Revisions

To make good decisions, our users need data as soon as possible. We balance speed of collection and processing with user needs for accuracy. BLS will often put out preliminary estimates, which are subject to revision. BLS data may undergo several types of revisions.

Revisions from Additional Data

Some BLS programs revise estimates after initial publication. These revisions include additional information that was not available at the time of the initial release. BLS continues to collect reports from the businesses in the samples and uses this information to make the revised estimates more accurate. Why doesn’t BLS wait until it has all the reports to make the estimates and avoid revisions? Data users are intensely interested in the earliest possible look at labor market developments and price changes. The initial estimates are generally accurate, and the revisions improve their accuracy. Revisions do not mean there are problems with the survey design.

The following programs have regular, planned revisions to incorporate additional data:

• Current Employment Statistics (National and State and Area)
• Local Area Unemployment Statistics (State and Metro Area)
• Producer Price Indexes
• Import and Export Price Indexes
• Productivity and Costs (Quarterly and Annual measures)
• Job Openings and Labor Turnover
• Quarterly Census of Employment and Wages
• Business Employment Dynamics

Revisions from Recalculation of Seasonal Adjustment Factors

BLS recalculates seasonal adjustment factors each year. These revisions are planned and announced ahead of implementation. These revisions may cause revisions up to 5 years before the current year. These revisions allow for seasonal adjustment models to incorporate the latest data and identify emerging seasonal patterns.

Revisions from Benchmarking or Re-Anchoring

Benchmarking, re-anchoring, or population control revisions are planned, annual revisions that align estimates to a population count. Benchmarking ensures the estimates reflect the full population counts for the nation. Check out What is “Benchmarking” of Bureau of Labor Statistics Employment Data?

The Current Employment Statistics program benchmarks employment counts to the full employment count from the Quarterly Census of Employment and Wages and other sources.

The Job Openings and Labor Turnover survey re-anchors their employment levels to the benchmarked Current Employment Statistics.

The Current Population Survey re-anchors labor force measures to population counts from the Census Bureau.

The Productivity and Costs programs revise based on revisions from the Current Employment Statistics program, the Current Population Survey, and the Bureau of Economic Analysis revisions to Gross Domestic Product (GDP).

Revisions from classification updates

Other less frequent types of revision may come from new geographic, industry, or occupational definitions. These types of revision tend to occur on a regular, announced schedule.