The Bureau of Labor Statistics (BLS) is changing how it announces data after an incident in which a data release was delayed but the information was shared with some external users who called or emailed to ask about it.
This incident — which centered on a 32-minute delay in the release of preliminary payroll employment data on Aug. 21 — was caused by an IT issue, by staff having to load the data manually and by a lack of BLS policy about how to respond to public inquiries, according to information provided to PYMNTS by a Labor Department spokesman on Wednesday (Aug. 28).
In addition, the announcement that was delayed had been categorized as an update to a web page rather than as a “news release,” so it had different protocols, according to the information. One difference is that there are several backups to ensure news releases are available on time, while there were no similar measures for updates.
One of the changes the BLS has made is that it will no longer differentiate between “website releases” and “news releases.” All data with an announced release time will be subject to the same protocols and safeguards.
The agency will also make several other changes, per the information provided. BLS will have multiple ways to distribute each data release, including via social media posts; will verify data is available to the public before responding to any public inquiries about it; and will document these activities for each data release.
BLS will announce more actions in the coming days, according to the information provided.
The Aug. 21 incident sparked fresh criticism of the BLS’ procedures from Wall Street firms that use the data to spot trends in the U.S. economy, Bloomberg reported Wednesday.
In another incident earlier this year, the agency inadvertently published some data 30 minutes early. Months before that, a BLS economist may have shared data that was not widely available with Wall Street firms, according to the report.
The Aug. 21 data release said that the U.S. economy employed 818,000 fewer individuals (equating to 0.5% of nonfarm employment) that what was initially reported. The data is preliminary; the final stats will be released in early 2025.