| Summary: | This thesis studies price behaviour, the determinants of inflation, threshold effects, and policy rules in Indonesia. Using data from January 2002 to April 2008, the study reveals that the relative law of one price holds. This is proven by the variability of one product across cities being lower than the variability of all products in one city. The variability of the product is explained well by the cost of transportation and the level of development. Employing panel regression, this study also shows that prices of 35 sub-categories within 45 cities in Indonesia exhibit convergence. The average speed of convergence for perishable goods is about 9 months, for non-perishable goods 32-36 months, and for services 18-19 months. Inflation in Indonesia is found to be affected by expected inflation (backward-looking expectations and forward-looking expectations), the output gap, exchange rate depreciation, and money growth. Backward-looking inflation expectations dominate the form of inflation expectations. Using data from 1980:1 to 2008:12 and employing Generalized Method of Moments, this study finds non-linearity in the Phillips curve. The effect of exchange rate depreciation on CPI inflation is found to be linear. However, the effect of money growth on inflation is not linear; there are two threshold values identified. The study shows that the higher the growth of money, the less the impact on inflation is. To guide policymakers, this thesis derives simple, but efficient policy rules. Using monthly data from 1980 to 2008 and simulating deterministically the small open macroeconomic model, the study reveals that the inflation forecast-based rule with contemporaneous output gap (IFBG) is the most efficient rule for Indonesia. The rule suggests that the central bank should react strongly on the inflation deviation from the target, react moderately on the output gap and smooth the interest rate. The optimal horizon is 3-4 quarters. Including exchange rate in the policy rule causes deterioration in economic performance.
|
|---|
