Younghoon PARK, Kyung Hee University, Korea (Republic of)
This poster aims to analyse the spatial pattern of the South Korean presidential election from 2002 to 2017. Election results have attracted numerous political scientists, journalists, and data professionals. However, in analysing the results, spatial factors such as demographic characteristics and neighbourhood effects were excluded. Moreover, the analysis was focused on the visualization of a single election result. In particular, even after the electoral system was changed to a direct election in 1987, South Korea has undergone massive changes in politics, society, and economy. There also have been many changes in both demographic structure and spatial structure, such as rapid suburbanization and ageing of the population. Nonetheless, the spatial patterns that have changed accordingly have not been deeply investigated. Therefore, this poster puts emphasis on the voting patterns of South Korea, which have not been researched in terms of space. How has voting behaviour changed over time? Do neighbouring regions show similar voting pattern? Is there a specific demographic background for voting results? To answer these questions, this poster will analyse the spatial autocorrelation of the four presidential elections in the 21st century on the municipal scale, and find out which demographic factors influenced the results by conducting a spatial regression. For spatial autocorrelation, both Global Moran's Index and Local Indicator of Spatial Association will be considered. Lastly, both the ordinary least-squares model and the spatial error model will be considered in spatial regression analysis.
Mots clés : South Korea|South Korean presidential election|Regional support|Electoral support|Spatial autocorrelation
A103729YP