Thanks Sam! I am interested in identifying systematic issues so that
approaches to improve the overall quality of the data can be made. The
recent discussions here, and on the lehd-ltd listserv have indicated a
problem with School Districts and employment at individual schools.
to hear that Ohio has improved the ES-202 file to address this! I
like to know if there are other industry categories that have similar
problems and what approaches to solving the problem have worked!
The Multiple Worksite Report http://www.bls.gov/cew/cewmwr00.htm
is a VOLUNTARY component of the QCEW/ES-202 program.
There are 2 papers that may be of interest to you, linked to the FHWA
Census page: http://www.fhwa.dot.gov/planning/census/lehd.htm
research sponsored by the Bureau of Transportation Statistics examined
from Florida and Illinois. Results for Illinois found that the LED 2001
data had much fewer shares of people working in their county of
compared to ACS. Please see the paper by Wende Mix presented at the
TRB Conference on Census Data for Transportation Planning. This was
attributed to over-assignment to headquarters locations.
Another early paper which discussed multiple worksite businesses (also
linked to the FHWA website) is by Julia Lane and Marc Roemer which
indicates that there are 2 main issues: 1) a business is NOT
identified as having
multiple sites, and 2) even if the business is identified as having
multiple sites, the list of those sites may be incomplete.
Lastly, just as a reminder, the universe of the QCEW/ES-202 is different
from Census or ACS household surveys.
1. QCEW does not include self-employed. LEHD program is exploring use
Census Business Register to add self-employed. In the Census 2000,
about 10% of workers reported they were self-employed.
2. QCEW does not include military or federal government or postal
workers. LEHD program is working with OPM to get federal employees
3. The coverage in the Agriculture & Mining sectors is not always
complete depending on the state.
4. LEHD data count both jobs and workers. LEHD uses the "primary" job
concept, i.e. the job with the highest earnings in each quarter, to
assign a single workplace geography to a worker.
I hope I am not sounding negative about LEHD! I think it has a lot of
potential, after more data cleaning particularly with multiple site
businesses at the front end, and integration with additional files.
FHWA Office of Planning
206-220-4460 (in Seattle)
For those of you who didn't already know, the US Dept of Education
maintains a national geo-codable database of K-12 schools at
(go to Data Resources>Build A Table), where you
select a particular geographic area and fields include both school
and geocodable street address. I've used this site on several occasions
supplement ES202 data to geocode education employment throughout the
of Ohio and for multi-state metro areas. FYI, our ES202 employment
in this state beginning in 2005 switched from coding all public K-12
employment to the district administrative offices to the individual
Ohio DOT, Office of Technical Services
1980 W. Broad Street, Columbus, OH 43223
Phone: 614-644-6796, Fax: 614-752-8646
"You can observe a lot just by watching." - Yogi Berra