Modeling of Poverty Determinants in Sumatera Island (Panel Regression Approach)

This paper examines to determine whether there is an effect of the human development index (HDI), the open unemployment rate, economic growth, and regional GRDP per capita on poverty levels in Sumatera Island. In general, the trend of the poverty rate by the province in Sumatera decreased during 2010-2019 period, however, there are five provinces which on average have a poverty rate above 11 percent (above the national average poverty rate), namely Aceh, Bengkulu, Lampung, South Sumatera, and North Sumatera. This study used a dynamic panel approach with the Random Effect Model (REM). The data source was obtained from Statistics Indonesia (BPS) for the period 2010-2019. The results showed that the HDI variables and economic growth had a negative and significant effect on poverty levels in Sumatera Island. Meanwhile, the variables of the open unemployment rate and regional GDP per capita have a negative and insignificant effect on poverty levels in Sumatera Island. Simultaneously, it is concluded that the independent variables overall affect the poverty level. The coefficient of determination is 65.6370 percent, which means that variations in the level of poverty can be explained by the independent variables by 65.6370 percent and the remaining 34.3630 percent is explained by other factors out of the model.