Census Tracts

Census tracts are small, relatively stable statistical areas that generally contain between 1,500 and 8,000 people, with an optimal population of 4,000. They belong to a hierarchical system for organizing the territory of the United States for census-taking purposes. The country is divided first into regions and subdivisions, then into states and counties, and finally into census tracts, block groups, and blocks. Census tracts are delineated by local census statistical area committees following U.S. Bureau of the Census guidelines. Their spatial sizes vary widely depending on the density of settlement. Census tract boundaries are defined with the idea that they will be maintained over many decades so that comparisons can be made from one census to the next, but physical changes in street patterns and new development may result in boundary revisions. In areas of rapid growth, census tracts often are split; in areas of substantial population decline, they are sometimes combined.

In 1906, the idea of collecting census information for small areas was first put forth by Walter Laidlaw, who studied neighborhoods in New York City. In response to his request, the Bureau of the Census tabulated information from the 1910 census by census tracts for eight cities with populations larger than 500,000: New York, Baltimore, Boston, Cleveland, Chicago, Philadelphia, Pittsburgh, and St. Louis. Data for the same eight cities were again tabulated in the 1920 census. In 1930, the number of cities was increased to 18. Data were not published but were made available by the Bureau of the Census for purchase. Beginning with the 1940 census, the Bureau of the Census established the census tract as an official geographic unit and published the tabulations for large cities. By 1990, census tracts were delineated for most metropolitan areas and other densely populated counties, and six states (California, Connecticut, Delaware, Hawaii, New Jersey, and Rhode Island) paid a fee to have complete tract coverage. The 2000[Page 32] census was the first in which the entire United States was covered by census tracts.

Census tract boundaries are available from the Bureau of the Census's Topologically Integrated Encoding and Referencing (TIGER) files. Files from the TIGER database can be downloaded directly into geographic information systems (GIS) software and used to map a variety of census tract features, including demographic (e.g., age, race, ethnicity, gender), social (e.g., education, place of birth, ancestry), economic (e.g., income, occupation, employment, poverty status), and housing (e.g., size, age, type, value, presence of plumbing and heating facilities) characteristics. Census tract maps commonly are used to represent intraurban variation because tracts generally are small enough to be somewhat similar in income, housing, and other characteristics but large enough to avoid a visually unmanageable number of spatial units. In addition, the use of tracts usually avoids missing data problems stemming from census disclosure rules. Census participants are promised that the information collected about them as individuals and households will remain confidential.