Introduction

According to a report released by the United Nations, the global urban population accounts for 57.5% in 2023, and this figure will reach 68.4% by 2050Footnote 1. With the concentration of global population towards cities, a series of social problems known as “urban diseases” heve gradually emerged, such as traffic congestion, housing shortages, energy shortages, environmental pollution, and employment difficulties (Tsekeris and Geroliminis, 2013; Neirotti et al. 2014; Shi L et al. 2016; Huang et al. 2015; Chen, 2007). These make urban construction and urban development in a state of imbalance and disorder, resulting in a huge waste of resources, a decline in the quality of life of residents and an increase in the cost of economic development. Consequently, the sustainable development of cities is hindered to a certain extent (Ng W S et al. 2010; Johnson B, 2008; Newton and Meyer, 2012; Cui and Shi, 2012).

To cope with them, the construction of smart cities in many countries, including the United States, Sweden, Spain, Germany, France, Singapore, Japan and South Korea, has begun to take effect (Wang et al. 2021; Sakuma et al. 2021; Fernández-Güell et al. 2016; Mora et al. 2019). Following them, China, whose urban population accounts for 65.18%, has begun to explore the construction of smart cities since 2008 (Harrison et al. 2010). As a supply-side policy solution, smart cities are regarded by the Chinese government as a panacea for urban problems. A number of cities across the country have rapidly emerged to start planning and constructing smart cities (Shen et al. 2018; Li et al. 2023). In September 2010, Ningbo became the first city in China to launch a comprehensive smart city plan.

In order to avoid the problem of local governments rushing forward and blindly repeating construction, the central government began to provide policy guidance and supervision. The Ministry of Industry and Information Technology, the National Development and Reform Commission, and the Ministry of Science and Technology have begun to jointly promote the research of smart cities, develop smart city technologies and industry standards, and provide development programs for local governments. In 2012, the Ministry of Housing and Urban-Rural Development began to designate smart city pilots and provide them with 80 billion yuan in financial supportFootnote 2. Since then, the scale of investment has increased every year. Approximately, the total investment scale of China’s smart city construction in 2020reached 2.4 trillion yuanFootnote 3.

According to the official statement of China, smart city is a new concept and new mode to promote intelligent urban planning, construction, management and service by using the new generation of information technologies such as Internet of Things infrastructure, cloud computing, big data, geospatial information integration, etc.Footnote 4. Its purpose is to achieve urban industrial upgrading and livelihood protection, and the tourism industry is one of the transformation objects (Su et al. 2011; Li et al. 2017).

As one of the strategic pillar industries of the national economy and an important part of the modern service industry, the added value of China’s tourism and related industries in 2021 was 4548.4 billion yuan, a year-on-year increase of 12%Footnote 5. Tourism has the characteristics of strong relevance and high added value, and is often regarded as a green driving force for economic development (Saarinen, 2006; Chou, 2013). However, under the impact of the global COVID-19, the number of tourists in China has declined precipitously, and some tourism enterprises are facing the choice of life and death (Zhong et al. 2021; Huang et al. 2021). At the same time, personalized, fragmented and diversified tourism demands are also forming, making the tourism market more complex (Eichhorn et al. 2008 ; Tang, 2022).

Fortunately, the construction of smart city seems to provide a good opportunity for the transformation and upgrading of China’s tourism industry. Roughly speaking, it provides a technical basis for the circulation of tourism production materials, the innovation of product system and the matching of market information (Tang, 2022; Gössling, 2021; Gretzel and Stankov, 2021). By sharing the existing achievements of smart cities, tourism may become an important subsystem of information network and industrial development in smart cities (Eichhorn et al. 2008; Gretzel et al. 2015).

The construction of smart city is closely related to tourism economy. The application of smart city technology enhances the level of urban infrastructure and public services, improves the livability and attractiveness of cities, and attracts more tourists to travel (Neirotti et al. 2014). The intelligent transportation, intelligent security, intelligent tour and other functions of smart cities have improved the convenience and safety of tourist destinations, and enhanced the satisfaction and experience of tourists (Batty et al. 2012). With the continuous improvement of the digital and intelligent level of cities, the popularization and sharing of tourism information and intelligent tourism services have provided new opportunities for the innovative development of tourism enterprises (Caragliu A et al. 2011). At the same time, the construction of smart cities can break the barrier between traditional cities and rural areas, promote the interconnection of information and resources between urban and rural areas, promote the sharing and development of urban and rural tourism resources, and realize the integrated development of urban and rural tourism. However, most of these theoretical clues are based on qualitative analysis, which is difficult to be convincing and frequently questioned (Buhalis and Amaranganna, 2015), due to the lack of relevant quantitative researches.

Thus, a series of potential problems are raised. China’s smart cities has been built for more than 10 years, has it become a new driving force for tourism development? If the answer is yes, to what extent does the improvement of tourism performance come from the impact of smart city designation? What is its internal mechanism? Is there a spatial spillover effect? Existing studies have not given a clear answer to these questions. Specifically, the study of the relationship between smart city construction and tourism economy can help to further explore the interaction mechanism between urban development and tourism development, and provide a scientific basis for urban planning and tourism management. In addition, in-depth research on the impact of smart city construction on the tourism economy can provide important guidance for the sustainable development of smart cities and the innovative development of tourism, and promote the growth of urban economy and the quality and efficiency of tourism.

Therefore, this study collects panel data of 297 prefecture-level cities spanning 22 years, and constructs a time-varying difference-in-difference (DID) model by taking the smart city designated by the Chinese government as a quasi-natural experiment to empirically test the impact of smart city construction on tourism development in China. Secondly, this study summarizes some possible paths through which smart city construction affects tourism development and verifies them with data, in order to open the “black box” of the relationship between them. Thirdly, the efficient transmission of information in smart cities is profoundly changing the traditional geospatial isolation. Hence, this study empirically analyzes the spatial spillover effect of smart city construction on tourism development by combining the spatial durbin model and the DID model.

The remaining chapters of this paper are organized as follows. The second chapter reviews the influencing factors of tourism economy and the role of smart city construction in it, and then constructs the theoretical framework. The research design is presented in the third chapter, including econometric modeling, data and variables. The analysis of empirical results are shown in the fourth chapter, and the fifth chapter summarizes the conclusions and provides relevant recommendations.

Literature review and theoretical analysis

Literature review and institutional background

Clearly, the growth of tourism economic is affected by many factors. Among them, the factors that have more significant constraint effect and have been widely studied are the structure of tourism industry, the degree of economic development, the support of government policies, the accessibility of transportation, the advantages of tourism resources, the characteristics of tourism products, and the scale and potential of tourism market (Higgins, 1996; Prideaux, 2000; Khan et al. 2020; Ma et al. 2015; Turner and Witt, 2001).

In the 21st century, with the popularization of information and communication technology (ICT) and the popularity of the digital economy, high-tech has penetrated and applied in the tourism industry, gradually forming a trend of integrated development (Gössling, 2021; Gretzel and Stankov, 2021). In view of the fact that the industrial structure of any era is a comprehensive reflection of a certain level of technology, the evolution of ICT has led the tourism industry into an era of innovative digital applications (Anser et al. 2022; Adeola and Evans, 2020). The impact of informatization on tourism and the prospect of smart tourism have also become increasingly popular research topics in the field of tourism management (Tang, 2022).

ICT has changed the way of connection between tourism market players and reduced the transaction costs of tourism operations (Lee et al. 2021). The Internet improves the production mode and organizational management mode of tourism enterprises, and innovates the tourism marketing and service system, thus enhancing the performance and productivity of tourism enterprises (Ali and Frew, 2014; Bano et al. 2022). As the driving force of emerging industries, the digital economy expands the tourism demand through the deep integration of “digitization + tourism”, integrates various discrete tourism resources and elements and innovates tourism products and formats through the penetration effect and information channel effect (Ye et al. 2022; Gössling and Micheal Hall et al. 2019).

In fact, ICT-centric technology is a relatively early view in smart city theory (van den Buuse and Kolk, 2019; Li et al. 2017). In recent years, the smart city theory has begun to focus on the importance of people, technology use and information (Neirotti et al. 2014; Angelidou, 2014). The cutting-edge view gradually believes that technology and hardware facilities are only means, while human and social learning is the core of developing smart cities (Fernández-Güell et al. 2016; kraus et al. 2015). As Caragliu et al. (2011) argued, cities are “smart” only when people, social capital, transportation infrastructure, and ICT work together to promote economic growth and resource allocation. Obviously, for smart cities, both objective technology and subjective initiative are important and indispensable (Camero and Alba, 2019).

At the end of last century, two trends of thought, New Urbanism and Smart Growth, became popular. The former advocated the construction of pedestrian communities that could accommodate a variety of different housing and jobs and the improvement of the environment (Talen E, 2000; Song and Knaap, 2003). The latter demonstrated promoting the sustainable and equitable urban development through measures such as compact and intensive land use, encouraging public transport and non-motorized travel, protecting urban open space and surrounding agricultural land, and developing multi-type mixed residences (Batty et al. 2012; Ye et al. 2005). As a continuation of the two trends of thought in the 21st century, the concept of smart city has been marketed worldwide as a comprehensive urban solution by IBM, Cisco and other technology companies (van den Buuse and Kolk, 2019).

Many organizations or scholars have interpreted the meaning of smart city, including Siemens, Hitachi, IBM, Cisco and other world-renowned enterprises, as well as government and civil organizations such as Intelligent Community Forum (ICF), International City/County Management Association (ICMA), Urban and Regional Innovation Research (URENIO), and academic research institutions such as the Massachusetts Institute of Technology (MIT) and Vienna University of Technology (Vienna UT). The concept of smart city has also been expanded from different perspectives (Caragliu et al. 2011; Hollands, 2008; Lombardi et al. 2012). Although no consensus has been reached, it can be summarized that smart city is a high-level stage of urban development, which is supported by information and communication technologies such as computers, IT, and the Internet. The smart city achieves the scientific management of resources by investing in human and social capital, transportation, and communication facilities. Meanwhile, it establishes a collaborative mechanism for command decision-making, real-time response and coordinated operation for various departments of the city (Angelidou, 2014; Camero and Alba, 2019; Neirotti et al. 2014).

As one of the six pillars of smart city: the important content of smart life (Al Nuaimi et al. 2015), the academic community has not yet given a final conclusion on the prospect of tourism development under the construction of smart city. Some scholars believe that people can apply the concept of smart city to urban or rural areas, develop innovative tourism destinations based on the big data of residents and tourists, and further optimize the resource allocation of tourism destinations, which can effectively promote the sustainable development of tourism (Gretzel et al. 2015; Li et al. 2017). Other scholars hold different views. They argue that the automatic capture of data by sensors may infringe on the privacy of tourists and create a crisis of trust (Afolabi et al. 2021; Gong and Schroeder, 2022). What’s worse, the learning cost of smart application software keeps some older tourists away (Buhalis and Amaranggana, 2015), and tourism destinations with inadequate infrastructure face the problem of digital divide (Torabi et al. 2023). However, there is no existence of empirical studies on the relationship between smart city construction and tourism development. Exploring the actual impact of smart city construction on tourism can provide empirical support for optimizing urban planning and improving the quality and level of tourism development, provide scientific basis for policy formulation and practice, promote the deep integration of urban intelligence and tourism development, and promote sustainable urban development. Due to the lack of relevant research and the necessity of carrying out research, we try to make some efforts.

Since 2008, the Chinese government has begun to explore policies related to the construction of smart cities. The first pilot work of smart cities was officially launched in 2012, and the second and third batches of smart cities were designated in 2013 and 2014, respectively. In 2012, the Ministry of Housing and Urban-Rural Development (MOHURD) issued the “Interim Management Measures for the National Smart City Pilot” for the organization and management of smart city creation. This Management Measure did not impose rigid constraints on conditions such as basic technical facilities in the pilot creation standards, and did not exclude remote and backward small towns from the pilot list, thus effectively guaranteeing the fairness and justice of smart city creation. In addition, the MOHURD also issued the “National Smart City Pilot Indicator System (Trial)”, which clearly explained the construction objectives, contents and components of the smart city pilot. Among them, the first-level indicators cover smart industry and economy, smart construction and livability, smart management and service, and security system and infrastructure. During the pilot period, the MOHURD provided support to the pilot cities in terms of construction planning, organization and coordination of policy and regulatory departments, investment and financing channels, and operation management in combination with the focus of smart cities. In 2014, the National Development and Reform Commission (NDRC) issued the “National New Urbanization Plan (2014–2020)” to correctly guide the direction of smart city construction. Subsequently, the NDRC issued the “Guiding Opinions on Promoting the Healthy Development of Smart Cities”, which further clarified the specific implementation path of smart city construction in various places from the aspects of development concept, planning ideas, key tasks, and safety protection. In the report of the 19th National Congress of the Communist Party of China, smart cities rose to the national strategy for the first time, together with the network power, the science and technology power, the transportation power, the digital China, etc., and continuously innovated new ideas for construction. According to the assessment of the National Information Center, the average score of China’s smart cities rose from 58.03 in 2017 to 68.16 in 2019Footnote 6. In 2020, China’s smart city investment scale increased by 12.7 percent year-on-year, reaching $25.9 billion, thus becoming the second largest smart city investor after the United States.

At the same time, smart tourism has become an important goal for China to promote tourism industrialization and informatization. The government has long been committed to the coordinated development of smart cities and smart tourism. In 2014, The State Council’s “Several Opinions on Promoting the Reform and Development of Tourism” clearly pointed out that it is necessary to formulate tourism informatization standards, accelerate the construction of smart scenic spots and smart tourism enterprises, and improve the tourism information service system. In 2015, the Tourism Bureau issued the “Guiding Opinions on Promoting the Development of Smart Tourism”, demonstrating that smart tourism is a new driving force and new trend for the innovative development of tourism driven by tourist market demand and modern information technology. In 2017, the “13th Five-Year Plan for National Tourism Informatization” was promulgated, aiming to promote the application of information technology in the tourism industry, further meet the needs of tourists and the market for informatization, and help the vigorous development of tourism.

After more than a decade of exploration and construction, the construction of smart cities in various regions has been relatively mature and perfect. As far as we know, the existing literature on smart cities and smart tourism still remains at the level of conceptual analysis and case studies, which are difficult to avoid the problem of insufficient persuasion. Hence, this study attempts to take China’s progressive designated smart city pilot as a quasi-natural experiment, and empirically analyze the impact of smart city construction on tourism development by using cutting-edge econometric methods and 22-year Chinese urban data, in order to fill the relevant gaps.

Theoretical analysis

Direct impact

Smart cities use digital information technology to intelligently enhance urban management and services (Camero and Alba, 2019; van den Buuse and Kolk, 2019), mainly involving smart infrastructure, public services, industrial systems, resource integration, security, and human construction (Sakuma et al. 2021; Neirotti et al. 2014). Thereby, the scientific development, efficient management and a better citizen life of the city can be realized (Shen et al. 2018; Lombardi et al. 2012). The network infrastructure and digital technology used in the process of smart city construction have driven the service industries such as data services, software development and business services (Pencarelli, 2020), which not only cultivates new economic growth points, but also opens up new market opportunities for tourism (Kraus et al. 2015; Kummitha, 2019).

Tourism is an open flow between different spaces. The tourism process has great uncertainty and unpredictability. Obtaining information anytime and anywhere is an important way to improve the quality of tourism experience. Therefore, travelers’ demand for information services is gradually increasing (Li et al. 2017). Meanwhile, the growing individual traveler market has made self-help and individual travel a major way to travel (Prillwitz and Barr, 2011). In the future, the market share of individual travelers will continue to expand, so the demand for more convenient and efficient intelligent, personalized and information-based services will continue to expand (Batty et al. 2012).

In the context of smart city construction, information technologies such as the Internet of Things and cloud computing are used to comprehensively sense tourism elements, integrate and analyze tourism information, intelligently respond to tourism needs, and realize intelligent tourism experience, service and management decision-making (Gretzel et al. 2015; Mora et al. 2019). For example, traffic information and load data of tourist distribution center are transmitted to the tourist side in real time, so that people can arrange and adjust their travel plans in time (Caragliu and Del Bo, 2016; Harrison et al. 2010; Wang et al. 2021). With the maturity of smart city construction, more and more digital technology products are transforming the tourism industry. In addition to authentication, video recording, translation, and navigation, smart cities are constantly creating new tourism application scenarios, expanding existing cognitive boundaries, and providing infinite possibilities for the broad needs of the tourism market (Kraus et al. 2015). Smart finance also facilitates the consumption experience of passengers and promotes tourism consumption (Wensheng, 2020).

Smart cities can improve the quality of tourism services and cultivate new forms of services. The feedback from tourists in a series of tourism activities is helpful for the tourism suppliers to improve their services, better meet the needs of tourism consumers, and form a benign interactive relationship (Li et al. 2017; Um T and Chung, 2021). Relying on the construction of smart cities, the tourist attractions build Internet sales platforms and create an honest shopping market. For example, China Telecom has set up a one-click complaint window on the “Tour Yunnan” program to ensure that tourists are provided with information consultation and after-sales service 24 h a day.

In addition, smart cities connect the physical infrastructure, information infrastructure, social infrastructure and commercial infrastructure in the tourism process, becoming a new generation of intelligent infrastructure. Hence, among different departments and systems involved in tourism, information sharing and collaborative operations can be realized, resources are utilized more rationally, in order to make the best tourism management decisions and timely predict, and respond to emergencies and disasters (Harrison et al. 2010).

Smart cities focus on the use of information and communication technologies in all aspects to achieve efficient governance of the city, which improves the local business environment (Caragliu and Del Bo, 2016; Angelidou, 2014). The introduction of digital technologies and new management methods can standardize the operation and management of tourism industry and enhance the ability of third parties to manage and monitor the tourism industry (Fernández-Güell et al. 2016). The construction of smart cities improves the content of monitoring and information tracking, breaking the balance between crime costs and benefits, ensuring the safety of passengers and enhancing travel satisfaction (Fernández-Güell et al. 2016).

Based on this, the first hypothesis of this study is proposed:

H1: The construction of smart cities promotes the development of China’s tourism industry.

Impact mechanism

The experience of smart cities in developed countries shows that continuous attraction of capital inflows is a common phenomenon in the process of smart city construction (Caragliu and Del Bo, 2016). Enterprises in the smart cities can enjoy various preferential policies and perfect infrastructure, which drives investment to gather in the smart cities, improves the local investment level, and makes the smart cities entrepreneurial cities with tourism business opportunities (Kraus et al. 2015).

Sensors laid in smart cities collect a large amount of data that can be utilized by entrepreneurs to understand the tourism market (Al Nuaimi et al. 2015). The real-time dissemination of digital information and the online tourism communities provide inspiration for entrepreneurs to timely and effectively understand tourism consumption needs and implement entrepreneurial activities. Based on data, tourism enterprises integrate technology, material, capital and talents, use digital technology to accurately predict tourism consumption demand, reduce low-end ineffective supply, and improve the quality of tourism supply, which gives rise to various digital tourism enterprises and drives traditional tourism enterprises to accelerate digital transformation (Kummitha, 2019). The intensification of competition forces the stock capital of enterprises to be used for innovative development (Li et al. 2023). Moreover, smart finance provides effective support for the financing of tourism enterprises, increasing investment in tourism innovation and the entrepreneurial activity of tourism-related enterprises (Caragliu et al. 2011).

Tourism enterprises comprehensively use 5 G, Internet of Things, artificial intelligence and other intelligent technologies to realize intelligent management of tourism destinations, and gradually form a digital technology service system based on tourism cloud platform (Ardito et al. 2019). Tourism enterprises accurately lock target customers, scientifically analyze customers’ needs, and redefine the supply chain of tourism products, so as to optimize the process of tourism product production and marketing, improve the operational efficiency of tourism enterprises, and broaden their income channels. Smart technology is applied to all aspects of the tourism industry chain to achieve efficient connections and enhance the value of each link (Caragliu et al. 2011). Large-scale and diversified data bring them valuable information, provide technical support for tourism product development and service innovation, transform their original profit and operation mode, further stimulate industry innovation, create new cultural and tourism scenes such as immersive experience, digital scenic spots, digital hotels, and create economic value for the tourism industry (Kraus et al. 2015).

The construction of smart city is penetrating into the whole scene of tourism. Smart cities innovatively apply information technology to build smart tourism platforms, providing convenience for tourism enterprises to carry out smart tourism construction (Camero and Alba, 2019). The smart tourism platform based on cloud computing provides services to tourism enterprises and saves investment of information construction and operating costs (Hamid et al. 2021). The smart tourism platform promotes the integration of tourism resources and product innovation to maintain customer stickiness. Smart cities improve the efficiency of tourism development, making the cultural and tourism products have the characteristics of replicability, stability, low transaction costs, and fast transmission, finally accelerating the improvement of tourism benefits (Luo and Zhou, 2021). Further, smart cities can promote the density of tourism agglomeration and generate economies of scale.

In this context, a group of creative and capable people have gathered in the smart cities to join the tourism service industry (Hollands, 2008). For example, through a crowdsourcing business mode, the tour guide service platform “Gold Medal Guide” attracts part-time guides such as cultural celebrities, university experts, Internet celebrities, primary school students to create tour guide content for scenic spots (Xu et al. 2018). Also, through a flexible and highly adaptable business mode, the travel platform “Ai Bi Ying” enables landlords and experience experts to become tourism service providers, creating personalized travel experiences for tourists (Cao et al. 2022).

Based on the above analysis, the second hypothesis is proposed to be tested:

H2: Smart city construction promotes tourism development by attracting tourism investment, stimulating tourism entrepreneurial activities, improving tourism operational efficiency, and attracting relevant talents to work in the tourism industry.

Modeling, data and variables

Model specification

Benchmark regression model

China proposed the concept of smart city construction in 2009, and officially designated the national smart city pilots in three batches from 2012 to 2014. The distribution map of the three batches of designated smart cities is shown in Fig. 1. It can be found that the spatial distribution of smart cities is irregular, implying that the designation is not influenced by specific geographical factors.

Fig. 1: Distribution of designated smart cities.
figure 1

This figure shows the distribution of three batches of smart city pilots in three degrees of blue on the zoning map of Chinese cities.

Full size image

The pilot policy of smart cities is regarded as a quasi-natural experiment, and the DID model is employed to examine the relative differences of tourism economic changes between the treatment group and the control group before and after the implementation of the policy, in order to obtain the net benefits of the policy implementation. The time-varying DID model is constructed as follows.

$${\rm{TourDev}}_{{it}}={\beta }_{0}+{\beta }_{1}{\rm{SmartCity}}_{{it}}+\gamma {Z}_{{it}}+{\mu }_{i}+{\nu }_{t}+{\varepsilon }_{{it}}$$
(1)

Where the city and year are denoted by \(i\) and \(t\), respectively. The development level of urban tourism is represented by \({\rm{TourDev}}\) and the variable of smart city pilot is denoted by \({\rm{SmartCity}}\). A series of control variables affecting the development of tourism are denoted by \(Z\). The fixed effects of city and year are represented by \(\mu\) and \(\nu\), respectively, which are used to control the influence of factors that do not change with time and time trends. The policy effect of smart city construction is denoted by \({\beta }_{1}\) and needs to be focused on. The constant term is denoted by \({\beta }_{0}\), and the parameter to be estimated for the control variable is denoted by \(\gamma\). The residual term is represented by \(\varepsilon\).

Parallel trend test and dynamic effects

The premise of the consistency of the DID estimation results is that the control group and the treatment group meet the parallel trend assumption. That is, in the absence of policy intervention, the control group and the treatment group have the same tourism development trend. The Event Study is introduced to estimate the following model.

$${\rm{TourDev}}_{{it}}={\beta }_{0}+\mathop{\sum }\limits_{k=1}^{K}{F}_{k}{\rm{SmartCity}}_{i,t-k}+\mathop{\sum }\limits_{m=0}^{M}{L}_{m}{\rm{SmartCity}}_{i,t+m}+\gamma {Z}_{{it}}+{\mu }_{i}+{\nu }_{t}+{\varepsilon }_{{it}}$$
(2)

Where the precedence term for the kth (k = 1,…, K) period of the smart city pilot, and the postposition term for the mth (m = 1,…, M) period of the smart city pilot. If the estimated coefficient \({F}_{k}\) of the precedence term is not significant, and the estimated coefficient \({L}_{m}\) of the postposition term is significant or partially significant, the model passes the parallel trend test. The meaning of the remaining variables is the same as Eq. (1). Moreover, Eq. (2) can also be regarded as a dynamic DID model, and the dynamic policy effect of smart city construction over time can be captured by observing the size of \({L}_{m}\).

Data and variables

In order to achieve the research goal, the relevant data of 297 cities above prefecture level in mainland China were collected from 2000 to 2021. The data of tourism revenue and tourist arrivals are from the CEIC database, other data come from the China City Statistical Yearbook and the The Yearbook of China Tourism Statistics and the statistical bureaus of each city over the years, which are compiled by authoritative departments. The list of three batches of smart city pilots comes from the Ministry of Housing and Urban-Rural Development of the P.R.C. In order to eliminate the impact of price, all the data related to price in this study are deflated with the base period of 2000. After a series of data cleaning work, a panel data with a sample size of 6534 across 22 years was generated. It should be noted that some observations in the sample are missing, so the actual sample size entering the model may be reduced.

The key variables used in this study are as follows:

  1. (1)

    Tourism development. The output indicators of tourism development are mainly expressed as the output indicator of tourism revenue and the scale indicator of tourist arrivals (Tang, 2022; Chou, 2013; Bano et al. 2022). The tourism revenue and tourist arrivals are taken to measure the level of tourism development. Among them, tourism revenue includes domestic tourism revenue (RMB) and foreign exchange tourism revenue (US dollar). The average exchange rate over the years is used to convert foreign exchange tourism revenue into RMB and sum it with domestic tourism revenue. The number of tourists is the sum of domestic tourists and inbound tourists. In order to avoid the influence of individual extreme values and the problem of model heteroscedasticity, these two data are logarithmized during regression.

  2. (2)

    Smart city construction. From 2012 to 2014, the Ministry of Housing and Urban-Rural Development successively identified three batches of smart city pilots, including 38 cities above the prefecture level in 2012, 41 cities above the prefecture level in 2013, and 28 cities above the prefecture level in 2014. A value of 1 is assigned to the samples in the year when the smart city was established and the years afterward, otherwise the value is assigned to 0.

  3. (3)

    Control variables. In order to avoid the bias of model estimation caused by the omission of key variables, some factors that may affect the development of regional tourism are added to the model, including human capital, per capita GDP, government intervention, openness to the outside world, industrial structure, transportation and public health events (Adeola and Evans, 2020; Huang et al. 2021; Khan et al. 2020; Lee et al. 2021). Human capital is represented by the proportion of the number of students enrolled in regular higher education institutions to the total population of the region. Per capita GDP is logarithmized. Government intervention is expressed as the proportion of local fiscal expenditure to GDP. Openness to the outside world is represented by the proportion of the amount of foreign capital actually used in the current year to the GDP. The industrial structure is denoted by the proportion of the added value of the tertiary industry to the GDP. Transportation is expressed as the mileage of highway per unit area. Public health events mainly refer to the SARS epidemic in 2003 and the COVID-19 in 2020. Combined with the actual start and end time of the epidemic, a value of 1 is assigned to the samples in 2003, 2020 and 2021, otherwise the value is assigned to 0.

For better comparison, the statistics of smart cities and non-smart cities are described separately, and the statistical characteristics of each variable are shown in Table 1. From the perspective of tourism development indicators, the average tourism revenue of smart cities (23.845) is much higher than that of non-smart cities (22.39) and has smaller internal fluctuations. The average number of tourists in smart cities (17.281) is also much higher than that in non-smart cities (16.028). In terms of control variables, except for the indicator of openness to the outside world, the mean values of other control variables of smart cities are slightly higher than those of non-smart cities. Of course, simple descriptive statistics are not sufficient to show that tourism development is caused by the pilot policy of smart cities. Thus, it’s necessary to carry out rigorous econometric analysis.

Table 1 Descriptive statistics of main variables.
Full size table

Results

Benchmark regression

The tourism revenue and tourist arrivals are taken as dependent variables to estimate Eq. (1). The estimation results of the tourism revenue equation are shown in columns 1–2 of Table 2. The city fixed effect and year fixed effect are controlled in column 1, and control variables are added in column 2. Similarly, the estimated results of the tourist arrival equation are shown in columns 3–4. It can be seen that the estimated coefficients of smart city in column 1 and column 3 are significantly positive at the level of 1%, indicating that the construction of smart city has significantly increased local tourism revenue and tourist numbers. After adding a series of control variables, the smart city variable still significantly increases the local tourism revenure at the level of 5%, with an increase of approximately 24.3%. At the same time, the construction of smart cities also improves the number of local tourists by about 17.2% at the level of 1%. It can be found that the construction of smart city is empowering the tourism industry. When smart technology penetrates into all aspects of the tourism industry chain, tourism enterprises realize innovation in the competition and provide more convenient and humanized services for tourists (Kraus et al. 2015). Tourists can also optimize their travel plans and enhance their travel experience with the help of smart city sensors (Bano et al. 2022). Further, it can be calculated that the tourism revenue per person has increased by 6%, demonstrating that the construction of smart city is helpful for tourists to travel deeply. This study reveals the positive impact of smart city construction on the tourism industry, although the pilot policy of smart cities is not promulgated specifically for the development of tourism. The results present that the construction of smart cities has increased tourism revenue by 24.3% and tourist arrivals by 17.2%. This means that the construction of smart cities has brought significant economic growth and development opportunities to the tourism industry, and promoted the prosperity and development of the tourism industry. This finding has important economic implications that smart city construction can not only improve the quality of urban life, but also inject new impetus into the urban economy, and promote the upgrading of the urban industrial structure and sustainable economic growth. In short, tourism has experienced rapid development in the process of smart city construction. The hypothesis of this study is preliminarily verified.

Table 2 Benchmark regression results.
Full size table

In terms of control variables, per capita GDP represents the level of urban economic development, and the more developed the economy, the more vigorous the urban tourism industry. Government financial intervention helps to regulate the operation of the macro economy, thus significantly promoting the development of tourism industry at the level of 5% (Tang, 2022). Cities with a high degree of openness to the outside world are more likely to attract foreign tourists. Cities with a larger proportion of the tertiary industry have richer tourism supporting resources, which can provide a better industrial environment and market source for the development of tourism economy (Ma et al. 2015). Transportation facilities provide tourists with convenient travel experiences (Wang et al. 2021). The epidemic greatly restricts the development of tourism. Overall, the regression coefficients of the control variables are consistent with expectations.

The premise of achieving unbiased estimation of DID is that the control group and the treatment group meet the parallel trend assumption. That is, without the intervention of the smart city pilot policy, the tourism development trend of the control group and the treatment group are the same. The test results based on the Event Study are shown in Fig. 2, where 95% confidence intervals are used for the upper and lower bounds of the estimated coefficients.

Fig. 2: Parallel trend test and dynamic effect.
figure 2

This figure depicts the dynamic impact of the pre- and post-terms of the pilot policy on tourism development.

Full size image

It is seen from Fig. 2 that before the implementation of the policy, the estimated coefficients of each year did not reach a significant level. It indicates that before the release of the smart city pilot, there was no significant difference in the level of tourism development between the pilot cities and non-pilot cities. Thereby, the parallel trend assumption is met. After the implementation of the policy, the estimated coefficients of each year are significantly positive at least at the level of 95%, demonstrating that the construction of smart cities has a positive effect on the development of tourism industry, and the policy effect will become increasingly significant over time.

For the positive dynamic policy effects shown in Fig. 2, this study attributes this phenomenon to the Solow Paradox. In the early stage of the smart city construction, tourists’ insufficient perception and trust in the functionality of smart technologies limited the pulling effect of smart city pilot policies on tourism consumption (Solow, 1956). With the gradual penetration of smart infrastructure into all aspects of residents’ production and life, tourists are increasingly involved in the collection, sharing and interaction of online tourism-related information. Tourism network publicity and word-of-mouth communication grows geometrically, which contributed to the greater conversion of the potential tourism consumption demand into tourism consumption practice and the sustained growth of the tourism economy (Li, 2014). The innovative application of smart technology in tourism products continuously breeds products that meet the diversified needs of tourists and have high added value. The more mature the construction of smart cities, the lower the marginal cost of integrated development with tourism. In addition, the demonstration effect of tourism innovation achievements attracts more market entities to participate in tourism economic activities, further expanding the scale of tourism market (Kumar et al. 2019).

Robustness test

Handling selection bias: PSM-DID

Although the pilot policy of smart cities is regarded as an exogenous shock, the central government may have a tendency to choose smart city pilots, such as giving priority to cities with good economic development and high digital level. Therefore, the selection of smart city pilots may not be random, resulting in the estimated parameters in the benchmark regression not reflecting the real effect of smart city construction on tourism development. In order to make the smart city and non-smart city comparable, the propensity score matching (PSM) combined with DID is introduced to test the benchmark regression.

Specifically, covariates and Logit models are employed to estimate the probability of each city becoming a smart city pilot, that is, the propensity score value. Based on the propensity score value, we use the Nearest-neighbor Matching within caliper and Kernel Matching to match each treatment group sample to a specific control group sample, so that there are similar endowment characteristics between smart cities and non-smart cities, basically achieving the effect of a random experiment. The estimated results using PSM-DID are shown in columns 1–4 of Table 3.

Table 3 Robustness test.
Full size table

It can be found that the estimation coefficient of Smart City is positively significant at the level of 1%, further confirming the significant tourism economic growth effect of smart city construction. From the perspective of policy effect intensity, the estimated coefficients based on PSM-DID are 0.286 and 0.157, while the benchmark regression coefficients are 0.243 and 0.172, which are not much different, demonstrating that the selective bias will not seriously affect the core conclusions of this study.

Excluding municipality samples

The data used in this study involves 293 prefecture-level cities and 4 municipalities directly under the central government. Due to the unique political and economic development levels of Beijing, Tianjin, Shanghai and Chongqing compared with other prefecture-level cities, the four municipalities are excluded and re-regressed to verify the robustness of the results. The results are presented in columns 5–6 of Table 3. It can be observed that compared with the benchmark regression results, the construction of smart cities has a greater effect on the development of tourism after excluding the samples of municipalities directly under the central government.

Placebo test

In response to the potential threat of missing variables or unobservable factors to the estimated results, this paper draws on Cantoni et al. (2017) to construct false policy shocks for the placebo test. We randomly selected 38 city samples from 2012, 41 city samples from 2013 and 28 city samples from 2014 as the treatment group, and generated dummy policy variables to be included in the benchmark regression model for regression. In theory, if the benchmark regression is not affected by missing variables or other random factors, the estimated coefficient of the dummy policy variable should have no significant difference from 0. That is, the randomly generated smart city will not have a significant impact on the development of the tourism industry. The above process is repeated for 500 times, and the scatter plots of the estimated coefficients and kernel density curves of regression using dummy policy variables for 500 times are plotted in Fig. 3.

Fig. 3: Placebo test.
figure 3

This figure shows the coefficients, P-values, and kernel density curves obtained by 500 regressions of tourism revenue and tourist arrivals to the dummy policy variables, respectively.

Full size image

It is not difficult to find from Fig. 3 that whether the dependent variable is tourism revenue or tourist arrival, the estimated coefficients of the dummy smart city variables are approximately normally distributed. Most of the coefficients are concentrated around 0, far from the real estimated values. Meanwhile, the estimated coefficients of dummy smart cities mostly do not have statistical significance, which can be seen from their p value greater than 0.1. Consequently, there is a significant difference between the placebo test results and the real model results. It can be concluded that the tourism development effect of smart cities in the benchmark regression is not caused by other unobservable factors.

Mechanism analysis

According to the above theoretical framework, the construction of smart cities can attract tourism investment, stimulate tourism entrepreneurship activities, improve the efficiency of tourism operation, absorb relevant talents to work in tourism, and other benefits. Combined with data availability, four main mechanisms are selected for analysis. We collected the operation data of tourism enterprises in 23 cities from 2000 to 2016 from the Yearbook of China Tourism Statistics over the years. This is all the data of tourism enterprises at the city level that we can find so far, because the statistical departments of most cities do not investigate or publish them. The data with a sample size of 391 is matched to the pilot list and control variables, and the DID estimation is conducted. Similar to the benchmark regression, the actual sample size entering the model is reduced due to the absence of some variables. The results of mechanism analysis are presented in Table 4.

Table 4 Mechanism analysis results.
Full size table

In the estimation, the investment in fixed assets of tourism enterprises is taken to measure tourism investment, the number of tourism enterprises is chosen to describe the entrepreneurial and competitive state of urban tourism. Moreover, the overall labor productivity of tourism enterprises is used to denote the economic efficiency of tourism, and the number of employees in tourism enterprises is taken to characterize the flow of talents to tourism. As can be seen in Table 4, the estimated coefficients of Smart City in all four columns are significantly positive at least at the level of 5%. The construction of smart cities allows tourism enterprises to see the development prospects and increase investment in fixed assets, attracting more market players to enter the tourism industry. They launch more excellent tourism products in the market competition to help the development of the tourism industry (Li et al. 2023). Further, tourism enterprises use smart technology to narrow the distance between themselves and tourists and reduce the degree of information asymmetry, resulting in the improvement of enterprises’ labor productivity (Li et al. 2017). In the context of the smart city construction, more and more human resources are flowing to the tourism industry, engaged in online marketing, offline reception, technology research and development (Xu et al. 2018). Based on this, the tourism economy of smart cities has grown rapidly.

Spatial spillover effect

The construction of smart cities is a process of integrated development of digital infrastructure and network economy, with strong externalities. It can break through the limitations of time and space, break the spatial barriers of existing market segmentation and factor flow, and then make the development of inter-regional tourism interdependent and interpenetrating (Li et al. 2020). Thereby, while promoting the development of tourism in pilot cities, the construction of smart cities may also have a positive impact on the tourism economy in other regions, namely the spatial spillover effect.

The traditional DID model implies the assumption of the stability of individual treatment effect, which requires individuals to be independent of each other when influenced by policies. To this end, this paper further draws on Chagas et al. (2016) to construct a spatial dubin DID model (SDM-DID) for estimation. Compared with the traditional DID model, the SDM-DID model relaxes the assumption of mutual independence among spatial individuals. The model not only considers the direct effect of the policy on the local area, but also takes into account the spatial spillover effect of the policy on other regions. So it can evaluate the policy impact effect of spatial dependence between variables, and comprehensively estimate the tourism economic spillover effect of smart city construction. The SDM-DID model of this study is set as follows:

$$\begin{array}{l}{\rm{TourDev}}_{{it}}={\beta }_{0}+\rho {W}_{{ij}}{\rm{TourDev}}_{{it}}+{\beta }_{1}{\rm{SmartCity}}_{{it}}\\\qquad\qquad\quad+{\beta }_{2}{W}_{{ij}}{\rm{SmartCity}}_{{it}}+\gamma {W}_{{ij}}{Z}_{{it}}+{\mu }_{i}+{\nu }_{t}+{\varepsilon }_{{it}}\end{array}$$
(3)

Where the coefficient of the spatial lag term of the dependent variable is denoted by \(\rho\). The spatial weight matrix is represented by \({W}_{{ij}}\). Three types of spatial weight matrices are employed to estimate. (1) The geographic distance matrix \(W1=\left\{\begin{array}{l}\frac{1}{{d}_{{ij}}^{2}}\quad{i}\,\ne\, j\\\,\, 0\quad\,{i}\,=j\end{array}\right.\), the distance between city \(i\) and city \(j\) is denoted by \({d}_{{ij}}\). (2) The economic distance matrix \(W2=\left\{\begin{array}{ll}\frac{1}{|{\rm{PGDP}}_{i}-{\rm{PGDP}}_{j}|}\quad{i}\,\ne \,j\\\,0\quad\qquad\quad\,{i}=j\end{array}\right.\), the per capita GDP of city \(i\) and city \(j\) are represented by \({\rm{PGDP}}_{i}\) and \({\rm{PGDP}}_{j}\), respectively. (3) The geo-economic nested matrix \(W3=\left\{\begin{array}{ll}W1\times W2\quad{i}\,\ne\, j\\ 0\qquad\quad{i}=j\end{array}\right.\). The meanings of other variables are the same as Eq. (1). It should be pointed out that the use of spatial econometric models requires no missing values in panel data. Due to the lack of tourism economic data in 2000, 2001 and 2021, the data from 2002 to 2020 is used in this section. And the interpolation method is taken to complete the missing values of these observation samples.

The premise of using the spatial econometric model is that the variables have spatial autocorrelation. Accordingly, this study uses Moran’s I to examine the spatial correlation of the development level of urban tourism. The Moran’s I distribution of tourism revenue and tourist arrivals over the years is drawn in Fig. 4. It can be found that Moran’s I based on the three matrices are all positive, and their corresponding p values are all less than 0.05, indicating that there is spatial autocorrelation in urban tourism economy, which is suitable for spatial econometric analysis.

Fig. 4: Distribution of Moran’s I.
figure 4

This figure shows the Moran’s index of tourism revenue and tourist arrivals under different spatial weight matrices over the years.

Full size image

We test whether spatial durbin model (SDM) can be degenerated into spatial lag model (SLM) or spatial errors model (SEM). Both the Wald test and the LR test significantly reject the null hypothesis that SDM can be degenerated into SLM or SEM. Thus, it is appropriate to choose SDM for estimation in this study. Also, according to the Hausman test, the two-way fixed SDM model is finally selected to estimate the economic spillover effect of tourism in smart city construction.

Table 5 reports the spatial spillover effect of smart city construction on the economic impact of urban tourism after using three spatial weight matrices. It can be found in Table 5 that whether using geographic distance matrix, economic distance matrix or geographic economic nested matrix, the coefficients of Smart City and its spatial lag are significantly positive. In addition, the spatial lag coefficient ρ of urban tourism development is also significantly positive, indicating that the construction of smart cities has a significantly positive spatial spillover effect on urban tourism development. The estimated coefficients of the SDM model cannot directly reflect the marginal effect of the explanatory variables, so we further decompose the spatial effects into direct effects, indirect effects and total effects (Elhorst, 2010). The direct effect is the impact of smart city construction on the development of local tourism industry, the indirect effect is the influence of smart city construction on the development of tourism industry in other areas, and the total effect is the average effect of smart city construction on the overall development of regional tourism industry.

Table 5 Spatial spillover effect.
Full size table

The results show that taking into account the spatial effect, the construction of smart city not only significantly promotes the development of the tourism industry in pilot cities, but also drives the growth of tourism economy in neighboring cities and cities with similar economic development. It plays a demonstration role in the joint development of “center and periphery”, generates the tourism economic effect under spatial spillovers, and promotes the overall development of the tourism industry.

The underlying logic of spatial spillover effect lies in that the innovation and optimization of tourism product formats, business processes and organizational management models in smart cities can derive more segmented industries and broaden the entrepreneurial space for tourism entrepreneurs in different regions. Digital information channels can share local entrepreneurial inspiration and resources with neighboring areas and provide entrepreneurial templates for neighboring areas, effectively increasing the number of tourism enterprises in surrounding areas. The popularization of smart technology can effectively complete the joint push of tourism products around the destination, improve the exposure of neighboring areas, and drive the cross-regional flow of tourists, in order to realize the development of tourism industry in neighboring areas. Local and neighboring service providers can use smart platforms to achieve reciprocal norms. They jointly design routes and promote products, so that tourists’ attention can cover market areas that were difficult to reach in the past, laying a foundation for the interconnection of inter-regional passenger flow (Xiang et al. 2015).

Conclusions and implications

Smart cities are in full swing in many countries around the world. The smart transformation of traditional tourism is one of the core purposes of smart city construction. The potential benefits of tourism development can become an important economic pillar of cities. However, the existing literature lacks relevant empirical evidence on the development of tourism under the construction of smart cities. This study focuses on the tourism economic effect of smart city construction. Taking China’s designated smart city as a quasi-natural experiment, we use the data of 297 cities from 2000 to 2021 to construct a time-varying DID model to empirically test the impact of smart city designation on tourism economic development. Furthermore, the influence mechanism and spatial spillover effect are deeply explored, which to some extent makes up for the shortcomings of current research.

This study finds that the construction of smart cities has significantly increased tourism revenue and tourist arrivals, promoting the upgrading and transformation of the tourism industry. The above conclusions are still valid in the case of handling selection bias, excluding samples from municipalities directly under the central government, and placebo tests. Based on the theoretical framework and data availability, this paper summarizes and confirms that smart cities promote urban tourism economic growth by attracting tourism investment, stimulating tourism entrepreneurship activities, improving tourism operation efficiency, and absorbing relevant talents to work in the tourism industry. To a certain extent, it opens the “black box” of the relationship between the smart city construction and tourism development. Further, the construction of smart cities not only stimulates the development of tourism in the city, but also has a spillover effect on surrounding cities and cities with similar economic development.

Based on the research conclusions, the following efforts can be made in the future. First of all, attaching importance to and supporting the construction of smart cities to face the new challenges of the current tourism economy is a feasible option for regional economic policies. The designated smart cities need to give full play to their pilot role, actively explore and accumulate experience, and gradually promote them. For smart cities with low levels of tourism resources or factor development, both central and local governments need to further tilt their policies or resource factors accordingly. Secondly, the potential of China’s future tourism industry should be centered on promoting the advancement of smart technology, focusing on the research and development investment in smart technology, in order to better play the tourism innovation effect of smart cities. At the same time, we should pay attention to the resource allocation effect and scale economy effect of smart cities, take the improvement of tourism efficiency as a breakthrough point, and accelerate the input of relevant factors, especially the utilization rate of new factors such as knowledge, information and technology. Accordingly, the tourism industry is empowered to transform and upgrade to digitization, networking, and intelligence. Thirdly, the tourism sector should establish a policy system that matches the construction of smart cities, including but not limited to aggregating tourism information resources to achieve accurate matching of tourism supply and demand, reducing the entry threshold of the tourism industry, simplifying the approval procedures for tourism entrepreneurship, and providing relevant skills training in the tourism industry. Finally, a joint marketing mechanism can be established between different cities to attract tourists to carry out multi-destination tourism by shaping a unified brand image and creating characteristic tourist routes, in order to share the dividends of tourists’ corss-regional flow.