The Effect of Local Public Spending on Poverty Reduction in Indonesia

This paper aims to examine the relationship between local public spending and poverty reduction in Indonesia from 2001 to 2004. Using balanced panel data of 273 districts in Indonesia, linear and log specification model, and random effect Generalized Least Square (GLS) regression as the methodology, we find that local government spending can be a powerful instrument in reducing poverty headcount, poverty gap and poverty severity in the following year. In addition, using log specification model, we find that per capita GRDP is significant in reducing poverty headcount, poverty gap and poverty severity. But there is no correlation between per capita GRDP and poverty in the linear specification model. Regarding the district characteristics, we find that household size and percentage of rural population have positive relationship with poverty reduction. On the other hand, clean water access is the only variable which can contribute to reducing all three poverty measures. Own house variable is only significant in reducing poverty headcount, but not other poverty indices. Meanwhile, district split can reduce poverty headcount and poverty gap, but not poverty severity.