A study on simultaneous centralized and decentralized urban growth

Abstract

Korea, a nation whose development is led by the state, has industrialized through the growth pole strategy intent on creating the raindrop effect. This effect was continuously expected throughout the selection and concentration method, and the concentration of population to growth pole has continued. However, from a mathematical view, the distribution of cities by Zipf's law shows that it changed from a concentrated pattern in some cities to a dispersed pattern. In other words, the growth of the cities and regions went through the process of centralization or decentralization based on the effects of the agglomeration economy. Therefore, in the economic growth of these Korean cities, this research tried to multilaterally analyse the dispute between centralization and decentralization of urban growth. The research results are as follows. First, the distribution of cities was decentralized from a mathematical view, but centralized from a spatial view. Second, the distribution of greater economic areas (GEAs) maintained a concentrated pattern from mathematical and spatial views. Last, the mathematical distribution of all cities intra-GEA was decentralized. So, both the centralization and decentralization processes of growth occurred in the cities at the same time.

Keywords:

• Zipf's law
• urban growth
• regional disparity
• greater economic area
• Moran's I

View correction statement:

Corrigendum

Acknowledgements

This work was supported by the 2012 Research Fund of the University of Seoul.

Notes

The analysis of the urban rank-size distribution is analysed with 82 cities, the total cities in Korean is 84 in 2010. The Jeju-si and Seoguipo-si, which are the district cities to move in or out, are excluded, and the Integration Changwon-si, integrated in 2010 is analysed by classifying into the previous administrative district, Changwon-si, Masan-si, and Jinhae-si.

The GEA means the economically homosphere region and Korea, in 2008, classified the whole country as 5+2 GEAs. The GEAs are Seoul-centred SGA (Seoul, Incheon, Gyeonggi-do), Daejeon-centred CGA (Daejeon, Chungcheongbuk-do, Chungcheongnam-do), Gwangju-centred HGA (Gwangju, Jeollabuk-do, Jeollanam-do), Daegu-centred DGA (Daegu, Gyeongsangbuk-do), Busan-centred BGA (Busan, Ulsan, Gyeongsangbuk-do), GGA (Gangwon-do) and JGA (Jeju-do). It is the metro area reflecting the local features and economic environment not limited to the administrative area of the existing cities and states.

The spatial autocorrelation coefficient follows that whether it applies any type of space weighting matrix or not, the result is changed. In this paper, the spatial autocorrelation was analysed after reflecting the polygon contiguity on the ArcMap 10.0, which is the analytical tool, and implementing the horizontal standardization, the sum of each area's horizontal influence becomes 1.

Anselin Local Moran's I, the visual index considering the local change of the spatial autocorrelation, shows the type of spatial clustering. This method's equation is I _i =z _i ∑ _j w _ij z _i . The w _ij means space weighted value, z _k means deviation of average.