Forecasting the Unemployment Rate: Application of Selected Prediction Methods

Purpose: Unemployment rate prediction has become critically significant, because it can be used by governments to make decision and design accurate policies. The paper's main objective is to compare the most significant predictive methods for modeling the unemployment rate. Design/Methodology/Approach: In this work, the selected predictive methods (naive method, regression model, ARIMA, Holt model and Winters model) were described, developed and compared using data collected by Central Statistical Office. Findings: The considered methods enable to predict the unemployment rate with high accuracy. The results of experiments allow to conclude that the most suited methods of forecasting the unemployment rate are the quadratic regression model and the Winters multiplicative model. Practical Implications: Forecasting the unemployment rate is one of the important elements in economy and presented methods can be easily used by labor market entities to predict and verify the situation in the market. Originality/Value: Forecasting the unemployment rate is an extremely difficult and demanding task, but on the other hand, it can be an effective tool that supports planning processes. The conducted research showed the quadratic regression model and the Winters multiplicative model provide high accuracy in terms of modeling the unemployment rate