National indicator series as quantitative predictors of small region monthly employment levels

This paper investigates the usefulness of several readily available national indicator series as transfer function drivers to forecast monthly employment levels in small regions. The severe data constraints characteristic of small regions often preclude forecasting practitioners from constructing traditional econometric models. Given these regional data constraints, the ability to generate accurate forecasts of regional variables from readily available national series would be useful to regional planners and other users of regional forecasts. Results of this study suggest that national indicator series can be combined with the transfer function methodology to provide accurate monthly employment forecasts over horizons extending to 12 months. For the composite leading indicator input series, mean absolute percentage errors in forecasting regional employment levels were less than one percent at all forecast horizons up to 12 months.