Volume 9, No. 2


Go Home

Featured Articles:


An Examination of the Illinois and United States Economy Using Current Employment Statistics Data

Demographic Issues Facing the “New” Workforce

New Standards and Geographic Definitions for Metropolitan Statistical Areas

Important Changes to Illinois Statewide and Sub-State Labor Force Estimates


Archives

 

Important Changes to Illinois Statewide and Sub-State Labor Force Estimates

By: Rich Reinhold

 

For more than 30 years, the Illinois Department of Employment Security has worked in cooperation with the U.S. Bureau of Labor Statistics (BLS) to produce Illinois statewide and local area labor force estimates, including unemployment rates. This federal-state cooperative initiative is known as the Local Area Unemployment Statistics (LAUS) program.

Several years ago, the BLS started work on the LAUS Redesign, a multi-year project designed to improve the accuracy of state and local area labor force estimates and introduce changes related to the 2000 Census. The key components of the LAUS redesign project include: new, enhanced statistical models for producing labor force estimates; benchmarking statewide monthly labor force estimates in "real time"; improved methods for estimating sub-state employment and unemployment estimates and introducing 2000 Census based data inputs. This article will describe the LAUS Redesign project and how its implementation in 2005 will impact Illinois Statewide and sub-state labor force estimates. These changes were also recently announced in a Jan. 25, 2005 Federal Register Notice as well as in on-line question and answer documents at the BLS Web site.

New Labor Force Statistical Models
Statistical models have been used to produce Illinois statewide labor force estimates since 1996. The monthly, household-based Current Population Survey (CPS), conducted by the U.S. Census Bureau for the BLS, is the official source for national labor force estimates. However, the CPS household sample is not large enough to support direct estimation at the state level - hence the need for modeling statewide labor force estimates. The state models include data from monthly CPS but also state supplied inputs including total non-farm jobs and Unemployment Insurance (UI) benefits claims. The statewide models estimate the true monthly labor force values and remove sampling error from the CPS.

In 2005, the current labor force statistical models for all 50 states, New York City and the Los Angeles Metropolitan Area will be replaced with new, third generation statistical models. Among the advantages of the new models are:

  • The new statistical models are able to capture large changes in the economy sooner than the current statewide models. Examples include economic recessions or non-economic events such as the September 11 terrorist attacks.
  • The labor force estimates developed from the new statewide statistical models include monthly error or reliability measures. The current statewide model does not produce reliability measures.
  • Analysis of the monthly labor force estimates is enhanced with improved seasonal adjustment procedures. Employment and unemployment estimates are adjusted for seasonal patterns in the labor force within the model structure and not externally.
  • The new statewide models will produce comparable monthly and annual labor force estimates back to 1976. The current model has estimates back to 1978.
  • New statistical models will also be introduced for six additional metropolitan areas, including the Chicago-Naperville-Joliet Metropolitan Division. The Chicago area model will replace the current handbook or building block method for producing labor force estimates.

Real Time Benchmarking of Statewide Labor Force Estimates
Under current methodology, Illinois monthly model-based labor force estimates are benchmarked at the end of each year to labor force estimates from the Illinois Current Population Survey for the most recent calendar year. The CPS labor force estimates serve as a control for the monthly model estimates, ensuring that the revised estimates do not average above or below the annual CPS benchmark. One drawback of this approach is that error from the CPS household survey is placed back into the monthly statewide labor force estimates.

  • The new statistical model benchmarks statewide labor force estimates to national labor force estimates each month (or in real time) instead of at the end of each year. The monthly national CPS labor force estimates are more reliable than annual statewide CPS benchmarks.
  • Real-time benchmarking ensures that large or sudden changes in the economy are captured during the production year and not postponed until annual revisions are completed.
  • The sum of statewide labor force estimates will equal national labor force estimates. The divergence between the sum of states and the nation widened in 2001 and 2002 following the recession and the September 11 terrorist attacks.
  • Annual revisions to monthly labor force estimates will be smaller.

Improved Procedures for Adjusting sub-State Worker Data for Residency and Multiple Job Holding
The LAUS program develops employed estimates based on the geographic area of residence, as opposed to the location of jobs or the place of work. Employed persons are counted just once, even if they hold multiple jobs. At the sub-state level, employed estimates are developed by adjusting total non-farm jobs for worker residency and multiple job holding.

  • The current method relies on single adjustment ratio for each sub-state estimating area, including resident employed data from the most recent decennial Census and non-farm jobs totals for the same time period.
  • The main limitation of the current method is that it does not take into account job growth or decline that occurs outside the estimating area.
  • The premise of the new procedure is that resident employed is not only tied to the relationship between employed residents and jobs located within the same area, but also the commutation of employed residents to jobs located in adjacent or nearby areas. Thus, resident employed will be impacted by job expansion and losses in nearby counties.
  • Under the new method, multiple residency adjustment ratios are produced for each estimating area. Residency adjustment ratios are developed for residents who both live and work in the estimating area as well for residents in the estimating area who commute to adjacent or nearby counties.
  • The new residency adjustment method will be used to develop sub-state employed estimates for years 2000 and later.

Improved Method for Estimating sub-state Unemployed Labor Force Entrants
Unemployed labor force entrants include those who have either entered the labor force for the first time (new entrants) or have returned to the labor force following a period of absence (reentrants). These two groups represent approximately 40 percent of total unemployed.

  • The current method for estimating sub-state labor force entrant unemployed relies primarily on national seasonal patterns in labor force entrant unemployed, adjusted to reflect concentration of youth in the estimating area to experienced labor force in the same area.
  • The current method underestimates the level of labor force entrant unemployed. This underestimation is now addressed by controlling sub-state unemployed levels to statewide levels of unemployed. The statewide adjustment, however, can distort monthly patterns in labor force entrant unemployed at the sub-state level.
  • Under the new method, monthly statewide estimates of labor force entrant unemployed are produced in an economic model that uses historical statewide data from the CPS household survey. The CPS reports monthly statewide estimates of various categories of unemployed, including labor force entrants.
  • State model-based estimates of new entrant unemployed are allocated to sub-state areas based on each area's share of population ages 16-19.
  • State model-based estimates of reentrant unemployed are allocated to sub-state areas based on each area's share of population age 20 and older.
  • Monthly labor force entrant unemployed estimates for years 2000 and later will be developed under the new method.

2000 Census Updates for Sub-State Agriculture Workers, Self Employed, Unpaid Family Workers and Private Household Workers (Domestics)
The LAUS program reports data for all employed persons including those who do not work in non-farm business establishments. These individuals represent less than 10 percent of total employed and include agricultural workers, the self- employed, unpaid family workers and private household workers (domestics).

  • State and national employed data for agricultural workers, the self employed, unpaid family workers and private household workers are captured through the CPS.
  • At the sub-state level, employed estimates for these groups are produced using data from the most recent decennial Census and monthly seasonal factors developed by the BLS.
  • In 2005, the benchmark level of employed for agricultural workers, the self employed, unpaid family workers and private household workers will be updated using data from the 2000 Census.
  • The 2000 Census employment updates will be made to monthly employed estimates for years 2000 and later.

For more information contact:
Rich Reinhold
e-mail: Richard.Reinhold@Illinois.gov
Phone (312) 793-5896

 
 

 

 

LMI Source homepage |
IDES homepage |
E-mail for Questions about
Labor Market Information