EMSI releases new data set
Update 2 (5/1/2008): EMSI has released the revised dataset, now incorporating state projections from New York (the last state to provide them).
Update (4/16/2008): We have temporarily reverted to the previous version of our data pending a review, testing, and revision of the new industry/occupation projections. The new data, with revised projections, will be re-released on Thursday, May 1, 2008. We apologize for any inconvenience to our users.
Economic Modeling Specialists Inc. (EMSI) is pleased to announce the release of its spring 2008 data set. All of our industry, occupation, demographic, indicator, and regional input-output modeling data have been rebuilt using our most sophisticated integration process ever and the latest available data from our 80-plus government sources. This update affects past, current, and projected data, so numbers will differ slightly from those contained in past EMSI data releases.
Here is an overview of the additions and changes in the new data (note that a few elements are still being added but will be available by the end of the week):
New County-Level Indicators
In this data release, we have continued to expand our portfolio of county-level economic indicators. We’ve added:
- Mortgage data from the Home Mortgage Disclosure Act (HMDA), including number of loans and total loan amounts from regular and subprime lenders (identified by the US Dept. of Housing and Urban Development).
- (Coming soon) Overall workforce indicators from the Census LEHD data set, including job churn (turnover) and percentage of workforce by age group and gender.
- Social Security beneficiaries and benefit amounts, both total and for retirees only.
- Building permits issued by year for both single- and multi-family units.
- Annual and monthly overall unemployment rate from the Bureau of Labor Statistics (BLS) - monthly figures coming soon.
- Estimates of total persons in poverty by county, children in poverty, and median household income, all from the Census’s Small Area Income and Poverty Estimates (SAIPE) program.
- (Coming soon) Personal income information from the US Bureau of Economic Analysis (BEA), including dividend/interest/rent income, transfer payments (Social Security, welfare payments, veterans’ benefits, government grants and loans, unemployment benefits, etc), and overall per-capita personal income.
Methodology Updates for Published Industry/Occupation Data
- We use the latest data sources, especially 1st and 2nd quarter 2007 covered employment from the Bureau of Labor Statistics (BLS).
- The Economic Impact input-output model uses updated national 2002 benchmark and 2006 annual tables from the BEA, inclusion of Census of Governments data, and an improved regionalization methodology for a more updated and accurate model.
- More sophisticated combination of covered employment (BLS) and complete employment (BEA), to account for BEA’s occasional re-classifications of jobs from one industry or county to another. This results in a more accurate count of proprietors and non-covered wage and salary jobs in cases where the BEA’s “complete” employment numbers are actually lower than the BLS’s covered employment numbers. This creates better local data, although it means that our state and national totals are no longer strictly benchmarked to BEA totals.
Improved, More Localized Projections
In the past, we controlled our own local projections first to statewide and then to national projections. We have since found that this sometimes places too much weight on national trends, which could override local trends. We now put more weight on our own local projections, official sub-state area projections (where available), and official statewide projections. (We still use national projections to adjust our own local projections initially, mainly because national projections are the most up-to-date and have the most industry-by-industry detail.) Overall, this results in more localized industry/occupation projections.
Finally, we have made minor adjustments to our demographics projection methodology to improve projections in certain counties where different racial/ethnic populations have rapidly divergent growth rates compared to statewide trends.
