One thing economists can agree on is that inflation is hard to forecast. This column argues that in this context, the idea that ‘core inflation’ may be a useful predictor is very appealing, especially for central banks that need to know where inflation is heading. Evaluating the ability of core to forecast headline inflation for OECD and some non-OECD countries, it seems that sizable predictability emerges for a very small subset of countries, but it is either subtle or undetectable for most other economies.
http://www.voxeu.org/article/evasive-predictive-ability-core-inflation