When the Bureau of Economic Analysis estimates the nation’s gross domestic product on March 28, it will officially be a “third estimate” – even though we’ve only estimated GDP for that quarter once before.
Wondering where the second estimate went? Don’t worry, we didn’t lose count. The unusual naming sequence reflects changes to our release schedule caused by the lapse in federal funding earlier this year. Many of the principal source data BEA uses to produce our first two estimates of GDP for each quarter were substantially delayed.
As a result, BEA combined the first two GDP estimates for the fourth quarter of 2018 (and the 2018 annual estimate) into a single release called the “initial estimate.” This initial estimate, published Feb. 28, combined some data and methods typically used for a first quarterly release of GDP, known as the “advance,” with some normally available by the second release. The initial estimate was based on data that were incomplete and subject to further revision by the source agency.
The source data normally used for a third estimate will be available for our March 28 release of GDP. Therefore, this release is keeping its original name – Gross Domestic Product, Fourth Quarter and Annual 2018 (Third Estimate). It is being released on its originally scheduled date, completing this cycle of GDP estimates for the fourth quarter and year.
Corporate profits for the fourth quarter and year of 2018 also will be released March 28 as originally scheduled. In April, BEA will move on to its advance estimate of GDP for the first quarter of 2019, set for release April 26.
Why does BEA normally release GDP estimates three times after the end of each quarter? Each subsequent estimate, coming about a month apart, includes new or improved source data from government surveys, private data collectors, and other sources that weren’t available for the previous estimate.
In addition, quarterly and yearly statistics are revised again during BEA’s annual updates each summer. That’s when we incorporate a substantial influx of data not available earlier, as well as any improvements in methodology.