Data Sources and Limitations


for Official Oregon Series vs Official BLS Series

Current Employment by Industry

Estimates of current employment by industry are made based on a sample of businesses that are surveyed each month. Data for the most recent month is considered preliminary, and is revised on the following month. Each year, monthly data for at least the past two years are revised based on information gathered from employers' payroll tax records. Employment figures represent employment by place of work, therefore multiple job holders are counted more than once. Workers involved in labor-management disputes are excluded from the nonfarm payroll employment totals.

What are the differences between the official Oregon series and the official Bureau of Labor Statistics (BLS) series?

Summary

The primary difference is that the official Oregon series incorporates universe covered employment and wages (QCEW) data soon after that data is available. Since QCEW data is updated once every three months with the most recent quarter's data, the official Oregon series incorporates this new data once every three months. In this way, the official Oregon series is likely to better reflect the level of employment on an ongoing basis.

In contrast, the official BLS series updates its employment levels with the latest QCEW data once each year. In the worst-case situation for example, the official BLS estimates for the data month of December 2018 will be anchored to September 2017 QCEW data. This is a 15-month lag. In contrast, for this situation, the official Oregon series will be anchored to June 2018 QCEW data, which will result in only a six-month lag.

Since the CES estimates are based on a survey of establishments, the data can often diverge from the total employment trends of all establishments in a given data series. The quarterly benchmarking process incorporated into the official Oregon series mitigates this problem by bringing the estimates closer in line with reality as soon as the universe QCEW data becomes available.

Geographic Coverage

The official Oregon series includes monthly employment estimates for counties, metropolitan statistical areas (MSAs), and for Oregon-statewide. The official BLS series includes the MSAs and statewide, but not the other counties.

Timing of Revisions

The official Oregon series and the official BLS series both have a preliminary and a final version for each month of data. For example, for the data month of August, a preliminary figure is published in the second half of September. Then an August revised figure is published in the second half of October. Note that this second version of the August employment data may be referred to as "August revised" or as "August final." However, it really isn't final in the sense that the August employment data will be revised again at a later date, once the sample-based estimates have been replaced (also known as "benchmarked") with universe data from the covered employment and wages program.

For the official Oregon series, data for recent months is revised using updated QCEW data four times per year. For the official BLS series, data is updated with QCEW only once per year. The following tables display the timing of these data revisions. The right-most column shows which data months are revised effective with the timing of the publication each month on QualityInfo.org.

Official Oregon Series-Data Revisions Schedule
 

Date of Data Release for Statewide Data

Date of Data Release for County and MSA Data

Most Recent Data Month (this will be a preliminary estimate for this data month.)

Latest Month of QCEW Data Used in the Revision Process

Data Months Revised at this Time

March 6, 2018

March 13th

January

September 2017

April 2016 thru December 2017

March 20, 2018

March 27th

February

September 2017

January

April 17, 2018

April 24th

March

September 2017

February

May 15, 2018

May 22nd

April

December 2017

October 2017 thru March 2018

June 12, 2018

June 19th

May

December 2017

April

July 17, 2018

July 24th

June

December 2017

May

August 14, 2018

August 21st

July

March 2018

April 2017 thru June 2018

September 18, 2018

September 25th

August

March 2018

July

October 16, 2018

October 23rd

September

March 2018

August

November 14, 2018

November 20th

October

June 2018

April 2018 thru September 2018

December 18, 2018

December 27th

November

June 2018

October

January 15, 2019

January 22nd

December

June 2018

November

March 5, 2019

March 12th

January

September 2018

April 2017 thru December 2018

March 19, 2019

March 26th

February

September 2018

January

April 16, 2019

April 23rd

March

September 2018

February

May 14, 2019

May 21st

April

December 2018

October 2018 thru March 2019


Notes for the above table:
At least six months of data are revised at the time of the press releases in May and November. At the time of the August press releases, 15 months of data are revised, to incorporate the effects of the first quarter non-economic code changes. At the time of the first March press releases, up to 24 months of data are subject to revision, due to the more-comprehensive annual benchmarking processing.

At the time of the March and April press releases, for statewide and MSA data, the official Oregon series exactly matches the official BLS series since both series have benchmarked using the prior September QCEW data.

Official BLS Series-Data Revisions Schedule
 

Date of Data Release for Statewide Data (on QualityInfo.org)

Date of Data Release for MSA Data

Most Recent Data Month (this will be a preliminary estimate for this data month.)

Latest Month of QCEW Data Used in the Revision Process

Data Months Revised at this Time

March 6, 2018

March 13th

January

September 2017

April 2016 thru December 2017

March 20, 2018

March 27th

February

September 2017

January

April 17, 2018

April 24th

March

September 2017

February

May 15, 2018

May 22nd

April

September 2017

March

June 12, 2018

June 19th

May

September 2017

April

July 17, 2018

July 24th

June

September 2017

May

August 14, 2018

August 21st

July

September 2017

June

September 18, 2018

September 25th

August

September 2017

July

October 16, 2018

October 23rd

September

September 2017

August

November 14, 2018

November 20th

October

September 2017

September

December 18, 2018

December 27th

November

September 2017

October

January 15, 2019

January 22nd

December

September 2017

November

March 5, 2019

March 12th

January

September 2018

April 2017 thru December 2018

March 19, 2019

March 26th

February

September 2018

January

April 16, 2019

April 23rd

March

September 2018

February

May 14, 2019

May 21st

April

September 2018

March


Notes for the above table:
At the time of the March and April press releases, for statewide and MSA data, the official Oregon series exactly matches the official BLS series since both series have benchmarked using the prior September QCEW data.

How the official Oregon series is calculated

The calculation method used for the release of data in May, August, and November incorporates the most recently available QCEW quarterly data.

For example, with the release of data in May 2018, the 4th quarter QCEW data is available; therefore, that data is incorporated into the monthly employment estimates. Continuing with this example, essentially the October 2017 revised data uses the September 2017 and October 2017 QCEW data for each series to produce an over-the-month QCEW link. This link is applied to the September CES data. Thus, the newly revised October 2017 CES figure follows the trend of the universe QCEW data.

Then the next two months of November 2017 and December 2017 are also based on QCEW over-the-month change links.

For the subsequent months, the rounded over-the-month difference in the CES estimates is applied additively to the previous month's CES estimate. This is the procedure used for statewide and the MSAs to ensure that the over-the-month changes exactly match the official BLS series. Please note that although the over-the-month changes mirror the official BLS series, the levels for a specific data series are likely to be either higher or lower, due to the revision process that incorporated the latest QCEW data for prior months.

Also, note that the county data uses a multiplicative link for post-benchmark months, rather than the additive link required by BLS for statewide and the counties. Oregon economists determined that the multiplicative link approach was warranted for the county estimates.

For the data release in August, the official Oregon series incorporates the non-economic code changes effective for first quarter QCEW data. The formulas for incorporating these adjustments are the same as those used by BLS during their annual benchmarking process.

For the data releases in March and April, the official Oregon series exactly matches the official BLS series for statewide and the MSAs. During that period, both series have incorporated September QCEW as the latest month in the replacement process.