The Impact of the "Belt and Road" Initiative on Tourism Economic Development in Chinese Regions along the Route---An Empirical Analysis Based on DID Model

: As an important component of modern service industry, tourism is regarded as a "green" driver to regional economic growth and industrial upgrading. Applying the "quasi-natural experi-ment" advocated by the “The Belt and Road” Initiative, this paper, based on the method of DID, assessed the influence of this initiative on regional tourism development by analyzing panel data collected from 286 prefecture-level cities in China over the period of 2007-2018. It is found that the "The Belt and Road" initiative has stimulated market dynamics and promoted tourism development in the regions along the route, and has significantly improved indicators such as per capita domestic and inbound tourism revenue and per capita domestic and inbound tourist receipt. Also, the degree of contribution of domestic tourists to the tourism economy of the regions where the policy has been implemented is stronger than that of inbound tourists. In sum, the results of this paper not only confirm that the “The Belt and Road” Initiative has significantly contributed to the high -quality tourism development of Chinese provinces along the routes, but also provide important policy enlightenment for the upgrading of regional tourism industry and regional economic balance.


Introduction
In 2015, China released the "Vision and Action on Jointly Building Silk Road Economic Belt and the 21st Century Maritime Silk Road" (hereinafter referred to as the "Vision and Action"). It is a programmatic policy document that maps out the blueprint for the development of the "The Belt and Road" and proposes to strengthen tourism cooperation. For the tourism industry, on the one hand, it enhances the domestic regional coordination and promotes the development of domestic tourism; on the other hand, it pushes forward the opening of tourism, attracts international tourists to visit Chinese attractions, and accelerates the pace of foreign tourism cooperation [1,3]. Thus, it can help the development of tourism economy along the route.
Based on this, this DID-based paper analyzes the panel data of 286 cities at prefecture level and above from 2007 to the end of 2018 as a way to explore the impact of "The Belt and Road" initiative on regional tourism. The paper finds that it has boosted tourists' expenditure in the regions along the route, and the cities affected by the policy have significantly increased tourism revenue and the number of visitors. Both the number of domestic tourists and their relevant spending indicate the positive correlation to the Initiative, but inbound tourists show less significant levels than domestic tourists. These results hold true after controlling for provinces. Further analysis of overnight travel reveals that the main path to growth in inbound travel is through an increase in average tourism spending from foreign visitors, rather than an increase in arrivals. The regression results pass the parallel trend test.
The contributions of this paper are as follows: First, it adds a new perspective to the study of "The Belt and Road" by exploring the impact of "The Belt and Road" on regional tourism in terms of domestic and foreign tourist performance, which enriches the research elements of regional tourism and expands the scope of the research of this initiative. Second, the DID method analyzes the degree of influence of "The Belt and Road" on regional tourism, and the difference between the control group and the experimental one makes the results more credible. Third, this paper assessed the impact of "The Belt and Road" on tourism, which can positively influence the prosperity of regional tourism, and can drive the development of regional tourism through the cities where the initiative is implemented.

Institutional Background
In March 2015, the National Development and Reform Commission (NDRC), the Ministry of Foreign Affairs and the Ministry of Commerce jointly released the "Vision and Action on Jointly Building Silk Road Economic Belt and the 21st Century Maritime Silk Road", announcing for the first time the overall top-level design and strategic plan of the "One Belt and One Road". Subsequently, this initiative was included in the main objectives of the thirteenth Five-Year Plan and was gradually put into in-depth implementation. So far, the "The Belt and Road" initiative has realized the evolution of "concept -strategy -planning -implementation". The vision of the initiative is to "jointly build an open, inclusive, balanced and universally beneficial regional economic cooperation structure, and generally form a network of high-standard free trade zones". As the initiator of "The Belt and Road", China enjoys long history, rich culture and landscape resources, ideal geographic location and huge throughput of tourists, all of which can promote high-quality tourism development from different dimensions. In addition, this initiative can promote sustainable tourism development in the countries along the route, improve the welfare of local communities, stimulate investment, protect the cultural and natural heritage and provide tourists with an unforgettable experience of the Silk Road. The development of tourism of the Belt and Road will go through a process of integration from a loose alliance, and the future goal of this tourism is to establish a tourism free trade zone [2], thus promoting the future development of tourism industry within the region.

Theoretical Hypotheses
In theory, the "The Belt and Road" initiative can form a multifaceted effect on the tourism economy of the domestic provinces along the route. First, policy incentives. The domestic provinces along the route will choose to respond the Belt and Road initiative to increase tourism investment in the region and promote local tourism supply, which is supposed to have a significant impact of enriching local tourism products and optimizing industrial structure [3]. Second, transportation improvement. Factors such as transportation infrastructure affect the agglomeration and diffusion of the regional tourism economy, and are also the key concerns in the transformation and upgrading of the tourism industry and the improvement of quality and speed [4,20]. The implementation of the "The Belt and Road" has further improved the transportation infrastructure in the provinces along the route, which has increased the spatial effect of tourism specialization. Third, the promotion to the flow of tourism production factors. "The Belt and Road" can strengthen tourism cooperation within the policy-implemented region, optimize the efficiency of the tourism industry, promote the use and flow of economic factors, improve the utilization efficiency of tourism resources and coordinate the development of regional tourism [5]. In conclusion, the "The Belt and Road" Initiative strengthens the interconnection of the tourism economies of the domestic cities where the policy is implemented [6], and through the interconnection of tourism resources between cities, a tourism circle is thus formed, which in turn promotes the growth of the local tourism economy. Accordingly, we propose the following hypotheses: Hypothesis 1: "The Belt and Road" stimulates the market dynamics and drives tourism development and tourism economic growth in policy-influencing areas.
At the same time, the "The Belt and Road" policy will also attract foreign tourists to visit China in several ways, which can drive tourism development in the regions along the route. First, policy interoperability. The countries involved in the "The Belt and Road" Initiative will form a more practical cooperation, coordination and effective policy support mechanism. The coordination of tourism-related policies can help connect tourists and tourist destinations to promote travel and mutual recognition, including visa-friendly policies to facilitate the arrival of tourists and mutual acceptance of industry standards, which will strongly support the development of high-quality tourism [1,5,16]. Second, unimpeded trade. The customs cooperation in information exchange, mutual recognition of supervision and assistance in law enforcement in countries along the route widens trade scopes and optimizes trade structure; also, all of these increases the demand for materials, technologies, and resources in domestic areas along the "Belt and Road", further promoting the upgrading of regional industrial structure, provides a good industrial environment for tourism development and significantly increase business tourism and other market sources. Third, financial support. The convenient capital settlement of the countries along the route make the trade including inbound tourism more convenient, and also the elimination of investment and trade barriers diversify the investment and financing channels of the domestic regions along the route, promote the efficient flow of capital factors among industries and provide factor support for the high-quality tourism development. Fourth, cultural communication. The development of cultural projects and activities in countries along the route provides a communication platform for tourism development, which can shorten the psychological distance of tourists. Tourism cooperation, sports activities and other projects promote the development of domestic exhibition tourism, festival tourism and other new industries; harmonious and friendly cultural ecology and public opinion environment is fundamental to the development of "The Belt and Road" tourism [2,14]. In conclusion, one of the objectives of the "Belt and Road" is to promote the opening up of the region and to carry out deeper cooperation and exchange with the international communities; and the implementation of the strategy has contributed to the tourism cooperation of the countries along the route [2,3,7], allowing China's own unique historical and cultural and natural landscapes to attract tourists from all over the world. The "The Belt and Road" Initiative helps the cities where the policy is implemented to be tourist-attracting while maintaining the stable development of domestic tourism, giving full play to the potential of regional tourism development.

Hypothesis 2: "The Belt and Road" attracts foreign tourists, increasing the number of tourist arrivals in China and boosting the development of regional tourism.
However, in reality, this initiative has had different impacts on domestic and international tourists. Firstly, visitors to China face higher costs of travel and information gathering. Therefore, under the premise of information asymmetry and the pursuit of maximizing the utility of tourism experience, international tourists tend to choose those Chinese tourist destinations with high quality and reputation, and thus have limited effect on the tourism economy of cities without well-known tourist attractions, but domestic tourists are not likely to be affected by this above situation. Secondly, for traveling abroad, tourists face higher travel costs and also need to spend more time and energy to prepare for the trip, and inbound tourists will also need longer touring time to maximize the utility [8], therefore, it will also form a certain obstacle to the number of tourists coming to China, but the increase of inbound tourists is lower than that of domestic tourists. Accordingly, we propose the following hypothesis.

Identification Strategy
The "Belt and Road" Initiative began in 2015, involving 140 prefecture-level cities, and this paper uses the DID method of time and region to analyze its impact on regional tourism. By the end of 2018, there were 298 prefecture-level cities, sub-provincial cities, and municipalities directly under the Chinese central government, and from 2007 to 2018, the number of Chinese prefecture-level cities has changed, so to ensure consistency, the data of Bijie, Chaohu, Sansha, Haidong, Danzhou, Turpan, Hami, Rikaze, Changdu, Linzhi, Shannan, and Naqu are deleted, and finally the data of 286 cities analyzed in the paper. Among them, there are 140 cities involved in "The Belt and Road" Initiative, which are the experimental group; cities in other regions constitute the control group. Specifically, this paper implements the DID method through two-way fixed effects. The model is, In which Tour represents tourism development; BR represents whether the city is the participant of "The Belt and Road" Initiative; Control is the control variable; i represents the order of respective city, and t represents the t-th year; represents the time fixed effect and represents the regional fixed effect.

Data, Variables and Descriptive Statistics
To study the impact of "The Belt and Road" Initiative on the development of regional tourism, this paper introduces domestic tourism revenue, domestic tourist receipts, inbound tourism revenue, inbound tourist receipt, total domestic and inbound tourism revenue and total domestic and inbound tourist receipt to indicate the level of regional tourism development; and uses "The Belt and Road" to indicate the cities participating in this initiative. A series of control variables are also used, including the number of star hotels per 10,000 people, the number of taxis per 10,000 people, the per capita consumption of electricity, the level of medical care, the level of higher education, the per capita GDP, the percentage of population in the tertiary sector and the greening rate. In this paper, per capita domestic tourism revenue, per capita domestic tourist receipts, inbound tourism revenue per capita, inbound tourist receipts per capita, total domestic and inbound tourism revenue per capita and total domestic and inbound tourist receipts per capita are selected to characterize the level of Tour(tourism) development. The core explanatory variable Belt and Road (BP) is denoted by BR. The 140 provinces that participated in Belt and Road in 2015 were assigned the value of 1 from that year and 0 for the others.
In order to control the influence of other factors on the development of regional tourism, a series of variables were selected as control variables in this paper based on the literature findings and reality. In the process of local tourism development, per capita GDP, which represents the level of economic development, and electricity consumption per capita play a very important role [9,15]; the ability of tourist attractions to provide good services to tourists is also an important part of local tourism development, so this paper uses the variable of the number of star hotels per 10,000 people to reflect the local service reception capacity [10,18]; the transportation directly affect the accessibility of tourist attractions [11,17,20], therefore, the number of taxis per 10,000 people is used to reflect the accessibility of local transportation; higher education means more convenient cultural exchange and is more likely to attract foreign tourists [12,18,19], so the "number of higher education schools/total population at the end of the year" is adopted to reflect the local higher education penetration rate. In addition, travelers have a strong preference for mountainous and forested areas with high greenery rate, and volcanoes and deserts with low greenery rate, so the quadratic term of greenery rate and greenery rate is used to reflect the attractiveness of natural environment to travelers. This paper uses Chinese provincial data (with deletions) from 2007 to 2018 to analyze the impact of the Belt and Road on tourism. Data sources are China City Statistical Yearbook, China Tourism Statistical Yearbook, and provincial economic statistical yearbooks. This paper uses the software, stata12 for regression analysis of the data and to prevent errors, time effects and individual effects are controlled. The regression results are shown in Table 2 and Table 3.

Preprints
Before controlling the regional economic development level, traffic accessibility, higher education level, medical care level and landscape preference, the regression results show that the coefficients of different explanatory variables reflecting tourism development level, such as domestic tourism revenue and domestic tourist arrivals, are positive and significant, and the regression results thus support hypothesis 1. The coefficients of inbound tourism revenue and inbound tourist arrivals are positive and significant, and the regression results thus support hypothesis 2. Comparing inbound tourism and domestic tourism, the coefficients of inbound tourist arrivals and revenues are significantly smaller than those of domestic tourism, and the degree of significance is also lower, and therefore, the regression results thus support hypothesis 3.
After adding controlled variables, the regression results still confirm the positive impact of Belt and Road on regional tourism development, as shown in Table 2.

Inclusion of Joint Geographic-temporal Fixed Effect
As different tourism economic policies are adopted in different years in the central region (CR), eastern region (ER) and western region (WR), and their attention and investment in tourism vary, the characteristics of regional changing over time can affect the development of tourism in each province. Considering that regional tourism development can be affected by overall tourism planning and policies at different geographic levels, this paper further considers joint geographic-temporal fixed effects in the robustness tests to capture the policy effects of each province over time. The impact of geographic differences for each province is examined.
The regression results are shown in Table 3. The coefficients of domestic tourism trips and domestic and inbound tourist arrival are significantly positive after adding the geographical control, and the other results are not significantly different from Table 2, i.e., the coefficients of inbound tourism and the degree of significance are smaller than domestic, which further supports Hypothesis 3.

Parallel Trend Test
The DID analysis assumes that the experimental and control groups have the same trend in the absence of event influence and that there is no ex ante error. The analysis in this paper assumes that tourism development in the control and experimental groups can maintain the same trend per se. If tourism development in policy-implementing and nonpolicy-implementing cities is systematically different, the premise of DID will not exist.
Parallel trend tests were conducted on the tourism development levels of the control and experimental groups in the first two and last four years of the Belt and Road, and the regression coefficients were plotted. The results are shown in Figures 1 to 6. In the year of before1, before2 and the present year of current, the curves fluctuate above and below the value of 0, indicating that there is no systematic difference. In the year of after1, after2, and after3, the curves rise significantly, indicating that "One Belt, One Road" has brought positive impact on regional tourism development, and the impact is gradually deepened in these three years. The curve declines sharply in the year of after4, indicating that the positive impact declines in the fourth year after the implementation of this initiative. The impact of "One Belt, One Road" is generated after a one-year delay and deepens over the next three years, before declining in the fourth year. However, it should be noted that domestic tourism revenue, domestic tourist receipts, inbound tourism revenue, inbound tourist receipts, and total domestic and inbound tourism revenue all show the above trends and satisfy the requirements of the parallel trend test. But the total domestic and foreign tourist receipts did not pass the test, so this indicator should not be used as a variable for DID.   In practical situations, there may be extreme cases affecting the regression results. For example, some cities have data that have a great positive impact on the regression results due to rich tourism resources, smooth transportation environment, and good policy conditions; while some cities have the opposite. The occurrence of extremes can bring errors to the regression results. In order to exclude the influence of extreme values, this paper uses 1% and 5% winsorization to the data to exclude the extreme values above this degree in the variables respectively, and then the results are tested. The new regression results are similar to the above, and the effects of extreme values can be excluded in the data used in this paper.

Heterogeneity Analysis
The market operation mechanism is based on the existing tourism resources in the region, and the degree of integration and utilization of tourism resources in the region is different, and the link between the richness of regional tourism resources and the impact of "One Belt, One Road" on the regional tourism economy is not consistent [14]. The impact of "Belt and Road" on regional tourism economy should also be considered.
In this paper, the number of 5A scenic spots is used to represent the richness of regional tourism resources. 5A is the highest rating of scenic spots in China and represents the most authoritative distribution of tourism resources in China. The data used are from the 5A scenic spots list of 2020, because the formation of the landscape requires a long period of accumulation, even though some attractions are excluded in 201 for cross-regional scenic spots (some scenic spots are distributed across a wide area, making it difficult to identify the area to which their tourism revenue belongs) and scenic spots not under the jurisdiction of prefecture-level municipalities (tourism revenue of scenic spots not under the jurisdiction of prefecture-level municipalities is not within the scope of prefecture-level municipal tourism revenue and are thus not included in provincial statistics As shown in Table 4, the impact coefficients of the cross term on domestic tourism revenue, domestic tourist arrival, inbound tourist arrival, and domestic and per capita inbound tourism revenue are significant, but the impact coefficients on inbound tourist arrival are not significant; moreover, the coefficients of the six regressions are all negative. This indicates that the richer the resources of 5A scenic spots, the smaller the positive impact of "One Belt, One Road" on regional tourism economy. It can also be interpreted that the regions with abundant tourism resources are less dependent on the tourism economy promoted by the "Belt and Road"; but the regions with insufficient tourism resources can gain the favor of domestic tourists through the "Belt and Road", including the increase in both of the number of domestic tourists and the spending of domestic tourists, which will bring new growth points for their tourism economies. This is in line with the above analysis that the impact of Belt and Road on domestic tourists is greater than that on inbound tourists. Meanwhile, this paper uses excellent tourism cities as representative variables of tourism resources included in the analysis, and its cross-term regression results are similar to those of the 5A scenic spots and thus will not be repeated.

Conclusion
The regression results of this paper support the hypothesis that "One Belt, One Road" has boosted the development of tourism economy in the regions along the route. The "Belt and Road" has significantly increased domestic tourism revenue per capita, inbound tourism revenue per capita, domestic tourist receipt per capita, foreign tourist receipt per capita, and domestic and inbound tourism revenue per capita, while the contribution of domestic tourists to the tourism economy of the regions where the policy is implemented is stronger than that of inbound tourists. The "Belt and Road" Initiative has promoted the development of tourism in the cities where the policy is implemented. And domestic and inbound tourists show a higher level of interest in tourism in the cities where the Initiative is implemented, as evidenced by higher average spending. The stronger attraction of the Belt and Road Initiative cities to domestic tourists may stem from their excellent performance in terms of tourism economic linkages, unique tourism resources and tourism transportation conditions and advantages that accelerate the chances of tourism cooperation and integration of regional tourism effects. The increased interest of Chinese tourists in these cities is reflected in the increase of the average spending of the intrinsically interested groups, while they do not significantly increase the attractiveness of the originally uninterested groups.

Policy Enlightenment
First, the analysis suggests that the Belt and Road has had a positive impact on tourism in the cities where the policy is implemented. China's implementation of the Belt and Road has not only strengthened its ties with foreign countries and increased its openness to the rest of the world, but has also greatly contributed to the development of tourism in the related cities. Therefore, Belt and Road cities should actively integrate and accelerate the implementation of specific rules and regulations to take advantage of the Belt and Road. As China's economy is facing the double pressure of eliminating backward production capacity and upgrading industries, the Belt and Road has strengthened domestic and international ties and contributed to breaking down regional barriers and releasing economic influence.
Second, the results of this paper show that the positive impact of Belt and Road on tourism development in policy-implementing cities exceeds that of non-policy-implementing cities which should actively integrate into the Belt and Road development strategy, using policy-implementing cities as core circles to expand their tourism influence and seek to absorb the benefits brought by the Belt and Road. Many cities, especially nonpolicy implementation cities in central and western China, have rich natural landscapes and unique humanistic landscapes, but due to their geographical location, they are unable to give full play to their tourism resources. Therefore, the policy should focus on narrowing the time distance between the "Belt and Road" to boost the potential of the local tourism industry and thus the utilization rate of tourism resources can be improved while satisfying the diversified tourism needs of people. The interaction of tourism elements between cities implementing the policy promotes closer tourism economic ties and achieves synergistic development between coastal and inland cities implementing the policy.
Third, the results of this paper show that the "One Belt, One Road" has had a significant impact on the average spending of visitors to China, but not on the number of tourist arrivals. In response to this feature, the cities implementing the policy should consider providing higher convenience to meet the spending needs of tourists coming to China. By developing a more flexible foreign exchange conversion mechanism and more extensive promotion, the existing tourist attraction groups should be deepened and their spending needs can be explored.

Data Availability Statement:
The data presented in this study are not publicly available due to the intention of further research. The data are available on request from the corresponding author.
Funding: This research received no external funding.