Applicability of the transfer function approach to forecasting employment levels in small regions

Abstract

The problem investigated in this paper is that of forecasting employment levels (both raw and seasonally adjusted) in small regions, regions typically lacking data sufficient for the construction of simultaneous equation econometric models. The method employed is transfer function analysis using readily available national variables as drivers. The results suggest that the transfer function approach is capable of providing accurate forecasts of employment levels in small regions, forecast accuracy being measured in terms of mean absolute percentage errors and root mean squared errors. We conclude on the basis of this demonstrated accuracy that this approach is a viable method for forecasting selected variables in a small region context.