Charting Pollution Effects on Tourism: A Regional Analysis

Abstract

This study presents empirical findings highlighting the role of pollution control measures in shaping the trajectory of local tourism market development. Analysis of provincial-level panel data underscores the pronounced impact of water pollution compared to air pollution. While water pollution tends to manifest as a localized issue, air pollution transcends provincial boundaries, posing challenges that extend across multiple regions simultaneously. The results demonstrate how imperative it is for northern China’s heavily polluted provinces to redouble efforts aimed at ameliorating their negative image within the tourism market. In contrast to their southern counterparts, these regions currently face hurdles in attracting tourists, underscoring the disparity in tourism acceptance between northern and southern provinces. Drawing insights from the successful implementation of tourism initiatives centered on the small-town economy in southern China, this research advocates for a paradigm shift in policy formulation for northern provinces. By leveraging lessons learned from southern counterparts, policymakers can chart a course toward sustainable tourism development tailored to the unique characteristics of each region. In delineating the differential impacts of air and water pollution on the Chinese tourism market, this study contributes to a nuanced understanding of tourism dynamics across provinces. The findings serve as a foundational framework for guiding future tourism market development strategies tailored to the heterogeneous landscape of Chinese provinces.

1. Introduction

Tourism development is an attractive option for local economies transitioning from manufacturing-focused production to more diversified service-oriented businesses [1,2]. In emerging markets, the emphasis on production often results in environmental challenges, such as factory emissions and logistics-related pollution [3,4]. These issues become more pronounced when economic growth is prioritized over environmental protection, highlighting the importance of corporate social responsibility (CSR) [5]. The costs associated with environmental protection can be significant. In emerging markets, political pressures often favor reinvestment and income growth over addressing long-term environmental costs [6]. This shortsighted approach can hinder tourism development, which relies on sustainable practices to attract visitors and diversify local income sources.

Tourists prioritize the environment and safety when planning their trips. Many seek to explore different cultures, traditions, and natural landscapes, often driven by curiosity or recommendations from acquaintances [7,8]. Tourism satisfaction is influenced by numerous factors, including the availability of attractions, quality of hospitality services, and environmental and public health conditions at the destination [9,10,11].

This study focuses on China, a market historically centered on manufacturing, to analyze the impact of pollution on tourism consumption. China has developed its infrastructure after opening up its economy for foreign capital and encouraging international trade [12]. The well-developed infrastructures originally designed for manufacturing and trade, including highways, airports, and the railway system, benefit the tourism industry as well. Most of the nearby city tourism attractions have well-connected highway, bus, and coach services [13]. The fast-developed high-speed railway connects different cities and significantly reduces inter-city travel time [14]. The airports in major cities have hundreds of planned flights to connect each other, and almost all smaller provincial-level cities have local airports [15].

China faces substantial air and water pollution due to its prioritization of economic development over sustainability [16,17]. Since 2010, there has been a growing recognition of the adverse effects of pollution on public health and environmental safety, prompting stricter pollution control measures [18,19,20]. Policies include higher emission standards, penalties for violations, and support for green innovations [21,22,23]. The introduction of green bonds and loans aims to reduce the credit risk of environmentally conscious firms and support green transitions in heavy industries [24,25]. Local governments are increasingly viewing tourism development as a key sector for future growth, as it can enhance environmental quality and income, thereby addressing pollution issues [26].

Despite existing studies on pollution and tourism sustainability in China, most focus on environmental impacts rather than tourism consumption; our research fills in this gap. This research provides empirical evidence on the differential effects of air and water pollution control on tourism consumption across various Chinese provinces. We show that tourists recognize the negative effects of air and water pollution differently. We also argue that the southern Chinese provinces have made some successful strategies to combine the local culture and characteristics with environmental protection and pollution control to successfully attract visitors. This research aims to offer policy recommendations for improving tourism development and attractiveness through effective pollution control.

2. Literature Review and Hypotheses

2.1. Pollution and Tourism Decisions

Pollution significantly threatens tourism development [27,28]. Tourists, regardless of their specific interests, seek clean, hygienic, and safe environments for their holidays. Pollution not only degrades the tourism environment but also poses health risks to visitors. As local governments aim to boost economic growth, particularly in smaller cities, maintaining a positive local reputation is crucial for attracting tourists and fostering a thriving tourism industry. Thus, pollution control has become a critical issue for local authorities.

Tourists, like other consumers, exhibit significant attention bias, prioritizing easily observable factors over less apparent ones. Both air and water pollution negatively impact tourism attitudes [29,30], but water pollution is more directly noticeable to visitors. Air pollution, however, can affect broader regions due to wind dispersion, making its effects more diffuse and less directly attributable to a single destination [31,32,33]. When regions are similar, variations in water pollution become more apparent and significantly influence tourism reputation and satisfaction [34,35,36].

Air pollution typically stems from transportation and industrial emissions [37,38], reflecting structural issues inherent to manufacturing-focused cities [39,40]. Water pollution often results from illegal industrial discharges [41,42]. During China’s high-growth period, low penalties for environmental violations led to increased future public health costs, prompting a shift towards emphasizing environmental protection and green innovation [43,44]. Local governments now recognize the need for sustainable development, viewing ecotourism as a viable path.

While both air and water pollution control are vital, air pollution requires broader macro-policies for industrial upgrades. Water pollution control involves stricter regulatory enforcement and monitoring. At the provincial level, water pollution control tends to be more effective and visible to tourists. Therefore, we propose:

H1. Water pollution control contributes more to local tourism development than air pollution control.

2.2. Tourism Reputation and Image

Pollution levels vary across China, with the northern region experiencing greater pollution due to its concentration of heavy industries [45]. Southern regions, engaged in lighter industries and with a historical emphasis on pollution control, fare better ecologically [46]. The geographical advantages of southern and southeastern China facilitate international trade and diversified economic development while maintaining environmental protection.

Southern Chinese cities, particularly those along the Yangtze River, have well-preserved historical sites and traditional cultures [47,48]. Local governments have promoted village and small-city tourism, enhancing the tourism infrastructure and services in these areas [49]. The milder climate in southern China also attracts tourists, especially those seeking to escape the harsh northern winters [50,51].

The development of the high-speed railway reduces inter-city travel time and increases the attractiveness of the small towns for both domestic and international visitors [52,53]. The water in the city is part of the historical characteristics of southern Chinatown [54]. Many small southern cities enlarge their cultural attractiveness by controlling water pollution and providing a better water environment [55]. Those small towns usually have a long history and unique southern Chinese architecture and yards. It becomes a unique experience for foreigners who want to experience old Chinese culture and attracts people living in larger cities to enjoy a relaxed weekend [56]. Such small-town tourism development also helps the local rural commercial and local product marketing developments [57].

Given the environmental, cultural, and climatic advantages of southern China, we hypothesize:

H2. On average, the tourism economy is better developed in southern China than in northern China.

3. Data and Methodology

3.1. Data

After opening up its economy, China started developing its travel market and introducing Chinese culture to foreign countries to increase its international impact. The country has abundant travelling resources, and Chinese history attracts foreigners for a complete taste. The number of foreign visitors visiting China has steadily increased, and the consumption of travelers benefits the local economy of many Chinese cities.

Table 1. Number of foreign visitors.

	2011	2012	2013	2014	2015	2016	2017	2018	2019
Number of foreign visitors (tens of thousands)	2711.2	2719.16	2629.03	2636.08	2598.54	3148.38	4294.2	4795.11	4913.36

Resource: China Statistical Yearbook.

below shows the total number of foreign visitors per year during our sample period, and Figure 1 shows the clear increase trend.

The consumption behavior data of foreign tourists visiting China are collected from the China Statistical Yearbook. Foreign visitors were chosen to minimize destination preference biases seen in local tourists, who might travel based on factors unrelated to environmental quality. The sample period spans 2011 to 2019, excluding 2015 due to missing data for several provinces.

Note that, in line with related research on the relationship between the environment and tourism in China [58,59], the period following the COVID-19 outbreak of 2020 is excluded from our analysis. China, like many other countries around the world, implemented domestic and international travel bans between 2020 and 2023 [60] and the tourism industry in China was decimated to the point that it has since required government subsidies to remain afloat [61]. Domestically, lockdowns in many Chinese cities made travel impossible and even in areas where travel was still legal, unnecessary travel such as tourism was naturally avoided during the pandemic [62,63]. Across the country, hotels in China were repurposed as hospital wards or quarantine facilities to deal with the public health crisis [64]. International travel was perhaps even more strictly prohibited. In March 2020, China’s Ministry of Foreign Affairs announced a ban on international visitors to China and rendered all previously issued travel visas invalid [65]. Even after visa applications were allowed again from 15 March 2023 [66], complicated, rigorous COVID-19 testing was required and long quarantines were strictly enforced upon arrival. This disruption of domestic and international travel certainly distorts the underlying relationships between environmental factors and tourism, making that period inappropriate for empirical analysis. Furthermore, data on the number of foreign tourists visiting China between 2020 and 2023 is unavailable.

The final dataset is an imbalanced panel sample. Air pollution is measured by sulfur dioxide emissions, and water pollution by chemical oxygen demand (COD), both adjusted for monetary output from the secondary sector. Higher values indicate more severe pollution. To address endogeneity, we use the year-over-year first differences in pollution indicators. Chinese tourism authorities classify attractions with “A” ratings, ranging from one “A” to five “A”s, based on visitor feedback and numbers.

Table 2. Variable definitions.

Variable	Symbol	Variable Treatment
Percentage of the total cost allocated to direct travel expenses, for example, entrance tickets	Travel	Directly observable from database
Percentage of the total cost allocated to accommodation	Accm	Directly observable from database
Number of five “A” tourism attractions	Five	Directly observable from database
Number of four “A” tourism attractions	Four	Directly observable from database
Number of three “A” tourism attractions	Three	Directly observable from database
Number of two “A” tourism attractions	Two	Directly observable from database
Number of one “A” tourism attractions	One	Directly observable from database
Number of total tourism attractions	Total	Directly observable from database
Number of five-star hotels	Fivestar	Directly observable from database
Number of four-star hotels	Fourstar	Directly observable from database
Number of three-star hotels	Threestar	Directly observable from database
Air pollution indicator	Air	The yearly first difference (later minus earlier) of COD/total sector two output in yuan
Water pollution indicator	Water	The yearly first difference (later minus earlier) of sulfur dioxide/total sector two output in yuan
Total tourism cost per day	Spend	Directly observable from database
Whether the province is located in northern China	North	Dummy variable: if the province is located in northern China, it equals 1

defines the variables used in the methodology and

Table 3. Summary Statistics.

Statistic	N	Mean	St. Dev.	Min	Pctl(25)	Pctl(75)	Max
Travel	248	5.265	1.954	1.600	3.800	6.400	13.800
Accm	248	13.662	3.854	5	10.9	15.9	32
Five	248	6.661	4.108	0	3	9	24
Four	248	87.907	56.070	6	48	118	269
Three	248	120.698	102.847	7	50.8	162.2	657
Two	248	65.500	64.152	0	19.8	93.8	396
One	248	3.645	5.867	0	0	4	28
Total	248	284.411	201.954	29	126	395.5	1292
Fivestar	246	24.150	24.300	0.000	8.250	26.750	107.000
Fourstar	248	75.351	42.891	10	44	107	188
Threestar	248	161.617	98.161	20	95	199.5	567
Air	248	−14.585	43.008	−300.101	−11.308	−0.377	71.580
Water	247	−3.613	11.336	−93.864	−3.327	−0.111	20.190
Spend	248	199.828	30.125	151.630	179.800	211.835	309.660
North	248	0.484	0.501	0	0	1	1

provides summary statistics of all the data used in the analysis to follows.

The final data is on the provincial level, from the year 2011 to 2019, excluding the missing data in the year 2015. The final pooled data in 8 sample years constitute the provincial-level panel data with a total of 248 observations. It is worth emphasizing that since the pollution data is the year-over-year first difference adjusted for monetary output if the later year pollution is smaller, the later year minus the earlier year would be a negative number. This explains the mean value of the ‘air’ and “water” pollution terms both have a negative mean value, such negative mean value indicating better overall provincial-level environmental protection.

This sample period was chosen because of the fast development of local transportation and infrastructure, which better hospitalizes foreign visitors. The Chinese high-speed train started its trial operation around the year 2010. The initial stage was the short-distance inter-city high-speed train travel between Beijing and Tianjin. In early 2010, the Guangzhou to Wuhan section started operating, and the formal expansion of the high-speed railway network started. The high-density railway network is built around the Yangtze River, Pearl River, and Bohai Bay region within the next eight years. Such a high-speed railway network reduced the inter-city travel time among the above regions and connected all satellite cities to the three regional centers: Shanghai, Shenzhen, and Beijing. The reduced travel time increases the accessibility of small cities and, crucially, helps the development of the regional tourism market [67,68].

3.2. Methodology

To analyze the causal relationship between pollution and direct travel expenses, as well as the effect of pollution control on accommodation costs and overall tourist consumption, we will estimate six specified regression models. According to Hypothesis 1, water pollution is expected to have a more highly statistically significant impact on tourism travel than air pollution. In the equations to follow, this would be reflected in a larger and more statistically significant coefficient estimate of the main parameter of interest, ${\beta }_1$, on the independent variable, ${Water}_{i,t}$, in Equations (1), (3) and (5), than on the independent variable, ${Air}_{i,t}$ in Equations (2), (4) and (6), below.

The first empirical tests concern the causal relationship between pollution and direct travel expenses. The test logic follows the past evidence which indicates the better environment improves the tourism revenues [69]. Equations (1) and (2) examine the relationship between pollution control and direct travel expenses, using province and time controls to account for local and temporal variations, isolating the effect of pollution on tourism-related variables. Accommodation sharing, price, and services could influence the tourism attraction [70]. The famousness or reputation and number of the local tourism attractions within a close distance could also affect the number of foreign visitors. If the number of tourist attractions is large, the overcrowding risk is low, and it increases the level of overall tourism satisfaction [71]. The number of tourism attractions with different ratings is also included as a control variable to account for the abundance of tourism resources, while the star ratings of hotels are used as a control variable to represent local hospitality service levels.

$$\begin{gathered} {Travel}_{i,t }={\beta }_0+{\beta }_1{Water}_{i,t}+{\beta }_2{Five}_{i,t}+{\beta }_3{Four}_{i,t}+{\beta }_4{Three}_{i,t}+{\beta }_5{Two}_{i,t}+{\beta }_6{One}_{i,t}++{\beta }_7{Fivestar}_{i,t} \\ +{\beta }_8{Fourstar}_{i,t}+{\beta }_9{Threestar}_{i,t}+\sum IND+\sum Year+{\epsilon }_{i,t} \end{gathered}$$

(1)

$$\begin{gathered} {Travel}_{i,t }={\beta }_0+{\beta }_1{Air}_{i,t}+{\beta }_2{Five}_{i,t}+{\beta }_3{Four}_{i,t}+{\beta }_4{Three}_{i,t}+{\beta }_5{Two}_{i,t}+{\beta }_6{One}_{i,t}++{\beta }_7{Fivestar}_{i,t} \\ +{\beta }_8{Fourstar}_{i,t}+{\beta }_9{Threestar}_{i,t}+\sum IND+\sum Year+{\epsilon }_{i,t} \end{gathered}$$

(2)

Next, to test the effect of environmental conditions on tourism, we examine a behavior related to tourists’ preferences, specifically whether tourists choose to spend a night in a province. If many tourists prefer to stay in a given province, accommodation costs are expected to increase. Equations (3) and (4) reflect the relationships between environmental quality and the accommodation fees charged to visitors.

$$\begin{gathered} {Accm}_{i,t }={\beta }_0+{\beta }_1{Water}_{i,t}+{\beta }_2{Five}_{i,t}+{\beta }_3{Four}_{i,t}+{\beta }_4{Three}_{i,t}+{\beta }_5{Two}_{i,t}+{\beta }_6{One}_{i,t}++{\beta }_7{Fivestar}_{i,t} \\ +{\beta }_8{Fourstar}_{i,t}+{\beta }_9{Threestar}_{i,t}+\sum IND+\sum Year+{\epsilon }_{i,t} \end{gathered}$$

(3)

$$\begin{gathered} {Accm}_{i,t }={\beta }_0+{\beta }_1{Air}_{i,t}+{\beta }_2{Five}_{i,t}+{\beta }_3{Four}_{i,t}+{\beta }_4{Three}_{i,t}+{\beta }_5{Two}_{i,t}+{\beta }_6{One}_{i,t}++{\beta }_7{Fivestar}_{i,t} \\ +{\beta }_8{Fourstar}_{i,t}+{\beta }_9{Threestar}_{i,t}+\sum IND+\sum Year+{\epsilon }_{i,t} \end{gathered}$$

(4)

Finally, we explore Hypothesis 2, that on average, the tourism economy is better developed in southern China than in northern China. Northern China is more focused on heavy industry. Despite local government investments and efforts to control pollution, the region, which includes famous tourist destinations like Beijing, has a less favorable reputation compared to southern China. Many studies indicate a negative regional ecological system reputation in the northern Chinese provinces [72]. This is particularly evident when discussing pollution and environmental issues, as the overall perception among visitors is negative. Compared with the north, southern Chinese provinces strategically developed their tourism market in smaller cities around the big cities, and the average travel time from the big cities are significantly reduced after the high-speed railway appears, even the benefits to different provinces could differ [73]. The southern provinces also melt their unique cultural and architectural characteristics into the urban planning and such well-designed small cities become highly attractive to not only foreign visitors, but also the residents from the big cities around those touristic small towns [74]. In our analysis, we exclude the individual province control and introduce a dummy variable, ${North}_{i,t}$, that takes the value of one for provinces located in the north. The key parameter estimates of interest are ${\beta }_2$, the coefficient estimate on the ${North}_{i,t}$ dummy variable, and the interaction term between the ${North}_{i,t}$ dummy variable and the pollution indicators discussed above: ${Water}_{i,t}$ and ${Air}_{i,t}$. We expect that northern provinces will exhibit lower overall tourist consumption, so that the parameter estimates for ${\beta }_2$ are negative with statistical significance, indicating that pollution controls in the Northern regions of China are ineffective in increasing tourism consumption as foreign visitors tend to prefer southern China regardless. Equations (5) and (6) illustrate the corresponding tests.

$$\begin{gathered} {Spend}_{i,t }={\beta }_0+{\beta }_1{Water}_{i,t}+{\beta }_2{North}_{i,t}+{\beta }_3{Five}_{i,t}+{\beta }_4{Four}_{i,t}+{\beta }_5{Three}_{i,t}+{\beta }_6{Two}_{i,t}+{\beta }_7{One}_{i,t} \\ +{\beta }_8{Fivestar}_{i,t}+{\beta }_9{Fourstar}_{i,t}+{\beta }_{10}{Threestar}_{i,t}+{\beta }_{11}\left({Water}_{i,t}*{North}_{i,t}\right)+\sum Year+{\epsilon }_{i,t} \end{gathered}$$

(5)

$$\begin{gathered} {Spend}_{i,t }={\beta }_0+{\beta }_1{Air}_{i,t}+{\beta }_2{North}_{i,t}+{\beta }_3{Five}_{i,t}+{\beta }_4{Four}_{i,t}+{\beta }_5{Three}_{i,t}+{\beta }_6{Two}_{i,t}+{\beta }_7{One}_{i,t} \\ +{\beta }_8{Fivestar}_{i,t}+{\beta }_9{Fourstar}_{i,t}+{\beta }_{10}{Threestar}_{i,t}+{\beta }_{11}\left({Air}_{i,t}*{North}_{i,t}\right)+\sum Year+{\epsilon }_{i,t} \end{gathered}$$

(6)

4. Results

4.1. The Main Results

The results of our test of Hypothesis 1, based on an estimation of Equations (1) and (2), are presented in

Table 4. Impact of water and air pollution on direct travel cost percentages.

	Dependent Variable
	Travel
	(1)	(2)	(3)
Water	−0.020 *	−0.023 **
	(0.012)	(0.011)
Air			−0.003
			(0.003)
Five	−0.061	−0.051	−0.056
	(0.046)	(0.106)	(0.107)
Four	0.001	0.005	0.005
	(0.004)	(0.007)	(0.007)
Three	−0.002	0.001	0.001
	(0.002)	(0.003)	(0.003)
Two	−0.003	−0.002	−0.003
	(0.003)	(0.005)	(0.005)
One	−0.012	0.032	0.034
	(0.020)	(0.032)	(0.032)
Fivestar	−0.036 ***	−0.021 *	−0.019
	(0.008)	(0.011)	(0.011)
Fourstar	0.009 *	−0.001	−0.003
	(0.006)	(0.010)	(0.010)
Threestar	0.002	0.006 *	0.005
	(0.002)	(0.003)	(0.003)
Constant	4.995 ***	4.584 ***	4.922 ***
	(0.405)	(1.266)	(1.266)
Province	N	Y	Y
Time	Y	Y	Y
Observations	246	246	246
R2	0.285	0.535	0.527
Adjusted R2	0.236	0.427	0.418
Residual Std. Error	1.713 (df = 229)	1.482 (df = 199)	1.495 (df = 199)
F Statistic	5.718 *** (df = 16; 229)	4.975 *** (df = 46; 199)	4.820 *** (df = 46; 199)

Note: ***, **, and * denote the statistical significance at the 1%, 5%, and 10% levels; standard errors are shown in parentheses.

. Recall from Table 2 above that the independent variables of interest, “Water” and “Air” indicate the change in the amount of water and air pollution, respectively. In the summary statistics in Table 3 above, we saw that on average, both types of pollution are decreasing, so the mean observation in both cases is a negative number. Thus, the negative and, once we control for province-level fixed effects, statistically significant coefficient estimates on water pollution reported in columns (1) and (2) indicate that as water pollution is reduced over time in the sample, there is an increase in the dependent variable, “Travel”, which represents the percentage of total direct travel costs in that province.

In Column (3), the coefficient estimate on air pollution is also negative, albeit statistically insignificant, aligning with our initial expectations. Air pollution tends to be a regional issue spanning multiple provinces rather than confined to a single province. Consequently, foreign visitors may feel compelled to visit regions affected by air pollution due to factors beyond their control.

Overall, the results reported in Table 4 support Hypothesis 1, that water pollution control makes a more significant contribution to local tourism development than does air pollution control, which is a more regional issue.

Table 5. Impact of water and air pollution on accommodation cost percentages.

	Dependent Variable
	Accm
	(1)	(2)	(3)
Water	−0.034 *	−0.039 **
	(0.020)	(0.017)
Air			−0.006
			(0.005)
Five	−0.103	−0.224	−0.231
	(0.079)	(0.166)	(0.167)
Four	−0.006	0.020 *	0.020 *
	(0.007)	(0.011)	(0.011)
Three	−0.009 ***	−0.005	−0.006
	(0.004)	(0.005)	(0.005)
Two	0.016 ***	−0.001	−0.001
	(0.005)	(0.007)	(0.008)
One	−0.049	0.127 **	0.130 **
	(0.034)	(0.050)	(0.050)
Fivestar	0.071 ***	0.013	0.016
	(0.013)	(0.018)	(0.018)
Fourstar	0.012	−0.006	−0.007
	(0.010)	(0.016)	(0.016)
Threestar	−0.011 ***	−0.005	−0.005
	(0.003)	(0.005)	(0.005)
Constant	12.828 ***	9.626 ***	10.135 ***
	(0.697)	(1.978)	(1.978)
Province	N	Y	Y
Time	Y	Y	Y
Observations	246	246	246
R2	0.456	0.708	0.703
Adjusted R2	0.418	0.641	0.634
Residual Std. Error	2.946 (df = 229)	2.316 (df = 199)	2.336 (df = 199)
F Statistic	12.012 *** (df = 16; 229)	10.495 *** (df = 46; 199)	10.238 *** (df = 46; 199)

Note: ***, **, and * denote the statistical significance at the 1%, 5%, and 10% levels; standard errors are shown in parentheses.

presents the results of the empirical estimation of Equations (3) and (4), which might be interpreted as a robustness test of Hypothesis 1. Here, the dependent variable, which above in Table 4 was direct travel to the province as a percentage of total travel costs, is replaced with the accommodation cost in the province as a percentage of total travel costs. The parameter estimates on the change in water pollution continue to be negative in column (1), and statistically significant in column (2) after controlling for province-level fixed effects. Again, confirming the results reported above, the parameter estimates on air pollution continue to exhibit negative coefficients, albeit with statistical insignificance. These findings corroborate the outcomes of our baseline model reported in Table 4.

We next turn to a consideration of Hypothesis 2, which considers the heterogeneities of tourism attractions in northern and southern China.

Table 6. Pollution in northern China.

	Dependent Variable
	Spend
	(1)	(2)
Water	−0.068
	(0.200)
Air		0.034
		(0.065)
North	−9.548 ***	−10.548 ***
	(3.525)	(3.583)
Five	−1.282 *	−1.412 **
	(0.663)	(0.665)
Four	−0.203 ***	−0.199 ***
	(0.059)	(0.059)
Three	0.093 ***	0.092 ***
	(0.027)	(0.027)
Two	0.076 **	0.081 **
	(0.036)	(0.036)
One	0.645 **	0.594 **
	(0.258)	(0.260)
Fivestar	0.630 ***	0.629 ***
	(0.101)	(0.101)
Fourstar	0.079	0.076
	(0.073)	(0.074)
Threestar	−0.056 **	−0.053 **
	(0.026)	(0.026)
Water * North	0.486 *
	(0.255)
Air * North		0.052
		(0.071)
Constant	187.817 ***	187.600 ***
	(5.491)	(5.446)
Time	Y	Y
Observations	238	238
R2	0.482	0.477
Adjusted R2	0.440	0.434
Residual Std. Error (df = 219)	21.968	22.089
F Statistic (df = 18; 219)	11.338 ***	11.081 ***

Note: ***, **, and * denote the statistical significance at the 1%, 5%, and 10% levels; standard errors are shown in parentheses.

reports the results of the estimations using Equations (5) and (6). The results demonstrate that when individual provincial controls are omitted, both water and air pollution cease to be independently significant. However, the dummy variable indicating all provinces in the north of China, “North”, exhibits a statistically significantly negative coefficient estimate, indicating that tourists’ per-night spending is lower in northern provinces with statistical significance. Notably, both water and air pollution levels are more pronounced in northern China compared to southern regions, leading to a corresponding negative disparity in tourism consumption as indicated by the regional dummy variable.

Furthermore, the interaction term between northern China and the water pollution indicator is positive and significant. This finding suggests that even as environmental conditions improve in northern cities, they are less preferred by visitors. This paradox arises because cities experiencing substantial improvements according to pollution indicators still grapple with more severe pollution issues. Unless sustained long-term improvements occur, and overall environmental quality matches or exceeds that of southern Chinese cities and provinces, changing entrenched visitor perceptions is likely to remain challenging.

4.2. Summary of Findings

The

Table 7. Summary of findings.

Hypotheses	Validation	Reasoning
H1. Water pollution control contributes more to local tourism development than air pollution control.	Supported by regression results.	The results in Table 4 and Table 5 show that water pollution control contributes to the direct travel-related costs of foreign visitors. Since the variable “water” is the first difference, the higher water pollution protection is indicated by a negative number, so the negative statistically significant coefficients in columns (1) and (2) in both Table 4 and Table 5 show a positive marginal effect to the dependent variable, the direct travel related costs, and accommodation costs.
H2. On average, the tourism economy is better developed in southern China than in northern China.	Supported by regression results.	In Table 6, the dummy variable “North” has significant negative coefficients, and the interactive coefficient of “North” with “Water” is positive and significant. Such results indicate that tourism consumption per day is lower in northern provinces. The local pollution image is deep in visitors’ minds, and a higher level of water protection does not alleviate such a bad image. The attractiveness from the aspect of tourism consumption in northern China is much underperformed if compared with southern China.

below summrizes the finds of the paper. The short reasons are give to explain the validation of the hypotheses.

5. Discussion

The Chinese tourism market has experienced rapid growth following the nation’s initiatives in clean energy development and emphasis on industrial green renovations. Both local governments and residents recognize the unsustainability of prioritizing economic growth at the expense of environmental safety. The relationship between economic growth and the environment experiences the inverted U-shape relation; the increase in living standards motivated the residents to consider the future environment and health [75]. Pollution not only harms ecosystems but also poses significant health risks. The considerable costs required for ecological restoration were often overlooked during periods of excessive economic focus. Effective governance and pollution control efforts necessitate a long-term commitment.

The government’s environmental initiatives and ecological restoration projects are pivotal drivers of tourism industry development. The Chinese tourism policies put sustainability as one of the policy priorities [76]. Local governments perceive environmental pollution control as conducive to tourism development, which, in turn, can enhance local employment and economic growth.

Despite tourism being a relatively low-polluting industry, its development can still impact local environments adversely [77]. While our study demonstrates a preference for destinations with minimal water pollution, the influx of tourists can paradoxically exacerbate local environmental degradation. Moreover, increased tourist numbers strain public transportation systems and affect local residents’ quality of life [78]. The local residents may feel that tourism takes their resources [79]. Sometimes, since the tourists are not familiar with the local cultures and values, their behaviors could make the local residents feel uncomfortable [80]. Local political tensions may arise if competition for resources emerges between residents and tourists, particularly considering the cultural differences between foreign visitors and locals.

To foster harmonious green tourism development, local governments must identify potential risks and communicate both the environmental and economic benefits to residents, seeking their social consensus. Some evidence shows that green innovations have high conversion efficiency in the tourism market [81]. The development of technology has led to electric cars becoming an efficient way of controlling carbon emissions in tourism development [82,83].

Small Chinese cities stand to benefit economically from tourism development, leveraging existing infrastructure and transportation networks [84]. China’s extensive high-speed rail network and well-connected airports facilitate travel, while hotels cater to both business and leisure visitors in the more economically developed regions [84]. Enhancing local transportation services can further improve tourist experiences and attract more visitors [85].

Southern China’s allure, owing to favorable weather and environmental conditions, persists. Efforts by northern cities to shed their image of heavy pollution have been less successful, compounded by the challenges of industrial pollution control, particularly in heavy industry-dominated regions. Conversely, southern regions, especially the southeastern coastal areas, have adopted strategies centered on small-town economies, leveraging their cultural heritage to offer unique experiences to visitors. Luxurious accommodations and local businesses thrive, catering to city dwellers seeking a slower-paced, authentic experience. Some of the nearby small cities collaborate together to increase their overall influence and attract more visitors [86,87]. The local rural tourism developments in southern China reduce the local city and rural differences [88].

In contrast, northern Chinese cities boast significant tourist attractions but struggle to alter negative perceptions. The past heavy pollution has a negative image in people’s minds, and travelers have become cautious when making destination decisions [89]. Developing distinctive cultural tourism experiences could be a viable approach, highlighting the unique cultures and traditions of regions like Inner Mongolia, the Ningxia Hui Autonomous Region, and the Xinjiang Uyghur Autonomous Region. Improving ecological environments and showcasing cultural diversity may help attract tourists and reshape environmental perceptions positively.

6. Conclusions and Directions for Future Research

This study, utilizing provincial-level data, contributes empirical evidence indicating that water pollution holds paramount importance for foreign tourists when making travel decisions in China. Conversely, while air pollution remains a concern, its impact is attenuated by its occurrence across larger regions spanning multiple provinces, resulting in a less discernible effect at the provincial level. Notably, our findings underscore a substantial disparity in average tourist consumption between northern and southern China, with the latter exhibiting a more conducive environment for tourism development.

Southern China’s strategic advantage lies in its thoughtful planning, particularly its implementation of the small-town economic strategy. This approach has facilitated successful economic transitions, transforming small towns previously reliant on light manufacturing, which posed environmental risks, into vibrant tourism hubs. Such initiatives not only bolster local economic sustainability but also contribute to the restoration of ecological environments.

Water protection plays an important role in Southern China’s culture and small-town tourism strategies. Unlike air pollution, which usually involves multiple provinces or even national-level protection requirements, water pollution protection has smaller affected regions, and when there is an economic development incentive, the nearby municipals could collaborate and develop the local tourism market. The smaller towns around the Suzhou region provide the best examples of collaborated regional tourism development [49]. The small towns are well-designed to show the local culture, and the tourism development helps to develop other tourism-related businesses [90]. Green innovations, including the development of electric cars, have reduced carbon emission levels and further extended the development of the local tourism market [91].

The current research focuses on the effect of water and air pollution on the decision-making of tourism consumers regarding consumption. However, from the perspective of local sustainability development, the impact of tourism helps the local economic growth but also negatively affects the local environment. For instance, even better water pollution control attracts tourists, but the tourists could significantly increase the water quality risk, especially if the number of tourists is large. Such environmental risk is usually non-linear and increases quickly if the number exceeds a certain threshold level. Also, if the data allows, there is a need to consider the negative tourist overcrowding effect should be considered when evaluating the consumption expenditure and incentives. Overcrowding would lower the incentive for tourists to visit.

Ideally, future research on this topic would expand to include an analysis of the carbon footprint as an environmental indicator. While this would add depth to future assessments of the environmental impacts of tourism, data availability is a challenge. Perhaps more realistically, future research avenues could explore domestic tourism preferences, given their prominence in tourism consumption. However, delineating these preferences presents challenges, notably due to seasonal migrations of northern residents to southern regions for climatic reasons, which complicates the distinction between tourism and non-tourism activities. Moreover, the prevalence of business travelers among domestic tourists further complicates this delineation. Nevertheless, the burgeoning trend of small-town tourism offers an avenue to disentangle leisure and business travelers, given the predominance of leisure-oriented visits to such destinations. By delving into domestic tourism preferences, local governments can refine their tourism policies, fostering sustainable and inclusive tourism development initiatives.