Missouri Census Data Center

What's New at the MCDC in 2016

Go to most recent posting(s).

Posted 1-6-2016

New Cycle of Data for the beareis Filetype

We have processed the latest economicn time series data at the state and county levels from the Bureau of Economic Analysis. Access these in the beareis data directory. The new data sets contain time series starting in 1969 and going through 2014. (Except there was no update to the wag (wage and salary) data sets. The BEA explained that all the items in this data set were also available in the PRO data sets (economic profile).

Posted 1-6-2016

New Cycle of Data for kidscnt Filetype

We have processed a number of resources (mostly from Missouri state agencies) to create an updated version of the database used in the Missouri Kids Count project. See the 2015 new data collection. .

Posted 1-11-2016

New Inter-censal State and County Pop Estimates 1980-1989

We added these data after a user request for some 1980 data by age, race and sex. The new data sets are in the popests2 data directory. The new data sets are usageracesex8089 which has county level data by 18 5-year age cohorts for 6 race/sex combos (white, black and other race by sex), and usageracesex8089Summaries, which contains summaries of the data in the first data set but summarized at various levels including nation, state and county geographic levels, and combinations of race and sex. The variable smrytyp is the key to understanding what is being summarized.

Posted 1-26-2016

Problem With Missing 2k PUMA on MABLE14

Discovered and fixed a problem with the MABLE14 database, affecting the geocorr14 web application. The 2000 PUMA and 2000 Pop count variables were missing from the mable data sets. Found and fixed the problem.

Posted 1-27-2016

Comparing Post and Intercensal Estimates for 2005-2009

There is an issue that has been raised recently regarding problems with comparing 5-year ACS periuod estimates for the non-overlapping (for the first time!) 2005-2009 and 2010-2014 data. The problem has to do with the sampling and weighting done when processing these two sets of data are different. The 2005-2009 data were weighting based on the then-best-available estimates at the county level with some combinations of age, race, sex. Hispanic. For the 2010-2014 they used the post-2010 estimates. So the latter period base estimates were "informed" by the results of the 2010 census, while the base numers used in the earlier period were not. The 2010 census had been taken yet. But in hindsight we have the "adjusted" intercensal estimates for these years. This is where the Bureau goes out and applies adjustments to the estimates so that they conform to what we found in April, 2010.

We ran a setup to compare the post- and intercensal estimates for the same 2005-2009 5-year period, taking averages of the 5 values in 14 key demographic categories based on age, race and Hispanic origin. We then ran a series of comparison reports, one per state, showing how these estimates differ from one another. See these reports at http://mcdc.missouri.edu/webrepts/post_intercensal0509/. A Readme file is provided.

Posted 2-13-2016

IRS Taxes Data for 2013 Added

We downloaded the IRS individual income tax summary data for states, counties and ZIP codes based on tax year 2013. These are complete data. (The IRS announced preliminary 2014 TY data but that has not been put into the format we use and finalized (i.e. based on a complete calendar year vs a first-9-months preliminary). Access the data at http://mcdc.missouri.edu/cgi-bin/uexplore?/pub/data/irstaxes/. These data are similar to what we have for 2011 and 2012 but there were 19 new items added for this cycle. These items have been included on the standard converted sets and have also been processed to appear on our enhanced "plus" data sets. See the Readme.htnl file for details.

Posted 3-1-2016

Data Set and Report Show Post-2010 Changes in Tract and BG Level Geography

Ideally the Census Bureau creates new definitions of census blocks, block groups and census tracts once every 10 years. One of the beauties of using these units for trend analysis within a decade (across decases is a whole other issue)is that (unlike ZIP codes or places, for example, which not only can change at any time but usually do so without changing their codes) they do not (or at least should not) change their definitions. But in the real world stuff happens, and the Census Bureau does wind up having to make changes to these definitions. Most of the "ajustments" have to do with errors made in the original definitions. These changes are documented on the Bureau's web page at http://www2.census.gov/geo/pdfs/reference/Geography_Notes.pdf. Using these notes the MCDC has created a 2010 block level data set where we have both old and new geocodes. Many of the changes are trivial and do not affect any blocks with population. We ignored those changes and focused on the more substantial ones -- in Pima Conty, AZ, Los Angeles County, CA and 3 counties in New York state (Oneida, Madison and Richmond). The results of our efforts can be found in the data set alttract_bg in the mable14 data directory. The data set can, of course, be accessed via Dexter.

We also did a special report showing 3 basic data items - 2010 population and housing unit counts and land area in square miles - for the affected tracts and block groups in Los Angeles and New York state. (We did not bother doing it for the Arizona tracts since those were just simple code changes with no boundary adjustments.) The report shows the value of the 3 items for both the "old" (as defined in the 2010 census and used in those decennial data products) and the new ("alternate") areas which are now being used in the ACS data products.

Posted 3-9-2016

New Tutorial: Data Allocation Using Geographic Equivalency Files

This rather long and fairly technical document describes the steps that can be used to transform geographic data from one geographic unit to another. A simplified case study (example) walks through the steps and displays intermediate and final results, as well as the SAS(r) code that creates them. A second case study shows the "real life" example of how we created a set of 20005-2009 5-year period estimates for 2010 census tracts, using the data published for 2000 tracts as the source. View the main document at UsingGeographicEquivalencyFiles.htm . Or you can use our uexplore app to access the directory where this document and related files are stored.

Posted 3-9-2016

New ACS Data Facilitates Tract Level Trend Analysis

Users who had been waiting anxiously back in December for the 2010-2014 5-year period estimates to become available because with that release we had non-overlapping 5-year data for small geographies. Such as census tracts and ZCTAs and counties of any size. But the tract option turned out to be problematic because when the Bureau created those data products for vintage 2009 back in 2010 they used the then-current 2000 census tracts as the units for tabulating tract level data. Which means doing comparisons at the tract level would not work unless there was some way to get those 2009 5-year tact estimates aggregated to 2010 tracts. We have tried to do this using a process that is at least well documented in the tutorial reference in the previous article where the process is described in detail. We used the Census Bureau's block level relationship files (stored in our blkrel10 data directory) to create a tract-to-tract equivalency file across the decade. We then did an allocation of the old 2000-tract estimates and arrived at a new set of 2010-tract summaries. We have done this for all 50 states plus DC but users should note that there are ten states (Alabama Alaska Colorado Iowa Maryland Massachusetts Michigan Minnesota Texas and Virginia) where there were unmatched geographies that have led to some loss of data during the allocation process. For a more detailed explanation see the tutorial or at least view the conversion log file where the unmatched tracts are documented. (There is also a summary version of this log file if you want to see the bigger picture without all the detail.)

Access the new data sets in the acs2009 data diretory. Access the new metadata for the Missouri data set -- which contains a link to invoke Dexter for that data set.

Posted 3-24-2016

New County Estimates for 2015 With Components of Change (since 2010)

The Census Bureau released their county and metropolitan area level population estimates today. The MCDC has completed their standard processing of these data, creating new uscom15 and mocom15 data sets and a series of reports based upon these data. These data sets can be accessed via Uexplore/Dexter at //mcdc.missouri.edu/data/popests/Datasets.html. (The first six rows reference the six new data sets.) We have also generated a series of reports based upon these data that can be accessed at the population estimates Reports index page.

Posted 3-24-2016

Two County Changes in Alaska and South Dakota

As part of our processing of the new county estimates we discovered two county changes. They are:

Posted 4-25-2016

New County Business Patterns Data

The Census Bureau released new County Business Patterns data for the US, states, counties and metro areas on 4-21-16. We have downloaded the data and created the usual collection of data sets in the cbp data directory. The new data sets are ustots14, usstates14, uscos14 and uscbsas14. These data can also be accessed on the Bureau's web site on their CBP page.

The companion ZIP Business Patterns data is scheduled for release in May.
(The ZIP level data was downloaded and processed 5-21-16 per note below).

Posted 5-17-2016

Health Insurance Coverage Data (SAHIE)

We have downloaded and processed two new sets of SAHIE data for 2013 (updated) and 2014. These datasets are stored in the saipe data directory. There are 6 new data sets: moshaie13, mosahie14, sahie14 and sahietotals14. Naming conventions and data content are essentially the same as we have stored in this time series that goes back to 2006.

Posted 5-19-2016

Sub-county Population Estimates

We downloaded and processed the latest sub-county population estimates for 2010-2015 from the Census Bureau. We did our standard processing, creating US and Missouri data sets in the popests data directory and generating a new set of reports based on these data. See our population estimates Reports page with links to the new reports and to the new data sets (Dexter access) on which they are based.
The new Excel file with these data has not yet been created.

Posted 5-21-2016

2014 ZIP Business Patterns

As promised back in April when we processed County Business Patterns data for counties and other levels we have now received and processed the corresponding data at the ZIP Code level. We store these data in the saem filetype data directory as the county level data. The easiest way to access these data is via the Datasets.html page in the cbp data directory . (Access here is, or course, via Uexplore/Dexter).

Go to the top of this page.


See the whatsnew page for last year.

This file last modified Saturday May 21, 2016, 09:38:36


Site Map    |    Our URL    |   
The Missouri Census Data Center is a sponsored program of the Missouri State Library within the office of the Missouri Secretary of State. The MCDC has been a partner in the U.S. Census Bureau's State Data Center program since 1979.

Questions/Comments regarding this page or this web site are strongly encouraged and can be sent to