Conversion to North American Industry Classification System (NAICS) and Other Changes, June 2003
With the release of May 2003 data on June 6, 2003, the CES National Nonfarm Payroll
series underwent a number of changes. The basis for industry classification changed from
the 1987 Standard Industrial Classification System (SIC) to the 2002
North American Industry Classification System (NAICS). In addition, the sample redesign
was completed for the remaining private sector industries. The CES series also
incorporated concurrent seasonal adjustment. Lastly, the federal government series was revised
slightly in scope and definition.
NAICS Conversion
NAICS replaced the SIC system. The Current Employment
Statistics (CES) survey published National data on a NAICS 2002 basis with the release of
May 2003 data on June 6, 2003. SIC-based data will no longer be produced or published; it
will be still be available but not updated past April 2003. Historical time series were
reconstructed as part of the NAICS conversion process. The NAICS-based reconstruction effort
covered all CES published datatypes. All NAICS series have history back to at least 1990.
Other series that were highly interchangeable with an SIC series have history going
farther back in time.
Visit the BLS NAICS webpage for more information about implementing NAICS at BLS.
Completion of the CES sample redesign
June 6, 2003 marked the completion of the CES sample redesign phase-in.
The redesign converted the CES from a quota-based sample to a probability-based sample. The final stage of sample redesign
phase-in resulted in level shifts for average weekly hours, average hourly earnings, production worker, and women worker series.
New levels for these series were computed from NAICS/probability sample-based averages. For further information, read the
section "Completion of probability sample redesign" in the article Recent changes
in the national Current Employment Statistics Survey by BLS Economist Teresa Morisi.
Concurrent seasonal adjustment
Beginning in June 2003, the CES program converted from its current practice of updating
seasonal factors twice a year to updating them every month. Concurrent seasonal adjustment
is technically superior to semiannual updates because it uses all available monthly estimates,
including those for the current month, thereby eliminating the need to project the seasonal
factors. With the introduction of concurrent seasonal adjustment, BLS no longer publishes
seasonal factors for CES national estimates. For more information, please read the following
paper on concurrent seasonal adjustment. (HTML)
(PDF 182K)
Changes to federal government series
The CES series for federal government employment was revised slightly in scope and definition
due to a change in source data and estimation methods. The previous national series was
an end-of-month federal employee count produced by the
Office of Personnel Management that excluded some workers, mostly employees who work in
Department of Defense-owned establishments such as military base commissaries. Beginning in June
2003, the CES national series includes these workers. Also, federal government employment
is now estimated from a sample of federal establishments, is benchmarked annually to counts
from unemployment insurance tax records, and reflects employee counts as of the pay period
including the 12th of the month, consistent with other CES industry series. The historical
time series for federal government employment was revised to reflect these changes.
Last Modified Date: June 27, 2008