The Impact of High-Speed Rail on High-Quality Economic Development: Evidence from China

Abstract

Utilizing data from 282 prefecture-level cities in China from 2005 to 2021, this study constructs an evaluation index system for high-quality economic development across the following five dimensions: innovation, coordination, green, openness, and sharing. A continuous difference-in-differences approach is employed for regression analysis to empirically examine the impact of high-speed rail on high-quality economic development, further exploring its mechanisms and spatial spillover effects. The findings reveal that (1) HSR significantly promotes high-quality economic development; (2) with the development of HSR, from 2005 to 2021, China’s high-quality economic development showed an evolutionary trend of overall improvement, with a gradual optimization of spatial patterns; (3) it facilitates high-quality economic development by enhancing capital and labor mobility, strengthening industrial chain resilience, and advancing industrial structure upgrading; (4) high-speed rail development in neighboring regions generates positive spatial spillover effects on local urban economic quality; and (5) the impact of high-speed rail on high-quality economic development exhibits significant heterogeneity across cities with different regions, tiers, scales, and resource endowments. These results confirm the positive role of high-speed rail in fostering high-quality economic development.

1. Introduction

The 2022 National Congress of the Communist Party of China, convened from October 16 to 22, highlighted that the contemporary socialist nation’s overall building requires prioritizing high-quality development. HQE essentially entails the harmonious integration of quantity and quality, encompassing economic, social, environmental, and innovative dimensions [1]. Economic growth does not equate to economic development [2]. In the four decades since the initiation of its reform and opening-up policy, China has witnessed remarkable economic expansion, but increasing resource and environmental constraints have rendered the traditional extensive development model unsustainable [3]. The pressure for green transformation is substantial, making it imperative to promote HQE [4].

Since the launch of the Beijing–Tianjin Intercity Railway in 2008, China’s HSR infrastructure has experienced swift expansion, amassing the most extensive operational mileage globally and accounting for nearly 67% of the international total [5]. This extensive rail network has bolstered regional interconnectivity [6]. With the implementation of the “Eight Vertical and Eight Horizontal” plan, China has fully embraced the “high-speed rail era” [7]. The increasing density of the HSR network has significantly boosted regional connectivity, effectively promoting the movement of people, goods, information, and capital along its routes [8]. This has injected new momentum into regional economic development, driving transformative improvements in economic activity quality, efficiency, and green transformation [9]. Beyond stimulating regional economic growth, HSR also curbs urban pollutants such as sulfur dioxide and wastewater, reduces carbon emissions, and enhances urban ecological efficiency [10], thereby influencing undesirable outputs in economic development and delivering the economic effects of both emission reduction and efficiency enhancement [11]. As a result, the HSR expansion has become a crucial driver for high-quality urban development [12].

Has more than a decade of HSR development spurred HQE? What are the mechanisms through which HSR drives HQE? This study explores several key questions, as follows: Does HSR still have a positive effect on HQE after accounting for other policies and confounding factors? Do regional differences exist in the impact of HSR on HQE? Does HQE create spatial spillover effects, and if so, what are the boundaries of these effects? To address these questions, this research constructs an evaluation system for HQE in China, covering the following five dimensions: innovation, coordination, green development, openness, and sharing. The HQE index for 282 cities from 2005 to 2021 is calculated using the entropy-weighted TOPSIS method. A continuous difference-in-differences model is employed to empirically examine the role of HSR in promoting HQE. By reducing spatial and temporal barriers, HSR accelerates the flow of labor and capital, enhances industrial chain resilience, promotes industrial structure upgrading, and improves green innovation levels, all of which contribute to HQE. The impact of HSR shows significant heterogeneity across regions, city tiers, transportation hub status, and resource endowments. It has a significantly positive effect on high-quality development in non-transportation hub cities, peripheral cities, and eastern and western regions, as well as growth-stage and mature resource-based cities. However, it shows no significant impact on growth-stage and renewable resource-based cities, central cities, and transportation hub cities, while exerting a significantly negative effect on declining resource-based cities. Additionally, HQE generates positive spatial spillover effects, with an attenuation boundary of 1000 km.

The primary contributions of this paper are mainly reflected in the following aspects: (1) innovatively selecting 33 effective indicators and employing the entropy-weighted TOPSIS method to measure the economic quality indices of 282 prefecture-level cities; (2) comprehensively validating the significantly positive impact of HSR on various mechanism variables, including labor mobility, capital mobility, industrial structure upgrading, and industrial chain resilience; (3) from the perspective of urban spatial connectivity, utilizing a continuous difference-in-differences model and a spatial Durbin model to comprehensively analyze the impact of HSR on urban HQE and its spatial spillover effects, with the spatial spillover boundary estimated at 1000 km, providing valuable insights for future large-scale HSR construction; and (4) to some extent, it enriches the research on HSR and HQE and provides an in-depth analysis of the spatial evolution characteristics of high-quality urban development.

The rest of the paper is structured as follows: the second section presents the literature review; the third section outlines the research method; the fourth section analyzes the empirical analysis; the fifth section presents the discussion of findings; and the sixth section concludes the paper.

2. Literature Review

The measurement of HQE indices in the existing literature is primarily divided into the following two approaches: single indicators and comprehensive evaluation index systems. Among these, single indicators often employ total factor productivity [13] and green total factor productivity [14] as proxy variables, while comprehensive evaluation index systems construct measurement frameworks from multiple dimensions. According to research, HSR’s introduction dramatically improves urban high-quality development [15]. The development of evaluation index systems has been examined by academics from some of the following angles: creating metrics for high-quality urban development that take into account infrastructure, social, ecological, and economic factors [16]; developing indicators for high-quality economic growth that are grounded in industrial restructuring, overall factor productivity, technological innovation, ecological environment, and enhancements in the standard of living for residents [17]; creating a HQE index based on the dimensions of development momentum, structural improvement, methodological strategies, and development results [18]; proposing a framework encompassing economic structure, economic scale, economic efficiency, and coordination [19]; assessing HQE through economic efficiency, stability, and sustainability; and innovatively constructing a HQE index based on the five development concepts of innovation, coordination, green development, openness, and sharing [20].

China’s interregional population mobility has been spatially constrained by natural geographical factors, leading to labor market segmentation and hindering the free allocation of high-skilled labor [21]. Geographical, cultural, and specialization differences across regions have resulted in heterogeneous labor under market segmentation [22]. The dual characteristics of HSR, namely “time–space compression” and “boundary breakthrough”, spatially dismantle labor segmentation caused by natural factors and expand the matching radius between firms and labor [23]. On the supply side, HSR lowers the expense of moving between regions for people [24] and increases the accessible distance for job seekers [25]. The integration and complementarity of heterogeneous labor with diverse knowledge and skills significantly optimize urban labor resource allocation and reduce frictional costs [26]. On the demand side, HSR expands urban boundaries, creating more job opportunities and increasing urban labor demand, thereby attracting higher-quality human capital [27]. Thus, HSR accelerates interregional population mobility [28], enhances labor matching efficiency [29], and fosters HQE through its resource allocation effects [30].

Green development necessitates stringent control of environmental pollution and the restoration of damaged ecosystems, objectives that rely heavily on green technologies [31]. Following the opening of HSR, the enhanced accessibility of cities has increased the mobility of innovation resources [11]. The cross-regional flow of high-level human capital induced by HSR [32] facilitates green innovation through technological advancements, knowledge spillovers, and efficiency improvements [33]. Furthermore, HSR strengthens interregional learning and collaboration, promoting the diffusion and sharing of intangible elements [34]. By elevating regional green innovation levels, it enhances urban GTFP [35] and accelerates the flow of innovation factors [36], amplifying communication and scale effects, thereby fostering regional green innovation and green transformation [37] and propelling high-quality economic growth in cities [38].

Following the inauguration of HSR, listed companies within cities have exhibited a greater propensity to establish subsidiaries in other regions, thereby facilitating cross-regional capital mobility [39]. The establishment of these subsidiaries not only increases fixed asset investments [40], but also drives local economic development through optimized resource allocation [41]. Transportation infrastructure induces the spatial redistribution of capital, altering the geographic distribution of economic activities and subsequently influencing the reduction in regional economic disparities [42]. Given that the per capita capital stock in developed regions significantly exceeds that in less developed areas, the law of diminishing marginal returns in neoclassical economic theory suggests that the return on capital in developed regions will be lower than that in less developed regions [43]. As improved transportation conditions reduce barriers to cross-regional capital flows, they promote the movement of capital from developed to less developed regions, thereby narrowing regional economic disparities [44].

The industrial chain is a linear structure composed of multiple enterprises at different production stages, interconnected through nodes, with various chains interwoven into a network. Damage to any node can have a cascading impact on the entire chain [45]. As a fundamental conduit for the circulation of products across industries, HSR directly facilitates intra-regional industrial division and collaboration by reducing the costs of product, factor, and technology flows, thereby enhancing regional production efficiency [46]. As a significant investment activity, building HSR encourages the growth of associated industries and increases regional industrial diversification [47]. Given that different industries exhibit varying demand elasticities and trade orientations [48], a diversified industrial structure can effectively mitigate and disperse external risks when faced with shocks, thereby strengthening the resilience of the industrial chain [49].

HSR primarily focuses on passenger transport, accelerating labor mobility in cities along its routes, particularly benefiting industries heavily reliant on human resources. Simultaneously, the expansion of HSR networks tends to increase rental costs and property prices [50], thereby reducing the proportion of manufacturing sectors sensitive to land prices and production costs and consequently lowering the share of the secondary sector in the economy while promoting the growth of the service industry. Employment in the service industry is greatly increased with the advent of HSR, such as tourism, accommodation, and catering within cities, but has a limited impact on non-service employment [51]. HSR notably boosts employment in the hospitality industry [52], and sectors that are intensive in information and knowledge, like finance, education, and tourism, gain more from the connectivity that HSR offers [53,54,55]. It is anticipated that HSR will facilitate industrial transformation in relevant cities.

3. Research Method

3.1. Variable Selection and Explanation

(1) Dependent variable: As shown in

Table 1. Construction of high-quality economic development indicators.

Primary Indicators	Second-Class Indicators	Third-Class Indicators	Basic Indicators	Unit	Type of Indicator
Innovation	Innovation input	Innovation input	R&D expenditure/GDP	%	+
			Science and technology expenditure/local general public budget expenditure	%	+
	Innovation output	Innovation output	Urban Innovation Index	-	+
		Patent applications granted per 10,000 persons	Number of patents granted/population of the region		+
	Innovative foundations	Scale of education	The count of students attending general higher education programs	Tens of thousands	+
		Investment in education	Share of education expenditure in general budget expenditure of local finance	%	+
	Innovation efficiency	Labor productivity	GDP/average annual number of employees	%	+
		Capital productivity	GDP/total investment in fixed assets	%	+
Coordination	Industrial structure	Rationalization of industrial structure	Percentage of tertiary sector output	%	+
		Advanced industrial structure	Tertiary sector output/secondary sector output	%	+
	Urban and rural coordination	The ratio of disposable income per capita for urban and rural residents	Urban disposable income per capita/rural disposable income per capita	%	-
		Urbanization rate	Permanent urban population/(permanent urban population + permanent rural population)	%	+
	Financial structure	Financial risk	Balance of deposits and loans/GDP	%	-
Greenness	Energy consumption	Electricity consumption per unit of output	Industrial Electricity Consumption/GDP	Kilowatt-hours/billion dollars	-
		Wastewater discharge per unit of output	Industrial wastewater discharge/GDP	Tons/billion dollars	-
		Exhaust emissions per unit of output	Smoke emissions per unit of output	Tons/billion dollars	-
		Emissions of smoke and dust per unit of output	Industrial soot and dust discharges/GDP	Tons/billion dollars	-
	Pollution emission	Haze pollution	Annual average PM2.5 concentration	μg/m3	-
Openness	Level of foreign trade	Degree of openness	Total Trade Imports and Exports/GDP	-	+
	Introduction of foreign capital	Effectiveness of openness	Total utilized foreign capital/GDP	-	+
	Tourism openness		International tourism revenue/GDP	-	+
	Tourism openness	International tourism revenue/GDP	Domestic tourism revenue/GDP	-	+
		Domestic tourism revenue/GDP	The number of inbound tourists received	10,000 people	+
Sharing	Income distribution	Per capita income	Real GDP per capita	-	+
		Remuneration for labor	Average wages of employees	Yuan	+
	Consumption level	Share of consumption	Social Retail Consumption/GDP	%	+
	Urban–rural sharing	Engel coefficient of urban households	Household food expenditure accounts for the proportion of urban consumption expenditure	%	-
		Engel coefficient of rural households	Household food expenditure accounts for the proportion of rural consumption expenditure	%	-
	Public services	Per capita expenditure on education	Education expenditures/total population	%	+
		Cultural resources	Public library collection per capita	Number of books/million people	+
		Health resources	Hospital bed count per 10,000 individuals	Number	+
		Employment effect	Registered urban unemployment rate	%	-
	Data sharing	Internet penetration	The number of internet broadband access users among 100 people	%	+

, the HQE index, which is derived through the entropy-weighted TOPSIS method, encompasses the following five key dimensions: innovation, coordination, green development, openness, and sharing.

(2) Core explanatory variable: The HSR variable is quantified by the count of HSR lines.

(3) Control variables: The level of human capital is indicated by the proportion of college and university students relative to the total year-end population; urban economic density is assessed by the ratio of regional GDP to the administrative land area; passenger turnover is determined by taking the natural logarithm of the aggregate passenger volumes from highways, waterways, and civil aviation; the level of transportation infrastructure is reflected by the natural logarithm of highway mileage; and the level of economic development is shown by the natural logarithm of regional GDP.

(4) Mechanism variables: Labor mobility is assessed using the natural logarithm of the number of employed individuals in urban units at year-end; capital mobility is represented by the natural logarithm of total fixed asset investment; industrial structure upgrading is calculated as the ratio of the added value of the tertiary sector to that of the secondary sector; industrial chain resilience is obtained from a composite index computed via the entropy weight method based on resistance recovery and transformation renewal indices; and green innovation is measured by the natural logarithm of the quantity of green invention patents issued in the current year.

3.2. Data

This study analyzes 282 prefecture-level cities from 2005 to 2021, encompassing 4794 observations, taking into account data availability and dependability. The China Statistical Yearbook and the City Statistical Yearbook are the sources of the variable data.

Table 2. Description of variables.

Variable Types	Variable Symbol	Variable Name	Description and Measurement Method
Dependent variable	HQE	High-quality economic development	Comprehensive index calculation
Explanatory variable	HSR	High-speed rail	Number of HSR lines opened
Control variables	HCI	Level of human capital	Logarithm of the number of employees at year end
	UED	Urban economic density	Logarithm of total fixed asset investment
	PTK	Passenger turnover	Logarithm of the sum of road passenger traffic, water passenger traffic, and civil aviation passenger traffic
	Road	Level of transportation infrastructure	Logarithm of highway mileage
	Economy	Level of economic development	Logarithm of regional gross domestic product
Mediating variables	Labor	Labor mobility	Logarithm of the number of employees at year end
	Capital	Capital mobility	Logarithm of total fixed asset investment
	Industrial structure	Industrial structure upgrading	Ratio of the added value of the tertiary industry to the added value of the secondary industry
	Resilience	Industrial chain resilience	Composite index of industrial chain resilience
	Green innovation	Green innovation	Logarithm of the number of green invention patents

offers definitions of the variables, while

Table 3. The primary variables’ descriptive statistics.

	N	Mean	Standard Deviation	Minimum	Maximum
HQE	4794	2.1865	1.8612	0.5477	12.4517
HSR	4794	0.7795	1.2323	0	9
HCI	4794	0.0201	0.0264	0	0.1238
UED	4794	0.2445	0.4190	0.004	2.7076
PTK	4794	12.897	1.8406	8.3087	17.8353
Road	4794	6.0009	0.5870	4.4886	7.1892
Economy	4794	16.2875	1.0383	13.9946	19.0017

displays descriptive statistics.

3.3. Construction of the Continuous Difference-in-Differences Model

Traditional difference-in-differences models ignore the diverse levels of influence brought about by different levels of HQE and only take into account the binary effects of HSR on cities along its lines. In order to examine the effect of HSR on the HQE of cities along its routes, this study uses the number of HSR lines as a continuous proxy variable for the HSR effect. It carries this out by using a continuous difference-in-differences model. Equation (1) represents the baseline DID regression model, as follows:

$${hqed}_{it}={\lambda }_0+{\lambda }_1{hsr}_{it}+X_{it}^{\prime }\gamma +f_i+{\delta }_t+{ϵ}_{it}$$

(1)

In the equation, ${hqed}_{it}$ reflects the degree of HQE in city i during period t; ${hsr}_{it}$ is a dummy variable representing the number of HSR lines in city i in year t, serving as the core explanatory variable to capture the rail effect. If ${\lambda }_1$ is significant, it indicates that HQE substantially influences the high-quality economic growth in cities along its routes. $X_{it}^{\prime }$ denotes the set of control variables; $f_i$ and ${\delta }_t$ represent individual and time effects, respectively; and ${ϵ}_{it}$ represents the random error term. This study employs city–time clustering to adjust the standard errors of the estimated coefficients, addressing potential heteroskedasticity and spatial autocorrelation issues that may affect the parameter estimates.

3.4. Spatial Evolution Characteristics of High-Quality Economic Development

To perform a thorough examination of the spatial evolution features of urban HQE, this research visualizes the regional distribution of HQE between 2005 and 2021 using ArcGIS 10.8 software, utilizing the natural breaks classification method to categorize development levels, thereby facilitating year-specific analyses. As illustrated in Figure 1, from 2005 to 2021, China’s HQE exhibited an evolutionary trend characterized by “overall advancement and enhanced regional coordination”, with a progressively optimized spatial pattern. In 2005, the vast majority of regions across the country were in a low-level development stage, with HQE indices converging between 0.547 and 1.665. The geographic dispersion of HQE was highly uneven, with central and western regions predominantly occupying low-value zones. There were only a handful of developed cities, forming isolated high-value “islands”. By 2021, with the sustained implementation of national strategies promoting innovation-driven growth, green development, and integrated development, the HQE index experienced widespread improvement. Medium–high-value regions (above 3.088) increased significantly, indicating that most areas had transitioned from quantitative growth to qualitative growth. Simultaneously, the developmental gap between western and central regions narrowed, reflecting a shift from an “east—high, west—low” pattern to a “gradient progression and regional synergy” trend, indicating the start of a new stage of HQE.

From a spatial perspective, the HQE index in 2005 displayed a distinct “east—strong, west—weak, point-based distribution” pattern. High-value zones were primarily concentrated in a few first-tier cities and provincial capitals along the eastern coast, forming limited high-quality development “core nodes”, while vast central and western regions, along with some border areas, remained low-value zones, highlighting pronounced spatial disparities. By 2021, the geographical allocation of HQE underwent significant transformation, manifesting as “areal diffusion and multi-core-driven growth”. Regional disparities tended to diminish and spatial equilibrium strengthened, indicating that the spatial organizational structure of China’s HQE is evolving toward greater rationality, efficiency, and sustainability.

4. Results

4.1. Baseline Regression

In the baseline regression model, HSR is the primary explanatory factor, while HQE is the dependent variable. Model (1) presents regression outcomes without considering control variables, while Models (2) and (3) progressively incorporate control variables. The positive and statistically significant coefficients of HSR service intensity in Models (1) to (4) suggest that the development of HSR has a significantly positive effect on urban HQE. Detailed regression results are provided in

Table 4. Baseline regression.

	(1)	(2)	(3)	(4)
	HQE	HQE	HQE	HQE
HSR	0.0422 ***	0.0779 ***	0.0748 ***	0.0742 ***
	(0.0123)	(0.0135)	(0.0135)	(0.0135)
HCI		2.3578 **	2.1194 *	1.9819 *
		(1.0903)	(1.0931)	(1.0923)
UED		−0.4401 ***	−0.4678 ***	−0.4989 ***
		(0.0668)	(0.0672)	(0.0677)
PTK			−0.0742 ***	−0.0771 ***
			(0.0182)	(0.0182)
Road			0.0982	0.0713
			(0.1111)	(0.1112)
Economy				0.2005 ***
				(0.0556)
Constant	1.7866 ***	1.7748 ***	2.0476 ***	−0.8141
	(0.0358)	(0.0428)	(0.7011)	(1.0587)
id	YES	YES	YES	YES
Year	YES	YES	YES	YES
N	4794	4794	4794	4794
$R^2$	0.2582	0.2659	0.2688	0.2709

Standard errors in parentheses. * p < 0.1, ** p < 0.05, and *** p < 0.01.

.

4.2. Parallel Trend Test for High-Speed Rail

This research adopts the continuous difference-in-differences model with city–year double fixed effects. The premise of this paper’s application of the model is to comply with the parallel trend, i.e., before the policy occurs, where the trend of change is the same for both the realization group and the control group, and in this paper, if there are variations between the experimental and control groups before the HSR’s opening, then it shows that the variation in HQE is not attributed to the establishment of HSR. In this research, if the control group and the experimental group show inconsistent trends before the HSR’s opening, then it indicates that the change in HQE is not induced by the construction of HSR. The findings reveal no noticeable distinction between the treated and untreated groups before the commencement of HQE, while in the later stage of building HSR, as shown in Figure 2, the outcome is affirmative and notable, signifying compliance with the parallel trend assessment.

4.3. Robustness Checks

As shown in

Table 5. Robustness test results I.

	(1)	(2)	(3)	(4)	(5)
	HQE	HQE	HQE	HQE	GTFP
HSR	0.0733 ***	0.0667 ***	0.0670 ***	0.0664 ***	0.0118 ***
	(0.0155)	(0.0147)	(0.0131)	(0.0139)	(0.0017)
Control variables	YES	YES	YES	YES	YES
Constant	−0.6575	−0.0834	−1.9733 *	−0.1753	0.1399
	(1.0615)	(1.1662)	(1.0219)	(1.0902)	(0.1324)
id	YES	YES	YES	YES	YES
Year	YES	YES	YES	YES	YES
N	4794	4230	4726	4512	4794
$R^2$	0.2696	0.2718	0.2881	0.2756	0.2751

Standard errors in parentheses. * p < 0.1 and *** p < 0.01.

, (1) the regression results further validate the robustness of the baseline regression conclusions by demonstrating a considerably favorable influence on HQE when HSR opening is used as the independent variable. (2) After excluding the special years of the 2008 financial crisis and the 2020 COVID-19 pandemic, the regression results remain significant, indicating that the positive effect of HSR on HQE is not disrupted by extreme economic events. (3) The findings indicate that HSR continues to considerably support high-quality economic growth in ordinary prefecture-level cities, even when municipalities are removed from direct central government control, eliminating the additional influence of policy priority and resource allocation. (4) One period behind the primary explanatory variable, the findings remain robust, demonstrating that HSR opening not only has an immediate positive effect on HQE, but also exhibits a significant lagged effect, continuously driving HQE in the long term. (5) Using green total factor productivity as a substitute for the HQE index, the regression results remain significant.

When assessing the effect of HSR on HQE, regression results are presented to exclude potential policy interferences. As shown in

Table 6. Robustness test results II.

	(1)	(2)	(3)	(4)
	HQE	HQE	HQE	HQE
HSR	0.0746 ***	0.0744 ***	0.0709 ***	0.0736 ***
	(0.0135)	(0.0135)	(0.0135)	(0.0135)
Smart City Pilot	−0.0225
	(0.0352)
Low-Carbon City Policy		−0.0080
		(0.0352)
Broadband China Pilot Program			0.1112 ***
			(0.0365)
New Energy Demonstration City				0.0482
				(0.0420)
Control variables	YES	YES	YES	YES
Constant	−0.8472	−0.8188	−0.9774	−0.7406
	(1.0601)	(1.0591)	(1.0591)	(1.0606)
id	YES	YES	YES	YES
Year	YES	YES	YES	YES
N	4794	4794	4794	4794
$R^2$	0.2709	0.2709	0.2724	0.2711

Standard errors in parentheses. *** p < 0.01.

, Model (1) excludes the effect of the smart city pilot policy, and the empirical results demonstrate that the positive effect of HSR on HQE is still statistically significant. Model (2) excludes the low-carbon city policy, and the regression analysis continues to reveal a favorable effect of HSR. Model (3) excludes the Broadband China pilot policy, and the positive effect of HSR remains unchanged and significant. Model (4) excludes the new energy demonstration city policy, and the regression results continue to demonstrate that HSR significantly facilitates HQE.

To confirm the credibility of the research results, this research addresses the potential influences of terrain fluctuation and resident population size on HSR construction by introducing interaction terms between terrain fluctuation and time, as well as resident population and time, as control variables to mitigate sample selection bias. As shown in

Table 7. Robustness test results III.

Description	Alleviate Sample Selection Bias			Instrumental Variable Method
Model	(1)	(2)	(3)	(4)	(5)
Variables	HQE	HQE	HQE	HQE	HQE
HSR	0.065 *** (0.016)	0.036 ** (0.017)	0.040 ** (0.017)		0.0787 *** (0.0199)
IV core explanatory variable lag 1				0.8432 *** (0.009)
Constant term	−10.002 *** (3.246)	−21.282 *** (5.220)	−32.281 *** (6.190)	-	-
Terrain × Time	YES	NO	YES	NO	NO
Population × Time	NO	YES	YES	NO	NO
Control variables	YES	YES	YES	YES	YES
id	YES	YES	YES	YES	YES
Year	YES	YES	YES	YES	YES
F-test	-	-	-	4132.87	-
R-square	0.271	0.272	0.274	-	0.275
N	4794	4794	4794	4512	4512

Standard errors in parentheses. ** p < 0.05 and *** p < 0.01.

, Column (1) presents regression results with the terrain fluctuation–time interaction term fixed, showing that HSR continues to significantly promote HQE. Column (2) reports regression results with the resident population–time interaction term fixed, indicating that the favorable influence of HSR is not significantly affected by population dynamics. Column (3) simultaneously fixes the interaction terms of terrain fluctuation time and resident population time, demonstrating that the effect of HSR on HQE remains robust. Columns (4) and (5) show instrumental variable regression results, with Column (4) confirming the validity of the instrumental variable and passing the weak instrument test and Column (5) showing that HSR construction exerts a positive impact on HQE at the 1% significance level, consistent with the results in Column (1), further validating the robustness of the findings.

4.4. Mechanism Analysis

Table 8 reports the results of the mediating effect test of HSR opening on the quality of economic growth, with the following model constructed:

$$M_{it}={\beta }_0+{\beta }_1{hsr}_{it}+{\beta }_2{control}_{it}+{\theta }_i+{\gamma }_t+{\epsilon }_{it}$$

(2)

Equation (2) in Model (3) bears resemblance to the specification of the benchmark regression model in Model (1). In this paper, five indicators are selected as the dependent variables, denoted as Mit.

Table 8. Mechanism analysis.

	(1)	(2)	(3)	(4)	(5)
	Asset Liquidity	Labor Mobility	Industrial Structure Upgrading	Industrial Chain Resilience	Green Innovation
HSR	0.3442 ***	0.4716 ***	0.0230 ***	0.0084 ***	0.0702 ***
	(0.0157)	(0.0671)	(0.0050)	(0.0006)	(0.0141)
Control variables	YES	YES	YES	YES	YES
Constant	−19.2800 ***	−0.9734	9.0130 ***	0.4505 ***	−7.1813 ***
	(1.2327)	(5.2587)	(0.3950)	(0.0489)	(1.1079)
id	YES	YES	YES	YES	YES
Year	YES	YES	YES	YES	YES
N	4794	4794	4794	4794	4794
r2	0.6273	0.2018	0.5764	0.4242	0.7019

Standard errors in parentheses. *** p < 0.01.

Table 8. Mechanism analysis.

	(1)	(2)	(3)	(4)	(5)
	Asset Liquidity	Labor Mobility	Industrial Structure Upgrading	Industrial Chain Resilience	Green Innovation
HSR	0.3442 ***	0.4716 ***	0.0230 ***	0.0084 ***	0.0702 ***
	(0.0157)	(0.0671)	(0.0050)	(0.0006)	(0.0141)
Control variables	YES	YES	YES	YES	YES
Constant	−19.2800 ***	−0.9734	9.0130 ***	0.4505 ***	−7.1813 ***
	(1.2327)	(5.2587)	(0.3950)	(0.0489)	(1.1079)
id	YES	YES	YES	YES	YES
Year	YES	YES	YES	YES	YES
N	4794	4794	4794	4794	4794
r2	0.6273	0.2018	0.5764	0.4242	0.7019

Standard errors in parentheses. *** p < 0.01.

Column (1) reveals that the regression coefficient for HQE is 0.3442, which passes the 1% significance level test, indicating that HSR significantly facilitates labor mobility and enhances the quality of economic development by optimizing human capital allocation. Column (2) demonstrates that the regression coefficient for HQE is 0.4716, passing the 1% significance level test, suggesting that HSR promotes capital mobility, providing financial support for economic growth. Column (3) shows that the regression coefficient for HQE is 0.0230, which also passes the 1% significance level test, indicating that HSR advances industrial structuring and optimizes the economic structure. Column (4) indicates that the regression coefficient for HQE is 0.0084, passing the 1% significance level test, suggesting that HSR strengthens the resilience of industrial chains, thereby ensuring economic stability and competitiveness. Column (5) reveals that the regression coefficient for HQE is 0.0702, passing the 1% significance level test, demonstrating that HSR fosters green innovation, driving green transformation and sustainable development.

4.5. Heterogeneity Analysis

This study uncovers notable distinctions in the promotional impact of HSR on the HQE of the eastern, central, western, and northeastern regions. As shown in

Table 9. Heterogeneity regression results I.

	(1)	(2)	(3)	(4)
	Eastern Region	Central Region	Western Region	Northeastern Region
HSR	0.0560 **	0.0107	0.1053 ***	0.0628
	(0.0243)	(0.0261)	(0.0257)	(0.0505)
	(2.5680)	(2.5571)	(1.7817)	(4.0934)
id	YES	YES	YES	YES
Year	YES	YES	YES	YES
N	1445	1343	1428	578
$R^2$	0.2191	0.3572	0.3341	0.2245

Standard errors in parentheses. ** p < 0.05 and *** p < 0.01.

, Column (1) shows that the impact of HSR on the HQE of the eastern region is positively significant at the 5% level, potentially benefiting from its solid economic foundation and optimized industrial structure. Column (3) demonstrates that the effect of HSR on the western region is positively significant at the 1% level, likely due to its underdeveloped infrastructure, as HSR has markedly improved transportation conditions. Columns (2) and (4) reveal that while the impacts of HSR on the central and northeastern regions are positive, they are not statistically significant, potentially related to their economic foundations, structures used in industry, and the extent of HSR network coverage, which have yet to completely realize the economic benefits of HSR. Thus, there exists significant diversity in the effect of HSR on the HQE of various regions.

There are two categories for the sample cities, as follows: hub cities and non-hub cities. According to

Table 10. Heterogeneity regression results II.

	(1)	(2)	(3)	(4)
	Transportation Hub City	Non-Transportation Hub City	Central City	Peripheral City
HSR	0.0420	0.0679 ***	0.0412	0.0962 ***
	(0.0576)	(0.0141)	(0.0446)	(0.0145)
Constant	4.9098	−2.1140 *	−1.0299	−1.5988
	(4.4179)	(1.1102)	(3.8566)	(1.0852)
id	YES	YES	YES	YES
Year	YES	YES	YES	YES
N	323	4471	612	4182
$R^2$	0.3129	0.2872	0.1878	0.3192

Standard errors in parentheses. * p < 0.1 and *** p < 0.01.

, Model (1) shows that the effect of HQE on the HQE of hub cities is positive but not statistically significant, potentially due to their already advanced infrastructure and economic development levels, resulting in smaller implications of HSR on the margins. Model (2) reveals that HSR exerts a significantly positive influence on the HQE of non-hub cities at the 1% significance level, likely because their infrastructure is relatively weaker, and HSR notably enhances transportation convenience and resource mobility. When cities are categorized into central cities and peripheral cities, Model (3) demonstrates that HSR positively impacts the HQE of peripheral cities, with statistical significance at the 1% level, potentially because HSR improves transportation accessibility, facilitating factor mobility and industrial agglomeration. Model (4) shows that HSR has a positive but insignificant effect on central cities, possibly because they possess strong economic foundations and resource agglomeration capabilities, with HSR primarily manifesting as a resource agglomeration factor rather than a new growth driver.

The samples are classified into resource-based and non-resource-based cities, with resource-based cities further divided into growing, mature, declining, and regenerative types. The results in

Table 11. Heterogeneity regression results III.

	Resource-Based City
	(1)	(2)	(3)	(4)	(5)	(6)
	Resource-Based City	Non-Resource-Based City	Growing	Mature	Declining	Rejuvenating
HSR	0.0659 ***	0.0770 ***	0.1875 ***	0.0744 **	−0.1341 ***	0.0433
	(0.0254)	(0.0168)	(0.0659)	(0.0368)	(0.0259)	(0.1322)
Control variables	YES	YES	YES	YES	YES	YES
Constant	2.7866 *	−2.7713 *	6.4608	2.4413	1.4723	25.9972 **
	(1.4884)	(1.4809)	(8.6322)	(2.0037)	(2.9810)	(11.0512)
Urban FE	YES	YES	YES	YES	YES	YES
Year FE	YES	YES	YES	YES	YES	YES
N	1904	2890	238	1020	391	255
$R^2$	0.3388	0.2501	0.6914	0.3427	0.6350	0.2916

Standard errors in parentheses. * p < 0.1, ** p < 0.05, and *** p < 0.01.

indicate that Models (1) and (2) show that the impact of HSR on the HQE of both resource-based and non-resource-based cities is significantly positive at the 1% level. Model (3) reveals that the influence of HSR on growing cities is significantly positive at the 1% level, potentially due to its acceleration of resource exploitation, industrial optimization, and market expansion. Model (4) demonstrates that the effect of HSR on mature cities is significantly positive at the 5% level. Model (5) shows that the impact of HSR on declining cities is significantly negative at the 1% level, likely because it has fails to effectively address resource depletion and industrial transformation issues. Model (6) indicates that the effect of HSR on regenerative cities is positive but not significant, possibly because they are in the early stages of transformation, and the positive effects of HSR have not yet fully materialized.

4.6. Analysis of Spatial Spillover Effects

Inter-city interactions indicate that the HQE of one city may be influenced by other cities, suggesting the presence of spatial correlation. Therefore, this study constructs a spatial difference-in-differences model to further examine the spatial impact effects of HSR on HQE.

4.7. Spatial Correlation Test

If the global Moran’s I index is greater than zero and the p-value is statistically significant, this indicates the presence of positive spatial effects. As shown in

Table 12. The global Moran’s I index results for HQE.

Economic Geography Weight Matrix
Year	Moran’ I	p-Value	Year	Moran’ I	p-Value
2005	0.2966	0.0000	2013	0.3147	0.0000
2006	0.3105	0.0000	2014	0.3035	0.0000
2007	0.3391	0.0000	2015	0.3004	0.0000
2008	0.3420	0.0000	2016	0.2869	0.0000
2009	0.2907	0.0000	2017	0.2931	0.0000
2010	0.3419	0.0000	2018	0.2554	0.0000
2011	0.3292	0.0000	2019	0.2324	0.0000
2012	0.3241	0.0000	2020	0.3021	0.0000
			2021	0.3305	0.0000

, under the economic geography weight matrix, spatial correlation is significant, which suggests that the use of a spatial econometric model is appropriate.

In the localized Moran scatter plot, the first quadrant represents regions with high levels of HQE surrounded by similar regions, while the third quadrant denotes low-level regions surrounded by analogous areas, both indicating positive spatial autocorrelation. The second quadrant illustrates low-level regions encircled by high-level regions, and the fourth quadrant depicts high-level regions surrounded by low-level regions, both reflecting negative spatial autocorrelation. The results demonstrate a significant spatial dependence between HSR and HQE, with high-level regions adjacent to other high-level regions and low-level regions neighboring other low-level regions. As shown in Figure 3 and Figure 4, the fitted regression line suggests a positive relationship.

This research uses a spatiotemporal fixed-effects Durbin model to examine the spatial correlation between HSR and HQE. By utilizing partial differentiation, the regression results are decomposed into direct, indirect, and total effects, as presented in the table. As shown in

Table 13. Estimation results of the spatial Durbin difference-in-differences model effects.

Variable	Main	Direct	Indirect	Total
HSR	0.0746 ***	0.0767 ***	0.0944 **	0.1711 ***
	(0.0131)	(0.0135)	(0.0444)	(0.0475)
rho	0.1373 ***
	(0.0271)
Control variables	YES	YES	YES	YES
Urban FE	YES	YES	YES	YES
Year FE	YES	YES	YES	YES
N	4794	4794	4794	4794

Standard errors in parentheses. ** p < 0.05 and *** p < 0.01.

, the direct effect of the economic–geographical weight matrix is positive and statistically significant at the 1% level, in accordance with the baseline regression results, indicating that the introduction of HSR fosters HQE in local areas. Furthermore, the indirect effect is also highly significant and positive at the 1% level, suggesting that HQE exerts a substantial beneficial influence on the high-quality economic growth of neighboring cities.

4.8. Spatial Distance Decay Effect

Having established the significant spatial spillover effects of HSR on urban HQE, this study employs a spatial Durbin difference-in-differences model with a geographically weighted matrix to further analyze the spatial decay boundary of HSR’s impact on HQE. The spatial spillover coefficients of HSR service intensity are examined by setting varying distance thresholds. If the geographical distance, dij, between city i and city j exceeds the specified threshold, the corresponding spatial weight matrix element is assigned a value of 1/ij d; otherwise, it is set to 0. The following is the precise calculation method:

$$f\left(x\right)=\left\{\begin{array}{c}1/d_{ij}, \mathrm{When} d_{ij} \mathrm{is} \mathrm{outside} \mathrm{the} \mathrm{distance} \mathrm{threshold} \\ 0 , \mathrm{When} d_{ij} \mathrm{is} \mathrm{inside} \mathrm{the} \mathrm{distance} \mathrm{threshold}\end{array}\right.$$

(3)

First, the initial distance threshold for $1/d_{ij}$ is set at the minimum inter-city distance of 100 km, incrementing by 100 km, with the step size reduced near the peak. Second, regression analysis is conducted using the spatial Durbin difference-in-differences model.

Figure 5 illustrates the spillover effect coefficients and their 95% confidence intervals across varying geographical distances. As shown in Figure 5, the spatial spillover coefficients exhibit a “U-shaped” distribution. In particular, there are four distinct intervals in the connection between the geographical distance threshold and the HSR spillover impact coefficient, as follows: (1) HQE in nearby regions favorably helps local high-quality economic growth when the geographical distance threshold falls between 100 and 200 km, demonstrating a positive spatial spillover effect; (2) when the threshold lies between 200 and 300 km, the spillover effect is not statistically significant; (3) when the threshold spans from 300 to 740 km, the development of HSR in surrounding areas exerts a restraining effect on local high-quality economic growth; and (4) when the threshold ranges from 740 to 1000 km, the development of HSR in surrounding areas again promotes local high-quality economic growth. This phenomenon may arise because, within this distance range, HSR begins to demonstrate its advantages as a long-distance rapid transportation mode.

5. Discussion

Our empirical results demonstrate that HSR significantly promotes HQE in China, with mechanisms including enhanced labor and capital mobility, industrial structure upgrading, and green innovation. However, the heterogeneity in effects across regions and city types suggests that the benefits of HSR are not uniformly distributed. For instance, the lack of a significant impact in central and northeastern regions may reflect infrastructural or policy gaps that need addressing. Additionally, the spatial spillover effects, while positive, exhibit a decay boundary of 1000 km, indicating that the economic benefits of HSR are geographically bounded. These findings underscore the importance of tailored regional policies to maximize HSR’s potential and ensure equitable development. Future research could explore the long-term dynamics of HSR’s impact and the role of complementary policies in amplifying its benefits.

6. Conclusions

This research, relying on city-level panel data from 2005 to 2021, employs a continuous difference-in-differences approach to empirically analyze the impact of HSR on urban HQE, further analyzing its mechanisms and spatial spillover effects. The findings reveal the following: First, HQE significantly enhances high-quality economic growth. Second, HSR enhances urban HQE by facilitating labor mobility, capital flow, industrial structure upgrading, supply chain resilience, and green innovation. Third, HQE generates favorable spatial spillover effects on the high-quality economic growth of neighboring cities, with a distinct spatial decay boundary of 1000 km. Fourth, the effect of HSR on urban energy efficiency exhibits significant heterogeneity across regions, city tiers, hub statuses, and resource endowments. Specifically, HSR positively and significantly influences HQE in eastern regions at the 5% level and in western regions at the 1% level, while its impact on central and northeastern regions, though positive, is not statistically significant. HQE positively affects hub cities, but not significantly, while its impact on non-hub cities and peripheral cities exhibits a significantly positive impact at the 1% level. Its influence on core cities is positive but not significant. In addition, this study reveals a remarkable spatial evolution of China’s HQE from 2005 to 2021, transitioning from an uneven “point-based distribution” to a more coordinated “multi-core driven” pattern.

In light of these findings, this study puts forward the following policy implications: First, given the significant role of HSR in driving HQE, the optimization of HSR network layouts should be prioritized, with accelerated construction in central, western, and remote regions to enhance regional accessibility and foster coordinated economic development. Second, as HSR boosts high-quality urban economic development by facilitating labor and capital mobility, advancing industrial upgrading, strengthening supply chain resilience, and stimulating green innovation, local governments should strengthen industrial park development along HSR routes, attract high-end manufacturing and modern service industries, and encourage green technological innovation to drive economic green transformation. Third, while HSR exhibits positive spatial spillover effects on neighboring cities within a 1000 km boundary, regional economic collaboration along HSR corridors should be strengthened to establish a coordinated industrial division system, enabling resource sharing and mutual benefits. Fourth, given the heterogeneous impacts of HSR across regions and city types, with more pronounced effects in eastern and western regions and greater benefits for non-hub and peripheral cities, policies should be more targeted. For instance, HSR resource allocation in eastern and western areas can be optimized to amplify economic benefits while exploring distinctive development pathways for HSR economies in central and northeastern regions based on regional strengths.