Analysis of spatial-temporal pattern, dynamic evolution and influencing factors of health tourism development in China

Measurement of development index

Table 2 reports the development index and rate of increase in health tourism for the whole country, three major regions, and 31 provinces from 2010 to 2019. First, from the overall national situation, the health tourism development index improved from 0.12 in 2010 to 0.23 in 2019, with an increased rate of 7.67%, which is a fast development rate and reflects the rising trend of health tourism development in China. The reasons for this: the rapid economic and social development and the rapid rise of middle-income class people stimulate people’s potential consumer demand for health tourism and enhance their invisible consumption ability, and tourism consumption is undergoing structural changes in the process of transformation and upgrading. However, judging from the current development index (0.12–0.23), health tourism still has much room for development, and it is expected to develop further as people’s health awareness continues to increase and government support for health tourism continues to grow. Secondly, from the perspective of the three regions, the average development index and development speed of health tourism are not consistent. In terms of the development index, the development index of eastern China (0.23) is higher than that of central China (0.15), and the development index of central China is higher than that of western China (0.13). The top five provinces in the health tourism development index are all located in eastern China, and four of the bottom five provinces are located in western China. For a long time, eastern China has been the center of gravity of China’s economic development, with great advantages in technology and capital. Although the abundance of resources in central China is higher than that in eastern China, there is still a big gap between the economic and social development of central China and eastern China. Although western China has obvious advantages in resources, the relative backwardness of transportation, communication and other infrastructures largely limits the development level of health tourism. In terms of growth rate, eastern China (6.75%) is lower than central China (7.34%), and central China is lower than western China (8.74%), which reflects the momentum of catching up with relatively advanced regions in the backward area of health tourism development. With the implementation of strategies such as the development of western China and the rise of central China, as well as the decline in the influence of factors such as technology and capital and the increase in the influence of factors such as resources and environment, the development advantages of western China and central China have gradually emerged, and the growth rate has accelerated significantly.

Difference in development index

The previous content has analyzed the development index of health tourism at different spatial scales in China as a whole, three major regions, and 31 provinces from 2010 to 2019. However, it was not possible to identify the difference between the development of health tourism across China as a whole and the three largest regions from the statistical data results, and it was also challenging to accurately determine the direction of further in-depth study after measuring and analyzing. Therefore, with the help of the Theil Index, this paper calculates the difference in the health tourism development index and presents it visually. It mainly calculates the overall difference index of China, the difference index of intra-region and inter-region, and the overall difference index of eastern China, central China, and western China, to prepare for future research based on the measurement results of development difference. The summary results are summarized in Table 3.

Firstly, when looking at China as a whole, the intra-regional difference is always greater than the level of inter-regional difference and maintains a relatively stable development despite fluctuations (Fig. 2). For instance, in 2010 and 2019, the difference between intra-regional and inter-regional was 0.046 and 0.048, respectively. Meanwhile, the intra-regional difference in 2010 was 0.087, accounting for 68.36%, and the inter-regional difference was 0.040, accounting for 31.64%. In 2019, the intra-regional difference was 0.081, accounting for 70.99%, and the inter-regional difference was 0.033, accounting for 29.01%. Based on the above data results, it can be seen that the intra-regional difference occupies a dominant position in the whole study period and has a weak strengthening trend to the end of the investigation, while the influence of inter-regional difference is somewhat weakened. With the rapid development of health tourism, eastern China has strengthened the transformation and upgrading of its industrial structure mainly based on its economic advantages, western China has strengthened the utilization of its good natural foundation and abundant resources, and central China has strengthened its interaction and cooperation with the eastern China and western China, which has made the impact of inter-regional difference among eastern China, central China, and western China become smaller. At the same time, the impact of intra-regional differences plays a major role due to the similarity of social environment, economic situation and resource base within each region of China, which is more likely to be affected by other factors such as policies of provinces and municipalities. In addition, from 2010 to 2012, the change trend of the overall difference was consistent with the intra-regional difference, while the change trend of the overall difference was opposite to the inter-regional difference, and after 2012, the change trend of the overall difference was generally consistent with the inter-regional difference and opposite to the intra-regional difference. Before 2012, China and its relevant departments still paid limited attention to health tourism, and the policy support was relatively low, which did not form a strong development momentum and joint force. The Ministry of Health issued the “Health China 2020” strategic research report in 2012, and the State Council proposed the “Opinions on Promoting the Development of Health Services” in 2013, and the release of these “policy signals” is of great benefit to the development of health tourism. In 2016, documents such as the “Health China 2030 Planning Outline”, the “Notice of the State Council on Issuing the “Thirteenth Five-Year Health and Health Plan” and the Standard of National Health Tourism Demonstration Base have become important “catalysts” for the development of health tourism. Overall, the development trend of health tourism is good.

Comparing the overall Theil Index at the national spatial level only provides the most comprehensive macro-view of health tourism but does not enable further analysis of the actual situation in the area. Therefore, it is necessary to interpret the spatial scale difference between the three regions. From the comparison of the three regions (Fig. 3), it can be found that the overall difference between eastern China and western China is significantly higher than that of central China. In terms of the specific magnitude of difference, the overall difference in western China was greater than that in eastern China before 2012, after which the difference in eastern China began to be greater than that in western China and continued until 2019, and the difference between the overall difference between eastern China and western China increased, the difference between western China and central China decreased, and central China was the region with the smallest overall difference in all the study periods. Finally, the overall difference between Eastern China and Central China both tends to increase and is more significant in Eastern China, while the overall difference in Western China tends to decrease significantly. The transfer of capital and technology from eastern China and the severe lag in economic development in western China have caused the overall difference between the two regions to be higher than that of central China. In addition, the overall difference between eastern China and central China has increased with the decline of economic and technological elements in health tourism development, while the implementation of western China development strategy and the full exploitation of quality health tourism resources has led to a reduction in the overall difference in western China.

Thiel index of three regions in China.

The above analysis shows that intra-regional difference has always been the dominant factor in the overall difference of health tourism development in China during the study period, while the influence of inter-regional difference is relatively small and the degree of the effect of intra-regional difference tends to increase, while the influence of inter-regional difference decreases, but the overall change is not significant. In terms of different regions, the overall difference between eastern China and central China tends to increase, we should pay attention to this phenomenon, investigate the causes, and create effective measures to reduce the level of difference in the development of health tourism and realize the comprehensive development of health tourism in China.

Spatial-temporal pattern

The test of Moran’s I index shows that the spatial correlation of health tourism development in China is significant (Table 4), and the development of health tourism in each province of China displays a state of agglomeration in space.

The global Moran’s I index identified a positive correlation of the development of health tourism above the global space, but it is unclear how each province’s above-spatial correlation pattern should be interpreted. It is necessary to interpret local autocorrelation status to better comprehend the spatial correlation pattern and spatiotemporal pattern of the development of health tourism. Through the calculation of local autocorrelation, a Moran scatter diagram can be drawn and the clustering characteristics of health tourism development can be described spatially through the LISA diagram. The four quadrants of the Moran scatter diagram represent different spatial management patterns, with high–high (H–H) and low–low (L–L) clustering in the first and third quadrants, respectively, and there is a positive spatial correlation between the provinces in these two quadrants; the second and fourth quadrants are low–high (L–H) agglomeration and high-low (H–L) agglomeration, respectively, and there is a negative correlation between the provinces in these two quadrants⁵⁴. Through sorting out the scatter chart, the local spatial autocorrelation types of health tourism development in China’s provinces during 2010–2019 are summarized in Table 5.

From the statistics of the table, H–H agglomeration is mainly concentrated in the eastern coastal region, L–L agglomeration is mainly concentrated in western China, and the distribution of H–L agglomeration and L–H agglomeration is more scattered, among which, H–L agglomeration is more prominent in provinces such as Tianjin, Inner Mongolia, and Chongqing, and L–H agglomeration is more prominent in provinces such as Shandong, Hubei, and Hunan. The LISA clustering map is still necessary for verification because the Moran scatter plot can only give us a preliminary assessment of the general state of spatial agglomeration. The LISA clustering map of the health tourism development index in 31 Chinese provinces from 2010 to 2019 was created using ArcGIS software (Fig. 4).

Local autocorrelation LISA plot of China’s health tourism development index (some years). Note: This map is based on the standard map of the standard map service system of the Ministry of Natural Resources of China (review number: GS (2019) 1822), and the base map is unmodified.

As can be seen from Fig. 4, the number of provinces where the local spatial autocorrelations of health tourism development pass the significance test is substantially less than the number listed in Table 5. H–L agglomeration, L–H agglomeration, and L-L agglomeration are the three main types of spatial agglomeration that were present in 2010, 2013, 2016, and 2019 respectively. Overall, homogeneous characteristics predominate in the spatial association pattern of health tourism development in China over the four-time transects and are complemented by heterogeneous characteristics; local spatial non-association is particularly noticeable. Specifically, Jiangsu, Henan, and other provinces and regions have always been the hot spot of health tourism development, while Gansu Province has always been the cold spot of health tourism development. The above analysis shows that although the development index of China’s health tourism is constantly improving, there is still a large distance to reach the ideal level of development.

Dynamic evolution

The development of health tourism in China is not independent of each other in terms of geography, and the development of health tourism in each province is influenced by the region in which it is located, with strong spatial agglomeration and spatial interaction effects, as reflected in the previous analysis on the spatial-temporal pattern. Additionally, the spatial Markov chain method can be used to determine the impact of various neighborhood types on the likelihood of shifting provincial health tourism types, providing a clear picture of the spatial dynamic evolution. Table 6 presents the results of the spatial Markov transfer probability significance test for health tourism development in China. The results once more demonstrate the existence of spatial effects in this process. Therefore, based on the constructed spatial weight matrix and with the help of spatial Markov chain analysis method to further quantify this spatial effect, a comparative analysis of the dynamic evolutionary characteristics of health tourism development is conducted to explore the dynamic evolutionary process of China’s health tourism in the geographical and economic context.

The measurement results of the spatial Markov chain are given in Table 7. Firstly, under the condition of considering spatial lag, the stability probabilities are 96.8%, 88.2%, 85.0%, and 25.0% for type 1, 76.9%, 91.7%, 45.5%, and 75.0% for type 2, 44.4%, 89.5%, 75.9%, and 80.0% for type 3, and type 4, respectively were 100%, 100%, 91.7%, and 100%. When the neighboring spatial lag type is 4, the stability probability of type 1 drops sharply from 96.8 to 25.0%, indicating that the provinces with low development index may break through the current development dilemma through their development under the positive “spillover effect” of the provinces with high development index. Secondly, the probability of type 1 with low development index moving to type 2 with a low to medium level under four domain types is 3.2%, 11.8%, 15.0%, and 75%, respectively, which also indicates that the proximity to provinces with high development index is conducive to improving the health tourism development index of this province and region, i.e., there is a positive “spillover effect” in geographic space. When the neighboring spatial lag type decreases from high development index type to medium–high development index type, the probability of type 3 shifting downward increases from 0 to 3.4%, and the probability of type 4 shifting downward increases from 0 to 8.3%, indicating that neighboring with provinces with lower health tourism development index will have certain negative effects on the health tourism development of this province, i.e., there is a geospatial negative “spillover effect”. The “the better is better, the worse is worse” polarization tendency is presented as a result of the “Matthew effect” characteristics of the evolution of China’s health tourism development index distribution. The analysis presented above demonstrates that the transfer of each type of development index is not geographically isolated but frequently influenced by the development of health tourism in nearby areas.

In summary, the spatial Markov transfer probability matrix shows that the probability values that are not immediately adjacent to the diagonal values are nearly all zero, indicating that it is unlikely, based on taking span into account, that the health tourism development index of each province will achieve leapfrogging, regardless of the index of the neighboring provinces’ health tourism development. There is no leapfrogging and the transfer of health tourism development index types only happens between adjacent rank types. The type of spatial lag in nearby areas tends to have an impact on how China’s health tourism development index evolves, but under the traditional rough tourism development model, China’s health tourism development has a strong path dependency, making it challenging to achieve short-term leapfrogging within a year. The evolution of China’s health tourism development index shows a partial “Matthew effect”, therefore, it will show a certain clustering distribution pattern in geographic space.

Influencing factors

Selection of influencing factors

The development of health tourism is affected by multiple factors, such as resource status, market demand, traffic condition, economic level and tourism investment, which are important driving forces for the development of health tourism in China^23,55. Based on this reference to related literature, eight factors such as economic situation, social situation, traffic condition, talent situation, industrial structure, industrial base, consumer demand and employment demand are selected as the explanatory variables for the four major driving forces of health tourism development such as safeguarding force, supporting force, pushing force and pulling force^29,56,57,58, and the analytical framework of the influencing factors of health tourism development is shown in Fig. 5. Specifically, the higher the per capita GDP, the more developed the regional economy, so the per capita GDP is chosen to measure the economic situation. The higher the level of urbanization of the population, the higher the level of population literacy, the more harmonious social atmosphere, so the level of urbanization of the population is used to measure the level of social development. The accessibility of transportation in the development of health tourism is an important support, so the per capita area of the city road is used to illustrate the traffic condition. The more students in higher education, the more favorable it is to cultivate more professionals for the development of health tourism, so the average number of students in higher education per 100,000 people was chosen to represent the talent situation. The tertiary industry can shape a good environment for the development of health tourism, so the proportion of tertiary industry in GDP is chosen to indicate the industrial structure. The state of natural resources and medical conditions is an important foundation for the development of the health tourism industry different from that of the general tourism industry, so the sum of the scores of per capita water resources and the number of practicing (assistant) physicians per 10,000 people are chosen to represent the industrial base. The higher the consumer demand for health tourism and the more employment it absorbs, the more it can stimulate the development of health tourism, so the share of domestic tourism revenue in GDP and the number of people employed in the culture, sports and entertainment industry were chosen to assess the consumer demand and employment demand respectively.

Analytical framework of factors influencing the development of health tourism.

Analysis of influencing factors

According to Moran’s I index, we find a significant spatial correlation between the health tourism development of Chinese provinces. In order to further explore the influencing factors of health tourism development and its spatial effects, the spatial econometric model was chosen for the follow-up study. The test results of specific model selection are shown in Table 8.

From the test results in the table above, SDM with spatial fixed effect is more suitable. For comparison, SAR and SEM were also estimated and tested as well and the results are presented in Table 9. From the estimation results, \(\sigma^{2}\) is significant at the 1% significance level, indicating a good fit of SDM.

In order to accurately judge the influencing factors and their spatial spillover effects on the development of health tourism in China, the estimation results of the direct, indirect and total effects are summarized in Table 10, where the direct effect mainly reflects the influence of the influencing factors on the development of health tourism in the region, the indirect effect mainly reflects the spatial spillover effects of the influencing factors on the development of health tourism in neighboring regions, and the total effect is the sum of the direct and indirect effects.

First, in terms of direct effect. Nationally, factors such as social situation, traffic condition, talent situation, industrial structure, industrial base, consumer demand and employment demand all have a positive impact on the development of health tourism. People in areas with good social situation tend to pay more attention to physical and mental health, and are more likely to accept and support the development of health tourism. Transportation accessibility changes the breadth and depth of economic ties between health tourism-related industries in the region to a certain extent. The greater the number of professionals, the more beneficial it is to the high-quality development of health tourism. A good industrial structure can promote the comprehensive and sustainable development of health tourism. The development of health tourism emphasizes “health”, and the health industry and health resources and other industrial bases are particularly important. In addition, the higher the consumer demand and the greater the benefits, the more it can directly stimulate the growth of health tourism-related investment. The more employment population absorbed and the stronger the employment demand, the more conducive to the expansion of the scale of health tourism-related industries. In terms of subregions, the positive influencing factors in eastern China are social situation, talent situation, industrial base and employment demand; in central China, the positive influencing factors are economic situation, social situation, industrial structure and consumer demand; and in western China, the positive influencing factors are social situation, traffic condition, talent situation and consumer demand. The difference in the influencing factors for the development of health tourism in the three regions also reflects the difference in the development of these regions in many areas, including the economy, society and culture. For example, the factors that have a greater impact on the development of health tourism in western China are the lower level of social development, the more inconvenient traffic condition and the lack of talent reserves, which cause the direct effect of these factors on the development of health tourism in the western region to be more significant.

Second, the indirect effect, i.e., the spatial spillover effect aspect. For the whole country, the positive spatial spillover effect of industrial structure on the development of health tourism in neighboring regions is the highest, and the indirect effect of industrial structure is greater than the direct effect. Liu &Tang (2022) pointed out that the industrial structure can characterize the level of specialization of regional tourism to a certain extent, which affects the efficiency of its resource utilization and economic benefits⁵⁹. A good industrial structure not only promotes tourism development within the region, but also drives the improvement of tourism resource utilization efficiency and tourism specialization level in neighboring regions to a certain extent. The spatial spillover effect of the industrial base on the development of health tourism in the neighboring regions is lower than that of the industrial structure, a phenomenon that may be explained by the relatively limited radiation of natural and human resources of health tourism to the neighboring regions and the strong territoriality of these resources. Moreover, the indirect effect of the six influencing factors, including economic situation, social situation, traffic condition, talent situation, consumer demand and employment demand, do not pass the significance test. In terms of subregions, the influencing factors with spatial spillover effect in eastern China are economic situation, industrial base, traffic condition and employment demand, the influencing factors with spatial spillover effect in central China are industrial structure, and the influencing factors with spatial spillover effect in western China are economic situation and consumer demand. As mentioned above, there are reasons behind the difference in the spatial spillover effect of the factors influencing the development of health tourism in the three major regions. For example, the overall economy of western China is relatively backward, and the enhancement of its economic situation can also promote the development of health tourism in neighboring regions, and as western China further strengthens its advantages in resources, it is inevitable that it will also weaken the attractiveness of the health tourism in the neighboring regions.

Third, the total effect, i.e. the sum of direct and indirect effects. For the whole country, the total effect of industrial structure is the highest, followed by industrial base, economic situation, and the five factors of social situation, traffic condition, talent situation, consumer demand and employment demand are not significant. In terms of subregions, the factors influencing the total effect in eastern China are economic situation, talent situation, industrial structure, consumer demand and employment demand, the factors influencing the total effect in central China are economic situation and industrial structure, and the factors influencing the total effect in western China are social situation, traffic condition, talent situation and consumer demand.

Finally, in order to avoid endogeneity problems between the data, the spatial weight matrix of economic adjacency was applied instead of the geo-economic adjacency matrix to test the robustness of the factors influencing the development of health tourism (Table 11). The results show that the benchmark regression results remain unchanged, indicating that the analysis is robust.