Analysis of spatial determinants of poverty in Kelantan

This study examines socio-demographic effects on poverty and measures spatial patterns in poverty risk looking for high risk of areas. The poverty data were counts of the numbers of poverty cases occurring in each 66 districts of Kelantan. A Poisson Log Linear Leroux Conditional Autoregressive model with different neighbourhood matrices was fitted to the data. The results show that the contiguity neighbour was performed nearly similar to Delaunay triangulation neighbourhood matrix in estimate poverty risk. Apart from that, the variables average age, number of non-education of household head and number of female household head significantly associated with the number of poor households head. Kursial was found as the highest risk area of poverty among 66 districts in Kelantan.