United States - Total Employment Non-Ag

For the U.S., the BLS "Current Employment Statistics" (CES) program, also known as the payroll survey or the establishment survey, is a monthly survey of approximately 140,000 businesses and government agencies representing approximately 440,000 worksites throughout the United States. From the sample, CES produces and publishes employment, hours, and earnings estimates for the nation, states, and metropolitan areas at detailed industry levels. All national CES employment estimates exclude employees in Puerto Rico. However, the BLS cooperates with both Puerto Rico and the U.S. Virgin Islands to collect data and publish employment estimates independent of national estimates.

Series covering all employees’ hours and earnings were officially added by CES on February 5, 2010, with estimates beginning in March 2006. Historically, CES hours and earnings series covered only production and nonsupervisory employees.

Contents

The Current Employment Statistics Program provides employment, hours of work, and earnings information on a national basis. The data include series for total employment, number of women employed, number of production or nonsupervisory workers, average hourly earnings, average weekly hours, average weekly overtime hours in manufacturing industries. The table below lists the prefixes of the available data series. In order to download a series, simply substitute the "E" in the mnemonics on the following pages with the appropriate prefix from the table below. Some series, even at the total employment level, are available only for states or MSAs. Employment of production workers, average weekly earnings, average weekly hours and average hourly earnings are available for the U.S., some states and some MSAs (the availability is the same as is denoted in the following pages). Series listed below that have an asterisk are available at the national level only.

Hours of work

• Industry classification: NAICS 2012
• Measurement: Billions of hours (Bil. Hrs.)
• Adjustment: Seasonally adjusted, annualized rate (SAAR)
• Native frequency: Quarterly
• Start date: 1964Q1

Net birth-death model

Active:

• Industry classification: NAICS 2012
• Measurement: Thousands (Ths. #)
• Adjustment: Not seasonally adjusted (NSA)
• Native frequency: Monthly
• Start dates:
  • 1999m4 for total
  • 2002m4 for industry branches

Predecessors:

• NAICS, prior to 2019 benchmark - as early as 1999m4 to 2019m12
• SIC - as early as 2000m4 to 2002m3

Industrial classification

The employment series currently cover the 2012 NAICS industrial classification. All major industry sectors include only privately-owned establishments, except for 90-910000 federal government, 90-920000 state government, and 90-930000 local government. For information on the composition of each industry, see the link below.

Concepts

The CES employment series are estimates of nonfarm wage and salary jobs, not an estimate of employed persons; an individual with two jobs is counted twice by the payroll survey.  The CES employment series excludes employees in agriculture and private households and the self-employed.

For more information, see the Concepts section of Chapter 2 of the BLS Handbook of Methods (see link below); this section includes definitions of the types of data available from the survey. 

Uses

A wide array of public and private policy makers use CES data because it is one of the earliest indicators of economic conditions each month. Major users of CES data include many government agencies and entities, financial markets in the United States and around the world, and other business and academic analysts, researchers, and forecasters

For more information, see the Uses section of Chapter 2 of the BLS Handbook of Methods.

Net Birth-Death Model

The CES estimates are adjusted each month by a statistical model in order to reduce a primary source of non-sampling error. This error is the inability of the sample to capture employment growth generated by new business formations in a timely manner. There is an unavoidable lag between a business opening and its appearance in the sample. These new firm births generate a portion of employment growth each month so non-sampling methods must be used to estimate this growth.

Effective with the release of January 2026 data, the establishment survey will change the birth-death model by incorporating current sample information each month. The change follows the same methodology applied to the April through October 2024 forecasts during the 2024 post-benchmark period (see question 9 in the CES Birth-Death Model Frequently Asked Questions).

Moody's Analytics supplements

For the net birth-death model, we produce a seasonally adjusted version of the "total nonfarm" series.

Seasonal factor revision of CPS

For the "Current Population Survey (CPS)" (the household survey, which is reported in the "Employment Situation" news release), the seasonal factor revision is conducted each year in January and revises the most recent five years of SA data. The event may be chronicled in Data Buffet News.

Population controls of CPS

Each year in February (upon the January results), new population controls are applied, from the annual update of population estimates by the U.S. Census Bureau. Historical data are not revised, consequently, there is a break between December and January, such that the CPS results are not directly comparable.

In January, the specifics are announced and the effect of the new controls is tabulated in The Employment Situation.

Benchmark and seasonal factor revisions of CES

Each year in February, two annual processes are conducted: update of seasonal adjustment factors (which revises the prior five years of SA data) and benchmarking (which revises the prior two years of NSA data, if not more).

The specifics are announced in the January issue of The Employment Situation.

Phased revisions of CES

The "Current Employment Statistics (CES)" (the establishment survey, which is reported in the "Employment Situation" news release) is published monthly, in four phases (initial, first revision, second revision, annual benchmark).

Each year in August, the BLS issues an estimate of the benchmark revision that will be made six months hence; this is not itself a revision.

Each year, a benchmark revision is conducted. The sample estimates from the CES are adjusted to universe counts of employment from the QCEW. Benchmarks are principally derived from a separate BLS program that aggregates the employment data reported on unemployment insurance (UI) tax reports that virtually all employers file each quarter with their state employment security agencies. This event is chronicled in Data Buffet News.

The benchmark process revises all SA series and introduces new seasonal factors to be used in subsequent months; these extend back as much as five years. If geographic or industrial classifications are redefined, the revision may impact the entire time series.

To summarize, the phases are:

• Phase 1: First preliminary estimate (initial estimate or "first print"), approximately three weeks after the reference period. Includes employment, hours, and earnings.
• Phase 2: Second preliminary estimate (first revision), one month after initial.
• Phase 3: Third preliinary estimate (second revision), two months after initial.
• Phase 4: Annual benchmark revision:
  • August: Preliminary estimate
  • February: For U.S.-level data
  • March: For states and metro areas

The BLS currently seasonally adjusts one-digit national and state figures and Moody's Analytics adjusts the one-digit MSA figures.

While the Office of Management and Budget defines 342 MSAs (including PMSAs and NECMAs), the BLS only reports data for approximately 275 of these areas.

The BLS790 databank includes redefinitions of geographies and industries as reported by the Bureau of Labor Statistics. Data in this databank are not altered to reflect new geographic definitions beyond those changes made by the BLS. Data redefined, and fixed for breakpoints, are available to subscribers of the Industry Services Database.

Concurrent with dissimination of April 2020 data for the CES Net-Birth-Death Model, the BLS made changes to the methdology due to the COVID-19 pandemic. The changes made were:

1. A portion of both reported zeros and returns from zero in the current month from the sample were used in estimation to better account for the fact that business births and deaths will not offset.
2. Current sample growth rates were included in the net birth-death forecasting model to better account for the changing relationships between business openings and closings.

Effect of severe weather

If there is an unusual circumstance that could potentially impact the response rate to the two surveys, such as a hurricane during the survey period, it is described in a box in the "Employment Situation" press release. Because the survey question is phrased "worked or received pay for any part of the pay period that includes the 12th of the month" (emphasis ours), a brief interruption (i.e., less than one pay period, however defined by the particular respondent) will not be detected.

Federal shutdown 2025

The September 2025 news release was published with a delay. The October news release was skipped. For November, the collection period for the two surveys was extended, and the news release was delayed. For the household survey, 2025Q4 results were skipped (because October was missing) and the annual 2025 results used 11-month averages excluding October.

In 2026, the annual population control adjustments will be applied with the February results intead of with January. January will later be revised.

Further reading

At the source:

• Monthly Employment Situation report: Quick guide to methods and measurement issues
• CES Published Series - Concordance between CES industry code, NAICS code, etc.
• Industries in alphabetical order
• Handbook of methods: Employment, hours, and earnings from the Establishment Survey (pdf)
• Can severe weather, natural disasters, government shutdowns, or other catastrophic events affect the CES survey?

At the source, for CES net birth-death model:

• Historical net birth-death forecasts
• Technical notes
• FAQ

On Data Buffet:

• BLS employment programs compared

• 7 Jul 2023, Phillip Thorne - Revisions.
• 9 Oct 2024, Phillip Thorne - Effect of severe weather.
• 24 Nov 2025, Phillip Thorne - Upcoming change (Jan 2026) to birth-death model. September, October and November news releases.
• 15 Jan 2026, Phillip Thorne - Federal shutdown 2025.