Measurement and Spatio-Temporal Evolution Analysis of the Business Environment in the Guangdong–Hong Kong–Macao Greater Bay Area

Abstract

Cultivating the best business environment ecosystem is important for advancing market-oriented reforms and achieving sustainable industrial transformation. How to quantify the business environment is also a relatively complex topic. Based on urban ecology theory, this study constructs a comprehensive evaluation framework for assessing urban business environment development. Using the entropy weight method and spatial autocorrelation analysis, we examine the time series and spatial evolution of the business environment in the Guangdong–Hong Kong–Macao Greater Bay Area from 2008 to 2021. Meanwhile, we further explore the main factors that influence the development level of the business environment. Finally, some suggestions are put forward to improve the business environment. The results show that (1) the development level of the business environment has gradually improved during the sample period, with stable growth from 2008 to 2015, followed by rapid development after 2015; (2) from different dimensions, there is an imbalance in the business environment development among cities within the Greater Bay Area, with core cities performing better than others; (3) from a spatial perspective, the business environment presents a “core-periphery” pattern, with higher levels clustered around the Pearl River Estuary, indicating strong spatial agglomeration. This research provides theoretical support and policy recommendations for the Three-Year Action Plan for Creating a World-Class Business Environment in the Greater Bay Area.

1. Introduction

Economic globalization is encountering significant challenges during the post-pandemic recovery phase, particularly the increasing prevalence of protectionism and unilateralism. The evolving geopolitical landscape, characterized by deglobalization trends and techno-economic fragmentation, has precipitated systemic volatility in China’s industrial upgrading processes, which has affected the multilateral recovery framework led by the World Trade Organization. The Chinese economy has transitioned into a “new normal” phase, shifting from high-speed expansion to a more sustainable and higher-quality development path. Many studies have shown that a favorable business environment can promote economic growth [1,2,3,4,5]. It plays a central role in shaping social capital flows and driving economic progress, acting not only as a significant factor but also a crucial determinant of these processes [6]. The Chinese government has prioritized the business environment as a key driver of sustainable development. In recent years, it has implemented a series of reforms focusing on simplifying administrative procedures, reducing enterprise burdens, and improving administrative efficiency and transparency. These initiatives aim to create a more favorable operating environment for market entities and have achieved remarkable results. According to the World Bank’s Doing Business Report, China has become one of the most open economies and significantly improved its global ranking in the world. This progress demonstrates the government’s commitment to building a more favorable business environment that supports enterprise growth and innovation.

The business environment encompasses a complex and multifaceted set of factors that collectively shape market dynamics. It represents a systemic transformation that transcends sectoral boundaries through institutional interdependencies, simultaneously influencing market participants across various stages of their lifecycle. At the macro level, the business environment encompasses six various dimensions, such as political governance, economic trends, social and cultural environment, technological innovation, legal–regulatory environment, and environmental sustainability. These dimensions form a dynamic ecosystem that encompasses regulatory policies, competitive landscapes, and socioeconomic conditions, which collectively shape the experiences of market participants throughout their entire lifecycle, such as entry, operation, and exit [7,8]. From a micro perspective, enterprises constitute the primary engine of aggregate economic prosperity. Improving the business environment leads to a decrease in the institutional transaction costs faced by market participants, including entry costs and operational expenses. This strategic reallocation would facilitate the efficient transfer of key resources like labor, capital, and technology from less productive sectors to high-performing segments across interconnected industrial value chains. This operational transformation does not merely improve the quality of production factors but also enhances allocative efficiency through strategic resource orchestration, directing these inputs toward advanced productivity orientations. This systemic alignment establishes an equitable competitive environment while fostering sustainable industrial upgrading. Through structural optimization and advanced transformation of industrial ecosystems, it enhances industrial productivity and establishes a fundamental pathway for advancing new quality development [9,10,11,12]. The business environment ecosystem is an important carrier and measure for enhancing regional carrying capacity and promoting the innovation and entrepreneurship of market entities [13]. It plays a crucial role in enhancing regional total factor productivity while advancing the new quality development model through innovative resource management.

Since the 21st century, research on the business environment has primarily focused on large urban agglomerations. The Guangdong–Hong Kong–Macao Greater Bay Area (GBA) is an outstanding representative of China’s urban agglomerations. With its economic foundation, abundant innovation resources, and a mature market system, it has become a pioneering area for exploring new models of economic growth. When benchmarked against global counterparts like New York, San Francisco, and Tokyo Bay areas, the GBA demonstrates notable comparative advantages in population size and demographic scale. By establishing a world-class business environment, the region seeks to open new avenues for innovation and expansion. In 2021, the Chinese government officially launched a pilot program to reform and optimize the business ecosystem, designating Guangzhou and Shenzhen as pilot cities for business environment innovation. This initiative assigns these cities a leading role in driving reforms and pioneering experiments. Meanwhile, municipal cities like Dongguan, Foshan, and Zhaoqing are undertaking comprehensive institutional reforms aimed at optimizing their commercial ecosystems. These cities have introduced a series of laws, regulations, and policy measures to promote the overall business environment. The core objective is to stimulate innovation and growth among market entities. By deepening reforms, expanding openness, and increasing policy support, the initiative will enhance domestic economy vitality through institutional impetus. This approach aims to accelerate transformation toward high-value-added economic development and resource-efficient growth paradigms within the GBA.

Although domestic scholars have made extensive research on the business environment, many existing studies often adopt a single perspective and theoretical foundation, lacking systematic integration of multi-dimensional external factors in macro-environmental analysis. This paper focuses on a region characterized by distinct economic and geographical features and innovatively employs urban ecosystem theory as an analytical framework to construct a comprehensive evaluation system. This system not only encompasses a multi-dimensional business environment but also systematically incorporates the spatial–temporal modeling of regional economic development. Then, this paper conducts an in-depth analysis of the temporal and spatial characteristics of the business environment within the GBA, examining the evolutionary patterns and development trends at different time points and spatial layouts. Through this comprehensive analysis, a more in-depth understanding of the development status and regional disparities in the business environment can be achieved. Finally, in order to further reveal the spatial correlation and clustering characteristics of the business environment, this paper uses Moran scatter plots for spatial autocorrelation analysis. This methodological method not only enriches the analytical framework of business environment research but also provides a more scientifically accurate basis for policy formulation and practical operation.

The organizational structure of this paper is presented as follows. The second section reviews the evolution patterns of the business environment and establishes a theoretical foundation for empirical analyses. By summarizing and organizing the relevant literature, it outlines the theoretical context and development process, clarifying the research focus and innovations. The third section constructs a comprehensive set of development indicators and measurement methods based on the objectives and conditions of the business environment. This system considers the economic characteristics, industrial structure, and policy orientation of the GBA to ensure accurate and effective measurement. The fourth section explores the indicator measurement system from temporal and spatial evolution perspectives. In terms of time series analysis, this study reveals the development trend and evolution of the business environment by longitudinally comparing historical data. In the spatial evolution analysis, the Moran index is used to examine the dynamic characteristics of the business environment. Through quantitative analysis of spatial autocorrelation patterns, this study examines the spatial agglomeration trajectories and identifies key determinants influencing the regional business environment. The findings provide empirical foundations to optimize the business environment across nine mainland cities and two special administrative regions.

This study may make three theoretical contributions. First, it employs the entropy method and spatial autocorrelation analysis to construct a comprehensive and scientifically analytical framework for evaluating the business environment. The entropy method objectively assigns weights based on the information entropy, avoiding biases caused by subjective weighting and offering a more accurate reflection of the importance of multi-dimensional indicators such as infrastructure, policy support, market vitality, and innovation capacity in the comprehensive evaluation. Compared with traditional methods that rely on subjective judgments, the spatial autocorrelation analysis reveals regional differences and spatial patterns, helping to identify core–edge structures and provide coordinated regional development. Unlike static city-level evaluations, this study explores the mechanisms behind spatial clustering, providing broader insights for policymaking. Second, the study selects the Greater Bay Area as a sample and introduces localized factors based on its unique geo-economic characteristics and policy background. This improves the accuracy of reflecting dynamic changes in the business environment and providing more practical policy guidance. Meanwhile, using time series analysis from 2008 to 2021, the research identifies two major stages: the stable development period from 2008 to 2015 and the rapid development stage after 2015, providing forward-looking suggestions. Third, this research links urban ecology theory with empirical analysis, providing new perspectives for the interdisciplinary research between urban ecology and regional economics. By applying ecosystem theory to the business environment of the Greater Bay Area, this study not only verifies its relevance but also proposes a new perspective on how urban business systems evolve. This integration enriches urban ecology and provides the development of regional economics through new methodologies and theoretical insights.

Furthermore, at the practical value level, the study examines the development level of the business environment and its spatial distribution characteristics within the Greater Bay Area. This helps one to understand how different regions of the GBA are performing and evolving in terms of business conditions. It provides empirical support for the “Three-Year Action Plan to Create a World-Class Business Environment” and proposes a series of specific policy recommendations. These suggestions are not only applicable to the Greater Bay Area but also serve as valuable references for other regions aiming to optimize their business environments. By strengthening infrastructure, improving policy support, promoting digital transformation, and enhancing regional collaboration, the Greater Bay Area and other regions can further optimize their business environments and drive high-quality regional economic development.

2. Literature Review

2.1. The Connotation of Business Environment

The launch of the World Bank’s Doing Business project in 2002 represented a turning point in global economic policy discussions. It introduced a standardized system for assessing regulatory environments and spurred international comparative research on optimizing business environments. Since then, numerous domestic and international institutions have carried out extensive studies focusing on various aspects of the business environment, including its definition, indicators construction, identification of influencing factors, and the provision of policy recommendations. Although the definitions of the business environment have been widely debated in the academic community, no consensus has been reached. This ongoing discussion highlights the complexity of multi-dimensional assessments of the business environment.

According to the existing research [14,15,16], the business environment can generally be divided into broad and narrow definitions. The broad definition encompasses multiple factors, such as politics, economy, society, and law, which collectively constitute the macro environment for business operations. These factors have a profound and multi-dimensional impact on various aspects of a business lifecycle, including enterprise establishment, production and operation, import and export trade, contract fulfillment, and bankruptcy application [17]. In contrast, the narrow definition refers to government regulations, policies, and institutional mechanisms that directly affect the economic activities of market entities. This reflects the overall development level of a country or region regarding policy environment, market conditions, legal systems, and social environment. The business environment is closely related to various factors such as economic growth, the political system, social development, market conditions, legal environment, and infrastructure [18,19,20,21]. The business environment assessment needs to adopt a systemic approach that considers the interactions and influences among various factors [22]. This conceptual framework views the business environment as an integrated ecosystem of external factors that influence corporate decision-making in areas such as entrepreneurial activities, research and development efforts, capital procurement processes, and financial portfolio management. It includes resource conditions and environmental factors that enterprises face, like capital, talent, technology, policies, and institutions.

2.2. Evaluation Indicators for Business Environment

Following the definition of the business environment, the academic community began to conduct quantitative assessments of its development status. The scientific selection of evaluation indicators is crucial for improving measurement effectiveness. The varying theoretical approaches have led to a dual-axis classification in evaluation systems, differentiating between operation-focused metrics from the enterprise perspective and governance indicators related to institutional frameworks. The first category consists of an indicator system developed by the World Bank that directly reflects the convenience of business through eleven aspects, spanning business entry procedures, construction permit acquisition, electricity availability, access to credit, safeguarding minority investors, tax compliance, cross-border operations, enforcement of contracts, resolution of insolvency, and workforce employment. This evaluation system is globally recognized as an important standard for measuring the business environment. Furthermore, there are other indicators, such as the marketization index [23], the cost of doing business in taxation [24], and the institutional policy system [25], which also reflect the convenience and costs of business operations. The second category increasingly focuses on constructing a macro-indicator system to assess the business environment’s overall status and development trend. The Economist Intelligence Unit has developed a multi-dimensional evaluation framework integrating eleven dimensions, including political landscape, macroeconomic conditions, market prospects, free market and competition regulations, foreign investment strategies, foreign trade and exchange rate management, tax rate, financing, labor market, and infrastructure [26]. This system provides a powerful tool for macro-level assessment of the business environment.

Through a systematic synthesis of the existing scholarly literature on index design methodologies, this study proposes an integrated approach to construct a multi-dimensional evaluation framework that combines quantitative metrics with qualitative indicators. This paper also found that, when analyzing the business environment from an urban perspective, the majority of scholars have established extensive evaluation index systems based on the urban ecosystem theory [7,24]. These systems encompass multiple dimensions, such as the natural geographic environment, economic environment, social environment, financial environment, infrastructure environment, and government legal environment. By incorporating the direct convenience of enterprise operations and macro-level urban development factors, these indicator systems provide strong support for the comprehensive assessment of the urban business environment [27,28,29]. Furthermore, academic and governmental organizations like the China Urban Business Environment Evaluation Project and the Guangzhou Greater Bay Area Research Institute have developed comprehensive macro-level evaluation systems for business environments. These frameworks provide essential analytical tools that support both policy formulation and academic research. The establishment of these macro-indicator systems not only enriches the content and methods of evaluation but also promotes a deeper understanding and development of the field.

2.3. The Measurement of Business Environment

The World Bank’s Doing Business Report has emerged as a critical benchmark for cross-border evaluations of national business environments, with its findings attracting academic and policy attention. This framework not only facilitates international comparisons across economies but also contributes to policy reform and academic discourse through methodological innovations. Drawing on the analytical framework developed in this study, scholars from academic and international organizations have conducted extensive studies on the business environment changes using a three-dimensional analysis method. This approach explores national policy structures, regional institutional arrangements, and enterprise operational contexts. To improve methodological precision, various quantitative analytical techniques have been adopted, including entropy-based weighting methods [30], multi-criteria decision analysis like TOPSIS, principal component analysis, and the analytical hierarchy process [31,32,33]. The World Bank’s business environment assessment indicators have been reorganized into three main categories: business facilitation, legal framework, and global integration [34]. By enabling systematic international comparisons, this methodology generates an integrated evaluation framework that surpasses conventional assessment methods in terms of methodological precision.

From a national perspective, the overall quality of the urban business environment shows an upward trend. However, this optimization process is relatively slow, and the internal indicators are unevenly developed, resulting in challenges in the transfer of development level ranking [30]. Regional studies have shown that the relative advantages of the business environment in some municipalities directly administered by the central government, separately planned cities, and provincial capitals are becoming gradually prominent. This trend is especially evident in key regions such as the Greater Bay Area, the Yangtze River Delta, and the Beijing–Tianjin–Hebei regions, where the urban business environment generally performs well [35]. The three municipal capitals of Beijing, Shanghai, and Guangzhou occupied the top three positions nationally, whereas East China demonstrated regional dominance, with a composite index 15.8% higher than the national average [36]. Utilizing World Bank’s Doing Business survey datasets (2018–2022), studies reveal statistically significant correlations between regulatory reforms and corporate R&D investment trajectories, with GDP expansion effects observed across 28 provinces [37,38,39].

3. Construction and Measurement Methods of Business Environment Development Indicators

3.1. Theoretical Foundation

In order to examine the business environment from an urban perspective, the academic community has increasingly adopted an ecosystem framework. The concept of an ecosystem was first introduced into the field of biology by British ecologist Arthur Tansley in 1935 [40]. He defined it as an interactive system between a biotic community and its surrounding habitat. It was not until the 1990s that the notion of a “business ecosystem” was formally proposed by American scholar Moore in the Harvard Business Review. Moore viewed enterprises as integral members of a business ecosystem, emphasizing that, while competition exists among firms, collaboration with suppliers, customers, and even competitors is essential for mutual development [41]. According to Moore’s definition, an ecosystem is characterized as a complex network composed of multiple entities, including companies, suppliers, customers, and partners. The interdependent dynamics among these entities generate synergistic effects that collectively drive the evolutionary trajectory of systemic development. This perspective transcends the limitations of traditional industry-based competitive strategy theories, thereby helping enterprises to better adapt to a rapidly changing business landscape.

Building upon the conceptual foundation, scholars have established theoretical frameworks for entrepreneurial ecosystems and innovation ecosystems [41,42,43], integrating ecosystem theory into economic research through conceptual synthesis. This extension emphasizes environmental determinants as critical points that influence corporate innovation and entrepreneurial activities, focusing on four resource dimensions: material infrastructure, market dynamism, financial capital accessibility, and human capital endowments. Participants rely on the continuous supply of these resources to maintain the stability and operation of the entire ecosystem [44,45,46]. An ecosystem typically includes various participants and interconnected environmental factors [47].

The traditional approaches to constructing a business environment index mainly focus on a single dimension, like streamlined administrative approvals. However, from an ecosystem viewpoint, the business environment can be viewed as a multi-dimensional ecosystem that integrates multiple interactions among key stakeholders, such as the government, enterprises, banks, and universities [48]. Among these entities, enterprises are the core component, which may be influenced by various factors throughout their entire process, from establishment to potential bankruptcy. These factors include government policies, institutional quality, capital accumulation, financial scale, level of openness, and access to innovation resources [49]. Based on the ecosystem theory, scholars have proposed the concept of a business environment ecosystem, which views the external conditions under which enterprises operate, including entrepreneurship, innovation, financing, and investment, as a comprehensive ecosystem. This framework highlights the interdependencies among various market entities, as well as the interconnected relationships between internal system components and external environmental factors. In essence, each city functions as a relatively independent business environment ecosystem that collectively contributes to broader provincial and regional ecosystems.

By analyzing the ecosystem theory and categorizing it based on participating entities, the business environment ecosystem can be classified into three hierarchical levels: micro, meso, and macro. At the micro level, the primary participants are enterprises and government agencies, with the core focus being the interaction between corporate business activities and governmental administrative functions. At the meso level, more entities have been introduced, such as third-sector organizations, industry associations, intermediaries, and financial institutions, thereby increasing the complexity and diversity of the system. At the macro level, the business environment ecosystem integrates broader factors, such as political, market, legal, and talent environments, offering a comprehensive view of the various influences on enterprise development. Given its relevance to theory with practical applications, this study adopts a macro-level perspective of the business environment ecosystem, which aligns with academic consensus and is widely used in policy formulation.

At the same time, cities like Beijing, Shanghai, Guangzhou, and Shenzhen provide valuable references for optimizing the business environment. In recent years, they have introduced a series of optimization measures, which are derived from real operational challenges faced by enterprises and are important for evaluating business conditions. On one hand, academic research has clearly defined the commercial ecosystem, integrating institutional frameworks and operational mechanisms that govern the financial operations of enterprises (Shanghai, 2020). Guangzhou further defined key elements of the business environment, including the market context, governmental framework, and legal system. On the other hand, these regulations propose multi-dimensional measures to optimize the business environment, covering aspects like market conditions, government services, and legal guarantees (Beijing, 2020), as well as specific indicators like market vitality, government services, financing convenience, open innovation, and legal protection (Shenzhen, 2020; Guangzhou, 2021). These first-tier cities highlight innovation, market, financing, talent, and government supports as key areas for advancing business environment initiatives, providing practical insights to construct a city-level evaluation system.

3.2. Indicators System Design

This study defines the business environment as a comprehensive ecosystem that enterprises interact with during their operations and adopts a dual-perspective approach to construct a city-level evaluation system. First, it emphasizes institutional and supply-side factors, including innovation resources, financial services, human capital, government services, and infrastructure. These elements directly influence enterprise behavior and are essential for operational efficiency. Second, enterprise production and sales activities are also influenced by external demand-side factors like market environment. Changes in market conditions influence enterprise strategies, which in turn affect the overall operations and profitability. This paper integrates institutional frameworks and stakeholder interactions to define urban business environments. Considering data availability, this paper constructs a six-component model: technological innovation, market openness, human capital, financial development, governance services, and infrastructure. It includes seventeen specific indicators that reflect the structural characteristics of regional business ecosystems, as shown in

Table 1. The indicators system for assessing the development level of the business environment.

Indicator	Symbol	Specific Indicators	Measurement Methods
Business environment	BE	BE development level	The index is constructed from six environmental components, weighted by the entropy method
Innovation environment	TE	R&D spending intensity	The R&D expenditure/GDP
		Technological innovation	Three types of patent grants
Market environment	ME	Market size	Aggregate retail sales of consumer products/GDP
		The degree of reliance on foreign trade	Aggregate imports and exports/GDP
		The actual use of foreign capital share	The total amount of foreign capital utilized/GDP
		Level of economic development	Per capita GDP
Talent environment	RE	Education expenditure intensity	Education expenditure/general government expenditure
		Level of education	Number of university students
			Personnel of education employment
Financial environment	FE	The ratio of financial deepening	Deposit–loan ratio of financial institutions/GDP
		The level of finance	Number of employees in the financial industry
Government environment	GE	Government service efficiency	Local general public finance expenditure/GDP
		Tax burden level	Local general public finance revenue/GDP
Infrastructure environment	IE	Per capita road length	Total road length/resident population
		Port handling capacity	Port cargo throughput
		Level of infrastructure investment	Infrastructure investment/fixed asset investment
		Per capita Internet level	Mobile Internet users per 10,000 people

.

(1) The technological innovation environment (TE) is mainly characterized by two dimensions: research and development (R&D) investment, and intellectual property output. Empirical data demonstrate a positive association between the ecosystem maturity and R&D funding. A well-developed innovation environment significantly increases R&D capital expenditure. This improvement is reflected in more R&D projects, higher capital expenditure, better research equipment upgrades, and the efficiency of the enterprises’ innovation process. Consequently, these factors lead to increased innovation output, which ultimately drives the growth of enterprise performance [50].

(2) The market environment is fundamentally associated with the overall economic conditions of a region, encompassing market size, import and export activities, trade liberalization policies, and the number of business organizations. These elements are critical for assessing economic opportunities and formulating regional strategic plans. They also reflect the market demand conditions and marketization level faced by enterprises. A higher quality of the market environment significantly influences production decisions and potential operating profits [51]. In high-GDP-per-capita areas, enterprises can benefit from optimized resource allocation, which supports innovation-driven growth while maintaining fair regulation and a competitive equilibrium.

(3) Talent is an important element of the innovation process and the fundamental driving force behind the rapid growth of regional economies. As one of the primary production factors, human resources play a vital role in supporting entrepreneurial activities [52]. The human resource index serves as a core indicator that is mainly used to measure the talent supply in the city’s labor market. A high-quality human resource market not only enables enterprises to access excellent talent and reduces recruitment and employment costs but also offers adequate support for the evolution and enhancement of their strategic goals.

(4) The financial environment serves as a crucial indicator of a region’s financial efficiency, the number of financial practitioners, and the financing scale. It also serves as a critical mechanism for the adjustment of industrial structures. The finance scale is a key instrument of regional financial development, reflecting its ability to provide superior financial services to enterprises. The complexity of financial market infrastructure shows an inverse correlation with corporate liquidity constraints, leading to lower financing costs. This mechanism encourages strategic investment activities, especially in venture capital distribution and R&D capitalization ratios [53]. Furthermore, the improvements in the financial environment positively affect foreign capital inflows [54]. Therefore, when creating a supportive business environment, it is essential to integrate financial system stability into assessment standards to ensure long-term enterprise sustainability.

(5) The government environment is a comprehensive concept that encompasses government quality, responsiveness, and efficiency. As a critical institutional player in market economies, the government undertakes important functions such as strategic planning, resource allocation, public service provision, and regulatory supervision. These activities directly shape urban business ecosystems and enhance market efficiency [55]. This paper indicates that better business environments not only enhance public confidence in governance systems but also foster a positive cycle between institutional trust and citizen well-being.

(6) The infrastructure environment is a foundational element for various resource flows in regional innovation activities and provides a supportive framework for industrial upgrades. The improvement of the business environment requires reasonable infrastructure development investment. A city with well-developed infrastructure fosters synergies among industries with varying levels of technical efficiency, thereby contributing to the optimization of its industrial structure.

3.3. Data Sources

This study utilizes original annual data from the statistical yearbooks of eleven cities in the Greater Bay Area, covering the period from 2008 to 2021. This dataset encompasses nine urban centers in the Pearl River Delta (PRD), with their annual reports offering multi-dimensional coverage, such as indicators of economic productivity, indices of social welfare, and demographic dynamics. The data for Hong Kong are sourced from the Hong Kong Statistical Yearbook, whereas the data for Macao are obtained from the Statistics and Census Service of Macao. Due to the unavailability of certain data in official yearbooks, supplementary information was collected from databases such as Foresight and CEID. To ensure data uniformity and comparability, the paper adopts the following aspects. In terms of exchange rate system and data calibration, Hong Kong and Macao adopt a linked exchange rate system while mainland China implements a floating exchange rate system. Considering that the currency units used by cities in the Pearl River Delta, Hong Kong, and Macao are not the same, in order to facilitate the unification and calculation of indicators, the exchange rates of each year are converted into CNY for indicator statistics and analysis so as to eliminate the errors caused by different currency units and ensure consistency with the data of mainland China.

3.4. Research Methodology

3.4.1. The Entropy Method

(1)

Theoretical foundation of the entropy method

The construction of the business environment index should be built with scientificity and objectivity principles. Especially when dealing with multi-dimensional indicators, the reasonable weight allocation plays a critical role in influencing the accuracy of the evaluation results. Shannon (1948) introduced the concept of information entropy, establishing a robust theoretical foundation for the entropy method [56]. Subsequently, numerous scholars have extensively applied this approach in the domain of multi-criteria comprehensive evaluation. Some scholars demonstrated that the entropy method addresses the issue of weight allocation in multi-criteria data, thereby minimizing biases arising from subjective weighting. This feature reduces the impact of human biases in subjective weighting methods, thereby ensuring the objectivity of weight distribution [57,58,59,60,61,62].

Firstly, as an objective weighting method, the entropy method overcomes the randomness and subjectivity problems compared with traditional subjective methods such as the analytic hierarchy process and Delphi method, thereby improving the rigor and credibility of the evaluation system. This method calculates the weights based on data discreteness to minimize human interference, which meets the requirement of objectivity and fairness in business environment evaluations. Specifically, the more dispersed an indicator’s data distribution, the greater the sample differences, the more information that it contains, and the lower its corresponding information entropy. Therefore, this indicator carries a higher weight in the overall evaluation. On the contrary, if the data tends to be more concentrated, the information entropy increases, resulting in a corresponding decrease in the weight. This automatic adjustment in weights based on data fluctuations makes the evaluation results more dynamic and adaptive. Although the principal component analysis is also an objective method, it mainly identifies key variables through dimensionality reduction, which may ignore some indicators that have less information but are still important for the overall assessment.

Secondly, the entropy method is applicable to various multi-criteria decision-making scenarios, making it highly suitable for evaluating the regional business environment. The business environment is a comprehensive concept covering multiple dimensions, like economy, law, administration, and market. Its evaluation system usually includes multiple dimensions and sub-indicators, such as market access convenience, enterprise start-up efficiency, financing channel availability, tax policy transparency, legal protection level, and government service efficiency. These indicators not only reflect the operation of the regional economy but also directly affect enterprise operating costs and development potential. Due to the multi-indicator nature and complexity of business environment assessments, the entropy method becomes a more appropriate choice.

In addition, the policy effects of the business environment change over time and across regions, requiring dynamic weight adjustments to reflect the latest development trends. The entropy method converts real-time fluctuations in business environment indicators into weight adjustment signals through the quantification mechanism of information entropy, solving the problem of traditional methods lagging behind dynamic changes. For example, when a region implements a series of policy measures to optimize the business environment in a short period, the values of relevant indicators may change significantly. In such cases, the entropy method can rapidly capture these changes and adjust the weights accordingly, ensuring that the evaluation results are closer to reality.

Finally, the entropy method effectively addresses information overlap among multiple indicator variables during the evaluation process. This is particularly important for assessing regional business environments, where factors like policy, market conditions, and rule may be interrelated. By using its unique weight allocation mechanism, the method reduces the impact of such overlaps. Therefore, it has been widely used across various fields. For example, it helps to evaluate regional development quality and optimize industrial structures. In the field of environmental science, it is used in scenarios such as ecological quality evaluation and pollution control analysis. These applications demonstrate the method’s stability and reliability in dynamic weight evaluation [58]. Especially when dealing with complex systems or multi-dimensional indicator systems, the entropy method shows strong adaptability and explanatory capability.

(2)

The calculation steps of the entropy method

This study adopts the entropy-based weighting method to quantify the dispersion of attributes within the business environment evaluation system of the GBA. This approach enhances the accuracy of reflecting each indicator’s contribution to the assessment of the commercial environment, thereby providing a strong basis for further detailed analysis. Building upon the analytical framework, this research constructs a six-dimensional structure for evaluating the business environment, which includes technological innovation, market conditions, talent availability, financial resources, government influence, and infrastructure quality. Using the entropy-based methodology, the study measures the spatio-temporal evolution of the business ecosystem across eleven cities during the period of 2008 to 2021. The mathematical derivation below is used to establish the computational framework for this method.

First, collect and sort the initial data. The data come from eleven cities within the GBA. X_tij is set as the first index of cities i in year t.

Second, standardize the indicators. Due to the extensive range of indices, it is necessary to standardize the indicators to eliminate the impact of differences in units and dimension changes. The formula utilized for standardizing the forward index is presented as follows:

$$x_{tij}^{\prime }=\left(x_{tij}-x_{min}\right)/\left(x_{max}-x_{min}\right);$$

The formula for standardizing the negative index is as follows:

$$x_{tij}^{\prime }=\left(x_{max}-x_{tij}\right)/\left(x_{max}-x_{min}\right);$$

Third, calculate the proportion of each item y within the index system:

$$y_j=\left({x^{\prime }x}_{tij}\right)/\left({\sum }_1^{t,i}{x}_{tij}^{\prime }\right);$$

Next, we calculate the information entropy e_j of the index:

$$j(k>0,k=1/L_{ni});e_j=-k/\left({\sum }_1^{t,i}{L}_{ni}{y}_j\right);$$

Calculate the information utility value d_j = 1 − e_j and the weight w_j for the j terms. The weighting scheme for evaluation indices is established through a quantitative derivation of the information entropy coefficients, incorporating standardized data preprocessing and information quantification. A higher coefficient value signifies a higher index weight, which has a more pronounced influence on the evaluation result. The weight w_j of the value of the j term is as follows:

$$w_j=d_j/\left({\sum }_1^j{d}_j\right);$$

Finally, calculate the index score $P_j:{P_j=w}_j{y}_j$, and then compute the overall evaluation score of the assessment object U. A higher score indicates a more favorable advantage of the evaluation object. The main results are illustrated in

Table 2. Summary of entropy method for calculating business environment weights.

Sub-Dimension	Specific Indicators	Information Entropy	Information Utility Value	Weight Coefficient
Innovation environment	R&D expenditure intensity	0.9591	0.0409	2.90%
	Number of patents granted	0.8338	0.1662	11.80%
Market environment	Retail sales of consumer goods	0.9829	0.0171	1.21%
	Total imports and exports	0.9158	0.0842	5.98%
	Utilization of foreign capital	0.8619	0.1381	9.80%
	Per capital GDP	0.9138	0.0862	6.12%
Talent environment	Education expenditure intensity	0.9879	0.0121	0.86%
	Number of university students	0.8178	0.1822	12.93%
	Education employment personnel	0.9120	0.0880	6.25%
Financial environment	Financial deepening ratio	0.9258	0.0742	5.27%
	Number of employees in the financial industry	0.8236	0.1764	12.52%
Government environment	Degree of government intervention	0.9593	0.0407	2.89%
	Level of tax burden	0.9870	0.0130	0.92%
Infrastructure environment	Per capital road length	0.9222	0.0778	5.52%
	Port throughput	0.9029	0.0971	6.89%
	Infrastructure investment	0.9697	0.0303	2.15%
	Number of mobile Internet users	0.9157	0.0843	5.98%

.

$$U={\sum }_1^j{w}_j{y}_j;$$

The normalized indicator weights derived through the entropy-based weighting algorithm are tabulated in Table 2.

As presented in Table 2, the third-order metric with the highest relative importance is patent authorization counts, assigned an entropy weight of 11.8%. In addition, higher education enrollment (12.93%) and the financial sector workforce indicator (12.52%) exhibit secondary significance within the hierarchical evaluation framework. The differences in weighting coefficients confirm that patent authorization exhibits greater discriminative power compared to tax burden indices and education expenditure measures, as evidenced by the entropy-based sensitivity analysis.

3.4.2. Kernel Density Estimation

Kernel density estimation (KDE) is a non-parametric statistical method used to estimate the underlying probability density function of a stochastic variable by superimposing kernel functions. In this approach, the bandwidth selection plays a crucial role in balancing estimator variance and bias. By leveraging kernel smoothing techniques, KDE eliminates the need for distributional assumptions and allows for the derivation of continuous density estimates from discrete data samples. By analyzing the horizontal position, peak height, width, distribution extensibility, and the number of peaks of multiple kernel density curves over time, KDE can effectively identify the dynamic evolution of the distributional characteristics of the research subject. The kernel density estimation uses a smooth kernel function to fit the data distribution, thereby describing the dynamic trend of variables in spatial and temporal distribution, especially the absolute difference of the random variable. Specifically, for a given dataset {x₁, x₂, ……, x_n}, the kernel density estimator for the continuous random variable X can be mathematically formulated as follows:

$$f\left(x\right)={\displaystyle \frac{1}{nh}}\sum _{i=1}^{n}K\left({\displaystyle \frac{X_i-\overline{X}}{h}}\right);K(x)={\displaystyle \frac{1}{\sqrt{2\pi }}}exp\left(-{\displaystyle \frac{x^2}{2}}\right)$$

In this context, n denotes the sample size of the dataset, $\overline{X}$ represents the arithmetic mean of observations, K specifies the kernel density function, and h corresponds to the bandwidth parameter governing the smoothing intensity in the non-parametric density estimation framework. Building upon the existing literature, this study selects the Gaussian kernel as the specific functional form of kernel density estimation, aiming to obtain a more accurate and reliable description of spatial density distribution.

3.4.3. Spatial Autocorrelation Analysis

Spatial autocorrelation analysis is an important statistical technique employed to identify clustering patterns and regional disparities. By quantifying spatial distribution characteristics, this method enables the effective identification of core–periphery structures within geographic regions, thereby offering a theoretical foundation for the interpretation of spatial phenomena. Moran initially introduced the Moran’s I statistic as a measure of spatial autocorrelation [63]. Subsequently, some scholars expanded upon this framework by proposing the concepts of global and local spatial autocorrelation analyses. Specifically, the global Moran’s I index assesses overall spatial clustering tendencies across the entire study area, whereas the local Moran’s I index facilitates the detection of localized clusters and spatial heterogeneity. The spatial autocorrelation analysis reveals core–periphery structures and spatial clustering. The research finds that this method provides a solid basis for regional coordinated development, helping policymakers to optimize resource allocation and improve regional performance [64]. In business environment research, this analytical approach effectively uncovers spatial agglomeration effects and highlights the radiating influence and developmental impact of core urban centers on surrounding regions. For instance, in the Greater Bay Area, spatial autocorrelation analysis can clearly delineate the spatial clustering characteristics of the business environment within the Pearl River Estuary region, thereby providing robust scientific support for the formulation of regional coordinated development strategies.

This study utilizes geospatial econometric techniques to examine the spatial heterogeneity of business environment development. The Global Moran’s I index and the Local Moran’s I statistic are utilized. These indices help to identify spatial clustering patterns, evaluate regional interdependence, and quantify intra-regional disparities. Through the following analytical stage, the methodological framework is structured:

$$Gocal Moran’s I={\displaystyle \frac{\sum _{i=1}^n\sum _{j=1}^n{W}_{ij}(Y_i-\overline{Y}\left)\right(Y_j-\overline{Y})}{S^2\sum _{i=1}^n\sum _{j=1}^n{W}_{ij}}};Local Moran’s I={\displaystyle \frac{\sum _{i=1}^n\sum _{j=1}^n{W}_{ij}(Y_i-\overline{Y}\left)\right(Y_j-\overline{Y})}{\frac{1}{n} \sum _{i=1}^n{(Y_i-\overline{Y})}^2}}$$

In the formulation, $\overline{Y}$ and S denote the mean value and standard deviation of the business environment, where n corresponds to the number of cities and W_ij represents the spatial adjacency matrix quantifying inter-regional relationships. The global Moran’s I coefficient demonstrates a statistically significant positive spatial autocorrelation, suggesting a clustered distribution pattern of comprehensive business environment development levels across the region.

Specifically, cities with high development levels tend to cluster near other high-level cities, while low-level cities are usually located near other low-level cities. When the index is less than 0, this trend is opposite. Although the local Moran’s I and global Moran’s I indices measure spatial autocorrelation, they exhibit methodological distinctions in analytical focus. The global index quantifies systemic spatial interdependence across the entire area, whereas the local index decomposes regional heterogeneity by identifying localized clustering patterns and spatial outliers within sub-regional units. A global Moran’s I coefficient greater than zero demonstrates a clustered spatial distribution of business environment quality across spatial neighbors, where municipalities with superior institutional frameworks are typically located adjacent to other high-performing regions, forming high–high clusters, while underperforming regions exhibit spatially contiguous effects, forming low–low clusters. Conversely, negative Moran’s I values reveal pronounced spatial dissimilarity, characterized by economic governance disparities between a given region and its neighboring areas, resulting in either high–low or low–high spatial configurations. These spatial relationships achieve statistical significance through Z-score validation and p-value thresholds, confirming a non-random spatial structuring of institutional quality. This study uses a binary representation (1 and 0) to define the adjacency relationship in spatial attributes between city i and city j.

3.4.4. The Markov Chain Analysis

The Markov chain analysis is based on the construction of transition probability matrices, which facilitate a systematic investigation of the temporal evolution mechanisms within business environment regimes. This temporally oriented Markovian process allows for a rigorous analysis of cross-sectional state transitions across multiple time horizons, thereby revealing key path-dependent processes and convergence characteristics in regional governance. The method is mathematically formalized through the conceptualization of state transition mechanisms, where the temporal evolution dynamics of stochastic systems X(t), t ∈ T, where the transition probabilities of the business environment are assumed to depend only on states city i and city j, and not on n. Based on this assumption, a homogeneous Markov chain model can be constructed, as shown in Formula:

$$\mathrm{P}\left\{X_{n+1}=j|X_0=i_0,X_1=i_1,\cdots ,X_{n-1}=i_{n-1},X_n=i_n\right\}=\mathrm{P}\left\{X_{n+1}=j|X_n=i\right\}$$

where P represents the transition probability of the business environment development level, where any elements P_ij ≥ 0 is calculated as follows: ij ∈ L. The specific calculation method for P_ij is given by P_ij = n_ij/n_i, where n_ij represents the number of transitions from type i to type j during the sample period and n_i denotes the total number of occurrences of type i.

4. The Measurement and Analysis of the Business Environment Development Level

4.1. The Development Level of the Business Environment in the Greater Bay Area

4.1.1. Comprehensive Development Level

Building upon the evaluation paradigms from prior studies, this research applies the entropy weight method to quantify the business environment development from 2008 to 2021. The evaluation system adopts a multi-dimensional framework that incorporates indicators such as innovation capacity, market dynamism, talent attraction, financial development, governance efficacy, and infrastructural investment, constructing a comprehensive business environment development index. The spatial and temporal patterns of this index are visually depicted through GIS mapping in Figure 1, revealing distinct regional development patterns within the GBA.

The temporal dynamics of the business environment development index in the GBA reveal a consistent upward trend over the fourteen-year period from 2008 to 2021. As illustrated in Figure 1, the composite index increased from an initial value of 1.636 to a final value of 2.724, resulting in a compound annual growth rate of 1.088. This trend indicates a sustained enhancement in the business environment development. From the viewpoint of development dynamics and speed, the overall development level can be categorized into two distinct stages: a stable growth period from 2010 to 2015 and a rapid growth stage from 2016 to 2021, excluding the early growth observed in 2008 and 2009.

During the stable development period, the index rose from 1.882 in 2010 to 2.075 in 2015, corresponding to a growth rate of 0.193. This shows a relatively stable upward trend. In contrast, during the rapid growth period, the development level experienced a significant improvement, with the index rising from 2.05 in 2016 to 2.724 in 2021, representing an increase of 0.674. This accelerated improvement can be attributed to the deepening of the “delegation, regulation, and service” reform in recent years, which has become a central policy priority at all governmental levels. These reforms have effectively promoted the enhancement and refinement of the business environment, thereby contributing to its overall development. Overall, the business environment in the GBA exhibited a stable and positive trend during the study period, characterized by minor fluctuations.

4.1.2. Development Levels of Six Sub-Environment Dimensions

As shown in Figure 2, from a dimensional view, the indices related to the innovation environment, talent environment, and financial environment have exhibited relatively rapid development. Specifically, the innovation index experienced a significant increase from 0.087 in 2008 to 0.482 in 2021, reflecting a total growth of 0.395, with a particularly marked acceleration following its lowest value in 2008. Similarly, the talent environment index in the Greater Bay Area exhibited steady growth, rising from 0.244 in 2008 to 0.482 in 2021, demonstrating a cumulative growth of 0.239. The financial index showed consistent growth, climbing from 0.225 in 2008 to 0.486 in 2021, with a total growth of 0.261.

In contrast, the developmental trends of the market environment, governance efficacy, and infrastructural capacity indices were comparatively slower when compared to other institutional dimensions. The market index experienced a decline, decreasing from 0.5028 in 2016 to 0.488 in 2020, which was affected by the dual impacts of the international environment and the COVID-19 pandemic. Despite this downturn, the market environment index remains the leading dimension among the six indices.

Since 2012, the government has placed a high priority on optimizing the business environment, intensifying its efforts in the areas of technological innovation, talent attraction, financial support, government efficiency, and infrastructure development. As the progress and improvements in these areas continue, there remains considerable potential for further enhancing the business environment.

4.2. Timing Evolution Analysis

4.2.1. The Overall Level of Business Environment Across Different Cities

Table 3. The business environment index of Guangdong–Hong Kong–Macao GBA from 2008 to 2021.

Year	Shen Zhen	Guang Zhou	Zhu Hai	Fo Shan	Hui Zhou	Dong Guan	Zhong Shan	Jiang Men	Zhao Qing	Hong Kong	Macao
2008	0.232	0.277	0.150	0.098	0.108	0.139	0.085	0.079	0.049	0.330	0.089
2009	0.240	0.298	0.151	0.109	0.113	0.135	0.087	0.096	0.054	0.343	0.086
2010	0.260	0.346	0.198	0.117	0.108	0.144	0.093	0.100	0.058	0.375	0.085
2011	0.277	0.336	0.165	0.119	0.102	0.151	0.093	0.092	0.058	0.394	0.090
2012	0.276	0.357	0.158	0.121	0.107	0.157	0.094	0.103	0.060	0.398	0.097
2013	0.266	0.346	0.178	0.128	0.114	0.166	0.100	0.109	0.065	0.414	0.102
2014	0.271	0.365	0.180	0.133	0.111	0.170	0.103	0.108	0.068	0.418	0.110
2015	0.276	0.382	0.182	0.130	0.110	0.178	0.104	0.109	0.070	0.415	0.118
2016	0.274	0.393	0.170	0.127	0.109	0.174	0.106	0.104	0.074	0.398	0.121
2017	0.290	0.409	0.164	0.141	0.114	0.176	0.116	0.108	0.082	0.429	0.123
2018	0.351	0.460	0.165	0.142	0.117	0.191	0.123	0.115	0.081	0.425	0.128
2019	0.392	0.496	0.179	0.166	0.136	0.207	0.120	0.112	0.082	0.435	0.133
2020	0.435	0.521	0.169	0.172	0.137	0.220	0.132	0.122	0.091	0.451	0.186
2021	0.460	0.545	0.173	0.183	0.145	0.232	0.132	0.126	0.083	0.469	0.176

presents the business environment (BE) analysis across eleven cities within the GBA from 2008 to 2021. During the sample period, the business environment index showed a steady upward trend, rising from 1.636 in 2008 to 2.724 in 2021. This trend reflects a continuous optimization of the BE level across these cities. It is worth noting that Hong Kong demonstrates a considerable advantage, with its BE index increasing from 0.33 in 2008 to 0.469 in 2021, exceeding other cities and consistently ranking among the top within the GBA. In contrast, Macao’s BE index has increased from 0.089 in 2008 to 0.176 in 2021, exceeding 0.1 in 2013. However, the overall BE level remains relatively low.

Among these cities, Guangzhou experienced the most substantial improvement, with its business environment index increasing from 0.277 in 2008 to 0.545 in 2021, with a total increase of 0.269. Shenzhen followed closely, with an increase of 0.228 units from 0.232 in 2008 to 0.46 in 2021. Compared to Hong Kong, although there is still a gap in the business environment levels between Shenzhen and Guangzhou, the pace of improvement in these two cities is steadily accelerating, leading to a gradual gap reduction. Furthermore, besides these major cities, Dongguan, Foshan, and Macao have shown similar levels of improvement in their BE indices, increasing by 0.093, 0.086, and 0.088 units, respectively. In contrast, other cities, such as Zhongshan, Zhuhai, and Huizhou, increased by fewer than 0.05 units, highlighting the urgent need for further improvements in these areas.

4.2.2. Development Levels of Six Sub-Environment Dimensions Across Different Cities

The evolution of the business ecosystem within the GBA was systematically assessed using a multi-dimensional analytical framework that incorporates six key dimensions of the business environment: technological innovation; market openness; talent attraction; financial development; governance intervention; and infrastructure investment. By employing the entropy weight method, this study constructed composite indices for each dimension. Due to space limitations,

Table 4. The business environment index scores and rankings of GBA in 2021.

City	BEDI		IE		ME		TE		FE		GE		IE
	S	R	S	R	S	R	S	R	S	R	S	R	S	R
Shenzhen	4.30	3	0.86	1	0.70	4	0.45	3	0.95	2	0.21	3	1.144	2
Guangzhou	5.53	2	0.52	2	0.44	7	1.99	1	0.66	3	0.18	7	1.749	1
Zhuhai	2.38	4	0.22	6	0.79	3	0.25	7	0.20	6	0.23	2	0.691	5
Foshan	1.89	6	0.36	4	0.35	10	0.37	4	0.22	5	0.16	9	0.435	6
Huizhou	1.63	8	0.20	7	0.50	6	0.22	9	0.09	10	0.20	4	0.427	7
Dongguan	2.44	5	0.38	3	0.60	5	0.30	5	0.19	7	0.16	10	0.822	4
Zhongshan	1.49	9	0.29	5	0.42	8	0.19	10	0.13	6	0.19	6	0.276	10
Jiangmen	1.48	10	0.17	8	0.37	9	0.22	8	0.12	9	0.19	5	0.407	8
Zhaoqing	0.98	11	0.08	9	0.20	11	0.22	6	0.04	11	0.14	11	0.226	11
Hong Kong	5.69	1	0.07	10	2.20	1	0.46	2	1.95	1	0.23	1	0.836	3
Macao	1.64	7	0.010	11	0.85	2	0.09	11	0.25	4	0.16	8	0.286	9

Note: S stands for score, R stands for ranking.

only presents the comprehensive business environment index and its rankings in 2021, and the specific index scores in the six dimensions. To better understand the strengths and weaknesses of each city in the GBA, this study conducted a detailed evaluative analysis based on these indicators. As shown in Table 4, the top three regions for each dimension are highlighted in bold font for clear emphasis, while the bottom three regions are marked in italic bold font, which illustrates that there are notable disparities in business environment parameters.

In terms of the comprehensive BEDI index scores, Hong Kong, Guangzhou, and Shenzhen claimed the leading three positions in 2021. As core driving cities of the GBA, they exhibit rapid economic development, strong technological innovation, and high levels of openness and marketization. These cities also possess notable advantages in talent attraction, financial development, and infrastructure construction. In contrast, Zhaoqing, Jiangmen, and Zhongshan ranked lowest and are classified as underdeveloped areas. Compared with more developed regions, these cities continue to face substantial opportunities for improvement in their business environment, as well as potential risks of losing investments, projects, and talent.

In terms of the innovation environment index, Shenzhen demonstrated outstanding performance in 2021, with an innovation index score of 0.86, ranking first among the cities in the GBA, followed by Guangzhou and Dongguan. This achievement highlights Shenzhen’s strong advantage in cultivating a world-class innovation environment, which has become a defining feature of the city’s development strategy. Its innovation leadership stems from a comprehensive and forward-thinking approach. Shenzhen has established a robust ecosystem that integrates advanced R&D, high-level manufacturing capabilities, and supportive policies. The city also benefits from its substantial investment in research infrastructure, which includes modern labs, innovation centers, and collaboration platforms that connect universities, industries, and government. However, Hong Kong’s innovation environment index was only 0.073, placing it at a relatively low rank. This may be attributed to land constraints that restrict the large-scale industrialization of technology-driven enterprises and innovative entities. This situation further illustrates that cultivating an innovation environment is a complex and long-term process that requires coordinated efforts among research institutions, government agencies, and enterprises.

In terms of the market environment index, Hong Kong shows a significant lead, with a score of 2.139, considerably surpassing Macao (0.85) in second place and Zhuhai (0.791) in third place. This advantage is mainly due to its strong capacity to attract foreign investment and its well-developed import–export activities. In contrast, cities such as Zhaoqing, Jiangmen, and Foshan exhibit relatively low market environment indices. This can be primarily attributed to their limited capacity to attract foreign direct investment and a generally low level of market openness. Compared with more developed cities, these municipalities lack the necessary infrastructure for international business, such as modern industrial parks, adequate logistics networks, and administrative approval procedures for foreign investors. Furthermore, the regulatory environment includes market access barriers like complex procedures and insufficient incentives for overseas enterprises.

In terms of the talent environment index, Guangzhou achieved a score of 1.992, significantly ahead of both Shenzhen (0.447) and Hong Kong (0.461). This advantage is largely attributed to Guangzhou’s abundant higher education resources, which provide a substantial talent pool for urban development. Although Shenzhen and Hong Kong offer higher average wages, diverse industries, and greater sector demands, they face talent competition pressure. Guangzhou hosts a large number of universities, research institutions, and vocational training centers that continuously produce highly skilled professionals and innovators. These educational institutions improve the overall workforce quality and support a competitive talent environment. Moreover, Guangzhou has implemented proactive policies aimed at promoting education, innovation, and talent retention, contributing to sustainable urban growth. As a result, the city’s human capital has become a key driver of its economic expansion, technological progress, and regional competitiveness. In contrast, cities like Macao, Zhongshan, and Huizhou exhibit relatively low talent environment indices due to structural challenges. These include limited access to higher-quality education, fewer innovation platforms, and weaker policy support for talent attraction and retention. Macao’s economy heavily depends on gaming and tourism, which limits job diversity and reduces demand for technical and managerial talent. Likewise, Zhongshan and Huizhou lack important universities and innovation ecosystems, making it difficult to cultivate top-tier professionals. Consequently, these cities struggle to compete with larger urban centers like Guangzhou and Shenzhen in attracting investment and fostering innovation-driven growth.

In terms of the financial environment index, Hong Kong ranks the highest, with an impressive score of 1.952, showing its exceptional strengths across multiple aspects of the financial sector. This achievement highlights the city’s considerable advantage in corporate financing efficiency, where businesses benefit from streamlined processes, accessible capital, and advanced financial infrastructure. Additionally, Hong Kong’s overall financial environment is further strengthened by a robust regulatory framework, a stable monetary system, and a globally connected financial network. These factors collectively reinforce its position as a leading global financial center, attracting international corporations and investors. In contrast, cities like Zhaoqing and Huizhou have lower financial indices, indicating relatively underdeveloped financial ecosystems compared to major urban areas. These cities face challenges such as limited access to financial resources, less specialized financial talent, and less mature institutional support. As a result, businesses in these regions may encounter difficulties in securing financing solutions that are both timely and cost-effective, thereby hindering their potential growth and competitiveness.

In terms of the governmental environment index, Hong Kong, Zhuhai, and Shenzhen rank in the top three positions, reflecting their outstanding performance in creating a favorable administrative and regulatory environment. This achievement stems from a series of proactive measures taken by local governments, which focus on improving public services quality, administrative procedures, and governmental efficiency. These efforts have resulted in more transparent, responsive, and business-friendly policies that reduce administrative barriers and provide strong support for market participants. Moreover, these cities have also made substantial progress in strengthening their connections with market entities. By fostering closer collaboration with enterprises, entrepreneurs, and investors, they have established mechanisms that align government actions with market demands. This includes simplified approvals, financial incentives, and consistent regulatory enforcement, all of which contribute to building trust and confidence among market participants. Conversely, cities such as Zhaoqing, Dongguan, and Foshan ranked relatively low, indicating that their government service and operational efficiency need to be improved.

In terms of the infrastructure environment index, Guangzhou, Shenzhen, and Hong Kong hold the top three positions, reflecting their leadership and excellence in creating world-class infrastructure systems. These cities have consistently demonstrated their commitments to advancing high-quality infrastructure that supports economic growth, urban development, and regional connectivity. Cities like Dongguan and Zhongshan also attach importance to infrastructure investment, thereby continually improving their service capabilities. Dongguan is upgrading its road networks and public transport systems to better integrate with the GBA’s transportation system. Benefiting from its strategic location between Guangzhou and Zhuhai, Zhongshan is improving its rail links and logistics infrastructure to support industrial growth. These efforts reflect a growing regional trend, where cities are increasingly recognizing the critical role of infrastructure in driving long-term economic competitiveness and urban sustainability.

Overall, there are obvious differences between the cities with the highest and lowest business environment scores in key sub-environment dimensions such as market openness, talent attraction, financial development, and urban infrastructure. These disparities reflect the varying development levels and competitiveness across different cities, highlighting the necessity for targeted strategies to address these gaps.

4.3. Temporal Evolution Analysis of the Business Environment

The spatial evolution of business environment development levels was examined using kernel density estimation, a non-parametric statistical method that reduces biases associated with distributional assumptions through adaptive bandwidth optimization. This method mainly examines three key aspects: distribution position, distribution shape, and polarization. The analytical outcomes are demonstrated in Figure 3.

Based on the positional distribution of the kernel density curve, the primary peak is located on the left side. This demonstrates that the comprehensive development level of the business environments in most cities within the GBA remains relatively underdeveloped. However, it is worth noting that, from 2008 to 2021, the kernel density distribution curve experienced a significant rightward shift. The temporal analysis of institutional maturity indices indicates a consistent upward trend in business environment development scores over the 2008–2021 period, as supported by panel data analysis. An increasing number of cities across the GBA have initiated efforts to improve their business environments, thereby contributing to overall regional progress.

Regarding the distribution pattern, the kernel density estimation curve shows a distinct right tail. From 2008 to 2021, the amplitude of the primary peak exhibited variation, concurrent with a steady expansion in distribution breadth that ultimately produced an elongated rightward tail. These changes indicate that the absolute differences in the urban business environments are continuously expanding, with certain cities demonstrating significantly more favorable conditions than others. This difference may stem from the varying performances of different cities concerning policy support, economic foundations, and innovation capabilities. Some cities have greatly improved their business environment levels with favorable conditions and proactive policies while others may have experienced slower progress due to more constraints. In the context of polarization phenomena, the kernel density curve presents a single peak distribution. This indicates that, although certain cities within the region demonstrate markedly higher levels of the business environment compared to others, the area has not developed a clear two-tiered differentiation structure. This pattern can be attributed to the implementation of regional coordinated development strategies in the GBA, along with the robust inter-city collaboration that has been fostered.

4.4. Spatial Correlation Evolution Analysis

4.4.1. Spatial Evolution Analysis

To examine the spatial evolution characteristics of the business environment among eleven cities from 2008 to 2021, this study uses ArcGIS 10.6 to generate spatial distribution maps and applies the natural breakpoint method to divide the development levels of the business environment into five categories: lower-value zone, low-value zone, middle-value zone, high-value zone, and higher-value zone. The business environment index for the years 2008, 2012, 2017, and 2021 is selected for visualization (Figure 4). The color depth reflects the index size of the comprehensive index, with darker colors indicating a higher level of the business environment. Overall, the findings reveal that there is a significant regional heterogeneity in the business environment scores, exhibiting a “core-periphery” spatial distribution pattern, with core areas progressively exceeding peripheral regions in terms of business environment level.

From a spatial distribution viewpoint, the business environment levels show obvious spatial agglomeration characteristics, with higher-value areas mainly concentrated around Guangzhou, Shenzhen, and Hong Kong near the Pearl River Estuary. Specifically, in 2008, the overall development level of the business environment was relatively low, with most cities falling into mid-value and low-value areas. Notably, these three core cities of Guangzhou, Shenzhen, and Hong Kong were categorized as higher-value area, Zhuhai as a high-value area, Foshan, Huizhou, and Dongguan as middle value areas, while the remaining four cities were primarily concentrated in the low-value and lower value areas. In 2012, the business environment levels in the three core cities remained in the higher value area, and Zhuhai maintained its position in the high value area. Other cities experienced improvements in their business environment levels, with Dongguan rising to the high value area and Jiangmen transitioning to the middle value area. In 2017, the development levels of the business environment across various cities improved significantly, leading to an increase in the number of cities falling into the middle value and high value areas. Among them, Zhongshan and Macao reached the middle-value zone, Foshan rose to the high-value zone, while Zhaoqing entered the low value area for the first time.

Overall, although the disparity of business environment index among cities remains significant, especially when compared with regional economic leaders, the gap is gradually narrowing. Notably, the most significant transformations are primarily taking place within the internal cities of the GBA. In 2021, the distribution pattern of the business environment development levels among cities was generally stable with small fluctuations. These three core cities continued to lead in business environment levels, while other cities showed overall improvements, with Macao entering the high value area for the first time. Over time, the gap between the business environment levels of core higher value cities and other cities has become increasingly wider. The large and medium-sized cities show a certain degree of stability in their business environment, while middle and high value cities have improved. For the comprehensive integration of the GBA, cities should strengthen the cooperation to promote the transfer and sharing of resources such as industry, talent, technology, and capital. This will enhance the radiation effects of core cities, reduce siphoning effects, and improve the overall business environment level.

4.4.2. Spatial Correlation Analysis

The analysis shows that the business environment development levels of eleven cities in the GBA display significant spatial clustering, with higher levels concentrated in major metropolitan areas. This spatial configuration reflects a hierarchical diffusion pattern, wherein core cities serve as innovation and resource concentration nodes, while secondary cities exhibit gradually decreasing developmental parameters along a centrifugal spatial gradient. This phenomenon reveals a certain spatial correlation in the business environment among eleven cities, and this agglomeration pattern is essential for regional economic development. Accordingly, this research adopts a methodological framework that integrates global and local Moran’s I indices to systematically examine the spatial distribution characteristics and hierarchical clustering mechanisms of business environment development across the Greater Bay Area. The analytical approach combines global spatial autocorrelation metrics, which identify overarching spatial pattern with localized spatial association measures to detect nodal clusters and dispersion gradients within the metropolitan network.

(1)

Global Spatial Correlation analysis

The global Moran’s I index functions as a key quantitative indicator for assessing the spatial autocorrelation patterns of variables within geographical systems. As shown in

Table 5. The Global Moran’s I index of the business environment of GBA cities from 2008 to 2021.

Variable	Moran Index	E(I)	Z(I)	p-Value
2008	0.227	−0.100	1.365	0.086
2009	0.187	−0.100	1.202	0.115
2010	0.074	−0.100	0.716	0.237
2011	0.183	−0.100	1.182	0.119
2012	0.171	−0.100	1.130	0.129
2013	0.154	−0.100	1.078	0.140
2014	0.133	−0.100	0.987	0.162
2015	0.122	−0.100	0.929	0.177
2016	0.113	−0.100	0.894	0.186
2017	0.133	−0.100	0.980	0.163
2018	0.152	−0.100	1.038	0.150
2019	0.157	−0.100	1.055	0.146
2020	0.169	−0.100	1.096	0.136
2021	0.181	−0.100	1.147	0.126

, this study indicates that the business environment development has consistent positive spatial autocorrelation from 2008 to 2021.The Moran’s I index has consistently remained positive throughout the entire period, with the spatial autocorrelation of the business environment fluctuating between approximately 0.07 and 0.25. This finding indicates a positive spatial correlation in the development of the business environment among the GBA cities during this period. Specifically, high-value areas are more likely to be spatially adjacent to other high-value areas, while low-value zones tend to exhibit spatial proximity to other low-value zones. In other words, high-level areas tend to cluster with other high-level areas, whereas low-level areas are more likely to be surrounded by other low-level areas. Furthermore, except for the year 2010, Moran’s I values are positive at the 20% significance level for all other years. This statistical significance further confirms the critical role of spatial factors in shaping business environment development, indicating that spatial interdependence and mutual influence should be taken into account in regional planning and policy formulation.

(2)

Spatial Correlation Analysis

The global Moran’s I index is insufficient for reflecting detailed regional differences. To thoroughly examine intra-regional spatial disparities, it is necessary to utilize local spatial autocorrelation methods for a more comprehensive analysis. The Moran scatter plot serves as an effective spatial analytical tool that can reveal spatial differentiation and clustering patterns within regions. Through the analysis of Moran scatter plots for the business environment development across the GBA in 2008, 2012, 2017, and 2021, this study reveals key spatial correlations within the urban agglomeration, as shown in Figure 5. The urban agglomeration displays two significant patterns of local spatial clustering: low-high clustering in the second quadrant and low-low clustering in the third quadrant.

Firstly, in the first and fourth quadrants, Shenzhen and Hong Kong have been in a high-high clustering zone, indicating that their business environment levels are not only high but also positively correlated with the business environment levels of surrounding cities, thereby generating a beneficial regional synergistic effect. In contrast, although Guangzhou maintains a relatively high level of business environment, it has not formed an effective high-high clustering pattern, primarily due to being surrounded by cities with comparatively lower business environment levels, such as Foshan and Huizhou.

Secondly, in the second quadrant, cities of Huizhou and Dongguan have always been in a low-high clustering pattern, indicating that their business environment levels are relatively low. Despite these cities being surrounded by core cities with higher business environment levels like Shenzhen and Guangzhou, they have not been able to effectively absorb the spillover effect. Instead, they may be affected by the siphoning effect, leading to slower improvements in their business environment levels. Zhongshan was close to a low-low clustering pattern in 2008. However, due to its advantageous geographic location and proximity to Shenzhen, it transitioned into a low-high clustering pattern in 2012. This indicates that although Zhongshan’s business environment level is relatively low, it has begun to benefit from the influence of core cities.

Thirdly, in the third quadrant, cities such as Zhuhai, Foshan, Jiangmen, Zhaoqing, and Macao are classified as low-low clustering groups. This classification indicates that their business environment levels are relatively low and show a positive spatial correlation with each other. In 2008, Macao exhibited a low-high clustering characteristic, while Zhuhai displayed a high-low clustering pattern. Despite this, both cities experienced significant advancements in their business environments, which is largely due to their proximity to urban areas with lower levels of business development. Consequently, their overall business environment levels remain relatively low, and both cities have been in a low-low clustering state from 2012 to 2021. By comparing the Moran scatter plots across different years, it can be observed that most cities have not experienced significant changes, suggesting that the spatial pattern of business environment development remains stable overall. However, this stability might also imply that some cities lack adequate momentum and vitality to further enhance their business environment.

4.4.3. Markov Analysis of the Business Environment

This study explores the internal dynamics of the business environment and its transition characteristics by utilizing a Markov transition probability matrix. A four-tier classification framework is established to quantitatively categorize the evolutionary trajectory of business environment development across the Greater Bay Area, thereby distinguishing among low, medium-low, medium-high, and high. Quartile divisions are employed to construct Markov chain analysis, based on which the transition probability matrix of the various development levels is calculated.

As presented in

Table 6. Markov Transition Probability Matrix of the Business Environment.

	Low	Medium-Low	Medium-High	High	n
Low	0.8205	0.1795	0.0000	0.0000	39
Medium-Low	0.0571	0.8286	0.1143	0.0000	35
Medium-High	0.0000	0.0294	0.9118	0.0588	34
High	0.0000	0.0000	0.0286	0.9714	35

, the probabilities of cities maintaining their original development levels after one year are 82.05% for the low level, 82.86% for the medium-low level, 91.18% for the medium-high level, and 97.14% for the high level. These data suggest that cities in the high-level category have the highest likelihood of maintaining their current business environment status. Additionally, the probability of cities at the low, medium-low, and medium-high levels transitioning to the next higher level after one year are 17.95%, 11.43%, and 5.88%, respectively. This suggests that while there is a small probability of cities at lower business environment levels moving to higher levels in the next year, there are no significant transitions across intervals. The business environment development levels show a relatively stable trend, characterized by less mobility and stronger sustainability. This trend becomes more pronounced as the business environment level improves.

4.5. Regression Analysis of Spatial Econometric Model

In recent years, spatial econometric models have been widely applied in empirical research in fields such as economics and sociology. Currently, the mainstream spatial econometric models mainly include the Spatial Lag Model (SLM), the Spatial Error Model (SEM), and the Spatial Durbin Model (SDM). Compared with SLM and SEM, SDM is more systematic and inclusive in its theoretical framework. It not only incorporates the spatial endogenous interaction effect and the spatial exogenous effect but also can effectively identify the direct impact of external variables on the dependent variable. Therefore, it has been more widely used in spatial econometric analysis. The specific model setting is as follows:

$$Z_{it}=\propto +\rho W_{ij}{Z}_{it}+\beta X_{it}+\sigma W_{ij}{X}_{it}+{\mu }_i+{\gamma }_t+{\epsilon }_{it}$$

where $Z$ is the level of business environment development, α is the constant term, X is the independent variable, W is the row-standardized spatial contiguity weight matrix, ρ is the spatial autocorrelation coefficient, β is the regression coefficient of the variable, $\sigma$ is the spatial lag regression coefficient, $\mu$ and $\gamma$ represent the city and time fixed effects, respectively, while ε denotes the random error term. The business environment is a complex system shaped by multiple interacting factors. Based on the business environment indicator system, this paper selects variables from

Table 7. The Meanings and Calculation Methods of Variables.

Variable Name	Variable Symbols	Calculation Method
The level of technological innovation	TIL	R&D investment intensity/population size
The degree of opening up	OPD	Total retail sales of consumer goods/GDP
The level of financial development	FDL	Natural logarithm of financial industry employees
The degree of government intervention	GED	Government fiscal expenditure/GDP
The level of infrastructure investment	IEI	Infrastructure investment/Fixed asset investment
The level of labor force	LF	Natural logarithm of the employed people
The educational level	EL	Education investment intensity/population size

as explanatory variables to quantitatively identify key factors influencing business environment development.

4.5.1. Descriptive Statistical Analysis

Table 8. Descriptive statistics of the sample.

Variable	Mean	SD	Min	Max	Obs
BE	0.191	0.121	0.049	0.545	154
TIL	0.195	0.176	0.005	0.951	154
OPD	0.344	0.081	0.122	0.492	154
FDL	10.20	1.070	8.810	12.53	154
GED	0.119	0.081	0.048	0.399	154
IEI	0.229	0.075	0.089	0.429	154
LF	15.16	0.636	13.80	16.34	154
ELF	0.418	0.557	0.017	2.472	154

summarizes the statistics characteristics of the data collected from 2008 to 2021, presenting a comprehensive overview of the data attributes. The business environment variable exhibited a mean value of 0.191, with a range from 0.049 to 0.545. Specifically, the average value of the technology innovation level (TIL) was 0.021, while the market opening (OPD) and financial development level (FDL) have average values of 0.344 and 10.20, respectively. In the context of the government intervention and infrastructure investment, the respective average values were recorded as 0.119 and 0.229. Additionally, the educational level (EL) exhibited an average value of 15.16. Furthermore, the labor force (LF) value was recorded as 0.418, suggesting that the majority of the resident population are of working age and is actively engaged in the labor force.

4.5.2. Correlation Statistical Analysis

To provide a more comprehensive understanding of the relationships among the variables, this study utilizes the Pearson coefficient test to perform a correlation analysis. This statistical method is widely used to measure the linear relationship between two variables and is particularly suitable for continuous data. The analysis encompasses the variables, aiming to uncover the degree and significance of their interrelationships. The statistical results are displayed in

Table 9. Correlation statistical analysis.

Variable	BE	TIL	OPD	FDL	GED	IEI	LF	EL
BE	1.000
TIL	0.692 ***	1.000
OPD	0.086	0.092	1.000
FDL	−0.039	−0.057	−0.360 ***	1.000
GED	−0.090	−0.093	−0.519 ***	0.089	1.000
IEI	0.310 ***	0.227 ***	0.423 ***	−0.042	−0.340 ***	1.000
LF	0.537 ***	0.415 ***	−0.119	−0.012	−0.017	0.179 **	1.000
EL	0.214 ***	0.109	0.032	0.039	0.009	−0.028	0.163 **	1.000

Note: ***, ** indicate significance at the 1%, 5% levels respectively.

, which presents the correlation coefficients along with their significance levels. The data reveal that the explained variable exhibits a relatively strong and significant correlation with both the explanatory variables and the control variables.

4.5.3. The Regression Results of Spatial Model

To select an appropriate spatial econometric model, this paper conducted LM, RLM, LR, and Wald tests, and ultimately chose the Spatial Durbin Model (SDM) for the research. Meanwhile, in order to deeply explore the influence of each factor on the development level of the business environment and interacts with others, this paper decomposed the spatial effect into direct and indirect effects. The direct effect reflects the role of influencing factors in the local area, while the indirect effect measures the spillover effect to neighboring regions. The regression results and the decomposition effect results are presented in

Table 10. Spatial model regression results.

	(1)	(2)	(5)	(6)	(7)
Variables	Main	WX	Direct	Indirect	Total
TIL	0.420 *** (7.78)	0.275 *** (3.21)	0.421 *** (6.67)	0.037 (0.62)	0.458 *** (6.36)
OPD	0.191 ** (2.36)	0.073 (0.52)	0.223 ** (2.42)	−0.131 (−1.23)	0.092 (0.87)
FDL	−0.005 (1.08)	−0.002 (−0.22)	−0.006 (−1.21)	0.001 (0.10)	−0.005 (−0.90)
GED	−0.027 (−0.41)	0.006 (0.006)	−0.025 (−0.32)	0.011 (0.15)	−0.015 (−0.17)
IEI	0.286 *** (3.72)	0.053 (0.48)	0.310 *** (3.81)	−0.074 (−0.72)	0.237 ** (2.09)
LF	0.054 *** (4.76)	0.083 *** (4.63)	0.041 *** (3.11)	0.046 *** (3.51)	0.087 *** (5.49)
EL	0.037 ** (2.15)	−0.049 (−1.12)	0.052 ** (2.46)	−0.061 (−1.51)	−0.010 (−0.32)
rho	0.538 *** (5.85)
${\sigma }^2$	0.003
N	154
R2	0.717
Log-L	223.636

Note: ***, ** indicate significance at the 1%, 5%, levels respectively, with z-values in parentheses.

.

As can be seen from Table 10, the spatial autocorrelation coefficient of the business environment development is statistically positive, indicating that there is a significant spatial spillover effect among regions. In other words, improvements in one region can help to enhance the business environment in surrounding regions. This spatial correlation reflects the trend of mutual influence and coordinated development among regions in terms of technological innovation, levels of opening up to the outside world, infrastructure investment, educational level, and labor resource allocation. Specifically,

(1) Technological innovation capacity is positively correlated with the business environment at the 1% significance level, indicating that regional improvements in innovation can effectively enhance the business environment. As a key driver of industrial transformation and upgrading, technological innovation improves enterprise efficiency and market responsiveness, and enhances the regional business environment through mechanisms such as technology diffusion and knowledge sharing. Further analysis shows that the direct effect of technological innovation is significantly positive, suggesting that stronger local innovation capabilities substantially improve the local business environment. Although the spatial spillover effect is not statistically significant, it is positive, implying a potential positive influence on neighboring regions. Additionally, the total effect is significantly positive, confirming the critical role of technological innovation in promoting regional integration and overall business environment improvement.

(2) The level of opening up to the outside world shows a positive relationship at the 5% significance level, indicating that enhancing the opening-up level of a region can significantly promote the optimization of its business environment. Further analysis reveals that the direct effect of openness is also positive and statistically significant at the 5% level, suggesting that a strong positive impact on the local business environment. However, both the spatial spillover effect and the total effect are not statistically significant, indicating that the driving effect of this region on the surrounding areas is relatively limited. This may be due to the fact that the resources brought about by opening up to the outside world have not yet formed a strong regional linkage mechanism, thus preventing neighboring regions from fully benefiting.

(3) The regression coefficient of the financial development level is negative and not passed the significance test, indicating that improvements in regional financial development have not yet generated the expected positive impact on the business environment. Effect decomposition further shows that both the direct effect and total effects are negative, but not statistically significant. This suggests that the financial system has not effectively played a positive role in promoting the business environment, although the potential inhibitory effect has not reached a clearly identifiable level. Theoretically, higher financial development is conducive to enhancing the efficiency of resource allocation, strengthening the financing capacity of enterprises, and reducing transaction costs, thereby having a positive impact on the business environment. However, the results show a negative trend, which may suggest existing inefficiencies in financial resource allocation and limited coverage of financial services in certain regions.

(4) The regression coefficient of the degree of government intervention is negative but not statistically significant, indicating that its overall impact on the business environment is not yet statistically detectable. Effect decomposition further shows that both the direct and indirect effects show a negative trend, but they are not statistically significant. This suggests that although government intervention in economic activities may theoretically have a certain inhibitory effect on the operation of market mechanisms and thus affect the improvement of the business environment. In other words, government intervention has not shown a clear impact on the business environment in this study. Moreover, this result may be influenced by multiple factors such as the effectiveness of policy implementation, regional heterogeneity, and industry-specific characteristics, leading to a certain degree of complexity and uncertainty.

(5) The infrastructure level shows a significantly positive relationship at the 1% significance level, indicating that increasing infrastructure investment can effectively promote the overall optimization of the business environment. Effect analysis shows that both the direct effect and total effects of infrastructure investment are statistically significant and positive, suggesting that infrastructure construction not only promotes local business conditions but also drives the development of related regions through industrial and transportation networks. However, in terms of spatial spillover effects, the estimated coefficient is negative, although it does not reach the statistical significance level, implying that the benefits of infrastructure investment remain largely localized, with limited influence on surrounding areas. These findings suggest that future infrastructure planning should not only focus on local regions but also emphasize regional coordination and resource integration to promote cross-regional connectivity and broader, more sustainable benefits.

(6) The labor force level shows a significant positive relationship at the 1% significance level, indicating that enhancing labor quality and supply capacity can effectively promote the business environment. This suggests that a region with higher-quality and more skilled labor resources will help improve the production efficiency, innovation capacity and service level of enterprises, thereby creating a better business climate. Effect analysis shows that both the direct effect and the spatial spillover effect of the labor force level are statistically significant and positive. This indicates that improving local labor force not only improves regional business environment but also has a positive impact on surrounding areas through various mechanisms such as knowledge transfer and industrial upgrading. As labor markets develop, inter-regional economic ties also strengthen, contributing to broader improvements in the business environment. This spatial spillover highlights the growing interconnectivity and collaboration in regional economic development.

(7) The education level shows a positive relationship at the 5% significance level, indicating that higher educational attainment helps improve the optimization of the business environment. The direct effect of the education is positive and statistically significant, indicating that educational investment has a clear positive impact on the local business conditions. However, the indirect effect and total effects are both negative and not statistically significant. This may reflect an uneven distribution of educational resources and the concentration of talent flows in major cities. For example, high-quality education and skilled workers in some regions may be drawn to more developed areas, leaving surrounding regions with weaker talent pools and limited innovation capacity, thus hindering regional coordination and development.

5. Research Conclusions and Policy Recommendations

5.1. Research Conclusions

With the acceleration of globalization, the development of the business ecosystem is no longer restricted to individual regions but is influenced by multiple factors. These factors include the global economic structure, international trade regulations, and the strategic deployment of multinational corporations. As a major gateway for China’s international engagement and a key driver of global economic growth, the Guangdong-Hong Kong-Macao Greater Bay Area holds significant strategic value. Its business environment development not only plays a vital role in China’s regional economic transformation but also exerts a measurable influence on the evolution of the global business landscape. This study constructs a business environment evaluation framework and an indicator system from six dimensions, including innovation environment, market environment, talent environment, financial environment, governmental environment, and infrastructure environment. By employing entropy weight methods and spatial autocorrelation, the research assesses the development levels of the business environment across eleven cities from 2008 to 2021, revealing their temporal and spatial evolution patterns. The findings indicate that: Guided by the principles of urban ecological systems theory, this study proposes a novel business environment evaluation framework through the development of a multi-dimensional analytical structure encompassing six strategic dimensions: maturity of the innovation ecosystem, efficiency of market mechanisms, dynamics of human capital, robustness of financial intermediation, quality of governance institutions, and connectivity of infrastructure. By integrating advanced computational techniques such as entropy-based weight determination and geospatial autocorrelation analysis, the research performs a longitudinal assessment of business environment performance across 11 metropolitan nodes from 2008 to 2021. Through multi-scale spatio-temporal modeling and spatial econometric diagnostics, the investigation reveals critical evolutionary trajectories characterized by the following aspects:

First, from a comprehensive perspective, the business environment showed a steady upward trend from 2008 to 2021, which was characterized by low volatility. The composite business environment index exhibited a progressive upward trend, rising from 1.636 in 2008 to 2.724 in 2021, thereby indicating statistically significant improvements over the past 13 years. Second, at the city level, the business environment levels of the eleven cities showed an upward trend with relatively stable rankings. Despite the minor fluctuations, the overall development level improved. This indicates that while pursuing high-quality economic development, each city has also given priority to optimizing and improving its business environments. Furthermore, judging from the changes across different dimensions, there are notable differences in the development levels among the 11 cities during the observation period. The business environment development levels of core cities, such as Guangzhou, Shenzhen, and Hong Kong, exhibit markedly higher levels of the business environment development compared to other cities. This is partly due to their stronger financial, innovation, and talent environments, as well as their strategic geographical locations. In contrast, cities like Zhaoqing, Jiangmen, and Huizhou have relatively low business environment development indexes due to their weak finance, innovation, and talent environments, as well as geographical restrictions.

In addition, from the perspective of spatial distribution, the development levels of business environment display obvious hierarchical characteristics. Core cities show relatively higher development levels, with significant disparities between high-value and low-value areas. In terms of geographical spatial patterns, the areas along the Pearl River Estuary maintain high-level development, while the central and western regions of the Greater Bay Area are relatively low. This phenomenon is common internationally, as many urban agglomerations face similar challenges of regional development imbalance. The Greater Bay Area promotes the rational flow of resources within the region through policy guidance and financial support and enhances the business environment in peripheral cities. Other regions can also adopt similar strategies. For instance, in the Belt and Road countries, policy guidance helps to promote the rational distribution of industrial structures and determine industrial development zones, leading to more efficient resource allocation and improvements in the business environment in underdeveloped areas.

Finally, from a spatial correlation perspective, the development levels of the business environment show a positive spatial correlation, characterized by distinct spatial agglomeration effects between high-value areas and low-value areas. The business environment of the global Moran’s I index shows an upward trend with fluctuations, although the significance level is relatively low each year. This suggests that while there is a tendency for high-value and low-value areas to cluster, the strength of this clustering varies over time. However, the coordinated development level of the urban agglomeration has gradually improved in recent years, leading to an enhanced degree of spatial correlation between different cities and geographical areas. This improvement in coordinated development can be attributed to various factors, including enhanced infrastructure, increased economic cooperation, and more effective regional policies. As a result, the spatial agglomeration effects have become more pronounced, further strengthening the positive spatial correlation of the business environment. This indicates that the development level of the business environment shows a significant spatial correlation, which has important implications for regional development strategies. For other countries, these findings suggest that spatial correlation should be included in business environment strategies. Regions with similar development levels tend to cluster, and policies can take advantage of this agglomeration. Regions with high development levels can benefit from clustering effects such as knowledge spillovers, economies of scale, and enhanced innovation capabilities. Conversely, regions with low development levels may require targeted policies to improve their business environment and reduce spatial disparities. Understanding and using spatial correlation can improve resource allocation and enhance the business environment globally. By recognizing and leveraging spatial agglomeration effects, policymakers can design more effective strategies to promote economic growth and development across different regions.

5.2. Policy Recommendations

5.2.1. Adopt Measures to Local Conditions and Strengthen Cooperation with Core Cities

Studies have shown that, in the development pattern of the Greater Bay Area, there are significant regional differences in the development level of the business environment, which presents obvious local spatial agglomeration characteristics. There exist both development differences and functional linkages between core cities and surrounding cities. Based on the previous analysis, this section proposes targeted optimization strategies for core cities, low–low agglomeration cities, and low–high agglomeration cities. These recommendations aim to solve the existing spatial imbalance of business environment development and promote coordinated regional growth.

First, for cities with a high level of business environment development, such as core cities like Guangzhou, Shenzhen, and Hong Kong, research shows that these cities have significant advantages in talent, innovation, and financial environments. To further consolidate their growth advantages and enhance their leading role, these cities should benchmark against international bay areas and actively learn from global best practices to enhance their global competitiveness and attractiveness. Specifically, Guangzhou can rely on its outstanding advantages in the talent environment by fostering talent exchanges and cooperation with surrounding cities, building a more open and inclusive talent development ecosystem. This will help to attract more high-end talent and drive continuous optimization of the business environment. Shenzhen should continue to increase R&D investment to advance its innovation infrastructure, deepen cooperation with various innovation entities, and accelerate the process of technology transfer. By optimizing the innovation ecosystem, strengthening intellectual property protection, and encouraging enterprises to increase R&D investment, it can further enhance its independent innovation capabilities and promote the high-quality development of high-tech industries. Hong Kong can leverage its unique advantages in financial and legal services to enhance its position in the international financial market. By deepening financial cooperation with the Pearl River Delta, promoting innovation in financial products and services, and expanding cross-border financial services, it can further consolidate its important status as a global financial hub.

Second, for cities with a lower level of business environment development, especially those characterized by low–low agglomeration cities, such as Zhaoqing, Jiangmen, and Huizhou, their development is mainly constrained by weak foundations in technology, market, finance, and talent environment. These cities continue to face certain limitations in cultivating a robust business environment. Therefore, it is necessary to increase policy support and systematically improve their business environment. Specific measures include enhancing innovation technology capabilities, expanding the level of openness, improving the service efficiency of government departments, and leveraging policy guidance and resource allocation, helping these cities to break through development bottlenecks and significantly improve the business environment. In addition, core cities like Guangzhou and Shenzhen should be encouraged to transfer some industrial resources to cities like Zhaoqing and Huizhou. Given its geographical connection to Shenzhen, Huizhou has already begun to absorb some industries relocated from Shenzhen and should continue to strengthen this collaborative trend. Similarly, Zhaoqing can take advantage of the industrial spillover from Guangzhou by actively developing complementary industrial systems to enhance its local industrial competitiveness.

Third, for cities with a medium-to-high level of business environment development, especially those characterized by low–high agglomeration cities, such as Dongguan, Foshan, Zhongshan, and Zhuhai, they should fully rely on their geographical advantages adjacent to core cities. These cities should strengthen collaborative cooperation and communication with surrounding core cities in order to promote higher-quality development of their local business environments. Taking Dongguan as an example, the city is located in the core area of the Greater Bay Area, between core cities like Guangzhou and Shenzhen, and thus benefits from strategic location advantages that enable it to capture regional development spillovers. It should strengthen coordinated development with these core cities, actively expand regional cooperation opportunities, and make full use of its strong manufacturing foundation, comprehensive industrial system, and mature supporting infrastructure. By effectively absorbing innovation resources and industrial transfers from Guangzhou, Shenzhen, and Hong Kong, Dongguan can significantly enhance the overall competitiveness of its business environment.

5.2.2. Strengthen Strategies for Optimizing the Business Environment to Improve Policy Operability

This study finds that the growth rates, like the market, government, and infrastructure environment, are relatively slow, which hinders the overall level of the business environment during the optimization process. To address this issue, targeted measures should be implemented. In terms of the market environment, efforts should be made to expand the openness degree at home and abroad and actively build a dual circulation market. This can be achieved by deepening market-oriented reforms, breaking down regional and industry barriers, and promoting the free flow and optimal allocation of resources. At the same time, it is necessary to enhance market regulation in order to maintain fair competition and create a more equitable, transparent, and predictable environment for all types of market entities. Cities are developing specific market-oriented reform strategies in accordance with their distinct characteristics. Core cities such as Hong Kong and Shenzhen can expand international market access to attract multinational corporations and enhance their global competitiveness. These cities also deepen strategic collaboration by leveraging their respective strengths in finance, technology, and market infrastructure to explore synergistic development models. For cities with a medium development level, it is essential to strengthen regional cooperation, break down regional barriers, and facilitate the efficient flow of resources.

Within the framework of the governmental environment, this study proposes the application of strategic policy instruments and the optimization of administrative service mechanisms, aiming to build synergistic political ecosystems. This includes simplifying the administrative approval procedures, improving the efficiency of government services, and reducing operational costs of businesses. By streamlining the administrative approval process, the government can significantly reduce the bureaucratic obstacles that enterprises often encounter and improve the quality and satisfaction of services provided to enterprises. Moreover, reducing the operational costs of businesses is a significant focus. By optimizing administrative service mechanisms and leveraging e-government platforms, businesses can benefit from lower costs associated with compliance and service access. This is further supported by the continuous efforts to simplify administrative procedures, such as the abolition of unnecessary registration fees and the integration of service platform. Additionally, the promotion of a legal system for the business environment is identified as a critical guarantee. A robust legal framework ensures that the rights of enterprises are safeguarded and that market confidence is enhanced among market entities. This includes the establishment of clear regulations and the enforcement of standards that promote fair competition and protect intellectual property rights.

With regard to the infrastructure environment of the GBA, it is imperative to enhance interconnectivity and expedite the integration of transportation systems. This can be achieved by increasing investment, improving the layout of transportation networks, and enhancing the service capacity and transportation facilities quality. Specifically, the government should allocate substantial financial resources to support the development of transportation infrastructure, particularly in underdeveloped regions of the GBA, to help to bridge the gap between different areas and promote balanced economic growth. Additionally, to improve the layout of transportation networks, the density of road and railway networks should be increased to strengthen the spatial connection between the GBA and its surrounding cities, shortening travel times and forming a complete transportation loop. In addition, the government should focus on linking up unconnected road, removing transport bottlenecks, and improving road safety facilities as well as management facilities and equipment. Cities should also formulate targeted development strategies in accordance with their own geographical location advantages and industrial development needs. For example, cities along the Pearl River Estuary like Shenzhen and Guangzhou strengthen the construction of ports and logistics infrastructure to further enhance regional interconnectivity. Cities with medium levels of development should concentrate on improving their transportation links with core cities and other surrounding cities to promote regional transportation integration. For instance, Zhuhai and Zhongshan can rely on the Hong Kong–Zhuhai–Macao Bridge to reduce logistics transportation costs and enhance overall accessibility within the Greater Bay Area. By strengthening such transportation connections, it will be more conducive to bettering the spillover effects of resources from core cities, thereby improving the sustainable development of the local business environment.

5.2.3. Learning from GBA’s Business Environment to Enhance Policy Transferability

Although the Greater Bay Area has a unique geopolitical and economic background, it has implemented a series of strategic initiatives to optimize the business environment in many aspects, such as technological innovation, financial development, talent attraction, and infrastructure construction, showing significant global reference value. These measures not only address globally relevant challenges but also can be adapted to other regions. In order to better integrate the GBA’s experience with local conditions, it is suggested that policymakers focus on the following aspects when formulating business environment strategies. The GBA has implemented coordinated development strategies under the “core-edge” spatial structure. Through policy guidance and resource allocation, it enhances the coordinated development between core cities and surrounding cities, thereby narrowing regional disparities. This approach offers valuable insights for other regions facing unbalanced regional development, such as emerging economies in Southeast Asia. These regions can learn from the GBA’s experience by strengthening infrastructure connectivity and fostering industrial collaboration to promote integrated regional development.

Moreover, the GBA’s advancements in the fields of technological innovation and green economy also provide valuable references for achieving global sustainable development goals. Shenzhen’s leadership in 5G communication, new energy vehicles, and artificial intelligence, as well as Guangzhou’s initiatives in green finance and ecological urban planning, can provide replaceable models for other regions. These practices not only enhance regional competitiveness but also support the integration of the digital economy with traditional industries and accelerate the development of a green circular economy. However, it is necessary to emphasize that, when transferring the GBA’s experience in business environment optimization to other regions, adaptive adjustments need to be made based on local conditions. Differences in the geopolitical situation, economic development levels, and industrial structure should be fully considered. For example, infrastructure construction in Southeast Asia is relatively weak, and priority should be given to strengthening its infrastructure to support economy development.

5.3. Limations and Future Research

This study has several limitations. Due to inconsistencies in statistical indicators between Hong Kong, Macao, and the mainland, as well as data availability, the research focuses on the Greater Bay Area as a representative sample, which restricts the comprehensive analysis of all factors influencing the business environment. Furthermore, during the development of the business ecosystem assessment framework, the limited number of selected indicators makes it difficult to systematically establish an evaluation indicator system. Therefore, the conclusions may not accurately reflect actual conditions. In addition, previous studies have mainly focused on city-level data over the past decade. Future research could extend the time horizon and integrate district-level analysis to further refine the indicator system of the business environment, thereby enhancing the accuracy and systematic nature of data measurement.

The research can be further strengthened in the following ways. First, it is necessary to continue advancing the assessment of the urban business environment in order to measure each city’s level in a more comprehensive and accurate way. Second, from an ecosystem perspective, it is important to investigate the specific impacts of various environmental factors and their coupling relationships on the business environment, as well as the underlying theoretical mechanisms and empirical validations. Finally, as understanding of the business environment evolves, it is necessary to re-evaluate its impact on corporate behavior. In particular, analyzing the business environment from multiple perspectives can help to clarify how different mechanisms influence corporate operation.