Working Paper: NBER ID: w1939
Authors: N. Gregory Mankiw; Matthew D. Shapiro
Abstract: This paper studies the nature of the errors in preliminary GNP data, \nIt first documents that these errors are large. For example, suppose the \nprelimimary estimate indicates that real GNP did not change over the \nrecent quarter; then one can be only 80 percent confident that the final \nestimate (annual rate) will be in the range from -2.8 percent to +2.8 \npercent. The paper also documents that the revisions in GNP data are not \nforecastable, This finding implies that the preliminary estimates are the \nefficient given available information. Hence, the Bureau of Economic \nAnalysis appears to follow efficient statistical procedures, in making its \npreliminary estimates.
Keywords: GNP revisions; economic forecasting; statistical efficiency
JEL Codes: No JEL codes provided
Edges that are evidenced by causal inference methods are in orange, and the rest are in light blue.
| Cause | Effect |
|---|---|
| preliminary estimates (C13) | revisions (Y20) |
| method of estimation (C13) | variability of revisions (L15) |
| preliminary estimates (C13) | uncertainty in revisions (D80) |
| revisions are not forecastable (C53) | preliminary estimates do not predict final values (C13) |