Using Univariate Time Series Models to Forecast Unemployment Rates: Evidence from State of Maine

Unemployment rate has been always one of the major issues that policy makers deal with. Having a future forecast of unemployment rate can really help policy makers in making right decision. Forecasting is a process that determines what will happen to the desired variable in future, which involves lots of uncertainties and risks, by studying and analyzing the past and the present. This paper uses ARIMA models to constructing a time series model based on historical data and then projecting it to the future to have a forecast on unemployment rate in the state of Maine. ARIMA models have the advantage of relatively low research costs compared to other econometric models and also a good efficiency in short-term forecasting. The time interval for the data is 1996M1-2015M4 and the data are taken from the “Center for Workforce Research and Information” website