This article is available as a PDF file with all images and graphics. Click here to download.

Census 2000: Trends and Issues in Counting the People


Nancy P. Gallavan

Taking a count of the people has been a federal government responsibility since our origin as a nation. The requirement to count the population was written into the U. S. Constitution, which specified that the “actual Enumeration shall be made within three Years after the first Meeting of the Congress of the United States, and within every subsequent Term of ten Years, in such Manner as they shall by Law direct.” The first decennial, or ten-year, census took place in 1790. The upcoming 21st census holds the special import of coinciding with the millennial year 2000.

In mid-March 2000, the Census Bureau will mail forms to approximately 118 million housing units in the United States to account for an estimated 275 million U.S. residents. (It is important to remember that the census is not limited to citizens, but counts all people defined as residing in the United States.) The requested date for returning census forms is April 1, after which census takers will attempt to visit all housing units whose residents did not return their forms. All field work will be completed by the end of November 2000; congressional apportionment counts will be delivered to the President of the United States on December 31, 2000; and on April 1, 2001, all states will receive redistricting counts in time for their next state elections.

The purpose of the decennial census is to collect essential data describing the number, location, and types of residents who live in the United States. These data constitute an information infrastructure that assists governments at all levels—federal, state, and loca#151;to plan and provide for the needs of the citizenry. For example, the data is used to:

> reapportion the number of Congressional seats to which each state is entitled for representation in the U.S. House of Representatives

> redraw the boundaries of state legislative districts and some local voting districts

> allocate funds, including more than $180 billion per year in federal program funds, for distribution to state and local governments

> help local communities grow and thrive by providing the information needed to plan for public services

> provide help in times of special need, such as weather disasters requiring emergency rescue operations or disease outbreaks requiring immediate and preventive health measures

> help the economy to be productive by assessing levels of productivity and consumer needs

> know more about ourselves as communities and individuals

This article looks at three aspects of Census 2000. First, it outlines the content of, and notes changes in, the short and long forms used to take the census. Second, it examines changes in the questions about race and ethnicity, and describes the rationale for asking these questions. Third, it probes the reasons behind miscounts (undercounts and double-counts) and emphasizes the crucial importance of obtaining a complete count of the population. Finally, it offers a series of suggested activities for teaching about Census 2000.


The Long and Short of Taking the Census

The first 15 decennial censuses (from 1790 through 1930) asked the same list of questions of all households in the United States. An essential change took place in 1940, with the introduction of sampling techniques that allowed the division of the census into short and long forms. Both forms contain a series of basic questions that constitute the “100-percent count”of the population. The long form includes additional questions and is distributed to a portion of the population selected to be representative of all.


The Census Short Form

In Census 2000, five out of every six housing units in the United States—representing about 83% of the population—will receive the census short form. This four-page questionnaire, which seeks six pieces of information about every person living in a household, is the shortest census form in 180 years. The Census Bureau anticipates that it will take approximately 10 minutes or less to complete.

Specifically, the short form asks the name, sex, age, relationship, possible Hispanic origin, and race of every person living in a household. Additionally, it asks for the number of persons living in the household and the type of housing tenure (whether the home is owned or rented). Each of these questions has been included because it is required by law to manage or evaluate federal programs or to meet federal case law requirements.

Household members include everyone living in a household on April 1, 2000. This includes all family members, housemates and roommates, foster children, roomers and boarders, and live-in employees (including all documented and undocumented domestic help) who stay there four out of seven nights of the week. The household count does not include college students living away while attending college; members of the Armed Forces; people living in institutions such as a nursing home, mental hospital, or correctional facility; or persons who stay at another place most of the time. These people will be counted as part of those institutions or other households.


The Census Long Form

One in six housing units in the United States—representing about 17% of the population—will receive the census long form. This includes the same 100-percent questions posed on the Census 2000 short form, plus 26 additional population questions and 20 additional housing questions. The additional questions are designed to provide socioeconomic data related to a wide range of government programs and federal requirements. In some rural areas of the United States, as many as every other housing unit may receive the long form, as a larger sample is needed to ensure that these towns and counties receive the same detailed information as do more densely populated areas.

The long form asks for an individua#146;s educational attainment, residence five years ago, language spoken at home, disability, labor force status, occupation, place of work, journey to work, income, telephone service, number of vehicles, and shelter costs, among other questions. The responses help us to understand long-term trends in the economy, migration patterns, labor force changes, housing needs, and other indicators of the nation’s well-being. The Census Bureau estimates it will take about 38 minutes to complete the long form.


Changes Since 1990

Five items from the 1990 Census have been moved from the short form to the long form on the basis that such sampling can provide the information needed reliably. One is the question of marital status. Four others involve people living in multi-housing units. No new items were added to the Census 2000 short form, and only one new item has been added to the Census 2000 long form (in addition to the five items moved from the short form). This item seeks information about grandparents as caregivers, reflecting a current trend in child care in the United States. The question was mandated by Congress to collect information needed for the Personal Responsibility and Work Opportunity Act (“welfare reform act”) of 1996.

(Note: Samples of the Census 2000 short and long forms are available at


Questions About Race and Ethnicity

Questions about race and Hispanic origin are part of the 100-percent count. The identification of race has been included in every census since 1790, although the form and purpose of the question has varied greatly over time. The question of Hispanic origin was introduced on the long form in 1970, moved to the short form in 1980, and has been further modified in Census 2000. A direct question regarding ancestry was introduced on the long form in 1980. However, questions that pertain to ancestry have been asked in different forms over time.


Hispanic Origin

The Census 2000 questions on race and ethnicity have been reorganized, in part to help clarify data on ethnicity and race in persons of Hispanic origin. The question on Hispanic origin now precedes the question on race. It asks whether the respondent is Spanish/Hispanic/Latino, to which all respondents must answer No or Yes. The possible Yes answers further delineate ethnic categories, as well as allowing for the respondent to self-identify. The response choices to the question of Hispanic origin are:

> No, not Spanish/Hispanic/Latino

> Yes, Mexican, Mexican American, Chicano

> Yes, Puerto Rican

> Yes, Cuban

> Yes, other Spanish, Hispanic, Latino (with space to self-identify)

Census 2000 considers that people of Hispanic origin may be of any race. Persons of Hispanic origin may select one or more of the race categories described in the subsequent question on race.



The second question on race and ethnicity asks the respondent to identify what race this person considers himself/herself to be. Fifteen choices are available, including the choice to self-identify.

> White

> Black, African American, or Negro

> American Indian or Alaska Native (with space to self-identify)

> Asian Indian

> Chinese

> Filipino

> Japanese

> Korean

> Vietnamese

> Other Asian (with space to self-identify)

> Native Hawaiian

> Guamanian or Chamorro

> Samoan

> Other Pacific Islander (with space to self-identify)

> Some other race (with space to self-identify)

For the first time in census-taking history, respondents may select more than one racial category to indicate all self-identified racial heritages. However, the federal government will continue to record individuals as belonging to one of six groups, and will collapse responses into the following categories in order to compare and contrast previous census data:

> White

> Black or African American

> American Indian and Alaska native

> Asian

> Native Hawaiian and other Pacific Islanders

> Some other race



Still another question related to ethnicity appears on the census long form and involves an individua#146;s ancestry. The ancestry question is open-ended and asks the respondent to write in his/her lineage or ancestry. This permits respondents to identify groups not listed in the Hispanic origin and race questions. Census Bureau statisticians consider that this sample is large enough to produce reliable information for all but the smallest groups. The answers to these questions are key to knowing who we are and how the composition of our nation continues to change over time.


Why Ask These Questions?

The data derived from questions on race and ethnicity are critical for the implementation of various federal policies and programs. Two essential purposes involve monitoring compliance with the Voting Rights Act and providing the information needed to meet legislative redistricting requirements. These data are used in many other ways to meet community needs and promote equity in our society. All levels of government need this information to implement and evaluate programs required, for example, under the Equal Employment Opportunity Act, the Civil Rights Act, the Public Health Act, the Healthcare Improvement Act, the Job Partnership Training Act, the Equal Credit Opportunity Act, and the Fair Housing Act.


The Importance of a Complete Count

It is the Census Bureau’s primary mission to achieve the most complete count of the population possible. Unfortunately, a steady decline in the return of forms has occurred over the past three censuses. When people do not get counted, the communities they live in are denied both an equal voice in the U. S. government and an equal distribution of federal and state funding—vital funds most often needed for the very people represented by the undercount. Some communities must take funds from their “wants” list to provide underfunded services on their “needs” list. A decreased undercount could thus help everyone in the community.


People Who Respond Most Frequently

The historical trend in responding to the Census has shown that householders 55 years of age and older are more likely than younger adults to mail back their census questionnaires and get counted. Specialists at the Census Bureau speculate that the higher response rate among older people can be attributed to their belief that completing the census is their civic duty and was written into the U.S. Constitution to ensure fair representation in Congress. They may think that returning the census form promptly saves the government tax money by eliminating the need to send a census taker to individual households who do not mail back their forms. Older people may also better understand the relationship between the census count and the distribution of funds to their communities, particularly for services they and their families currently receive.


A Challenge: Where to Count People

One challenge for the Census Bureau involves where to count people in our mobile population. It is important to note that anyone staying in the United States on the average of four of seven nights of the week should be included in the Census 2000 count. Many older people live in different cities or states during various seasons of the year, and may wonder which location to consider their permanent residence. These people tend to follow a yearly cycle and should be counted at the residence where they spend most of their time during any given year.

The same rule applies to all people who divide their time among more than one residence, including children who may live in more than one home. For example, people receiving short-term care at hospitals or other health care facilities are asked to count their household as their permanent residence, while those receiving long-term health care will be counted by census takers at the health facility. Many other people living in special places and group living quarters need to be counted.

To count those who aren’t living in their usual residences from April to June, 2000, census takers will interview people staying at campgrounds, fairgrounds, marinas, and so on. For people living in remote areas of the U. S., such as parts of Alaska, Census 2000 will be conducted prior to the spring thaw when these individuals begin their annual migrations to other locations. Likewise, migrant and other seasonal farm workers will be located and counted where they live and work, including undocumented laborers in camps parked near work sites along unnamed roads. Census takers also will visit places where people without housing receive services, such as emergency and transitional shelters, food banks, and soup kitchens. All persons contacted or interviewed will have the opportunity to report their permanent addresses.


Undercount and Double-count=Miscount

It is estimated that 85% of all census forms were completed and returned in 1970, 75% were completed and returned in 1980, but only 65% were completed and returned in 1990. The decreasing response rate has been attributed to several factors, among them:

> apathy—misunderstanding and/or a lack of caring about the importance of collecting and using census data

> anger—irritation due to the perception that the government has too much information about individuals and/or the government is not distributing funds adequately

> fear–distrust and suspicion of the ability of governmental agencies to access data regarding immigration, taxes, and/or law enforcement, and to consequently deport, arrest, incarcerate, or disqualify individuals from various social welfare programs

> language barriers—difficulties with English, illiteracy issues, and lower education levels; the United States has a 21% functional illiteracy rate that affects the completion of the census count

> fatigue—tiredness among some citizens who feel the census constitutes redundant reporting and that they have completed similar forms through other agencies on a regular basis; in fact, the Census Bureau does not access data reported through other measures taken by states, counties, schools, etc. Fatigue is also experienced by census takers attempting to locate and interview all citizens, and some may not conduct the census thoroughly and accurately.

Although an undercount was evident among all racial, ethnic, and socioeconomic segments of the U.S. population in the 1990 Census, it was most acute—in fact, the highest ever recorded since 1940—among specific racial and ethnic groups. It is estimated that 4.4% of all African Americans, 5% of all Americans of Hispanic origin, 2.3% of Asians and Pacific Islanders, and more than 12% of Native Americans living on reservations constituted what is known as the “differential undercount” in 1990.

The differential undercount is obtained by comparing the regular census count to a second, separately conducted census whereby enumerators visit all housing units from which census forms were not returned. This is called the nonresponse followup (NFRU). Based on previous censuses, those who do not return forms are more likely to be urban, low income, and minority people. A proposal by the Census Bureau to use sampling to better estimate the undercount in Census 2000 was challenged by the House of Representatives in 1999. The case went to the U. S. Supreme Court, which decided against such use of sampling to correct the undercount [see “Looking at the Law: U. S. Supreme Court Trends” in Social Education 63, No. 5 (October 1999)].

Another form of miscount is double-counting. This may occur, for example, when residents living in special places or group living quarters are identified by their families as well as by the institutions. It is estimated by the General Accounting Office (GAO) that in 1990, the total miscount—including persons undercounted or double-counted—was more than 12 million people. That number equals the total populations of the states of Ohio, Michigan, and most of Illinois. Moreover, it is estimated that of the total persons missed, 50% were children under the age of 18.

In order to counter the downward trend in census returns, the Census Bureau is working hard to overcome barriers perceived as affecting the return of forms.


Ensuring Confidentiality

The issue of confidentiality is of grave concern to both the U.S. Census Bureau and most American citizens. The Census Bureau notes that census data are protected by Title 13 of the U. S. Civil Code, and cannot be shared with any others, including the Internal Revenue Service (IRS), the Immigration and Naturalization Service (INS), welfare agencies, courts, police, and the military services. The law protects responses from being sold or given to people who want to contact citizens in person, by telephone, or through the mail.

Census workers are sworn to secrecy prior to employment and must pass security and job reference checks; moreover, they cannot currently work as tax collectors, assessors, or law enforcement officials. Breaking the law is punishable in fines of $5,000 and a five-year prison term. The Census Bureau reports that their record of excellence for protecting the public’s privacy remains unbroken. This includes the denial of information about persons of Japanese origin for use in the Japanese relocation program during World War II.

Although individual records are held confidential for 72 years, people can request certificates from past censuses to use as proof in establishing age, relationships, and other information needed to qualify for pensions, citizenship, inheritance, and so on. Many people interested in family history use census data to learn more about their own families and communities.


Offering Language Options

The Census Bureau believes that many census forms were not returned in past decades due to language barriers. To meet the various language requirements within the United States, it will send postcards to all housing units throughout the nation a few weeks before Census 2000 forms are distributed. The postcards will ask respondents to inform the bureau if they need a census form in some language other than English. Census forms will be available in English, Spanish, Chinese, Tagalog, Vietnamese, and Korean. Language Assistance Guides will be available in more than 30 other languages, and will incorporate visual guides to assist respondents in completing the questionnaires. The Census Bureau’s network of Questionnaire Assistance Centers will also work to ensure that all residents receive a Census 2000 form written in their fluent language.


Working with Communities to Overcome Barriers

The Census Bureau recognizes that a complete count cannot occur unless people understand, and do not fear, the census. To achieve this understanding, the Bureau is forging cooperative partnerships with local communities and neighborhoods. Many communities are preparing to help by forming Complete Count Committees to help educate the public about Census 2000. These committees are being coordinated through state, regional, county, and city government offices, and are working with community institutions—such as schools and libraries, public service agencies, religious congregations, and business and labor groups—to encourage people to return their census forms. Special emphasis is being placed on especially hard-to-reach groups that are historically undercounted.

The Census Bureau is seeking to staff its network of Questionnaire Assistance Centers with community leaders and other local volunteers in order to bolster confidence in the process. Census takers know that the most effective tool for promoting and understanding the census is through individual conversations and word of mouth. The Bureau thinks this concept of “Each One, Reach One” probably will have the greatest influence on increasing the Census 2000 response rate.

Teaching About the Census

Teaching about the U.S. Census offers a multitude of fascinating explorations into our society’s history, geography, economics, and politics. Such study can easily be integrated across the curriculum at any grade level. Teaching materials for “Making Sense of Census 2000,” developed by the Scholastic Census in Schools Project, are available in kits or on the Web (see the box in this article). The U.S. Census Bureau website also has “Drop-in” sheets for teaching about various aspects of the census. How students can learn by volunteering to help with the census is described by Kathleen Kennedy Townsend in “Census 2000 and Service Learning” in this issue of Social Education.

The following are recommended activities for teaching about the census:

1. Teachers could begin by having students take a mock census count of the classroom using copies of the actual Census 2000 forms (available on the Web at After completing the questionnaires, students could discuss the questions on both the short form and the long form, the kinds of data collected, and why these particular data might be important for us to know. Depending on level, students could:

> develop maps, graphs and charts to “make sense” of census data

> research the uses of census information within their own local communities

> compare and contrast data collected in various censuses throughout our history

> form discussion groups or debate teams to probe difficult issues regarding the census

2. Guest speakers from public agencies could talk to students about how they make use of census data in providing services to the community. Or, students could make a field trip to one such agency in order to gain first-hand experience. Students could discuss or analyze the kinds of data needed to carry out particular community purposes. They could also discuss how the census undercount might affect these purposes. Students may want to become involved with Census 2000 by making posters or participating in community outreach projects.

3. Students could conduct historical research to discover the methods involved in conducting the census since its beginning in 1790. For example, they might be interested to know that the invention of an electric tabulating machine for reading data entered on punch cards that revolutionized the process of taking the census between 1880 and 1890. Inventor Herman Hollerith, a former Census Bureau employee, went on to become one of the founders of IBM. Another milestone in census taking occurred following the 1950 Census. It was determined that the use of census takers going door-to-door was too costly and introduced additional error into the process. For these reasons, the Census Bureau initiated the use of the mail in the 1960 Census. These and other changes could be identified on a timeline and related to other historical developments.

Teachers could integrate math and social studies skills in assigning students to develop maps, graphs, or charts that illustrate the findings of a particular census or a number of censuses over time. Depending on their level, students could create maps that show population densities or migration patterns, or develop charts or graphs using socioeconomic indicators to demonstrate important trends in the nation.

5. Elementary and middle school students could read about living conditions associated with a variety of times and places in U. S. history; they could write imaginary letters to people in the past posing questions about what it was like to live in this place or time. They could likewise write letters to imaginary students of the future telling them about the United States at the turn of the millennium. Students of all age could construct time capsules with significant information about their world to be opened at some point in the future.

6. High school students could explore such important census issues as the undercount and the importance of confidentiality in census returns. The census has at times been controversial (for example, limiting the scope of the census was an issue for some Southerners before the 1850 Census, and for some members of Congress—largely Republicans—before the 1970 Census.) The proposed use of sampling techniques to correct the undercount has been a controversial issue involving court challenges in the years preceding Census 2000. Students could find out why and how politics is involved in this issue. They could also interview members of their families or the community at large to discover how a sampling of adults feel about this issue.

Teaching about the census can engage K-12 teachers and students in a host of authentic learning experiences, including active participation in Census 2000 through service learning. Easily all ten of the National Council for the Social Studies strands of knowledge can be addressed through lessons integrating Census 2000.



Bureau of the Census


Census Catalog


Census 2000 Initiative


50 States & Capitals


Leadership Conference Education Fund


Map Man


Statistical Abstract of the U.S.


U.S. House of Representatives


Nancy P. Gallavan is assistant professor of Elementary School Social Studies and K-12 Multicultural Education at the University of Nevada, Las Vegas. She worked at the U. S. Bureau of the Census in summer 1999 in an Educator Externship.

©1999 National Council for the Social Studies. All rights reserved.