Search

Forecast errors in energy consumption are compared for IEA-countries in 1985, 1990 and 1995. Countries having small/big forecast errors in 1985 do not have small/big forecast errors in 1990 and 1995. Almost no significant correlation coefficient is found between forecast errors in the 3 years. Correspondingly, no significant correlation coefficient is found between forecast errors in the 3 main sectors. Therefore, a relatively small forecast error is not caused by a relatively small forecast error in all 3 sectors.

Every year Policy of IEA Countries includes a forecast of the energy consumption in the member countries. Forecasts concerning the years 1985, 1990 and 1995 can now be compared to the actual values. The second oil crises resulted in big positive forcast errors. The oil price drop in 1986 did not have a similar effect. A correction for economic growth reduces forecast errors during the second oil crises but not elsewhere. Industry has a relatively big positive forecast error while transportation has a negative forecast error. Even when the forecast error is small, the results are not so "nice" because the small value is often the sum of large positive and negative errors.