Mapping Decennial Census or American Community Survey Data with TIGER GIS Files

The 2010 Census

The US Constitution requires the Federal government to conduct a decennial census of the US population for purposes of reapportioning seats in the House of Representatives.  At the start of each decade, the Census Bureau surveys every household, nursing home, prison, homeless shelter, etc., by mail, with followup visits by enumerators. 

In 2010, the Census Bureau's questionnaire was a 10-question form asking how many people live in the residence, whether the residence is owned by an occupant with or without a mortgage or rented, and asking each household member's the sex, age, ethnicity (Hispanic, Latino or Spanish origin) and race.  The form included seven questions per additional resident.

The collected data are summarized and cross-tabulated via various hierarchies of geography by "summary levels" (SUMLEV).  Commonly-used summary levels that cover all areas of the US include state (040), county (050), Census tract (140), block group (150) and block (101). 

The diagram below shows this principal hierarchy: blocks > block groups > tracts > counties > states, etc. There are alternative hierarchies with different summary level codes such as blocks > Congressional Districts > states, etc.  Census blocks are the fundamental geographic units on which all hierarchies are built:

The "Summary File 1" tabulations of Census data are distributed from the Census Bureau's website at 2010.census.gov/2010census/data or via the American FactFinder web interface.  Here's a quick overview of each:

  • You can download raw block- and higher-level SF1 2010 Census data tabulations by state from 2010.census.gov/2010census/data and format these yourself based on the Data Dictionary section of the technical documentation.  This is more challenging than simply downloading pre-formatted tables from the American FactFinder (see below), but the process will teach you a lot about how Census data are structured.  The data fields are broken up into ASCII (plain text) file segments containing sets of data fields, and generally require formatting before they can be used in a GIS.  Here's a typical set of unzipped SF1 file segments, including a geo file for linking the data files to geographic polygons created from TIGER data (see below).

    Each segment has the same number of records, with each record given a unique Logical Record Number (LOGRECNO) so that you can join fields and match records from different segments. 

    The technical documentation includes detailed explanations of the geographic hierarchies, summary level codes, file structures and data elements.  Note that the SF1 data cover all levels of geography within the state in the same files.  Each record includes its summary level (SUMLEV).  Open any file segment in Excel using the text import wizard; the technical documentation will tell you which fields it contains.  If you add a header row containing the appropriate variable names, you can then add the worksheet ply to the GIS for joining to geography polygons.

    The join process actually requires two steps, however.  Each data record is identified by its ID, the LOGRECNO field.  Each geography polygon is identified by its FIPS code, usually the GEOID field.  You will need to format the geo file as well, which matches GEOID to LOGRECNO.  Join the geo file into the polygon attribute table by GEOID, then you can join the data files into the polygon attribute table by LOGRECNO.

  • Alternately, you can download structured Census data tables for any specific geography level from the Bureau's American FactFinder web server.  This web interface lets you search and download all sorts of data collected by the Census Bureau since 2000--the Decennial Census, the American Community Survey (see below), the Economic Census, etc.--at various levels of geography.  You can search data tables by geography, keyword, year, topic, etc., and you can view the table structures before you download them in Excel or another convenient format. 

    If you "Select Geographies" by "Name," you can extract block- or block group-level SF1 Census data.

The American Community Survey

In the 2000 census, every household answered the "short form" questions about occupants' genders, ages, races, etc.  But a large proportion of households received a "long-form" questionnaire containing the short-form questions plus additional questions regarding income, schooling, employment, marital status, etc.  Summary data compiled from the short-form ("100% count") questions were called Standard Format 1, or "SF1" data.  Summary data compiled from the 2000 Census long-form ("Sample") questions were called "SF3" data.

The 2010 Census did not include a long-form questionnaire, so there will be no 2010 SF3 data release.  More or less equivalent information is now collected as part of the ongoing American Community Survey, which releases annual 1-year, 3-year and 5-year estimates for various geographic areas   These data, including mean and standard error estimates, are also distributed via the American FactFinder

Each year's 1-year data release summarizes survey data from the prior year, and only covers geographic areas with populations of 65,000 or more, including about 800 counties and 500 metro statistical areas. 

Each year's 3-year data release summarizes survey data from three prior years, and covers all geographic areas with populations of 20,000 or more, including about 1,800 counties and 900 metro statistical areas. 

Each year's 5-year data release summarizes survey data from five prior years, and covers all geographic areas down to the Census tract level. 

TIGER Geodata

The Census Bureau also publishes GIS data, known as TIGER (Topologically Integrated Geographic Encoding and Referencing) files, that are used to create polygon layers of states, counties, tracts, block groups, and blocks, as well as water areas.  TIGER data also include data to create line layers representing roads, railways and streams, or point layers representing various landmarks.  Each unit of Census geography is identified by a unique FIPS (Federal Information Processing Standard) code:

  • Each state (SUMLEV = 040) has a 2-digit FIPS ID; Delaware's is 10.
  • Each county (SUMLEV = 050) within a state has a 3-digit FIPS ID, appended to the 2-digit state ID. New Castle County, Delaware, has FIPS ID 10003.
  • Each Census Tract (SUMLEV = 140) within a county has a 6-digit ID, appended to the county code. The Tract in New Castle County DE that contains most of the the UD campus has FIPS ID 10003014502.
  • Each Block Group (SUMLEV = 150) within a Tract has a single digit ID appended to the Tract ID. The center of campus in the northwest corner of the tract is Block Group100030145022.
  • Each Block (SUMLEV = 750) within a Block Group is identified by three more digits appended to the Block Group ID. Pearson Hall is located in Block 100030145022009.
The Census Bureau releases summarized SF1 data down to the Block level.  State-, county-, tract-, block-group- and block-level Census data tables extracted from these data files can be joined to these polygons via their matching FIPS codes.

Lab Exercises

Thematic Mapping of US Counties
TIGER and SF1 Data for Delaware


"The Dude abides."