Revisiting the Phillips curve trade-off: evidence from Tanzania using nonlinear ARDL approach

The spotlight of this study is to re-examine the presence and nature of the long run relationship between inflation and unemployment. The topic has been of much interest to researchers and policy makers as it has significant implications on macroeconomic stabilization policies. Using expected inflation rates and expected unemployment rates generated by Autoregressive integrated moving average (ARIMA), the study tests whether the relationship between the variables is symmetrical or asymmetrical in both short run and long run. Applying the autoregressive distributed lags model (ARDL) and Nonlinear ARDL approaches proposed by Pesaran et al. (2001) and Shin et al. (2014), we confirm the presence of long run equilibrium relationship between expected inflation and expected unemployment. Findings tend to indicate that the long run relationship is symmetrical whereas evidence is in support of asymmetrical short-run trade-off between the variables. CUSUM and CUSUMSQ tests confirm the stability of the coefficients. The study contributes to the literature in three ways. First it uses expected as opposed to actual variables, second it employs recent methodology of Nonlinear ARDL and third it presents new evidence from a Highly indebted poor country-HIPC (Tanzania) using data from 1991 to 2017.