Does world heritage site initiation promote tourism? A difference-in-difference approach

Abstract

Extant studies have found that the inscription of World Heritage sites (WHS) has no or negative effect on tourism in China, which does not explain why Chinese local governments are eager to apply for WHS. This study employs the difference-in-difference approach and city-level data from China to assess the effect of WHS over a more extended period, starting from initiation instead of inscription. The results indicate that WHS initiation can promote tourism, while WHS inscription reduces tourism revenue. Furthermore, heterogeneity analysis reveals that WHS initiation has a more positive effect on domestic tourism revenue in the eastern region of China. In contrast, WHS inscription has a more negative effect on tourism revenue in the western region.

Keywords

China difference-in-difference world heritage world heritage initiation world heritage inscription

Introduction

The impact of World Heritage Sites (WHS) inscription on tourism has been the subject of previous research. Many studies have found that WHS inscription has no or negative effect on tourism in China. For instance, Huang et al. (2012) reported that WHS inscription does not influence long-term tourism development. Yan and Morrison (2008) used case studies to demonstrate that the positive effect of the inscription on tourist arrivals is not particularly pronounced. Empirical evidence from Gao and Su (2019) reveals that, despite China having the second largest number of WHS in the world, such inscriptions do not increase tourism revenue or visitor arrivals.

However, the Chinese government attaches great importance to economic development. Given that the increase in the gross domestic product (GDP) is one of the critical indicators for promoting local officials, using WHS to build famous tourist attractions and develop tourism becomes a primary motivation for the Chinese government to pursue WHS inscription (Vigneron, 2016), and the government puts in a lot of effort for the application (Zhang, 2017; Zhang and Brown, 2022). For example, during the promotion meeting for the application of the WHS at the Zhouyuan Heritage and Qinyongcheng Heritage, the deputy mayor of Baoji City emphasized three aspects of the application’s significance: creating Baoji’s historical and cultural logo, promoting the integration of culture, tourism, and business, and accelerating the construction of a sub-center. Similarly, at the second phase of the China Yellow (Bohai) Sea Migratory Bird Habitat (Phase II) WHS Application Work Conference, the mayor of Dalian City highlighted that the application for WHS should be considered as an important means to accelerate the construction of ecological civilization and promote high-quality district development. However, the fact that the inscription has no effect or negative impact on tourism (Gao and Su, 2019; Huang et al., 2012; Yan and Morrison, 2008) contradicts the original intent behind local governments’ applications for WHS. The existing literature has not explained why local government officials continue to show enthusiasm for WHS applications despite the focus on GDP.

According to the Convention Concerning the Protection of the World Cultural and Natural Heritage, WHS should have outstanding universal value. Local governments invite experts with adequate research expertise to explore cultural and natural attributes and values for a World Heritage submission and hope the outstanding universal value they find and the successful application of WHS can boost tourism (Zhang and Brown, 2022). However, the government’s hope of tourism exploitation brings conflict with property protection (Drost, 1996; Landorf, 2009; Nicholas et al., 2009; Zhang et al., 2015). Research found that old management models, inappropriate tourism operations, and ineffective regulation led to this conflict (Zhang et al., 2015). Successful inscription of WHS means that local governments must comply with requirements of the United Nations Educational, Scientific and Cultural Organization (UNESCO), one of which is protecting, maintaining, and preserving WHS. This requirement aggravates the conflict and results in a drop in tourism revenue (Gao and Su, 2019). The reason for this inconsistency of purpose and result is worth further research.

To examine the impact of WHS more comprehensively, it is important to consider a longer period, specifically starting from the initiation stage rather than just the inscription. WHS initiation refers to the time when the government decides to begin preparing for WHS application. It occurs before the submission of declaration documents to UNESCO and marks the beginning of the planning stage. Previous literature has primarily focused on the effect of inscription (Cuccia et al., 2016; Gao and Su, 2019; Huang et al., 2012) but has overlooked that the preparation process for WHS is lengthy and uncertain, which may also affect tourism. According to UNESCO’s regulations, the duration from document submission to the announcement of new inscription results typically spans approximately 18 months. However, the time required for preparation varies significantly between cities in China, ranging from three to 13 years. Therefore, this study aims to assess this extended and uncertain process, starting from initiation to the present, to gain deeper insights into its impact on tourism and reevaluate the impact of inscription.

Our research offers three distinct theoretical contributions. First, we provide an analysis of WHS with a longer time period, starting from initiation to the present. To the best of our knowledge, this is the first empirical study that examines the impact of WHS initiation on tourism revenue while also comparing it to the relationship between WHS inscription and tourism development. Previous scholars have extensively studied the impact of inscription (Cuccia et al., 2016; Gao and Su, 2019; Huang et al., 2012; Patuelli et al., 2013), but they have overlooked the fact that WHS inscription does not occur overnight, and the preparation process is time-consuming and challenging. Our study aims to explain why the Chinese government eagerly applies for WHS, even though WHS inscription has no or negative effect on tourism, particularly under the focus on GDP.

Second, our research compares the impact of WHS initiation and inscription in regions that exhibit varying levels of economic development. Previous studies have demonstrated that tourism development can be influenced by several factors, including the level of economic development and the availability of tourism infrastructure (Arezki et al., 2012; Chaudhary, 2000; Said et al., 2020). Given the notable disparity in economic development across China, examining the differential impact of WHS initiation and inscription on tourism across regions is imperative.

Third, our research investigates the impact of WHS in China, a country renowned for its rich cultural history, diverse landscapes, and rapid economic development. As the world’s largest developing nation, China places significant emphasis on economic growth and has incorporated GDP as one of the metrics for evaluating government officials, which differs from the Western context. Consequently, local governments hold high expectations for utilizing WHS to stimulate tourism. By extensively contextualizing our study, we expand the scope of research on WHS to a novel and unique context.

Related literature and hypotheses

The existing literature on how WHS initiation influences tourism is insufficient. While some scholars argue that WHS can effectively boost tourism in the short term, leading to the government’s willingness to apply for it (Alzua et al., 1998; Herbert, 2001; Vigneron, 2016), limited research supports this expectation. Only Li et al. (2020) focused on WHS initiation in China, specifically examining the case of Tulou. When the local government initiates the application preparation work, various stakeholders, including governments, residents, and tourists, begin planning and implementing the necessary steps for the WHS application (Yang et al., 2009). For example, the local government places significant emphasis on economic benefits rather than protection and long-term development, seeking professional suggestions and evaluations from tourism planners and scholars. This involves building holiday resorts, designing roads to the heritage site, and determining locations for guest hotels and public toilets (Li et al., 2020). Residents seek additional financial resources by running small businesses and working in local tourism companies, shifting away from traditional farming work (Li et al., 2020). During this application preparation process, the infrastructure level, commercialization level, and public awareness of the candidate heritage site experienced significant improvement. Consequently, initiated sites witness rapid tourism development compared to attractions that have not commenced application preparation. This transformation turns ordinary scenic spots into famous ones, attracting more tourists and enhancing tourism revenue. Based on interviews and questionnaire investigation conducted by Li et al. (2020), both domestic and international tourists show a willingness to experience the unique culture of Tulou, spend more money on souvenirs, and provide positive feedback on their trip to Tulou. Therefore, WHS initiation is viewed as a viable strategy for stimulating the local tourism economy. Building on this understanding, we propose the following hypotheses:

H1a: The initiation of WHS increases the domestic tourism revenue of cities.

H1b: The initiation of WHS increases the foreign tourism revenue of cities.

Scholars have conducted extensive research on the impact of WHS inscription on tourism. Based on the Convention concerning the Protection of the World Cultural and Natural Heritage adopted in 1972, countries are mandated to safeguard WHS. As a result, governments prioritize conservation efforts over tourism development and may implement protective measures such as limiting the number of tourists to prevent heritage oxidation and corrosion. They may also temporarily close attractions for repairs, which can potentially lead to a decrease in domestic tourism revenue. Zhang and Brown (2022) found that residents cannot gain economic or social benefits from WHS, contrarily, they undertake the burden of protection. Gao and Su (2019) found that WHS inscription in China does not promote tourism revenue or tourist arrivals. This is because WHS inscription often leads to the heightened fame of heritage sites, attracting a large influx of tourists. Consequently, the popularity of other tourist attractions in the same city may be overshadowed or replaced (Gao et al., 2022). This can result in a reduction in tourist inflows and revenue for the city. Frey and Steiner (2011) argue that the WHS inscription is an additional honor for originally famous tourist attractions. An Italian study also shows that despite the cultural and environmental advantages that can enhance the competitiveness of tourism destinations, WHS inscription may have a negative impact (Patuelli et al., 2013). Therefore, we propose the following hypotheses:

H2a: The inscription of WHS decreases the domestic tourism revenue of cities.

H2b: The inscription of WHS decreases the foreign tourism revenue of cities.

Methodology

Sample selection and data sources

To address our research question, we gathered panel data at the city level spanning from 2000 to 2019 from various sources. Our first source included the China Regional Economic Yearbook, the Provincial Statistical Yearbook, and the China Tourism Statistical Yearbook, from which we obtained data on tourism outcomes, such as domestic and foreign tourism revenue. Data on WHS in China was collected from UNESCO’s official Web site. The year of each WHS initiation was extracted from news articles titled like ‘Heritage Site Application Scheduling Meeting’ retrieved from the Bing search engine. We used the earliest news time as the initiation time. To ensure accuracy, all authors conducted separate searches for the same WHS, cross-checked and communicated their findings, and reached a consensus. When a WHS was connected to multiple cities, we only included cities with WHS-related attractions as listed on https://Ctrip.com, one of China’s largest travel websites. Other city characteristic data, such as GDP per capita, were also extracted from the China Stock Market & Accounting Research (CSMAR) database. It is important to note that our analysis was limited to mainland Chinese cities and municipalities due to data constraints. After removing missing data, we constructed an unbalanced panel data set comprising 285 cities from 2000 to 2019.

Variable definitions

Dependent variables

This study considers two dependent variables: domestic tourism revenue (LnDTR_i,t) and foreign tourism revenue (LnFTR_i,t). Both variables are measured in million RMB and transformed into natural logarithms. Specifically, LnDTR_i,t denotes the natural logarithm of the domestic tourism revenue generated by the city i in year t, while LnFTR_i,t corresponds to the natural logarithm of the foreign tourism revenue of the city i in year t. According to the National Bureau of Statistics of China, domestic tourism revenue includes domestic tourists’ spending on transportation, visiting, accommodation, catering, shopping, and entertainment in the process of traveling; and foreign tourism revenue includes inbound tourists’ spending on transportation, visiting, accommodation, catering, shopping, and entertainment during their sightseeing.

Independent variables

The study incorporates Initiation_i,t and Inscription_i,t as independent variables. When exploring the effect of WHS initiation on tourism revenue, it is important to note that the application preparation process is ongoing and long-term. For example, increasing public awareness of site protection, enhancement of tourism management, and growth in heritage popularity during the preparation for WHS can have a lasting impact on tourism revenue beyond WHS application or inscription. Similarly, when examining the impact of WHS inscription on tourism revenue, it is essential to recognize that the effect is persistent. Building upon Beck, Levine, and Levkov’s study (2010), Initiation_i,t is a binary variable that takes on a value of one if the city i in year t is in or after the year of WHS initiation, and 0 otherwise. Inscription_i,t is a binary variable that takes a value of 1 if the city i in year t is in or after the year of WHS inscription, and 0 otherwise.

Control variables

The study incorporates several control variables in its analysis, including the number of 5A-level sites, GDP per capita, high-speed railway and international airport connections, population density, number of employees, general budget expenditure of local finance, the proportion of employees in the tertiary industry, and the number of passengers.

Within empirical investigations of the determinants of tourism demand, GDP per capita has emerged as a reliable measure of local income, offering a comprehensive depiction of regional development (Albalate et al., 2017). Additionally, population density is utilized to gauge the intensity and connectivity of human activity, thereby generating a range of spillover effects that entice tourists to visit and explore (Gao et al., 2019). Acknowledging the significant contribution of public expenditure to the creation of appealing and populous tourist destinations, the provision of public support is critical to the establishment of tourist amenities and the preparation for WHS inscription (Mules, 2005). Finally, in the present study, we have incorporated two transportation variables, specifically airport and railway connections, which are important predictors of tourism outcomes (Gao and Su, 2019).

In this study, we have employed the natural logarithm to calculate several key variables, including GDP per capita and population density. For the presence of a high-speed railway in the city i during year t, a binary indicator variable is utilized, where the high-speed railway connection takes a value of 1 if present and 0 otherwise. Similarly, the presence of an international airport in the city i during year t is represented by a binary variable, where the international airport connection takes a value of 1 if present and 0 otherwise. The number of employees is measured in ten thousand units. The general budget expenditure of local finance and the number of passengers is measured in ten thousand units and calculated using the natural logarithm.

Models

The present study utilizes the Difference-in-Difference (DID) approach to evaluate the impact of WHS initiation and inscription on tourism. The DID approach is suitable for scenarios where specific groups are subjected to treatment while others are not. It applies when there are two groups and two periods. In the first period, neither group is subjected to treatment. In the second period, only one group is exposed to treatment while the other is not. The implementation of DID involves computing two differences between group means in a particular way. The first difference is the difference between the mean of the outcome variable across the two time periods for each group. The second difference measures the variation in outcome change between the two groups, representing the explanatory variable’s causal impact. The identification strategy of DID relies on the assumption that group-specific means differs in the absence of treatment. However, if this difference remains constant over time (without any treatment), it can be eliminated by subtracting the group-specific means of the outcome variable under consideration. Thus, the remaining difference between these group-specific differences indicates the causal impact of interest (Schwerdt and Woessmann, 2020). In this study, when assessing the impact of initiation, all cities are divided into two groups: cities that initiate WHS and cities that do not initiate WHS. Similarly, when assessing the impact of inscription, all cities are also divided into two groups: cities with WHS inscription and cities without WHS inscription. Furthermore, using fixed-effect estimation can help mitigate some issues from omitted variables. Accordingly, we estimate the following regression model:

y_{i, t} = α_{0} + α_{1} x_{i, t} + α^{'} Z_{i, t} + C i t y_{i} + Y e a r_{t} + ε_{i, t}

(1)

In this equation, the dependent variable y_i,t is either LnDTR_i,t or LnFTR_i,t, which are the natural logarithms of domestic and foreign tourist revenue of the city i in year t, respectively. As previously explained, the independent variable x_i,t can be either Initiation_i,t or Inscription_i,t. The vector Z _i,t, as described earlier, represents the control variables. The city-fixed effect and year-fixed effect are denoted by City_i and Year_t, respectively. ε_i,t represents the error term. The parameter α₁ estimates the effect of initiation or inscription on tourism revenue.

Results

Descriptive statistics

The summary statistics for the entire sample are presented in Table 1. The mean value of the natural logarithm of domestic tourism revenue for all cities is 8.622, and the mean value of the natural logarithm of foreign tourism revenue for all cities is 2.978.

Table 1.

Summary statistics of whole sample.

Variables	Mean	S.D.	Min	Max
Domestic tourism revenue	8.622	1.746	3.908	12.59
Foreign tourism revenue	2.978	2.030	0.0100	8.518
5A-level sites	0.302	0.744	0	9
GDP per capita	10.07	0.893	7.920	12.05
High-speed railways connection	0.244	0.430	0	1
International airports connection	0.208	0.406	0	1
Population density	5.738	0.860	2.894	7.723
Number of employees	40.41	47.36	5.360	382.3
General budget expenditures	13.88	1.255	11.02	17.37
Employees in the third industry	52.34	13.29	19.07	85.60
Number of passengers	8.527	0.936	5.861	11.49

Baseline results

Table 2 presents the baseline findings derived from equation (1) estimation. The results indicate a statistically significant increase in domestic tourism revenue from WHS initiation at a 5% significance level, as shown in column (1). However, no discernible impact on foreign tourism revenue was detected, as demonstrated in column (2), which is consistent with Zhang, Cheng, and Zhang’s study (Zhang et al., 2023). Additionally, the findings reveal that the inscription of WHS leads to a statistically significant reduction in domestic tourism revenue at a 5% significance level, as indicated in column (3), while having no discernible effect on foreign tourism revenue, as demonstrated in column (4). Consequently, hypotheses H1a and H2a should be accepted, whereas hypotheses H1b and H2b should be rejected. These findings suggest that WHS initiation can promote tourism development, but tourism revenue decreases once a site has been inscribed.

Table 2.

Baseline results from DID.

Variables	(1)	(2)	(3)	(4)
Variables	Domestic tourism revenue	Foreign tourism revenue	Domestic tourism revenue	Foreign tourism revenue
Initiation	0.3073^**	0.0254
	(0.149)	(0.134)
Inscription			−0.1442^**	0.1234
			(0.069)	(0.141)
5A-level sites	−0.0550^**	−0.0218	−0.0340	−0.0231
	(0.026)	(0.044)	(0.026)	(0.043)
GDP per capita	0.4497^***	0.2881^**	0.4436^***	0.2876^**
	(0.066)	(0.124)	(0.066)	(0.124)
High-speed railways connection	−0.0401	0.1175^*	−0.0365	0.1186^*
	(0.033)	(0.063)	(0.033)	(0.063)
International airports connection	0.0042	−0.1969	0.0198	−0.1992
	(0.092)	(0.144)	(0.094)	(0.145)
Population density	0.0427	−0.1333^**	0.0423	−0.1374^**
	(0.032)	(0.066)	(0.033)	(0.067)
Number of employees	−0.0031^***	0.0001	−0.0031^***	0.0001
	(0.001)	(0.001)	(0.001)	(0.001)
General budget expenditures	0.2384^***	0.0524	0.2317^***	0.0535
	(0.082)	(0.124)	(0.082)	(0.124)
Employees in the third industry	0.0059^***	0.0050	0.0057^***	0.0052
	(0.002)	(0.004)	(0.002)	(0.004)
Number of passengers	−0.0092	−0.0058	−0.0044	−0.0056
	(0.023)	(0.045)	(0.023)	(0.045)
Year-fixed effect	Yes	Yes	Yes	Yes
City-fixed effect	Yes	Yes	Yes	Yes
(Constant)	−0.4467	−0.7350	−0.3299	−0.7343
	(1.126)	(1.699)	(1.124)	(1.698)
N	3520	3746	3520	3746
R²	0.905	0.611	0.904	0.611
Adj. R²	0.904	0.608	0.904	0.608
Number of cities	285	285	285	285

Note: Values in the table are unstandardized regression coefficients. Standard errors are in parentheses. ^*p < .1, ^**p < .05, ^***p < .01.

Robustness tests

Parallel trends test

The parallel trend assumption is the prerequisite for applying the DID method to achieve causal inference. This assumption asserts that the treatment and control groups exhibit the same trajectory before the natural experiment (Schwerdt and Woessmann, 2020). If dissimilarities exist between the treatment and control groups before the experiment, the DID results may not accurately depict the effect of the experiment, and other factors may likely influence the experiment’s outcome. To assess the dynamics of the treatment effect, we employed the approach proposed by Beck et al. (2010) and Flammer and Kacperczyk (2016). We included interactions between time and experiment dummy variables in the regression model. Non-significant coefficients of the interactions before the experiment indicate the presence of parallel trends. The parallel trends test evaluates the variation between two groups, where one period is the benchmark, and all other periods are compared with this benchmark. Since the current period and all periods after the experiment are affected by the experiment, they cannot serve as benchmarks. To overcome this issue, we implemented dummy variables to represent 4 years before and after the treatment, using our study’s first year of the sample time as the benchmark.

Table 3 presents the findings of the parallel trends test. Column (1) reveals that the periods preceding the initiation time are statistically insignificant, whereas those following this period are statistically significant, with positive coefficients. On the other hand, column (3) demonstrates that periods before the inscription time are insignificant, but those lasting 3 years or more after the inscription time are significant, with negative coefficients. These results indicate that the parallel trend assumption holds.

Table 3.

Parallel trends test results.

Variables	(1)	(2)	(3)	(4)
	Initiation		Inscription
	Domestic tourism revenue	Foreign tourism revenue	Domestic tourism revenue	Foreign tourism revenue
−4	0.3578	0.2390	0.0470	0.0880
	(0.337)	(0.155)	(0.078)	(0.098)
−3	0.2243	0.1516	−0.0184	0.0439
	(0.300)	(0.144)	(0.080)	(0.096)
−2	0.4383	0.0486	−0.0644	0.1073
	(0.319)	(0.186)	(0.091)	(0.111)
−1	0.4992	0.0266	−0.0930	−0.1928
	(0.327)	(0.174)	(0.082)	(0.208)
0	0.4957^*	0.0864	−0.0889	−0.1746
	(0.283)	(0.166)	(0.083)	(0.182)
+1	0.5946^**	0.1883	−0.1266	0.1362
	(0.284)	(0.187)	(0.108)	(0.184)
+2	0.5749^**	0.1517	−0.1387	0.1762
	(0.290)	(0.197)	(0.122)	(0.185)
+3	0.4908^*	0.1268	−0.3449^***	0.1990
	(0.283)	(0.202)	(0.116)	(0.188)
+4 (+)	0.5079^*	0.0932	−0.2213^*	0.3716^*
	(0.287)	(0.224)	(0.119)	(0.191)
Control variables	Yes	Yes	Yes	Yes
Year-fixed effect	Yes	Yes	Yes	Yes
City-fixed effect	Yes	Yes	Yes	Yes
(Constant)	−0.3734	−0.7076	−0.3142	−0.7616
	(1.121)	(1.704)	(1.122)	(1.695)
N	3520	3746	3520	3746
R²	0.905	0.611	0.905	0.613
Adj. R²	0.905	0.608	0.904	0.610
Number of cities	285	285	285	285

Note: Values in the table are unstandardized regression coefficients. Standard errors are in parentheses. ^*p < .1, ^**p < .05, ^***p < .01.

Placebo test

The present study includes distinct initiation and inscription times for each city in the treatment group. We employed a random selection approach to assigning cities to the treatment group. We repeated this process 500 times to assess the statistical significance of the coefficients of the fabricated initiation and inscription dummy variables. The analysis evaluated the association between WHS initiation and domestic tourism revenue and the relationship between WHS inscription and domestic tourism revenue.

Figure 1 depicts that the majority of coefficients of the fabricated initiation and inscription dummy variables are either equal to or in proximity to zero, differing from the actual estimate of the DID method. As their p-values exceed 10%, these coefficients lack statistical significance. Consequently, our results are highly improbable to have occurred by chance and are improbable to have been influenced by extraneous policies or arbitrary factors.

Figure 1.

Placebo test of the impact of WHS’ initiation and inscription on domestic tourism revenue. Note: The figure plots the effect of fake WHS’ initiation and inscription on the natural logarithm of domestic tourism revenue. The left vertical coordinate axis represents the natural logarithm of domestic tourism revenue, and the right vertical coordinate axis represents the p-value. The curve is the kernel density distribution of the estimated coefficients, blue dots are p-values corresponding to the estimated coefficient, and the horizontal blue dashed line is the significance level of 10%. Figure 1 (1) shows WHS initiation results on domestic tourism revenue, and the red vertical dashed line is the true estimated value of the DID method (0.3073). Figure 1 (2) is the results of WHS inscription on domestic tourism revenue, the red vertical dashed line is the true estimated value of the DID method (−0.1442).

OLS test

We employed the OLS method to validate our DID results. Specifically, we established two dummy variables, Ini_i,t and Ins_i,t, and incorporated them into a single regression. The variable Ini_i,t indicates whether the city i initiates WHS preparation work in year t. If so, it is denoted as 1; otherwise, it is 0. The variable Ins_i,t signifies whether the site in the city i has been inscribed for WHS in year t. If yes, it is denoted as 1; otherwise, it is 0. The findings using the OLS method are presented in Table 4. Column (1) reveals that WHS initiation significantly and positively impacts domestic tourism revenue, while WHS inscription negatively influences domestic tourism revenue. Furthermore, column (2) suggests that both initiation and inscription do not affect foreign tourism revenue. Hence, these are consistent with the DID results.

Table 4.

OLS test results.

Variables	(1)	(2)
Variables	Domestic tourism revenue	Foreign tourism revenue
Ini _i,t	0.3202^**	0.0157
	(0.146)	(0.131)
Ins _i,t	−0.1635^**	0.1226
	(0.065)	(0.141)
5A-level sites	−0.0523^**	−0.0239
	(0.025)	(0.044)
GDP per capita	0.4484^***	0.2881^**
	(0.066)	(0.124)
High-speed railways connection	−0.0422	0.1183^*
	(0.033)	(0.063)
International airports connection	0.0089	−0.1998
	(0.090)	(0.145)
Population density	0.0442	−0.1374^**
	(0.032)	(0.067)
Number of employees	−0.0031^***	0.0001
	(0.001)	(0.001)
General budget expenditures	0.2361^***	0.0539
	(0.082)	(0.124)
Employees in the third industry	0.0058^***	0.0052
	(0.002)	(0.004)
Number of passengers	−0.0082	−0.0059
	(0.023)	(0.045)
Year-fixed effect	Yes	Yes
City-fixed effect	Yes	Yes
(Constant)	−0.4137	−0.7413
	(1.117)	(1.698)
N	3520	3746
R²	0.905	0.611
Adj. R²	0.904	0.608
Number of cities	285	285

Note: Values in the table are unstandardized regression coefficients. Standard errors are in parentheses. ^*p < .1, ^**p < .05, ^***p < .01.

Alternative measures for dependent variables

To verify our results, we use domestic and foreign tourism arrivals to measure tourism impact. The data about domestic and tourism arrivals is from the Provincial Statistical Yearbook and the China Tourism Statistical Yearbook. We repeat the exercise and report results in Table 5, showing that WHS initiation positively affects domestic tourism revenue, while WHS inscription negatively affects domestic tourism revenue.

Table 5.

Robustness tests by measuring tourism impact as tourism arrivals.

Variables	(1)	(2)	(3)	(4)
Variables	Domestic tourism arrivals	Foreign tourism arrivals	Domestic tourism arrivals	Foreign tourism arrivals
Initiation	0.1828^**	0.2585
	(0.089)	(0.194)
Inscription			−0.1589^**	0.0835
			(0.081)	(0.157)
Controls	Yes	Yes	Yes	Yes
Year-fixed effect	Yes	Yes	Yes	Yes
City-fixed effect	Yes	Yes	Yes	Yes
(Constant)	4.6910^***	−5.2059^**	4.983^***	−0.5099^**
	(0.966)	(2.234)	(0.893)	(2.231)
N	3055	3055	3236	3236
R²	0.888	0.553	0.888	0.552
Number of cities	282	282	279	279

Note: Values in the table are unstandardized regression coefficients. Standard errors are in parentheses. ^*p < .1, ^**p < .05, ^***p < .01.

Alternative measure to determine initiation years

To verify the impact of WHS initiation on tourism, we use the year of inscription in the Tentative List as the starting point of initiation to conduct a robustness check. The Tentative List of World Heritage Sites in China is assessed and published by the National Cultural Heritage Administration, and submitted to the UNESCO World Heritage Centre. Table 6 reports the results that use the year of inscription in the Tentative List as the starting point of initiation.

Table 6.

Robustness tests by using the year of inscription in the Tentative List as the starting point of initiation.

Variables	(1)	(2)
Variables	Domestic tourism revenue	Foreign tourism revenue
Initiation	0.2651^**	0.1861
	(0.142)	(0.144)
Controls	Yes	Yes
Year-fixed effect	Yes	Yes
City-fixed effect	Yes	Yes
(Constant)	0.2450	−0.4076
	(1.281)	(1.129)
N	3520	3746
R²	0.899
Number of cities	285	285

Note: Values in the table are unstandardized regression coefficients. Standard errors are in parentheses. ^*p < .1, ^**p < .05, ^***p < .01.

Using DID with multiple time periods as robustness tests

As the time periods of treatment in our study differ, the effect of participating in the treatment may vary with the length of exposure to the treatment. To verify our results, we use DID with multiple time periods referring to Callaway and Sant’Anna’s study (2021). We control the same covariates as in the main results, but exclude the number of passengers, employees, or GDP per capita to ensure that covariates are independent. Table 7 shows the results. From Rows of Single Parameters, we can see that the results of DID with multiple time periods are consistent with those of standard DID.

Table 7.

Robustness tests by using DID with multiple time periods.

Initiation- domestic tourism revenue
	Simple weighted average		Group-specific effects			Event study			Calendar time effects
Partially aggregated			g = 2001	0.640^***	(-0.049)	e = 0	0.075	(-0.051)	g = 2001	0.016	(-0.029)
			g = 2004	0.318^***	(-0.097)	e = 1	−0.039	(-0.058)	g = 2002	0.109^***	(-0.034)
			g = 2005	0.415^***	(-0.048)	e = 2	0.079^**	(-0.077)	g = 2003	0.272^***	(-0.049)
			g = 2006	−0.053	(-0.035)	e = 3	0.109	(-0.104)	g = 2004	0.392^***	(-0.088)
			g = 2007	0.414^*	(-0.032)	e = 4	0.133^*	(-0.146)	g = 2005	0.426^***	(-0.11)
			g = 2009	−0.004	(-0.074)	e = 5	0.139	(-0.138)	g = 2006	0.342^**	(-0.142)
			g = 2013	0.051^**	(-0.017)	e = 6	−0.039	(-0.147)	g = 2007	0.408^***	(-0.088)
			g = 2015	0.041^*	(-0.03)	e = 7	0.326^*	(-0.183)	g = 2008	0.313^**	(-0.185)
			g = 2017	0.037	(-0.069)	e = 8	0.298	(-0.21)	g = 2009	0.109^*	(-0.106)
						e = 9	0.225	(-0.325)	g = 2010	0.144	(-0.123)
						e = 10	0.565^**	(-0.227)	g = 2011	−0.096	(-0.124)
						e = 11	0.588^**	(-0.233)	g = 2012	0.122^*	(-0.132)
						e = 12	0.899^***	(-0.071)	g = 2013	0.097	(-0.127)
						e = 13	0.828^***	(-0.074)	g = 2014	0.128^**	(-0.147)
						e = 14	0.796^***	(-0.081)	g = 2015	0.111^*	(-0.126)
						e = 15	0.561^***	(-0.348)	g = 2016	0.162	(-0.235)
						e = 16	0.608^**	(-0.293)	g = 2017	0.390^**	(-0.19)
						e = 17	0.211^*	(-0.078)	g = 2018	−0.135	(-0.037)
						e = 18	0.079	(-0.093)	g = 2019	0.049	(-0.064)
Single parameters	0.154^*	(-0.111)		0.070^**	(-0.031)		0.310^***	(-0.054)		0.189^**	(-0.078)
Partially aggregated			g = 2001	−0.062	(-0.069)	e = 0	0.074	(-0.082)	g = 2001	−0.03	(-0.092)
			g = 2004	−0.348^***	(-0.134)	e = 1	0.109	(-0.098)	g = 2002	0.023	(-0.125)
			g = 2005	1.269^***	(-0.162)	e = 2	0.034	(-0.137)	g = 2003	−0.1	(-0.086)
			g = 2006	−0.766^***	(-0.086)	e = 3	0.031	(-0.155)	g = 2004	−0.037	(-0.261)
			g = 2007	−0.758^***	(-0.052)	e = 4	−0.038	(-0.178)	g = 2005	−0.003	(-0.243)
			g = 2009	0.244^***	(-0.052)	e = 5	−0.002	(-0.218)	g = 2006	−0.022	(-0.248)
			g = 2013	−0.011	(-0.165)	e = 6	0.065	(-0.243)	g = 2007	−0.116	(-0.228)
			g = 2015	0.165^***	(-0.053)	e = 7	0.134	(-0.366)	g = 2008	0.089	(-0.237)
			g = 2017	0.212^***	(-0.032)	e = 8	0.066	(-0.392)	g = 2009	0.032	(-0.152)
						e = 9	0.315	(-0.432)	g = 2010	0.058	(-0.164)
						e = 10	−0.174	(-0.429)	g = 2011	0.031	(-0.182)
						e = 11	0.246	(-0.403)	g = 2012	−0.009	(-0.193)
						e = 12	−0.047	(-0.188)	g = 2013	0.077	(-0.206)
						e = 13	−0.052	(-0.202)	g = 2014	0.231	(-0.248)
						e = 14	−0.013	(-0.082)	g = 2015	0.154	(-0.278)
						e = 15	−0.116	(-0.075)	g = 2016	0.098	(-0.304)
						e = 16	−0.048	(-0.072)	g = 2017	0.088	(-0.283)
						e = 17	0.07	(-0.129)	g = 2018	0.136	(-0.21)
						e = 18	−0.056	(-0.059)	g = 2019	0.091	(-0.138)
Single parameters	0.05	(-0.176)		0.07	(-0.073)		0.004	(-0.015)		0.029	(-0.163)
Partially aggregated			g = 2001	−0.114^*	(-0.062)	e = 0	−0.009	(-0.027)	g = 2001	−0.123^***	(-0.046)
			g = 2003	−0.244^*	(-0.129)	e = 1	−0.027	(-0.031)	g = 2002	−0.252^***	(-0.084)
			g = 2004	0.042	(-0.047)	e = 2	−0.162^***	(-0.06)	g = 2003	−0.131	(-0.288)
			g = 2006	−0.018	(-0.031)	e = 3	−0.106^**	(-0.052)	g = 2004	−0.047^*	(-0.099)
			g = 2007	−0.214^***	(-0.045)	e = 4	−0.083	(-0.056)	g = 2005	0.032	(-0.059)
			g = 2008	0.335^**	(-0.027)	e = 5	−0.066^**	(-0.087)	g = 2006	−0.013	(-0.048)
			g = 2009	−0.195^***	(-0.053)	e = 6	−0.072^*	(-0.108)	g = 2007	−0.059	(-0.061)
			g = 2010	−0.069	(-0.083)	e = 7	−0.009	(-0.136)	g = 2008	−0.032^*	(-0.07)
			g = 2012	−0.090^***	(-0.016)	e = 8	−0.109	(-0.124)	g = 2009	−0.04	(-0.079)
			g = 2014	−0.008	(-0.021)	e = 9	−0.103^*	(-0.141)	g = 2010	−0.113^*	(-0.067)
			g = 2016	−0.092^***	(-0.036)	e = 10	−0.149	(-0.116)	g = 2011	−0.116^**	(-0.059)
						e = 11	−0.133^*	(-0.181)	g = 2012	−0.159^***	(-0.053)
						e = 12	−0.01	(-0.117)	g = 2013	−0.126^*	(-0.067)
						e = 13	−0.066	(-0.095)	g = 2014	−0.052	(-0.061)
						e = 14	0.124	(-0.184)	g = 2015	−0.039^***	(-0.081)
						e = 15	−0.186^*	(-0.106)	g = 2016	−0.0480^**	(-0.078)
						e = 16	−0.321^**	(-0.142)	g = 2017	−0.032^**	(-0.12)
						e = 17	−0.221^***	(-0.079)	g = 2018	−0.05	(-0.045)
						e = 18	−0.361^***	(-0.08)	g = 2019	−0.021	(-0.017)
Single parameters	−0.079^*	(-0.049)		−0.051^***	(-0.016)		−0.063^***	(-0.02)		−0.085^**	(-0.038)
Partially aggregated			g = 2001	−0.033	(-0.058)	e = 0	0.074	(-0.08)	g = 2001	0.198	(0. 201)
			g = 2003	0.650^*	(-0.173)	e = 1	0.082	(-0.095)	g = 2002	−0.173	(-0.197)
			g = 2004	1.049^**	(-0.072)	e = 2	0.074	(-0.155)	g = 2003	0.23	(-0.238)
			g = 2006	0.152^*	(-0.081)	e = 3	0.163	(-0.151)	g = 2004	−0.19	(-0.191)
			g = 2007	0.055	(-0.372)	e = 4	0.223	(-0.176)	g = 2005	0.169^**	(-0.049)
			g = 2008	0.316	(-0.446)	e = 5	0.134	(-0.251)	g = 2006	0.186^**	(-0.042)
			g = 2009	−0.497^**	(-0.075)	e = 6	0.265	(-0.304)	g = 2007	−0.098	(-0.184)
			g = 2010	−0.184^*	(-0.061)	e = 7	0.567^*	(-0.232)	g = 2008	0.27	(-0.37)
			g = 2012	−0.084^*	(-0.033)	e = 8	0.466	(-0.271)	g = 2009	−0.432^*	(-0.247)
			g = 2014	0.18	(-0.078)	e = 9	0.761^**	(-0.252)	g = 2010	−0.348^**	(-0.197)
			g = 2016	−0.054	(-0.049)	e = 10	0.858^**	(-0.319)	g = 2011	0.239	(-0.122)
						e = 11	0.699^*	(-0.294)	g = 2012	−0.220^*	(-0.179)
						e = 12	0.554	(-0.225)	g = 2013	0.206	(-0.18)
						e = 13	−1.016^***	(-0.098)	g = 2014	0.188	(-0.167)
						e = 14	−1.148^***	(-0.1)	g = 2015	0.235	(-0.188)
						e = 15	−0.228	(-0.174)	g = 2016	0.218	(-0.191)
						e = 16	0.065	(-0.085)	g = 2017	0.354	(-0.164)
						e = 17	−0.043	(-0.049)	g = 2018	0.164	(-0.179)
						e = 18	0.056	(-0.058)	g = 2019	0.127	(-0.185)
Single parameters	0.214	(-0.149)		0.16	(-0.082)		−0.005	(-0.023)		0.217	(-0.121)

Note: Values in the table are unstandardized regression coefficients. Standard errors are in parentheses. ^*p < .1, ^**p < .05, ^***p < .01.

Using quantile DID as robustness tests

To provide an enriched analysis, we use Quantile DID to do robustness tests. Given Callaway and Li’s research (2019) and Nchare and Makioka’s research (2023), we add the same covariates as in the main results, including the number of passengers, the number of employees, or GDP per capita, 5-A level sites, high-speed railway and international airport connections, population density, general budget expenditure of local finance, the proportion of employees in the tertiary industry. The results of Quantile DID show that WHS initiation is significantly and positively correlated with domestic tourism revenue (β = 1.4165, p = .000), and has no discernible effect on foreign tourism revenue (β = 0.1078, p = .518). WHS inscription is significantly and negatively correlated with domestic tourism revenue (β = −1.2574, p = .000), and has no significant effect on foreign tourism revenue (β = 0.1181, p = .311). Figure 2 also reports these results.

Figure 2.

Robustness tests by using Quantile DID.

Deleting demand covariates as robustness tests

Referring to Card and Krueger’s study (1993), we consider that the covariates must be independent. So, we delete demand covariates including the number of passengers, the number of employees, and GDP per capita, and repeat the exercise. Table 8 shows the results, indicating that our baseline results of DID still hold.

Table 8.

Robustness tests by deleting demand covariates.

Variables	(1)	(2)	(3)	(4)
Variables	Domestic tourism revenue	Foreign tourism revenue	Domestic tourism revenue	Foreign tourism revenue
Initiation	0.2425^*	−0.0214
	(0.149)	(0.137)
Inscription			−0.1754^**	0.0518
			(0.075)	(0.146)
5A-level sites	−0.0743^***	−0.0097	−0.0754^***	−0.012
	(0.026)	(0.039)	(0.026)	(0.039)
High-speed railways connection	−0.0389	0.0944	−0.0728^**	0.0942
	(0.035)	(0.063)	(0.035)	(0.063)
International airports connection	0.2008^**	−0.2445^*	−0.0090	−0.2483^*
	(0.098)	(0.148)	(0.104)	(0.147)
Population density	0.2506^***	−0.1465^*	0.0030	−0.1484^*
	(0.039)	(0.087)	(0.051)	(0.087)
General budget expenditures	0.7474^***	0.2241^*	0.4984^***	0.2260^*
	(0.076)	(0.087)	(0.090)	(0.118)
Employees in the third industry	0.0068^***	0.0050	0.0081^***	0.0051
	(0.002)	(0.003)	(0.002)	(0.003)
Year-fixed effect	Yes	Yes	Yes	Yes
City-fixed effect	Yes	Yes	Yes	Yes
(Constant)	−4.1881^***	−0.2845	0.2187	−0.3012
	(0.908)	(1.514)	(1.12)	(1.517)
N	3816	4200	3816	4200
R²	0.907	0.622	0.909	0.622
Number of cities	285	285	285	285

Note: Values in the table are unstandardized regression coefficients. Standard errors are in parentheses. ^*p < .1, ^**p < .05, ^***p < .01.

Heterogeneity analysis

Previous studies have revealed that the economic development level of a city has a significant impact on tourism. Merely possessing UNESCO WHS is insufficient to attract tourists to a city; instead, destination competitiveness is necessary to leverage tourism for local economic development (Farid, 2015). Urban branding strategies, such as the development of tourism-related infrastructure, the enrichment of historical heritage, the promotion of local lifestyles (Anjomrouz et al., 2021), and the enhancement of residents’ cultural intelligence (Zaman and Aktan, 2021), also exert considerable influence on boosting the economic contribution of tourism. Conversely, some underdeveloped cities face insufficient tourism infrastructure, support services, local healthcare, and other factors (Ristić et al., 2020; Said et al., 2020) that impede their attractiveness as tourist destinations.

While China has the world’s second-largest economy, there are significant discrepancies in economic development among various regions of the country. The eastern region of China is the most advanced, housing prominent and well-developed cities such as Beijing and Shanghai. In contrast, the western region is relatively less developed. Cities in more developed regions generally possess greater resources to pursue UNESCO WHS inscription, and their governments have a better understanding of how to promote the economy through tourism.

To explore the effect of spatial heterogeneity, this study examines the relationship between the initiation and inscription of WHS on domestic tourism revenue in cities in the eastern, western, and central regions of China. Table 9 presents the results, which show that the initiation for WHS significantly boosts tourism revenue in the eastern region at a significance level of 5% (see column (1)). In contrast, no such effect is observed in the western and central regions (see columns (2) and (3)). Compared with other regions in China, the eastern region has a higher level of economic development, and local governments there have a higher level of management. Therefore, when governments in the eastern region initiate WHS preparation, they have more funding spent on hiring experts to explore sites’ outstanding universal value and have more experience and capability to develop more effective tourism management policies. With good tourism management, sites’ unique characteristics are conserved, sustainable tourism development is conducted, and attractions become more famous, so more and more tourists visit these attractions, and tourism revenue increases.

Table 9.

Results of spatial heterogeneity analysis for domestic tourism revenue.

Variables	(1)	(2)	(3)	(4)	(5)	(6)
Variables	Eastern region	Western region	Central region	Eastern region	Western region	Central region
Initiation	0.4692^**	0.2909	0.0416
	(0.210)	(0.216)	(0.053)
Inscription				−0.0917	−0.3914^**	−0.0458
				(0.108)	(0.155)	(0.077)
Controls	Yes	Yes	Yes	Yes	Yes	Yes
Year-fixed effect	Yes	Yes	Yes	Yes	Yes	Yes
City-fixed effect	Yes	Yes	Yes	Yes	Yes	Yes
(Constant)	0.6686	0.5923	−2.2020	−0.0412	1.3226	0.5071
	(1.545)	(3.619)	(2.692)	(1.632)	(3.621)	(2.023)
N	1485	1011	1030	1535	1011	1074
R²	0.893	0.890	0.938	0.886	0.890	0.944
Adj. R²	0.891	0.887	0.937	0.884	0.887	0.943
Number of cities	121	84	80	121	84	80

Note: Values in the table are unstandardized regression coefficients. Standard errors are in parentheses. ^*p < .1, ^**p < .05, ^***p < .01.

Moreover, the inscription of WHS negatively affects domestic tourism revenue in the western region at a significant level of 5% (see column (5)) but has no significant impact on revenue in the eastern and central regions (see columns (4) and (6)). After the inscription, local governments need to obey the Convention concerning the Protection of the World Cultural and Natural Heritage, protect and preserve WHS, which costs a lot of money. However, for governments in the western region, their financial budget and funding are limited, and external revenue resources are fewer than in other regions, so they may choose some simple and low-cost methods, such as limiting tourist numbers to avoid oxidation and corrosion of heritage, and temporarily closing attractions for repairs, which result in a decline in domestic tourism revenue. Also, in the initiation process, tourism development is sufficient, so it is hard to find a new tourism revenue growth point after the inscription.

Discussion and conclusion

China has the second world’s highest number of World Heritage sites. Despite the Chinese government’s pursuit of economic development (Vigneron, 2016), WHS inscription surprisingly has a negative effect on tourism (Frey and Steiner, 2011; Gao et al., 2022; Gao and Su, 2019). To explore the cause of this contradictory phenomenon, this study focuses on an extended time period of WHS, starting from initiation instead of inscription, and analyzes the impact of WHS initiation and inscription on tourism.

Consistent with the expectations of the Chinese government (Li et al., 2020), our study empirically shows that WHS initiation can positively affect the development of domestic tourism. However, the inscription of WHS is found to have a negative impact on domestic tourism revenue. In contrast, WHS’s initiation and inscription do not significantly affect inbound tourism. Our heterogeneity analysis suggests WHS initiation can promote domestic tourism in the eastern region, whereas WHS inscription decreases tourism revenue in the western region. Our findings suggest that during the preparation process for WHS, the tourism economy is likely to grow, particularly in the eastern region of China. However, following the inscription of WHS, cities in the western region of China are more negatively affected.

Our study makes significant theoretical contributions to the existing literature. Specifically, we provide a comprehensive analysis of WHS in China over a longer period (i.e., from initiation to present). While Li et al. (2020) reported in their case study that the inscription of WHS is a viable strategy for boosting tourism, our study using the DID method reveals that WHS initiation positively impacts tourism for most cities with WHS. Moreover, we contribute to the literature by clearly delineating the negative effect of the inscription on tourism, consistent with the findings of Gao and Su’s study (2019). Furthermore, our study enhances understanding of the impact of WHS initiation and inscription in Chinese regions with varying levels of economic development.

Our study has significant practical implications. The WHS initiation and inscription process is crucial as it has the potential to not only protect WHS but also promote tourism. The development of WHS tourism must be sustainable, as highlighted in previous research (Imon, 2017; Landorf, 2009; Li et al., 2008). Therefore, it is recommended that stakeholders involved in WHS, including local governments, residents, and tourism companies, work collaboratively to achieve a balance between conservation and tourism. This approach will ensure the long-term preservation of WHS while also contributing to the local economy through tourism.

Our research has a notable limitation in terms of generalizability, as our findings were derived solely from data collected in China. Given China’s distinct approach to managing WHS compared to other countries, it is not possible to extrapolate our results to other contexts without further theoretical and empirical investigation. Therefore, additional research is necessary to draw conclusive and generalizable conclusions.

Footnotes

Declaration of conflicting interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

ORCID iDs

Shouming Chen

Ahmed Rabeeu

Yueqi Wang

Author biographies

Yueqi Wang is a full-time PhD student at Tongji University. Her research interest includes tourism development and strategic management.

Shouming Chen is a Full Professor of Innovation & Strategy in the School of Economics and Management at Tongji University in China. He received his PhD. in Management from Fudan University. His research interests include strategic management, innovation, corporate social responsibility, and business ethics. His work has been published in such journals as Journal of Business Ethics, Journal of Business Research, Human Resource Management, Asia Pacific Journal of Management, International Entrepreneurship and Management Journal, Social Indicators Research, and International Journal of Quality & Reliability Management.

Ahmed Rabeeu is a full-time PhD scholar at Tongji University. His research interests include sustainable tourism, smart tourism, heritage tourism and CSR. He holds a Bachelor’s (Hons) degree in Business and Management Studies from Cardiff Metropolitan University, UK and MBA from Shanghai International Studies University, China. He has a proven multinational record with over 11 years of extensive managerial level experience across different industries, having served for the leading utility service provider and manufacturer in the Maldives, an international property management company based in Shanghai, different United Nations agencies and the tourism sector of the Maldives.

References

Albalate

Campos

Jiménez

(2017) Tourism and high speed rail in Spain: does the AVE increase local visitors? Annals of Tourism Research 65: 71–82.

Alzua

O’Leary

Morrison

(1998) Cultural and heritage tourism: identifying niches for international travellers. Journal of Tourism Studies 9(2): 2–13.

Anjomrouz

Dehyadegari

Hosseini

, et al. (2021) The role of urban branding to attract foreign tourists. Independent Journal of Management & Production 12(1): 310–328.

Arezki

Cherif

Piotrowski

(2012). In: Licciardi

Amirtahmasebi

(eds) UNESCO World Heritage List, Tourism, and Economic Growth. Washington: World Bank Inst. Available at: https://www-webofscience-com-s.web.bisu.edu.cn/wos/alldb/full-record/WOS:000327133200009 (accessed 27 February 2022).

Beck

Levine

Levkov

(2010) Big bad banks? The winners and losers from bank deregulation in the United States. The Journal of Finance 65(5): 1637–1667.

Callaway

(2019) Quantile treatment effects in difference in differences models with panel data. Quantitative Economics 10(4): 1579–1618.

Callaway

Sant’Anna

PHC

(2021) Difference-in-Differences with multiple time periods. Journal of Econometrics 225(2): 200–230 (Themed Issue: Treatment Effect 1).

Card

Krueger

(1993) Minimum wages and employment: a case study of the fast food industry in New Jersey and Pennsylvania. American Economic Review 84(4): 772–793.

Chaudhary

(2000) India’s image as a tourist destination - a perspective of foreign tourists. Tourism Management 21: 293–297.

10.

Cuccia

Guccio

Rizzo

(2016) The effects of UNESCO World Heritage List inscription on tourism destinations performance in Italian regions. Economic Modelling 53: 494–508.

11.

Drost

(1996) Developing sustainable tourism for world heritage sites. Annals of Tourism Research 23(23): 479–484.

12.

Farid

(2015) Tourism management in World Heritage Sites and its impact on economic development in Mali and Ethiopia. Procedia - Social and Behavioral Sciences 211: 595–604.

13.

Flammer

Kacperczyk

(2016) The impact of stakeholder orientation on innovation: evidence from a natural experiment. Management Science 62(7): 1982–2001.

14.

Frey

Steiner

(2011) World Heritage List: does it make sense? International Journal of Cultural Policy 17(5): 555–573.

15.

Gao

(2019) Is the World Heritage just a title for tourism? Annals of Tourism Research 14: 102748.

16.

Gao

Wang

(2019) Does high-speed rail boost tourism growth? New evidence from China. Tourism Management 12: 29.

17.

Gao

Zang

(2022) Does Regional Tourism Benefit from the Official Quality Rating of Tourist Attractions? Evidence from China’s Top-grade Tourist Attraction Accreditations. Journal of China Tourism Research. Routledge. Epub ahead of print 3 April 2022. DOI: 10.1080/19388160.2020.1822975.

18.

Herbert

(2001) Literary places, tourism and the heritage experience. Annals of Tourism Research 2(28): 312–333.

19.

Huang

C-H

Tsaur

J-R

Yang

C-H

(2012) Does World Heritage list really induce more tourists? Evidence from Macau. Tourism Management 33: 1450–1457.

20.

Imon

(2017) Cultural heritage management under tourism pressure. Worldwide Hospitality and Tourism Themes 9(3): 335–348.

21.

Landorf

(2009) Managing for sustainable tourism: a review of six cultural World Heritage Sites. Journal of Sustainable Tourism 17(1): 53–70.

22.

Cai

(2008) Tourism development of world heritage sites in China: a geographic perspective. Tourism Management 29: 308–319.

23.

Lau

(2020) Heritage tourism stakeholder conflict: a case of a World Heritage Site in China. Journal of Tourism and Cultural Change 18(3): 267–287.

24.

Mules

(2005) Economic impacts of national park tourism on gateway communities: the case of Kosciuszko National Park. Tourism Economics 11(2): 247–259 (Sage Publications Ltd).

25.

Nchare

Makioka

(2023) Quantile difference in differences with time-varying qualification in panel data. Journal of Econometric Methods 12(1): 105–116.

26.

Nicholas

Thapa

(2009) Residents’ perspectives of a World Heritage Site: the pitons management area, ST. Lucia. Annals of Tourism Research 36(3): 390–412.

27.

Patuelli

Mussoni

Candela

(2013) The effects of World Heritage Sites on domestic tourism: a spatial interaction model for Italy. Journal of Geographical Systems 15(3): 369–402.

28.

Ristić

Vukoičić

Nikolić

, et al. (2020) Tourism value assessment model of ‘UNESCO-listed’ monasteries: Kosovo and Metohija. Current Issues in Tourism 23(17): 2098–2102.

29.

Said

Ab Majid

Pey

, et al. (2020) Ascertaining the business sustainability of heritage properties in Malaysia. International Real Estate Review 23(4), 1163–1185. (Rockville: Global Social Science Inst).

30.

Schwerdt

Woessmann

(2020) Chapter 1 - empirical methods in the economics of education. In: Bradley

Green

(eds) The Economics of Education. 2nd edition. Academic Press, 3–20. Available at: https://www-sciencedirect-com-443.web.bisu.edu.cn/science/article/pii/B978012815391800001X (accessed 28 May 2023).

31.

Vigneron

(2016) From local to World Heritage: a comparative analysis. The Historic Environment: Policy & Practice 7(2–3): 115–132 (Routledge).

32.

Yan

Morrison

(2008) The influence of visitors’ awareness of world heritage listings: a case study of Huangshan, Xidi and Hongcun in Southern Anhui, China. Journal of Heritage Tourism 2(3): 184–195.

33.

Yang

Zhao

(2009) Preliminary study on the national strategy to improve the tentative list for Chinese mixed World Heritage. Chinese Landscape Architecture 6: 9.

34.

Zaman

Aktan

(2021) Examining residents’ cultural intelligence, place image and foreign tourist attractiveness: a mediated-moderation model of support for tourism development in Cappadocia (Turkey). Journal of Hospitality and Tourism Management 46: 1–12.

35.

Zhang

(2017) World heritage listing and changes of political values: a case study in west lake cultural landscape in Hongzhou, China. International Journal of Heritage Studies 23(3): 215–233 (Routledge).

36.

Zhang

Brown

(2022) Benefit or burden? The world heritage listing of libo karst, China. International Journal of Heritage Studies 28(5): 578–596 (Routledge).

37.

Zhang

Fyall

Zheng

(2015) Heritage and tourism conflict within world heritage sites in China: a longitudinal study. Current Issues in Tourism 18(2): 110–136.

38.

Zhang

Cheng

Zhang

(2023) Does World Heritage inscription promote regional tourism? Evidence from China. Tourism Economics 29(4): 929–951.