Introduction

In 2002, the World Bank launched the Doing Business project, introducing the concept of the business environment for the first time. The project defined the business environment as a set of external conditions that enterprises cannot control but that significantly impact their entire lifecycle. Following this, the World Bank established a business environment evaluation system to comprehensively rank the business environments of different countries. In May 2023, the World Bank introduced the B-READY framework, a new system designed to assess the legal framework and regulatory quality of business environments. According to the 2024 report based on this framework, the top ten economies with the best business environment indices include Singapore, Estonia, Georgia, Rwanda, Hungary, Portugal, Czechoslovakia, Bulgaria, Hong Kong (China), and New Zealand. Notably, the ranking features few developing countries or emerging markets, highlighting a significant gap. For emerging markets, optimizing the business environment has become a critical issue for enabling markets to play a decisive role in resource allocation and enhancing international competitiveness.

As the largest developing country, China ranked 91st in the comprehensive business environment index in 2005. However, in the World Bank’s 2020 Doing Business Report, China’s ranking surged to 31st place, making it one of the top ten economies with the most significant improvements in the business environment globally. The western region of China, due to geographical constraints, economic development disparities, and institutional limitations, has struggled with inefficiencies in its soft institutional environment. In recent years, while the overall business environment in western China still lags that of the central and eastern regions, its improvement has been more pronounced than in the central and eastern regions1.

On the one hand, this progress can be largely attributed to the Chinese government’s policies and initiatives aimed at optimizing the business environment and promoting economic development. Table 1 provides a detailed overview of these policy measures. On the other hand, China’s smart city pilot program has expanded to 290 cities, forming a new urban model in the Smart City 2.0 era. Among the 94 cities in western China, 37 have been designated as smart city pilots, accounting for nearly 40% of the region. As an innovative model of urban development2, smart cities leverage intelligent computing technologies to enhance the intelligence, coordination, and efficiency of key urban infrastructure and services. Additionally, smart city development has significantly improved urban ecological efficiency, and the policy effects have strengthened over time3. When measuring urban ecological efficiency, key indicators typically include resource inputs, economic benefits, and environmental impacts, all of which are closely linked to the business environment. This suggests that smart city development plays a crucial role in optimizing the business environment.

From the perspective of institutional economics, institutions serve as effective constraints on economic behavior. Their core function is to reduce transaction costs and institutional uncertainty, thereby improving institutional efficiency and enhancing economic performance4. SCP as a form of institutional reform, are not merely about building digital infrastructure; more importantly, they represent a systemic transformation of governance models and institutional arrangements5. By leveraging digital technologies such as big data, cloud computing, and artificial intelligence, smart city improves information structures and governance capacity across multiple dimensions. This, in turn, helps optimize market operations and regulatory order, providing institutional support for the enhancement of the business environment.

Based on the above analysis, this study empirically examines the impact of SCP on the optimization of the business environment in western China from 2006 to 2021. To understand how smart city policies affect the business environment, this study integrates insights from institutional economics, information economics, and the theory of national competitive advantage. Based on these perspectives, it constructs an analytical framework centered on the core mechanisms of marketization, internationalization, and legal institutionalization. Additionally, it investigates whether SCP and business environment optimization interact positively from the perspective of Fintech advancement. Finally, a spatial model is constructed to examine the spillover effects of smart city program on the business environment.

This research makes several key contributions: First, for emerging economies, optimizing the business environment has become a critical factor in enhancing international competitiveness and capturing global market share. China, as a leader in business environment reforms, serves as a benchmark for other emerging economies seeking to improve their local business environment. By analyzing the business environment in western China, this study offers valuable insights and lessons for other developing economies worldwide. Second, Existing research primarily qualitatively analyzes smart city development, lacking empirical validation to assess how smart cities impact economic practices6. China launched its first batch of smart city pilot projects in 2012, followed by a second and third batch in 2013 and 2014, respectively. As a result, 37 cities in western China were eventually selected as smart city pilots. Leveraging the gradual rollout of SCP, this study empirically investigates how smart cities, through digital infrastructure development, contribute to business environment optimization and economic growth. Finally, Existing research primarily examines how the business environment influences macroeconomic development7,8 and corporate operations9,10, but few studies explore the driving factors behind business environment optimization. Western China, with underdeveloped infrastructure and weak institutional frameworks, faces significant challenges in both hardware and software support for business-friendly policies. However, SCP enhances digital infrastructure, attracts skilled talent, and fosters business environment optimization. By simultaneously improving hardware environment and software environment, smart city program contributes to economic growth in western China.

Policy background and theoretical analysis

Smart city policy

The concept of smart cities was initially proposed to address the various challenges brought by urban sprawl in Western countries. Smart city development is built on digital technology as its core foundation, with the primary goal of enhancing interconnectivity and intelligence within urban operational systems, thereby fostering a new model of urban innovation and development. As the digital era unfolds, countries worldwide have launched smart city development strategies aimed at optimizing urban governance, improving resource allocation efficiency, and enhancing residents’ quality of life. In 2004, South Korea introduced the “U-Korea” development strategy, dedicated to building an eco-friendly, digitalized, and seamlessly mobile-connected smart city. In 2006, the European Union initiated the European Smart Cities Network, and in June of the same year, Singapore formulated a ten-year plan, “Intelligent Nation 2015,” to advance the nation’s smart city infrastructure. In 2009, the city of Dubuque established the first smart city in the United States, leveraging IoT technology to interconnect various public resources and integrating big data analytics to intelligently address the diverse needs of residents. China first introduced the “smart city” concept in 2009, and in 2012, the Ministry of Housing and Urban–Rural Development issued the “Interim Administrative Measures for National Smart City Pilot Programs”, announcing the first batch of 90 pilot smart cities at the county level. By 2017, 95% of provincial-level cities and 76% of county-level cities had either formulated or implemented smart city development plans. As of today, China has launched more than 500 smart city pilot programs, exceeding the total number in all other regions worldwide. Compared with traditional urban development policies, SCP take a more human-centric approach by deeply integrating artificial intelligence with sustainable economic development. As a result, SCP have become a key policy tool driving China’s green economic growth (Guo et al., 2022). Current research on smart cities primarily focuses on three core areas: (1) The impact of smart city development on the ecological environment. (2) The role of smart city development in fostering technological innovation. (3) The contribution of smart city program to improving economic and ecological efficiency.

First, the green industrial park theory suggests that reducing energy consumption through intelligent resource management and industrial symbiosis can drive the green transformation of industries. Studies have shown that smart city development significantly reduces urban carbon emissions11, improves energy efficiency6, and enhances urban green total factor productivity12, ultimately contributing to higher levels of urban green and low-carbon development2. Most research finds that smart cities, supported by information and communication technology (ICT) applications, reduce urban energy consumption through multiple channels, particularly in smart transportation and smart grid systems13. Moreover, ICT technology enables demand-side management and low-carbon scenario design through smart monitoring systems and data visualization technologies14. Second, according to Porter’s hypothesis, smart city development fosters technological innovation, particularly in energy-saving and clean production technologies, which enhance resource efficiency and reduce pollution emissions. The systemic governance framework of smart cities, especially through cross-departmental collaboration and multi-stakeholder participation, has effectively promoted urban innovation and transformation15. Caragliu16 further argues that SCP not only directly impact smart city-related technologies but also generate technological spillover effects, facilitating broader technological innovation through technology cooperation and localized applications. Finally, Fromhold-Eisebith5 emphasizes that SCP are not merely technology-driven urban solutions but also processes that drive governance transformation. While SCP may not always fully achieve their technological objectives, they can still promote sustainable urban development through institutional reforms. Similarly, Nam et al.[17](https://www.nature.com/articles/s41598-025-05630-9#ref-CR17 “Nam, T. & Pardo, T. A. Smart city as urban innovation: focusing on management, policy, and context. In Proceedings of the 5th International Conference on Theory and Practice of Electronic Governance 185–194 (ACM, 2011). https://doi.org/10.1145/2072069.2072100

.“) argue that smart cities should not be viewed solely as a technological concept but rather as a broader socio-economic development strategy. Since urban green development, technological innovation, and economic growth contribute to an improved business environment, smart city development plays a crucial role in optimizing the business environment.

theoretical analysis

Institutional theory posits that institutions function as constraints on the behavior of economic actors. Their core role is to reduce transaction costs and institutional uncertainty, thereby enhancing institutional efficiency and economic performance4. SCP as a form of institutional reform5, represent not only the development of digital infrastructure but also a systemic transformation of governance arrangements. By applying advanced digital technologies such as big data, cloud computing, and artificial intelligence, SCP improve the quality and accessibility of market information18, mitigating moral hazard and adverse selection. Through the creation of data-sharing platforms, SCP reduce information asymmetries between financial institutions and enterprises, lowering transaction costs and uncertainty. At the same time, they establish more efficient and accessible communication channels for cross-border trade, thereby reducing communication and logistics costs. The development of digital infrastructure not only enhances trade facilitation but also contributes to the expansion of international trade19. Moreover, SCP strengthen institutional performance through the construction of digital governance platforms, intelligent regulatory systems, and information-sharing mechanisms. These tools help mitigate information asymmetries between governments and market participants, improve policy implementation efficiency, and reduce enforcement costs. As a result, SCP enhance institutional predictability and transparency, support the standardization of market operations, and create a business environment that is more convenient, efficient, and equitable. Based on this, the study proposes the following research hypothesis.

Hypothesis 1

SCP promote the optimization of the urban business environment.

Smart city is not merely the accumulation of digital infrastructure; rather, it constitutes a profound process of institutional transformation. By enabling technological empowerment, institutional restructuring, and governance enhancement, smart cities promote the improvement of the urban business environment across multiple dimensions. To systematically analyze the underlying mechanisms, this study draws upon information economics, the theory of competitive advantage, and institutional economics, constructing an analytical framework centered on three key pathways: marketization, internationalization, and legal institutionalization.

First, according to information economics, the non-rivalrous nature of information implies that information sharing can enhance overall market welfare20. When information search costs are high, market participants are less able to make optimal decisions. Smart cities address this problem by developing digital platforms—such as big data systems and blockchain networks—that promote market-wide information sharing. These platforms reduce the information costs faced by financial institutions and mitigate information asymmetry. As a result, reduced asymmetry between banks and enterprises lowers firms’ borrowing costs, decreases banks’ operational risks, and improves the efficiency of financial resource allocation, thereby fostering a more effective and accessible market environment18.

Second, the theory of national competitive advantage suggests that firm competitiveness is shaped not only by internal capabilities but also by the surrounding institutional and infrastructural context21. Smart cities support the construction of cross-border digital infrastructure—such as smart ports and international data exchange platforms—which improves customs efficiency, reduces transaction costs, and creates new trade opportunities19,22. Furthermore, the emergence of digital trade regulations provides firms with a more predictable and supportive institutional framework. The interplay of digitalization, connectivity, and globalization also enhances cities’ ability to attract international capital and enterprises.

Finally, Acemoglu and Robinson23 argue that inclusive institutions are central to long-term economic development, while extractive institutions are more prone to rent-seeking and corruption. Smart city construction fosters inclusive governance through the integration of digital technologies with public administration. Platforms such as “Internet + Government Services” enhance citizen oversight and participation by enabling open access to government data and the digital supervision of administrative procedures24. These initiatives increase institutional transparency, improve regulatory efficiency, and help curb corruption25. Moreover, they strengthen the rule of law by enhancing legal accountability for both government and businesses, thereby cultivating a business environment that is fair, transparent, and conducive to sustainable growth. Based on this, the study proposes the following hypothesis. The mechanism proposed is illustrated in Fig. 1.

Fig. 1

Research framework diagram. The diagram outlines the overall structure and methodology of the study.

Hypothesis 2

SCP influence the business environment by advancing marketization, legal institutionalization, and internationalization.

The innovation diffusion theory describes the process through which a new idea or concept is disseminated and adopted. As of now, only two-fifths of cities in western China have been selected as pilot cities for smart city development, and their digital transformation fosters local technological innovation. Compared with earlier communication technologies, the information technologies used in smart cities are better equipped to transcend geographical distance26,27. First, population mobility and the convenience of intelligent transportation in these pilot cities, technological talent is likely to migrate to neighboring cities, facilitating knowledge and technology spillovers from central cities to surrounding areas. Second, this spillover effect is not only reflected in technology and infrastructure development but may also profoundly influence regional economic development models. As smart city program advance, the experience and resources accumulated by central cities in digital economy, smart manufacturing, and technology services will gradually diffuse to surrounding regions through industrial collaboration and supply chain linkages, thereby promoting regional industrial upgrading and economic integration. Thus, SCP not only directly enhance the economic vitality of central cities but also stimulate high-quality development in surrounding cities through spillover effects, ultimately optimizing the overall regional business environment.

Hypothesis 3

SCP generate positive spatial spillover effects.

Methodology

Regression model

The impact of SCP on the business environment includes time effects that evolve over time and policy treatment effects resulting from the implementation of specific policies. This study focuses primarily on examining the policy treatment effects of smart city program on the urban business environment. The difference-in-differences (DID) model is particularly effective in distinguishing between these two effects, providing a robust analysis of the treatment effects caused by the implementation of SCP while also helping to mitigate potential endogeneity issues. The specific model structure is as follows:

$$BE_{i,t} = \beta_{0} + \beta_{1} SCP_{i,t} + \beta_{2} X_{i,t} + \mu_{i} + \lambda_{t} + \varepsilon_{i,t}$$

(1)

Here, (i) represents the city, and (t) represents the year. The dependent variable, (BE_{i,t}), denotes the business environment of city (i) in year (t). (SCP_{i,t}) is a dummy variable that equals 1 if city (i) is selected as a national smart city pilot in year t or any subsequent years, and 0 otherwise. The coefficient (\beta_{1}) measures the impact of smart city development on the business environment. A positive and significant (\beta_{1}) indicates that smart city development enhances the urban business environment, while a negative and significant (\beta_{1}) suggests that it worsens the business environment. (X_{i,t}) represents a series of control variables, including urban economic density, urbanization rate, expenditure on education level, fiscal investment effort, Internet penetration rate, economic development. (\mu_{i}) and (\lambda_{t}) capture city fixed effects and time fixed effects, respectively.(\varepsilon_{i,t}) represents the random error term.

To better verify the mechanism by which SCP impacts the business environment, this study adopts the research methodology from28. This method effectively avoids endogeneity issues commonly found in traditional mechanism analyses, thereby making the test results more robust and persuasive. The study establishes Eq. (2) as follows:

$$M_{i,t} = \beta_{0} + \beta_{3} SCP_{i,t} + \beta_{4} X_{i,t} + \mu_{i} + \lambda_{t} + \varepsilon_{i,t}$$

(2)

Here, (M_{i,t}) represents the mechanism variable, while the meanings of the other variables remain consistent with those in Eq. (1). Together, Eqs. (1) and (2) constitute the mechanism testing model. The coefficients of the key explanatory variables in Eqs. (1) and (2) are denoted as (\beta_{1}) and (\beta_{3}), respectively. If both (\beta_{1}) and (\beta_{3}) are statistically significant, the proposed mechanism pathway is confirmed. Furthermore, this study seeks to determine whether Fintech plays a moderating role in the relationship between SCP and the business environment. Therefore, we establish Eq. (3) as follows:

$$BE_{i,t} = \beta_{0} + \beta_{5} SCP_{i,t} + \beta_{6} X_{i,t} + \beta_{7} Z_{i,t} {*}D_{i,t} + \mu_{i} + \lambda_{t} + \varepsilon_{i,t}$$

(3)

Here, (Z_{i,t}) represents the moderating variable, and (\beta_{7}) denotes the coefficient of the interaction term between the moderating variable and the core explanatory variable, which reflects the regulatory role of Fintech on policy effects. The meanings of the other variables remain consistent with those in Eq. (1). If (\beta_{7}) is statistically significant, it indicates that the moderating effect holds. SCP has driven technological progress and facilitated resource agglomeration. Cities with higher administrative rankings leverage their capacity to concentrate resources and technology, generating regional spillover effects that influence the emission reduction performance of neighboring areas11. Based on this premise, we construct a spatial econometric model to examine whether SCP exerts a spatial spillover effect on the business environment. Accordingly, we formulate Eqs. (4)–(6).

$$BE_{i,t} = \beta_{0} + \theta W + \beta_{8} SCP_{i,t} + \beta_{9} X_{i,t} + \mu_{i} + \lambda_{t} + \varepsilon_{i,t}$$

(4)

$$BE_{i,t} = \beta_{0} + \gamma WBE_{i,t} + \beta_{10} WSCP_{i,t} + \beta_{11} X_{i,t} + \mu_{i} + \lambda_{t} + \varepsilon_{i,t}$$

(5)

$$BE_{i,t} = \beta_{0} + \gamma WBE_{i,t} + \beta_{12} WSCP_{i,t} + \rho WSCP_{i,t} + \beta_{13} X_{i,t} + \beta_{14} WX_{i,t} + \mu_{i} + \lambda_{t} + \varepsilon_{i,t}$$

(6)

Equation (4) represents the Spatial Error Model (SEM). It assumes that unobservable factors exhibit spatial correlation. This correlation is transmitted through the error term to identify the impact of spatial variables on the measured space. In this model, (\theta) denotes the spatial error coefficient, which captures the extent to which the unobserved characteristics of neighboring areas influence the local business environment. (W) represents the spatial weight matrix, which based on the reciprocal of geographic distance. Equation (5) corresponds to the Spatial Autoregressive Model (SAR). In this model, it is assumed that a city’s business environment is influenced not only by its own SCP but also by the business environment levels of neighboring cities. (\gamma) denotes the coefficient of the spatially lagged dependent variable, reflecting how the business environment of neighboring areas affects the local business environment. (WBE_{i,t}) represents the weighted average influence of the business environment of neighboring cities, and (WSCP_{i,t}) represents the weighted average impact of the neighboring cities’ smart city policies. Equation (6) represents the Spatial Durbin Model (SDM). It is employed to capture both the direct effect of SCP on the local business environment and the spatial spillover effects on neighboring regions. Here, (\rho) denotes the coefficient of the spatial lag term of the SCP, reflecting the extent to which the promotion of SCP in neighboring areas indirectly impacts the local business environment. (WX_{i,t}) captures the influence of neighboring cities’ characteristics on the local outcomes.

The prerequisite for constructing a spatial econometric model to examine whether SCP has a spatial spillover effect on the business environment is to first verify the spatial correlation of the business environment. To achieve this, we employ Moran’s I index to analyze the spatial correlation of the business environment. The calculation formulas for Moran’s I are presented in Eqs. (7)–(9).

$$Moran{^{{\prime}}s{ }I{ } = \frac{{\mathop \sum \nolimits_{a = 1}}{n} \mathop \sum \nolimits_{b = 1}^{{n} \left( {BE_{a} - \overline{BE} } \right)\left( {BE_{b} - \overline{BE} } \right)}}{{S}{2} \mathop \sum \nolimits_{a = 1}^{{n} \mathop \sum \nolimits_{b = 1}}{n} W_{ab} }}$$

(7)

$$S^{2} = \frac{1}{n}\left( {BE_{a} - \overline{BE} } \right)$$

(8)

$$\overline{BE} = \frac{1}{n}BE_{a}$$

(9)

(S^{2}) represents the variance of BE, (\overline{BE}) denotes the mean value of BE, and (n) is the number of sample cities. If the Moran’s I index is greater than 0, it indicates a positive spatial correlation between provinces, whereas a value less than 0 suggests a negative spatial correlation. To further analyze spatial relationships, we construct a geographical distance matrix, denoted as (W), which is formulated as shown in Eq. (10).

$$d_{ij} = \sqrt {(x_{i} - x_{j} )^{{2} - \left( {y_{i} - y_{j} } \right)}{2} }$$

(10)

(d_{ij}) represents the distance between city (i) and city (j), where ((x_{i} ,y_{i} )) and ((x_{j} ,y_{j} )) are the geographical coordinates of city (i) and city (j), respectively.

Variable description

Business environment

Scholars have varied approaches to constructing and measuring the business environment index system. To more comprehensively and accurately assess the business environment, the number of indicators included in the index system has increased, and the methods of measurement have become more complex. The World Economic Forum ranks countries’ business environments based on the average of the Global Competitiveness Index, which was developed by Martin and Elsa through a stepwise weighted approach. The World Bank employs the “frontier distance method”, while other scholars have utilized methods such as weighted averaging, principal component analysis, factor analysis, and entropy methods1,24,[29](#ref-CR29 “Yin, X., Yuan, Y. & Zhou, J. Evaluation and optimization of business environment based on sustainable development perspective: Exploring the role of the reform of government functions. Sustain. Dev. sd.2655. https://doi.org/10.1002/sd.2655

(2023).“),30,31. As China increasingly emphasis on optimizing business environment, Chinese economic research institutes have launched the “China Urban Business Environment Assessment” research project. Drawing on the indicator design by Li and Yin et al.[29](https://www.nature.com/articles/s41598-025-05630-9#ref-CR29 “Yin, X., Yuan, Y. & Zhou, J. Evaluation and optimization of business environment based on sustainable development perspective: Exploring the role of the reform of government functions. Sustain. Dev. sd.2655. https://doi.org/10.1002/sd.2655

(2023).“),32 and the “China Urban Business Environment Evaluation” project team from the Economic Research Institute of Management World, this study constructs a business environment evaluation system based on six key dimensions—public services, human resources, market environment, innovation environment, financial services, and government environment—along with 18 secondary indicators. The entropy weight method assigns weights based on data variability. The smaller the entropy of an evaluation index, the greater the degree of variation, which means more information is provided, thus increasing the weight of the comprehensive evaluation. Since the business environment indicators consist of multiple components, the entropy weight method can avoid the interference of subjective human judgment, effectively enhancing the objectivity of the overall evaluation. The process involves first calculating the weights of tertiary indicators under each secondary indicator, then using these weights to compute the comprehensive scores of the secondary indicators. Next, the process is repeated to calculate the comprehensive scores of the primary indicators, and finally, the overall business environment score is obtained through the same method. The specific process and results are as follows:

Step 1 Standardizing the raw data for comparability which ensure the comparability of different indicators. To prevent the occurrence of ln0, a small constant (0.00000001) is added to the data. For positive indicators, Eq. (11) is applied. For negative indicators, Eq. (12) is used. Here, (i) represents the sample size, and (j) represents the number of indicators.

$$X_{ij}{\prime} = \frac{{x_{ij} - {\text{min}}\left{ {x_{ij} } \right}}}{{max\left{ {x_{ij} } \right} - min\left{ {x_{ij} } \right}}} + 0.00000001$$

(11)

$$X_{ij}{\prime} = \frac{{max\left{ {x_{ij} } \right} - x_{ij} }}{{max\left{ {x_{ij} } \right} - min\left{ {x_{ij} } \right}}} + 0.00000001$$

(12)

Step 2 Calculating the entropy value of each indicator.

$$r_{ij} = \frac{{X_{ij}{\prime} }}{{\mathop \sum \nolimits_{i = 1}^{304} X_{ij}{\prime} }}$$

(13)

$$e_{j} = - \frac{1}{ln304}\mathop \sum \limits_{i = 1}^{304} r_{ij} lnr_{ij}$$

(14)

Step 3 Calculating the weight of each indicator.

$$w_{j} = \frac{{1 - e_{j} }}{{\mathop \sum \nolimits_{j = 1}^{n} 1 - e_{j} }}$$

(15)

Step 4 Calculating the comprehensive score for each sample.

$$F_{i} = \mathop \sum \limits_{j = 1}^{n} w_{j} X_{ij}{\prime}$$

(16)

Based on the above calculation method, the weight distribution of the regional business environment indicator system is presented in Table 2. Table 3 lists the business environment scores and rankings of the provincial capital cities in the western region for the years 2006 and 2021. From Table 3, it can be observed that while all cities have experienced a significant improvement in their business environment scores, the rankings of some cities have declined. This suggests that despite progress, the relative competitiveness of certain cities has weakened compared to others in the China’s western region.

Smart city policy

China first proposed the concept of smart cities in 2009 and officially introduced SCP which piloted in 90 prefecture-level and county-level cities in 2012. In this study, smart city pilot is treated as a quasi-natural experiment. Given that this study focuses on western China, the evaluation of the treatment effect of SCP is primarily based on cities in this region that were selected as smart city pilots between 2012 and 2014. To examine the impact of SCP, this study constructs a dummy variable, which comprises two components: (1) a treatment dummy variable (treated), where cities designated as smart city pilots are assigned to the treatment group (treated = 1), while non-pilot cities constitute the control group (treated = 0); and (2) a time dummy variable (time), where time = 1 if the year in which a city is selected as a pilot and all subsequent years, otherwise time = 0.

Control variables

In selecting control variables, we focus on factors that have a significant impact on the business environment but are not yet included in the business environment evaluation index system. Drawing on the findings of Li and Clarke et al.32,33, we incorporate additional city-specific characteristics that may influence the business environment to minimize the potential impact of omitted variable bias. These control variables include urban economic density, urbanization rate, expenditure on education level, fiscal investment effort, Internet penetration rate, economic development. The definitions of the key variables are presented in Table 4. Additionally, a correlation analysis was conducted for the control variables. As shown in Fig. 2, darker colors indicate a stronger positive correlation, while lighter colors indicate a stronger negative correlation. Some variables exhibit negative correlations, others show positive correlations, and certain variables display no significant correlation with each other.

Fig. 2

Correlation matrix of control variables. Heatmap of the correlation coefficients of control variables in the baseline regression.

Mediating and moderating variables

This study examines the mechanism through which smart city development policies enhance the business environment by fostering a market-oriented, globally integrated, and law-governed market system. It also explores the moderating effect of Fintech on the relationship between SCP and the business environment. To measure Fintech, this study utilizes the Digital Financial Inclusion Index compiled by the Peking University Internet Finance Research Center. This index that based on data provided by Ant Financial assesses the level of Fintech at the provincial and city levels across China (excluding Hong Kong, Macau, and Taiwan). The index is employed as a proxy variable for Fintech development and has been normalized for analysis.

Data source

This study selects 82 cities in western China from 2006 to 2021 as the research sample, resulting in a total of 1312 observations. Considering data availability and reliability, we exclude data prior to 2006 due to its limited representativeness and do not include post-2021 data due to substantial missing values. Consequently, we conduct empirical analysis using data from 2006 to 2021. For variables with minor missing values, we employ interpolation methods for imputation. The business environment is measured using the entropy method, incorporating six dimensions and 18 indicators. The primary data sources include the China Urban Statistical Yearbook, the China Environmental Statistical Yearbook, and the Wind Database. The rule of law indicator is derived through web scraping and systematic compilation of policy documents published by prefecture-level governments. Specifically, we assess the presence of digital economy-related keywords in these documents, ultimately determining the indicator based on the proportion of relevant keywords to the total word count.

Results and discussions

Baseline regression results

We use Eq. (1) to examine the relationship between SCP and the business environment with the regression results presented in Table 5. In Column (1), no regional characteristic variables are controlled, and a fixed-effects model is not applied. Column (2) controls for regional characteristics while incorporating city fixed effects, but without year fixed effects. Column (3) controls for regional characteristics and simultaneously includes both city and year fixed effects. Across all specifications, the regression results consistently indicate a significant positive impact at the 10% level. This suggests that the development of smart cities effectively enhances the urban business environment, providing empirical support for Hypothesis 1.

On one hand, SCP provide assurance that market participants operate in a fair and equitable competitive environment by building digital government platforms that enable information sharing and intelligent supervision. At the same time, the extensive use of IoT-enabled big data, artificial intelligence, and blockchain technologies in smart cities not only enhances government governance capacity and efficiency34 but also establishes communication bridges between the government, enterprises, and citizens, thereby promoting government-society cooperation24. Moreover, smart cities are not limited to intelligent governance but also encompass a smart economy, smart transportation, smart living, and a smart environment. The multidimensional application of ICT improves the quality and efficiency of urban services, reduces corporate costs, and further optimizes the business environment35.

On the other hand, smart city systems drive urban technological innovation and transformation by fostering collaboration among governments, enterprises, and universities5,15, thereby enhancing urban competitiveness36. Meanwhile, intelligent resource management facilitates the rational allocation of energy, reduces energy consumption, and drives the green transformation of industries, ultimately improving the level of urban green and low-carbon development. Increased technological innovation, industrial upgrading, and optimized resource allocation enhance urban ecological efficiency, which further contributes to the optimization of the business environment[3](https: