Economic Decentralization and High-Quality Urban Development: Perspective from Local Effect and Spatial Spillover in 276 Prefecture-Level Cities in China

Abstract

The advancement of high-quality urban development is of paramount importance for the enhancement of sustainable development and competitiveness at the city level. The economic decentralization system represents a pivotal institutional driving force in this regard. This paper examines the impact of decentralization on the high-quality development of Chinese cities. It constructs the high-quality urban development index ($HUDI$) through the entropy weight method and analyzes the mechanisms and spatial correlations between fiscal and financial decentralization on the high-quality development of cities through the establishment of panel and spatial regression models. The findings indicate that fiscal and financial decentralization exert a positive influence on urban high-quality development. However, the two forms of decentralization do not exhibit synergies but rather exert an inhibitory effect on one another. Furthermore, decentralization has a considerable positive spatial spillover effect on urban high-quality development. Heterogeneity analyses demonstrate that the impact of the economic decentralization system varies across different regions, reform periods, and cities with varying administrative levels. The robustness test of this paper provides further evidence of the reliability of the research findings. This paper offers theoretical support and policy recommendations for optimizing economic decentralization systems and promoting high-quality urban development.

1. Introduction

Since her reform and opening-up, China’s economy has undergone a significant acceleration in both speed and quantity of growth. Despite the significant challenges posed by the global economic downturn and the emergence of COVID-19, China’s economy demonstrated remarkable resilience in 2020, with its gross domestic product (GDP) surpassing the 100 trillion CNY, marking a distinctive chapter in the country’s economic growth history. Nevertheless, a number of issues have emerged alongside this impressive growth. These include income inequality [1], rising environmental pollution [2], an imbalanced industrial structure [3], and a reliance on resource-based industries [4]. In light of these developments, it is imperative that China undergo a transformation in its economic development model, with a view to promoting economic transformation and upgrading. This will enable the country to achieve its long-term goal of sustainable development. In 2017, China’s 19th National Congress explicitly declared that the country has entered a new stage of high-quality development [5]. This new stage is characterized by a shift in the driving forces of development, with innovation becoming the primary driver, coherence becoming an endogenous feature, greening becoming a universal form, openness becoming a necessary path, and sharing becoming the fundamental purpose. The “innovation, coherence, greening, openness, and sharing” constitute China’s new development philosophy [6]. For cities to achieve this ambitious blueprint for high-quality development, it is clear that they must prioritize several key factors. Firstly, cities must foster innovation to drive industrial upgrading and technological progress. Secondly, cities must promote the efficient allocation of resources between urban and rural areas, ensuring balanced development across different regions. Thirdly, cities must prioritize the construction of green and low-carbon urban development models. Fourthly, cities must deepen their openness to the outside world and engage in more robust forms of cooperation. Finally, cities must ensure that the benefits of urban development are more widely and more equitably shared with all citizens. What are the principal factors that will determine the achievement of the aforementioned objectives?

In examining the various factors that contribute to the high-quality advancement of regions, scholars have conducted comprehensive research from a multitude of perspectives. These include the role of entrepreneurship [7], the influence of urban land and population size [8], the impact of the digital economy [9], the implications of environmental regulation [10], and the significance of industrial integration [11]. It is, however, important to note that the majority of these studies have not fully considered the far-reaching implications of China’s distinctive economic institutional framework, particularly the pivotal role played by the economic decentralization system within it. In light of China’s current period of economic transformation, it is evident that failing to acknowledge the distinctive institutional framework embedded within the economic system will inevitably impede the realization of institutional dividends and the optimization of policy effectiveness. This, in turn, may impede the successful attainment of the objective of high-quality development. It is, therefore, pertinent to inquire whether economic decentralization can be conceptualized as a system that enables high-quality urban development.

The fundamental elements of China’s economic decentralization system are constituted by two key pillars: fiscal decentralization and financial decentralization. Fiscal decentralization affords local governments autonomy in the allocation of fiscal resources, enabling them to make decisions with greater flexibility in accordance with local needs and development strategies. Financial decentralization, furthermore, reinforces local governments’ control over financial resources, providing robust financial support for local economic development. The implementation of this economic decentralization system has significantly expanded the discretionary scope of local governments in economic activities, becoming a pivotal driving force behind the diversified and differentiated growth of the Chinese economy. Nevertheless, there is no consensus among scholars regarding the impact of economic decentralization, particularly fiscal decentralization, on the quality of urban development.

On the one hand, it has been argued that fiscal decentralization may have a detrimental impact on the governance effectiveness and the quality of development pursued by city governments. For instance, Shleifer and Vishny [12] indicate that within a multilevel government structure, each level of government is inclined to maximize the extraction of resources from citizens, which may have a detrimental impact on governance effectiveness. Additionally, studies by Gerring and Thacker [13] and Fan et al. [14] demonstrate that fiscal decentralization may result in the decentralization of decision-making authority at multiple levels, leading to ambiguity in the attribution of responsibilities and increasing the complexity of governance. Furthermore, Shon and Cho [15], based on an analysis of data from 50 states in the U.S., found that fiscal decentralization is associated with an increase in corruption. Additionally, Lin and Zhou [16] proposed that fiscal decentralization may induce incentive distortions, which are not conducive to the effective governance of public environments.

Nevertheless, another school of thought is optimistic, arguing that economic decentralization, by granting local governments economic autonomy and decision-making power, has significantly stimulated the endogenous dynamics and innovative vitality of local economic development. Kyriacou and Roca-Sagalés [17] argued that fiscal decentralization actually enhances the overall quality of government governance. Adam et al. [18] proposed that this enhancement effect may exhibit an inverted U-shaped pattern. The promotion is significant at a certain stage, but it may weaken beyond a certain threshold. Cheng et al. [19] demonstrated that fiscal decentralization combined with technological innovations effectively improves the quality of the environment. Lu et al. [20] also support the positive role of fiscal decentralization in enhancing resource efficiency in OECD countries.

In light of these considerations, it is imperative to inquire as to the genuine impact of fiscal decentralization on the advancement of superior urban development. Concurrently, what is the function of financial decentralization in this process? What is the nature of the interaction mechanism between different decentralization systems? This paper selects China as the subject of investigation to analyze how fiscal and financial decentralization contribute to high-quality urban development in cities. Three key factors support this choice. First, China has a clearly articulated concept of high-quality development, especially the five principles of “Innovation, Coherence, Greening, Openness, and Sharing” outlined in the 19th National Congress Report. These principles provide a comprehensive framework for evaluating high-quality development, enabling a systematic and practical approach to studying this developmental path. Second, China’s fiscal decentralization system has distinctive characteristics. Since the implementation of the tax-sharing system in 1994, governmental fiscal authorities and responsibilities have been more clearly defined, resulting in a unique decentralization model [21]. This system not only grants local governments greater financial autonomy to stimulate local economic development but also allows for quantifiable measures of decentralization, facilitating meaningful comparisons and analyses across cities. Finally, as a rapidly growing emerging economy, China’s experience in pursuing high-quality development and adjusting decentralization policies offers valuable insights. This makes China a compelling case for academic study and a critical reference for policy formulation in other countries.

Using panel data from 276 prefecture-level cities in China (2006–2022), we construct a city quality development index through the entropy weight method to comprehensively measure urban development quality. We then integrate fiscal and financial decentralization, as well as their synergies, into a unified framework to assess the multidimensional effects of economic decentralization on high-quality urban development. Additionally, a spatial Durbin model (SDM) is constructed to explore spatial effects among cities under the decentralization system. The study finds that both fiscal and financial decentralization significantly boost high-quality urban development, particularly in promoting innovation, openness, and sharing. However, fiscal decentralization contributes little to green development, while financial decentralization hinders green progress and shows weak performance in coherence. Furthermore, synergy between the two types of decentralization is limited, only becoming noticeable when the impact of either diminishes. Spatially, decentralization generates positive spillover effects, where neighboring cities benefit from local practices, though negative synergistic effects persist within individual cities. The heterogeneity analysis indicates that, geographically, eastern cities gain the most from decentralization, followed by central and western regions. Recent reforms have strengthened local and spatial spillover effects. Additionally, provincial capitals benefit more from financial decentralization, while non-provincial capitals gain more from fiscal decentralization. Ultimately, even after reconstructing the development and decentralization indicators and adjusting the spatial weighting matrix, the findings remain robust.

This paper makes significant contributions to the field at both theoretical and practical levels. From a theoretical perspective, it offers a novel analysis of the complex relationship between China’s economic decentralization system and high-quality urban development, situating these dynamics within economic institutional frameworks. This approach addresses existing research gaps and provides a robust empirical foundation for academia. First, this study advances the level of analysis from the provincial to the city level, enhancing the precision of insights into the effects of economic decentralization on urban development. Second, it goes beyond examining fiscal decentralization alone by incorporating financial decentralization, making it the first study to systematically analyze the synergistic mechanisms between fiscal and financial decentralization in promoting high-quality development. Additionally, using the spatial Durbin model, this paper investigates the spatial spillover effects of economic decentralization policies, shedding light on their interactive effects among neighboring cities and offering fresh insights into regional coordinated development.

From a practical perspective, this paper provides essential guidance for policymakers. By clarifying the various pathways through which economic decentralization drives cities toward high-quality development, it offers actionable policy recommendations. Specifically, the paper proposes that decentralization policies should be adjusted according to the level of regional development and industrial characteristics in order to optimize resource allocation and stimulate local innovation. Concurrently, by reinforcing regional collaboration, the government can establish a constructive interplay between fiscal and financial decentralization, thereby facilitating inter-regional resource allocation and leveraging complementary advantages, and advancing the sustainable growth of cities. In summary, the findings of this study not only expand the research scope on the nexus between economic decentralization and urban development but also offer robust policy guidance for establishing a more prosperous and sustainable urban development paradigm.

The rest of the paper is structured as follows: part two analyzes China’s economic decentralization system, starting with a review of existing research, followed by the construction of a theoretical framework and testable hypotheses. Part three introduces the high-quality urban development index ($HUDI$) and develops both a panel regression model and a spatial econometric model based on the hypotheses. It also covers sample selection and key variable definitions. Part four examines the spatio-temporal characteristics of the $HUDI$ and decentralization indicators, followed by a discussion of the econometric model results. Part five conducts a heterogeneity analysis to reveal regional, temporal, and city-type differences in decentralization effects and performs robustness tests. Finally, part six provides targeted policy recommendations and suggests directions for future research.

2. Research Background, Theoretical Analysis, and Hypothesis Construction

2.1. Research Background

2.1.1. Background of China’s Economic Decentralization System

In the context of China’s economic expansion, the phenomenon of economic decentralization has emerged as a notable phenomenon. In this process, economic decentralization is closely intertwined with the performance appraisal mechanism of officials, which together constitute the dual incentives that drive local governments to actively engage in economic construction. The Chinese government has adopted an innovative approach to viewing the vast economic system as a comprehensive political ecosystem. Decentralization is not only a key incentive for maintaining the system’s efficient functioning, but it also provides local governments with the opportunity to receive direct positive feedback by promoting local economic growth [22]. China’s economic decentralization can be broadly classified into two categories: fiscal decentralization and financial decentralization. Fiscal decentralization refers to an institutional arrangement whereby the central government delegates certain taxing powers and spending responsibilities to local governments, thereby enabling them to independently determine the size and structure of budgetary expenditures. Following the implementation of the tax-sharing reform in China during the 1990s, the system of fiscal decentralization was formalized, and the institutional arrangements became increasingly clear [23]. The primary objective of the Chinese government in implementing this system is to enhance fiscal efficiency. The decentralization system confers greater autonomy upon local governments with respect to fiscal revenue and expenditure. This effectively encourages local governments to pursue revenue-generating activities while facilitating the optimal allocation and utilization of fiscal resources. Secondly, local governments are able to formulate policies that align with their specific circumstances, thereby better addressing the needs of their local communities [24]. Ultimately, the fiscal decentralization system engenders a natural competition for resources among local governments, thereby motivating them to pursue institutional innovation and policy adjustment. Simultaneously, governments may also pursue collaboration and facilitate inter-regional cooperation and coordinated development by leveraging local advantages. This has undoubtedly become a significant driving force for high-quality urban development.

An additional economic decentralization system that is complementary to fiscal decentralization is financial decentralization. In China, the core of financial decentralization is reflected in two dimensions: the division of power between the central and local governments, and the reshaping of the boundaries between the government and market. In a market-oriented economic system, the vitality of the local economy is often contingent upon the competitive pursuit of financial resources. Consequently, local governments endeavor to obtain greater financial authority from the central government, thereby establishing a foundation for the establishment and control of local financial institutions and financial markets [25]. This process directly contributes to the financial decentralization pattern between the central government and local governments. Concurrently, the operational strategies of local state-owned enterprises serve as an extension of the local government’s economic policy, reflecting a delicate balance between the local government’s control and efficiency. An excessive degree of centralization may impede the innovative dynamism of enterprises and the regulatory capacity of the market, thereby constraining economic growth. Conversely, an excessive degree of decentralization may give rise to irrational prosperity in the financial market, which, in turn, may lead to the accumulation of systemic risks. These two forms of financial decentralization function in conjunction within the financial system, with the objective of facilitating the optimal allocation of resources, enhancing the efficiency of financial services, and preventing and resolving financial risks. Nevertheless, as the essence of fiscal decentralization hinges on the delineation of the fiscal resource distribution between the central and local governments, this paper focuses on defining financial decentralization with regard to the financial resource distribution between the central and local governments.

2.1.2. Literature Review

One of the initial key topics to be addressed in this paper is how to measure the level of high-quality urban development. In the field of urban development measurement, two primary approaches have emerged within academic discourse. One such approach is the utilization of the data envelopment analysis (DEA), which seeks to assess the efficiency of urban development. This method is founded upon the principles of linear programming and is primarily employed for the evaluation of the relative efficiency of decision-making units with multiple inputs and outputs. For example, Iribarren et al. [26] utilized the DEA methodology to assess socio-economic indicators in accordance with sustainability criteria. Ren et al. [27] employed the DEA method to assess the green total factor energy efficiency of 30 provinces in China. In addition, Neykov et al. [28] examined the efficiency of micro and small wood-processing enterprises in the European Union with the DEA model. However, this method places an emphasis on the transformational efficiency of inputs and outputs. However, in the context of urban high-quality development, the input factors are numerous and cannot be fully quantified. Furthermore, the specific connotation of high-quality development has been clearly defined in China’s 19th National Congress. Consequently, the objective value assignment (OVA) method is a more suitable choice for this paper. The OVA is founded upon the intrinsic attributes and interrelationships of the data, which are employed to determine the weight of each sub-indicator, including methods of entropy weight, principal component analysis (PCA), coefficient of variation (CV), and criteria importance through inter-criteria correlation (CRITIC).

To begin with, the entropy weight method employs information entropy, defined as the uncertainty of information, to calculate the weight of each indicator. These weights are then aggregated to derive a comprehensive indicator. Sahoo et al. [29] employed the entropy weight method to assess the overall water quality of the Brahmani River, resulting in the creation of a composite index based on six water quality indicators. Wang et al. [30] constructed an evaluation index system for the green development of the city by using the entropy weight method from five aspects such as improvement of human habitat as well as pollutant management and utilization. Chen et al. [31] used the entropy weight method to measure and rank the ecological level of cities in China’s Yellow River Basin. Secondly, PCA represents another OVA method that is widely utilized. The fundamental concept of PCA is the mapping of high-dimensional data to a lower-dimensional indicator through the application of dimensionality reduction techniques, with the objective of retaining the essential characteristics of the data as much as possible. For instance, Costa et al. [32] employed PCA to assess the economic and financial performance of Italian social cooperatives, with a particular focus on the efficiency and profitability of these organizations. Hu and Xu [33] employed remote sensing data of Fuzhou, China, to develop an ecological quality evaluation index for the city through the PCA technique. Zhang et al. [34] employed PCA to integrate 16 sub-indicators of economic, social, and ecological development in order to assess the environmental carrying capacity of water resources in urban areas within the Xiangjiang River Basin in China. Thirdly, the CV method is employed to ascertain the coefficient of variation of each indicator, thereby determining its significance within the evaluation system. Long et al. [35] utilized the CV method to assess Beijing’s water environmental carrying capacity. Finally, the CRITIC method determines the weight of each indicator in a scientific and objective manner. This is performed by comprehensively considering the comparative strength of the evaluation indicators and the conflict between the indicators. Li et al. [36] used the CRITIC method to construct indicators to assess the age-friendliness of the smart city. As the quantification of the sub-indicators of urban high-quality development is also the objective of this paper, this paper employs the entropy weight with linear combination method to assess urban high-quality development. Additionally, this paper utilizes the TOPSIS, PCA, CV, and CRITIC methods to reconstruct the index, which is then subjected to a robustness test.

Although there is a paucity of research examining the role of an economic decentralization system in fostering high-quality urban development, numerous scholars have delved deeply into the specific impacts of decentralization on corporate innovation, regional coordination, resource sharing, and green development. These economic outcomes are particularly pertinent to high-quality urban development. For example, Busemeyer [37] identified a robust positive relationship between fiscal decentralization and the level of total expenditure on education. This finding suggests that fiscal decentralization facilitates the sharing of educational resources. Kyriacou et al. [38] demonstrated that fiscal decentralization effectively reduces regional development imbalances based on panel data from 23 OECD countries. Similarly, Canavire-Bacarreza et al. [39] found that an increase in either the fiscal expenditure or revenue share of local governments significantly contributes to economic output. In addition, Khan et al. [40] investigated the influence of fiscal decentralization on carbon dioxide emissions in a sample of seven OECD countries. Their findings indicated that a decentralized fiscal system is conducive to enhanced environmental quality. A case study comprising data from 18 countries for the period of 2011–2017 by Hung and Thanh [41] indicated that fiscal decentralization plays a significant role in the growth of national economies. Based on panel data of non-financial listed companies in China, Li and Qi [42] found that fiscal decentralization has a positive effect on promoting firms’ innovative activities. The study by Satrovic et al. [43] examined the relationship between fiscal decentralization and environmental degradation in nine EU member states over the period of 1995 to 2018. The findings indicate that decentralization contributes to environmental degradation and that EU member states must empower local governments to reduce pollution through innovative technologies in order to strengthen the sustainable development goals. The preceding research findings substantiate the assertion that fiscal decentralization exerts an influence on the specific dimensions of urban high-quality development, including innovation, coherence, greening, openness, and sharing. Nevertheless, there is a paucity of research examining the influence of financial decentralization, and even fewer studies have explored the potential synergies between these two forms of economic decentralization. This paper introduces a novel perspective on the role of financial decentralization in China’s economic decentralization. It also presents a new approach to understanding the relationship between decentralized systems and spatial spillover effects, thereby contributing to the existing literature on this topic.

2.2. Theoretical Analysis and Hypothesis Construction

2.2.1. The Economic Decentralization System and High-Quality Urban Development

The impact of fiscal decentralization on high-quality urban development may be observed in the following aspects. Firstly, local governments are able to enhance the level of social welfare of residents through the provision of public products and public services, which, in turn, promotes local high-quality development. Local governments can leverage their informational advantage to influence sectoral outputs through fiscal resources, tailor public products and services to local preferences, and enhance the efficiency of public product supply, thereby increasing societal output and promoting economic growth [22]. Secondly, local governments at all levels are driven by a robust incentive to stimulate economic growth, particularly in light of the considerable political pressure exerted by economic championships, promotional mechanisms for officials, and intergovernmental competition [44]. The advancement of the regional economy is conducive to the enhancement of economic indicators and the advancement of local officials. Fiscal decentralization permits local governments to leverage the fiscal resources at their disposal to proactively stimulate economic growth. Concurrently, the institutional framework of fiscal decentralization markedly encourages intergovernmental competition for economic advancement, encompassing both horizontal and vertical dimensions. This competitive environment has prompted local governments at all levels to implement active measures to stimulate economic growth in their respective jurisdictions [45]. Thirdly, the fiscal decentralization system allows for the coordination of economic development at all levels of government. On one side, from the perspective of enhancing the governance capacity and performance, local governments can utilize the substantial fiscal revenues afforded by the fiscal decentralization system to implement preferential policies, such as tax reductions and fee waivers, with the objective of attracting exemplary enterprises to relocate. This strategy guides the factors of production, including capital, labor, and technology, to congregate in the pivotal economic regions, influencing the shifts in the industrial structure and the economic structure and, thus, promoting the optimization and upgrading of the economic structure and its advancement towards a higher level of development. On the other side, the central government is also able to adjust and optimize the overall economic layout through transfer payment tools and fiscal policies under the fiscal decentralization system. This enables the redistribution of resources from a national perspective, allowing the local economy to leverage its advantages and thereby enhance the quality of its development. In conclusion, this paper proposes the following initial hypothesis:

Hypothesis 1 (${\mathit{H}}_{\mathbf{1}}$). 

Fiscal decentralization will enhance the high quality of urban development.

In addition, financial decentralization may facilitate the high-quality development of the economy through the following channels. Firstly, financial decentralization may facilitate the effective utilization of financial resources by local governments to supplement their existing financial resources. Financial decentralization will enable local governments to obtain the right to deploy financial resources. In situations where fiscal resources are constrained, local governments will utilize credit financing through market-oriented operations, thereby transforming financial resources into resources for economic development and promoting the high-quality development of their cities. Secondly, financial decentralization encourages local governments to engage actively in the development of the financial market. Local governments will engage in horizontal competition for limited financial resources, subsequently developing their own financial markets, expanding the scale of bank savings and credit, and improving the level and efficiency of investment [46]. This will undoubtedly facilitate high-quality economic development. Thirdly, the implementation of a decentralized financial system will facilitate the realization of the multiplier effect inherent to capital operations. In comparison to the direct utilization of financial resources, local governments are able to leverage the multiplier effect to enhance the scale and impact of financial resource deployment. Financial institutions facilitate the provision of financing to governments, enterprises, and individuals in a manner that addresses the discrepancy between the supply and demand of funds through market-based mechanisms. Ultimately, financial decentralization is characterized by a market-oriented approach. The market-oriented resource allocation method, which is a characteristic of the financial decentralization system, can regulate the problem of unbalanced distribution of financial resources. Furthermore, it can rationally allocate the reasonable and orderly flow of financial resources within the region, thereby promoting high-quality economic development. Accordingly, this paper puts forward the second hypothesis:

Hypothesis 2 (${\mathit{H}}_{\mathbf{2}}$). 

Financial decentralization will enhance the high quality of urban development.

2.2.2. Synergies Between Economic Decentralization Systems in Promoting High-Quality Urban Development

As two significant financing instruments for local government economic development, the actions of one decentralized system inevitably affect the other. Given that economic phenomena are not independently affected by either fiscal or financial policies, but rather by both, it is imperative to inquire whether the two decentralization systems are synergistic in their promotion of high-quality urban development. In other words, does the implementation of one decentralization system result in an increase or a decrease in the impact of the other? From a theoretical standpoint, the two decentralization systems may exhibit a complex relationship that is simultaneously “complementary” and “conflicting”. The potential for a synergistic relationship between the two is initially explored from the perspective of complementarity. For one thing, from the perspective of China’s current economic system and market regulations, while the delineation of rights and responsibilities between the central and local governments with respect to fiscal revenues is relatively clear, the allocation of financial expenditures at the local level is primarily determined by local governments. Local governments have strong incentives to overspend in order to promote economic growth and safeguard people’s livelihoods [47]. The financial resources owned by the government under the financial decentralization system can effectively alleviate the government’s fiscal constraints. For another, the advancement of the economy necessitates a greater allocation of financial resources, with the local government assuming a pivotal role as both a source and user of these funds. The local government can facilitate investment and borrowing through financial institutions through various channels, such as investment and financing platforms. However, these actions require a robust fiscal foundation, particularly in the context of a fiscal decentralization system. Therefore, fiscal and financial decentralization are mutually reinforcing, reflecting a complementary character in the process of promoting high-quality urban development.

Nevertheless, from the perspective of “conflict”, fiscal and financial forms of decentralization are frequently ambivalent. First, the financial decentralization system is not as clearly defined as the fiscal system. This lack of clarity frequently gives rise to unhealthy competition among local governments for limited financial resources, which, in turn, gives rise to local protectionism. Second, excessive financial decentralization compels local governments to intervene in the market through economic policies, which, in turn, results in the distorted allocation of market resources and an uneven distribution of financial resources between state-owned and private enterprises. This, in turn, affects market innovation [48]. This situation can result in a reduction in the efficiency of financial resource supply, thereby limiting the ability of citizens to benefit from the positive effects of rapid development. Finally, intense local government competition and promotion pressures have compelled officials to vie for credit resources through various avenues, thereby heightening the incentives for local governments to assume additional debt. The accumulation of significant local debt has the potential to impact not only the long-term fiscal sustainability, but also the stability of the financial system. When risks reach a certain threshold, they can give rise to regional or systemic financial risks, thereby undermining the positive effects of decentralization on high-quality urban development.

Consequently, the synergies of decentralization on high-quality urban development are evidenced by a conjunction of “complementarities” and “conflicts”. Should the complementarity exceed the conflict, the resulting synergy will be positive. Conversely, the synergistic effect is detrimental. Hence, this paper puts forth a third set of controlling hypotheses:

Hypothesis 3 (${\mathit{H}}_{\mathbf{3}}$). 

The decentralization system shows synergy in the process of enhancing the high-quality urban development.

Hypothesis 4 (${\mathit{H}}_{\mathbf{4}}$). 

The decentralization system does not show synergy in the process of enhancing the high-quality urban development.

2.2.3. Spatial Spillover Effects of Economic Decentralization on High-Quality Urban Development

According to the first law of geography, nearby entities are more closely related than distant ones [49]. Thus, when analyzing the impact of decentralization on urban development, it is crucial to account for spatial factors, as decentralization in one city can influence high-quality development in neighboring cities. This phenomenon can be explained by three factors. First, the mobility effect: fiscal and financial resources generated by decentralization are mobile, and higher decentralization in one city may cause these resources to spill over to nearby cities, boosting their development. Second, the demonstration effect: cities with greater fiscal and financial autonomy may innovate in resource use—collaborating with financial institutions, offering subsidies, or reducing taxes—which neighboring cities can adopt, creating positive spillover effects. Third, the competitive effect: cities with similar resources may compete for fiscal and financial decentralization, with highly decentralized cities potentially siphoning resources from surrounding areas, leading to negative spillovers. The combined influence of these effects determines the direction of spatial spillovers. Therefore, when studying decentralization’s impact on urban development, spatial correlation must be considered, leading to the fourth set of control hypotheses in this paper:

Hypothesis 5 (${\mathit{H}}_{\mathbf{5}}$). 

The economic decentralization system has a positive spatial spillover effect on high-quality urban development.

Hypothesis 6 (${\mathit{H}}_{\mathbf{6}}$). 

The economic decentralization system has a negative spatial spillover effect on high-quality urban development.

Figure 1 provides a visual representation of the theoretical analysis presented in this paper, elucidating the relationship between economic decentralization and local high-quality development.

3. Indicator Construction, Model Setting, and Data Selection

3.1. High-Quality Urban Development Index Construction

The adoption of green total factor productivity as a benchmark for evaluating the quality of urban development has gained significant recognition among academic researchers [50]. Nevertheless, the implementation of this method is contingent upon the accuracy of the modeling. The use of disparate statistical bases and measurement methods can readily result in the introduction of bias into the assessment results. In light of the aforementioned considerations, this paper proposes a comprehensive multidimensional high-quality urban development index ($HUDI$) system, based on the five core development concepts of innovation, coherence, greening, openness, and sharing. This paper uses the entropy weight with a linear combination method to calculate the comprehensive scores for the HUDI and each sub-dimension. This method objectively determines indicator weights based on data dispersion and allows for the division of the overall indicator into multiple sub-indicators, enabling detailed analysis of each dimension of high-quality urban development. In contrast, the PCA, CV, and CRITIC methods may result in information loss or computational complexity in dimensionality reduction and weight assignment, limiting the precision of sub-indicator analysis. Thus, the entropy weight with a linear combination method is well-suited for comprehensive and detailed multi-dimensional analysis in this study. The selection and construction of the $HUDI$ are presented in

Table 1. Construction of $HUDI$.

Tier 1 Indicator	Tier 2 Indicator	Calculation Method	Unit of Measure	Direction
$Innovation$ $U_1$	Investing in Innovation $X_1$	Government expenditure on science divided by expenditure on the general government budget	%	+
	Education Development $X_2$	Number of students enrolled in general higher education institutions	Person	+
	Patent Innovation $X_3$	Number of patents obtained	Number of Cases	+
$Coherence$ $U_2$	Balanced Urban and Rural Development $X_4$	Per capita disposable income of urban residents divided by per capita disposable income of rural residents	%	−
	Balanced Industrial Structure $X_5$	Value added of the tertiary sector divided by value added of the secondary sector	%	+
	Urbanization Level $X_6$	Non-agricultural population divided by household population	%	+
$Greening$ $U_3$	Sulfide Emission $X_7$	Industrial Sulfur Dioxide Emission	Ton	−
	Clean Living $X_8$	Harmless treatment rate of domestic rubbish	%	+
	Dust Emission $X_9$	Emission of industrial dust	Ton	−
$Openness$ $U_4$	Foreign Investment Openness $X_{10}$	Amount of foreign capital actually used in the year	Ten Thousand US Dollars	+
	Import and Export Openness $X_{11}$	Total import and export of goods	Ten Thousand CNY	+
	Foreign Enterprises Openness $X_{12}$	Number of foreign-invested enterprises	Number of Enterprises	+
$Sharing$ $U_5$	Education Resources Sharing $X_{13}$	Total number of books in public libraries	Thousands of Copies	+
	Medical Resources Sharing $X_{14}$	Number of beds in hospitals and health centers	Number of Beds	+
	Network Resource Sharing $X_{15}$	Number of Internet users	Number of Users	+

Source of data: China Statistical Yearbook, China Urban Statistical Yearbook, China Rural Statistical Yearbook, and Chinese Research Data Service Platform. Note: The use of a plus sign (+) in Direction indicates that a larger variable is preferable, whereas the use of a minus sign (−) indicates that a smaller variable is preferable.

.

This paper carefully selects representative indicators based on data availability to assess a city’s $HUDI$. For innovation vitality and capacity, key indicators include government investment in science and technology, education system development, and patent applications. For coherence, sub-indicators such as the urban–rural development gap, balanced industrial structure, and urbanization rate are chosen to reflect improvements in quality of life, industrial optimization, and urban–rural integration. Regarding greening, the focus is on clean production and green living, with sulfur dioxide and soot emissions, as well as harmless waste treatment, used to measure environmental performance. The openness dimension considers the city’s engagement with the global economy, using foreign investment, foreign-owned enterprises, and import–export activities as indicators. Lastly, the sharing dimension emphasizes social equity, evaluating access to education, healthcare, and network resources to measure a city’s efforts in promoting well-being for all citizens.

3.2. Model Setting

3.2.1. Baseline Regression Model

In order to test Hypotheses 1–4, this paper constructs a panel model as shown in Equation (1) as the baseline regression model.

$${HUDI}_{it}=\alpha +{\beta }_1{Dfis}_{it}+{\beta }_2{Dfin}_{it}+{\beta }_3{Cross}_{it}+{\beta }_3{X}_{it}+{\mu }_i+{\nu }_t+{\epsilon }_{it}$$

(1)

In Equation (1), ${Dfis}_{it}$ denotes the level of fiscal decentralization of city i in year t, while ${Dfin}_{it}$ denotes financial decentralization. The coefficients ${\beta }_1$ and ${\beta }_2$ denote the extent to which decentralization affects the $HUDI$ to test Hypotheses 1 and 2, respectively. ${Cross}_{it}$ denotes the cross-sectional term between fiscal and financial decentralization, and the coefficient ${\beta }_3$ is used to denote whether or not decentralization creates synergies in promoting the $HUDI$, i.e., discussing the scenario in Hypotheses 3 and 4. $X_{it}$ denotes the set of control variables. ${\mu }_i$ denotes city fixed effects, ${\nu }_t$ denotes year fixed effects, and ${\epsilon }_{it}$ denotes the random error term.

3.2.2. Spatial Econometric Model

In order to test Hypotheses 5 and 6, which states that economic decentralization has spatial spillovers, this paper sets up a spatial econometric model for estimation. As posited by LeSage and Fischer [51], the spatial correlation has its origins in different sources. If the source is the error term, this is referred to as the spatial error model (SEM). If the source of the spatial correlation is the dependent variable, the appropriate model is the spatial auto-regressive model (SAR). If the spatial correlation of the dependent and independent variables is all considered, this is the SDM model. This paper employs the SDM model (the SEM and SAR models were also constructed in this paper, and the model settings and regression results are shown in Appendix A), as illustrated in Equation (2). The rationale for selecting the SDM model is that it is capable of incorporating the spatial lag effects of both dependent and independent variables in a comprehensive manner. From the perspective of urban development, the quality of growth in different cities is influenced by a complex network of interdependencies. This underscores the importance of accounting for the spatial lag effect of the dependent variable, namely the HUDI, in the model. Furthermore, the implementation of economic decentralization policies exerts influence not only on local cities but also on neighboring cities through spatial spillover effects. Therefore, it is essential to consider the spatial lag effect of the independent variable, namely the economic decentralization system. The SDM model is capable of incorporating the spatial lag terms of the dependent and independent variables simultaneously, thereby capturing the spatial correlation between the dependent and independent variables and accurately reflecting the impacts of economic decentralization on the high-quality development of cities.

$$\begin{array}{cc}\hfill & {HUDI}_{it}=\alpha +{\beta }_1{Dfis}_{it}+{\beta }_2{Dfin}_{it}+{\beta }_3{Cross}_{it}+{\beta }_3{X}_{it}+\rho {{\sum }_{j}W}_{ij}{HUDI}_{it}\hfill \\ \hfill & +{\gamma }_1{{\sum }_{j}W}_{ij}{Dfis}_{it}+{\gamma }_2{{\sum }_{j}W}_{ij}{Dfin}_{it}+{\gamma }_3{{\sum }_{j}W}_{ij}{Cross}_{it}+{\mu }_i+{\nu }_t+{\epsilon }_{it}\hfill \end{array}$$

(2)

In Equation (2), $W_{ij}$ is the matrix that represents the spatial correlation between city i and j. In this paper, the types of spatial correlation are considered comprehensively, and three types of spatial weighting matrices are set, which are defined as geographic adjacency matrix $W_{0–1}$, distance matrix $W_d$, and economic matrix $W_e$ (please refer to Appendix B for detailed instructions on how to define the three matrices). The three spatial weighting matrices measure the spatial relevance in terms of city geographic proximity, distance, and economic intensity. Moreover, $\rho$ denotes the effect of the spatial lag term of the independent variable, $\gamma$ denotes the effect of the spatial lag term of the decentralized indicator and the cross term, and the other variables are set as in Equation (1).

3.3. Variable Selection

This paper employs a panel data set comprising 276 prefecture-level cities in China for the period between 2006 and 2022 (due to differences in administrative levels and the caliber of data statistics, the sample does not include the four municipalities directly under the central government of Beijing, Tianjin, Shanghai, and Chongqing, nor does it include the Hong Kong and Macao Special Administrative Regions and the municipalities in Taiwan Province. Furthermore, due to the absence of data, this study excludes the prefecture-level cities of Bijie, Changdu, Haidong, Linzhi, Nagchu, Shigatse, Shannan, Tongren, Chaohu, Laiwu, Sansha, Danzhou, Turpan, Hami, Wuzhong, Guyuan, Zhongwei, Urumqi, and Karamay). Other variables are defined as follows.

3.3.1. Fiscal Decentralization

This paper makes reference to the works of Liu et al. [52], Yu et al. [53], and other scholars in order to quantify the degree of municipal fiscal decentralization from the perspectives of decentralization of fiscal revenues ($DFR$) and decentralization of fiscal expenditures ($DFE$), respectively. The calculation of $DFR$ is based on the per capita fiscal revenues of the prefecture-level city, divided by the sum of per capita fiscal revenues of the municipal, provincial, and central governments. The calculation of $DFE$ entails dividing the per capita fiscal expenditure of prefecture-level cities by the sum of the per capita fiscal expenditures of municipal, provincial, and central governments. Similarly, Wang et al. [54] put forth a methodology wherein the per capita fiscal revenue is divided by the per capita fiscal expenditure, thereby indicating the degree of autonomy local governments possess in the utilization of fiscal resources. This paper employs the decentralized degree of fiscal freedom ($DFF$) as a measure of this autonomy. This paper presents only the empirical results for $DFR$. The robustness of the fiscal decentralization indicators is tested using $DFE$ and $DFF$.

3.3.2. Financial Decentralization

The prevailing methodology for measuring financial decentralization entails a comparison between the per capita loan balance of prefecture-level cities and the national per capita loan balance [55]. This approach is employed in this study to construct the decentralization of financial loan ($DFL$) indicator. However, considering that the large amount of deposits absorbed by financial institutions may also be converted into financial resources available to local governments, this paper also innovatively proposes the decentralization of financial deposit ($DFD$) by dividing the per capita deposit balance of prefectural cities by the national level, and this indicator is used as a robustness test of the $DFL$.

3.3.3. Control Variables and Data Sources

Considering that there are other variables that may affect the high-quality development of cities, this paper refers to the studies of scholars such as Bai et al. [56], as well as Peng et al. [57], and selects regional population ($Pop$), degree of financial development ($Fin$), investment in fixed assets ($Ast$), level of industrialization ($Ind$), level of human capital ($Hum$), transport infrastructure ($Tsp$), and market scale ($Mkt$) as control variables (please refer to Appendix C for details regarding the calculations of the control variables). The data sources for the independent variables are China Statistical Yearbook and China Urban Statistical Yearbook. The descriptive statistics of all variables are shown in

Table 2. Descriptive statistics for variables.

Variable	Observation	Mean	Std.Dev.	Min	Max
$HUDI$	4692	3.132	3.681	0.455	40.911
$Innovation$	4692	1.141	1.701	0.008	18.172
$Coherence$	4692	0.686	0.336	0.204	2.632
$Greening$	4692	0.145	0.028	0.024	0.474
$Openness$	4692	0.640	1.634	0.005	24.635
$Sharing$	4692	0.521	0.560	0.011	12.222
$DFR$	4692	0.343	0.163	0.018	0.870
$DFE$	4692	0.453	0.132	0.021	0.912
$DFF$	4692	0.551	0.271	0.023	8.390
$DFL$	4692	2.477	2.343	0.214	57.693
$DFD$	4692	2.443	1.842	0.236	18.993
$Pop$	4692	4.637	0.722	2.703	6.942
$Fin$	4692	6.182	4.618	0.652	102.282
$Ast$	4692	10.517	1.002	5.031	13.894
$Ind$	4692	0.468	0.121	0.086	0.888
$Hum$	4692	0.059	0.060	0	0.681
$Tsp$	4692	12.774	8.824	0.308	108.327
$Mkt$	4692	15.329	1.097	11.959	18.450

.

4. Indicator Analysis and Empirical Results

4.1. Analysis of $HUDI$ and Decentralization Indicators

4.1.1. Analysis of $HUDI$ and Sub-Indicators

In this paper, the entropy weight method is employed to assign weights to the indicators enumerated in Table 1. These weights are then utilized in a linear combination to derive the $HUDI$ and the five sub-indicators it encompasses. Figure 2 depicts the temporal trend of these indicators in terms of the national average between 2006 and 2022. As illustrated in Figure 2, the $HUDI$ demonstrates a persistent and consistent upward trajectory, substantiating the assertion that China’s cities are progressing along a trajectory of high-quality development and exhibiting a relatively stable growth pattern. A detailed examination of the sub-indicators reveals that the growth trend of the $Innovation$ indicator is the most pronounced, with its growth contributing the most to the $HUDI$. The rate of growth for $Sharing$ is 196.33% in the sample year. The $Openness$ indicator also demonstrates a notable increase, with its mean value exceeding twofold, from 0.55 in 2006 to 1.23 in 2022. The $Coherence$ and $Greening$ indicators, though smaller in magnitude than the others, also demonstrate growth rates of 48.35% and 43.70%, respectively. This reflects China’s emphasis on balance and sustainability while pursuing comprehensive development.

An analysis of the $HUDI$ reveals significant regional disparities in China. Figure 3 visually represents the average $HUDI$ levels across cities during the sampling period. Figure 3a highlights the total indicator average, showing a clear geographic pattern: higher $HUDI$ levels in the eastern region, a slight decline in the central region, and relatively low levels in the west. This reflects the uneven progress in high-quality development across China’s regions. Figure 3b–f further illustrate this disparity through color shading, with darker shades representing higher levels and lighter ones indicating lower levels. All sub-indicators follow the same pattern of “deep in the east, medium in the center, and shallow in the west”, underscoring regional heterogeneity in urban development. This can be attributed to the fact that the eastern region exhibits a more developed economy, superior infrastructure, and more sophisticated production technologies in comparison to the central and western regions. Based on these findings, this paper uses subregional regression in the heterogeneity test, dividing China into eastern, central, and western regions to explore differences in how decentralization affects high-quality urban development.

4.1.2. Analysis of Fiscal and Financial Decentralization Indicators

After measuring the level of fiscal and financial decentralization for each prefecture-level city within the sample, this paper reveals significant regional differences in the spatial distribution of these two variables. Figure 4 presents the mean values of the decentralization indicators for each city over the sample period. Figure 4a focuses on the spatial distribution of $DFR$, which clearly depicts that cities in the eastern region have higher levels of fiscal decentralization. This implies that these local governments have stronger dominance over fiscal resources. In contrast, cities in the central region exhibit a moderately diminished level, while the western region demonstrates the lowest level. Figure 4b presents the spatial distribution of $DFL$, which exhibits highly similar characteristics to Figure 4a. This phenomenon may be attributed to the intrinsic constraints of the economic development process in the central and western regions. These regions exhibit a comparatively constrained capacity to attract external investment, which renders them more reliant on the central government’s comprehensive planning and resource allocation at the national level. The central government aids with urban construction and development in the central and western regions through the transfer of fiscal funds or the provision of financial support. As a result of this process, the capacity of local governments to regulate fiscal and financial resources within their respective regions is constrained, thereby indicating a diminished degree of decentralization.

4.2. Empirical Results

4.2.1. Baseline Regression Model Results

The paper presents the findings of the baseline regression of the model outlined in Equation (1), as illustrated in

Table 3. Baseline regression results.

	$\mathit{HUDI}$		$\mathit{Innovation}$	$\mathit{Coherence}$	$\mathit{Greening}$	$\mathit{Openness}$	$\mathit{Sharing}$
	(1)	(2)	(3)	(4)	(5)	(6)	(7)
$DFR$	3.393 **	3.884 ***	1.932 **	0.133 **	−0.006	1.161 ***	0.664 ***
	(2.54)	(2.91)	(2.49)	(1.97)	(−0.64)	(2.79)	(2.98)
$DFL$	0.311 *	0.333 **	0.213 **	−0.002	−0.003 ***	0.069 *	0.056 **
	(1.83)	(2.12)	(2.29)	(−0.34)	(−3.37)	(1.69)	(2.13)
$Cross$	−1.185 **	−1.159 **	−0.726 **	0.008	0.006 ***	−0.240 *	−0.207 **
	(−2.26)	(−2.37)	(−2.49)	(0.63)	(2.86)	(−1.88)	(−2.57)
$Pop$	——	0.768 ***	0.417 ***	−0.005	−0.003	0.217 *	0.142 ***
		(3.08)	(3.33)	(−0.20)	(−0.67)	(1.88)	(3.15)
$Fin$	——	0.001	−0.003	0.002	0.000	0.001	0.001
		(0.05)	(−0.58)	(1.18)	(1.52)	(0.24)	(0.26)
$Ast$	——	−0.076	0.036	−0.005	0.001	−0.108	0.000
		(−0.45)	(1.05)	(−0.65)	(0.57)	(−0.74)	(0.02)
$Ind$	——	−0.965	−0.176	−0.778 ***	0.010	0.042	−0.063
		(−1.63)	(−0.69)	(−10.46)	(0.82)	(0.12)	(−0.66)
$Hum$	——	−0.418	0.335	0.012	−0.003	−0.262	−0.500
		(−0.26)	(0.36)	(0.10)	(−0.11)	(−0.64)	(−1.17)
$Tsp$	——	0.000	0.005	−0.002 **	0.000	−0.003	0.000
		(0.04)	(0.86)	(−2.57)	(0.74)	(−0.68)	(0.20)
$Mkt$	——	0.633 ***	0.303 ***	−0.057 **	−0.001	0.280 ***	0.107 ***
		(3.34)	(3.04)	(−2.15)	(−0.39)	(3.10)	(3.16)
Constant	1.249 ***	−10.234 ***	−6.390 ***	1.800 ***	0.126 ***	−3.791 **	−1.980 ***
	(3.76)	(−3.29)	(−4.21)	(5.33)	(3.54)	(−2.25)	(−3.91)
Fixed Effect	Yes	Yes	Yes	Yes	Yes	Yes	Yes
N	4692	4692	4692	4692	4692	4692	4692
$R^2$	0.374	0.392	0.370	0.609	0.371	0.089	0.384
F-statistic	16.776	14.283	16.314	40.055	20.392	3.944	23.721

Note: t-statistics are in parentheses. ***, **, and * indicate that the coefficients are significant at 1%, 5%, and 10% significance, respectively, and subsequent graphs and tables are presented accordingly unless otherwise indicated.

. The results account for city and year fixed effects and employ robust standard errors for city clustering, thereby ensuring the reliability of the estimates. In particular, columns (1) and (2) of Table 3 concentrate on the $HUDI$ as the dependent variable. Column (1) does not incorporate control variables, whereas column (2) does. In both models, the coefficients of $DFR$ and $DFL$ are significantly positive, indicating that decentralization plays a pivotal role in fostering high-quality urban development. This finding substantiates the assertions put forth in Hypotheses 1 and 2. Moreover, the markedly negative coefficients of the cross-terms of the decentralization indicators indicate that, while each of the two types of economic decentralization exerts a positive influence on high-quality urban development, they do not demonstrate a synergistic enhancement of one another, but rather a mutually inhibitory relationship. In other words, the operation of one decentralization system impairs the effectiveness of the other, indicating that the relationship between decentralization systems is a conflict rather than complementarity. This finding corroborates Hypotheses 3 and 4. Furthermore, it is notable that the explanatory power of the model is enhanced by the inclusion of control variables. In particular, the coefficients of $Pop$ and $Mkt$ are significantly positive, which further underscores the importance of population growth and market mechanisms in facilitating sustained and healthy urban development. From the perspective of population size, the research of Glover and Simon [58] demonstrated that an increase in population density can facilitate the growth of economic efficiency. Concurrently, the expansion of the population may incite the government to furnish more convenient and efficacious public services to satisfy the mounting demand, which, in turn, propels the urban development in the direction of scientization. For example, Chu [59] indicates that a high population density has a positive impact on the advancement of transportation technology. In terms of market size, Head and Mayer [60] found that market potential is a significant driver of per capita income growth. Shan et al. [61] suggests that market size in African countries is a crucial factor in attracting foreign capital inflows, and that these developments have contributed to the high quality of urban development. The findings of this paper are consistent with those of the aforementioned scholars.

The results of the regression with the five sub-indicators as dependent variables are presented in columns (3) through (7) of Table 3. The results demonstrate that decentralization exerts a significant positive influence on all three pivotal indicators, namely $Innovation$, $Openness$, and $Sharing$. Moreover, its synergistic effect is also markedly negative. This finding is consistent with that of column (2), which suggests that decentralization has a significant positive effect on the core areas that drive urban development. However, the impact of decentralization on the two sub-indicators where growth is more lagging, namely $Coherence$ and $Greening$, shows significant differences. With respect to the $Coherence$ indicator, fiscal decentralization has a positive effect on boosting its growth, while financial decentralization has a relatively limited impact. The study conducted by Chen et al. [62] revealed that fiscal decentralization plays a pivotal role in narrowing the income disparity between urban and rural regions. Meanwhile, Nirola et al. [63] provide an illustrative example of India and demonstrate that fiscal decentralization has the effect of reducing the differences between Indian states. These findings are consistent with the conclusion that a fiscal decentralization system has a positive impact on $Coherence$ indicators, as presented in this paper. In contrast to fiscal decentralization, financial decentralization is characterized by a market-oriented approach that may prioritize efficiency and returns in resource allocation. This may, to some extent, contradict the goal of coherence. Consequently, the combined effect of the two in promoting the growth of the $Coherence$ indicator is not significant. Conversely, the decentralization system exerts a discernible inhibitory effect on the $Greening$ indicator. It may be due to the fact that the decentralization system enhances the control of resources by local governments, which, driven by political incentives, are more inclined to invest resources in areas with significant economic benefits, such as heavy industrial sectors. Such sectors are frequently accompanied by elevated pollution emissions, which are not conducive to the green development of cities. A closer examination of the impact of decentralization on environmental pollution reveals that both revenue and expenditure decentralization have contributed to the exacerbation of local environmental pollution problems, as documented in the study by Guo et al. [64]. Another study by Chen and Liu [65], based on Chinese provincial data from 2003 to 2017, also found that fiscal decentralization had a positive impact on environmental pollution. Additionally, the study identified a phenomenon it termed a “race to the bottom” among local governments. Furthermore, when both decentralization regimes exert a minimal influence on the $Greening$ indicator, there is no evident contradiction between them, but rather some constructive interdependencies, albeit relatively modest ones.

4.2.2. Spatial Econometric Model Results

• SDM results for $HUDI$. The global and local Moran’s I confirm that the $HUDI$ exhibits significant positive spatial correlation (the results of Moran’s I are shown in Appendix D), thereby supporting the rationale for constructing the SDM model.

  Table 4. SDM regression results.

  	$\mathit{HUDI}$
  	${\mathit{W}}_{\mathbf{0}\mathbf{–}\mathbf{1}}$	${\mathit{W}}_{\mathit{d}}$	${\mathit{W}}_{\mathit{e}}$
  	(1)	(2)	(3)
  $DFR$	3.520 ***	3.612 ***	3.169 ***
  	(9.78)	(10.06)	(9.91)
  $DFL$	0.320 ***	0.327 ***	0.257 ***
  	(9.40)	(9.60)	(8.36)
  $Cross$	−1.119 ***	−1.139 ***	−0.933 ***
  	(−15.24)	(−15.50)	(−14.05)
  $W\times DFR$	1.826 ***	4.474 ***	2.504 **
  	(3.05)	(3.94)	(2.43)
  $W\times DFL$	0.048	0.421 ***	0.294 **
  	(0.80)	(2.65)	(2.52)
  $W\times Cross$	−0.098	−1.104 ***	−1.250 ***
  	(−0.78)	(−2.82)	(−4.38)
  $\rho$	0.084 ***	0.055	0.688 ***
  	(4.11)	(1.40)	(28.91)
  Control Variables	Yes	Yes	Yes
  Fixed Effect	Yes	Yes	Yes
  N	4692	4692	4692
  $R^2$	0.327	0.328	0.344
  Log L	−7.2 × ${10}^3$	−7.2 × ${10}^3$	−6.8 × ${10}^3$

  Note: t-statistics are in parentheses. ***, and ** indicate that the coefficients are significant at 1%, and 5% significance, respectively, and subsequent graphs and tables are presented accordingly unless otherwise indicated.

  presents a comprehensive overview of the regression results for the SDM models constructed using Equation (2) under three distinct spatial weighting matrices. All models are spatial and year-fixed. The initial observation is that the coefficients of both $DFR$ and $DFL$ in columns (1)–(3) of Table 4 are significantly positive, while the coefficients of the cross terms are significantly negative. This outcome aligns with the findings of the baseline regression analysis. This suggests that Hypotheses 1, 2 and 4 remain valid when spatial correlation is considered. Secondly, the $\rho$ values in columns (1) and (3) of Table 4 are significantly positive, indicating that the $HUDI$ indicator has a significant positive spatial spillover effect. This implies that the high-quality development of neighboring cities can positively promote the development of the local city. It is particularly noteworthy that this positive spillover effect is most significant under the $W_e$ matrix, reaching 0.688. This indicates that the propagation of this spatial impact is mainly realized through economic channels. Conversely, the spatial spillover effect under the $W_d$ matrix is not significant, suggesting that the geographic distance between cities is not a key factor in determining the strength of the spillover effect. Finally, in all three models, the coefficients of $W\times DFR$ and $W\times DFL$ are positive, whereas the coefficient of $W\times Cross$ is predominantly negative. This suggests that decentralization in neighboring cities has a beneficial impact on local high-quality development, but the collective effect of decentralization has a detrimental effect locally. It is important to note, however, that the impact of decentralization in neighboring cities on their own high-quality development and the spatial spillover effects associated with the $HUDI$ imply that the coefficients of the spatial lagged terms of the independent variables do not directly correspond to the strength of the spatial spillover effects. Accordingly, the estimated coefficients of the SDM model are decomposed using the partial differential decomposition method (the results of the decomposition of the effects of the SDM model are presented in Appendix E). The results demonstrate that the decentralization system of neighboring cities exerts a positive influence on the city’s high-quality urban development, thereby substantiating Hypothesis 3 regarding the presence of a positive spatial spillover effect of decentralization.
• SDM results for sub-indicators. This paper also develops SDM models for the five sub-indicators and presents a summary of the results in

  Table 5. SDM regression results of sub-indicators with $W_e$.

  	$\mathit{Innovation}$	$\mathit{Coherence}$	$\mathit{Greening}$	$\mathit{Openness}$	$\mathit{Sharing}$
  	(1)	(2)	(3)	(4)	(5)
  $DFR$	1.624 ***	0.106 ***	−0.003	0.957 ***	0.574 ***
  	(9.64)	(3.45)	(−0.54)	(5.48)	(7.72)
  $DFL$	0.177 ***	−0.004	−0.002 ***	0.048 ***	0.048 ***
  	(10.91)	(−1.46)	(−4.25)	(2.88)	(6.71)
  $Cross$	−0.618 ***	0.016 **	0.005 ***	−0.177 ***	−0.186 ***
  	(−17.66)	(2.47)	(4.14)	(−4.89)	(−12.03)
  $W\times DFR$	1.525 ***	0.609 ***	−0.004	0.106	0.934 ***
  	(2.81)	(6.22)	(−0.19)	(0.19)	(3.88)
  $W\times DFL$	0.215 ***	0.013	−0.003	0.055	0.073 ***
  	(3.50)	(1.13)	(−1.55)	(0.86)	(2.69)
  $W\times Cross$	−0.889 ***	−0.097 ***	0.017 ***	−0.251	−0.283 ***
  	(−5.89)	(−3.62)	(3.23)	(−1.62)	(−4.24)
  $\rho$	0.537 ***	0.126 ***	0.226 ***	0.749 ***	0.263 ***
  	(19.19)	(4.13)	(7.41)	(33.50)	(8.01)
  Control Variables	Yes	Yes	Yes	Yes	Yes
  Fixed Effect	Yes	Yes	Yes	Yes	Yes
  N	4692	4692	4692	4692	4692
  $R^2$	0.337	0.100	0.013	0.060	0.355
  Log L	−3.7 × ${10}^3$	4301.467	1.2 × ${10}^4$	−4.0 × ${10}^3$	114.096

  Note: t-statistics are in parentheses. ***, and ** indicate that the coefficients are significant at 1%, and 5% significance, respectively, and subsequent graphs and tables are presented accordingly unless otherwise indicated.

  . From the perspective of local economic decentralization, the results demonstrate that the local decentralization system significantly and positively contributes to the development of the three dimensions of $Innovation$, $Openness$, and $Sharing$, with a more moderate effect on $Coherence$. However, it negatively contributes to $Greening$ development. It is of particular importance to note that when the influence of decentralization is strong, both factors are in conflict and, thus, no longer exhibit a synergistic effect. When the influence of the decentralized system is relatively limited, the complementary effect between the two is predominant, resulting in effective synergy. These findings are in alignment with the results of the baseline regression analysis.
  With regard to the spatial effects of the neighboring decentralization system, firstly, the positive values of $\rho$ in all models indicate that the growth of neighboring regions in various indicators has a significant promotion effect on the local area, reflecting the positive spatial spillover effect in the process of high-quality urban development. Of these, the $\rho$ values of the $Innovation$ and $Openness$ indicators are particularly noteworthy, indicating a pronounced driving effect of these two factors in the context of inter-regional relations. Secondly, the fiscal decentralization of neighboring cities is demonstrated to significantly contribute to the improvement of local $Innovation$, $Coherence$, and $Sharing$ levels, as evidenced by the considerable positive value of $W\times DFR$. This may be attributed to the fact that the augmented fiscal autonomy of neighboring local governments can more effectively facilitate scientific and technological advancement and elevate the standard of living for residents. The phenomenon of fiscal decentralization is observed to exhibit robust positive spatial spillovers, which can be attributed to the mobility of technological and human resources, as well as the positive externalities associated with scientific and technological progress. In particular, the $Innovation$ indicator is most significantly affected by the spatial spillover of fiscal decentralization, with a coefficient of 1.525. The results of other researchers corroborate the findings of this paper. For example, De Siano and D’uva [66] demonstrated that local governments are able to benefit from increased public spending by neighboring city governments, thereby illustrating positive spillover effects. Similarly, Zallé and Bakouan [67] investigated the influence of fiscal decentralization on the accessibility of public social services for communities in Burkina Faso. The findings of this study further substantiate the spillover impact of fiscal decentralization in facilitating the delivery of public social services. These findings indicate that fiscal decentralization in neighboring cities may facilitate local coordination and resource sharing. In contrast, the growth of the $Greening$ and $Openness$ indicators is more dependent on specific local industrial, energy, and trade policies due to differences in resource endowment and industrial structure, less affected by the flow of fiscal resources and policies of neighboring cities. Finally, the decentralization of financial resources in neighboring cities has a considerable and positive impact on local $Innovation$ and $Sharing$ indicators. This phenomenon may be attributed to the high mobility of financial capital, which, in conjunction with the potential for development and profitability in technology and livelihood sectors, attracts financial resources to these sectors, thereby contributing to an improvement in these two indicators.

5. Heterogeneity Analysis and Robustness Tests

5.1. Heterogeneity Analysis

5.1.1. Regression Results for Eastern, Central, and Western China

Due to the considerable spatial variability of the $HUDI$ and decentralization indicators, this paper adopts the strategy of subdividing the prefecture-level cities into three major regions, namely, eastern, central, and western regions, according to geographic location, and conducting regression analyses separately. The results of the regression analyses are summarized in columns (1)–(3) of

Table 6. Results of the heterogeneity test with $W_e$.

	$\mathit{HUDI}$
	Eastern China	Central China	Western China	2005–2015	2016–2022	Provincial Capitals	General Cities
	(1)	(2)	(3)	(4)	(5)	(6)	(7)
$DFR$	7.840 ***	2.112 ***	0.545	1.726 ***	1.971 ***	3.809	3.239 ***
	(9.00)	(4.56)	(0.80)	(6.25)	(4.01)	(1.44)	(14.34)
$DFL$	1.170 ***	0.192 ***	0.018	0.104 ***	0.210 ***	0.518 **	0.266 ***
	(10.00)	(4.34)	(0.18)	(4.01)	(2.67)	(2.00)	(12.02)
$Cross$	−2.606 ***	−0.666 ***	−0.141	−0.442 ***	−0.496 ***	−0.438	−0.948 ***
	(−14.31)	(−4.84)	(−0.94)	(−8.03)	(−3.83)	(−0.86)	(−19.71)
$W\times DFR$	8.113 ***	1.883	−5.926 ***	−2.659 ***	1.761	8.134	0.212
	(3.99)	(1.48)	(−3.32)	(−2.61)	(1.08)	(1.30)	(0.28)
$W\times DFL$	0.833 ***	0.592 ***	−1.160 ***	−0.005	−0.354	0.739	0.107
	(2.66)	(4.71)	(−4.03)	(−0.06)	(−1.52)	(1.36)	(1.25)
$W\times Cross$	−2.045 ***	−2.199 ***	3.405 ***	−0.016	−0.235	−2.701 **	−0.661 ***
	(−3.66)	(−5.17)	(5.21)	(−0.06)	(−0.45)	(−2.38)	(−3.00)
$\rho$	0.272 ***	0.408 ***	0.161 ***	0.509 ***	0.700 ***	0.497 ***	0.309 ***
	(7.00)	(8.78)	(2.69)	(15.25)	(18.25)	(9.56)	(8.67)
Control Variables	Yes	Yes	Yes	Yes	Yes	Yes	Yes
Fixed Effect	Yes	Yes	Yes	Yes	Yes	Yes	Yes
N	1649	1683	1360	2760	1932	459	4233
$R^2$	0.476	0.429	0.221	0.093	0.513	0.694	0.455
Log L	−2.6 × ${10}^3$	−2.1 × ${10}^3$	−2.1 × ${10}^3$	−2.3 × ${10}^3$	−2.3 × ${10}^3$	−902.099	−4.5 × ${10}^3$

Note: t-statistics are in parentheses. ***, and ** indicate that the coefficients are significant at 1%, and 5% significance, respectively, and subsequent graphs and tables are presented accordingly unless otherwise indicated.

. (This paper presents only the regression results of the SDM model under the $W_e$ matrix. The conclusions for other matrices are identical and are not reiterated here.) A review of the data in Table 6 reveals that the impact coefficient of fiscal decentralization in the eastern region is 7.840, while that of financial decentralization is 1.170. Both of these coefficients are significantly higher than those observed in the central and western regions. Additionally, the negative effect of the interaction term between the two is also the most significant. This phenomenon can be attributed to the high degree of decentralization of local governments in the eastern region, which has a strong ability to deploy resources and a rich reserve of urban construction resources, thus promoting a higher level of high-quality urban development. In contrast, the efficacy of local governments in fostering high-quality urban development in western cities is constrained by the comparatively lower degree of decentralization.

A further examination of the spatial spillover effects reveals that the performance of cities in eastern and central China is generally consistent with the findings of the full sample analysis. However, in western China, the spatial spillover effects of the decentralization system are negative. This discrepancy may be attributed to the fact that, in the western region, where high-quality development is comparatively less advanced, cities are particularly resource-intensive, resulting in heightened competition among local governments. When neighboring cities have a high level of decentralization, the existence of a siphon effect may draw development resources from neighboring cities to themselves, thus adversely affecting local high-quality development. In this context, the “competition effect” among western cities is evidently more pronounced than the “mobility effect” and the “demonstration effect”. This ultimately gives rise to a negative spatial spillover effect.

5.1.2. Regression Results for the Pre- and Post-Reform Time Periods

In 2016, China’s State Council issued the Guiding Opinions on Promoting the Reform of the Division of Fiscal Affairs and Expenditure Responsibilities, aimed at restructuring and optimizing the fiscal relationship between central and local governments to support sustainable economic and social development. These reforms marked a new phase in fiscal decentralization, institutionalizing and standardizing the system to better promote urban development. Simultaneously, financial decentralization reforms continued, with local governments expanding inclusive financial services, boosting support for sectors like agriculture and rural areas. The government also strengthened financial regulation to ensure efficient use of financial capital for local economic transformation and growth.

In order to assess the specific effectiveness of this series of reform policies, this study takes 2016 as the time point, divides the sample data into two different time periods, and carries out group regression analysis. The results are detailed in columns (4) and (5) of Table 6. The results of the analysis demonstrate that since the implementation of the reform in 2016, the coefficient of the positive impact of the decentralization system on the $HUDI$ has been significantly enhanced, while the $\rho$ value has also achieved a significant increase. This finding substantiates the assertion that the reform measures have not only effectively promoted the positive effect of the decentralization system on the city’s high-quality development, but also significantly enhanced the radiation and driving effect of this development outcome on neighboring regions.

5.1.3. Regression Results for Cities at Different Administrative Levels

This paper presents a heterogeneity analysis that investigates the potential influence of diverse administrative structures within urban areas on the outcomes of the study. It places particular emphasis on provincial capitals, which serve as the political and economic hubs of their respective provinces. As these cities tend to enjoy more significant policy preferences and resource concentration advantages, the promotion of high-quality development by decentralized regimes may exhibit different characteristics from non-capital cities. For this reason, this paper employs a subgroup regression analysis and presents the results in columns (6) and (7) of Table 6.

The regression results indicate that the impact of $DFR$ is not statistically significant for provincial capital cities. Conversely, the impact of financial decentralization is significantly higher than that of other ordinary prefecture-level cities. This finding may be attributed to the fact that in more prosperous provincial capital cities with higher levels of development, the direct pull effect of local government fiscal resources may have reached a point of diminishing marginal efficiency, leading to a relatively weaker role of fiscal decentralization in promoting high-quality urban development. In contrast, financial decentralization, with its more flexible and market-oriented characteristics, is better suited to adapt to and serve the high-quality development needs of these cities. Furthermore, the capital cities exhibit elevated $\rho$ values, suggesting the presence of more pronounced spatial spillovers associated with their high-quality development levels.

5.2. Robustness Tests

5.2.1. Replacing Core Independent Variables

As previously indicated, fiscal decentralization can be represented by $DFE$ and $DFF$, while financial decentralization can be represented by $DFD$. To test the robustness of the construction of the decentralization indicators in this paper, we present the regression results after replacing the original independent variables with the different decentralization indicators in

Table 7. Replacement of independent variables in robustness test with $W_e$.

	$\mathit{HUDI}$
	(1)	(2)	(3)
$Dfis$	2.383 ***	1.090 ***	4.648 ***
	(8.00)	(7.94)	(13.58)
$Dfin$	0.381 ***	0.142 ***	0.519 ***
	(10.36)	(4.02)	(9.36)
$Cross$	−1.100 ***	−0.356 ***	−1.549 ***
	(−14.87)	(−8.24)	(−17.80)
$W\times Dfis$	1.223	3.052 ***	4.545 ***
	(1.20)	(5.59)	(4.23)
$W\times Dfin$	0.334 **	0.568 ***	0.736 ***
	(2.28)	(4.40)	(4.25)
$W\times Cross$	−1.076 ***	−1.053 ***	−2.220 ***
	(−3.43)	(−6.29)	(−6.47)
$\rho$	0.717 ***	0.684 ***	0.646 ***
	(31.12)	(27.93)	(25.84)
Control Variables	Yes	Yes	Yes
Fixed Effect	Yes	Yes	Yes
N	4692	4692	4692
$R^2$	0.339	0.340	0.380
Log L	−6.8 × ${10}^3$	−6.9 × ${10}^3$	−6.7 × ${10}^3$

Note: $Dfis$ in columns (1)–(3) refers to $DFE$, $DFF$, and $DFR$, respectively. $Dfin$ in columns (1) and (2) denotes $DFL$, while in column (3), it denotes $DFD$. t-statistics are in parentheses. ***, and ** indicate that the coefficients are significant at 1%, and 5% significance, respectively, and subsequent graphs and tables are presented accordingly unless otherwise indicated.

. As can be observed, all decentralization indicators exhibit a statistically significant positive effect at the 1% level, while the cross terms display a significant negative effect, thereby corroborating Hypotheses 1, 2 and 4. Furthermore, the coefficients $\rho$, associated with the spatial lag terms of decentralization indicators and cross terms, are identical to those presented in Table 4, thereby substantiating Hypothesis 3 proposed in this paper.

5.2.2. Replacing the $HUDI$ Indicator Construction Method

In order to ascertain the reliability of the urban high-quality development index, this paper employs a range of OVA methods to reconstruct the $HUDI$ indicator. Specifically, the weights of each indicator are determined by applying the entropy weight method, and the ${HUDI}_{topsis}$ indicator is constructed by combining the TOPSIS method. Furthermore, this paper employs the method of PCA, CV, and CRITIC to reconstruct the urban high-quality development indicators in accordance with the indicator system delineated in Table 1, resulting in the generation of ${HUDI}_{pca}$, ${HUDI}_{cv}$, and ${HUDI}_{critic}$. Subsequently, the SDM model is reconstructed to conduct regression analysis with these indicators as dependent variables. The results are presented in columns (1) to (4) of

Table 8. Replacement of dependent variables and spatial weighting matrix in robustness test.

	${\mathit{HUDI}}_{\mathit{topsis}}$	${\mathit{HUDI}}_{\mathit{pca}}$	${\mathit{HUDI}}_{\mathit{cv}}$	${\mathit{HUDI}}_{\mathit{critic}}$	$\mathit{HUDI}$
	${\mathit{W}}_{\mathit{e}}$				${\mathit{W}}_{\mathit{g}}$
	(1)	(2)	(3)	(4)	(5)
$DFR$	2.482 ***	0.855 ***	2.811 ***	2.007 ***	3.719 ***
	(6.43)	(9.43)	(9.24)	(5.82)	(10.54)
$DFL$	0.144 ***	0.057 ***	0.215 ***	0.146 ***	0.332 ***
	(3.86)	(6.56)	(7.34)	(4.39)	(9.90)
$Cross$	−0.584 ***	−0.229 ***	−0.787 ***	−0.516 ***	−1.138 ***
	(−7.29)	(−12.16)	(−12.47)	(−7.21)	(−15.75)
$W\times DFR$	3.143 **	1.014 ***	2.458 **	6.509 ***	5.249 ***
	(2.53)	(3.46)	(2.51)	(5.90)	(5.07)
$W\times DFL$	0.389 ***	0.068 **	0.256 **	0.293 **	0.280 *
	(2.77)	(2.05)	(2.31)	(2.34)	(1.89)
$W\times Cross$	−1.550 ***	−0.308 ***	−1.070 ***	−1.404 ***	−0.725 ***
	(−4.52)	(−3.81)	(−3.95)	(−4.61)	(−2.96)
$\rho$	0.588 ***	0.550 ***	0.747 ***	0.110 ***	−0.164 ***
	(23.25)	(20.08)	(33.59)	(3.39)	(−7.55)
Control Variables	Yes	Yes	Yes	Yes	Yes
Fixed Effect	Yes	Yes	Yes	Yes	Yes
N	4692	4692	4692	4692	4692
$R^2$	0.255	0.383	0.368	0.684	0.303
Log L	−7.7 × ${10}^3$	−855.491	−6.6 × ${10}^3$	−7.1 × ${10}^3$	−7.2 × ${10}^3$

Note: t-statistics are in parentheses. ***, **, and * indicate that the coefficients are significant at 1%, 5%, and 10% significance, respectively, and subsequent graphs and tables are presented accordingly unless otherwise indicated.

. The regression results demonstrate that Hypotheses 1, 2, 4 and 5 of this paper are effectively verified regardless of the $HUDI$ constructed using any OVA method. This indicates that the conclusions of this paper are robust and reliable.

5.2.3. Replacing the Spatial Weighting Matrix

In this paper, spatial weighting matrices are constructed around the three dimensions of adjacency, geographic distance, and economic impact, as previously studied. However, Ulucak et al. [68] and Zhao et al. [69] employed an innovative approach by utilizing the gravity model to quantify the degree of spatial closeness of association. Therefore, this paper establishes a spatial weighting matrix $W_g$ (the methodology employed in the construction of $W_g$ is detailed in Appendix F), which measures the attractiveness of high-quality development between cities, based on the $HUDI$, and employs it to construct the SDM model. The resulting results are presented in column (5) of Table 8. The analysis demonstrates that the decentralized system continues to demonstrate a positive effect, while the effect of joint implementation exhibits a dampening effect. This finding is consistent with the core theory of this paper. It is noteworthy that the $\rho$ values in column (5) are significantly negative, indicating an intriguing phenomenon: the process of high-quality development in one city does not contribute to the development of neighboring cities with similar levels of high-quality development but rather has a negative impact. This phenomenon may be attributed to the competition for resources among these cities due to the proximity of high-quality development levels. Furthermore, there may be a more intense competition for resources among local governments, which precisely verifies Hypothesis 5 in this paper. This hypothesis states that the existence of competition effects among local governments may lead to negative spatial spillover effects.

6. Conclusions and Discussion

This paper explores how China’s economic decentralization impacts high-quality urban development. Using data from 276 prefectural cities between 2006 and 2022, the study analyzes five core development dimensions: innovation, coherence, greening, openness, and sharing, applying panel and spatial models to evaluate both direct and spatial spillover effects. The key findings are as follows: (1) fiscal and financial decentralization positively impact innovation, openness, and sharing but have limited effects on greening and coherence. Financial decentralization, in particular, hampers green development. (2) The two decentralization types lack synergy, often conflicting unless one system weakens. (3) There is a notable “local-neighborhood” spillover effect, where decentralization in neighboring cities benefits local development, but synergy remains negative. (4) Regional disparities show that cities in the eastern region benefit most from decentralization, while the western region lags. Post-2016 reforms strengthened both local and spatial spillover effects. The provincial capital cities benefited more from financial decentralization, with other prefecture-level cities from fiscal decentralization. (5) Robustness tests confirm the findings, even after reconstructing the models and indices.

This paper recommends several policy measures to enhance the effectiveness of China’s decentralization system. First, to resolve the conflicts between fiscal and financial decentralization, the system design should be optimized to foster cooperation, allowing the two systems to complement each other. A cross-sectoral coordination mechanism should be implemented to regularly assess how fiscal and financial resources are utilized, ensuring efficient resource use and preventing financial risks. This approach would help achieve innovation, openness, and sharing while minimizing inherent conflicts between the two decentralization systems.

Second, decentralization policies should be regionally differentiated to account for varying levels of development across eastern, central, and western China. The eastern region should focus on advancing financial decentralization to promote capital flows and financial innovation, while the central region would benefit from a balanced mix of fiscal and financial decentralization to improve capital efficiency and project capacity. The western region should prioritize strengthening fiscal decentralization, focusing on infrastructure development and public service investment to reduce regional disparities.

Third, the government should enhance policies for green and coherent development. Since financial decentralization has a limiting effect on green development, policy adjustments should promote green financial products such as green bonds and green credit, providing low-cost financing for environmental protection initiatives. Strengthening environmental risk assessments for financial institutions would also help direct more funding toward green industries. In addition, support for coherent development should be increased by establishing a regional coherent development fund to encourage resource sharing and collaborative infrastructure development, especially between urban and rural areas.

Finally, the government should fully leverage the spatial spillover effects of decentralization to boost synergistic regional growth. By fostering cooperation and resource sharing among neighboring cities, through initiatives like regional development alliances, joint investment promotion, and industrial park development, regions can achieve shared economic growth. In particular, capital cities, given their pivotal role in financial decentralization and strong spillover effects, should be supported through targeted policies and resource allocation, enabling them to act as leaders in driving regional development and facilitating shared prosperity with surrounding areas.

While this paper presents meaningful findings regarding the role of China’s economic decentralization system in facilitating high-quality urban development, it also acknowledges several research limitations and potential directions for future studies. The first limitation is that the SDM model employed may not fully capture the complexity of urban interactions. As LeSage and Fischer [51] notes, the assumption of a linear relationship between cities may overlook the intricate, nonlinear dynamics that often characterize these interactions. Consequently, this linear assumption could lead to biased results. Moreover, the model does not address the issue of endogeneity. The relationship between fiscal and financial decentralization and high-quality urban development may be influenced by unobserved factors, resulting in biased estimates. Failure to account for endogeneity can impair the accurate assessment of policy effects, as highlighted by Pickup and Evans [70], thereby limiting the study’s validity. Additionally, this study does not consider the influence of prior high-quality urban development on current outcomes, omitting dynamic effects. Dynamic models can effectively capture the lagged impacts of policy changes, as Bollen [71] emphasizes, and neglecting these effects may lead to misinterpretations of causality. Lastly, this paper inadequately addresses the interaction between central and local governments, which is crucial for understanding policy impacts and implementation complexities. The coordination and conflicts between these levels of government can significantly affect policy effectiveness [72]. Future research should, therefore, focus on this interactive relationship to achieve a more comprehensive understanding of how economic decentralization influences high-quality urban development.

Given the limitations of this study, future research could benefit from several enhancements. First, to address the issue of endogeneity, the instrumental variable approach should be employed. By selecting appropriate instrumental variables, researchers can effectively eliminate endogeneity bias between dependent and independent variables, thereby improving the accuracy and reliability of estimation results. Second, the dynamic spatial Durbin model could be used to examine the impact of prior high-quality urban development on current outcomes. This model allows for the inclusion of lagged dependent variables, facilitating a comprehensive understanding of the temporal continuity of policy effects and revealing how dynamic changes influence urban development. Finally, to gain deeper insights into the interactions between central and local governments, future studies might employ dynamic stochastic general equilibrium modeling or structural equation modeling. These methodologies can effectively capture the complex interactions involved in policy implementation, enabling an analysis of how coordination and conflicts between different levels of government affect the achievement of high-quality development. Incorporating these enhancements will provide a more nuanced examination of the influence of economic decentralization on urban high-quality development, thereby offering a more empirically grounded foundation for policy formulation.