Analisis Spasial Kemiskinan Dengan Pendekatan Geographically Weighted Regression: Studi Kasus Kabupaten Pandeglang Dan Lebak

Abstract

Poverty is the main problem both at the national and regional development. Existing poverty alleviation programs have not paid attention to the spatial aspect. Thus the policies are often poorly targeted. This study aims to find spatial patterns of poverty in Pandeglang and Lebak districts. Geographically weighted regression (GWR) is used to analyze the poverty data in 2016. Based on the analysis, positive spatial autocorrelation is found and clustered in 25 sub-districts. Net enrollment rates tend to reduce poverty in all sub-districts. Meanwhile, village funds, electricity, and roads tend to reduce poverty rates in more than 80% of sub-districts. Independent variables have a different response in each sub-district. Therefore, the poverty alleviation program of each sub-district is adjusting to its influencing factor.

Factors affecting poverty using a geographically weighted regression approach (case study of Java Island, 2020)

“…The linear regression model only produces globally valid parameter estimates, while the GWR model produces localized model parameter estimates for each observation location (Widiyanti et al, 2014). Several studies on poverty have been carried out Sukanto et al (2019) found that 25 sub-districts in Pandeglang and Lebak Regencies have positive spatial autocorrelation. Village funding, electricity, and roads tend to have a poverty reduction effect in 80 percent of the sub-districts in Pandeglang and Lebak districts.…”