Modelling and Forecasting Unemployment Non-linear Dynamics Using Spectral Analysis
Movements in the unemployment in Croatia are largely permanent but transitory movements accounts for most of the unemployment dynamics after 2008. Unemployment series is non-stationary but mean reversing, non-linear with structural breaks and significant white and red noise in the data. This paper estimates Multivariate singular spectrum model (MSSA) to explain overall fluctuations in unemployment registered during 1998 – 2013. Unemployment behaviour in Croatia show evidence of cyclostationarity caused by seasonal employment effects. We use 88 time series (variable) to comprehensively explain observed fluctuations with our MSSA model explaining 76% of the total unemployment variance. Evidence of this study points that unemployment phenomena should be modelled by using non-linear model with multivariate singular spectrum models giving more robust and empirically valid results in relation to standard modelling techniques. A 5-6 years limit cycle in unemployment is isolated dominating unemployment behaviour in Croatia over last two decades.