2013 : PEMODELAN KEMISKINAN PERDESAAN DAN PERKOTAAN DENGAN PENDEKATAN GARIS KEMISKINAN MENGGUNAKAN MODEL PROBIT BINER BIVARIAT DI PROVINSI BENGKULU
Poverty is one of the biggest problems in development countries. The dilatory in poverty rate decrease and it’s highly disparity among regions has become major burden to the government. Thus it was important for them to identify significant factor that cause such problems, to find the best formula to handle it. Having different characteristic, the poverty causing factor must be differentiated between urban and rural areas. Such procedure applied in the current poverty rate determining method, which was using poverty line approach. In general, poverty line was the sum of poverty line in food and non food commodities, so that the person who has monthly expenditure below those lines is categorized to be the poor. In this papper, we determined the poverty characteristic in urban and rural areas in Bengkulu Province using Binary Bivariate Probit Regression Model. We used food and non food commodities poverty line as a different response variable and run it altogether in one model. we used 17 predictor variables in Susenas that has been proven to be significant in determining poverty characteristic on previous research. The goodness of fit model was measured by AIC criterion. As a result, the numbers of household members, the kinds of house lightning source and motorcylce ownership was significant on determining poverty in rural areas. Meanwhile, the age of household leader, numbers of household members, status of residential building, type of wall widest, the main fuel and motorcylce ownership was significant on determining poverty on urban area.