The NAIRU in Israel: An Unobserved Components Approach
The Non-Accelerating Inflation Rate of Unemployment (NAIRU) is estimated for the post-stabilization period, as an unobserved stochastic variable, using state-of-the-art State Space Models. The NAIRU is identified by a Phillips curve equation, and is assumed to follow a random walk. The basic model is augmented by an equation that captures the persistence of the unemployment gap. We also use the joint system first introduced by Apel and Jansson (1999) to estimate potential output and the NAIRU simultaneously. Confidence intervals around the NAIRU were computed by a jackknife technique. The results indicate that the actual variation of unemployment has only a minor effect on the NAIRU, which remained relatively stable throughout the sample period. The state variables have sufficiently stable characteristics to be successfully predicted at least one step ahead. However, policy implications that may be derived are sometimes limited, as there is substantial uncertainty around the estimated NAIRU. No evidence for hysteresis was found. The estimates show that the disinflation process during the 1990s did not cause an increase in the NAIRU.