Abstract
Although shift – share is an inexpensive and easy means of making regional forecasts, its accuracy depends on the method used in forecasting the competitive component of the region. As theory is unclear as to the sign, let alone the magnitude, of the competitive component of the next period, it would be appropriate to use time-series techniques to forecast it. Unfortunately the typical method of application of shift–share results in too few data points for the application of time-series analysis. In this paper an alternative method of calculating the competitive component that resolves that problem is presented. In an empirical test of the technique, time-series methods are used to forecast the competitive series for a metropolitan region. The results are found to compare favorably with the forecasts of other techniques.
Get full access to this article
View all access options for this article.
