Use the latest browser recommended by Microsoft
Get speed, security and privacy with Microsoft Edge
One of the major problems challenging policy makers in developing countries is coping with high and persistent fluctuation in the level of unemployment. The objective of this paper is to identify the best approach to forecasting unemployment in Barbados using seasonal autoregressive integrated moving average (SARIMA), Basic Structural Time Series (BSTS) and General Structural Time Series (GSTS) models. Applying quarterly data for Barbados from 1983Q1 to 2013Q4 to the rate unemployment, this study evaluates the forecasting performance of the three competing models, using forecast accuracy criteria, such as the root mean squared error (RMSE), mean absolute percentage error (MAPE) and Theil’s inequality coefficient. With respect to the techniques used the seasonal autoregressive integrated moving average (SARIMA) model produces superior results, as the forecasts horizon increases, but the General Structural Time Series model performs better in the shorter term. Thus for policy purposes, a seasonal autoregressive integrated moving average (SARIMA) model is relevant for decision-making.