An alternative measure of underlying inflation in Kenya
Measurement of core inflation has become important with countries adopting inflation targeting frameworks. The focus has been on those elements of inflation that are not transitory, that is, those elements that are permanent and not prone to non-monetary forces. In trying to get these permanent components, the KNBS excludes food and non-alcoholic drinks from the overall CPI basket. On the other hand, the CBK goes further to exclude, in addition to food and non-alcoholic drinks, fuel and power and transport and communications. This research paper tried to look for an alternative measure of core inflation by excluding only the most volatile commodities. The research found that 40 commodities were most volatile with a total weight of 40.661 per cent and most of the volatile items were in the group of food and non-alcoholic drinks. In the fuel and power group, electricity, paraffin and charcoal were found to be most volatile. A core inflation constructed was compared with other core inflation measures by the KNBS and the CBK. The tests showed many similarities between the constructed core inflation and that calculated by the KNBS. For instance, the mean core inflation by the KNBS was 6.9 per cent while the constructed core inflation was 6.7 per cent from January 2003 to December 2008. The deviations from tracking the trend of overall inflation and forecasting overall inflation using Root Mean Square Error (RMSE) was lowest as measured by constructed core inflation followed by core inflation by the KNBS. In the long run, the overall inflation is expected to converge to the core inflation. The highest speed of -0.17 was in the constructed core inflation followed by that of the core inflation as calculated by the KNBS. On the other hand, the highest coefficient on the short run impact of 1.36 was on the core inflation by the KNBS followed by that of the constructed core inflation. The core inflation as measured by the CBK failed in many statistical tests.