EMSI Data Release Notes: 4th Quarter 2010 Beta

November 30, 2010 by Emsi Burning Glass

image of Lightcast gradiant

We are pleased to announce the beta, or preview, release of our fourth quarter 2010 dataset. This release primarily reflects important updates from some of our 80+ data sources. Note: this release is a preview to be used for testing and feedback only. We do not yet recommend using this data release for other purposes until it is finalized, which is scheduled for mid-December.

Updated Data Sources in This Release

  • Quarterly Census of Employment and Wages, Q1 2010 and 2006-2009 revisions (BLS). These updates significantly influence our new 2010 job estimates. In addition, we now use 2006-09 QCEW data that has been revised by the BLS since its initial release.

  • State & Local Personal Income (BEA). The BEA has released 2009 State Personal Income data as well as revised historical data going back to 1990. These new and revised figures affect the non-covered portion of EMSI Complete, as well as agriculture and government sectors in both EMSI Covered and EMSI Complete.

  • Current Employment Statistics, January-September 2010 (BLS). This dataset helps us form an estimate of final 2010 annual job averages.

  • State Projections. We now use more recent (in most cases 2008-18) projections from Alaska, California, Delaware, Idaho, Kentucky, Louisiana, Michigan, Nevada, North Dakota, Oklahoma, Rhode Island, South Carolina, Tennessee, West Virginia, and Wisconsin. As more states release 2008-18 projections, they are weighted more heavily than the national 2008-18 projection from the BLS. Due to the complex interactions between various projection sources, EMSI final projections across all areas and industries are affected.

  • Annual Input-Output Tables, 2008 (BEA). EMSI’s input-output modeling system (Economic Impact module) now includes data from the 2008 annual release of the national I-O tables.

  • American Community Survey, 2009 (Census). We now use ACS microdata from 2007-2009 to inform non-covered staffing patterns and occupational employment.

Methodology Changes

Noncovered Staffing Patterns and Occupation Data: We have made significant improvements to our processes of generating non-covered staffing patterns and occupational employment estimates (primarily self-employed workers) from American Community Survey microdata. The primary changes are in how we assign estimates to detailed occupation and industry categories when only aggregate categories are available in the raw data. In addition, we now control national self-employment by occupation to percentages in the BLS’s national estimates of self-employment by occupation.

Special Notices

  • Widespread industry reclassification in the Energy sector: In Q1 2010 QCEW data from the BLS, thousands of private-sector jobs in several states have been reclassified from NAICS 22111 (Electric Power Generation) to NAICS 22112 (Electric Power Transmission, Control, and Distribution). About 65,000 jobs at the national level were affected. This change does not reflect an economic shift, but rather a difference in how employers self-identify within the public unemployment insurance system. EMSI does not currently correct for such changes.

  • Reminder on temporary Census worker employment: Our 2010 employment estimate for Federal Government (civilian) includes the effects of temporary Census worker jobs, which peaked in May. This effect is removed from projections. The 2010 federal government staffing pattern (and occupational data affected by it) does not currently account for this effect, due to lack of data on the occupational categorization of temporary Census workers.

Questions? Contact Us

If you have any questions or feedback about the new data, please contact Customer Solutions at (866) 999-3674 or via the online chat feature within EMSI Analyst.