Spatial autocorrelation analysis of tourist arrivals using municipal data: A Serbian example

Geographica Pannonica • Volume 21, Issue 2, June 2017

By Uglješa Stankov, Tanja Armenski, Michal Klauco, Vanja Pavluković, Marija Cimbaljević, and Nataša Drakulić-Kovačević

“Spatial autocorrelation methodologies can be used to reveal patterns and temporal changes of different spatial variables, including tourism arrivals. The research adopts a GIS-based approach to spatially analyse tourist arrivals in Serbia, using Global Moran’s I and Anselin’s Local Moran’s I statistics applied on the level of municipalities. To assess feasibility of this approach the article discusses spatial changes of tourist arrivals in order to identify potentially significant trends of interest for tourism development policy in Serbia.


Moran significance map for international tourist arrivals in Serbia (without the territory of Kosovo and Metohija) in 2001 and 2013.
Metohija) in 2001 and 2013

“There is a significant spatial inequality in the distribution of tourism arrivals in Serbia that is not adequately addressed in tourism development plans. The results of global autocorrelation suggest the existence of low and decreasing spatial clustering for domestic tourist arrivals and high, relatively stable spatial clustering for international tourists. Local autocorrelation statistics revealed different of domestic and international tourism arrivals. In order to assess feasibility of this approach these results are discussed in their significance to tourism development policy in Serbia.”

Read the paper