KONVERGENSI PENERIMAAN DAN PENGELUARAN PEMERINTAH PROVINSI DI INDONESIA: PENDEKATAN DATA PANEL DINAMIS SPASIAL TAHUN 2000-2012

• Main Author: hermada dekiawan
• Format: Article
• Language: Indonesian
• Published: Bank Indonesia 2018-09-01
• Series: Bulletin Ekonomi Moneter dan Perbankan
• Online Access: https://www.bmeb-bi.org/index.php/BEMP/article/view/4

The study analyse sigma and beta convergence of provincial government revenues and expenditures in Indonesia (APBD) by using panel data 30 provinces over the period 2000-2012. The variables used in this study consists of real income per capita and the revenues and expenditures variables in the APBD. The study also included a spatial weights matrix to analyse dependency among provinces. Spatial weights matrix used consists of two types of weights, the real income per capita and the geographical distance. Testing for the presence of spatial dependency performed using Moran's I. The model used in this study are spatial autoreregressive model (SAM) and the spatial error model (SEM). The models are estimated using panel least squares, fixed effects, random effects, as well as both GMM first difference (GMM-DIFF) and system GMM (GMM-SYS). Based on sigma convergence approach, the results of the study showed that during the period of 2000-2012 occurred convergence on total revenue, income, tax, fund balance, total spending, employee spending, and goods spending, but not for the real income per capita. Estimation with beta convergence approach conducted on four variables as each sample (in per capita value), namely: total income, tax, total spending, and goods spending. Estimation with beta convergence is done by using additional explanatory variables which include: economic growth, the degree of openness, as well as economic growth. Based on the beta convergence approach, the study results indicate the occurrence of convergence on total income, tax, total spending, and goods spending. Estimates also lead to the conclusion that there are spatial dependencies between provinces either use distance and income per capita weight.