Examining the quality of early GDP component estimates

In this paper we examine the quality of the initial estimates of headline GDP and 10 major components of both real and nominal U.S. GDP. We ask a number of questions about various characteristics of the differences between the initial estimates, available one month after the end of the quarter, and the estimates available three months after the end of the quarter. Do the first estimates have the same directional signs as the later numbers? Are the original numbers unbiased estimates of the later figures? Are any observed biases related to the state of the economy? Finally, we determine whether there is a significant difference between the vector of the 30-day estimates of the 10 major components and the vector of the 90-day estimates of the same components. We conclude that, under most circumstances, despite the existence of some bias, an analyst could use the early data to obtain a realistic picture of what had happened in the economy in the previous quarter.