Preliminary data in economic databases
Old preliminary data should not be removed when new data arrives, otherwise important information is lost. Economic research, especially modelling, needs this data. The costs of saving old preliminary data can be made very small by proper arrangement of data and with simple bookkeeping about when and from where the data came. As an example, the database of the Research Institute of the Finnish Economy (ETLA) is described.
KeywordsPreliminary Data Forecast Error National Account Preliminary Figure Final Figure
Unable to display preview. Download preview PDF.
- (1).Boucelham, J. and T. Teräsvirta (1989). How to use preliminary values in forecasting the monthly index of industrial production? The Research Institute of the Finnish Economy (ETLA), Discussion Paper No 284.Google Scholar
- (2).Jong, P. de (1987). Rational economic data revisions. Journal of Business & Economic Statistics 5, 539–548.Google Scholar
- (3).Trivellato, U. and E. Rettore (1986). Preliminary data errors and their impact on the forecast error of simultaneous-equation models. Journal of Business & Economic Statistics 4, 445–453.Google Scholar
- (4).See almost any issue of International Journal of Forecasting or Journal of Business & Economic Statistics.Google Scholar
- (5).Vajanne, H. (1989). Table manual (in Finnish, unpublished).Google Scholar