<![CDATA[Anti-poverty programs and policies are designed similar for all regions in Indonesia, disregarding the local socio-culture and the poverty spatial pattern of the regions. The approach is based on central government’s program and not based on each region’s locality. This generic programming approach caused the achievement of development goals decline. The effect of space on poverty can be identified by the presence of spatial autocorrelation, which is the link between the examined variable to itself in a spatial manner or commonly referred to as spatial dependence.The aim of this paper is to investigate the global and local spatial autocorrelation for micro poverty data set in Bengkulu City in order to identify spatial approach for its anti-poverty program. Global Moran Index (MI) tests identifies the overall occurrence of autocorrelation, meanwhile the local spatial test shows which subdistricts has the presence of autocorrelation. Global and local MI are popular tools utilized to calculate the spatial effect, particularly to present spatial dependencies. The relation between urban village linkages obtained an MI value of 0.322. This MI value indicates the presence of spatial autocorrelation for subdistricts located in cluster. In local spatial effect observation using Local Indicator of Spatial Autocorrelation (LISA), its discovered that there are several subdistricts having autocorrelation, meanwhile the rest are not significant. Cluster mapping on global MI and LISA shows high-high poverty districts are located in the south of the city, low-high poverty districts in the east, and low-low high-low poverty districts near the city center.]]>
Copyrights © 2017
