Abstract:

Employment is an important indicator closely followed by policy makers and scholars. Time series properties of employment variable play an important role in the validity of the forecasts and econometric models. We investigated stationarity properties of monthly employment data by conventional, structural break and seasonal unit root tests. Employing seasonally adjusted employment data or investigating stationarity without accounting for seasonality may lead to wrong conclusions. Our findings show that the conventional, structural break and seasonal unit root tests contradict each other. Conventional unit root tests and unit root test with structural breaks indicate varying results; however, seasonal unit root tests show that employment data has a tendency to be stationary. This result provides insights to studies that use employment in assessing the stationarity properties of the series. This study also shows that it may be more appropriate to use seasonal unit root tests for series that contain a seasonality component.

Keywords: