The effect of GNP announcements on fluctuations of GNP growth

This paper is an empirical study of the degree in which perceptions affect the evolution of the economy. We study the effects of the announcements that the government makes on GNP growth. These announcements are subject to a substantial degree of noise and its accuracy improves with time. A revised number is published several years after the first announcement was made public. We consider that the final revision is the “true” value of GNP growth. We show that once announcements are taken into account, the true value of GNP growth at time t has no predictive power in determining growth at any future time. All the predictive power lies in the announcements, and not in the true level of growth. Actually, we show that the variable that determines future growth is the unexpected part of the announcements. We also show that announcements affect growth via aggregate investment.