RE: [CTPP] RE: Question about PUMS data and Geography

I have experience using both the PUMS data and the LEHD data. They each have their strengths and weaknesses. The PUMS data is more comprehensive and more reliable, while the LEHD data has more detailed and flexible geographic scope. For the PUMS data, use the 5% data for the 2000 Census, or the 1% data from the more current ACS - they both rely on the same PUMS geography of 100,000 population Public Use Microdata Areas (aka PUMAs). The PUMS data is a sample of all industries, public sector and private, employees and proprietors alike. The geographic limitation can be a problem if you are seeking county-to-county commute flows involving counties of under 100,000. We have that problem here in Olympia, WA, where some of our flows are with small counties, some with large, leading to uneven data availability. Also, the PUMS data is subject to fairly large margins of error, so you must be cautious in drawing conclusions about trends. Go ahead and draw conclusions, but only after doing your proper margin of error analysis to look for statistical significance. The necessary techniques and formulas are spelled out in the documentation available on the Census ACS PUMS website. LEHD takes a combination of Census and unemployment insurance administrative data, then -- to maintain confidentiality -- creates a synthetic dataset and reports the results. This is powerful in that it shows detail down to the block group level. The weaknesses are that the process of synthesizing can cause weird results out at the margins. In our area, Census data shows about 5,000 workers (out of a workforce of about 100,000) commuting from the Olympia area about 60 miles north to the Seattle area (where there are about 1,300,000 jobs). LEHD creates an estimate of about 10,000 making the commute. Looking at the LEHD documentation, I conclude that the algorithms they use account for the doubling of the figure. To summarize, their algorithms take the known data for employment by location from unemployment insurance administrative data, combined with the known data for residences by location from Census, then stir it all up with modeling to produce estimated flows of workers, based in part on Census flow data. The small flows between my county and Seattle are down in the margin of error of the model, inconsequential to the relatively massive Seattle employment base universe. But that error in the flow estimate is much more consequential to our smaller local labor force universe. Examining other flows from LEHD, it appears that it works pretty well when you are not working out at the tail of the bell curve, the way we are. With one added caveat: LEHD does not include Federal employment. For us, this creates another limitation, since there is a large Army-Air Force base complex next door (Fort Lewis Army Base and McChord AFB, between Olympia and Tacoma), with thousands of military and civilian defense workers. They get left out. So LEHD says there are thousands fewer commuters to Pierce County (Ft. Lewis) than Census finds, and thousands more commuters to King County (Seattle) than Census finds. So my advice would be to look at both datasets and see which better meets your needs. You might even be able to create your own estimated flows by using the geographic power of the LEHD data to disaggregate the more comprehensive PUMS data, coming up with your own synthetic estimates. Pete Swensson, Senior Planner Thurston Regional Planning Council 2424 Heritage Ct. SW Olympia, WA 98502 (360) 741-2530 (direct line) (360) 956-7575 (main desk) (360) 956-7815 (fax) swenssp(a)trpc.org   This e-mail and any attachments are for the use of the addressed individual.  If you have received this e-mail in error, please notify our systems manager.  TRPC has taken responsible precautions to ensure no viruses are present in this e-mail, however we do not accept responsibility for loss or damage arising from the use of this e-mail or attachments. -----Original Message----- From: ctpp-news-bounces(a)chrispy.net [mailto:ctpp-news-bounces@chrispy.net] On Behalf Of Filipi, Mark Sent: Friday, March 27, 2009 8:28 AM To: 'ctpp-news(a)chrispy.net' Subject: [CTPP] RE: Question about PUMS data and Geography Martin, Rather than PUMS data, I would suggest you check out the LEHD data available through the Census website. Go to www.census.gov, then click on Local Employment Dynamics in the Business and Industry section. That will take you to the LEHD website. On the right side are a list of links. VirtualRDC will get you to a location that will explain how to download and use the LEHD data. This is data connects home locations-work locations based on data available from other sources. It is supposed to essentially include all employment except work-at-home, so it is a strong data set. Depending on what you are trying to illustrate, On The Map may suffice. You can find it on the LEHD website as well. Mark Filipi, AICP PTP Manager, Technical Planning Support Metropolitan Transportation Services Metropolitan Council 390 Robert Street North St. Paul, MN 55101-1805 651-602-1725 Fax: 651-602-1739 -----Original Message----- From: ctpp-news-bounces(a)chrispy.net [mailto:ctpp-news-bounces@chrispy.net] On Behalf Of Catala, Martin Sent: Wednesday, March 25, 2009 2:38 PM To: ctpp-news(a)chrispy.net Subject: [CTPP] Question about PUMS data and Geography Hello, I am looking to create some commute flow maps and I think I remember someone mentioning (between all the please remove me from this list e-mails) that flow maps can be created by using ACS data and the PUMS geography. I was at the Census site trying to download the PUMS geography files and found that there were two types 1% and 5%. Does anyone know which geography file I should use? The 1% or the 5%? Thanks in advance, Martin Catala GIS Manager CUTR _______________________________________________ ctpp-news mailing list ctpp-news(a)chrispy.net http://www.chrispy.net/mailman/listinfo/ctpp-news _______________________________________________ ctpp-news mailing list ctpp-news(a)chrispy.net http://www.chrispy.net/mailman/listinfo/ctpp-news