The Current Employment Survey provides the information on the increase or decrease in the number of jobs in the U.S. economy on a monthly basis. This measure is considered to be one of the leading economic indicators. The initial numbers may be revised due to firms that do not report in a timely manner leading to revisions in the initial numbers for several months afterwards. These revisions, up or down, can be significant at times, and may change the inferences made about the health of the economy.
This paper will examine the relationships between the characteristics and recent CES reporting and employment history of late responders over the last several years using several methods, including quantile regression. With this information in hand, models will be constructed that predict the revisions overall and by economic sector. The quality of these model predictions will be evaluated by adjusting the initial estimates and comparing them to the actual revised estimates.
|