Among the important economic data developed by state workforce agencies and the Bureau of Labor Statistics (BLS), labor force data (including unemployment estimates) for states and local areas are viewed as key indicators of local economic conditions. Currently, monthly estimates of resident employment, unemployment, and the unemployment rate are prepared for around 7,000 areas-regions, divisions, all states and the District of Columbia, metropolitan and small labor market areas, counties and many cities and towns. Among the many users of these data, state and local governments use the estimates for planning and budgetary purposes and as determinants of need for local services and programs. The state labor force estimates are one of the timeliest subnational economic measures, as figures are released by BLS and the states within five weeks after the reference week, and just two weeks after the national estimates' release.
BLS is responsible for the concepts and definitions, technical procedures, and review, analysis and publication of labor force estimates. State agencies (in Connecticut, the Labor Department's Office of Research) are responsible for the production of the estimates and analysis and dissemination of the data to their data users. A key element of the Bureau's approach to subnational labor force estimation is to ensure that these estimates are comparable to the official concepts and measures of the labor force as reflected in the Current Population Survey (CPS). The CPS is the monthly survey of households that is designed to provide reliable monthly labor force estimates for the nation. To support reliability of subnational estimates, the CPS employs a state-based sample design.
A hierarchy of estimation methods is used to produce state and local labor force data based in large part on the availability and quality of data from the CPS. The strongest estimating method--signal-plus-noise models for current estimation and annual average CPS benchmarks--is employed at the state level. While not reliable enough to use directly, the monthly CPS values are integral to the signal-plus-noise estimation. In order to ensure comparability across states, the annual average employment and unemployment levels from the CPS are used as the benchmarks for the modeled estimates.
Current Modeling and Benchmarking Procedure
A number of methodological and analytical issues have surfaced in the current estimation/benchmark procedures. These include reintroduction of sampling error to monthly estimates, discontinuities between December benchmarked and January model estimates, impaired comparability of data over the year, and inability to address, on a timely basis, "shocks" to the model such as the September 11 terrorist attacks and the onset of an economic recession.
In the current methodology, the state model estimates are developed independent of the national CPS. Although the monthly state CPS input data sum to the national measures, the sum of the state model output estimates generally do not equal the national CPS estimates. In general, the current method of model estimation results in an overestimate of employment and an underestimate of unemployment and the unemployment rate in states as compared to the national CPS estimates.
A Fiscal Year 2001 federal budget initiative provided BLS with resources to improve the methods used to develop state and area labor force estimates, including upgrading and enhancing the modeling approach, extending it to more areas, and incorporating decennial updates to procedures, data inputs, and geography. As part of this major redesign, BLS developed an innovative alternative to model benchmarking that will be part of improved monthly model-based estimation. This alternative addresses longstanding issues related to accuracy and end-of-year revision, and also enhances the analytical capability of the estimates. The redesigned method of estimation ensures that state estimates add up to the national estimates of employment and unemployment each month. In doing so, the benchmark will change from annual average state-level estimates of employment and unemployment to monthly national estimates of these measures, and will be part of current monthly estimation. In this way, economic changes will be reflected in the state estimates on a real-time basis (real-time benchmarking), and end-of-year revisions will be significantly smaller.
The BLS and states are now in a one-year dual estimation period that allows for the evaluation of the proposed methods and systems, and the impact on estimation.
General Methodological Approach
Under real-time benchmarking, a tiered approach to estimation is used. Model-based estimates are developed for the nine Census divisions that geographically exhaust the nation. Connecticut belongs to the New England division. The division estimates are benchmarked to the national levels of employment and unemployment on a monthly basis. The benchmarked division estimates are then used as the benchmarks for the states within each division. That is, state model-based estimates are controlled to add to the division's employment and unemployment. The distribution of the monthly benchmark adjustment to the states is based on each state's monthly model estimate. In this manner, the monthly state employment and unemployment estimates add to the national levels, precluding differences between the sum of states and the national estimates. Annual historical benchmarking will still continue for state estimates with the updating of model inputs, model re-estimation, and incorporation of updated population controls performed each year. However, the impact on the historical series of these benchmark activities is expected to be fairly small, especially in comparison with annual revisions using the current methodology.
Estimation Period and Implementation Plan
As part of implementation, a dual estimation period began with January 2004 data so that the proposed methodology and operational systems could be reviewed in a real-time environment and the impact on estimation could be evaluated. An analysis of the numbers produced by states so far indicates consistency with the redesign objectives of addressing issues in current estimation. In general, the new models with real-time benchmarking result in higher estimates of unemployment and the unemployment rate, and lower estimates of employment, and thus remedy the consistent under- and over-estimation mentioned earlier. (A comparison of Connecticut's redesign estimates to those made using the present methodology, also shows higher unemployment and unemployment rates and a mix of higher and lower employment estimates in the redesign figures for the months estimated so far in 2004.) The new estimates of both employment and unemployment of State residents are expected to be more accurate using the new methodology than with the current procedures.
The redesigned estimation methodology is planned to be implemented with labor force, employment and unemployment estimates for January 2005 to be published in March 2005. Historical series from January 1976 forward will be replaced with estimates based on the redesigned models. Additionally, revised data from 2000 forward will reflect Census population estimates updated to account for changes in births, deaths and migration that have occurred since the decennial Census.
Commissioner James F. Abromaitis of the Connecticut Department of
Economic and Community Development (DECD) announced that Connecticut communities authorized
862 new housing units in October 2004, a 23.1 percent decrease compared to October of
2003 when 1,121 units were authorized.
The Department further indicated that the 862 units permitted in
October 2004 represent a 23.1 percent decrease from the 1,121 units permitted in September 2004.
The year-to-date permits are up 15.1 percent, from 8,390 through October 2003, to 9,659 through October 2004.
Eight of the ten Labor Market Areas showed losses compared to a year ago. Bridgeport led all
municipalities with 46 units, followed by Danbury with 40 and Hartford with 28. From a county
perspective, only Hartford and Tolland counties showed year-to-date losses.
Return to Top