The determinants of inflation in South Africa : 1997-2011

Abstract

This study targeted the a cause-effect relationship pertaining to the CPI inflation in South Africa and the core variables causing it. The result is a set of scenarios of effects on CPI inflation due to the core variables, the corresponding transitional probability matrices and the resulting Markov's chains for forecasting CPI inflation. The core variables have been identified using the findings of an in-depth literature review. The core variables are: total government expenditures, total private expenditures, exchange rate (USO to ZAR) and taxation. A five year sample has been taken from RBSA and OECD sources. First of all, a Pearson's correlation testing has been conducted to estimate the approximate transitional probabilities. Thereafter, four scenarios were defined to arrive at the final Markov's chains. Based on the findings, the current scenario of CPI inflation in South Africa has been discussed, compared with the past, and a method of forecasting CPI inflation using the Markov's chains is proposed. In addition, a discussion of how the Markov's chains can be modified by moving the five year (or larger sample) ahead has been presented. In the future, it is proposed that a benchmark be created by conducting a similar analysis on the comparable countries such that a better view of national performance of South Africa can be gauged.