Data revisions are not well-behaved

We document the empirical properties of revisions to major macroeconomic vari-ables in the United States. Our \u85ndings suggest that they do not satisfy simple de-sirable statistical properties. In particular, we \u85nd that these revisions do not have a zero mean, which indicates that the initial announcements by statistical agencies are biased. We also \u85nd that the revisions are quite large compared to the original variables and they are predictable using the information set at the time of the initial announcement, which means that the initial announcements of statistical agencies are not rational forecasts. We also provide evidence that professional forecasters ignore this predictability