Search

There have been a number of empirical analyses of the accuracy of economic forecasts. These studies have shown that economists have generally underestimated increases in gross national product and its components. On the other hand, when these variables declined, the predictions frequently failed to forecast these movements, thus producing turning point errors. One of the studies also indicated that forecasters were unable to "identify" cyclical peaks for some time after the occurrence of the event. For instance, while the peak preceding the 1960 recession occurred in the third quarter of 1960, none of the three sets of forecasts analyzed in this study predicted this turn in advance. The best performance was turned in by the forecasters who recognized the turn near the end of the third quarter. However, the forecasts of this set had, earlier in the third quarter, indicated that the economy would grow more rapidly in that quarter than it had in the second period. The other two sets of forecasts did not recognize the turn until the beginning or end of the fourth quarter of 1960. This paper point errors in the vicinity of cyclical peaks and troughs and will advance one hypothesis to explain why the observed errors might have occurred. This knowledge might prove useful in improving the quality of future forecasts.