Digital Spatial Data

US Census

Overview

(Picture above: racial composition changes of Milwaukee City between 1980 and 2000, showing major racial group at block level)

The decennial Census is an excellent source for population, housing, economic and geographic information. It also gives us opportunities to see changes or identify trends over time because census has been taken every 10 years since the first Census conducted in 1790. Information collected by each decennial census varies, unfortunately, which makes historical analysis sometimes difficult. However, much information has been collected consistently.

In recent years, the Census Bureau has used two different types of questionnaire forms, called "short form - 100% count" and "long form - sample count" questionnaires. In the long form, additional questions were asked of a sample of persons and housing units (generally 1 in 6 households) to collect information on detailed socio- and economic states of the US population and housing.

The results from decennial censuses are available in a number of products. For example, five different types of data reports, or files, were resulted from the Census 2000. They are:

• Redistricting data (Public Law 94-171) - from 100% with limited variables; race and age.
• Summary File 1 (SF1) - from 100% data.
• Summary File 2 (SF2) - from 100% data, variables are the same as SF1 but repeated for different population groups.
• Summary File 3 (SF3) - from sample data.
• Summary File 4 (SF4) - from sample data, variables are the same as SF3 but repeated for different population groups.

In this section information available at the AGSL and the UWM Libraries and internet resources are discussed. 

Census Geography 

One factor that makes the census so special, other than its value as a historical dataset, is the way data is organized in several different geographic levels, or entities. In other words, census data is provided in the same format for many different geographic entities. Thus one can obtain the same information at different geographic levels, such as median household income data of a county or a census tract. Its geographic entities are generally structured in a hierarchical manner. The smallest geographic entity is a block and blocks combine to the level of block groups (or sometimes places), and block groups combine to the level of census tracts, census tracts combine to the level of counties, and so forth (see a figure below - figure is from the 1990 Census Bureau's Geographic Areas Reference Manual). 

Return to top of page

Comments for the AGSL? Send them to agsl@uwm.edu.