Home FAQ Hours & Earnings Advanced Technology


Frequently Asked Questions

Please send us an email with any questions you would like to have answered here!

1. What is a Metropolitan Statistical Area (MSA)?

2. What is the definition of CES Employment?

3. How are the hours and earnings data calculated?

4. How does a strike impact CES data?

5. How many times is CES data revised?

6. What is the benchmark process?

7. Why is employment data seasonally adjusted?

8. Where can I find historical CES data?


1. What is a Metropolitan Statistical Area?

A Metropolitan Statistical Area (MSA) is defined as a county or group of counties that has either (1) a city with a minimum population of 50,000 or (2) an urbanized area (minimum population of 50,000) and a total population of at least 100,000 in the component counties. An urbanized area is defined as a central city and any densely populated area adjacent to that central city. An urbanized area is required to have more than 50,000 inhabitants and be closely settled.

Colorado CES publishes employment data for 7 MSAs:

-Denver-Aurora MSA (Adams, Arapahoe, Broomfield, Clear Creek, Denver, Douglas, Elbert, Gilpin, Jefferson, and Park counties)

-Boulder MSA (Boulder County)

-Colorado Springs MSA (El Paso and Teller counties)

-Ft. Collins-Loveland MSA (Larimer County)

-Grand Junction MSA (Mesa County)

-Greeley MSA (Weld County)

-Pueblo (Pueblo County)

For additional information, read our pdf document on MSA's.

2. What is the definition of CES employment?

CES defines employment as the number of employees who receive pay for the pay period that includes the 12th of the month. This includes both full and part time employees, temporary employees, workers on paid sick or holiday leave, and those who worked for only part of the pay period. People included on multiple payrolls are counted at each establishment. CES data do not include: self-employed, farm workers, volunteers, unpaid family workers, employees on leave without pay, domestic workers, and those on strike for the entire pay period. Striking employees who work a portion of the survey period and are paid are counted in the survey.

3. How are the hours and earnings data calculated?

Colorado CES produces hours and earnings data (AWE: Average Weekly Earnings, AWH: Average Weekly Hours, and AHE: Average Hourly Earnings) for the Manufacturing and Construction industries at the statewide and Denver MSA level. The estimates are derived from reports of gross payrolls and total hours worked by production workers for the pay period that includes the 12th of the month. A production worker is defined as an employee who is closely associated with the operation of production; this includes both full and part time workers. Office workers and those employees whose primary responsibility is to supervise are not included.

Payroll data is reported before deductions of any kind are taken and includes pay for overtime, vacations, holidays, and sick leave. Any type of non-wage cash payments such as bonuses or commissions are not included unless they are paid regularly. Tips and payments for business expenses are not included. Total hours includes overtime hours and hours for paid holidays, vacations, and sick leave.

Average hourly earnings should not be confused with wage rates. Since average hourly earnings are based on gross earnings, they are subject to variable factors such as overtime pay and shifts in the number of higher paid or lower paid employees. Average weekly hours should not be confused with scheduled hours worked. Factors such as unpaid absenteeism, labor turnover, strikes, part-time work, changes in overtime worked, and other variables cause average weekly hours to differ from a firm's scheduled hours worked.

4. How does a strike impact CES data?

Since CES employment counts workers who receive pay for any portion of the reference period (the pay period which includes the 12th of the month), a strike affects the employment count only if the workers are on strike for the entire pay period (do not receive pay for any portion of that pay period).

Example: XYZ Inc is a manufacturing company that employs 2,000 workers. It has a monthly pay period where employees are paid at the end of every month for their work during the previous calendar month. On January 15, 1,000 XYZ Inc employees go on strike and do not return to work until March 1. January employment totals in the manufacturing industry will not be impacted since the workers who are on strike still receive some pay for their work in January. However, February employment totals will be lowered by 1,000 since the workers on strike will not receive any pay for February (they are absent for the entire pay period of February). March employment in Manufacturing will not be impacted by the strike and will include all 2,000 employees at XYZ Inc since the strikers have returned to work for the entire pay period in March.  

5. How many times is CES data revised?

When CES data for a given month is first released, it is referred to as preliminary data. These estimates have been calculated from the portion of the sample that has been received by a predetermined due date in order to have the data released in a timely manner. A few weeks after the preliminary data has been released to the public, the estimates are recalculated to include additional sample that has been received since the preliminary estimates were calculated. Such data is referred to as revised data and is released along with the following month's preliminary data. For example, from the main CES home page, data can be obtained on the most recently released monthly data. That data for the current month is preliminary data, while the previous month's data is revised.

Every year, all CES data goes through another revision, which is termed benchmark. During the benchmark process, all previously released revised data is replaced with benchmark numbers. Read more about the benchmark process under question 6.

6. What is the benchmark process?

One of the important aspects of Current Employment Statistics data is that it is released in a timely manner. The employment statistics are, just as the name implies, current. Since CES employment estimates are typically released only one month behind the reference month, they help to paint a current picture of the economy. However, since the monthly employment statistics are based on a sample of the population of business establishments, the estimates are subject to sampling and nonsampling error. Hence, there is a need to adjust the sample estimates periodically against full population counts.

During the annual benchmark process, the sample estimates are replaced with full population counts from Unemployment Insurance (UI) Tax records, which are filed quarterly by employers with State Employment Security Agencies. UI Tax data, while limited in its ability to provide current or timely information, is useful as a universe or population count since it covers nearly 97% of total non-farm employment. This recalibration of the sample data from UI tax records is done on an annual basis, and the results are released with the next year's January data. For example, January 2004 monthly CES data was released in March 2004, along with 2003 and 2002 benchmarked data. Typically, January's preliminary estimates, along with the previous years' benchmarked data, are released in the month of March.

7. Why is employment data seasonally adjusted?

A seasonal event is an event which occurs at the same time every year, and a data series that is affected by these events is said to exhibit seasonality. Christmas is a good example of a seasonal event; it occurs at the same time every year and affects economic series such as employment trends. Prior to December 25, retailers begin dramatically increasing their employment to account for the increase in shopping leading up to the holiday season. Then in January and February, retail employment declines as employers shed jobs that were established just for the Christmas season. This is something that occurs at the same time every year and can easily be seen by observing a graph of retail trade employment.

However, since one of the main purposes of CES is to gauge the direction of the economy, it is necessary to remove the seasonal component in order to observe the underlying economic trend. During the seasonal adjustment process, the seasonal component is removed from CES employment data so that the actual trends in Colorado's labor market can be examined.

8. Where can I find historical CES data?

Historical CES data can be accessed from the Bureau of Labor Statistics Public Data Query site. This site lets you create customized tables of historical CES data. Choose the historical data you want by selecting 1) the state 2) areas with published data within that state 3) industries 4) and seasonally adjusted or unadjusted data. This site has CES data for every state back to 1990 and also has graphing capabilities.

9. What are the upcoming changes to the CES survey?

 The Bureau of Labor Statistics (BLS) is in the process of making several changes to the Current Employment Statistics (CES) survey to improve its relevance to the needs of data users, as well as its value as an input to other key economic statistics. CES data on employment, hours, and earnings by industry are among the most visible and widely-used economic indicators at national, state, and metropolitan area levels. CES data are also among the timeliest economic indicators, with their release by BLS in the Employment Situation news release shortly after the end of the reference month. To implement the needed additions to the survey while maintaining the viability of the CES program as a high volume, quick turnaround, voluntary survey, BLS carefully reviewed the public’s use of CES data to determine if reductions could be made in some series as a tradeoff for significant overall data improvements that will result from adding several new data items.



Page last updated:  3-11-2015