5:15 PM - 7:15 PM
[HTT15-P08] Evolution of Tourism Destination Network: The Case of Chinese Package Tours to Japan from 2006 to 2016
Keywords:gravity model, inbound tourism, tourism flow, network analysis
Tourism flow, a key topic in tourism geography, follows an origin-link-destination structure. Since the 1990s, network analysis has been widely applied, including evolutionary dynamics. However, studies focusing on the evolution within receiving countries from an industry perspective remain limited. This study analyzes Chinese package tours to Japan (2006–2016), uncovering structural evolution and driving factors from the perspective of tour operators.
Key characteristics of international destination networks include declining centralization and the role of distance in shaping clusters and flow volume. Miguéns and Mendes (2008) suggest that international destination network resembles transportation and economic networks rather than social ones, where high-degree nodes mainly connect to low-degree nodes. Furthermore they hypothesize that a sublinear preferential attachment model explains cumulative degree distribution decay. This study examines assortative mixing, degree distribution changes, and where new links and destinations emerge.
The destination network consists of nodes (destinations), links (paths), direction (visit sequence), and weight (tour sample frequency). Destinations spanning multiple cities are treated as single entities. Evolutionary patterns are analyzed through assortative mixing, degree distribution, and sublinear preferential attachment i.e. 0 < α < 1 in the BA model (Jeong et al., 2003). Centralization and density track overall shifts, while community detection and an advanced core-periphery analysis identify destination changes. For the gravity model, flow volume between origin-destination pairs is the dependent variable, with population, distance, direct flights, and Michelin Green Guide ratings as independent variables. Interviews were conducted with four tour operators in 2024.
Analyzing 6,849 package tours from 2006 to 2016, network analysis and the gravity model yield several findings. In terms of broader network dynamics, this study contributes to understanding disassortativity formation. While Miguéns and Mendes (2008) argue tourism networks resemble transportation rather than social networks, this study finds a gradual decline in assortativity. This trend may be attributed to that popular destinations increasingly link to less popular ones, expanding the Golden Route. The degree distribution decays more slowly, reflecting more destinations and greater flow concentration. This trend is reinforced by rising weight of core destinations. Although tourism networks generally decay faster than other networks, this study highlights that decay rates vary by time. Sublinear preferential attachment was evident early but later disappeared, suggesting new links form with minor destinations rather than major ones.
Beyond these general findings, new destinations tend to emerge near existing ones, confirming an experiential strategy in tour management. Chinese package tours favor metropolitan areas, while interviews highlight the Golden Route-oriented strategy as an evolutionary pathway. Structural stability is evident, as network density remained stable, possibly due to transportation infrastructure. Although at international level centralization declined, this study finds little change at the national level. The gravity model reinforces these findings, showing distance decay reflects with rising transport cost, while population significance is shaped by business strategies. Interviews reveal the ‘2024 issue’ in logistics is reshaping inbound tourism, potentially narrowing travel range.
This study enhances understanding of tourism destination networks by highlighting the evolutionary patterns of Chinese package tours to Japan. Findings demonstrate that tourism networks evolve in a complex manner, shaped by business strategies, spatial and economic constraints.
References
H. Jeong, et al., Europhys. Lett. 61(4), 567 (2003).
J. I. L. Miguéns and J. F. F. Mendes, Physica A 387(12), 2963 (2008).
Key characteristics of international destination networks include declining centralization and the role of distance in shaping clusters and flow volume. Miguéns and Mendes (2008) suggest that international destination network resembles transportation and economic networks rather than social ones, where high-degree nodes mainly connect to low-degree nodes. Furthermore they hypothesize that a sublinear preferential attachment model explains cumulative degree distribution decay. This study examines assortative mixing, degree distribution changes, and where new links and destinations emerge.
The destination network consists of nodes (destinations), links (paths), direction (visit sequence), and weight (tour sample frequency). Destinations spanning multiple cities are treated as single entities. Evolutionary patterns are analyzed through assortative mixing, degree distribution, and sublinear preferential attachment i.e. 0 < α < 1 in the BA model (Jeong et al., 2003). Centralization and density track overall shifts, while community detection and an advanced core-periphery analysis identify destination changes. For the gravity model, flow volume between origin-destination pairs is the dependent variable, with population, distance, direct flights, and Michelin Green Guide ratings as independent variables. Interviews were conducted with four tour operators in 2024.
Analyzing 6,849 package tours from 2006 to 2016, network analysis and the gravity model yield several findings. In terms of broader network dynamics, this study contributes to understanding disassortativity formation. While Miguéns and Mendes (2008) argue tourism networks resemble transportation rather than social networks, this study finds a gradual decline in assortativity. This trend may be attributed to that popular destinations increasingly link to less popular ones, expanding the Golden Route. The degree distribution decays more slowly, reflecting more destinations and greater flow concentration. This trend is reinforced by rising weight of core destinations. Although tourism networks generally decay faster than other networks, this study highlights that decay rates vary by time. Sublinear preferential attachment was evident early but later disappeared, suggesting new links form with minor destinations rather than major ones.
Beyond these general findings, new destinations tend to emerge near existing ones, confirming an experiential strategy in tour management. Chinese package tours favor metropolitan areas, while interviews highlight the Golden Route-oriented strategy as an evolutionary pathway. Structural stability is evident, as network density remained stable, possibly due to transportation infrastructure. Although at international level centralization declined, this study finds little change at the national level. The gravity model reinforces these findings, showing distance decay reflects with rising transport cost, while population significance is shaped by business strategies. Interviews reveal the ‘2024 issue’ in logistics is reshaping inbound tourism, potentially narrowing travel range.
This study enhances understanding of tourism destination networks by highlighting the evolutionary patterns of Chinese package tours to Japan. Findings demonstrate that tourism networks evolve in a complex manner, shaped by business strategies, spatial and economic constraints.
References
H. Jeong, et al., Europhys. Lett. 61(4), 567 (2003).
J. I. L. Miguéns and J. F. F. Mendes, Physica A 387(12), 2963 (2008).