What Was New at the MCDC in 2009

Capturing the highlights that appeared on the MCDC home page during 2009

February || March || April || May || June || July || September || October || December

December 2009....

Census Bureau Releases State Level Estimates for 2009

The Census Bureau released new population estimates at the nation, region and state levels today (12/23). These files contain data for each year since the 2000 census and go up to July 1, 2009. These are the last post-censal estimates at the state level that we'll see for this decade. The data show that Missouri added over 31,000 people in the most recent year (July 1, 2008 to July 1, 2009), bringing it to a new estimate of 5,987,580. Missouri ranked 24th in terms of persons added to its population in the last year, but only 37th when that growth is measured as a percentage of the population. Looking at the trends over the decade, Missouri ranks 18th in terms of total increase and 26th in terms of percentage increase. The state has grown by almost 391,000 persons or 7% over the 9-year period.

Texas and California continue to lead the way in terms of absolute population growth, both in the most recent year and in the decade, with Texas ranking first and California second. Nevada leads in percentage growth for the decade, but slipped to 17th in percentage growth for the latest year (with Wyoming moving into the top ranking there with a 2.1% increase this year).

This is an excerpt from a report generated by the Missouri Census Data Center based upon the files made available by the Bureau. Click here to view the entire report, and/or here to access a csv file with the report data and more.

Uexplore/Dexter users can access the new data in dataset uscomnst09 in the /pub/data/popests data directory.

Introducing Rankster (Preview)

The pdf and csv files referenced above were generated using a new web utility application called rankster. Users who are familiar with our uexplore/dexter software will recognize rankster as a companion tool, with a comparable user interface and similar output formats and conventions. In fact, the easiest way to access rankster is via uexplore. You will navigate to a data set in our archive and when the Dexter custom form page displays there will be a new line near the top of the form that lets you specify that you want to run rankster to process this dataset. We use the future tense here because the code we are talking about is still in test mode. In order to use it at this point you will need to know how to edit a URL that appears in your browser window. When you have a Dexter query form displayed in your window (usually you get there by using uexplore) you are looking at the results of running an application that goes by the geek name "websas.uex2dex.sas". In order to run the test version and get access to rankster you need to change this name to "test.uex2dex.sas". How do you do that? By looking at the URL in your browser's address window (the one that has displayed /cgi-bin/broker?_PROGRAM=test.uex2dex.sas&...) You use your mouse to carefully highlight the word "websas" in this URL and then you type the word "test" (all lowercase, without the quotes). Then just hit enter to run the test version of the application. If you do it correctly the top of your new Dexter extracter page should look something like this: (If you are having trouble seeing the page image below in your browser because it is too small, try "zooming in" on the page by typing "ctrl +" - holding down the Ctrl key while simultaneously typing "+" - to enlarge the image.)

As you can see (*), you now have an option to "Skip dexter and go straight to rankster", which you can just click on. But there will be times when it will be helpful to first run dexter to create a new dataset with just the rows and columns that you are interested in using for your rankings. In these cases you will be able to choose an option to first run Dexter but then pass the output dataset to rankster. If you ignore the new Rankster option line then rankster is not invoked. We hope to have this application available in production within the next week or two. We are hoping to have some useful training materials ready when we announce that step.

October 2009....

Census Bureau Releases 3-year Period Estimates for 2006-2008

As promised the Census Bureau has delivered the latest set of summary data products based on the American Community Survye for the calendar years 2006 through 2008. See the
relevant press release at the Bureau's web site. The MCDC has updated our ACS Profiles web application so that it will now allow you to display profiles based on these data. (See the link to this app as the first item in our Quick Links navigation box - just to the right side of this page.) The data can also be accessed via the American FactFinder web site.

IRS Migration Data for 2007/2008 Available

The IRS has released a new year of state and county level migration flows based on tracking tax returns between 2007 and 2008. As usual, the Missouri Census Data Center has obtained these data and made them available in our public data archive, as well as via our interactive web application that displays a migration profile for any selected county in the U.S.

We used to have to purchase these data on cd-rom/DVD but starting this year the IRS has decided to make the raw files available via a free download site.

You can access our irsmig data collection via uexplore, which lets you access the Migration_Profiles menu page .

New for this year (on our site, at least) is the availability of state-to-state migration profiles. (Prior to this everything we had was county level data.) These reports display the same kind of data in essentially the same format as the county profiles but the data are all at the state level. So you can access a report for Missouri to see how many tax-return flows occurred between Missouri and any other state in the country. We have these data for the most recent four tax-year combinations: 2004-2005, 2005-2006, 2006-2007 and 2007-2008. To access these you need to go to the Migration_Profiles menu page and look for the menu item US State-to-State.html.

2008 ACS Single-Year Data Now Complete

Last week (late September) we noted here that the Census Bureau had released a new set of American Community Survey data products based on a single year's worth of surveys, 2008. (See our What's New page for details). We noted at that time that the data were not complete due to problems discovered by the Bureau in some of the tables relating to income and poverty. Those problems have now been fixed and a complete set of tables have been released. The Missouri Census Data Center has processed the refreshed tables and have used the new data to complete our ACS Profile Reports (see first link under Quick Links - just to the right). Better yet, click here to see a sample report with just the brand new economic sub-profile for the cities of Columbia and Springfield, Missouri and the U.S. Scroll to the bottom and note the last table in the profile: this is the eagerly anticipated data regarding insurance coverage that was just added to the ACS for 2008. Click on the graph icon to see these data as pie charts. You can also follow the links in the (rightmost) "Ref Table" column to display related complete base tables.

The 3-year period estimates data based on survey years 2006 through 2008 are scheduled to be released on October 27. That will give us data for geographic areas as small as 20,000. However, because the questions on insurance coverage have only been asked for 2008, there will not be any data on this topic in the these period estimates.

We have also downloaded the complete set of ACS detailed ("base") tables and have made them available via links within the ACS Profiles as well as via Uexplore/Dexter in the MCDC data archive. You can also access these tables at the Census Bureau's American FactFinder site.

September 2009....

2008 ACS Data Released

The Census Bureau has released a new set of American Community Survey data products. These are single-year data, based entirely on the 2008 surveys (available for geographic areas of 65,000+ population). (See the
Census Bureau Press Release regarding these data.)
The 3-year period estimates based on surveys taken over the calendar years 2006 through 2008 (available for geographic areas of 20,000+ population) are scheduled to be released next month (on October 27).

Unfortunately, the Bureau discovered problems with some of the data for this release, which has caused them to restrict access to those data. The problems affect tables dealing with income and poverty. The Bureau estimates that the corrected tables will be available in about a week (on September 29, next Tuesday).

In addition to the new data, MCDC has new software for displaying profile reports based upon the ACS data. This is reflected in a change to first two links (to "ACS Profile Reports" and "ACS Trends Reports") within our Quick Links navigation boxes (the light blue boxes like the one just above and to the right of what you are reading right now). There is little or no change in the content of the reports (just some changes in style), but there are significant differences in the interface that lets you select geography and time periods for your reports. The graphics associated with the tables have also been signficantly upgraded.

For those Uexplore/Dexter users wanting to access the raw data sets associated with this release, there will be a slight delay in making these available. We are working on making the usmcdcprofiles data set for 2008 (albeit with the economic data items missing) available. We plan to wait for the final set of detailed (base) tables to be released before putting our versions of those up in the data archive.

Population Estimates by Age

The National Center for Health Statistics has released "vintage 2008" population estimates for all counties in the U.S. These are the special "bridged" race estimates that attempt to replicate the four major race categories that were used by the Census Bureau and other federal agencies until the OMB-mandated shift to multi-racial categories in the late 90's. See the NCHS Web Site for more details and links to the downloadable data files.

As we have been doing for the last several years now, the Missouri Census Data Center had downloaded these estimates and made them available in our public data archive. (For Uexplore/Dexter users the datra are in the nchsbri subdirectory of the /pub/data/popests filetype directory). More importantly (for most users) the data have been summarized and are accessible via the MCDC's Population Estimates by Age web application. This utility app can dynamically generate estimates by age (including user-specified custom intervals), race, sex and Hispanic origin for all U.S. states and counties, for years 1990 to 2008 (the addition of 2008 data being the news item here).

These estimates are produced by the Census Bureau under a contract with the NCHS and are consistent with the Bureau's "casrh" estimates by age, sex, race and hispanic origin which were published (with the vintage 2008 updates) back in May. We have been monitoring the NCHS web site all summer waiting for this alternate version to become available. The major differences between these estimates and those published by the Bureau are the use of bridged race categories and the single year of age detail available in the NCHS files (vs. 5-year cohorts used in the Bureau's files).

The MCDC's Population Trend Report web application also makes use of these detailed estimates data, so you can now run that application and have new choices for the time periods going through 2008.

July 2009....

What's Coming

It's getting close again. We are less than a year away from the next Decennial Census day - April 1, 2010. Which means we are less than two years away from seeing the first tabulated results of that once-in-a-decade survey. Less than two years away from knowing with near certainty just how many people live in our counties, our cities and our neighborhoods, regardless of size. Not only will we know the head count but we'll also know important demographic details of the population: age, race, gender, hispanic origin, living in a household (and whether it's a family household) or in group quarters (what kind), renters or owners, etc. Or at least we'll know these things as they were circa that point in time - April 1, 2010. So we can easily spot trends by linking to comparable data gathered on April 1, 2000 and even April 1, 1990.

What else are we looking forward to getting from the new census? Won't we also be getting new data regarding things such as people's incomes, poverty status, education, occupation, employment status, house values, and disabilities? Well, no. Not from the decennial census, not this decade. ...

Read more in our essay The American Community Survey vs. the Decennial Census Long Form - Are We Better Off Now Than We Were a Decade Ago?.

It's Official: Columbia Population Goes Over 100,000

The Census Bureau has released new population estimates for 2008 for sub-county geographic areas. Among the interesting facts that can be gleaned from the new data:
  • The city of Columbia, MO grew by almost 1,800 people (1.8%) in the year ending July 1, 2008 resulting in a new official estimate of 100,733. Columbia has been experiencing consistent growth over the decade, averaging just under 2,000 new people per year since the 2000 census. It joins Kansas City, St. Louis, Springfield and Independence as one of five Missouri cities with over 100,000 population. Lee's Summit, with a new current estimate of 84,208 and with growth of 13,490 since 2000 is the next closest to crossing this threshold.

  • Other cities in the state showing significant growth in the latest year were O'Fallon (+1952, 2.6%) and Wentzville (+1358, 6.1%) in St. Charles county, Springfield and Lee's Summit, both with gains of about 1500. Pevely showed a surprising and dramatic increase of 1233 persons. That took Pevely's population to 5808 and was a 27% increase making it the top growth city in the state based on percentage increase.

  • Only seven cities in the state are estimated to have lost at least 100 people in the latest year. One of these is the City of St. Louis, which probably doesn't count since they traditionally are under-estimated by the Bureau and have to challenge the estimate each year. So their iniital estimate showing a loss of 1300 people will probably be turned into an increase of several thousand if they successfully challenge again (see our web page documenting how this has worked for the past five years). The other six population losers were all in St. Louis County: Florissant was the biggest loser with -274, followed by University City (-196), Chesterfield (-132), Webster Groves (-122), Ferguson (-119) and Maryland Heights (-117).

  • Nationwide, New York City had the largest numeric increase during the year with 53,498 new residents (.6%), while four Texas cities (Houston, San Antonio, Fort Worth, and Austin) all posted gains of over 18,000 to rank in the top 10. New Orleans showed the largest percentage increase of cities over 200,000 at 8.2% as victims of Katrina continued to return (the increase was almost 29,000 last year and just under 24,000 this year so the pace slowed a little); but the 2008 estimate of 311,853 is still almost 36% less than what it was in 2000 (484,674).

  • Among larger cities (200,000 or more) Cleveland and Detroit had the largest declines in population. Detroit has the greatest numeric decrease, losing another 4,874 persons, while Cleveland was second with a decline of 4,265. But since Detroit is roughly twice as large as Cleveland, the Ohio city would seem to merit the distinction of being the largest loser of the year.

View these numbers and more in a series of Top 50/Bottom 50 reports for Missouri only  | United States

To access the data along with a press release visit the Census Bureau web page . If you are not familiar with the Bureau's program for allowing cities to challenge their estimates you should be sure to visit the Estimates Challenges web page (which is linked to from the above-referenced cities page). Missouri data followers will be particularly interested in the successful challenges of the 2007 estimates for St. Louis City, Kansas City and St. Joseph. This is a first for Kansas City and St. Joesph, but is old hat for St. Louis. The Kansas City adjustment was over 25,000 which is more than five times as much as the adjustments for the other two Missouri cities. The challenge was not accepted until January 29 of this year and as a result it is not reflected in these latest estimates (which include revised values of the estimates for each year since the 2000 census). The higher numbers for St. Louis and St. Joseph are refelcted in the July 1 , 2007 estimate figures for those 2 cities. The MCDC has downloaded the data table shown on the Challenges page and created a data set that can be accessed via Uexplore/Dexter in our public archive. The dataset is named challenges2007.sas7bdat.

We have updated the Curmoests.xls Excel workbook (within the popests data directory), to reflect these latest estimates.

June 2009....

MCDC Workshops

MCDC staff (Katina Jones, Billy Earney, Jim Harlan, and John Blodgett) presented a series of workshops at Ellis Library and Stewart Hall on the University of Missouri Columbia campus on May 28th, 2009.

Population Trends

One of the first data products/tools demonstrated at the recent MCDC workshops was the poptrends dynamic web application. Using this tool (which can be viewed by going to the
Data Products/Inventory Showcase page and and then clicking on the Select menu (application) link in the first row of the matrix) we created a report that shows trends in selected demographic indicators for selected time periods (going back to 1990 and forward to 2007). We were able to provode a custom title for the report and even to specify that we wanted a blue color scheme (overriding the default green version.) Here is a partial snapshot of that report:

Click here to see the full report (in a scrollable sub-window). The report shows clearly how the state's population has been growing over the last 17 years. Looking at the most recent years of data (2006 to 2007) we see that the state grew by almost 41,000 in that year, with only about 22,000 of that accounted for by white non-hispanic persons. Hispanics accounted for almost 9,000 of the increase and African Americans for about 6,500.

These data will be refreshed later this month when we expect to get the special estimates data from the National Center for Health Statistics that we use as the source of our demographic trends data. We hope to have 2008 data added to the application before the start of summer.

May 2009....

2008 County Estimates by Age, Race, Sex, Hispanic

The Census Bureau has released county level numbers with detail by age, race, sex and hispanic origin.

See the Census Bureau estimates page for links to data sets and a presss release.

(Veteran users of these estimates will be pleased to note that we are getting these number almost 3 months sooner than in recent years.)
The MCDC has downloaded these data into our public archive where they can be accessed via Uexplore/Dexter in the popests data directory (datasets mocasrh08, ilcasrh08, etc - one dataset per state). We have also generated a series of summary reports at the county and metropolitan/micropolitan area levels for all states. To access the Missouri reports use this link, or go to the Population Estimates Reports by State page to access reports for any state in the nation.

MCDC To Present Workshops on May 28

The MCDC will present a series of three workshops on May 28 at the campus of the University of Missouri in Columbia. The workshops are being held in conjunction with the Missouri Government Documents Conference being held at the same site on the previous day (May 27). Space will be limited so you may want to register early. The three workshops (two of which will be presented concurrently) will cover the American Community Survey, an Introduction to GIS, and a guided tour of the MCDC web site.

SF3 Profiles at Census Tract and Block Group Levels

You can now access basic demographic profiles based on the 2000 census at the census tract and block group levels for the entire United States using the MCDC's SF3 Profiles (the 3rd item in our Quick Links menu box). Choose United States off the main menu and note that the two new entries on that menu page - Census Tracts and Block Groups.

April 2009....

Missouri Taxable Sales Data for 2008

The 2008 taxable sales data for Missouri counties with state totals have been added to the MCDC data archive in the
taxales filetype directory. This new set of data is in the identical structure with comparable content as the eight years of previous data already stored in the archive. What is new this year is a data set called taxsales_ranks where we have combined selected or calculated data from all nine years and then added some key ranking variables. We are hoping to develop a dynamic web application that will use the contents of this data set to let users see trends in the sales activity.

New Economic Data from the Bureau of Economic Analysis

The Bureau of Economic Analysis has released its annual update to the REIS (Region Economic Information System) economic data resource. As we have for the past several years now, the MCDC has downloaded the data and converted the files to a series of SAS data sets covering the country at the state and county levels. Uexplore/Dexter users can access the data sets in the beareis filetype directory. You can also access the data using our BEA Economic Profile web application (also available via applinks and the Missouri County Data Map). Here is an example of what such a profile looks like:

The data can also be accessed at the BEA's web site.

MCDC to Present Workshops in Late May

The MCDC will present a series of three workshops on May 28 at the campus of the University of Missouri in Columbia. The workshops are being held in conjunction with the Missouri Government Documents Conference being held at the same site on the previous day (May 27). Space will be limited so you may want to register early. The three workshops (two of which will be presented concurrently) will cover the American Community Survey, an Introduction to GIS, and a guided tour of the MCDC web site.

March 2009....

MABLE/Geocorr Update

Using the recently released 2008 county level population estimates along with the accompanying county-to-metro area equivalency files from the Census Bureau, the MCDC has updated the MABLE data base and geocorr2k processing module. Users can now access the latest CBSA and CBSA-related geographic codes (CSA's and Metrodivs) and can use 2008 population estimates as weight variables in order to look at how these geographies are related to many others. The previous data being used was vintage 2006.

ACS Training Session March 26, 2009

MCDC staff (Katina Jones, Billy Earney and John Blodgett) presented an ACS Training session at Thomas Jefferson Library on the UMSL campus in St. Louis on March 26th, 2009.

2008 County Level Population Estimates

The Census Bureau has released county level estimates of the population of states, counties and metro areas for the years 2000 to 2008. These estimates show not only the population estimates but also the components of population change -- births, deaths and net migration -- for each of the years. The Missouri Census Data Center has added these data to our public archive and has created several standard data products similar to what we have been doing with these updated estimates for each of the last several years. Specifically:

  • The curmoests.xls Excel spreadsheet file has been updated. There are 4 worksheets within this file and only 1 has been updated - Counties 2008 . These data are for Missouri only and show the revised estimates for the state and each county for each year since the 2000 census. The other 3 worksheets contain data at sub-county levels and these will not be updated until the 2007 sub-county estimates are released (usually in late June or early July).

  • The county data have been aggregated to create summaries at the metropolitan and micropolitan statistical area level for the entire United States. We have also aggregated the data to a number of county-based regions for the state of Missouri only. To view reports based upon these data go to our Population Estimates Reports page and note the revised entries in rows 3 thru 5. Notice that the left column contains a description/hyperlink to the report, while the right column contains the name of the archive data set used to generate the report and doubles as a hyperlink that will take you to the Dexter data extraction utility with that dataset selected for extracting.

Progress with Restoration of sf32000 Files

Last month we reported that due to a combination of a disk crash and a glitch in our backup system we lost our sf32000 data collection, containing the complete tables from Summary File 3, 2000 decennial census. This is far too important a collection to not recover these data. Here is where we are in our process of putting things back in place.
  1. We downloaded the state level csv data files for Missouri, Illinois and Kansas, as well as the national (US) data files.
  2. Using SAS code that we wrote back in 2002, which we were able to get from our colleagues in the State Data Center program (Larry Meyers at PSU and Roy Williams at U of Massachussetts), we were able to run conversions on these raw data files to put back our basic collection of datasets.
  3. We reran a custom aggregation setup to recreate the important uspumas data set with PUMA-level summaries for the entire U.S.
  4. We were able to use the Varlabs metadata files (also returned to us as part of our conversion code kit from same sources as noted above) we were able to rewrite a program to create a series of metadata data sets. This was critical because without these data sets our dexter and sf3tabgen dynamic web applications would not work. They are now back and working (as of 3-18-09).

We still have work to do, with a number of custom aggregations to rerun and some metadata files to recreate, but we are well on our way to a near full recovery. Fortunately, the large majority of our end-user usage of these data involve using the standard extracts derived from these complete table data sets. And since those extract datasets were stored in a different filespace they were not affected by the problem. Hence our SF3 Profiles web app has continued to work, although the "drill down to source tables" feature (which uses the sf3taben application) was not working until we were able to recreate the metadata files (yesterday).

February 2009....

SF32000 data lost

On Feb 24, 2009, we had a hard drive failure. We were able to recover all data except for the sf32000 data directory. We are in the process of trying to rebuild these data sets as well as the related programs and metadata. It may take a while.

ACS Profile Reports Offer "Drill Down" to Base Tables

The MCDC's ACS Profile reports (see first entry in the Quick Links menu box at right side) now provide users with links to the more detailed base tables from which the profile data are extracted. For example, the first table, AGE, contains the link
B01001. Clicking on this link takes you to a display of the more detailed Age by Sex base table which has slightly more detailed age cohorts and is broken down by gender. It displays this table for each geographic area you have selected. The base table display module is considered an "alpha" (test) version and we are continuing to work on refinements. These involve displaying multiple geographies across the page instead of down (consistent with how the profile data are displayed) and using fonts/colors to indicate levels of statistical uncertainty based on margin of error values.

2007 Census of Agriculture

The U.S. Department of Agriculture has released the results of the 2007 Census of Agriculture. They have created an extensive web site with easy access to voluminous state and county level profiles. See, for example, the Missouri County Level Data index page which lets you access the 56 county-level profile reports for Missouri. (See the state level index if you want to access comparable data for any state.) Or visit the Ag Census 2007 home page at http://www.agcensus.usda.gov/index.asp.

The MCDC has downloaded the data used in these profiles and is in the process of converting it to a form consistent with our data archive. Uexplore/Dexter users can access the new ag2007 filetype (data directory) in order to access these data in a database format. We are still working on the conversion process but do have a preliminary data set with the data included in Table 1: County Summary Highlights for all states and counties in the country.

New County Economic Profiles

You can now view the MCDC's economic profiles based upon the Bureau of Economic Analysis REIS profiles. The reports contain basic economic indicators for the years 1969 thru 2006 (with 2007 data scheduled to be available in the spring). See a sample report or access the front-end web page which allows you to specify the geographic areas, years, data detail, etc. for your customized report. You can also access the reports for Missouri counties using the Missouri County Data Map front end (see link in our Quick Links box), or you can follow the link at the bottom of that page to select any state in the U.S. and be taken to the applinks menu page for choosing a county. The Regional Economic Profile application is now the very first item shown on the applinks page.

Other "What's New" by Year: 2008 | 2007 | 2006 | 2005 | 2004 | 2003 | 2002