Last revised: July, 2007
Census 2000: Finding Data Online
Scope: This guide is intended to assist you in finding Census 2000 data online. If you need additional census help, contact us at govdocs@umd.edu or 301.405.9165.
Table of Contents
About Census 2000
Beginning in 1790, the U.S. government has performed a survey of its population every 10 years called the Decennial Census. Census 2000 is the 22nd Decennial Census. The principal topics covered by the censuses are population and housing but you will find data on peripheral subjects. Unlike older censuses, the Census 2000 has been made available in detail online through the Census Bureau's website.
Census 2000, like many of its predecessors, is divided in two distinct data groups. The "short form" is intended to count 100% of the people residing in the United States. Sample characteristics are also taken for those people who responded to the "long form." Approximately one in six residents receive the long form questionnaire. The U.S. Census Bureau compiles data from these questionnaires and publishes it in a number of different forms.
The tables that you see on the Census Bureau's web site are the result of a marriage of two entities--a data concept and a geographic area. In order to make the best use of the wealth of data available, it is important to match the census' data and geographic concepts with those of your particular research project. In addition to assisting you in using the Census Bureau's on line tools, this guide will help you locate and understand the data point and geographic area or concept that best fits your research needs.
Data Files: Finding the Data Point That Meets Your Needs
The census consists of two broad data sets--the 100% or short form set and the sample or long form set.
You will find the following for "100%" of the U.S. population:
- Household relationship
- Sex
- Age
- Hispanic or Latino origin
- Race
- Tenure (rent or own)
- Vacancy characteristics
You will find the following data for a sample of the U.S. Population:
Population
- Marital status
- Place of birth, citizenship, and year of entry
- School enrollment and education attainment
- Ancestry
- Migration (residence in 1995)
- Language spoken at home and ability to speak English
- Veteran status
- Disability
- Grandparents as caregivers
- Labor force status
- Place of work and journey to work
- Occupation, industry, and class of worker
- Work status in 1999
- Income in 1999
Housing
- Value of home or monthly rent paid
- Units in structure
- Year structure built
- Number of rooms and number of bedrooms
- Year moved into residence
- Plumbing and kitchen facilities
- Telephone service
- Vehicles available
- Heating fuel
- Farm residence
- Utilities, mortgage, taxes, insurance, and fuel costs
The subjects listed above become parts of different census data sets. In the on line environment, you can think of the data set as a particular publication where certain tables or information are represented for particular geographic areas. For example, the "Census 2000 Redistricting Data Summary File" consists of state population counts for race and Hispanic or Latino origin categories. This file was released March 2-31, 2001, and the lowest level of geography available is the block (more on blocks below).
The major data sets for Census 2000 are as follows:
- Census 2000 Redistricting Summary File
Basic data required for local redistricting for federal elections with detailed tabulations at the block level - Summary File 1
Basic cross tabulations for population and housing units with most data available only at the census tract level - Summary File 2
More detailed data on population and housing units with iterations for race and Latino origin and data available at the census tract level - Summary File 3
First release of sample data with some data available at the block level and other data available at the census tract level as well as data by Zip Code Tabulation Area and congressional district - Summary File 4
Sample data on population with iterations for race and Latino groups
If you are unsure of what data set may contain the data point of interest to your research, consult a complete listing at Census 2000 Data Products Listed by Subject. In this table, you will find each individual data point, the data set of which it is a part, and the release data or planned release date.
Geographic Areas and Terminology
The Census Bureau uses a number of different geographic concepts. Imagine a Venn diagram. The Census geographic areas are like circles on this diagram. Some areas exist side by side and do not overlap; others overlap partially, and some circles are wholly within others.
There are many different geographic areas. The main geographic areas are the nation, regions, states, counties, census tracts, block groups, and blocks (in order from largest to smallest). These broad categories are further broken down into other geographic areas. For example, states include school districts, congressional districts, and places, to name a few. Please note that census data are not tabulated to a particular geographic level if doing so would make is statistically possible to identify an individual. Individual data remains confidential for a period of 72 years. The following is a more detailed list of some of the more important geographic areas:
- Blocks are the smallest geographic area for which the Census Bureau tabulates data. They are bounded by streets, legal boundaries, and other features. Census 2000 includes approximately 8.5 million blocks.
- Block Groups are collections of blocks within a single census tract that share the same first digit of their four-digit identifying numbers.
- Census Tracts average about 4,000 persons and exist within counties or equivalent units. They have fairly stable boundaries from one census to the next. When they were first designed, they were intended to have relatively homogenous demographic characteristics.
- Counties and Equivalent Areas are the primary subdivisions of most states, Puerto Rico, and the Island areas. The equivalent areas are boroughs, parishes, independent cities, the District of Columbia, and other entities in Island areas.
- Metropolitan Areas have a large population nucleus of at least 50,000 persons plus adjacent communities that have a high degree of social and economic integration with that core. Metropolitan areas are composed of at least one county except in New England where cities and towns are the basic geographic units.
- Places are areas that have concentrations of people with legally prescribed boundaries, powers, and functions. These are related to Census-Designated Places, which are practically the same except that they are not creatures of law.
- Zip Code Tabulation Areas are approximate representations of five or three-digit postal zip codes. They are composed of census blocks and represent the majority zip code for addresses within the block(s). Not all housing units in a ZCTA will share the same zip code.
In searching for data, there is also a wealth of geographic areas for describing urban areas. The Census Bureau tabulates data for urbanized areas. These are densely settled areas of at least 50,000 persons. Urban clusters are densely settled areas of between 2,500 and 49,999 persons. A list of urbanized areas and a list of urbanized clusters are available on the Census Bureau's web page. You may also want to look at the state sorted list of urbanized areas and the state sorted list of urbanized clusters.
Special Features and Considerations: From Quickfacts to "Race"
- Available through: www.census.gov
State & County Quickfacts: The Census Bureau provides ready-made tables for each state and county with information about population, business, and geography. If you only need a few basic statistics, use this easy access interface.
Race Data: There are 63 racial categories in Census 2000. Understanding them can be very confusing. In answering the census questions, respondents were able to select more than one race to describe themselves for the first time. For example, a person might indicate that she is white, black, and Native American. In addition, she might indicate that she is Hispanic or Latina. Please visit the Census Bureau's FAQ's on race in order to get a complete understanding of how racial data and data on Hispanic or Latino origin are tabulated.
American Community Survey: Census 2000 marks the last long form that Americans will see. The American Community Survey will take the place of long form data and will be conducted in the years between Decennial Censuses. Not all geographic areas have American Community Survey data available yet. If you need up-to-date information, you may wish to check this data set.
Population Estimates of population are compiled by the Census Bureau between Decennial Censuses. Data series on births, deaths, and international migration are used to update the base counts taken for the last Decennial Census. These estimates are used to allocate federal funds, to provide controls for surveys, and to monitor demographic changes. Estimates are published each July and are available at the county level.
Population Projections are different from estimates. Estimates provide the best statistical guess for the population on a given date; projections attempt to determine what the population will be in the future. The Census Bureau prepares population projections for the nation, the states, and the District of Columbia. Population projections are prepared less frequently than estimates.
The accuracy of the Decennial Census became increasingly doubtful as the nation and the Census Bureau prepared for Census 2000. It has been shown that African Americans, immigrants, and other traditionally marginalized groups often go uncounted. Statisticians and other social scientists have argued that sampling the population actually produces a count of greater accuracy than attempting to count each person, as the short form purports to do. As a result, the Department of Commerce brought suit against the U.S. House of Representatives to enjoin the enforcement of any provisions of Census Act that might prevent the use of sampling techniques to obtain a "100%" count of the population. In Department of Commerce v. U.S. House of Representatives, 525 U.S. 316 (1999), the U.S. Supreme Court held that the Constitution does not allow the use of sampling techniques to obtain total counts of the population. Essentially, the court prevented the use of sampling techniques to determine total populations for the purpose of establishing congressional districts. As a result, under-counted areas are also under-represented in Congress.
Links: Census How to and Reference Corner
The following is an alphabetical list of links that you may find useful in using the Census web site (www.census.gov)
Briefs (issue briefs on popular topics)
County and City Data Book: 2000 (selected statistics for counties, cities, and places)
Data Products Listed by Subject
Freedom of Information Act Requests to the Census Bureau
Geographic Products Information
Glossary of Decennial Census Terms
Glossary of Geographic Terms (including explanation of Census 2000 changes)
Housing Unit Counts (for nation and states)
Index to Census Bureau Web Page
International Statistical Agencies
New on the Census Bureau Web Site
Population Density Data (available only for 1990 Census as of April 7, 2003)
Public Use Microdata Samples (PUMS): Housing units and characteristics
Questionnaires (long form and short form)
Final Census 2000 Response Rates
Rural and Urban Definitions and Data from 1790 to 1990
Small Area Income and Poverty Estimates (for states, counties, and school districts for intercensal years)
Statistical Abstract of the United States
