New 5-Year Data from the American Community Survey
The Census Bureau released the latest 5-year ACS period estimates (2009-2013) on December 5. You can access these data via Americna FactFinder
or you can now access it from the MCDC's web site. The easiest ways to access the data are via the two applications at the top of our Quick Links
menu box (just to the right on this and most other MCDC pages):
- The ACS Profile Reports
app now lets you choose 2009-2013 from the Select Period menu.
- The ACS Profile Extract
app now will let you pull extracts from the acs2013 filetype data collection using menus allowing you to choose geography type(s), state(s) or counties, and
Note that there is nothing new with the ACS Trends Reports application. That is because the earliest 5-year period for which we have
ACS data is 2005-2009, which overlaps with this latest period. This time next year we should be getting 2010-2014 data which we shall be able to
combine with the non-overlapping 2005-2009 data to create our first set of trend reports for 5-year ACS data.
Uexplore/Dexter users can also access the new set of preliminary 5yr base (summary) tables in the basetbls
New PUMS Data from the 2010 Census
Better late than never -- the Census Bureau has released a Public Use MicroSample file based on the 2010 census. They did so early last month
without much fanfare (we missed it and we are no most of the Bureau's alert lists). We discovered its existence on the Proximity 1 blog earlier this
week. The data are available in the Bureau's ftp site. This is perhaps a preliminary release since they do not yet have a complete set of
Technical Documentation. But a simple Excel file (available from our data archive as file 2010 PUMS Record Layout.xlsx) provides a layout as well as codebook (value labels). These data are lacking in subject matter (no long form in 2010, and not much of a short form either). But they
are a 1-in-10 sample so they can be valuable for studying rather small demographic groups. Geography is the usual for PUMS files with the smallest
area identified being a PUMA. (2010 PUMAs, of course).
The data are released (for now, at least) as a collection of state .zip files with two fixed format (.txt) files inside: one for the H(ousehold) records and one for the P(erson) records. We have written SAS code to convert these files to a pair of SAS data sets (e.g. mohrecs and moprecs) and a SAS view (mov) that merges the two together as if they were a single rectangularized data set. You can access the data via Uexplore/Dexter at
All we have so far are data for Missouri. Let us know if you need data for other states and we'll consider adding them.
If you are interested in seeing the SAS code we used to read the input files that is avaialble as the cnvtpums.sas file in the Tools subdirectory
New PUMA Master Dataset and Web Application
We recently generated a report showing which of the datasets in our public archive were getting accessed the most. We discovered that
by far the most frequently accessed was the zcta_master dataset in the /pub/data/georef data directory. This dataset is useful for
anyone working with ZCTA (or ZIP code) level geography, and now we have created a comparable data resource for users working with the new (2010) PUMA geography.
We understand that this is a much smaller audience than ZIP users, but there is (or at least should be) demand from users who want to use the PUMA level summary data from the
ACS (important because PUMAs are large enough that they get single-year data every year), or for those doing research with PUMS datasets
(both the more recent acspums datasets and the just-released 2010 PUMS data).
As we did for the zcta_master data, we have also created a new web application that allows
the user to choose a state and enter a PUMA code and get a display of all the relevant geographic info and ACS-based key indictors. The output page also contains links to
two map applications where users can go to actually see where the PUMA(s) are located.
You can now access the Display PUMA_Master application. Links at the bottom
of this page can be used to access the puma_master dataset (via Dexter) and to view the puma_master.Metdata.html file.