The Confluence of Change: 2005 Labor Force Estimates
In June, Indiana submitted revised estimates for January through May 2005 to the Bureau of Labor Statistics. These figures replaced earlier monthly estimates, reflecting the latest iteration of the 2004 benchmarked estimates and correcting some errors resulting from faulty ratios used as inputs. These labor force estimates, along with earlier, preliminary estimates for 2005, implemented a variety of changes discussed below, all designed to improve the quality and accuracy of our monthly estimates and of the re-estimated data for prior years.
“Normal” Benchmarking Revisions
In a normal year, the benchmarking process takes updated information into account, including such items as monthly claims data received after the estimates were completed and updated estimates of nonfarm employment from the Current Employment Statistics program that include six months of universe employment data. These updated data inputs generally create some revisions to the previously published monthly data in and of themselves.
Effective with the 2005 estimates, the Local Area Unemployment Statistics (LAUS) program adopted changes to the composition of Indiana’s metropolitan statistical areas (metros) that were announced by the Office of Management and Budget in the summer of 2003. Three new metros were introduced—Anderson, Columbus and Michigan City—and most others had changes to the counties which compose them, either adding or dropping one or more counties. Historical data for the revised metro definitions, incorporating the new modeling software and population controls discussed below will be available back to 2000 once all the benchmarking is complete.
A third generation of the LAUS modeling software was introduced for 2005, after a period of dual estimation during 2004. This new software, while continuing the use of most state-specific inputs, allows for “real-time” benchmarking using employment and unemployment estimates from the national Current Population Survey (CPS). The monthly state estimates will be adjusted to census region totals derived from the national estimates, which use the CPS (a detailed, monthly telephone survey of participant households’ work or job-search activities) as their methodology. Individual state portions of the CPS survey results are inadequate to use as a single source in estimating the labor force; however, state CPS data are used as one component for monthly estimation in both the old and new models. This improvement to methodology is expected to result in smaller revisions to monthly state estimates.
Population Controls Change
Estimates for 2005 (and benchmarked estimates for 2000 to 2004) reflect a change to 2000 as the base year in defining the population for the state, metros, counties and cities. The population and the ratio of employment to population derived from the census are key inputs in monthly estimate processing. Census data from 1990, supplemented by interim estimates, were used as inputs from 1995 to 2004.
Vicki Seegert, Manager
Advanced Economic and Market Analysis Group, Indiana Department of Workforce Development