Pursuing the Phillips curve in an African monarchy: The Swazi case

The purpose of this study is to examine whether we can identify a Philips curve fit for the Kingdom of Swaziland as a low middle income Sub-Saharan Africa monarchy using data collected between 1991 and 2016. In our approach we rely on the recently introduced nonlinear autoregressive distributive lag (N-ARDL) model to a variety of Phillips curve specifications. For robustness sake, we further employ three filters (one-sided HP, two-sided HP and Corbae-Oularis filters) to extract the gap variables necessary for empirical analysis. Our findings point to a linear, short-run traditional Philips curve whereas we find strong support for concave shaped unemployment-gap and output –gap based Phillips curve specifications. Given the specific form of concavity discovered in the Phillips curves, the low inflation rate experienced over the last couple of decades can be attributed to a worsening labour and goods markets. Moreover, our evidence also cautions Swazi policymakers of ‘overheating’ of the economy during economic booms in which stabilization tools are required to implemented in such instances. Given the overall absence of empirical studies establishing the Philips curve for the Swazi economy our study makes a valid contribution to the literature.