Charles F Manski, Wednesday, May 21, 2014 - 00:00

Many economic statistics move markets when first released, and move them again when they are revised. This column suggests ways of measuring the transitory statistical uncertainty in estimates of official statistics based on incomplete data and the permanent statistical uncertainty stemming from survey non-response. Government agencies would be doing the public and policymakers a service by being clear about these uncertainties.

CEPR Policy Research