30.45)	(−8.62)	(−8.89)	(−7.40)	(−0.15)
Year	No	No	Yes	Yes	Yes	Yes
City	No	No	Yes	Yes	Yes	Yes
Year * City	No	No	No	Yes	Yes	Yes
N	3374	3352	3352	3352	2354	998
Adj. R2			0.956	0.956	0.968	0.979

Note: t-statistics in parentheses; ***, **, and * indicate significance levels of 1%, 5%, and 10%.

. The regression in column (1) directly examines the impact of education investment on urban resilience without including control variables and without controlling for year and city fixed effects, and the estimated coefficients of education investment are significantly positive, indicating that education investment have a positive impact on urban resilience, and preliminarily verifying that education investment can significantly enhance the level of urban resilience. Subsequently, the control variables are added to repeat the benchmark regression, and (2) shows that the estimated coefficient of education investment is 0.012 and statistically significant, implying that education investment enhances urban resilience by 1.2%. Compared with the results of the test without the inclusion of control variables, the effect of education investment on urban resilience is substantially reduced, which is due to the fact that part of the effect is absorbed by other control variables, which proves the reasonableness of the selection of control variables. In column (3), the regression results are still significantly positive when controlling for both year and city two-way fixed effects based on the inclusion of control variables.

What needs special attention is that although this paper controls for year and city fixed effects in the benchmark regression model, it does not exclude the possibility that urban resilience shaping is also driven by its own factor endowment characteristics. In other words, urban resilience shaping cannot be separated from the support of urban population, capital, technology and other production resources, which are continuously adjusted and perfected along with the passage of time in the process of urban development, forming the endogenous driving force for the enhancement of urban resilience level, thus triggering potential endogeneity. Therefore, this paper sets the year and city fixed-effects interaction term in the regression in column (4) to control the endogenous impact of urban factor endowment characteristics on urban resilience over time, which effectively ensures that the results are robust and reliable. According to the results, the estimated coefficient of education investment after controlling for multiple fixed effects is 0.012, which passes the 1% significance level test, proving that the enhancement effect of education investment on resilience shaping exists substantially.

Further considering the shock of the COVID-19 epidemic, this paper divides the study sample into two time periods, 2011–2019 and 2020–2023, using 2020 as the time point. The regression results are presented in columns (5)–(6) of Table 4. The regression coefficient of education investment in the study sample of 2011–2019 is 0.005, which is significantly positive, implying that education investment in this period can significantly contribute to urban resilience enhancement. However, the regression coefficient of education investment in the 2020–2023 study sample is greater than 0, which fails the significance test indicating that there is no substantial correlation between education investment and urban resilience due to the impact of epidemic shocks.

5.2. Robustness Test

There are still potential endogeneity and result bias issues in the benchmark regression modeling, and this paper has carried out the following to ensure the robustness of the research results. First, considering the endogeneity issue, this paper employs an instrumental variable approach to endogeneity by introducing three exogenous instrumental variables using 2SLS. Second, the benchmark regression results are re-validated using a range of other methods, including replacing explanatory variables, controlling for time trends, and excluding municipality samples.

5.2.1. Instrumental Variable

In the benchmark regression, this paper uses stepwise regression to control for control variables and different fixed effects, which can effectively ensure that the test results are robust and reliable. However, in addition to omitted variables, endogeneity may also be caused by issues such as reverse causation, as well as measurement error. Firstly, urban resilience as a systematic project, its shaping process may also put forward corresponding requirements on urban education development, which may lead to the expansion of education investment in cities. Secondly, this paper constructs a comprehensive evaluation index of urban resilience, in which there may be some data bias of secondary and tertiary indicators that may interfere with the accuracy of the index. Therefore, this paper adopts the instrumental variable method for endogeneity testing. Specifically, this paper adopts lag one (IV1) and lag two (IV2) of education investment and the number of full-time teachers in urban general higher education schools (IV3) as instrumental variables, respectively. The reasons for the settings are as follows: first, according to the exogeneity requirement of instrumental variables, there is no direct correlation between instrumental variables and urban resilience. On the one hand, China’s urban education investment is determined at the end of the previous year through the government work report, so it is possible to take the previous year’s education investment as an ex ante variable, and there is no correlation between it and the urban resilience of the current year. On the other hand, for urban education development, the number of full-time teachers in general higher education institutions is mainly affected by the size of students in higher education institutions, which has been in a relatively stable state and has a limited impact on urban resilience. Therefore, the above three instrumental variables all meet the exogenous requirement. Secondly, the instrumental variables also have to satisfy the requirement of relevance; that is, the instrumental variables are correlated with education investment. On the one hand, education investment in the current year will be adjusted on the basis of the previous year’s education investment, so as to ensure the coherence of urban education development. On the other hand, as an important part of the urban teacher group, changes in the size of full-time teachers in general colleges and universities cause changes in salaries and wages, which further affects urban education investment. Therefore, the three instrumental variables also meet the correlation requirements.

The results of instrumental variable tests are presented in

Table 5. Robustness test I.

	(1)	(2)	(3)	(4)	(5)	(6)
	First	Second	First	Second	First	Second
	EDU	RES	EDU	RES	EDU	RES
IV1	0.034 ***
	(4.46)
IV2			0.017 **
			(2.24)
IV3					0.019 ***
					(5.08)
EDU		0.041 *		0.053 *		0.419 ***
		(1.67)		(1.86)		(5.01)
Eco	0.559 ***	0.007	0.568 ***	−0.012	0.577 ***	−0.211 ***
	(17.41)	(0.45)	(17.66)	(−0.36)	(17.86)	(−4.20)
Open	−0.002	−0.014 ***	−0.006	−0.013 **	0.04	−0.0107
	(−0.06)	(−2.68)	(−0.14)	(−2.38)	(0.10)	(−0.58)
Fin	−0.014	0.000	−0.156	0.001	−0.014	0.00656
	(−0.81)	(0.08)	(−0.89)	(0.30)	(−0.80)	(0.86)
Cap	−0.006	0.003	−0.006	0.003	−0.003	0.00542
	(−0.32)	(1.59)	(−0.35)	(1.51)	(−0.15)	(0.72)
Inv	0.046 ***	−0.003	0.049 ***	−0.004	0.055 ***	−0.0218 ***
	(3.15)	(−1.24)	(3.29)	(−1.26)	(3.68)	(−2.87)
Infras	0.010	0.000	0.012	−0.000	0.011	−0.00414
	(1.17)	(0.07)	(1.34)	(−0.22)	(1.31)	(−1.08)
Gov	0.996 ***	0.002	1.019 ***	−0.031	1.068 ***	−0.389 ***
	(7.93)	(0.08)	(8.08)	(−0.51)	(8.46)	(−3.82)
Urb	0.057	−0.002	0.057	−0.005	0.066	−0.0275
	(0.67)	(−0.28)	(0.67)	(−0.42)	(0.78)	(−0.75)
Const	4.006 ***	−0.453 ***	4.108 ***	−0.592 ***	3.289 ***	−2.038 ***
	(6.67)	(−3.70)	(6.79)	(−2.48)	(5.24)	(−5.04)
Observations	3341	3341	3341	3341	3287	3287
R-squared	0.025	0.016	0.016	0.208	0.131	0.117
Cragg-Donald Wald F statistic	19.901		25.008		25.790

Note: t-statistics in parentheses; ***, **, and * indicate significance levels of 1%, 5%, and 10%.

. In the first stage, the estimated coefficients of IV1, IV2, and IV3 are all significantly positive and pass the weak instrumental variable test, which proves the correlation between the above instrumental variables and education investment, and the instrumental variable settings are reasonable. In the second stage, the estimated systems of the core explanatory variable education investment in the three instrumental variable regression tests are also positive and highly significant, indicating that the enhancement effect of education investment on urban resilience shaping still exists significantly after adopting instrumental variables for the potential endogeneity problem. The baseline regression results are reliable.

5.2.2. Replacement of Explanatory Variables

This paper uses the city’s public education expenditure in the benchmark regression to measure education investment from the absolute scale dimension, which can generally verify the relationship between the scale of education investment and urban resilience. However, the scale of education investment is also constrained by the city’s economic development and resource endowment. Next, this paper replaces the index of urban education investment and adopts the ratio of urban public education expenditure to total general budget expenditure to measure the urban education investment intensity (EDUI) based on the relative level dimension, which is used as a new explanatory variable to repeat the benchmark regression. This can weaken the interference of economic development and resource endowment on regional public education to a certain extent. According to the test results in

Table 6. Robustness Test II.

	(1)	(2)	(3)
	Replace Explanatory Variables	Control of Time Trends	Exclude Municipalities
EDUI	0.062 ***
	(3.87)
EDU		0.011 ***	0.014 ***
		(4.46)	(5.85)
trend		−0.008 **
		(−2.44)
Eco	0.032 ***	0.024 ***	0.024 ***
	(8.80)	(6.22)	(6.41)
Open	−0.014 ***	−0.014 ***	−0.014 ***
	(−2.78)	(−2.86)	(−2.83)
Fin	−0.001	−0.001	−0.000
	(−0.51)	(−0.32)	(−0.02)
Cap	0.003	0.003	0.004 **
	(1.36)	(1.45)	(2.27)
Inv	−0.001	−0.001	−0.001
	(−0.33)	(−0.87)	(−0.49)
Infras	0.001	0.001	0.000
	(0.68)	(0.53)	(0.25)
Gov	0.054 ***	0.033 **	0.045 ***
	(3.77)	(2.32)	(3.19)
Urb	−0.001	−0.001	−0.000
	(−0.07)	(−0.08)	(−0.01)
Const	−0.522 ***	−0.491 ***	−0.596 ***
	(−8.80)	(−8.09)	(−10.17)
Year	Yes	Yes	Yes
City	Yes	Yes	Yes
Year * City	Yes	Yes	Yes
N	3352	3352	3288
Adj. R2	0.956	0.956	0.946

Note: t-statistics in parentheses; ***, **, and * indicate significance levels of 1%, 5%, and 10%.

, the regression coefficient of education investment intensity as the core explanatory variable is 0.063, which is statistically significant and proves that the positive impact of education investment on urban resilience is still significantly present.

5.2.3. Control of Time Trends

In addition, the research sample in this paper is city-level panel data, which contains both cross-sectional and time-series data. Although this paper controls for year fixed effects and city–year interaction fixed effects in the benchmark regressions to reduce the potential impact of time-varying trends. However, time-induced trend changes in urban governance, economic development, and environmental protection cannot be ruled out as having other effects on urban resilience shaping. Therefore, this paper incorporates time controls into the regression model as a robustness test to further mitigate the disturbance of time-varying trends. The results in column (2) of Table 6 show that after controlling for the time trend, the regression coefficient of trend is significantly negative, which proves that there is a substantial correlation between the time trend change and urban resilience, and that urban resilience shaping is indeed affected by the change in time trend. The estimated coefficient of education investment is still significantly positive, which is in line with the results of the benchmark regression, and it implies that there is a significant urban resilience shaping effect of the city’s education investment.

5.2.4. Resampling

The sample of this paper is drawn from 280 cities in China, including the four municipalities of Beijing, Shanghai, Tianjin, and Chongqing. These four cities, as the forerunners of China’s economy, are naturally favored by more policies and are ahead of other ordinary cities in terms of urban governance, resilience building, and other aspects, which may lead to sample selection bias and thus bias in the regression results. In order to eliminate this potential risk, this paper removes the sample of municipalities from the study and repeats the benchmark regression. The results show that the education investment regression results do not change significantly after excluding the sample of municipalities and are still significantly positive, which ensures the reliability of the benchmark regression results.

6. Discussions

6.1. Mechanism Analysis

Based on the above theoretical inferences, this paper conducts mechanism tests from four dimensions of human capital; science and technology innovation; and industry upgrading and labor efficiency, so as to verify the logic of the role of urban education investment in promoting resilience improvement.

6.1.1. Human Capital Test

Countries around the world generally combine the concept of “human-centeredness” with urban resilience, especially emphasizing the important value of human capital in the resilience shaping process. The innovative capacity, industry labor force rematching effect, and diversified externalities associated with abundant human capital all contribute to the response, adjustment capacity, and synergy of urban resilience systems, and inject vitality into resilience shaping. Based on this, this paper focuses on the correlation between urban education investment and human capital in the mechanism test. The test results show that the regression coefficient of education investment for human capital is 0.040, but it is not statistically significant, implying that the positive contribution of education investment to urban human capital accumulation does not hold in this study.

The reason for the divergence between theoretical analysis and empirical test is that, on the one hand, this paper uses the number of students enrolled in general higher education institutions as a proxy variable for human capital, but there is still a certain distance between students’ education and human capital formation, and not all students can be transformed into human capital immediately after receiving education, so differences in the setting of the indicators may lead to biased results. On the other hand, as the proverb says, “Ten years to grow trees, a hundred years to count people”: education investment to human capital accumulation is a long and slow process, and the correlation between education investment and human capital needs a longer period of time to be reflected. Although the test results of the sample in this paper do not prove the role of education investment in human capital, it is undeniable that education plays an important role in urban resilience enhancement, which reveals that we still need to continue to expand the positive role of education investment in the future to enhance the learning and adaptive capacity of individuals and cities, so as to provide sufficient support for resilience building.

6.1.2. Science and Technology Innovation Test

Theoretical analysis points out that education investment can help cities form technological advantages through innovation breakthroughs and technological applications, prioritizing the restoration of the original state or the adjustment of the transition state to show stronger resistance to external shocks. In this part of the paper, we use the number of city invention patent applications to measure innovation and conduct mechanism tests. The results are listed in column (2) of

Table 7. Mechanism analysis.

	(1)	(2)	(3)	(4)
	Human	Innovation	Upgrade	Efficiency
EDU	0.040	0.404 ***	0.019 ***	0.006
	(0.98)	(5.69)	(2.98)	(0.53)
Eco	−0.137 **	0.354 ***	−0.035 ***	−0.000
	(−2.26)	(3.35)	(−3.76)	(−0.02)
Open	−0.041	−0.085	0.035 ***	0.022
	(−0.56)	(−0.67)	(3.17)	(1.18)
Fin	−0.195 ***	0.118 **	−0.036 ***	−0.002
	(−6.42)	(2.24)	(−7.71)	(−0.32)
Cap	0.044	0.079	0.052 ***	−0.006
	(1.33)	(1.37)	(10.24)	(−0.73)
Inv	−0.003	0.020	0.011 ***	−0.006
	(−0.11)	(0.45)	(2.79)	(−0.88)
lnInfras	0.032 **	−0.017	−0.002	−0.003
	(2.16)	(−0.64)	(−0.91)	(−0.66)
Gov	−0.513 **	0.863 **	0.011	0.023
	(−2.25)	(2.18)	(0.32)	(0.40)
Urb	−0.114	0.485 *	0.052 **	0.030
	(−0.80)	(1.95)	(2.40)	(0.82)
Const	12.100 ***	−6.342 ***	1.856 ***	0.334
	(13.27)	(−4.00)	(13.29)	(1.43)
Year	Yes	Yes	Yes	Yes
City	Yes	Yes	Yes	Yes
Year * City	Yes	Yes	Yes	Yes
N	3352	3352	3352	3352
Adj. R2	0.973	0.950	0.834	0.946

Note: t-statistics in parentheses; ***, **, and * indicate significance levels of 1%, 5%, and 10%.

, and the regression coefficient of education investment is significantly positive, which proves that urban education investment has a strong promotional effect on innovation. It also means that science and technology assumes a substantial mediating effect in the process of influencing resilience construction of urban education investment.

6.1.3. Industry Upgrading Test

Much of the literature has found that industry upgrading can lead to the rise of industrial synergy, guide the rational allocation of resources, and inject a new stable foundation for urban resilience with the help of rationalized and stabilized industrial structure. At the same time, industry upgrading helps to break the traditional industrial path dependence, build new growth points for the urban economic system, enhance the ability of the urban economic system to absorb risks and restore development, and enhance the competitive advantage of the urban resilience system under the crisis impact. To test Hypothesis 4, this paper examines the impact of urban education investment on the resilience system. As shown in column (3) of Table 7, the regression coefficient of education investment is also significantly positive, indicating that education investment plays a significant incentive effect in industry upgrading. It can be seen that urban education investment promotes the strengthening of the resilience system through the path of industrial structure upgrading, and this mechanism is basically valid. This finding also provides new opportunities and development ideas for subsequent urban transformation and resilience shaping.

6.1.4. Labor Efficiency Test

Another path of education investment lies in efficiency improvement. By improving the match between labor supply and demand, it incentivizes the labor force to move from low-skill and low-value-added industries to high-skill and high-value-added industries, which improves labor efficiency and thus positively affects urban resilience. Based on the results, it is found that the regression coefficient of education investment on labor efficiency only has a positive coefficient, which fails the significance test, so the incentive effect of education investment on labor efficiency lacks evidence support, proving that the mediating effect borne by labor efficiency is more limited.

For this phenomenon, this paper explains from the following two points: First, the increase in education investment can indeed promote the progress of labor skills, but labor efficiency improvement does not rely on the supply of highly skilled labor alone; it also needs to have the corresponding labor market demand, to achieve a balanced match between labor supply and demand in order to form the effect of labor efficiency. Therefore, how to optimize the economic structure and realize the effective allocation of labor resources are the key issues to be concerned about in the process of urban resilience. Secondly, the spillover effect brought by education investment may need a long-term transformation stage. In particular, labor efficiency improvement involves structural transformation, market supply and demand allocation, and other aspects, and it is difficult to achieve the expected effect in the short term by simply relying on the increase in the scale of education funding.

6.2. Heterogeneity Test

The previous empirical results basically confirm that education investment can indeed promote the construction of urban resilience. This paper analyzes the heterogeneity of the effect of education investment on urban resilience based on three dimensions: economic level, administrative level, and regional input preference and geographic location.

6.2.1. Economic Level

This paper argues that the level of urban economic development directly affects the incentive strength of urban education investment on resilience shaping. From the perspective of urban education investment, it is obvious that the economic level is highly correlated with the city’s finance and public expenditure. Cities with higher levels of economic development usually have richer sources of fiscal revenue and maintain larger expenditures on public services, such as education and health care. Based on the perspective of urban resilience, the economic level of cities directly determines the layout of industrial structure. Along with the increase in economic level, the industrial structure of the city tends to be more diversified; not only can the labor force obtain more education, training, and employment opportunities, but they can provide more complete conditions for technological innovation. In sum, economic level can be regarded as both the basis for the recovery of urban resilience and the enhancement of the urban system to endow with a stronger ability to adapt and adjust.

Here the median city–year GDP is used as a criterion to distinguish the study sample into two main groups: high and low economic levels. The corresponding regression results are given in columns (1) and (2) of

Table 8. Heterogeneity test I.

	(1)	(2)	(3)	(4)	(5)	(6)
	High Economic Level	Low Economic Level	Core City	Ordinary City	High Regional Input	Low Regional Input
EDU	0.037 ***	0.000	0.045 **	0.001	0.007 *	0.002
	(5.29)	(0.67)	(2.29)	(1.09)	(1.71)	(0.83)
Eco	0.020 *	0.011 ***	0.079 ***	0.022 ***	0.022 ***	0.044 ***
	(1.91)	(8.62)	(2.83)	(5.64)	(4.23)	(6.73)
Open	−0.073 ***	0.002	−0.012	−0.068 ***	−0.008 *	−0.073 ***
	(−9.40)	(0.97)	(−0.28)	(−15.51)	(−1.73)	(−10.35)
Fin	0.024 ***	−0.000	0.044 **	0.002	0.002	0.000
	(4.34)	(−1.14)	(2.22)	(1.28)	(1.20)	(0.22)
Cap	0.006	0.001 *	0.046 **	0.002	0.003	0.006
	(1.14)	(1.88)	(2.45)	(1.47)	(1.31)	(1.25)
Inv	−0.023 ***	0.004 ***	−0.014	0.002	0.003	−0.002
	(−4.96)	(6.34)	(−1.37)	(1.00)	(1.55)	(−0.63)
Infras	−0.005 **	−0.000	−0.016 **	0.000	−0.001	0.001
	(−1.98)	(−1.21)	(−2.22)	(0.48)	(−1.51)	(0.73)
Gov	0.000	0.014 ***	0.439 ***	−0.001	0.024	0.029
	(0.00)	(3.02)	(2.74)	(−0.04)	(1.36)	(1.10)
Urb	−0.061 ***	0.015 ***	0.010	−0.033 ***	−0.006	−0.017
	(−2.93)	(4.24)	(0.23)	(−3.05)	(−0.58)	(−1.00)
Const	−0.799 ***	−0.167 ***	−2.691 ***	−0.324 ***	−0.432 ***	−0.748 ***
	(−5.04)	(−7.59)	(−6.55)	(−4.90)	(−5.51)	(−6.67)
Year	Yes	Yes	Yes	Yes	Yes	Yes
City	Yes	Yes	Yes	Yes	Yes	Yes
Year * City	Yes	Yes	Yes	Yes	Yes	Yes
N	1704	1648	429	2923	1766	1586
Adj. R2	0.933	0.938	0.913	0.932	0.951	0.942
Group	−0.037 ***		−0.044 ***		−0.005 *

Note: t-statistics in parentheses; ***, **, and * indicate significance levels of 1%, 5%, and 10%.

, which show that the estimated coefficients on education investment are significantly positive only in areas with higher economic levels, implying that the higher the economic level, the greater the incentives of education investment on the urban resilience, verifying the previous inference.

6.2.2. Administrative Level

This paper uses Chinese prefecture-level and city-level panel data from 280 cities. In terms of administrative level, the sample cities include both cities under the direct jurisdiction of the central government, such as Beijing and Shanghai, as well as provincial capitals, such as Guangzhou and Nanjing. These cities are usually identified as central cities at the administrative level, and have advantages in terms of urban policy support, resource tilting, and social development that many other small and medium-sized cities do not have, which are crucial for urban resilience shaping. On the contrary, for small and medium-sized cities, they are negatively affected by the siphoning effect of central cities, and often suffer from population loss, weak industries and other drawbacks, which constrain the improvement of urban governance and resilience. It is inferred that there is a differentiation in the role of education investment on the resilience system in the research samples at different administrative levels.

This paper divides the research sample into two subsamples of core cities and ordinary cities according to the administrative level, where the core cities include municipalities, provincial capitals, and planned cities, and all other cities are included in ordinary cities. The regression results of subgroups in columns (3) and (4) of Table 8 show that the estimated coefficients of education investment are significantly positive in the sub-sample of core cities. This suggests that the positive effect of urban education investment on resilience shaping is more pronounced in core cities, validating the manifestation of heterogeneity due to administrative level.

6.2.3. Regional Input Preference

The level of existing local government investment in and emphasis on education may also have an impact on the actual benefits of urban resilience. For example, districts with higher local investment preferences in education may make more efforts to improve the level of education provision with the support of specialized inputs and seek to supply more training courses and technical support with higher costs, while districts with lower local investment preferences in education may tend to choose courses with fewer requirements for local matching funds, which ultimately results in a weaker benefit from urban resilience in districts with low local investment preferences in education.

This paper uses the intensity of local educational inputs (EDUI) as a proxy for different local educational input preferences to explore whether there are differences in the urban resilience shaping effects of lower and higher education input preference areas grouped by the median of this ratio. The results show that the estimated coefficients of education investment are significantly positive in the sample with high education investment preferences, while on the contrary, the positive effect of education investment on urban resilience in areas with low education investment preferences fails the test. This also suggests that differences in education investment preferences between regions interfere with the effect of education investment on urban resilience.

6.2.4. Geographic Location

Combined with China’s urban development history, the problem of regional imbalance has existed for a long time. Relying on location advantages, eastern cities are taking the lead in market-oriented reforms to form a strong agglomeration effect, forming competitive advantages in talent, technology, and innovation resources, effectively promoting the upgrading of urban governance and laying a good foundation for a resilient system. On the contrary, the central and western regions are constrained by geographical limitations, and both education investment and urban resilience are still in a relatively backward state. Accordingly, this paper expects the enhancement effect of urban education investment on resilience shaping to show differentiated performance in different regions.

The results in

Table 9. Heterogeneity test II.

	(1)	(2)	(3)
	Eastern Region	Central Region	Western Region
EDU	0.015 **	0.008 ***	0.000
	(2.28)	(2.72)	(0.01)
Eco	0.031 ***	0.032 ***	0.025 ***
	(3.13)	(5.43)	(4.02)
Open	−0.078 ***	0.028 ***	0.001
	(−9.64)	(3.02)	(0.15)
Fin	0.016 **	0.000	−0.001
	(2.57)	(0.41)	(−0.29)
Cap	0.008	0.003	0.001
	(1.46)	(1.20)	(0.51)
Inv	−0.009 **	−0.002	−0.002
	(−2.09)	(−0.70)	(−0.63)
Infras	−0.005 *	−0.002	−0.001
	(−1.70)	(−1.47)	(−0.68)
Gov	−0.094 **	0.057 ***	0.006
	(−2.03)	(3.01)	(0.27)
Urb	−0.101 ***	0.092 ***	0.001
	(−4.32)	(7.20)	(0.08)
Const	−0.659 ***	−0.690 ***	−0.360 ***
	(−4.50)	(−7.06)	(−3.38)
Year	Yes	Yes	Yes
City	Yes	Yes	Yes
Year * City	Yes	Yes	Yes
N	1256	1124	972
Adj. R2	0.933	0.933	0.933
Group	−0.007 ***
		−0.008 ***

Note: t-statistics in parentheses; ***, **, and * indicate significance levels of 1%, 5%, and 10%.

show that the regression coefficient of education investment is only significant in the subsample of cities in the eastern and central regions, indicating that at this stage, education investment has a significant role in promoting the risk-resistant ability of the eastern and central regions, but the effect on the western region is more limited. For the western region, the long-term reliance on central financial transfers and the education investment gap between the east and central regions are difficult to reverse in a short period of time, limiting the incentive effect of education investment. In the future, it is still necessary to further tilt more education resources to the western region to narrow the regional education gap.

7. Conclusions and Suggestions

7.1. Research Conclusion

In recent years, in order to enhance urban resilience governance capacity, all sectors in China have actively promoted the construction of resilient cities and carried out a great deal of exploration and practice. As a basic public service of local governments, education investment is prominently characterized by permeability and comprehensiveness, providing all-round support for urban transformation. This paper adopts the panel data of Chinese prefecture-level cities from 2011 to 2023 as a research sample to examine the impact of urban education investment on urban resilience in order to reveal new paths of urban resilience shaping in the context of urban transformation challenges. Robust empirical evidence suggests that urban education investment significantly contributes to the level of urban resilience. Mechanism analysis reveals that urban education investment can help shape urban resilience through the path of strengthening science and technology innovation and promoting industry upgrading. Heterogeneity analysis shows that the differential impact of education investment on urban resilience is mainly determined by a combination of factors such as the city’s economic level, administrative level, regional input preference and geographic location. The findings of this paper are of great significance for the in-depth excavation and assessment of the impact effect of urban education investment on resilience shaping, and they provide a program that can be drawn upon for the promotion of the transformation of resilient cities.

7.2. Suggestions

First, a long-term mechanism for education investment should be established to ensure the adequacy and continuity of education investment. The findings of this paper fully validate the positive correlation between education investment and urban resilience; consequently, local government departments should pay great attention to the key role of education investment in the city’s response to the challenges of the crisis and the future transformation process, safeguard the basic status of education investment in public services, and continue to expand the scale of education investment with a long-term perspective.

Second, science and technology innovation should be encouraged and industry upgrading should be promoted. First of all, science and technology innovation, as an important intermediary channel for education investment to influence the resilience of cities, can transform the potential impact of education investment into more substantial production kinetic energy. Thus, local governments should insist on promoting independent innovation and creating a favorable policy environment for independent innovation. At the same time, they should improve the level and quality of foreign investment utilization, introduce, digest, and absorb advanced technologies and make full use of the latecomer’s advantage. Then, the intermediary role of industry upgrading is more reflected in the market demand side through the renewal and adjustment of industrial structure, absorbing and utilizing a large amount of human capital to form the output benefit, providing a shaping foundation for urban resilience. Local governments should comply with the trend in economic change; fully seize the opportunity of urban transformation; promote the industrial structure to dispose of the traditional low-end industrial lock with the help of a series of emerging production technologies, such as digital economy; break through the shackles of the crude development mode; lead the industry to shift to a new mode of low energy consumption and high output; and then realize the advanced industrial foundation, enhance the resilience of the industrial chain and supply chain, and comprehensively enhance the support capacity of industry upgrading for the resilience of the city.

Third, the reasonable distribution of education financial funds should be promoted to ensure the fairness of education investment. For a long time, the unbalanced development of education has become a real problem that restricts the role of education investment. Policymakers need to tailor their policies to local conditions and cities, and through reasonable and effective targeted policy measures, increase education investment in ordinary cities and cities in the central and western regions, narrow the gap in education investment between regions, and achieve balanced development and inclusive growth to maximize the resilience incentive effect of education investment in cities.

7.3. Research Outlook

The current study provides an important empirical basis for the relationship between education investment and urban resilience, but there are still some improvements in the depth of the mechanism. The main explanatory mechanism of this paper lies in scientific and technological innovation and industrial upgrading, both of which are based on traditional economic paths. Therefore, future research can consider the mechanism by which educational investment affects urban resilience through non-economic paths, such as social network and cultural identity, and introduce other models to test the synergistic effect of multiple paths.