Identifying hysteresis effect in unemployment rate with emphasis on second generation panel unit root and PANIC method

In the last decade, Iran is one of the numerous countries with persistent and high unemployment rate. Examining of unemployment trend in Iran shows that over the recent years, this variable has been constantly at a high level, without any tendency to converge to a certain amount. This issue justifies the necessity of examining the existence of hysteresis effect hypothesis in unemployment rate of Iranian economy. In fact, the term of hysteresis effect means dependency to the past. Therefore, hysteresis effect in unemployment rate implies the dependency between current and past unemployment rates. In this situation, all the shocks will have a permanent effect on the path of unemployment and the economy will never achieve long-term equilibrium because the equilibrium is changing constantly. Due to above, hysteresis effect hypothesis is counterpoint of natural rate of unemployment.
The reaction of the labor market to the shocks will be established the existence of natural rate of unemployment hypothesis (In the absence of unit root) or the alternative (if there is a unit root process), hysteresis effect hypothesis. When natural rate of unemployment hypothesis is established, after a direct (or indirect) shock to the labor market, unemployment rate will converge to the long-term value which is calld natural rate of unemployment without any kind of intervention while if there is hysteresis effect, the convergenc of unemployment rate to the long-term value after a shock will not take place. In this study the existing of hysteresis effect in unemployment rate of Iranian economy is investigated with using first and second generations of panel unit root tests with emphasis on second one in period 2005Q1-2015Q3.
In addition to the evolution of unit root tests in terms of the heterogeneity problem which ocurres in time series estimations, a second evolution has also heppend recently that takes the existence of cross-sectional dependence into account. One of the methods which categorize in the second generation of unit root tests is PANIC approach. We employ the PANIC procedurs of Bai and Ng (2004), which allows us to decompose the observed unemployment rate series into common factor and idiosyncratic components. This enables us to identify the source behind the hysteretic behavior which may be found. Using this method decreases the possible doubts on the results of the previous study in two ways. First, with increasing the sample size due to using of panel data instead of time series data. Second, with applying the new method which is used frequently due to more accurate output.
While the results of first generation tests show that the unemployment rate is stationary, the second generation refers to the non-stationary process and prove the existence of hysteresis effect in unemployment rate. Due to the higher efficiency of second generation when the cross-section dependency in the panel is identified, we rely on this method and then hysteresis effect hypothesis is accepted which is in accordance with the evidence of labor market. The existence of hysteresis effect suggests long-run policies rather than short-run solutions to stabilize the labor market. Furthermore, proving that there is a hysteresis effect in unemployment rate, it is suggested that the sources of this effect be analyzed in future studies. It should be noted that in this study four stochastic trend that led to nonstationarity and hysteresis effect were identified. Future studies can introduce these resources by studying precisely the methods for identifying the sources which generate hysteresis effect.