Estimating Nairu for the Turkish Economy Using Extended Kalman Filter Approach
This paper estimates NAIRU (Non-Accelerating Inflation Rate of Unemployment) for the Turkish economy as an unobserved stochastic variable by systems approach. Based on a Phillips curve equation combined with an Okun law for output gap and unemployment gap, the systems approach imposes stochastic laws of motion for NAIRU and potential output and also assumes the parameters to be time-varying. However, the requirement to simultaneously estimate parameters and to solve the state space problem introduces nonlinearity, which requires the use of Extended Kalman Filter (EKF). Estimation results suggest that NAIRU moves in tandem with the actual unemployment, but it follows a more volatile path than the latter. Accordingly, the estimated NAIRU series reacts more sharply to the crises than the actual unemployment. This observation is in line with the prior studies reporting the relatively persistent nature of actual unemployment in comparison to the non-accelerating inflation rate of unemployment. Furthermore, all of the derived series are plausible and capture the significant turning points of the economy. As for the coefficients, the time-varying parameters indicate a stable, yet quite a weak link between unemployment and inflation. Meanwhile, the coefficient of exchange rate in the Phillips curve equation suggests a declining, but significant pass-through to inflation. Moreover, estimation results also point to the presence of considerable inertia in inflation. To sum up, findings of this paper provide guidance for future research on NAIRU, which is an important tool for monetary policy. The paper findings also lay the basis for prospective work that may adopt EKF. But most importantly, this paper confirms the view that the conduct of monetary policy should be based on a flexible and comprehensive framework.