| Abstract: | ABSTRACT. Economic forecasting models are famous for performing well over short time periods and then suffering rapidly deteriorating performance when economic conditions change. This behavior makes composite forecasting models valuable in situations where large forecast errors cause considerable losses. A composite forecasting model for state-level employment is proposed here. This method is designed to protect state budget processes by producing robust forecasts of changes in employment and the related revenue collections. An application to Georgia nonagricultural employment is presented which demonstrates the benefits of this technique. The example shows that the method can forecast such series accurately without the forecaster having to choose in advance a single model specification to all economic conditions. |
