Abstract:
Inflation uncertainty causes macroeconomic ills and instability in the economy. This paper investigates if rising levels of inflation uncertainty serve as a source of higher inflation outcomes or vice versa, to determine if inflation uncertainty is potentially a self-fulfilling prophecy. In addition, this paper examines the impact of inflation targeting, implemented in South Africa in February 2000, on the level of inflation and inflation uncertainty. Using a generalised autoregressive conditional heteroskedasticity (GARCH) and GARCH-in-mean model and monthly data spanning the period 1970:01 to 2022:05, the empirical outcomes from this study suggest the existence of a bi-directional relationship between inflation and inflation uncertainty, with strong evidence in favour of the Friedman–Ball hypothesis and weaker evidence in support of the Cukierman–Meltzer hypothesis. This study also finds that inflation targeting has contributed significantly to reducing the level of inflation and inflation uncertainty since its adoption as policy framework. Time-varying Granger causality tests accounting for instabilities underscore the above results, namely that inflation uncertainty led to increased inflation uncertainty in the full pre-inflation targeting period, whereas increased uncertainty led to increased inflation only during the decade preceding inflation targeting. The results heed important policy implications, as it is imperative that inflation is kept low, stable and predictable.
