Using a Dynamic Spatial Difference-in-Differences estimator to evaluate the effect of High speed rail and tourist transit service on Tourism Demand

Using a Dynamic Spatial Difference-in-Differences estimator to evaluate the effect of High speed rail and tourist transit service on Tourism Demand | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Research Article Using a Dynamic Spatial Difference-in-Differences estimator to evaluate the effect of High speed rail and tourist transit service on Tourism Demand Tzu-Ming Liu This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-5219377/v1 This work is licensed under a CC BY 4.0 License Status: Published Journal Publication published 19 Dec, 2024 Read the published version in Transportation → Version 1 posted 9 You are reading this latest preprint version Abstract This study uses the Dynamic Spatial Difference-in-Differences model (Dynamic SDID) to analyze the impact of the Taiwan High-Speed Rail (THSR) on Taiwan's tourism demand. To control for spillover effects, the model incorporates the Taiwan Tourist Shuttle service (TSHU) as an alternative transportation option, the interactive effects between TSHU and THSR, and the spatial autocorrelation between TSHU and THSR. The analysis results indicate that controlling for spillover effects is crucial for analyzing the impact of the High-Speed Rail and tourist transit service on Tourism Demand, and the Dynamic SDID is a better analytical model for this purpose. The THSR has a significant positive impact on tourism demand, while its spatial autocorrelation effect is significantly negative. This suggests that the increase in tourist traffic brought about by THSR mainly comes from existing tourists in the surrounding areas rather than generating new tourism demand. The TSHU, on the other hand, has a negative but insignificant impact on tourism demand, but its interaction with THSR has a significant positive effect, indicating that the two services complement each other. Therefore, to enhance Taiwan's tourism demand, the focus should still be on improving the attractiveness of tourist destinations rather than solely relying on the construction of the High-Speed Rail. Additionally, while the TSHU does not contribute significantly to the development of specific individual tourist destinations, it does facilitate regional tourism development. Therefore, selecting TSHU routes based on actual market conditions can promote the growth of the tourism industry. Taiwan High Speed Rail Taiwan Tourist Shuttle service Dynamic SDID CIPS test Spillover effect Spatial econometric analysis Figures Figure 1 1. Introduction Transportation is an essential key element in the development of tourism. Improving transportation infrastructure can offer tourists safer, more comfortable, and efficient means of travel, attracting more visitors (Khadaroo and Seetanah 2007 ). Therefore, in theory, enhancing transportation infrastructure is expected to increase market accessibility, reduce travel costs, and make tourist destinations more attractive. However, the research on high-speed rail (HSR) does not completely support this argument. Hence, while France, Germany, Italy, Spain, Japan, and Taiwan have built HSR systems, countries like the U.K., United States, India, and Malaysia have faced skepticism in their planning and construction of HSR. Given the significant cost of HSR construction, understanding its actual impact on tourism and regional development is crucial to optimizing tourism resource allocation. This includes identifying whether the effects result from the characteristics of HSR itself or from issues in data analysis. Over the past two decades, many countries have constructed HSR as a significant means to boost tourism demand (Albalate and Fageda 2016 ). Some studies show that HSR promotes tourism demand (Deng, Gan et al. 2021 ), while others indicate it boosts tourism growth but reduces tourism revenue (Gao, Su et al. 2019 ), or that the effect is only significant in the first year after HSR introduction and diminishes in subsequent years (Kurihara and Wu 2016 ). Studies from the U.S. (Yu and Fan 2018 ), Spain (Pagliara, La Pietra et al. 2015 ), and France (Delaplace, Pagliara et al. 2014 ) reveal that only a few cities experienced an increase in tourists after the introduction of HSR. Bazin, Beckerich et al. ( 2006 ) argue that HSR may fail to ignite tourists' curiosity and have limited impact on leisure tourism. Albalate and Fageda ( 2016 ) attribute the differing research results to variations in econometric methods and the network design and accessibility of HSR, which relate to the spillover effects of tourism and HSR and the connections between them. Tourism spillover effects refer to the impact of tourism activities in one area on nearby areas (Liu 2020 ). Liu ( 2020 ) uses Becker's social capital theory (Becker and Murphy 2009 ) and tastes (Becker and Becker 2009 ) as the theoretical basis for exploring tourism spillover effects. Multidestination travel (de Oliveira Santos, Ramos et al. 2011 ) is also a source of tourism spillover effects. Multidestination travel, based on Lancaster's (1966) characteristics theory (Lancaster 1966 ), involves visiting more than one destination in a single trip, allowing tourists to maximize utility under cost and time constraints (Hwang and Fesenmaier 2003 ). Improving transportation infrastructure reduces the cost of traveling to different destinations, leading to more characteristics (destinations) included in tourists' travel decisions, thus increasing the number and possibilities of visiting multiple destinations. Spillover effects are not just theoretical predictions but have empirical support. Empirical studies show that spillover effects are essential explanatory variables in tourism demand models. For instance, tourists can take HSR to a central city in a region and then transfer to road transportation to reach their final destination. As a result, there will be transport spillover effects from the hub city to the destination, contributing to tourism spatial agglomeration (Masson and Petiot 2009 ). Improved transportation to specific tourist destinations creates more opportunities for joint promotion or marketing between that destination and neighboring ones, enhancing the overall tourism competitiveness of both the destination and surrounding areas, thus generating spillover effects of transportation construction on tourism (Gooroochurn and Hanley 2005 , Dehghan Shabani and Safaie 2018 ). However, HSR does not always produce positive spillover effects on tourism and may also have negative effects. The improved accessibility brought by HSR benefits the cities where stations are located due to their better transport resources, services, and facilities, which attract more tourists and also draw tourists from neighboring cities. The substantial demand for tourism generated by HSR further concentrates tourism resources, providing more specialized tourism services and attracting more tourists, intensifying the competition in the local tourism market and suppressing the tourism development of surrounding destinations. Therefore, HSR creates spillover effects, but the net impact of these effects is not universally applicable. HSR spillover effects are related to the aforementioned tourist transportation services. Tourist transportation services can help transport HSR-generated tourists to surrounding areas, resulting in positive spillover effects. On the other hand, they can also redirect tourists from surrounding destinations to the HSR destination, reducing tourism activities in those areas and leading to negative spillover effects. Public road transportation includes cars, buses, and coaches. Buses and long-distance coaches are generally operated by public transport networks or tourism agencies. An excellent road transportation system acts as a catalyst for the growth of the tourism industry. Many studies have confirmed the positive impact of road transportation on tourism development (Kanwal, Rasheed et al. 2020 ). Road transportation provides greater accessibility to tourist destinations (Masson and Petiot 2009 ), encourages local business activities (Khadaroo and Seetanah 2007 ), attracts tourists, and promotes new tourism destinations (Currie and Falconer 2014 , Virkar and Mallya 2018 ). Road transportation's advantages include convenience, flexibility, affordability, ease of access to tourist destinations, fewer baggage restrictions, and greater agency in personal travel experiences (Virkar and Mallya 2018 , Kanwal, Rasheed et al. 2020 ). However, there has been less attention in the academic community to tourist transportation shuttle bus services (TTSBS) compared to other forms of transportation (Peeters and Schouten 2006 , Prideaux 2018 ), and their interaction with HSR's impact on tourism is not well-studied, thus requiring further investigation. While DID and panel data techniques have been used in some papers to study the impact of HSR on tourism development (Albalate, Campos et al. 2017 , Fang 2021 , Di Matteo 2022 , Wang, Ma et al. 2023 ), there are still shortcomings in these studies. Some of the limitations include insufficient consideration of HSR's spillover effects on tourism development and the neglect of the impact of transferring to road transportation. Only a few studies have combined spatial econometric methods and DID in Spatial Difference-in-Differences (SDID) analysis of HSR's impact on tourism (Yang, Hu et al. 2021 ). Yang, Hu et al. ( 2021 ) use the spatial autoregressive lag (SAR) and spatial error term (SE) to represent HSR's spillover effects. Yu ( 2021 ) uses the spatial autoregressive lag (SAR) and spatial DID term to capture the spillover effect. The former, while SDID, only includes a simple dependent variable spatial lag term, and since the paper does not use the spatial weighting term of HSR, it cannot control for the spillover effect of the high-speed rail. The latter uses the spatial DID term but does not consider TTSBS or the spatial autocorrelation of the dependent variable lag. Both studies have omitted essential explanatory variables, leading to possible estimation biases. Considering the importance of HSR's impact on tourism industry development and the advantages of the DID method in policy analysis, this study addresses the previously mentioned research gaps and investigates the impact of TTSBS on HSR's effect on tourism activities. By supplementing important control variables not adequately explored in previous literature, this study aims to avoid omitted variable biases. To achieve this research goal, this study employs the SDID analysis with a dependent variable lag term to examine HSR's impact on regional tourism. The structure of this paper is as follows: The introduction presents the controversy surrounding HSR's impact on tourism development, proposes statistical analysis methods, and discusses the potential effects of omitting important control variables on the previous controversy, making a case for the rationality of using SDID as the analytical method. The second section introduces the research methodology, providing an overview of the traditional DID model and extending it to the Dynamic SDID. It also explains the data sources for the variables. The third section presents the results of the statistical analysis. The fourth section provides conclusions and policy recommendations. 2. Methodology 2.1 Introduction of Dynamic Spatial Difference-in-Difference model The Difference-in-Differences (DID) method is commonly used to estimate the effect of a specific policy (treatment) by comparing the changes in the outcome of the group affected by the policy (treatment group) and the group unaffected by the planned intervention (control group) over a specific period. As outcomes may be influenced by external factors and may change over time, simply observing the simple changes in outcomes before and after the treatment is not sufficient to draw causal conclusions, as factors other than the treatment might influence the results over time. Additionally, comparing participating and non-participating groups alone may lead to selection bias and differences in unobservable characteristics between the groups. DID combines these two approaches by comparing the differences in outcome changes between the treatment and control groups before and after the treatment, estimating the overall impact of the program. DID uses data from before and after the policy intervention, such as group or panel data (individual-level data that varies over time) or repeated cross-sectional data (individual or group-level data). This method eliminates biases caused by inter-group permanent differences in post-intervention comparisons and biases from trends in the intervention group caused by other outcome factors over time. Therefore, even without random selection of the affected sample, causal inferences can be made. DID calculates the before-and-after difference in outcomes for the treatment group, which is the first difference (Fig. 1 , Table 1 ). By comparing the same group with itself, the first difference can control for factors that remain constant within the group over time. Then, to capture time-varying factors, the difference in differences calculates the before-and-after difference for the control group, which is affected by the same environmental conditions as the treatment group but not by the policy. This is the second difference. Finally, DID "cleans out" all time-varying factors by subtracting the second difference from the first difference. This gives us the estimated effect, namely the Difference-in-Difference. Table 1 DID model coefficient relationship in regression Control Group (i = 0) Without THSR Stations Treatment Group (i = 1) With THSR Stations Difference (1) Pre-intervention (t = 0)(Before THSR operation) \(\:{\alpha\:}_{0}\) \(\:{\alpha\:}_{0}{+\alpha\:}_{1}\) \(\:{\alpha\:}_{1}\) Post-intervention (t = 1)(After THSR operation) \(\:{\alpha\:}_{0}{+\alpha\:}_{2}\) \(\:{\alpha\:}_{0}{+\alpha\:}_{1}+{\alpha\:}_{2}{+\alpha\:}_{3}\) \(\:{\alpha\:}_{1}{+\alpha\:}_{3}\) Difference(2) \(\:{\alpha\:}_{2}\) \(\:{\alpha\:}_{2}{+\alpha\:}_{3}\) \(\:{\alpha\:}_{3}\) Footnote: The simple DID estimator allows for the intercepts to vary between the treatment ( \(\:{\alpha\:}_{0}{+\alpha\:}_{1}\) ) and the control group ( \(\:{\alpha\:}_{0}\) ) and assumes constant outcomes within the two time periods ( \(\:{\alpha\:}_{1}\) ). \(\:{\alpha\:}_{1}\) : Average difference in outcome between the two groups that is common in both time periods. \(\:{\alpha\:}_{2}\) : Average change in outcome from the Pre-intervention to the Post-intervention time period that is common to both groups. \(\:{\alpha\:}_{3}\) : Average differential change in outcome from the first to the second time period of the treatment group relative to the control group DID requires outcome data for the policy treatment group and control group, as well as data before and after policy implementation. The following steps calculate the difference in differences (Fig. 1 ): Calculate the before-and-after difference in outcomes for the treatment group (B-A). Calculate the before-and-after difference in outcomes for the control group (D-C). Calculate the difference between the before-and-after differences in outcomes for the treatment group (B-A) and the control group (D-C). This is the DID: (B-A) - (D-C). Figure 1 and Table 1 also present the specific Stable Unit Treatment Value Assumption (SUTVA) of DID: Intervention unrelated to outcome at baseline, Treatment/intervention and control groups have Parallel Trends in outcome, Composition of intervention and comparison groups is stable for repeated cross-sectional design, No spillover effects. The traditional DID analysis of HSR policy impact can be represented using the following formula (Albalate and Fageda 2016 , Albalate, Campos et al. 2017 ). $$\:{y}_{i,t}={\alpha\:}_{0}+{\alpha\:}_{1}{HSR}_{i,t}+{\alpha\:}_{2}{Pos{t}_{construction}}_{i,t}+{\alpha\:}_{3}\left({HSR}_{i,t}\times\:{Pos{t}_{construction}}_{i,t}\right)+{X{\prime\:}}_{i,t}\theta\:+{ϵ}_{i,t}$$ 1 In Eq. ( 1 ), \(\:{HSR}_{i,t}\) and \(\:{Pos{t}_{construction}}_{i,t}\) are dummy variables: when i = 1, it indicates the treatment group, i = 0 indicates the control group; when t = 1, it indicates after the experimental treatment, and t = 0 indicates before the experimental treatment. The regression coefficient relationship is summarized in Fig. 1 and Table 1 : \(\:{\alpha\:}_{1}\) , B-A, shows the difference between different groups at the same time point; \(\:{\alpha\:}_{2}\) , D-C, shows the difference between the same groups at a different time point; \(\:{\alpha\:}_{3}\) , (B-A)-(D-C), represents the true experimental treatment effect when both treatment and after treatment are 1; if there is no effect of the experimental treatment, \(\:{\alpha\:}_{3}\) is 0. The variable \(\:{y}_{i,t}\) indicates the policy outcome of HSR , and X denotes the control variables. In this study, the policy outcome, \(\:{y}_{i,t}\) , represents tourism development, and we use the logarithm of tourist arrivals to characterize tourism development. Empirical evidence shows that tourist arrivals have significant spatial spillover effects (Liu 2020 ), and therefore, the standard DID model (Eq. 1 ) cannot yield consistent estimation results because SUTVA does not hold (Li and Li 2020 , Pan, Cong et al. 2020 , Tian, Yang et al. 2022 ). However, the spatial difference-in-differences (SDID) model can consider spatial effects (Chagas, Azzoni et al. 2016, Diao, Leonard et al. 2017, Ferman 2023). Accordingly, we construct the SDID model to accurately measure the influence of HSR on tourism. Previous studies have mainly considered the spillover effects of tourism arrivals, incorporating the spatial lag term of tourism development into the model (Liu 2020 ), as well as the time lag term of the dependent variable. In line with these previous studies, this paper combines the spatial autocorrelation model (SAR) with the DID model (Eq. 1 ) and constructs the SDID model as follows: $$\:{y}_{i,t}={\alpha\:}_{0}+{{\alpha\:}_{1}{HSR}_{i,t}+{\alpha\:}_{2}{Pos{t}_{construction}}_{i,t}+{\alpha\:}_{3}({HSR}_{i,t}\times\:{Post\_construction}_{i,t})+\beta\:}_{1}{y}_{i,t-1}+{\omega\:}_{1}W{Y}_{t-1}{+\omega\:}_{2}W{Y}_{t}+{\omega\:}_{3}W{HSR}_{t}+{X{\prime\:}}_{i,t}\theta\:+{ϵ}_{i,t}$$ 2 In Eq. 2 , \(\:{y}_{i,t-1}\) represents the time lag term of \(\:{y}_{i,t}\) , which refers to tourist arrivals, making Eq. 2 a Dynamic Difference-in-Differences model (Dynamic DID). The spatial connectivity matrix W quantifies the spatial spillover of tourism development. Furthermore, \(\:W{Y}_{t-1},\:W{Y}_{t},\) and \(\:W{HSR}_{t}\) are spatial lag terms of \(\:{Y}_{t-1},\:{Y}_{t},\) and \(\:{HSR}_{t}\) , respectively, used to control for the spatial spillover effects of tourism development, making Eq. 2 a Dynamic Spatial Difference-in-Differences model (Dynamic SDID). In this study, we also investigate whether TTSBS (tourist transportation shuttle bus services) affect HSR's effects on tourism development. Therefore, we extend Eq. 2 as follows: $$\:{y}_{i,t}={\alpha\:}_{0}+{{\alpha\:}_{1}{HSR}_{i,t}+{\alpha\:}_{2}{Pos{t}_{construction}}_{i,t}+{\alpha\:}_{3}({HSR}_{i,t}\times\:{Post\_construction}_{i,t})+\beta\:}_{1}{y}_{i,t-1}+{\omega\:}_{1}W{Y}_{t-1}{+\omega\:}_{2}W{Y}_{t}+{\omega\:}_{3}W{HSR}_{t}+{\omega\:}_{3}W{SHU}_{t}+{\gamma\:}_{1}{SHU}_{i,t}+{\gamma\:}_{2}{SHU}_{i,t}{HSR}_{i,t}+{X{\prime\:}}_{i,t}\theta\:+{ϵ}_{i,t}$$ 3 In Eq. 3 , \(\:{SHU}_{i,t}\) represents the number of TTSBS (tourist transportation shuttle bus services), and \(\:W{SHU}_{t}\) represents the spillover effects of \(\:{SHU}_{t}\) . The term \(\:{SHU}_{i,t}{HSR}_{i,t}\) represents how \(\:{SHU}_{i,t}\) affects HSR’s effects on tourism development. This extension allows us to explore the potential interaction between TTSBS and HSR in their combined impact on tourism development. 2.2 Introduction of Variables and Data Sources This study uses commonly used and significant explanatory variables from the literature, including income (Lim 2006, Song, Witt et al. 2009 ), price (Song and Witt 2006 ), population variables (Song, Witt et al. 2009 ), lag terms of the dependent variable (Garín-Muñoz 2009, Zhang, Kulendran et al. 2010, Yap and Allen 2011, Massidda and Etzo 2012, Liu 2020 ), as well as the variables related to the research topic, namely, Taiwan High Speed Rail (THSR) and Taiwan Tourist Shuttle service (TSHU). Commonly used and significant explanatory variables for tourism demand have been extensively applied in tourism demand studies, but due to space limitations, this paper does not provide detailed explanations. Interested readers can refer to the listed references. The following will elaborate on the transportation-related policy variables that this paper focuses on (Albalate and Fageda 2016 , Gao, Su et al. 2019 ). (1) Introduction of Taiwan High Speed Rail The idea of Taiwan High Speed Rail (THSR) originated from the "Development of Super Express Railway Project" proposed by the Taiwan Railway Administration in 1974. In response to the deteriorating transportation service quality and increasing congestion in the western region of Taiwan, the Ministry of Transportation conducted the "Feasibility Study of Taiwan Western Corridor High-Speed Rail" in 1987. On July 14, 1994, it was included in the government's priority list of major public construction projects and became the country's first significant national infrastructure project promoted through private investment. The THSR route spans 350 kilometers and includes 12 stations, such as Nangang, Taipei, Banqiao, Taoyuan, Hsinchu, Miaoli, Taichung, Changhua, Yunlin, Chiayi, Tainan, and Zuoying. The stations were opened for service on different dates: January 5, 2007, for Banqiao, Taoyuan, Hsinchu, Taichung, Chiayi, Tainan, and Zuoying; March 2, 2007, for Taipei; December 1, 2015, for Miaoli, Changhua, and Yunlin; and July 1, 2016, for Nangang. The THSR trains can reach a maximum speed of 300 kilometers per hour, significantly reducing the travel time between Taipei and Kaohsiung to 90 minutes. The daily passenger capacity of THSR exceeds 300,000. After the THSR was launched, it transformed Taiwan's spatial structure and turned the northern, central, and southern metropolitan areas into a balanced one-day living circle (Ministry of Transportation and Communications, Railway Bureau, 2020). (2) Introduction of Taiwan Tourist Shuttle service Taiwan Tourist Shuttle service (TSHU) is a bus service specifically designed for tourism, providing transportation for travelers from major Taiwan Railways and THSR stations to Taiwan's main tourist attractions. Its purpose is to offer convenient public transportation services to attract travelers, reduce the proportion of self-driving trips, and avoid traffic congestion, parking shortages, and related issues affecting travel quality. Additionally, it aims to increase the use of public transportation among locals to promote energy-saving and carbon reduction in tourism activities. After analyzing the rapid increase in the number of tourists traveling in Taiwan, the tendency of using private vehicles for travel, and the preference for free and independent travel, the Tourism Bureau of the Ministry of Transportation concluded that there was a need to establish transportation services to cater to the needs of independent travelers. Since 2009, they began planning and providing travel shuttle services for independent travelers (Tourism Bureau, Ministry of Transportation, 2012). The route selection criteria for TSHU are as follows: "(1) Select attractive destinations: As transportation services are essentially derived from travel needs, the key is still the attractiveness of the destinations. There must be compelling reasons for travelers to visit the locations so that the transportation services provided can meet the needs of travelers, and operators can obtain appropriate profits to provide long-term services. (2) Collaborate with multiple transportation modes: To promote the use of public transportation for travel in Taiwan, it is necessary to link the shuttle routes to nearby major long-distance transportation stations, such as Taiwan Railways and THSR stations, to facilitate travelers' transfers and usage. (3) Fast and stable transportation services: To attract independent travelers to take advantage of the convenient shuttle service and facilitate travel planning, the shuttle routes should be designed differently from the existing public bus services, with fewer stops, stable schedules, and reasonable intervals, providing fast and direct access to tourist attractions. (4) Integrated preferential ticketing (5) Short-term subsidies and long-term autonomous operation" After the evaluation process, TSHU officially started operating on April 5, 2010. Each year, based on the operation status of selected routes and new route applications, adjustments are made to the routes for the following year. The data for the aforementioned variables come from various statistical databases of the respective authorities. The number of tourists in each city and county is from the Tourism Statistics Yearbook of the Tourism Bureau, Ministry of Transportation. Data on real per capita national income and the total population are from the databases of the Directorate-General of Budget, Accounting and Statistics. Transportation-related policy variables are sourced from the Railway Bureau and the Tourism Bureau of the Ministry of Transportation. 2.3 Unit Root Test for Data This study utilizes panel data from 19 counties and cities in Taiwan from 2001 to 2020. The analysis of panel data depends on whether the data possesses unit root characteristics, such as the existence of long-term trends or cointegration relationships. If the data exhibits unit root characteristics, indicating the presence of long-term trends or cointegration, it is necessary to perform data transformations such as differencing to ensure the reliability of the analysis results; otherwise, hypothesis testing may become invalid. Therefore, before conducting the analysis, we follow the suggestion of González-Val and Silvestre ( 2023 ) and the approach of Udeagha and Muchapondwa ( 2023 ) for analyzing SDID, which involves conducting panel unit root tests on the tracking data. Commonly used panel data unit root tests include the Levin-Lin-Chu (LLC), Im-Pesaran-Shin (IPS), Fisher-type, and Hadri tests. However, these methods may not be suitable for the data in this study. LLC and IPS assume that each individual is independent, and if there is spatial correlation in the data, resulting in cross-sectional dependence, the mean of the t-statistic may not be zero, leading to non-zero bias and affecting the test results. While Fisher-type can tolerate a certain degree of cross-sectional correlation, if the cross-sectional correlation is too strong, the Fisher statistic may still become invalid. Hadri uses the OLS estimation method to calculate the LM statistic, but this test is also influenced by cross-sectional dependence, leading to an overestimation of the LM statistic. Since Taiwan's tourist arrivals exhibit spatial correlation (Liu 2020 ), indicating the presence of cross-sectional dependence, the LLC, IPS, Fisher-type, and Hadri tests may produce biased test results. When cross-sectional dependence is present in panel data, panel unit root tests that consider cross-sectional dependence should be used. Commonly used panel unit root tests considering cross-sectional dependence include the Pesaran ( 2007 ) CD test, Moon and Perron (2004) CD test, and Breitung (2000) test with cross-section dependence adjustment. However, the assumptions of these methods do not match the characteristics of the data in this study. Pesaran ( 2007 ) CD test has strict assumptions regarding cross-sectional dependence, and when the form of cross-sectional dependence is more complex, the test's performance may be affected. Moon and Perron (2004) CD test requires estimating the degree of cross-sectional dependence, which may encounter computational difficulties with high-dimensional panel data. Breitung (2000) test has strict assumptions about cross-sectional dependence, and its estimation of the degree of cross-sectional dependence may not be accurate enough. The data in this study exhibit complex forms of cross-sectional dependence due to spatial correlation, and with a large number of model variables, increasing the dimension of the panel data, thus making these unit root tests that consider cross-sectional dependence unsuitable for this study. Baltagi, Bresson et al. ( 2007 ) suggested using Pesaran ( 2007 ) CIPS test because it eliminates the influence of spatial and cross-sectional dependence. They compared this method with others using simulations and real data analysis. They simulated balanced or unbalanced panel data with or without unit roots and spatial dependence, and then performed various panel unit root tests on these datasets to calculate the statistical efficiency and rejection rates. They also applied these tests to actual panel data, including US state house price indices, European consumer price indices, and exchange rates, and compared the results and implications. The authors found that when spatial dependence is present, commonly used unit root tests like Levin-Lin-Chu (2002), Im-Pesaran-Shin (2003), Maddala and Wu (1999), Choi (2001), etc., have reduced statistical power and may produce incorrect conclusions. Pesaran ( 2007 ) CIPS test, on the other hand, does not have this issue and exhibits higher statistical efficiency. Considering the presence of spatial correlation in the data, and using the Dynamic Spatial Difference-in-Differences model for econometric analysis, the panel data suffer from cross-sectional dependence, causing biases in the results of the Levin-Lin-Chu (LLC), Im-Pesaran-Shin (IPS), Fisher-type, and Hadri tests. Moreover, the spatial correlation introduces a more complex form of cross-sectional dependence and a larger number of model variables, which do not align with the assumptions of panel unit root tests used for cross-sectional dependence. In line with the recommendation of Baltagi, Bresson et al. ( 2007 ), we utilize Pesaran ( 2007 ) CIPS test as the unit root test method for the data in this study. 3. Analysis Results 3.1 Descriptive Statistics for the Panel In this section, we present the descriptive statistics for the variables used in the econometric model. Descriptive statistics help the audience understand the data and assess the impact of its estimates (Song, Qiu et al. 2023 ). The descriptive statistics for the panel variables are shown in Table 2 . All continuous control variables are log-transformed. Table 2 Descriptive statistics Variable Mean Std. dev. Min Max Skewness Kurtosis lntourist 12.80 2.58 -11.51 16.15 -7.16 67.40 Wlntourist 14.59 0.93 4.03 16.84 -1.52 10.08 lnpop 16.95 0.02 16.92 16.97 -0.25 1.88 lnincome 10.68 0.18 10.52 11.39 2.42 8.33 Gasoline_Index 84.10 16.91 56.42 112.15 0.03 1.74 SARS 0.03 0.16 0.00 1.00 5.85 35.23 Financial Crisis 0.10 0.30 0.00 1.00 2.63 7.90 THSR 0.33 0.47 0.00 1.00 0.75 1.56 TSHU 0.68 1.09 0.00 5.00 1.46 4.16 3.2 Unit Root Analysis Our data exhibit spatial dependence and cross-sectional dependence characteristics. Following the recommendation of Baltagi, Bresson et al. ( 2007 ), we use the CIPS test proposed by Pesaran ( 2007 ) to test for unit roots. The results of all variable tests reject the null hypothesis of unit roots (Table 3 ). In addition to the CIPS test results, we also provide the results of the Breitung test, Fisher-type test, and IPS test for comparison. Table 3 Unit root test (at level) Variable CIPS test Breitung test Fisher-type IPS test lntourist -4.961*** -2.911*** 23.216*** -10.627*** Wlntourist -5.705*** -3.954*** 69.778*** -5.064*** lnpop 2.610** 4.990 -4.358 -7.7902*** lnincome 2.610** -8.453*** 152.7517*** -10.627*** Gasoline_index 2.610** -0.635 7.3613*** -1.498 TSHU -2.330*** 0.758 8.652*** -0.314 The symbols *, **, and *** refer to level of significance at 10%, 5%, and 1% respectively The unit root test results for each variable are inconsistent across the four methods. For example, all four methods reject the presence of unit roots in "lntourist"; both CIPS test and IPS test reject the unit root hypothesis for "lnpop," while Breitung test and Fisher-type test do not; CIPS test and Fisher-type test reject the unit root hypothesis for "Shuttle_Routes," while Breitung test and IPS test do not. Different tests yield different results and may influence the analysis method. As the CIPS test is suitable for data with spatial dependence and cross-sectional dependence, which is the case for our data, we adopt the results of the CIPS test and reject the presence of unit roots for all variables. 3.3 Analysis of Dynamic SDID Estimation Results This paper mainly applies Dynamic SDID to analyze the effect of the High-speed rail and tourist transit service on Tourism Demand. To evaluate whether Dynamic SDID outperforms the traditional DID and the use of DUMMY variables to represent the presence or absence of the High-speed rail, we also present the results of traditional DID and DUMMY analysis. The analysis models for traditional DID are DID model 1 (High-speed rail) and DID model 2 (High-speed rail and tourist transit service), and for Dynamic SDID are Dynamic SDID (High-speed rail and tourist transit service) (Table 4 ). The analysis models using DUMMY variables are DUMMY model 1 (High-speed rail), DUMMY model 2 (High-speed rail and tourist transit service), and DUMMY model 3 (High-speed rail and tourist transit service with controlling spillover effect) (Table 5 ). When not controlling for spillover effect, there is no clear advantage between the DID and DUMMY models; when controlling for spillover effect, the DID model performs better than the DID model without controlling for spillover effect; the DUMMY model with spillover effect control performs better than the DUMMY model without spillover effect control; the DID model with spillover effect control outperforms the DUMMY model with spillover effect control (Table 4 , Table 5 ). Therefore, controlling for spillover effect is crucial for analyzing the effect of the High-speed rail and tourist transit service on Tourism Demand, and after controlling for spillover effect, the DID model outperforms the DUMMY model. Consequently, the following analysis will discuss the results based on the DID model 3 with spillover effect control (i.e., Dynamic SDID). Table 4 Results of models using DID for HSR Variables DID model 1 DID model 2 DID model 3 Coefficient Std. err. Coefficient Std. err. Coefficient Std. err. Laglntourist 0.748*** 0.168 0.747*** 0.168 0.839*** 0.12 Wlntourist 0.565*** 0.031 LagWlntourist -0.038 0.032 lnNpop 29.764*** 5.638 28.529*** 6.788 11.94 8.47 lnincome 0.265*** 0.065 0.262*** 0.066 0.141** 0.057 Gasoline_Index 0.002 0.003 0.003 0.003 0.003 0.003 SARS -0.173 0.129 -0.173 0.129 -0.048 0.112 Financial_Crisis 0.023 0.111 0.023 0.112 0.011 0.105 treatTHSR -0.49 0.92 -0.509 0.926 -0.213 0.95 postTHSR -0.580*** 0.215 -0.493** 0.219 0.092 0.313 THSRDID 0.993*** 0.211 0.891*** 0.216 0.608** 0.248 TSHU -0.076 0.068 -0.063 0.076 TSHUTHSR 0.185** 0.088 0.236** 0.099 WTSHU 0.005 0.021 WTHSR -0.227* 0.115 _cons -124.805*** 24.078 -119.650*** 29.029 -32.009 23.047 Wald Test 181.65 186.339 549.722 F-Test 20.183 16.94 36.648 Root MSE (Sigma) 379.699 364.012 178.112 LLF -4489.57 -4503.58 -4388.5 AIC 0.758 0.765 0.718 SC 0.771 0.781 0.738 Table 5 Results of models using dummy variable for HSR Variables DUMMY model 1 DUMMY model 2 DUMMY model 3 Coefficient Std. err. Coefficient Std. err. Coefficient Std. err. Laglntourist 0.745*** 0.168 0.744 0.169*** 0.842*** 0.118 Wlntourist 0.565*** 0.031 LagWlntourist -0.036 0.031 lnNpop 21.159*** 4.648 19.251*** 5.915 13.111 8.042 lnincome 0.259*** 0.066 0.256*** 0.066 0.143** 0.056 Gasoline_Index 0.001 0.003 0.002 0.003 0.003 0.003 SARS -0.168 0.129 -0.17 0.129 -0.048 0.111 Financial_Crisis -0.035 0.108 -0.019 0.11 0.01 0.104 THSRDUMMY 0.933*** 0.153 0.887*** 0.163 0.816*** 0.201 TSHU -0.008 0.07 -0.019 0.077 TSHU THSRDUMMY 0.103 0.09 0.164 0.102 WTSHU 0.000 0.021 WTHSR -0.179** 0.075 _cons -89.010*** 20.043 -81.007** 25.597 -34.68 21.536 Wald Test 197.362 199.563 560.665 F-Test 28.195 22.174 43.128 Root MSE (Sigma) 269.801 245.538 195.481 LLF -4494.77 -4506.22 -4419.08 AIC 0.759 0.765 0.729 SC 0.77 0.778 0.747 The results of Dynamic SDID analysis show that the current tourism demand ("Wlntourist") exhibits significant positive spatial autocorrelation, and the spatial autocorrelation of the lagged tourism demand ("lagWlntourist") is negative, which aligns with the estimates of Liu ( 2020 ). The study's use of significant and commonly used explanatory variables for tourism demand is consistent with past research. The lagged tourism demand and income show significant positive impacts, while the effects of population, travel cost, and the 2008 financial crisis are positive but not significantly different from zero, and the impact of SARS is negative but not significantly different from zero. Although the research results differ from past studies, this difference may arise from tourism's apparent "decline-rebound-overshoot" effect (Jeong and Lee 2022 , Kei Kuok, Koo et al. 2023 ). Major economic and environmental changes lead to significant fluctuations in tourism activities in the short term, making short-term estimates of these variables statistically significant. However, in the long term, these fluctuations tend to level out, leading to non-significant estimates. The impact of THSR on tourism demand is significantly positive (DID = 0.608**), which aligns with intuition. The significant negative spatial autocorrelation effect of THSR (WHSR=-0.227*) indicates that THSR has negative spillover effects on tourism activities. The results suggest that the influx of tourists brought by THSR primarily comes from existing tourists in the surrounding regions rather than creating new tourism demand. Thus, from a national perspective on the tourism industry, THSR does not create new demand, which supports the conclusions of Delaplace, Pagliara et al. ( 2014 ), Kurihara and Wu ( 2016 ), Pagliara, La Pietra et al. ( 2015 ), and Yu and Fan ( 2018 ). The impact of TSHU on tourism demand is negative but not significant (TSHU=-0.063), which contradicts intuition. However, upon further investigation, this result is not surprising. The effect of TSHU on tourism growth depends on both its positive contribution and the tourism market trend in the destinations it connects to. If the tourism market in the destinations traversed by TSHU is in decline, TSHU's positive contribution may not be sufficient to reverse the decline, but it can slow down the pace of tourism decline. When the effects of TSHU and THSR on tourism activities are considered, the spatial autocorrelation effect of TSHU is positive but not significant (WSHU = 0.005), indicating a positive spillover effect of TSHU on tourism activities, implying that although TSHU may not significantly boost tourism development in specific and individual travel destinations, it contributes to the overall tourism development in the region. The interaction effect of TSHU and THSR is significantly positive (TSHUTHSR = 0.236**), indicating a complementary relationship between the two. When tourists use THSR and do not use personal transportation, the convenience of connecting transport depends on shuttle services, and providing shuttle buses to improve transport convenience contributes to promoting tourism activities. The estimated results of TSHUTHSR support the aforementioned argument. 4. Conclusion and Policy Recommendations This paper uses the Dynamic Spatial Difference-in-Differences (Dynamic SDID) model to analyze the effect of High-speed rail (HSR) and tourist transit service (TTSBS) on Tourism Demand. The transportation construction variables considered are THSR and TSHU. Since TSHU mainly serves as a connection between tourist attractions, we include the interaction effect between TSHU and THSR. Additionally, to verify whether the increased tourism from transportation construction comes from new tourism demand created by the construction or from substituting existing demand, we introduce spatial autocorrelation variables for THSR and TSHU. Considering the existence of spatial dependence and cross-sectional dependence in the data, we use the CIPS test for unit root testing. The research findings show that THSR and TSHU have different impacts on tourism. THSR has a positive impact on tourism demand, while TSHU has a negative impact. The interaction effect between TSHU and THSR is positive. The negative spatial autocorrelation effect of THSR indicates that it has negative spillover effects on tourism activities. This result suggests that the influx of tourists brought by THSR primarily comes from existing tourists in the surrounding regions rather than creating new tourism demand. Therefore, from a national perspective on the tourism industry, THSR does not create new demand and may indicate the presence of substitution between tourism attractions. This finding supports the conclusions of Delaplace, Pagliara et al. ( 2014 ), Kurihara and Wu ( 2016 ), Pagliara, La Pietra et al. ( 2015 ), and Yu and Fan ( 2018 ), indicating that HSR does not significantly contribute to promoting tourism. Therefore, to boost national tourism demand, the focus should be on enhancing the attractiveness of tourist destinations rather than solely relying on HSR construction. TSHU has a non-significant negative impact on tourism, and the reason lies in the criteria for selecting TSHU routes, such as selecting attractive tourist destinations and collaborating with various transportation modes, which may not have been fully implemented, or the data used for evaluation may not align with the actual conditions of the tourism market. Thus, TSHU should focus on linking important tourist attractions. Evaluating TSHU routes using data that reflects the actual conditions of the tourism market can improve transportation convenience and promote the development of the tourism industry. Previous literature mainly explored the impact of a single HSR on tourism demand (Yu and Fan 2018 , Gao, Su et al. 2019 , Deng, Gan et al. 2021 ) without considering the interaction effects between TTSBS and HSR or the spillover effects, leading to serious biases, such as overlooking important explanatory variables or using incorrect functions. Moreover, these studies did not consider the spatial correlation between tourism trips and transportation construction, leading to biased estimation results. In contrast, this study simultaneously considers THSR and TSHU, controls for the spatial correlation between tourism trips and transportation construction, and takes into account the interaction effect between THSR and TSHU, thereby excluding known factors that could cause biased estimates, resulting in more reliable and unbiased estimation results. The empirical and policy contributions of this study lie in demonstrating the interaction effects, spatial effects, and uncertainties of HSR and TTSBS on tourism demand. For HSR, this study shows that it can increase tourism demand in the destinations where HSR stations are located but may reduce tourism demand in surrounding areas. TTSBS can further enhance the impact of HSR on tourism demand. Policy measures, such as package tours and interregional transportation connections, should be implemented to mitigate the impact of HSR on surrounding tourist destinations. As for TTSBS, it does not have a significant impact on Taiwan's tourism demand, which may be due to the selection of TTSBS routes. The theoretical contribution of this study lies in simultaneously exploring the impacts of HSR and TTSBS on tourism, not limited to each HSR and TTSBS individually but also considering their interaction effects and spatial externalities. In terms of modeling, this study uses the Dynamic Spatial Difference-in-Differences model, which better suits the theoretical requirements of analyzing the causal relationship between HSR and tourism demand. The empirical results show that Dynamic SDID outperforms DID or DUMMY models, supporting Albalate and Fageda ( 2016 ) argument that differences in research results regarding the impact of HSR on tourism are due to different econometric methods and the network design and accessibility of HSR, providing a reference for future research in model settings. The use of CIPS test for unit root testing yields more consistent estimation results, making it a valuable reference for subsequent tourism studies. Declarations Author Contribution T.L. wrote the main manuscript text. Acknowledgments This work was supported by the National Science and Technology Council, Taiwan, under Grant MOST 110-2410-H-110 -073 -programs. Data Availability Data derived from public domain resources. References Albalate, D., J. Campos and J. L. Jiménez (2017). "Tourism and high speed rail in Spain: Does the AVE increase local visitors?" Annals of Tourism Research 65 : 71-82. Albalate, D. and X. Fageda (2016). "High speed rail and tourism: Empirical evidence from Spain." Transportation Research Part A: Policy and Practice 85 : 174-185. Baltagi, B. H., G. Bresson and A. Pirotte (2007). "Panel unit root tests and spatial dependence." Journal of Applied Econometrics 22 (2): 339-360. Bazin, S., C. Beckerich and M. Delaplace (2006). "Analyse prospective des impacts de la Ligne Grande Vitesse Est-Européenne dans l’agglomération rémoise et en région Champagne-Ardenne." Rapport final pour le Conseil Régional Champagne-Ardenne 495 . Becker, G. S. and G. S. Becker (2009). Accounting for tastes , Harvard University Press. Becker, G. S. and K. M. Murphy (2009). Social economics: Market behavior in a social environment , Harvard University Press. Currie, C. and P. Falconer (2014). "Maintaining sustainable island destinations in Scotland: The role of the transport–tourism relationship." Journal of Destination Marketing & Management 3 (3): 162-172. de Oliveira Santos, G. E., V. Ramos and J. Rey-Maquieira (2011). "A microeconomic model of multidestination tourism trips." Tourism Economics 17 (3): 509-529. Dehghan Shabani, Z. and S. Safaie (2018). "Do transport infrastructure spillovers matter for economic growth? Evidence on road and railway transport infrastructure in Iranian provinces." Regional science policy & practice 10 (1): 49-63. Delaplace, M., F. Pagliara, J. Perrin and S. Mermet (2014). "Can High Speed Rail foster the choice of destination for tourism purpose?" Procedia-Social and Behavioral Sciences 111 : 166-175. Deng, T., C. Gan, H. Du, Y. Hu and D. Wang (2021). "Do high speed rail configurations matter to tourist arrivals? Empirical evidence from China's prefecture-level cities." Research in Transportation Economics 90 : 100952. Di Matteo, D. (2022). "Does high-speed rail matter for tourism? Evidence from Italy." Research in Transportation Business & Management : 100881. Fang, J. (2021). "Impacts of high-speed rail on urban smog pollution in China: A spatial difference-in-difference approach." Science of The Total Environment 777 : 146153. Gao, Y., W. Su and K. Wang (2019). "Does high-speed rail boost tourism growth? New evidence from China." Tourism Management 72 : 220-231. González‐Val, R. and J. Silvestre (2023). "War and city size: The asymmetric effects of the Spanish Civil War." Journal of Regional Science . Gooroochurn, N. and A. Hanley (2005). "Spillover effects in long‐haul visitors between two regions." Regional Studies 39 (6): 727-738. Hwang, Y.-H. and D. R. Fesenmaier (2003). "Multidestination pleasure travel patterns: empirical evidence from the American Travel Survey." Journal of Travel Research 42 (2): 166-171. Jeong, J. Y. and K.-Y. Lee (2022). "A rebound in nature-based tourism intentions during the COVID-19 era." Journal of Vacation Marketing : 13567667221111566. Kanwal, S., M. I. Rasheed, A. H. Pitafi, A. Pitafi and M. Ren (2020). "Road and transport infrastructure development and community support for tourism: The role of perceived benefits, and community satisfaction." Tourism Management 77 : 104014. Kei Kuok, R. U., T. T. Koo and C. Lim (2023). "Economic policy uncertainty and international tourism demand: a global vector autoregressive approach." Journal of Travel Research 62 (3): 540-562. Khadaroo, J. and B. Seetanah (2007). "Transport infrastructure and tourism development." Annals of tourism research 34 (4): 1021-1032. Kurihara, T. and L. Wu (2016). "The impact of high speed rail on tourism development: A case study of Japan." The Open Transportation Journal 10 (1). Lancaster, K. (1966). "Change and innovation in the technology of consumption." The American Economic Review 56 (1/2): 14-23. Li, J. and S. Li (2020). "Energy investment, economic growth and carbon emissions in China—Empirical analysis based on spatial Durbin model." Energy Policy 140 : 111425. Lim, C. (2006). A survey of tourism demand modeling practice: issues and implications. International Handbook on the Economics of Tourism . L. Dwyer and P. Forsyth. Northampton, Edward Elgar Publishing : 45–72. Liu, T.-M. (2020). "Habit formation or word of mouth: What does lagged dependent variable in tourism demand models imply?" Tourism Economics 26 (3): 461-474. Masson, S. and R. Petiot (2009). "Can the high speed rail reinforce tourism attractiveness? The case of the high speed rail between Perpignan (France) and Barcelona (Spain)." Technovation 29 (9): 611-617. Pagliara, F., A. La Pietra, J. Gomez and J. M. Vassallo (2015). "High Speed Rail and the tourism market: Evidence from the Madrid case study." Transport Policy 37 : 187-194. Pan, H., C. Cong, X. Zhang and Y. Zhang (2020). "How do high-speed rail projects affect the agglomeration in cities and regions?" Transportation Research Part D: Transport and Environment 88 : 102561. Peeters, P. and F. Schouten (2006). "Reducing the ecological footprint of inbound tourism and transport to Amsterdam." Journal of sustainable tourism 14 (2): 157-171. Pesaran, M. H. (2007). "A simple panel unit root test in the presence of cross-section dependence." Journal of Applied Econometrics 22 (2): 265-312. Prideaux, B. (2018). "Tourism and surface transport." The Sage Handbook of Tourism Management. London: Sage : 297-313. Song, H., R. T. Qiu and J. Park (2023). "Progress in tourism demand research: Theory and empirics." Tourism Management 94 : 104655. Song, H. and S. F. Witt (2006). "Forecasting international tourist flows to Macau." Tourism Management 27 (2): 214-224. Song, H., S. F. Witt and G. Li (2009). The Advanced Econometrics of Tourism Demand . New York, Routledge. Tian, F., Y. Yang and L. Jiang (2022). "Spatial spillover of transport improvement on tourism growth." Tourism Economics 28 (5): 1416-1432. Udeagha, M. C. and E. Muchapondwa (2023). "Striving for the United Nations (UN) sustainable development goals (SDGs) in BRICS economies: The role of green finance, fintech, and natural resource rent." Sustainable Development . Virkar, A. R. and P. D. Mallya (2018). "A conceptual paper on the dimensions of the road transport system that influence tourist satisfaction." International Journal of Research in Social Sciences 8 (3): 296-303. Wang, F., L. Ma, B. Gao and Y. S. Liu (2023). "Does high-speed railway affect the cost behavior of tourism firms? Evidence from China." Tourism Economics : 13548166221150698. Yang, L., L. Hu, P. Shang and Y. Ye (2021). "Estimating the Impacts of High-speed Rail on Service Industry Agglomeration in China: Advanced Modelling with Spatial Difference-in-difference Models and Propensity Score Matching." Journal of Transport Economics and Policy (JTEP) 55 (1): 16-35. Yu, L. (2021). "Study on treatment effects and spatial spillover effects of Beijing–Shanghai HSR on the cities along the line." The Annals of Regional Science 67 (3): 671-695. Yu, M. and W. Fan (2018). "Accessibility impact of future high speed rail corridor on the piedmont Atlantic megaregion." Journal of Transport Geography 73 : 1-12. Additional Declarations No competing interests reported. Cite Share Download PDF Status: Published Journal Publication published 19 Dec, 2024 Read the published version in Transportation → Version 1 posted Editorial decision: Revision requested 08 Nov, 2024 Reviews received at journal 05 Nov, 2024 Reviews received at journal 18 Oct, 2024 Reviewers agreed at journal 11 Oct, 2024 Reviewers agreed at journal 10 Oct, 2024 Reviewers invited by journal 09 Oct, 2024 Editor assigned by journal 09 Oct, 2024 Submission checks completed at journal 08 Oct, 2024 First submitted to journal 07 Oct, 2024 You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-5219377","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":375852429,"identity":"708c3d79-6f3c-42d8-855a-ecd0c8d12ffb","order_by":0,"name":"Tzu-Ming Liu","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAAvklEQVRIiWNgGAWjYBACPmYogx9EJBQQoYWNmYGxgYHBgEGyAaTFgBgtDFAtBgdAXKK0sPOYP7rB8Cdx8/nViR8eGDDI84sdIOQwHsPmHAaDxG033m6WADrMcObsBOK05G67cXYDSEuCwW1itWyecXbzD9K0bODv3UasLWyFs3MMjOtn3ODdZpFgIEHYL/z8hzd8zqmQM+bvP7v55o8KG3l+aQJaIAAUHRJglRLEKIfbd4AU1aNgFIyCUTCSAAA+hTxh/WydagAAAABJRU5ErkJggg==","orcid":"","institution":"Institute of Marine Affairs National Sun Yat-sen University","correspondingAuthor":true,"prefix":"","firstName":"Tzu-Ming","middleName":"","lastName":"Liu","suffix":""}],"badges":[],"createdAt":"2024-10-07 15:38:25","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-5219377/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-5219377/v1","draftVersion":[],"editorialEvents":[{"content":"https://doi.org/10.1007/s11116-024-10576-4","type":"published","date":"2024-12-19T15:57:57+00:00"}],"editorialNote":"","failedWorkflow":false,"files":[{"id":69209304,"identity":"b486c7a9-b599-4d31-9ffa-5b03eaa165e7","added_by":"auto","created_at":"2024-11-18 04:50:15","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":99289,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eGraphical Illustration\u003c/strong\u003e of Difference-in-Difference (DID) estimation: In the absence of treatment, the treated should observe the same trend as the control group. If this assumption holds, we can then attribute the change in the trend of the treated, in the presence of treatment, to the treatment itself.\u003c/p\u003e","description":"","filename":"floatimage1.png","url":"https://assets-eu.researchsquare.com/files/rs-5219377/v1/8093870b29503a74683bf931.png"},{"id":72202530,"identity":"dbd409a0-e8ef-4f8e-a8ad-6021a53b5eb2","added_by":"auto","created_at":"2024-12-23 16:14:46","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":849311,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-5219377/v1/cd4357c3-9398-46f1-851e-d847c1e69b96.pdf"}],"financialInterests":"No competing interests reported.","formattedTitle":"Using a Dynamic Spatial Difference-in-Differences estimator to evaluate the effect of High speed rail and tourist transit service on Tourism Demand","fulltext":[{"header":"1. Introduction","content":"\u003cp\u003eTransportation is an essential key element in the development of tourism. Improving transportation infrastructure can offer tourists safer, more comfortable, and efficient means of travel, attracting more visitors (Khadaroo and Seetanah \u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e2007\u003c/span\u003e). Therefore, in theory, enhancing transportation infrastructure is expected to increase market accessibility, reduce travel costs, and make tourist destinations more attractive. However, the research on high-speed rail (HSR) does not completely support this argument. Hence, while France, Germany, Italy, Spain, Japan, and Taiwan have built HSR systems, countries like the U.K., United States, India, and Malaysia have faced skepticism in their planning and construction of HSR. Given the significant cost of HSR construction, understanding its actual impact on tourism and regional development is crucial to optimizing tourism resource allocation. This includes identifying whether the effects result from the characteristics of HSR itself or from issues in data analysis.\u003c/p\u003e \u003cp\u003eOver the past two decades, many countries have constructed HSR as a significant means to boost tourism demand (Albalate and Fageda \u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2016\u003c/span\u003e). Some studies show that HSR promotes tourism demand (Deng, Gan et al. \u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e2021\u003c/span\u003e), while others indicate it boosts tourism growth but reduces tourism revenue (Gao, Su et al. \u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e2019\u003c/span\u003e), or that the effect is only significant in the first year after HSR introduction and diminishes in subsequent years (Kurihara and Wu \u003cspan citationid=\"CR22\" class=\"CitationRef\"\u003e2016\u003c/span\u003e). Studies from the U.S. (Yu and Fan \u003cspan citationid=\"CR42\" class=\"CitationRef\"\u003e2018\u003c/span\u003e), Spain (Pagliara, La Pietra et al. \u003cspan citationid=\"CR28\" class=\"CitationRef\"\u003e2015\u003c/span\u003e), and France (Delaplace, Pagliara et al. \u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e2014\u003c/span\u003e) reveal that only a few cities experienced an increase in tourists after the introduction of HSR. Bazin, Beckerich et al. (\u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e2006\u003c/span\u003e) argue that HSR may fail to ignite tourists' curiosity and have limited impact on leisure tourism. Albalate and Fageda (\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2016\u003c/span\u003e) attribute the differing research results to variations in econometric methods and the network design and accessibility of HSR, which relate to the spillover effects of tourism and HSR and the connections between them.\u003c/p\u003e \u003cp\u003eTourism spillover effects refer to the impact of tourism activities in one area on nearby areas (Liu \u003cspan citationid=\"CR26\" class=\"CitationRef\"\u003e2020\u003c/span\u003e). Liu (\u003cspan citationid=\"CR26\" class=\"CitationRef\"\u003e2020\u003c/span\u003e) uses Becker's social capital theory (Becker and Murphy \u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e2009\u003c/span\u003e) and tastes (Becker and Becker \u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e2009\u003c/span\u003e) as the theoretical basis for exploring tourism spillover effects. Multidestination travel (de Oliveira Santos, Ramos et al. \u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e2011\u003c/span\u003e) is also a source of tourism spillover effects. Multidestination travel, based on Lancaster's (1966) characteristics theory (Lancaster \u003cspan citationid=\"CR23\" class=\"CitationRef\"\u003e1966\u003c/span\u003e), involves visiting more than one destination in a single trip, allowing tourists to maximize utility under cost and time constraints (Hwang and Fesenmaier \u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e2003\u003c/span\u003e). Improving transportation infrastructure reduces the cost of traveling to different destinations, leading to more characteristics (destinations) included in tourists' travel decisions, thus increasing the number and possibilities of visiting multiple destinations.\u003c/p\u003e \u003cp\u003eSpillover effects are not just theoretical predictions but have empirical support. Empirical studies show that spillover effects are essential explanatory variables in tourism demand models. For instance, tourists can take HSR to a central city in a region and then transfer to road transportation to reach their final destination. As a result, there will be transport spillover effects from the hub city to the destination, contributing to tourism spatial agglomeration (Masson and Petiot \u003cspan citationid=\"CR27\" class=\"CitationRef\"\u003e2009\u003c/span\u003e). Improved transportation to specific tourist destinations creates more opportunities for joint promotion or marketing between that destination and neighboring ones, enhancing the overall tourism competitiveness of both the destination and surrounding areas, thus generating spillover effects of transportation construction on tourism (Gooroochurn and Hanley \u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e2005\u003c/span\u003e, Dehghan Shabani and Safaie \u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e2018\u003c/span\u003e). However, HSR does not always produce positive spillover effects on tourism and may also have negative effects. The improved accessibility brought by HSR benefits the cities where stations are located due to their better transport resources, services, and facilities, which attract more tourists and also draw tourists from neighboring cities. The substantial demand for tourism generated by HSR further concentrates tourism resources, providing more specialized tourism services and attracting more tourists, intensifying the competition in the local tourism market and suppressing the tourism development of surrounding destinations. Therefore, HSR creates spillover effects, but the net impact of these effects is not universally applicable.\u003c/p\u003e \u003cp\u003eHSR spillover effects are related to the aforementioned tourist transportation services. Tourist transportation services can help transport HSR-generated tourists to surrounding areas, resulting in positive spillover effects. On the other hand, they can also redirect tourists from surrounding destinations to the HSR destination, reducing tourism activities in those areas and leading to negative spillover effects. Public road transportation includes cars, buses, and coaches. Buses and long-distance coaches are generally operated by public transport networks or tourism agencies. An excellent road transportation system acts as a catalyst for the growth of the tourism industry. Many studies have confirmed the positive impact of road transportation on tourism development (Kanwal, Rasheed et al. \u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e2020\u003c/span\u003e). Road transportation provides greater accessibility to tourist destinations (Masson and Petiot \u003cspan citationid=\"CR27\" class=\"CitationRef\"\u003e2009\u003c/span\u003e), encourages local business activities (Khadaroo and Seetanah \u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e2007\u003c/span\u003e), attracts tourists, and promotes new tourism destinations (Currie and Falconer \u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e2014\u003c/span\u003e, Virkar and Mallya \u003cspan citationid=\"CR38\" class=\"CitationRef\"\u003e2018\u003c/span\u003e). Road transportation's advantages include convenience, flexibility, affordability, ease of access to tourist destinations, fewer baggage restrictions, and greater agency in personal travel experiences (Virkar and Mallya \u003cspan citationid=\"CR38\" class=\"CitationRef\"\u003e2018\u003c/span\u003e, Kanwal, Rasheed et al. \u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e2020\u003c/span\u003e). However, there has been less attention in the academic community to tourist transportation shuttle bus services (TTSBS) compared to other forms of transportation (Peeters and Schouten \u003cspan citationid=\"CR30\" class=\"CitationRef\"\u003e2006\u003c/span\u003e, Prideaux \u003cspan citationid=\"CR32\" class=\"CitationRef\"\u003e2018\u003c/span\u003e), and their interaction with HSR's impact on tourism is not well-studied, thus requiring further investigation.\u003c/p\u003e \u003cp\u003eWhile DID and panel data techniques have been used in some papers to study the impact of HSR on tourism development (Albalate, Campos et al. \u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e2017\u003c/span\u003e, Fang \u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e2021\u003c/span\u003e, Di Matteo \u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e2022\u003c/span\u003e, Wang, Ma et al. \u003cspan citationid=\"CR39\" class=\"CitationRef\"\u003e2023\u003c/span\u003e), there are still shortcomings in these studies. Some of the limitations include insufficient consideration of HSR's spillover effects on tourism development and the neglect of the impact of transferring to road transportation. Only a few studies have combined spatial econometric methods and DID in Spatial Difference-in-Differences (SDID) analysis of HSR's impact on tourism (Yang, Hu et al. \u003cspan citationid=\"CR40\" class=\"CitationRef\"\u003e2021\u003c/span\u003e). Yang, Hu et al. (\u003cspan citationid=\"CR40\" class=\"CitationRef\"\u003e2021\u003c/span\u003e) use the spatial autoregressive lag (SAR) and spatial error term (SE) to represent HSR's spillover effects. Yu (\u003cspan citationid=\"CR41\" class=\"CitationRef\"\u003e2021\u003c/span\u003e) uses the spatial autoregressive lag (SAR) and spatial DID term to capture the spillover effect. The former, while SDID, only includes a simple dependent variable spatial lag term, and since the paper does not use the spatial weighting term of HSR, it cannot control for the spillover effect of the high-speed rail. The latter uses the spatial DID term but does not consider TTSBS or the spatial autocorrelation of the dependent variable lag. Both studies have omitted essential explanatory variables, leading to possible estimation biases.\u003c/p\u003e \u003cp\u003eConsidering the importance of HSR's impact on tourism industry development and the advantages of the DID method in policy analysis, this study addresses the previously mentioned research gaps and investigates the impact of TTSBS on HSR's effect on tourism activities. By supplementing important control variables not adequately explored in previous literature, this study aims to avoid omitted variable biases. To achieve this research goal, this study employs the SDID analysis with a dependent variable lag term to examine HSR's impact on regional tourism.\u003c/p\u003e \u003cp\u003eThe structure of this paper is as follows: The introduction presents the controversy surrounding HSR's impact on tourism development, proposes statistical analysis methods, and discusses the potential effects of omitting important control variables on the previous controversy, making a case for the rationality of using SDID as the analytical method. The second section introduces the research methodology, providing an overview of the traditional DID model and extending it to the Dynamic SDID. It also explains the data sources for the variables. The third section presents the results of the statistical analysis. The fourth section provides conclusions and policy recommendations.\u003c/p\u003e"},{"header":"2. Methodology","content":"\u003cdiv id=\"Sec3\" class=\"Section2\"\u003e \u003ch2\u003e2.1 Introduction of Dynamic Spatial Difference-in-Difference model\u003c/h2\u003e \u003cp\u003eThe Difference-in-Differences (DID) method is commonly used to estimate the effect of a specific policy (treatment) by comparing the changes in the outcome of the group affected by the policy (treatment group) and the group unaffected by the planned intervention (control group) over a specific period. As outcomes may be influenced by external factors and may change over time, simply observing the simple changes in outcomes before and after the treatment is not sufficient to draw causal conclusions, as factors other than the treatment might influence the results over time. Additionally, comparing participating and non-participating groups alone may lead to selection bias and differences in unobservable characteristics between the groups. DID combines these two approaches by comparing the differences in outcome changes between the treatment and control groups before and after the treatment, estimating the overall impact of the program. DID uses data from before and after the policy intervention, such as group or panel data (individual-level data that varies over time) or repeated cross-sectional data (individual or group-level data). This method eliminates biases caused by inter-group permanent differences in post-intervention comparisons and biases from trends in the intervention group caused by other outcome factors over time. Therefore, even without random selection of the affected sample, causal inferences can be made.\u003c/p\u003e \u003cp\u003eDID calculates the before-and-after difference in outcomes for the treatment group, which is the first difference (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e, Table\u0026nbsp;\u003cspan refid=\"Tab1\" class=\"InternalRef\"\u003e1\u003c/span\u003e). By comparing the same group with itself, the first difference can control for factors that remain constant within the group over time. Then, to capture time-varying factors, the difference in differences calculates the before-and-after difference for the control group, which is affected by the same environmental conditions as the treatment group but not by the policy. This is the second difference. Finally, DID \"cleans out\" all time-varying factors by subtracting the second difference from the first difference. This gives us the estimated effect, namely the Difference-in-Difference.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab1\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 1\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003eDID model coefficient relationship in regression\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"4\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003eControl Group (i\u0026thinsp;=\u0026thinsp;0) Without THSR Stations\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c3\"\u003e \u003cp\u003eTreatment Group (i\u0026thinsp;=\u0026thinsp;1) With THSR Stations\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c4\"\u003e \u003cp\u003eDifference (1)\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003ePre-intervention (t\u0026thinsp;=\u0026thinsp;0)(Before THSR operation)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{\\alpha\\:}_{0}\\)\u003c/span\u003e\u003c/span\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{\\alpha\\:}_{0}{+\\alpha\\:}_{1}\\)\u003c/span\u003e\u003c/span\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{\\alpha\\:}_{1}\\)\u003c/span\u003e\u003c/span\u003e\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003ePost-intervention (t\u0026thinsp;=\u0026thinsp;1)(After THSR operation)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{\\alpha\\:}_{0}{+\\alpha\\:}_{2}\\)\u003c/span\u003e\u003c/span\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{\\alpha\\:}_{0}{+\\alpha\\:}_{1}+{\\alpha\\:}_{2}{+\\alpha\\:}_{3}\\)\u003c/span\u003e\u003c/span\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{\\alpha\\:}_{1}{+\\alpha\\:}_{3}\\)\u003c/span\u003e\u003c/span\u003e\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eDifference(2)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{\\alpha\\:}_{2}\\)\u003c/span\u003e\u003c/span\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{\\alpha\\:}_{2}{+\\alpha\\:}_{3}\\)\u003c/span\u003e\u003c/span\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{\\alpha\\:}_{3}\\)\u003c/span\u003e\u003c/span\u003e\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003ctfoot\u003e \u003ctr\u003e\u003ctd colspan=\"4\"\u003eFootnote: The simple DID estimator allows for the intercepts to vary between the treatment (\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{\\alpha\\:}_{0}{+\\alpha\\:}_{1}\\)\u003c/span\u003e\u003c/span\u003e) and the control group (\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{\\alpha\\:}_{0}\\)\u003c/span\u003e\u003c/span\u003e) and assumes constant outcomes within the two time periods (\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{\\alpha\\:}_{1}\\)\u003c/span\u003e\u003c/span\u003e). \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{\\alpha\\:}_{1}\\)\u003c/span\u003e\u003c/span\u003e: Average difference in \u003cem\u003eoutcome\u003c/em\u003e between the two groups that is common in both time periods. \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{\\alpha\\:}_{2}\\)\u003c/span\u003e\u003c/span\u003e: Average change in \u003cem\u003eoutcome\u003c/em\u003e from the Pre-intervention to the Post-intervention time period that is common to both groups.\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{\\alpha\\:}_{3}\\)\u003c/span\u003e\u003c/span\u003e: Average differential change in \u003cem\u003eoutcome\u003c/em\u003e from the first to the second time period of the treatment group relative to the control group\u003c/td\u003e\u003c/tr\u003e \u003c/tfoot\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003cp\u003eDID requires outcome data for the policy treatment group and control group, as well as data before and after policy implementation. The following steps calculate the difference in differences (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e):\u003c/p\u003e \u003cp\u003e \u003col\u003e \u003cspan\u003e \u003cli\u003e \u003cp\u003eCalculate the before-and-after difference in outcomes for the treatment group (B-A).\u003c/p\u003e \u003c/li\u003e \u003c/span\u003e \u003cspan\u003e \u003cli\u003e \u003cp\u003eCalculate the before-and-after difference in outcomes for the control group (D-C).\u003c/p\u003e \u003c/li\u003e \u003c/span\u003e \u003cspan\u003e \u003cli\u003e \u003cp\u003eCalculate the difference between the before-and-after differences in outcomes for the treatment group (B-A) and the control group (D-C). This is the DID: (B-A) - (D-C).\u003c/p\u003e \u003c/li\u003e \u003c/span\u003e \u003c/ol\u003e \u003c/p\u003e \u003cp\u003eFigure \u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e and Table\u0026nbsp;\u003cspan refid=\"Tab1\" class=\"InternalRef\"\u003e1\u003c/span\u003e also present the specific Stable Unit Treatment Value Assumption (SUTVA) of DID: Intervention unrelated to outcome at baseline, Treatment/intervention and control groups have Parallel Trends in outcome, Composition of intervention and comparison groups is stable for repeated cross-sectional design, No spillover effects.\u003c/p\u003e \u003cp\u003eThe traditional DID analysis of HSR policy impact can be represented using the following formula (Albalate and Fageda \u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2016\u003c/span\u003e, Albalate, Campos et al. \u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e2017\u003c/span\u003e).\u003cdiv id=\"Equ1\" class=\"Equation\"\u003e\u003cdiv format=\"TEX\" class=\"mathdisplay\" id=\"FileID_Equ1\" name=\"EquationSource\"\u003e\n$$\\:{y}_{i,t}={\\alpha\\:}_{0}+{\\alpha\\:}_{1}{HSR}_{i,t}+{\\alpha\\:}_{2}{Pos{t}_{construction}}_{i,t}+{\\alpha\\:}_{3}\\left({HSR}_{i,t}\\times\\:{Pos{t}_{construction}}_{i,t}\\right)+{X{\\prime\\:}}_{i,t}\\theta\\:+{ϵ}_{i,t}$$\u003c/div\u003e\u003cdiv class=\"EquationNumber\"\u003e1\u003c/div\u003e\u003c/div\u003e\u003c/p\u003e \u003cp\u003eIn Eq.\u0026nbsp;(\u003cspan refid=\"Equ1\" class=\"InternalRef\"\u003e1\u003c/span\u003e), \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{HSR}_{i,t}\\)\u003c/span\u003e\u003c/span\u003e and \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{Pos{t}_{construction}}_{i,t}\\)\u003c/span\u003e\u003c/span\u003e are dummy variables: when \u003cem\u003ei\u003c/em\u003e\u0026thinsp;=\u0026thinsp;1, it indicates the treatment group, \u003cem\u003ei\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0 indicates the control group; when t\u0026thinsp;=\u0026thinsp;1, it indicates after the experimental treatment, and \u003cem\u003et\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0 indicates before the experimental treatment. The regression coefficient relationship is summarized in Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e and Table\u0026nbsp;\u003cspan refid=\"Tab1\" class=\"InternalRef\"\u003e1\u003c/span\u003e: \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{\\alpha\\:}_{1}\\)\u003c/span\u003e\u003c/span\u003e, B-A, shows the difference between different groups at the same time point; \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{\\alpha\\:}_{2}\\)\u003c/span\u003e\u003c/span\u003e, D-C, shows the difference between the same groups at a different time point; \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{\\alpha\\:}_{3}\\)\u003c/span\u003e\u003c/span\u003e, (B-A)-(D-C), represents the true experimental treatment effect when both treatment and after treatment are 1; if there is no effect of the experimental treatment, \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{\\alpha\\:}_{3}\\)\u003c/span\u003e\u003c/span\u003e is 0. The variable \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{y}_{i,t}\\)\u003c/span\u003e\u003c/span\u003e indicates the policy outcome of \u003cem\u003eHSR\u003c/em\u003e, and \u003cb\u003eX\u003c/b\u003e denotes the control variables.\u003c/p\u003e \u003cp\u003eIn this study, the policy outcome, \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{y}_{i,t}\\)\u003c/span\u003e\u003c/span\u003e, represents tourism development, and we use the logarithm of tourist arrivals to characterize tourism development. Empirical evidence shows that tourist arrivals have significant spatial spillover effects (Liu \u003cspan citationid=\"CR26\" class=\"CitationRef\"\u003e2020\u003c/span\u003e), and therefore, the standard DID model (Eq.\u0026nbsp;\u003cspan refid=\"Equ1\" class=\"InternalRef\"\u003e1\u003c/span\u003e) cannot yield consistent estimation results because SUTVA does not hold (Li and Li \u003cspan citationid=\"CR24\" class=\"CitationRef\"\u003e2020\u003c/span\u003e, Pan, Cong et al. \u003cspan citationid=\"CR29\" class=\"CitationRef\"\u003e2020\u003c/span\u003e, Tian, Yang et al. \u003cspan citationid=\"CR36\" class=\"CitationRef\"\u003e2022\u003c/span\u003e). However, the spatial difference-in-differences (SDID) model can consider spatial effects (Chagas, Azzoni et al. 2016, Diao, Leonard et al. 2017, Ferman 2023). Accordingly, we construct the SDID model to accurately measure the influence of HSR on tourism.\u003c/p\u003e \u003cp\u003ePrevious studies have mainly considered the spillover effects of tourism arrivals, incorporating the spatial lag term of tourism development into the model (Liu \u003cspan citationid=\"CR26\" class=\"CitationRef\"\u003e2020\u003c/span\u003e), as well as the time lag term of the dependent variable. In line with these previous studies, this paper combines the spatial autocorrelation model (SAR) with the DID model (Eq.\u0026nbsp;\u003cspan refid=\"Equ1\" class=\"InternalRef\"\u003e1\u003c/span\u003e) and constructs the SDID model as follows:\u003cdiv id=\"Equ2\" class=\"Equation\"\u003e\u003cdiv format=\"TEX\" class=\"mathdisplay\" id=\"FileID_Equ2\" name=\"EquationSource\"\u003e\n$$\\:{y}_{i,t}={\\alpha\\:}_{0}+{{\\alpha\\:}_{1}{HSR}_{i,t}+{\\alpha\\:}_{2}{Pos{t}_{construction}}_{i,t}+{\\alpha\\:}_{3}({HSR}_{i,t}\\times\\:{Post\\_construction}_{i,t})+\\beta\\:}_{1}{y}_{i,t-1}+{\\omega\\:}_{1}W{Y}_{t-1}{+\\omega\\:}_{2}W{Y}_{t}+{\\omega\\:}_{3}W{HSR}_{t}+{X{\\prime\\:}}_{i,t}\\theta\\:+{ϵ}_{i,t}$$\u003c/div\u003e\u003cdiv class=\"EquationNumber\"\u003e2\u003c/div\u003e\u003c/div\u003e\u003c/p\u003e \u003cp\u003eIn Eq.\u0026nbsp;\u003cspan refid=\"Equ2\" class=\"InternalRef\"\u003e2\u003c/span\u003e, \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{y}_{i,t-1}\\)\u003c/span\u003e\u003c/span\u003e represents the time lag term of \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{y}_{i,t}\\)\u003c/span\u003e\u003c/span\u003e, which refers to tourist arrivals, making Eq.\u0026nbsp;\u003cspan refid=\"Equ2\" class=\"InternalRef\"\u003e2\u003c/span\u003e a Dynamic Difference-in-Differences model (Dynamic DID). The spatial connectivity matrix W quantifies the spatial spillover of tourism development. Furthermore, \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:W{Y}_{t-1},\\:W{Y}_{t},\\)\u003c/span\u003e\u003c/span\u003e and \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:W{HSR}_{t}\\)\u003c/span\u003e\u003c/span\u003e are spatial lag terms of \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{Y}_{t-1},\\:{Y}_{t},\\)\u003c/span\u003e\u003c/span\u003e and \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{HSR}_{t}\\)\u003c/span\u003e\u003c/span\u003e, respectively, used to control for the spatial spillover effects of tourism development, making Eq.\u0026nbsp;\u003cspan refid=\"Equ2\" class=\"InternalRef\"\u003e2\u003c/span\u003e a Dynamic Spatial Difference-in-Differences model (Dynamic SDID).\u003c/p\u003e \u003cp\u003eIn this study, we also investigate whether TTSBS (tourist transportation shuttle bus services) affect HSR's effects on tourism development. Therefore, we extend Eq.\u0026nbsp;\u003cspan refid=\"Equ2\" class=\"InternalRef\"\u003e2\u003c/span\u003e as follows:\u003cdiv id=\"Equ3\" class=\"Equation\"\u003e\u003cdiv format=\"TEX\" class=\"mathdisplay\" id=\"FileID_Equ3\" name=\"EquationSource\"\u003e\n$$\\:{y}_{i,t}={\\alpha\\:}_{0}+{{\\alpha\\:}_{1}{HSR}_{i,t}+{\\alpha\\:}_{2}{Pos{t}_{construction}}_{i,t}+{\\alpha\\:}_{3}({HSR}_{i,t}\\times\\:{Post\\_construction}_{i,t})+\\beta\\:}_{1}{y}_{i,t-1}+{\\omega\\:}_{1}W{Y}_{t-1}{+\\omega\\:}_{2}W{Y}_{t}+{\\omega\\:}_{3}W{HSR}_{t}+{\\omega\\:}_{3}W{SHU}_{t}+{\\gamma\\:}_{1}{SHU}_{i,t}+{\\gamma\\:}_{2}{SHU}_{i,t}{HSR}_{i,t}+{X{\\prime\\:}}_{i,t}\\theta\\:+{ϵ}_{i,t}$$\u003c/div\u003e\u003cdiv class=\"EquationNumber\"\u003e3\u003c/div\u003e\u003c/div\u003e\u003c/p\u003e \u003cp\u003eIn Eq.\u0026nbsp;\u003cspan refid=\"Equ3\" class=\"InternalRef\"\u003e3\u003c/span\u003e, \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{SHU}_{i,t}\\)\u003c/span\u003e\u003c/span\u003e represents the number of TTSBS (tourist transportation shuttle bus services), and \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:W{SHU}_{t}\\)\u003c/span\u003e\u003c/span\u003e represents the spillover effects of \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{SHU}_{t}\\)\u003c/span\u003e\u003c/span\u003e. The term \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{SHU}_{i,t}{HSR}_{i,t}\\)\u003c/span\u003e\u003c/span\u003e represents how \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{SHU}_{i,t}\\)\u003c/span\u003e\u003c/span\u003e affects HSR\u0026rsquo;s effects on tourism development. This extension allows us to explore the potential interaction between TTSBS and HSR in their combined impact on tourism development.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec4\" class=\"Section2\"\u003e \u003ch2\u003e2.2 Introduction of Variables and Data Sources\u003c/h2\u003e \u003cp\u003eThis study uses commonly used and significant explanatory variables from the literature, including income (Lim 2006, Song, Witt et al. \u003cspan citationid=\"CR35\" class=\"CitationRef\"\u003e2009\u003c/span\u003e), price (Song and Witt \u003cspan citationid=\"CR34\" class=\"CitationRef\"\u003e2006\u003c/span\u003e), population variables (Song, Witt et al. \u003cspan citationid=\"CR35\" class=\"CitationRef\"\u003e2009\u003c/span\u003e), lag terms of the dependent variable (Gar\u0026iacute;n-Mu\u0026ntilde;oz 2009, Zhang, Kulendran et al. 2010, Yap and Allen 2011, Massidda and Etzo 2012, Liu \u003cspan citationid=\"CR26\" class=\"CitationRef\"\u003e2020\u003c/span\u003e), as well as the variables related to the research topic, namely, Taiwan High Speed Rail (THSR) and Taiwan Tourist Shuttle service (TSHU). Commonly used and significant explanatory variables for tourism demand have been extensively applied in tourism demand studies, but due to space limitations, this paper does not provide detailed explanations. Interested readers can refer to the listed references. The following will elaborate on the transportation-related policy variables that this paper focuses on (Albalate and Fageda \u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2016\u003c/span\u003e, Gao, Su et al. \u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e2019\u003c/span\u003e).\u003c/p\u003e \u003cp\u003e(1) Introduction of Taiwan High Speed Rail\u003c/p\u003e \u003cp\u003eThe idea of Taiwan High Speed Rail (THSR) originated from the \"Development of Super Express Railway Project\" proposed by the Taiwan Railway Administration in 1974. In response to the deteriorating transportation service quality and increasing congestion in the western region of Taiwan, the Ministry of Transportation conducted the \"Feasibility Study of Taiwan Western Corridor High-Speed Rail\" in 1987. On July 14, 1994, it was included in the government's priority list of major public construction projects and became the country's first significant national infrastructure project promoted through private investment. The THSR route spans 350 kilometers and includes 12 stations, such as Nangang, Taipei, Banqiao, Taoyuan, Hsinchu, Miaoli, Taichung, Changhua, Yunlin, Chiayi, Tainan, and Zuoying. The stations were opened for service on different dates: January 5, 2007, for Banqiao, Taoyuan, Hsinchu, Taichung, Chiayi, Tainan, and Zuoying; March 2, 2007, for Taipei; December 1, 2015, for Miaoli, Changhua, and Yunlin; and July 1, 2016, for Nangang. The THSR trains can reach a maximum speed of 300 kilometers per hour, significantly reducing the travel time between Taipei and Kaohsiung to 90 minutes. The daily passenger capacity of THSR exceeds 300,000. After the THSR was launched, it transformed Taiwan's spatial structure and turned the northern, central, and southern metropolitan areas into a balanced one-day living circle (Ministry of Transportation and Communications, Railway Bureau, 2020).\u003c/p\u003e \u003cp\u003e(2) Introduction of Taiwan Tourist Shuttle service\u003c/p\u003e \u003cp\u003eTaiwan Tourist Shuttle service (TSHU) is a bus service specifically designed for tourism, providing transportation for travelers from major Taiwan Railways and THSR stations to Taiwan's main tourist attractions. Its purpose is to offer convenient public transportation services to attract travelers, reduce the proportion of self-driving trips, and avoid traffic congestion, parking shortages, and related issues affecting travel quality. Additionally, it aims to increase the use of public transportation among locals to promote energy-saving and carbon reduction in tourism activities.\u003c/p\u003e \u003cp\u003eAfter analyzing the rapid increase in the number of tourists traveling in Taiwan, the tendency of using private vehicles for travel, and the preference for free and independent travel, the Tourism Bureau of the Ministry of Transportation concluded that there was a need to establish transportation services to cater to the needs of independent travelers. Since 2009, they began planning and providing travel shuttle services for independent travelers (Tourism Bureau, Ministry of Transportation, 2012). The route selection criteria for TSHU are as follows:\u003cdiv class=\"BlockQuote\"\u003e\u003cp\u003e\"(1) Select attractive destinations: As transportation services are essentially derived from travel needs, the key is still the attractiveness of the destinations. There must be compelling reasons for travelers to visit the locations so that the transportation services provided can meet the needs of travelers, and operators can obtain appropriate profits to provide long-term services.\u003c/p\u003e\u003cp\u003e(2) Collaborate with multiple transportation modes: To promote the use of public transportation for travel in Taiwan, it is necessary to link the shuttle routes to nearby major long-distance transportation stations, such as Taiwan Railways and THSR stations, to facilitate travelers' transfers and usage.\u003c/p\u003e\u003cp\u003e(3) Fast and stable transportation services: To attract independent travelers to take advantage of the convenient shuttle service and facilitate travel planning, the shuttle routes should be designed differently from the existing public bus services, with fewer stops, stable schedules, and reasonable intervals, providing fast and direct access to tourist attractions.\u003c/p\u003e\u003cp\u003e(4) Integrated preferential ticketing\u003c/p\u003e\u003cp\u003e(5) Short-term subsidies and long-term autonomous operation\"\u003c/p\u003e\u003c/div\u003e\u003c/p\u003e \u003cp\u003eAfter the evaluation process, TSHU officially started operating on April 5, 2010. Each year, based on the operation status of selected routes and new route applications, adjustments are made to the routes for the following year.\u003c/p\u003e \u003cp\u003eThe data for the aforementioned variables come from various statistical databases of the respective authorities. The number of tourists in each city and county is from the Tourism Statistics Yearbook of the Tourism Bureau, Ministry of Transportation. Data on real per capita national income and the total population are from the databases of the Directorate-General of Budget, Accounting and Statistics. Transportation-related policy variables are sourced from the Railway Bureau and the Tourism Bureau of the Ministry of Transportation.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec5\" class=\"Section2\"\u003e \u003ch2\u003e2.3 Unit Root Test for Data\u003c/h2\u003e \u003cp\u003eThis study utilizes panel data from 19 counties and cities in Taiwan from 2001 to 2020. The analysis of panel data depends on whether the data possesses unit root characteristics, such as the existence of long-term trends or cointegration relationships. If the data exhibits unit root characteristics, indicating the presence of long-term trends or cointegration, it is necessary to perform data transformations such as differencing to ensure the reliability of the analysis results; otherwise, hypothesis testing may become invalid. Therefore, before conducting the analysis, we follow the suggestion of Gonz\u0026aacute;lez-Val and Silvestre (\u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e2023\u003c/span\u003e) and the approach of Udeagha and Muchapondwa (\u003cspan citationid=\"CR37\" class=\"CitationRef\"\u003e2023\u003c/span\u003e) for analyzing SDID, which involves conducting panel unit root tests on the tracking data.\u003c/p\u003e \u003cp\u003eCommonly used panel data unit root tests include the Levin-Lin-Chu (LLC), Im-Pesaran-Shin (IPS), Fisher-type, and Hadri tests. However, these methods may not be suitable for the data in this study. LLC and IPS assume that each individual is independent, and if there is spatial correlation in the data, resulting in cross-sectional dependence, the mean of the t-statistic may not be zero, leading to non-zero bias and affecting the test results. While Fisher-type can tolerate a certain degree of cross-sectional correlation, if the cross-sectional correlation is too strong, the Fisher statistic may still become invalid. Hadri uses the OLS estimation method to calculate the LM statistic, but this test is also influenced by cross-sectional dependence, leading to an overestimation of the LM statistic. Since Taiwan's tourist arrivals exhibit spatial correlation (Liu \u003cspan citationid=\"CR26\" class=\"CitationRef\"\u003e2020\u003c/span\u003e), indicating the presence of cross-sectional dependence, the LLC, IPS, Fisher-type, and Hadri tests may produce biased test results.\u003c/p\u003e \u003cp\u003eWhen cross-sectional dependence is present in panel data, panel unit root tests that consider cross-sectional dependence should be used. Commonly used panel unit root tests considering cross-sectional dependence include the Pesaran (\u003cspan citationid=\"CR31\" class=\"CitationRef\"\u003e2007\u003c/span\u003e) CD test, Moon and Perron (2004) CD test, and Breitung (2000) test with cross-section dependence adjustment. However, the assumptions of these methods do not match the characteristics of the data in this study. Pesaran (\u003cspan citationid=\"CR31\" class=\"CitationRef\"\u003e2007\u003c/span\u003e) CD test has strict assumptions regarding cross-sectional dependence, and when the form of cross-sectional dependence is more complex, the test's performance may be affected. Moon and Perron (2004) CD test requires estimating the degree of cross-sectional dependence, which may encounter computational difficulties with high-dimensional panel data. Breitung (2000) test has strict assumptions about cross-sectional dependence, and its estimation of the degree of cross-sectional dependence may not be accurate enough. The data in this study exhibit complex forms of cross-sectional dependence due to spatial correlation, and with a large number of model variables, increasing the dimension of the panel data, thus making these unit root tests that consider cross-sectional dependence unsuitable for this study.\u003c/p\u003e \u003cp\u003eBaltagi, Bresson et al. (\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e2007\u003c/span\u003e) suggested using Pesaran (\u003cspan citationid=\"CR31\" class=\"CitationRef\"\u003e2007\u003c/span\u003e) CIPS test because it eliminates the influence of spatial and cross-sectional dependence. They compared this method with others using simulations and real data analysis. They simulated balanced or unbalanced panel data with or without unit roots and spatial dependence, and then performed various panel unit root tests on these datasets to calculate the statistical efficiency and rejection rates. They also applied these tests to actual panel data, including US state house price indices, European consumer price indices, and exchange rates, and compared the results and implications. The authors found that when spatial dependence is present, commonly used unit root tests like Levin-Lin-Chu (2002), Im-Pesaran-Shin (2003), Maddala and Wu (1999), Choi (2001), etc., have reduced statistical power and may produce incorrect conclusions. Pesaran (\u003cspan citationid=\"CR31\" class=\"CitationRef\"\u003e2007\u003c/span\u003e) CIPS test, on the other hand, does not have this issue and exhibits higher statistical efficiency.\u003c/p\u003e \u003cp\u003eConsidering the presence of spatial correlation in the data, and using the Dynamic Spatial Difference-in-Differences model for econometric analysis, the panel data suffer from cross-sectional dependence, causing biases in the results of the Levin-Lin-Chu (LLC), Im-Pesaran-Shin (IPS), Fisher-type, and Hadri tests. Moreover, the spatial correlation introduces a more complex form of cross-sectional dependence and a larger number of model variables, which do not align with the assumptions of panel unit root tests used for cross-sectional dependence. In line with the recommendation of Baltagi, Bresson et al. (\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e2007\u003c/span\u003e), we utilize Pesaran (\u003cspan citationid=\"CR31\" class=\"CitationRef\"\u003e2007\u003c/span\u003e) CIPS test as the unit root test method for the data in this study.\u003c/p\u003e \u003c/div\u003e"},{"header":"3. Analysis Results","content":"\u003cdiv id=\"Sec7\" class=\"Section2\"\u003e \u003ch2\u003e3.1 Descriptive Statistics for the Panel\u003c/h2\u003e \u003cp\u003eIn this section, we present the descriptive statistics for the variables used in the econometric model. Descriptive statistics help the audience understand the data and assess the impact of its estimates (Song, Qiu et al. \u003cspan citationid=\"CR33\" class=\"CitationRef\"\u003e2023\u003c/span\u003e). The descriptive statistics for the panel variables are shown in Table\u0026nbsp;\u003cspan refid=\"Tab2\" class=\"InternalRef\"\u003e2\u003c/span\u003e. All continuous control variables are log-transformed.\u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab2\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 2\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003eDescriptive statistics\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"7\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c5\" colnum=\"5\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c6\" colnum=\"6\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c7\" colnum=\"7\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\"\u003e \u003cp\u003eVariable\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003eMean\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c3\"\u003e \u003cp\u003eStd. dev.\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c4\"\u003e \u003cp\u003eMin\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c5\"\u003e \u003cp\u003eMax\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c6\"\u003e \u003cp\u003eSkewness\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c7\"\u003e \u003cp\u003eKurtosis\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003elntourist\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e12.80\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e2.58\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e-11.51\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e16.15\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e-7.16\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e \u003cp\u003e67.40\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eWlntourist\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e14.59\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.93\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e4.03\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e16.84\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e-1.52\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e \u003cp\u003e10.08\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003elnpop\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e16.95\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.02\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e16.92\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e16.97\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e-0.25\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e \u003cp\u003e1.88\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003elnincome\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e10.68\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.18\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e10.52\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e11.39\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e2.42\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e \u003cp\u003e8.33\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eGasoline_Index\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e84.10\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e16.91\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e56.42\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e112.15\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e0.03\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e \u003cp\u003e1.74\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eSARS\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e0.03\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.16\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.00\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e1.00\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e5.85\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e \u003cp\u003e35.23\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eFinancial Crisis\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e0.10\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.30\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.00\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e1.00\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e2.63\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e \u003cp\u003e7.90\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eTHSR\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e0.33\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.47\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.00\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e1.00\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e0.75\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e \u003cp\u003e1.56\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eTSHU\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e0.68\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e1.09\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.00\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e5.00\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e1.46\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e \u003cp\u003e4.16\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec8\" class=\"Section2\"\u003e \u003ch2\u003e3.2 Unit Root Analysis\u003c/h2\u003e \u003cp\u003eOur data exhibit spatial dependence and cross-sectional dependence characteristics. Following the recommendation of Baltagi, Bresson et al. (\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e2007\u003c/span\u003e), we use the CIPS test proposed by Pesaran (\u003cspan citationid=\"CR31\" class=\"CitationRef\"\u003e2007\u003c/span\u003e) to test for unit roots. The results of all variable tests reject the null hypothesis of unit roots (Table\u0026nbsp;\u003cspan refid=\"Tab3\" class=\"InternalRef\"\u003e3\u003c/span\u003e). In addition to the CIPS test results, we also provide the results of the Breitung test, Fisher-type test, and IPS test for comparison.\u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab3\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 3\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003eUnit root test (at level)\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"5\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c5\" colnum=\"5\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\"\u003e \u003cp\u003eVariable\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003eCIPS test\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c3\"\u003e \u003cp\u003eBreitung test\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c4\"\u003e \u003cp\u003eFisher-type\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c5\"\u003e \u003cp\u003eIPS test\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003elntourist\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e-4.961***\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e-2.911***\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e23.216***\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e-10.627***\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eWlntourist\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e-5.705***\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e-3.954***\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e69.778***\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e-5.064***\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003elnpop\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e2.610**\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e4.990\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e-4.358\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e-7.7902***\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003elnincome\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e2.610**\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e-8.453***\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e152.7517***\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e-10.627***\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eGasoline_index\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e2.610**\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e-0.635\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e7.3613***\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e-1.498\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eTSHU\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e-2.330***\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.758\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e8.652***\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e-0.314\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003ctfoot\u003e \u003ctr\u003e\u003ctd colspan=\"5\"\u003eThe symbols *, **, and *** refer to level of significance at 10%, 5%, and 1% respectively\u003c/td\u003e\u003c/tr\u003e \u003c/tfoot\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003cp\u003eThe unit root test results for each variable are inconsistent across the four methods. For example, all four methods reject the presence of unit roots in \"lntourist\"; both CIPS test and IPS test reject the unit root hypothesis for \"lnpop,\" while Breitung test and Fisher-type test do not; CIPS test and Fisher-type test reject the unit root hypothesis for \"Shuttle_Routes,\" while Breitung test and IPS test do not. Different tests yield different results and may influence the analysis method. As the CIPS test is suitable for data with spatial dependence and cross-sectional dependence, which is the case for our data, we adopt the results of the CIPS test and reject the presence of unit roots for all variables.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec9\" class=\"Section2\"\u003e \u003ch2\u003e3.3 Analysis of Dynamic SDID Estimation Results\u003c/h2\u003e \u003cp\u003eThis paper mainly applies Dynamic SDID to analyze the effect of the High-speed rail and tourist transit service on Tourism Demand. To evaluate whether Dynamic SDID outperforms the traditional DID and the use of DUMMY variables to represent the presence or absence of the High-speed rail, we also present the results of traditional DID and DUMMY analysis.\u003c/p\u003e \u003cp\u003eThe analysis models for traditional DID are DID model 1 (High-speed rail) and DID model 2 (High-speed rail and tourist transit service), and for Dynamic SDID are Dynamic SDID (High-speed rail and tourist transit service) (Table\u0026nbsp;\u003cspan refid=\"Tab4\" class=\"InternalRef\"\u003e4\u003c/span\u003e). The analysis models using DUMMY variables are DUMMY model 1 (High-speed rail), DUMMY model 2 (High-speed rail and tourist transit service), and DUMMY model 3 (High-speed rail and tourist transit service with controlling spillover effect) (Table\u0026nbsp;\u003cspan refid=\"Tab5\" class=\"InternalRef\"\u003e5\u003c/span\u003e). When not controlling for spillover effect, there is no clear advantage between the DID and DUMMY models; when controlling for spillover effect, the DID model performs better than the DID model without controlling for spillover effect; the DUMMY model with spillover effect control performs better than the DUMMY model without spillover effect control; the DID model with spillover effect control outperforms the DUMMY model with spillover effect control (Table\u0026nbsp;\u003cspan refid=\"Tab4\" class=\"InternalRef\"\u003e4\u003c/span\u003e, Table\u0026nbsp;\u003cspan refid=\"Tab5\" class=\"InternalRef\"\u003e5\u003c/span\u003e). Therefore, controlling for spillover effect is crucial for analyzing the effect of the High-speed rail and tourist transit service on Tourism Demand, and after controlling for spillover effect, the DID model outperforms the DUMMY model. Consequently, the following analysis will discuss the results based on the DID model 3 with spillover effect control (i.e., Dynamic SDID).\u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab4\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 4\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003eResults of models using DID for HSR\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"7\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c5\" colnum=\"5\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c6\" colnum=\"6\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c7\" colnum=\"7\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\"\u003e \u003cp\u003eVariables\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colspan=\"2\" nameend=\"c3\" namest=\"c2\"\u003e \u003cp\u003eDID model 1\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colspan=\"2\" nameend=\"c5\" namest=\"c4\"\u003e \u003cp\u003eDID model 2\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colspan=\"2\" nameend=\"c7\" namest=\"c6\"\u003e \u003cp\u003eDID model 3\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eCoefficient\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eStd. err.\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eCoefficient\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003eStd. err.\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003eCoefficient\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003eStd. err.\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eLaglntourist\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e0.748***\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.168\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0.747***\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e0.168\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e0.839***\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e0.12\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eWlntourist\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e0.565***\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e0.031\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eLagWlntourist\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e-0.038\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e0.032\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003elnNpop\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e29.764***\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e5.638\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e28.529***\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e6.788\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e11.94\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e8.47\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003elnincome\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e0.265***\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.065\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0.262***\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e0.066\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e0.141**\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e0.057\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eGasoline_Index\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e0.002\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.003\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0.003\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e0.003\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e0.003\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e0.003\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eSARS\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e-0.173\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.129\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e-0.173\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e0.129\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e-0.048\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e0.112\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eFinancial_Crisis\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e0.023\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.111\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0.023\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e0.112\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e0.011\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e0.105\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003etreatTHSR\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e-0.49\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.92\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e-0.509\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e0.926\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e-0.213\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e0.95\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003epostTHSR\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e-0.580***\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.215\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e-0.493**\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e0.219\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e0.092\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e0.313\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eTHSRDID\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e0.993***\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.211\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0.891***\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e0.216\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e0.608**\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e0.248\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eTSHU\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e-0.076\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e0.068\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e-0.063\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e0.076\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eTSHUTHSR\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0.185**\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e0.088\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e0.236**\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e0.099\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eWTSHU\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e0.005\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e0.021\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eWTHSR\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e-0.227*\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e0.115\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e_cons\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e-124.805***\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e24.078\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e-119.650***\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e29.029\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e-32.009\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e23.047\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eWald Test\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e181.65\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e186.339\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e549.722\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eF-Test\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e20.183\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e16.94\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e36.648\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eRoot MSE (Sigma)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e379.699\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e364.012\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e178.112\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eLLF\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e-4489.57\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e-4503.58\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e-4388.5\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eAIC\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e0.758\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0.765\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e0.718\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eSC\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e0.771\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0.781\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e0.738\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab5\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 5\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003eResults of models using dummy variable for HSR\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"7\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c5\" colnum=\"5\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c6\" colnum=\"6\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c7\" colnum=\"7\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\"\u003e \u003cp\u003eVariables\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colspan=\"2\" nameend=\"c3\" namest=\"c2\"\u003e \u003cp\u003eDUMMY model 1\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colspan=\"2\" nameend=\"c5\" namest=\"c4\"\u003e \u003cp\u003eDUMMY model 2\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colspan=\"2\" nameend=\"c7\" namest=\"c6\"\u003e \u003cp\u003eDUMMY model 3\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eCoefficient\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eStd. err.\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eCoefficient\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003eStd. err.\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003eCoefficient\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003eStd. err.\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eLaglntourist\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e0.745***\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.168\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0.744\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e0.169***\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e0.842***\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e0.118\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eWlntourist\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e0.565***\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e0.031\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eLagWlntourist\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e-0.036\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e0.031\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003elnNpop\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e21.159***\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e4.648\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e19.251***\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e5.915\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e13.111\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e8.042\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003elnincome\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e0.259***\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.066\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0.256***\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e0.066\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e0.143**\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e0.056\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eGasoline_Index\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e0.001\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.003\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0.002\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e0.003\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e0.003\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e0.003\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eSARS\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e-0.168\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.129\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e-0.17\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e0.129\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e-0.048\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e0.111\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eFinancial_Crisis\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e-0.035\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.108\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e-0.019\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e0.11\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e0.01\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e0.104\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eTHSRDUMMY\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e0.933***\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.153\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0.887***\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e0.163\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e0.816***\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e0.201\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eTSHU\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e-0.008\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e0.07\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e-0.019\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e0.077\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eTSHU THSRDUMMY\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0.103\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e0.09\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e0.164\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e0.102\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eWTSHU\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e0.000\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e0.021\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eWTHSR\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e-0.179**\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e0.075\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e_cons\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e-89.010***\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e20.043\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e-81.007**\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e25.597\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e-34.68\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e21.536\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eWald Test\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e197.362\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e199.563\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e560.665\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eF-Test\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e28.195\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e22.174\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e43.128\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eRoot MSE (Sigma)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e269.801\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e245.538\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e195.481\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eLLF\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e-4494.77\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e-4506.22\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e-4419.08\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eAIC\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e0.759\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0.765\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e0.729\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eSC\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e0.77\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0.778\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e0.747\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003cp\u003eThe results of Dynamic SDID analysis show that the current tourism demand (\"Wlntourist\") exhibits significant positive spatial autocorrelation, and the spatial autocorrelation of the lagged tourism demand (\"lagWlntourist\") is negative, which aligns with the estimates of Liu (\u003cspan citationid=\"CR26\" class=\"CitationRef\"\u003e2020\u003c/span\u003e). The study's use of significant and commonly used explanatory variables for tourism demand is consistent with past research. The lagged tourism demand and income show significant positive impacts, while the effects of population, travel cost, and the 2008 financial crisis are positive but not significantly different from zero, and the impact of SARS is negative but not significantly different from zero. Although the research results differ from past studies, this difference may arise from tourism's apparent \"decline-rebound-overshoot\" effect (Jeong and Lee \u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e2022\u003c/span\u003e, Kei Kuok, Koo et al. \u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e2023\u003c/span\u003e). Major economic and environmental changes lead to significant fluctuations in tourism activities in the short term, making short-term estimates of these variables statistically significant. However, in the long term, these fluctuations tend to level out, leading to non-significant estimates.\u003c/p\u003e \u003cp\u003eThe impact of THSR on tourism demand is significantly positive (DID\u0026thinsp;=\u0026thinsp;0.608**), which aligns with intuition. The significant negative spatial autocorrelation effect of THSR (WHSR=-0.227*) indicates that THSR has negative spillover effects on tourism activities. The results suggest that the influx of tourists brought by THSR primarily comes from existing tourists in the surrounding regions rather than creating new tourism demand. Thus, from a national perspective on the tourism industry, THSR does not create new demand, which supports the conclusions of Delaplace, Pagliara et al. (\u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e2014\u003c/span\u003e), Kurihara and Wu (\u003cspan citationid=\"CR22\" class=\"CitationRef\"\u003e2016\u003c/span\u003e), Pagliara, La Pietra et al. (\u003cspan citationid=\"CR28\" class=\"CitationRef\"\u003e2015\u003c/span\u003e), and Yu and Fan (\u003cspan citationid=\"CR42\" class=\"CitationRef\"\u003e2018\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eThe impact of TSHU on tourism demand is negative but not significant (TSHU=-0.063), which contradicts intuition. However, upon further investigation, this result is not surprising. The effect of TSHU on tourism growth depends on both its positive contribution and the tourism market trend in the destinations it connects to. If the tourism market in the destinations traversed by TSHU is in decline, TSHU's positive contribution may not be sufficient to reverse the decline, but it can slow down the pace of tourism decline. When the effects of TSHU and THSR on tourism activities are considered, the spatial autocorrelation effect of TSHU is positive but not significant (WSHU\u0026thinsp;=\u0026thinsp;0.005), indicating a positive spillover effect of TSHU on tourism activities, implying that although TSHU may not significantly boost tourism development in specific and individual travel destinations, it contributes to the overall tourism development in the region.\u003c/p\u003e \u003cp\u003eThe interaction effect of TSHU and THSR is significantly positive (TSHUTHSR\u0026thinsp;=\u0026thinsp;0.236**), indicating a complementary relationship between the two. When tourists use THSR and do not use personal transportation, the convenience of connecting transport depends on shuttle services, and providing shuttle buses to improve transport convenience contributes to promoting tourism activities. The estimated results of TSHUTHSR support the aforementioned argument.\u003c/p\u003e \u003c/div\u003e"},{"header":"4. Conclusion and Policy Recommendations","content":"\u003cp\u003eThis paper uses the Dynamic Spatial Difference-in-Differences (Dynamic SDID) model to analyze the effect of High-speed rail (HSR) and tourist transit service (TTSBS) on Tourism Demand. The transportation construction variables considered are THSR and TSHU. Since TSHU mainly serves as a connection between tourist attractions, we include the interaction effect between TSHU and THSR. Additionally, to verify whether the increased tourism from transportation construction comes from new tourism demand created by the construction or from substituting existing demand, we introduce spatial autocorrelation variables for THSR and TSHU. Considering the existence of spatial dependence and cross-sectional dependence in the data, we use the CIPS test for unit root testing.\u003c/p\u003e \u003cp\u003eThe research findings show that THSR and TSHU have different impacts on tourism. THSR has a positive impact on tourism demand, while TSHU has a negative impact. The interaction effect between TSHU and THSR is positive. The negative spatial autocorrelation effect of THSR indicates that it has negative spillover effects on tourism activities. This result suggests that the influx of tourists brought by THSR primarily comes from existing tourists in the surrounding regions rather than creating new tourism demand. Therefore, from a national perspective on the tourism industry, THSR does not create new demand and may indicate the presence of substitution between tourism attractions. This finding supports the conclusions of Delaplace, Pagliara et al. (\u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e2014\u003c/span\u003e), Kurihara and Wu (\u003cspan citationid=\"CR22\" class=\"CitationRef\"\u003e2016\u003c/span\u003e), Pagliara, La Pietra et al. (\u003cspan citationid=\"CR28\" class=\"CitationRef\"\u003e2015\u003c/span\u003e), and Yu and Fan (\u003cspan citationid=\"CR42\" class=\"CitationRef\"\u003e2018\u003c/span\u003e), indicating that HSR does not significantly contribute to promoting tourism. Therefore, to boost national tourism demand, the focus should be on enhancing the attractiveness of tourist destinations rather than solely relying on HSR construction.\u003c/p\u003e \u003cp\u003eTSHU has a non-significant negative impact on tourism, and the reason lies in the criteria for selecting TSHU routes, such as selecting attractive tourist destinations and collaborating with various transportation modes, which may not have been fully implemented, or the data used for evaluation may not align with the actual conditions of the tourism market. Thus, TSHU should focus on linking important tourist attractions. Evaluating TSHU routes using data that reflects the actual conditions of the tourism market can improve transportation convenience and promote the development of the tourism industry.\u003c/p\u003e \u003cp\u003ePrevious literature mainly explored the impact of a single HSR on tourism demand (Yu and Fan \u003cspan citationid=\"CR42\" class=\"CitationRef\"\u003e2018\u003c/span\u003e, Gao, Su et al. \u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e2019\u003c/span\u003e, Deng, Gan et al. \u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e2021\u003c/span\u003e) without considering the interaction effects between TTSBS and HSR or the spillover effects, leading to serious biases, such as overlooking important explanatory variables or using incorrect functions. Moreover, these studies did not consider the spatial correlation between tourism trips and transportation construction, leading to biased estimation results. In contrast, this study simultaneously considers THSR and TSHU, controls for the spatial correlation between tourism trips and transportation construction, and takes into account the interaction effect between THSR and TSHU, thereby excluding known factors that could cause biased estimates, resulting in more reliable and unbiased estimation results.\u003c/p\u003e \u003cp\u003eThe empirical and policy contributions of this study lie in demonstrating the interaction effects, spatial effects, and uncertainties of HSR and TTSBS on tourism demand. For HSR, this study shows that it can increase tourism demand in the destinations where HSR stations are located but may reduce tourism demand in surrounding areas. TTSBS can further enhance the impact of HSR on tourism demand. Policy measures, such as package tours and interregional transportation connections, should be implemented to mitigate the impact of HSR on surrounding tourist destinations. As for TTSBS, it does not have a significant impact on Taiwan's tourism demand, which may be due to the selection of TTSBS routes.\u003c/p\u003e \u003cp\u003eThe theoretical contribution of this study lies in simultaneously exploring the impacts of HSR and TTSBS on tourism, not limited to each HSR and TTSBS individually but also considering their interaction effects and spatial externalities. In terms of modeling, this study uses the Dynamic Spatial Difference-in-Differences model, which better suits the theoretical requirements of analyzing the causal relationship between HSR and tourism demand. The empirical results show that Dynamic SDID outperforms DID or DUMMY models, supporting Albalate and Fageda (\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2016\u003c/span\u003e) argument that differences in research results regarding the impact of HSR on tourism are due to different econometric methods and the network design and accessibility of HSR, providing a reference for future research in model settings. The use of CIPS test for unit root testing yields more consistent estimation results, making it a valuable reference for subsequent tourism studies.\u003c/p\u003e"},{"header":"Declarations","content":"\u003ch2\u003eAuthor Contribution\u003c/h2\u003e\u003cp\u003eT.L. wrote the main manuscript text.\u003c/p\u003e\u003ch2\u003eAcknowledgments\u003c/h2\u003e \u003cp\u003eThis work was supported by the National Science and Technology Council, Taiwan, under Grant MOST 110-2410-H-110 -073 -programs.\u003c/p\u003e\u003ch2\u003eData Availability\u003c/h2\u003e\u003cp\u003eData derived from public domain resources.\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\n\u003cli\u003eAlbalate, D., J. Campos and J. L. Jim\u0026eacute;nez (2017). \u0026quot;Tourism and high speed rail in Spain: Does the AVE increase local visitors?\u0026quot; \u003cu\u003eAnnals of Tourism Research\u003c/u\u003e \u003cstrong\u003e65\u003c/strong\u003e: 71-82.\u003c/li\u003e\n\u003cli\u003eAlbalate, D. and X. Fageda (2016). \u0026quot;High speed rail and tourism: Empirical evidence from Spain.\u0026quot; \u003cu\u003eTransportation Research Part A: Policy and Practice\u003c/u\u003e \u003cstrong\u003e85\u003c/strong\u003e: 174-185.\u003c/li\u003e\n\u003cli\u003eBaltagi, B. H., G. Bresson and A. Pirotte (2007). \u0026quot;Panel unit root tests and spatial dependence.\u0026quot; \u003cu\u003eJournal of Applied Econometrics\u003c/u\u003e \u003cstrong\u003e22\u003c/strong\u003e(2): 339-360.\u003c/li\u003e\n\u003cli\u003eBazin, S., C. Beckerich and M. Delaplace (2006). \u0026quot;Analyse prospective des impacts de la Ligne Grande Vitesse Est-Europ\u0026eacute;enne dans l\u0026rsquo;agglom\u0026eacute;ration r\u0026eacute;moise et en r\u0026eacute;gion Champagne-Ardenne.\u0026quot; \u003cu\u003eRapport final pour le Conseil R\u0026eacute;gional Champagne-Ardenne\u003c/u\u003e \u003cstrong\u003e495\u003c/strong\u003e.\u003c/li\u003e\n\u003cli\u003eBecker, G. S. and G. S. Becker (2009). \u003cu\u003eAccounting for tastes\u003c/u\u003e, Harvard University Press.\u003c/li\u003e\n\u003cli\u003eBecker, G. S. and K. M. Murphy (2009). \u003cu\u003eSocial economics: Market behavior in a social environment\u003c/u\u003e, Harvard University Press.\u003c/li\u003e\n\u003cli\u003eCurrie, C. and P. Falconer (2014). \u0026quot;Maintaining sustainable island destinations in Scotland: The role of the transport\u0026ndash;tourism relationship.\u0026quot; \u003cu\u003eJournal of Destination Marketing \u0026amp; Management\u003c/u\u003e \u003cstrong\u003e3\u003c/strong\u003e(3): 162-172.\u003c/li\u003e\n\u003cli\u003ede Oliveira Santos, G. E., V. Ramos and J. Rey-Maquieira (2011). \u0026quot;A microeconomic model of multidestination tourism trips.\u0026quot; \u003cu\u003eTourism Economics\u003c/u\u003e \u003cstrong\u003e17\u003c/strong\u003e(3): 509-529.\u003c/li\u003e\n\u003cli\u003eDehghan Shabani, Z. and S. Safaie (2018). \u0026quot;Do transport infrastructure spillovers matter for economic growth? Evidence on road and railway transport infrastructure in Iranian provinces.\u0026quot; \u003cu\u003eRegional science policy \u0026amp; practice\u003c/u\u003e \u003cstrong\u003e10\u003c/strong\u003e(1): 49-63.\u003c/li\u003e\n\u003cli\u003eDelaplace, M., F. Pagliara, J. Perrin and S. Mermet (2014). \u0026quot;Can High Speed Rail foster the choice of destination for tourism purpose?\u0026quot; \u003cu\u003eProcedia-Social and Behavioral Sciences\u003c/u\u003e \u003cstrong\u003e111\u003c/strong\u003e: 166-175.\u003c/li\u003e\n\u003cli\u003eDeng, T., C. Gan, H. Du, Y. Hu and D. Wang (2021). \u0026quot;Do high speed rail configurations matter to tourist arrivals? Empirical evidence from China\u0026apos;s prefecture-level cities.\u0026quot; \u003cu\u003eResearch in Transportation Economics\u003c/u\u003e \u003cstrong\u003e90\u003c/strong\u003e: 100952.\u003c/li\u003e\n\u003cli\u003eDi Matteo, D. (2022). \u0026quot;Does high-speed rail matter for tourism? Evidence from Italy.\u0026quot; \u003cu\u003eResearch in Transportation Business \u0026amp; Management\u003c/u\u003e: 100881.\u003c/li\u003e\n\u003cli\u003eFang, J. (2021). \u0026quot;Impacts of high-speed rail on urban smog pollution in China: A spatial difference-in-difference approach.\u0026quot; \u003cu\u003eScience of The Total Environment\u003c/u\u003e \u003cstrong\u003e777\u003c/strong\u003e: 146153.\u003c/li\u003e\n\u003cli\u003eGao, Y., W. Su and K. Wang (2019). \u0026quot;Does high-speed rail boost tourism growth? New evidence from China.\u0026quot; \u003cu\u003eTourism Management\u003c/u\u003e \u003cstrong\u003e72\u003c/strong\u003e: 220-231.\u003c/li\u003e\n\u003cli\u003eGonz\u0026aacute;lez‐Val, R. and J. Silvestre (2023). \u0026quot;War and city size: The asymmetric effects of the Spanish Civil War.\u0026quot; \u003cu\u003eJournal of Regional Science\u003c/u\u003e.\u003c/li\u003e\n\u003cli\u003eGooroochurn, N. and A. Hanley (2005). \u0026quot;Spillover effects in long‐haul visitors between two regions.\u0026quot; \u003cu\u003eRegional Studies\u003c/u\u003e \u003cstrong\u003e39\u003c/strong\u003e(6): 727-738.\u003c/li\u003e\n\u003cli\u003eHwang, Y.-H. and D. R. Fesenmaier (2003). \u0026quot;Multidestination pleasure travel patterns: empirical evidence from the American Travel Survey.\u0026quot; \u003cu\u003eJournal of Travel Research\u003c/u\u003e \u003cstrong\u003e42\u003c/strong\u003e(2): 166-171.\u003c/li\u003e\n\u003cli\u003eJeong, J. Y. and K.-Y. Lee (2022). \u0026quot;A rebound in nature-based tourism intentions during the COVID-19 era.\u0026quot; \u003cu\u003eJournal of Vacation Marketing\u003c/u\u003e: 13567667221111566.\u003c/li\u003e\n\u003cli\u003eKanwal, S., M. I. Rasheed, A. H. Pitafi, A. Pitafi and M. Ren (2020). \u0026quot;Road and transport infrastructure development and community support for tourism: The role of perceived benefits, and community satisfaction.\u0026quot; \u003cu\u003eTourism Management\u003c/u\u003e \u003cstrong\u003e77\u003c/strong\u003e: 104014.\u003c/li\u003e\n\u003cli\u003eKei Kuok, R. U., T. T. Koo and C. Lim (2023). \u0026quot;Economic policy uncertainty and international tourism demand: a global vector autoregressive approach.\u0026quot; \u003cu\u003eJournal of Travel Research\u003c/u\u003e \u003cstrong\u003e62\u003c/strong\u003e(3): 540-562.\u003c/li\u003e\n\u003cli\u003eKhadaroo, J. and B. Seetanah (2007). \u0026quot;Transport infrastructure and tourism development.\u0026quot; \u003cu\u003eAnnals of tourism research\u003c/u\u003e \u003cstrong\u003e34\u003c/strong\u003e(4): 1021-1032.\u003c/li\u003e\n\u003cli\u003eKurihara, T. and L. Wu (2016). \u0026quot;The impact of high speed rail on tourism development: A case study of Japan.\u0026quot; \u003cu\u003eThe Open Transportation Journal\u003c/u\u003e \u003cstrong\u003e10\u003c/strong\u003e(1).\u003c/li\u003e\n\u003cli\u003eLancaster, K. (1966). \u0026quot;Change and innovation in the technology of consumption.\u0026quot; \u003cu\u003eThe American Economic Review\u003c/u\u003e \u003cstrong\u003e56\u003c/strong\u003e(1/2): 14-23.\u003c/li\u003e\n\u003cli\u003eLi, J. and S. Li (2020). \u0026quot;Energy investment, economic growth and carbon emissions in China\u0026mdash;Empirical analysis based on spatial Durbin model.\u0026quot; \u003cu\u003eEnergy Policy\u003c/u\u003e \u003cstrong\u003e140\u003c/strong\u003e: 111425.\u003c/li\u003e\n\u003cli\u003eLim, C. (2006). A survey of tourism demand modeling practice: issues and implications. \u003cu\u003eInternational Handbook on the Economics of Tourism\u003c/u\u003e. L. Dwyer and P. Forsyth. Northampton, Edward Elgar Publishing\u003cstrong\u003e: \u003c/strong\u003e45\u0026ndash;72.\u003c/li\u003e\n\u003cli\u003eLiu, T.-M. (2020). \u0026quot;Habit formation or word of mouth: What does lagged dependent variable in tourism demand models imply?\u0026quot; \u003cu\u003eTourism Economics\u003c/u\u003e \u003cstrong\u003e26\u003c/strong\u003e(3): 461-474.\u003c/li\u003e\n\u003cli\u003eMasson, S. and R. Petiot (2009). \u0026quot;Can the high speed rail reinforce tourism attractiveness? The case of the high speed rail between Perpignan (France) and Barcelona (Spain).\u0026quot; \u003cu\u003eTechnovation\u003c/u\u003e \u003cstrong\u003e29\u003c/strong\u003e(9): 611-617.\u003c/li\u003e\n\u003cli\u003ePagliara, F., A. La Pietra, J. Gomez and J. M. Vassallo (2015). \u0026quot;High Speed Rail and the tourism market: Evidence from the Madrid case study.\u0026quot; \u003cu\u003eTransport Policy\u003c/u\u003e \u003cstrong\u003e37\u003c/strong\u003e: 187-194.\u003c/li\u003e\n\u003cli\u003ePan, H., C. Cong, X. Zhang and Y. Zhang (2020). \u0026quot;How do high-speed rail projects affect the agglomeration in cities and regions?\u0026quot; \u003cu\u003eTransportation Research Part D: Transport and Environment\u003c/u\u003e \u003cstrong\u003e88\u003c/strong\u003e: 102561.\u003c/li\u003e\n\u003cli\u003ePeeters, P. and F. Schouten (2006). \u0026quot;Reducing the ecological footprint of inbound tourism and transport to Amsterdam.\u0026quot; \u003cu\u003eJournal of sustainable tourism\u003c/u\u003e \u003cstrong\u003e14\u003c/strong\u003e(2): 157-171.\u003c/li\u003e\n\u003cli\u003ePesaran, M. H. (2007). \u0026quot;A simple panel unit root test in the presence of cross-section dependence.\u0026quot; \u003cu\u003eJournal of Applied Econometrics\u003c/u\u003e \u003cstrong\u003e22\u003c/strong\u003e(2): 265-312.\u003c/li\u003e\n\u003cli\u003ePrideaux, B. (2018). \u0026quot;Tourism and surface transport.\u0026quot; \u003cu\u003eThe Sage Handbook of Tourism Management. London: Sage\u003c/u\u003e: 297-313.\u003c/li\u003e\n\u003cli\u003eSong, H., R. T. Qiu and J. Park (2023). \u0026quot;Progress in tourism demand research: Theory and empirics.\u0026quot; \u003cu\u003eTourism Management\u003c/u\u003e \u003cstrong\u003e94\u003c/strong\u003e: 104655.\u003c/li\u003e\n\u003cli\u003eSong, H. and S. F. Witt (2006). \u0026quot;Forecasting international tourist flows to Macau.\u0026quot; \u003cu\u003eTourism Management\u003c/u\u003e \u003cstrong\u003e27\u003c/strong\u003e(2): 214-224.\u003c/li\u003e\n\u003cli\u003eSong, H., S. F. Witt and G. Li (2009). \u003cu\u003eThe Advanced Econometrics of Tourism Demand\u003c/u\u003e. New York, Routledge.\u003c/li\u003e\n\u003cli\u003eTian, F., Y. Yang and L. Jiang (2022). \u0026quot;Spatial spillover of transport improvement on tourism growth.\u0026quot; \u003cu\u003eTourism Economics\u003c/u\u003e \u003cstrong\u003e28\u003c/strong\u003e(5): 1416-1432.\u003c/li\u003e\n\u003cli\u003eUdeagha, M. C. and E. Muchapondwa (2023). \u0026quot;Striving for the United Nations (UN) sustainable development goals (SDGs) in BRICS economies: The role of green finance, fintech, and natural resource rent.\u0026quot; \u003cu\u003eSustainable Development\u003c/u\u003e.\u003c/li\u003e\n\u003cli\u003eVirkar, A. R. and P. D. Mallya (2018). \u0026quot;A conceptual paper on the dimensions of the road transport system that influence tourist satisfaction.\u0026quot; \u003cu\u003eInternational Journal of Research in Social Sciences\u003c/u\u003e \u003cstrong\u003e8\u003c/strong\u003e(3): 296-303.\u003c/li\u003e\n\u003cli\u003eWang, F., L. Ma, B. Gao and Y. S. Liu (2023). \u0026quot;Does high-speed railway affect the cost behavior of tourism firms? Evidence from China.\u0026quot; \u003cu\u003eTourism Economics\u003c/u\u003e: 13548166221150698.\u003c/li\u003e\n\u003cli\u003eYang, L., L. Hu, P. Shang and Y. Ye (2021). \u0026quot;Estimating the Impacts of High-speed Rail on Service Industry Agglomeration in China: Advanced Modelling with Spatial Difference-in-difference Models and Propensity Score Matching.\u0026quot; \u003cu\u003eJournal of Transport Economics and Policy (JTEP)\u003c/u\u003e \u003cstrong\u003e55\u003c/strong\u003e(1): 16-35.\u003c/li\u003e\n\u003cli\u003eYu, L. (2021). \u0026quot;Study on treatment effects and spatial spillover effects of Beijing\u0026ndash;Shanghai HSR on the cities along the line.\u0026quot; \u003cu\u003eThe Annals of Regional Science\u003c/u\u003e \u003cstrong\u003e67\u003c/strong\u003e(3): 671-695.\u003c/li\u003e\n\u003cli\u003eYu, M. and W. Fan (2018). \u0026quot;Accessibility impact of future high speed rail corridor on the piedmont Atlantic megaregion.\u0026quot; \u003cu\u003eJournal of Transport Geography\u003c/u\u003e \u003cstrong\u003e73\u003c/strong\u003e: 1-12.\u003c/li\u003e\n\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":false,"hideJournal":false,"highlight":"","institution":"","isAcceptedByJournal":true,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"transportation","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"port","sideBox":"Learn more about [Transportation](http://link.springer.com/journal/11116)","snPcode":"11116","submissionUrl":"https://submission.nature.com/new-submission/11116/3","title":"Transportation","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"em","reportingPortfolio":"Springer Hybrid","inReviewEnabled":true,"inReviewRevisionsEnabled":false},"keywords":"Taiwan High Speed Rail, Taiwan Tourist Shuttle service, Dynamic SDID, CIPS test, Spillover effect, Spatial econometric analysis","lastPublishedDoi":"10.21203/rs.3.rs-5219377/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-5219377/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eThis study uses the Dynamic Spatial Difference-in-Differences model (Dynamic SDID) to analyze the impact of the Taiwan High-Speed Rail (THSR) on Taiwan's tourism demand. To control for spillover effects, the model incorporates the Taiwan Tourist Shuttle service (TSHU) as an alternative transportation option, the interactive effects between TSHU and THSR, and the spatial autocorrelation between TSHU and THSR. The analysis results indicate that controlling for spillover effects is crucial for analyzing the impact of the High-Speed Rail and tourist transit service on Tourism Demand, and the Dynamic SDID is a better analytical model for this purpose. The THSR has a significant positive impact on tourism demand, while its spatial autocorrelation effect is significantly negative. This suggests that the increase in tourist traffic brought about by THSR mainly comes from existing tourists in the surrounding areas rather than generating new tourism demand. The TSHU, on the other hand, has a negative but insignificant impact on tourism demand, but its interaction with THSR has a significant positive effect, indicating that the two services complement each other. Therefore, to enhance Taiwan's tourism demand, the focus should still be on improving the attractiveness of tourist destinations rather than solely relying on the construction of the High-Speed Rail. Additionally, while the TSHU does not contribute significantly to the development of specific individual tourist destinations, it does facilitate regional tourism development. Therefore, selecting TSHU routes based on actual market conditions can promote the growth of the tourism industry.\u003c/p\u003e","manuscriptTitle":"Using a Dynamic Spatial Difference-in-Differences estimator to evaluate the effect of High speed rail and tourist transit service on Tourism Demand","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2024-11-18 04:50:10","doi":"10.21203/rs.3.rs-5219377/v1","editorialEvents":[{"type":"communityComments","content":0},{"type":"decision","content":"Revision requested","date":"2024-11-08T16:09:33+00:00","index":"","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2024-11-05T19:31:31+00:00","index":"hide","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2024-10-18T14:03:56+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"82493999558969085307710823423857137164","date":"2024-10-11T08:10:38+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"336520267632907585566507829561081526742","date":"2024-10-10T05:06:39+00:00","index":"hide","fulltext":""},{"type":"reviewersInvited","content":"","date":"2024-10-09T16:07:41+00:00","index":"","fulltext":""},{"type":"editorAssigned","content":"","date":"2024-10-09T08:54:39+00:00","index":"","fulltext":""},{"type":"checksComplete","content":"","date":"2024-10-08T13:17:35+00:00","index":"","fulltext":""},{"type":"submitted","content":"Transportation","date":"2024-10-07T15:30:10+00:00","index":"","fulltext":""}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"transportation","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"port","sideBox":"Learn more about [Transportation](http://link.springer.com/journal/11116)","snPcode":"11116","submissionUrl":"https://submission.nature.com/new-submission/11116/3","title":"Transportation","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"em","reportingPortfolio":"Springer Hybrid","inReviewEnabled":true,"inReviewRevisionsEnabled":false}}],"origin":"","ownerIdentity":"a991b19a-f06e-4061-a1a8-5d4f57c596c3","owner":[],"postedDate":"November 18th, 2024","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"published-in-journal","subjectAreas":[],"tags":[],"updatedAt":"2024-12-23T16:08:11+00:00","versionOfRecord":{"articleIdentity":"rs-5219377","link":"https://doi.org/10.1007/s11116-024-10576-4","journal":{"identity":"transportation","isVorOnly":false,"title":"Transportation"},"publishedOn":"2024-12-19 15:57:57","publishedOnDateReadable":"December 19th, 2024"},"versionCreatedAt":"2024-11-18 04:50:10","video":"","vorDoi":"10.1007/s11116-024-10576-4","vorDoiUrl":"https://doi.org/10.1007/s11116-024-10576-4","workflowStages":[]},"version":"v1","identity":"rs-5219377","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-5219377","identity":"rs-5219377","version":["v1"]},"buildId":"qtupq5eGEP_6zYnWcrvyt","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}