The Power of Knowledge: How Can Educational Competitiveness Improve Urban Energy Efficiency?

Abstract

With an economic model characterized by high energy consumption and low efficiency, China is facing serious energy shortages and environmental problems. However, education, as the cornerstone of social progress, has been overlooked in its role in improving energy efficiency. This study aims to enhance our understanding of the impact of educational competitiveness on urban green total factor energy efficiency (GTFEE), helping policymakers to achieve sustainable urban development. This study utilizes panel data from 20 major Chinese cities spanning from 2012 to 2022 and applies a two-way fixed effects model to investigate the relationship and pathways of educational competitiveness (Ec) on GTFEE. Our results show that the Ec index can enhance the major urban GTFEE. Among them, educational resource competitiveness, input competitiveness, efficiency competitiveness, and sustainable competitiveness can all enhance urban GTFEE, but the coefficient of the educational scale is not significant. In addition, Ec can effectively improve GTFEE by promoting green technological innovation, alleviating human resource mismatch, and driving industrial structure upgrading. Furthermore, the impact of Ec on GTFEE shows significant regional heterogeneity, with its effect weakening from the eastern coastal areas to the western inland regions.

1. Introduction

In 2015, the United Nations Sustainable Development Summit released 15 sustainable development goals, identifying improving energy efficiency as a key lever for sustainable development [1,2]. However, as one of the world’s highest greenhouse gas-emitting economies, the United States Energy Information Administration stated that fossil fuel combustion for energy accounted for 74% of the total US GHG emissions [3]. Similarly, European Union countries face significant energy consumption and environmental pollution challenges, with fossil fuels still dominating their energy mix, accounting for over 70% of the total energy supply in some member states. This integrated policy landscape underscores why GTFEE improvements serve as a critical climate change mitigation strategy, enabling the operationalization of emission reduction targets into concrete implementation pathways.

Improving green total factor energy efficiency is an important issue that needs to be solved urgently to realize green development in China’s economy, from the stage of high-speed growth to the stage of high-quality development [4]. With China’s rapid economic growth, the energy-intensive and inefficient development model has resulted in significant environmental pollution issues [5]. Based on the BP World Energy Statistical Yearbook 2022, China leads, consuming 26.5% of the world’s total energy globally. China’s energy structure is dominated by coal, with the total CO₂ emissions related to energy reaching as high as 10.588 billion tons, accounting for 31.2% of the global CO₂ emissions. Therefore, improving GTFEE stands as an essential decision for China’s high-caliber economic advancement [6], which can not only alleviate the problem of high energy consumption and low efficiency, but is also the primary way to reduce ecological pressure.

Due to the growing contradiction between energy demand and the accelerated pace of energy structure transformation, China has taken many measures to accelerate the energy revolution. Currently, the world has several important issues to address: high energy demand, the massive depletion of non-renewable energy resources, and the seriousness of local and global environmental pollution. To solve these problems, it is necessary to study the energy situation at this stage. Numerous scholars have researched the influencing factors of GTFEE and measures to improve GTFEE, including innovating energy production technologies [7], formulating energy and environmental policies [8], changing consumption patterns [9], enhancing energy conservation awareness [10], and so on. Nevertheless, while examining the factors of GTFEE, scholars have previously largely overlooked the role of educational competitiveness (Ec) in influencing GTFEE.

Indeed, Ec may significantly enhance urban GTFEE. Specifically, educational competitiveness positively affects GTFEE in three main ways: the optimization of human resource allocation [11], the improvement of technological innovation [12], and energy conservation awareness promotion [12]. Firstly, improving educational competitiveness produces highly qualified energy management professionals who drive the global transition to renewable energy and sustainable development. Secondly, researchers with quality education can promote energy technology innovation, effectively promote industry–university–research cooperation, and accelerate the transformation of energy research results into practical applications [13]. Finally, with the improvement of educational competitiveness, the awareness of energy conservation and the notion of sustainable development have now gained extensive dissemination, which is conducive to enhancing the public’s awareness of environmental protection, prompting citizens to adopt more efficient energy use behaviors, and contributing to the improvement of GTFEE at the social level [14]. Furthermore, the talent cultivated through enhanced educational competitiveness actively drives innovation in energy-related products and services. This fosters the emergence of new industries and markets, facilitating the transformation of traditional energy sectors into advanced, intelligent, and environmentally sustainable structures. Therefore, our study identifies three specific pathways through which Ec affects energy efficiency: green technology innovation, the optimization of human resource allocation, and industrial structure upgrading.

Therefore, is there a contribution of educational competitiveness to GTFEE? What are the potential mechanisms of its impact? Is there regional heterogeneity in the impact of educational competitiveness on GTFEE? Entering the new era of high-quality development, these problems in education and energy in China are worth studying. Thus, we address these questions by examining how Ec improves urban GTFEE in China.

Using panel data from 20 major Chinese cities, this study calculates the Ec index for these urban centers. Then, to investigate the causal link between Ec and urban GTFEE, we employ a two-way fixed effects model. We construct a mediating effect model, which enables us to quantify the indirect effects transmitted via three critical pathways: green technology innovation, human resource mismatch, and industrial structure upgrading. The results demonstrate that Ec directly improves urban GTFEE, with education resource competitiveness, input competitiveness, output competitiveness, and sustainable competitiveness all exhibiting statistically significant direct effects. Furthermore, this study investigates the underlying mechanisms through which Ec enhances GTFEE, including pathways of green technological innovation, human resource allocation optimization, and industrial structure upgrading. The study also examines the regional heterogeneity of these effects across different areas in China. The findings offer valuable insights for policymakers to leverage educational development as a strategic tool for promoting green transformation in urban energy systems.

This paper presents the following contributions: firstly, by gathering panel data from central Chinese cities for the years from 2012 to 2022, we construct an educational comprehensive index from five dimensions: education resources, education input, education scale, education output, and education sustainable development. The educational competitiveness index is assessed by calculating the weights of the indicators using the objective entropy value assignment method. Secondly, we investigate the causal link between Ec and GTFEE from the standpoint of the educational level. The study not only analyzes the direct effect of Ec on GTFEE in Chinese cities but also analyzes the heterogeneity in terms of the multidimensionality of Ec, as well as the regional geographic location. Thirdly, we delve into the mechanism of the impact of Ec on GTFEE. We employ a mediating model to uncover how it impacts Ec regarding GTFEE, while also examining the possible effects of green technology innovation, misalignment in human resources, and the enhancement of industrial structure. Drawing upon the research outcomes, this study proposes policy recommendations across three dimensions to improve Ec and foster sustainable energy development in urban areas.

2. Literature Review

Ec refers to the sustainable comparative advantages and capabilities demonstrated by a country’s or region’s overall educational development level and strength in competition with others. By nature, it falls within the domain of comprehensive national competitiveness. The advancement of Ec directly influences a nation’s scientific and technological competitiveness, as well as its corporate competitiveness, thereby serving as a critical factor in shaping comprehensive national power (or regional competitiveness) and long-term economic growth trends [15]. This has prompted scholars to pay close attention to its related research. In studies on the selection of indicators of Ec, most of the existing literature uses indicators such as the number of students [16,17], enrollment [18], education expenditure [19], and education quality [20], among other indicators. However, these evaluation indicators often measure only one or several aspects of Ec, and there is an obvious lack of systematization in the selection of indicators, making it difficult for the indicator system to reflect the real situation of Ec.

To assess Ec metrics, most researchers employ the TOPSIS approach to evaluate higher Ec levels [21,22]. However, the TOPSIS method usually defaults to the indicators being independent of each other and does not fully consider the possible correlation between the indicators, and this omission may lead to an incorrect assessment, so that the evaluation results can show a bias. In addition, some studies have used the analytic hierarchy process to decide the weights of the indicators [23], thus calculating the Ec index [24]. However, the AHP is greatly influenced by subjective factors, and the weights of the indicators depend on the subjective judgment of the experts, leading to insufficiently objective assessment results. Throughout the existing literature, most research assesses the Ec at the provincial level [25,26], but there is a notable lack of research at the municipal level. Furthermore, the current research is limited to higher education and fails to assess the competitiveness of the overall education system.

For the factors affecting GTFEE, existing studies are mainly attributed to technological innovation [27], industrial structure [28], environmental regulation [29], and trade policy. For example, Xue et al. [30] propose that industrial structure does have a promotional impact on GTFEE, but it is characterized by phases. Fan et al. [31] indicate that environmental regulations positively influence GTFP, with local regulations also playing a part in the GTFP of adjacent regions. A reduction in trade policy uncertainty can enhance GTFEE through two key channels: export expansion and technological progress acceleration. While the existing literature has examined various determinants of GTFEE from multiple perspectives, the role of education—a critical yet underexplored factor—remains insufficiently addressed in the current discourse [32].

In summary, the existing literature suffers from some shortcomings: first, a comprehensive evaluation index for urban Ec remains unformed in the urban dimension. Most scholars measure Ec in terms of one aspect of Ec, such as educational input or educational scale. Second, most of the studies center on the field of higher education and lack measurement of the entire education system. Due to the limitations of the research sample, the available literature on the Ec assessment is limited to the national, regional, or provincial level and lacks more detailed research at the prefecture level. Third, current research has overlooked the relationship between Ec and GTFEE, failing to explore how Ec improves GTFEE.

3. Research Hypothesis

This research concludes that Ec can notably boost urban GTFEE and exert an impact on GTFEE via three influencing channels: green technology innovation, human resource mismatch, and the upgrading of industrial structure, as presented in Figure 1.

Educational competitiveness plays an irreplaceable role in improving energy efficiency by providing human resources support [32], promoting energy technology innovation [33], and fostering public awareness of energy conservation [34]. This multidimensional influence has been widely recognized in the sustainability literature, where education is increasingly seen as a foundational driver of energy transitions [35]. First, according to the human capital theory, the enhancement of Ec can cultivate high-quality human resources with high skills and innovative capabilities. These high-quality human capitals can not only promote the creation of energy-saving technologies, but also effectively absorb and utilize new technologies, thus improving GTFEE [36]. Second, improving Ec also attracts enterprises to cooperate with universities and educational institutions. This model of cooperation not only enhances the usefulness of energy technologies but also accelerates the dissemination, adaptation, and integration of novel technological advancements within relevant fields in the market [37]. Third, education serves as a crucial function in cultivating the public understanding of energy conservation and sustainable development. By raising awareness, it encourages individuals to actively engage in environmentally friendly practices in their everyday lives, ultimately steering society toward a greener, low-carbon, and more sustainable future [38]. Following the analysis presented, we suggest:

Hypothesis 1.

Ec can significantly improve urban GTFEE.

Aside from having an immediate influence, Ec may also have an indirect influence on GTFEE by way of environmentally conscious innovation, mitigating human resource mismatch, and industrial structural upgrading. According to the endogenous growth theory, technological progress is a central factor influencing GTFEE [39]. Green technology innovation serves as a distinct form of environmentally biased technical advancement that enhances GTFEE and fulfills the “dual-carbon” objectives [40]. The improvement of Ec can promote the transformation of educational achievements into innovative outputs, produce more efficient and low-carbon green technologies, and effectively improve GTFEE in all aspects of production, transformation, and application [41]. Citizens’ green awareness and environmental protection consciousness could increase with the improvement of Ec, so that they are more capable of accepting green energy technologies and promoting the popularization of green technologies [42]. Therefore, we propose:

Hypothesis 2.

Ec enhances urban GTFEE through green technological innovation.

The workforce’s quality, the mechanism for sharing education information, and the degree of demand matching between education and the market are key factors affecting the mismatch of human resources [12]. When urban education competitiveness is minimal, the quality of the labor force is relatively low, and cities may also neglect the construction of information-sharing mechanisms in the process of the development of education, making the information between the education sector and the employment sector closed. This leads to the limited knowledge and skill level of the workers themselves, and due to the lack of information, they cannot understand the labor requirements of the market promptly; therefore, job mismatches occur, leading to the waste of human resources in the whole society [43]. On the contrary, as urban education competitiveness improves, workers acquire specialized knowledge and core skills in specific fields and are better adapted to the requirements of their jobs. At the same time, information sharing and cooperation between the education and employment sectors have been strengthened, so that workers can keep abreast of market demand, leading to a rational allocation of human resources. On this basis, industries that pay more attention to GTFEE, such as emerging industries and high-end manufacturing industries, can be adapted to relevant specialized talents. These industries adopt advanced technologies and management modes, which in turn improves productivity accompanied by the enhancement of energy utilization efficiency [44]. Then, we propose:

Hypothesis 3.

Ec improves urban GTFEE by mitigating human resource mismatch.

The improvement of Ec can effectively promote industrial structure upgrading, which can significantly improve GTFEE [45]. Firstly, improving Ec is a key way to enhance the quality of the labor force [46]. Schools and research institutions, as important vehicles for education, produce a large number of highly qualified workforce professionals who are committed to innovation in energy technologies, products, and services. The continuous emergence of new products and services promotes the generation of new industries and markets and drives the shift of conventional sectors toward advanced, intelligent, and green [47]. Secondly, industrial structural upgrading is essentially a shift of resource factors from the inefficient part to the high-efficiency sector, so that the proportion of the high-efficiency sector will increase, and ultimately the productivity of all industries could be jointly improved [48]. According to the resource allocation theory, limited resources are more reasonably allocated to different areas, so that equal amounts of energy factors can produce more economic output and enhance the efficiency of energy use. Therefore, we propose:

Hypothesis 4.

Ec indirectly affects urban GTFEE through industrial structural upgrading.

4. Research Design

4.1. Model Design

4.1.1. Baseline Model

To determine if Ec improves urban GTFEE, we construct a two-way fixed effects panel model as follows. This model is employed to establish the fundamental relationship between Ec and urban GTFEE. By incorporating both city- and year-fixed effects, it isolates the net impact by accounting for time-invariant regional characteristics and temporal trends common to all cities. This approach aligns with standard practice in GTFEE studies and provides a robust foundation for the subsequent mechanism analysis.

$$GTFE{E}_{jk}={\beta }_0+{\beta }_{1}Ed{u}_{jk}+\lambda X_{jk}+{\mu }_j+{\gamma }_k+{\epsilon }_{jk}$$

(1)

where the explained variable is $GTFE{E}_{jk}$, and subscripted $j$ and $k$ mean the $k$ city in the $j$ year; $Ed{u}_{jk}$ is the core explanatory variant; $X_{jk}$ is the control variable, which includes the level of economic development (Dev), the level of financial development (Fin), the degree of fiscal decentralization (Caz), infrastructure (Rod), and international direct investment (FDI); ${\mu }_j$ denotes city-fixed effects, ${\gamma }_k$ denotes year-fixed effects; ${\beta }_0$ is the constant term and ${\beta }_1$ is the estimated coefficient for each variable; and ${\epsilon }_{\mathrm{jk}}$ denotes a random error.

4.1.2. Mediating Effect Model

To uncover how Ec enhances GTFEE, we examine three transmission channels—green innovation, human capital allocation, and industrial upgrading—using a three-step mediation framework. This method quantifies the indirect effects of Ec through these pathways while statistically validating their significance, offering a comprehensive understanding of the education–energy efficiency relationship. This dual approach ensures methodological rigor while generating findings with clear theoretical and policy relevance.

$$W_{jk}={\alpha }_0+{\alpha }_{1}Ed{u}_{jk}+{\alpha }_2{X}_{jk}+{\mu }_{2j}+{\gamma }_{2j}+{\epsilon }_{2j}$$

(2)

$$GTFE{E}_{jk}={\rho }_0+{\rho }_{1}Ed{u}_{jk}+{\rho }_2{M}_{jk}+{\rho }_3{X}_{jk}+{\mu }_{3j}+{\gamma }_{3k}+{\epsilon }_{3jk}$$

(3)

In this case, $W_{jk}$ represents three mechanism variables. According to the principle of impact mechanism modeling, when the coefficients are significant, the underlying mechanism is verified.

4.2. Variable Selection

4.2.1. Explained Variable

GTFEE serves as the dependent variable in this research. Drawing on the research results of Li et al. [28], we use the SBM model, which accounts for undesirable outputs, to assess GTFEE at the city level from 2012 to 2022. Three pollutants, namely SO_2, exhaust gas, and wastewater, are classified as undesirable output indicators. This study utilized linear programming software in MATLAB 2018a to compute the GTFEE index for each city using the specified linear equations.

$$\begin{array}{c}\min \rho ={\displaystyle \frac{1-(1/a){\displaystyle {\sum }_{i=1}^a{v}_i^{-}/x_{i0}}}{1+\frac{1}{v_1+v_2}({\displaystyle {\sum }_{r=1}^{v_1}{v}_r^d/y_{r0}^d}+{\displaystyle {\sum }_{t=1}^{v_2}{v}_t^{ud}/y_{t0}^{ud}})}}\\ x_0=X\lambda +v^{-}\\ y_0^d={\gamma }^d\lambda -v^d\\ y_0^{ud}={\gamma }^{ud}\lambda -v^{ud}\\ \lambda \ge 0,v^{-}\ge 0,v^{ud}\ge 0,v^d\ge 0\end{array}$$

(4)

In this modeling setup, there are $n$ decision units. Each decision unit is equipped with $a$ inputs, $v_1$ expected outputs, and $v_2$ unexpected outputs. The input vector is $x\in R^m$, the desired output vector is $y^d\in R^{v_1}$, and the undesired output vector is $y^{ud}\in R^{v_2}$. Meanwhile, the vector $v^d$ is used to represent the shortage of desired outputs, and the vectors $v^{-}$ and $v^{ud}$ reflect the redundancy in inputs and undesired outputs, respectively. The objective function $\beta$ of this model takes a range of values in a specific interval $\left[0,1\right]$. The decision unit is SBM, and is valid if and only if the objective function $\beta$ value is 1 and $s^{-}$ = $s^d$ = $v^{ud}$ = 0. If the objective function $\beta$ value is less than 1, it means that the corresponding decision unit is not effective, and its input–output ratio can be further improved.

4.2.2. Core Explanatory Variable

The independent variable of this study is Ec. The study is based on the CIPP evaluation model [49], which includes the four dimensions of context evaluation, input evaluation, process evaluation, and product evaluation n [50], and innovatively adds the dimension of sustainable development, constructing a city education competitiveness index containing the five secondary indicators of education resources, input, scale, output, and sustainability, as well as 16 tertiary indicators. The weight values are calculated using the combination of the AHP subjective assignment method and the objective entropy value assignment method, and the education competitiveness index of each city is calculated by multiplying the weights.

Table 1. The weights of the urban education competitiveness index system.

Level 1 Indicator	Level 2 Indicators	Level 3 Indicators	References	Direction of Indicators	Indicator Weights
Urban education competitiveness	Education resource competitiveness	Books per 100 People in Public Libraries (Volume)	Lynch and Baine (2004) [51]	+	6.32%
		Teacher–student ratio (%)	Asfahani (2023) [52]	−	3.79%
		Qualified Full-Time Teacher Rate (%)		+	6.89%
	Education input competitiveness	Computers per Student (Units)	Liu and Xu (2017) [53]	+	6.46%
		Education Research and Experimental Development (R&D) Expenditure (10,000 Yuan)	Zanzig (1997) [54]	+	9.30%
		The proportion of Fiscal Education Expenditure in the Government Fiscal Expenditure (%)	Mahajan and Golahit (2020) [55]	+	6.14%
	Education scale competitiveness	Total number of schools at all levels and of all types (number of schools)	Latif and Marimon (2019) [56]	+	6.90%
		Number of full-time teachers at all levels and in all types of schools (persons)	Felgueira and Rodrigues (2020) [57]	+	4.19%
		Average students per 100,000 population at all levels of education (persons)	Segarra and Segarra (2016) [58]	+	5.43%
	Education efficiency and output competitiveness	Number of persons with higher education as a proportion of the city’s total population (%)	Mu and He (2024) [59]	+	5.51%
		Graduation rate at all levels and in all types of education (%)	Klumpp (2018) [60]	+	5.71%
		Number of years of education per capita (years)	Martínez-Campillo and Fernández-Santos (2020) [61]	+	3.76%
		Employment rate of people over 16 years of age (%)		+	7.99%
	Education sustainable development competitiveness	Sustainable Competitiveness of Education growth rate of education expenditure (%)	Brudermann et al. (2019) [62]	+	5.75%
		Contribution rate of science and technology (%)	Lai and Peng (2020) [63]	+	8.75%
		Talent contribution rate (%)	Krstić et al. (2020) [64]	+	7.11%

Note: The Third Plenary Session of the Twentieth Central Committee of the Communist Party of China underscored that “education, science and technology, and human resources are fundamental and strategic pillars for Chinese-style modernization.” The advancement and sustainable growth of education are inextricably linked to the contributions of science and technology, as well as the support of skilled individuals. We have developed a secondary index for “sustainable competitiveness of education” to assess the impact of science, technology, and talent on educational advancement. The “contribution rate of science and technology” serves as an indicator to assess the impact of science and technology on education, while the “number of invention patents granted per 10,000 individuals” functions as its proxy variable. The “talent contribution rate” is selected as a metric to assess the impact of skills on education.

displays the final constructions.

4.2.3. Mediating Variables

Green technology innovation (Gt). Referring to Yang et al. [65] and Gao et al. [66], we apply the number of green patents granted to express and take the logarithm to calculate the green technology innovation index.

Human resource mismatch (Hm). We measure the urban labor mismatch index, referring to Gao et al. [66]. If the index is negative, it indicates that human resource mismatch has an inhibitory effect on urban GTFEE.

Industrial structure upgrading (Iu). We refer to Muhammad et al. [67], Zhang et al. [68], and Gao et al. [69], who apply the ratio of the tertiary industry output to the secondary industry output as a metric.

4.2.4. Control Variables

In this paper, we aimed to minimize endogeneity issues by controlling for other variables that may have significant impacts, according to the studies of Chen et al. [70], Gao et al. [69], and Kendall et al. [71]. The study incorporated the subsequent control variables: (1) economic development level (Dev), indicated by GDP per capita; (2) financial development level (Fin), assessed through the year-end balance of deposits and loans of financial institutions relative to the city’s GDP; (3) fiscal decentralization (Caz), computed as the ratio of public revenues to public expenditures; and (4) infrastructure (Rod), quantified by per capita road mileage. International direct investment (FDI) is represented as a proportion of FDI relative to GDP.

4.3. Data Source

The research utilizes panel data from China’s central cities from 2012 to 2022 as the sample. Central cities have a stronger ability to integrate national educational resource elements and competitiveness in the region or even the country. Finally, data from Beijing, Shanghai, Wuhan, Shenzhen, Guangzhou, Nanjing, Hangzhou, Xi’an, Tianjin, Chengdu, Chongqing, Ningbo, Xiamen, Qingdao, Shenyang, Dalian, Changchun, Harbin, Zhengzhou, and Jinan, which are the 20 central cities, are selected as the research samples for this study. The data given in this study are from the National Bureau of Statistics of China “https://www.stats.gov.cn/english/ (accessed on 20 December 2024)”, the China Urban Statistical Yearbook, the China Regional Economic Statistical Yearbook “https://data.stats.gov.cn/ (accessed on 20 December 2024)”, and the China Education Statistical Yearbook “http://www.moe.gov.cn/ (accessed on 20 December 2024)”. Carbon emissions are taken from the “CEADs—China Carbon Accounting Database “https://www.ceads.net.cn/ (accessed on 20 December 2024)”.

Table 2. Descriptive statistics.

Variables	N	Mean	Median	S.D	Min	Max
GTFEE	220	0.672	0.652	0.144	0.374	1.000
Ec	220	0.359	0.352	0.123	0.141	0.693
Dev	220	11.288	11.308	0.436	9.800	13.056
Fin	220	3.689	3.540	1.001	1.912	6.400
Caz	220	0.799	0.825	0.158	0.384	1.541
Rod	220	2.888	2.931	0.441	0.468	3.514
FDI	220	0.015	0.006	0.021	0.000	0.121

Note: This table presents the descriptive statistics for the variables in the sample; the dependent variable is GTFEE, the independent variable is Ec, and the control variables are economic development level (Dev), financial development level (Fin), fiscal decentralization (Caz), infrastructure (Rod), and international direct investment (FDI).

presents the descriptive statistics.

5. Empirical Results

5.1. The Baseline Regression Results

The baseline test results are presented in

Table 3. Baseline regression results.

	(1)	(2)	(3)	(4)
Variables	GTFEE	GTFEE	GTFEE	GTFEE
Ec	0.829 ***	0.811 ***	0.555 **	0.558 **
	(0.218)	(0.208)	(0.273)	(0.231)
Dev		0.026 ***	0.013 **	0.015 *
		(0.009)	(0.006)	(0.008)
Fin		0.011	0.009 *	0.005 **
		(0.015)	(0.005)	(0.002)
Caz			−0.031	−0.031
			(0.047)	(0.047)
Rod			−0.110 ***	−0.109 ***
			(0.039)	(0.035)
FDI				−0.130
				(0.467)
Constant	1.374 ***	1.274 ***	0.949 **	0.631 *
	(0.278)	(0.435)	(0.457)	(0.352)
City FE	√	√	√	√
Year FE	√	√	√	√
R2	0.420	0.570	0.598	0.628
N	220	220	220	220

Notes: This table presents the results obtained in Equation (1), which explores the impact of Ec on GTFEE. The explanations of the variables are the same as above. Column (1) presents the results of the effect of Ec on GTFEE. *** p < 0.01, ** p < 0.05, and * p < 0.10; standard errors are in parentheses.

. Column (1) displays the regression results without any control variables, whereas columns (2)–(4) provide the regression findings that included the selected control variables Dev, Fin, Caz, Rod, and FDI in sequence. The results in column (1) show that the regression coefficient of Ec on GTFEE is 0.829, and it is significant at the 1% level. This indicates that Ec can boost in improvement of GTFEE. The results in columns (2)–(4) indicate that after including the control variables Dev, Fin, Caz, Rod, and FDI, the coefficient of Ec for GTFEE remains significant. Hypothesis 1 is preliminarily verified. This may be because a competitive education system produces talented people with an international outlook, who can participate in international energy project cooperation, academic exchanges, and the introduction of advanced energy concepts and technologies from abroad [72,73].

The control variables in this paper generally have an impact on GTFEE. Among them, Dev (economic development level) and Fin (financial development level) are positive at the 10% level, indicating that both have a promoting effect on GTFEE at the urban level. However, Rod (infrastructure, measured by per capita road mileage) is significantly negative at the 1% level, indicating that it has an inhibitory effect on GTFEE. This is in line with the findings of Li et al. [28].

Consequently, we regress each of the five secondary indices of Ec against GTFEE, with the findings presented in

Table 4. The results of the second-dimensional educational competitiveness analysis.

	(1)	(2)	(3)	(4)	(5)
	GTFEE	GTFEE	GTFEE	GTFEE	GTFEE
Resources	0.177 *
	(0.106)
Input		0.261 **
		(0.103)
Scale			0.206
			(0.173)
Efficiency				0.393 ***
				(0.111)
Devp					0.133 **
					(0.064)
Constant	1.262 ***	0.746 *	0.939 **	1.110 ***	0.863 *
	(0.433)	(0.397)	(0.445)	(0.385)	(0.446)
Control	√	√	√	√	√
City FE	√	√	√	√	√
Year FE	√	√	√	√	√
R2	0.324	0.431	0.223	0.427	0.524
N	220	220	220	220	220

Notes: This table presents the findings regarding how five secondary indicators of Ec improve GTFEE. The five secondary indicators are education resources competitiveness (Resources), education input competitiveness (Input), education scale competitiveness (Scale), education efficiency and output competitiveness (Efficiency), and education sustainable competitiveness (Devp). Columns (1)–(5) present the results of the above five secondary indicators on GTFEE. All other variables and related explanations are the same as above. *** p < 0.01, ** p < 0.05, and * p < 0.10.

. Columns (1)–(5) present the regression outcomes for educational resource competitiveness, input competitiveness, scale competitiveness, efficiency competitiveness, and sustainable competitiveness concerning GTFEE, respectively. Table 4 indicates that the coefficient for column (1) Resources on GTFEE is 0.177, which is significant at the 10% level, suggesting a substantial positive correlation between Resources and GTFEE; thus, the competitiveness of educational resources enhances urban GTFEE. The plausible explanation is that greater competition in educational resources correlates with an abundance of energy-related professional disciplines, professors, and research resources, which collectively establish a robust basis for the city’s energy development.

In column (2), the coefficient for Input concerning GTFEE is 0.261, demonstrating significance at 5%. This finding suggests that Input is positively correlated with GTFEE, meaning that educational input plays a role in enhancing urban GTFEE. This could be attributed to the fact that increased investment by the city in educational research provides universities and research institutions with greater financial resources to pursue innovative research in the energy sector. As a result, they are better equipped to develop efficient and eco-friendly energy technologies and equipment.

In column (4), the coefficient for Efficiency concerning GTFEE is 0.393. This shows a notable positive relationship between Efficiency and GTFEE, indicating that educational efficiency and output competitiveness can significantly improve GTFEE. The possible reason is that high-quality educational efficiency and output competitiveness allow the results of educational research to be transformed into practical energy innovation technologies and energy-saving products. For example, new energy technologies and equipment, such as efficient solar cells, wind turbines, energy storage systems, and so on, can enhance the efficiency and stability of new energy use [74].

In column (5), the coefficient for Devp concerning GTFEE is 0.133, with significance at 5%. This shows a strong positive relationship between Devp and GTFEE, suggesting that the concept of education sustainable competitiveness is crucial for enhancing GTFEE in cities. This may be because a high level of sustainable competitiveness in education feeds the city with sustainable educational concepts, content, and teaching methods [67], making the whole education system high-quality and long-lasting, and providing a sustainable impetus to the city’s energy future.

Nevertheless, the coefficient of Scale is not significant at 0.206. This may be attributed to certain cities blindly expanding the scale of education while neglecting the quality of education, which lacks effectiveness in contributing to urban GTFEE. Furthermore, Scale does not necessarily contribute to improving urban GTFEE. Overall, Ec significantly improves GTFEE.

5.2. Mechanism Analysis

Based on previous studies, Ec can boost urban GTFEE through three unique pathways: green technology innovation (Gt), human resource mismatch (Hm), and industrial structure upgrading (Iu). According to Gao et al. [66], this study examines the proposed mechanisms based on the mediating effect model described above.

First, the initial two columns of

Table 5. The results of the mediating effect models.

	(1)	(2)	(3)	(4)	(5)	(6)
Variables	Gt	GTFEE	Hm	GTFEE	Iu	GTFEE
Ec	0.121 ***	0.062 **	−0.165 **	0.158 **	0.512 ***	0.238 ***
	(0.042)	(0.031)	(0.066)	(0.079)	(0.150)	(0.085)
Gt		0.103 **
		(0.048)
Hm				−0.254 **
				(0.121)
Iu						0.288 ***
						(0.096)
Constant	2.159 ***	1.011 ***	1.653 ***	1.226 **	1.238 ***	0.593 *
	(0.443)	(0.314)	(0.274)	(0.523)	(0.316)	(0.311)
Control	√	√	√	√	√	√
City FE	√	√	√	√	√	√
Year FE	√	√	√	√	√	√
R2	0.370	0.471	0.264	0.496	0.398	0.539
N	220	220	220	220	220	220

Notes: This table displays the outcomes calculated using Equations (2) and (3). These two equations investigate the pathways through which Ec affects GTFEE via green technology innovation, human resource mismatch, and industrial structure upgrading. The mediating effect variables are green technology innovation (Gt), human resource mismatch (Hm), and industrial structure upgrading (Iu), all other variables and related explanations are the same as above. The results for Gt, Hm, and Iu as channels of influence are presented in columns (1)–(6). *** p < 0.01, ** p < 0.05, and * p < 0.10.

display results of the mechanism test on green technological innovation. The coefficient of the Ec on Gt in column (1) is 0.121, which indicates that Ec is positively correlated with the mechanism variable Gt, that is, enhancing Ec can facilitate the advancement of Gt. In column (2), both regression results for Ec and Gt are positive and significant at least at the 5% level. Therefore, Ec can promote Gt, thereby improving urban GTFEE. Gt serves as a significant medium of impact, thus confirming the validity of Hypothesis 2.

Second, we argue that there is an inhibitory effect of Hm on GTFEE. The results presented in column (3) of Table 5 indicate that the coefficient of Ec on GTFEE is −0.165, demonstrating a significant negative relationship at 5%. Similarly, the coefficient of Hm on GTFEE is −0.254, which also holds statistical significance at 5%. This indicates that the mechanism variable Hm is significantly negatively related to GTFEE, meaning that Hm hinders the improvement of urban GTFEE, and Ec can alleviate Hm and thus improve urban GTFEE. This may be because Hm may lead to excess or insufficient labor in energy-intensive industries, affecting the effective use of energy. Nonetheless, Ec has significantly enhanced the efficiency of human resource allocation by optimizing the structure of talent training. This, in turn, has played a key role in boosting operational efficiency [75]. Consequently, Hypothesis 3 has been affirmed.

We propose that Iu is an effective means through which Ec improves GTFEE. The final columns of Table 5 display the outcomes of the testing factors related to the mechanism Iu. In column (5), the coefficient of Ec on Iu is 0.512, indicating that Ec facilitates the enhancement of Iu. In the last column, the coefficients for Ec and Iu are 0.238 and 0.288, respectively, both of which are significant at the 1% level. Therefore, Hypothesis 4 has been validated. This indicates that Ec can promote the transition of conventional energy-intensive businesses to technology-intensive green economy sectors through improved optimization and innovation [76], thus improving GTFEE.

5.3. Robustness Test

To check the robustness of the results, we use the following four methods.

Table 6. Robustness test results.

	(1)	(2)	(3)	(4)
Variables	EBM_GTFEE	GTFEE	GTFEE	GTFEE
Ec	1.157 ***	0.406 ***	0.319 **	0.773 **
	(0.293)	(0.235)	(0.140)	(0.310)
L.GTFEE				0.085 **
				(0.034)
Constant	1.402 ***	2.118 ***	1.693 ***	0.976 **
	(0.286)	(0.642)	(0.535)	(0.459)
Control	√	√	√	√
City FE	√	√	√	√
Year FE	√	√	√	√
R2	0.495	0.596	0.513	0.609
N	220	160	176	200

Note: This table shows the results of the robustness tests, all other variables and related explanations are the same as above. Columns (1)–(4) show the results of the four tests, respectively. *** p < 0.01, and ** p < 0.05.

includes the test results.

5.3.1. Replacing the Explained Variables

We replace the measure of the dependent variable. The efficiency based measure (EBM) model is used to regress GTFEE, and the results are shown in Table 6. The EBM model innovatively combines both radial and non-radial approaches within a unified framework, thereby addressing the limitations of traditional DEA models that typically consider only one of these dimensions. This hybrid characteristic allows the EBM model to simultaneously account for proportional changes (radial component) and specific input/output adjustments (non-radial slacks), making it particularly suitable for measuring complex efficiency concepts. This hybrid framework enables the precise measurement of energy efficiency by simultaneously optimizing multiple dimensions of energy inputs and outputs. After modifying the calculation method of GTFEE, the coefficient of Ec on GTFEE is 1.157. This indicates that Ec competitiveness has a positive contribution to GTFEE, thus reinforcing the validity of the previous conclusion.

5.3.2. Excluding the Impact of the COVID-19 Pandemic

The COVID-19 pandemic caused unprecedented disruptions to economic activities and education systems worldwide, potentially distorting normal patterns of educational competitiveness and green development. Due to the large influence of the COVID-19 pandemic, the data for the years from 2020 to 2022 are therefore not included in this study. The competitiveness of education is re-evaluated concerning urban GTFEE, utilizing panel data from 2012 to 2019 in column (2) of Table 6. At the 1% level, the Ec GTFEE coefficient of 0.406 is statistically significant. These results align with those addressed in the preceding section.

5.3.3. Excluding the Special Sample

China’s four directly governed municipalities (Beijing, Tianjin, Shanghai, and Chongqing) enjoy distinct political and economic advantages as national-level administrative units. These cities typically receive preferential resource allocations in terms of fiscal policies, infrastructure investment, and talent attraction due to their unique administrative status, which may significantly distort regional comparisons in our analysis. To address this potential bias, we exclude these four municipalities and re-estimate the model using the remaining sample. The results in column (3) show a regression coefficient of 0.319, which is again significant in the 1% range, confirming the robustness of the empirical results.

5.3.4. Replacement Econometric Model

To better capture the temporal dynamics and address potential path dependency in the relationship between Ec and GTFEE, we employ a dynamic system GMM model for re-estimation. This approach not only accounts for the cumulative inertia observed in their co-evolution but also effectively mitigates the endogeneity concerns arising from reverse causality and omitted variable bias. Column (4) of Table 6 shows the finding that the variable Ec continues to hold a significantly positive relationship with GTFEE, maintaining significance at the 5% level. This strengthens the reliability of the earlier empirical results.

5.4. Heterogeneity Analysis

Owing to the uneven levels of economic development and resource endowment in different regions, Chinese cities exhibit diverse energy development levels. In addition, regional external conditions, such as educational resource endowment, innovation atmosphere, and market environment, should also be combined to promote Ec to empower the improvement of GTFEE according to local conditions. Therefore, according to Yi et al. [77], we categorize the 20 central cities in our study sample according to their regional geographic locations: coastal, inland, eastern, central, and western cities. The investigation of the variability of Ec in improving GTFEE across cities in different regions is further examined. The results of these analyses are encapsulated in

Table 7. The results of the heterogeneity analysis.

	(1)	(2)	(3)	(4)	(5)
	Coastal	Inland	Eastern	Central	Western
Score	0.829 ***	0.301	0.566 ***	0.267 **	−0.125
	(0.302)	(0.252)	(0.182)	(0.127)	(0.311)
Constant	1.278 ***	1.514 ***	1.069 ***	1.099 ***	0.953 **
	(0.411)	(0.342)	(0.207)	(0.237)	(0.397)
Control	√	√	√	√	√
City FE	√	√	√	√	√
Year FE	√	√	√	√	√
R2	0.624	0.274	0.522	0.335	0.184
N	155	65	143	44	33

Notes: Variables and related explanations are the same as above. *** p < 0.01, and ** p < 0.05.

.

Table 7 indicates that the coefficient for coastal cities, presented in column (1), is 0.892 and is statistically significant at the 1% level. This research shows that heightened educational competition in coastal communities correlates with significant enhancements in GTFEE. The coefficient for central cities, as shown in column (4) of Table 7, is 0.267 and is statistically significant at the 5% level. The eastern coastline cities and central cities of China have attained substantial economic growth and possess superior educational resources. The advantageous educational environment has led to the emergence of a significant number of exceptional talents. The density of universities, research institutions, and business R&D facilities in this region facilitates the application and realization of scientific research findings, thus advancing green technology innovation [78] and the optimization of industrial structure. Consequently, this enhances the city’s GTFEE.

Nonetheless, the coefficients for column (2) inland cities and column (5) western cities lack significance. The observed disparity in the impact of Ec on GTFEE between central and coastal cities may be due to the relatively underdeveloped economic status of the inland and western regions compared with the eastern coastal regions. This has led to difficulties in maintaining educational quality, and governmental policy support for education and GTFEE has been insufficient to fulfill the necessary standards for improving GTFEE. As a result, Ec has not matched that of the coastal and eastern regions in enhancing the city’s GTFEE.

6. Discussion

Based on the above empirical findings, this study reveals that Ec significantly enhances urban GTFEE. Specifically, the competitiveness of educational resources, educational investment, and educational efficiency and output, as well as educational sustainability, all demonstrate statistically significant positive effects on GTFEE improvement. These promoting effects are primarily mediated through three key channels: green technology innovation, human resource allocation optimization, and industrial structure upgrading. Furthermore, the impact of Ec on GTFEE exhibits notable regional heterogeneity, with varying degrees of influence observed across different geographical areas.

Compared with the existing literature, the similarity between our article and Guan et al. [79] and Dong et al. [80] lies in the heterogeneity analysis results. Both studies found that there were significant regional differences in the impact of education on energy efficiency, especially between the eastern and western regions. However, unlike most research viewpoints, such as Ayadi et al. [81], we found that the impact of educational scale competitiveness on energy efficiency was not significant, which might be related to the characteristics and development stage of educational resource allocation in China. It is indicated that simply expanding the scale of education may not directly improve energy efficiency. Instead, more attention should be paid to the quality of education and structural optimization [82].

Compared with the existing research, the original contribution lies in the following three points: first, departing from conventional approaches that examine isolated educational metrics (e.g., enrollment rates or expenditure), our comprehensive evaluation indicator system based on the CIPP model systematically integrates multiple dimensions of education resources, inputs, efficiency and outputs, and sustainability, providing a new framework for assessing the systemic impacts of Ec on GTFEE. Second, while the prior literature has extensively investigated technological innovation, environmental regulation, and trade policies as drivers of GTFEE, this study establishes Ec as a critical yet underexplored determinant. By empirically validating its direct and mediated effects, we broaden the theoretical boundaries of GTFEE research. Finally, these evidence-based approaches can help nations to tailor education reforms to their developmental stages, harnessing education’s potential to contribute to global sustainable energy development.

Although this study comprehensively analyzed the relationship between Ec and urban GTFEE, there are still shortcomings. In the evaluation of the competitiveness of urban education, we only measured the index of overall Ec and did not make a more detailed assessment of the different education stages. Moreover, owing to the constraints imposed by the study sample, data were collected from only 20 central cities in China for the empirical analysis. Future studies are recommended to encompass a more extensive range of urban areas.

7. Conclusions

Utilizing panel data from 20 major Chinese cities spanning from 2012 to 2022, we investigate the influence of Ec on GTFEE using a two-way fixed effects model. We further investigate how different dimensions of Ec influence GTFEE. A mediation effects model analyzes the transmission mechanisms involving green technology innovation, human resource misallocation, and industrial structure upgrading. Additionally, we explore the geographical heterogeneity of these effects across different regions. The main conclusions are as follows: (1) Ec demonstrates a statistically significant positive effect on urban GTFEE enhancement; (2) the analysis of Ec dimensions indicates that resource competitiveness, investment competitiveness, efficiency–output competitiveness, and sustainability competitiveness all contribute to improved GTFEE, while scale competitiveness shows no significant effect; (3) the pathways analysis confirms the significant mediating roles of green technology innovation, human resource optimization, and industrial upgrading in the relationship between Ec and GTFEE; (4) the regional heterogeneity analysis shows a diminishing effect of Ec on GTFEE from China’s eastern coastal regions to western inland areas.

We also propose recommendations to enhance the role of Ec in advancing energy sustainability. First, China’s strategic shift from educational scale expansion to quality enhancement for GTFEE improvement, exemplified by the Ministry of Education’s “Carbon Neutrality Innovation Action Plan” and Tongji University’s operational reforms, presents a scalable model for global application. The plan’s focus on redirecting research funding to priority energy projects demonstrates how developing economies can optimize limited resources for maximum impact. Tongji’s campus energy reduction of 32% through its sustainability initiative offers practical insights into how institutional practices can serve as living laboratories for urban energy solutions. These approaches hold particular relevance for Southeast Asian and African nations that are undergoing rapid educational expansion while facing urgent energy transition demands. The Chinese experience shows that even resource-constrained systems can achieve meaningful progress through targeted investments and the integration of sustainability principles across all university operations.

Second, to fully realize Ec’s potential in boosting GTFEE, China must leverage the synergistic effects across three critical pathways: green technology innovation, human capital optimization, and industrial structure upgrading. The NIO-USTC Battery Lab exemplifies successful industry–academic collaboration in green technology development, while Guangdong’s vocational reforms address human capital mismatches in emerging green sectors. When combined with Tianjin’s industrial policy incentives, these initiatives demonstrate how coordinated interventions can accelerate sustainable energy transitions. This integrated approach offers valuable lessons for industrialized nations seeking to align their innovation ecosystems with net-zero targets, as well as for developing countries working to overcome skills gaps in renewable energy sectors. The key insight is that education’s impact on energy efficiency is multiplied when technological, human resource, and industrial policies are strategically aligned.

Finally, China’s regional differentiation strategy, featuring Peking University’s global MOOC program and Ningxia’s solar technician training, presents valuable insights for geographically diverse nations. The MOOC program’s success in reaching international learners suggests how digital education can accelerate global knowledge transfer in clean energy technologies. Ningxia’s localized training model shows how tailored vocational programs can address specific regional energy transition needs while creating employment opportunities. This balanced approach—combining high-technology knowledge dissemination with grassroots skills development—offers a template for large federal systems and archipelagic nations managing regional disparities. The strategies demonstrate how education policies can be spatially calibrated to maximize their GTFEE impact while promoting inclusive development.