Households by Type: 1970–2003

Updated About content Print Article Share Article
views updated

Households by Type: 1970–2003


By: Jason Fields

Date: November 2004

Source: Fields, Jason. U.S. Census Bureau. "America's Families and Living Arrangements: Households by Type: 1970–2003." November 2004. 〈〉 (accessed July 23, 2006).

About the Author: Jason Fields is a family demographer and a member of the Fertility and Family Statistics Branch in the Population Division of the United States Census Bureau. The U.S. Census Bureau was established to conduct the ten-year census needed to reallocate members of Congress among the states. As of 2006, the agency employs 12,000 and conducts approximately one hundred annual surveys annually in addition to the Census itself.


The taking of a decennial census during the first year of every decade was mandated by Article 2, Section 1 of the United States Constitution: "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."

When the Census formally began (population counts and descriptions had occurred informally in the USA prior to this, with the first count taking place in Virginia in the 1600s) in 1970, it was brief and quite specific, detailing just a few population characteristics for each household: number of free white males from the age of sixteen and up; number of free white males under the age of sixteen; number of free white females regardless of age; and the number of free persons of any other skin color, including an enumeration of Native American Indians who were not exempted from paying taxes. Slaves were simply counted during the first census.

At the outset, the only precise data collected concerned the name, date and place of birth, and gender of the head of the household, and this household was rather loosely defined by the person who considered himself or herself as such. The specification of males by age had to do with ascertaining areas of potential military conscription, as any free white male who had attained the age of sixteen years was eligible to be pressed into military service. The remainder of the free persons living under the same roof—figuratively speaking, as an entire plantation, with many houses and assorted staff and workers might be considered one household for the purposes of the census—were either counted or estimated, but not specifically listed out by individual identity, although they were tabulated loosely by race and gender. For the first five rounds of census-taking, this remained the case. Commencing with the sixth decennial census in 1850, the names of every member of the household were recorded, and demographic information began to be collected.

Although the questions to be asked and the types of data to be collected were relatively specific, the means by which that occurred were determined by each census taker. Printed forms for the census were not created until 1830.



See primary source image.


From the outset, the census defined the nature of the household by the person who was considered to be its head, and counted the other members as subsidiary. Starting in 1800, it categorized household members, in addition to the head, by number of free white males and free white females below the age of ten, between ten and sixteen years of age, between sixteen and twenty-six, between twenty-six and forty-five, and above the age of forty-five, as well as any other persons in the household subject to taxation. Slaves were simply counted, not categorized by age or gender. In addition to the personal data, census takers also indi-cated the county, parish, town, township, or city in which the household lay. Starting in 1820, the census began to enumerate the free "colored" persons and slaves, distributed by age, for each household. According to the instructions given to the marshals, who were the census-takers, for the conduction of the data-gathering for the 1820 census, the head of the household was rather loosely defined—typically according to functionality or type of dwelling or business establishment. The head of household could be a shop steward, factory owner, overseer, plantation boss, or any person who was in charge of the running of the building, whether home or business establishment. For the purpose of clarity, a specific date was set for the commencement of the census, and all data gathered was to be a snapshot of reality for that day. That is, persons who were alive on the target date but died before the census was taken would be counted as living. Babies born after that date were not to be counted.

In 1830, the census underwent several significant changes: for the first time, a printed form with exact questions was used; each individual in a household was to be enumerated according to age category (the number of which were greatly expanded) and gender. Free slaves and "colored persons" were counted together; deaf and dumb persons were counted as a group, as were blind persons, without regard to race, age, or gender. White persons who were foreign-born and had not become United States citizens were also counted. Beginning in 1850, each person, not simply the head of the household, was listed by name and demographics, for the census. The population census was concerned with dwelling places; directors of institutions and wardens of correctional facilities were considered household heads, and inhabitants thereof were considered as though they were family members.

In 1850, an accounting of slaves was made, with listings of each person who held ownership rights to any individual slave listed out. Slaves were enumerated by age and gender, and assigned numbers rather than names. Family received its first formal definition for the purpose of census-taking in 1850. It was as follows: "By the term family is meant. Either one person living separately in a house, or a part of a house, and providing for him or herself, or several persons living together in a house, or in part of a house, upon one common means of support, and separately from others in similar circumstances. A widow living alone and separately providing for herself, or 200 individuals living together and provided for by a common head, should each be numbered as one family. The resident inmates of a hotel, jail, garrison, hospital, an asylum, or other similar institution, should be reckoned as one family." By 1880, it had been recognized that the residents of tenement or apartment houses ought to be enumerated as separate families, so long as they dwelled in distinct portions of the buildings. With the passage of time, more specificity was added to the description of the inhabitants of an abode—each person's relationship to the head of the household was to be delineated as exactly as possible, whether husband, wife, child, boarder, sister-in-law, lodger, inmate, pupil, or servant (to name but a few possibilities).

With the passage of time, changes in family composition and household definition have been created. Households are defined according to those (whether one or more) who live together in shared space. One person who is involved in the financial maintenance of the dwelling is self-designated as the "householder." Households are defined by whether they contain "family" or "non-family." "Family households" are considered those in which at least two persons, one of whom is designated as the householder, are related by birth, adoption, or marriage. "Nonfamily households" are composed either of people living alone or people sharing a dwelling who are not related to the householder. Family households have been grouped according to: married couples living together either with or without children present; other family members constituting a household; and adult women or men who live alone. The number of married-couple households has declined significantly since 1970 (from forty percent down to twenty-three percent), and the number of single-parent households or family households without spouses present has increased from eleven percent in 1970 to sixteen percent in 2003.



Census and Identity: the Politics of Race, Ethnicity, and Language in National Censuses, edited by David I. Kertzer and Dominique Arel. Cambridge, UK: Cambridge University Press, 2002.

Handbook of Family Diversity, edited by David H. Demo, Katharine R. Allen, and Mark A. Fine. New York: Oxford University Press, 1999.

All Our Families: New Policies for a New Century, edited by Mary Ann Mason, Arlene Skolnick, and Stephen D. Sugarman. New York: Oxford University Press, 2002.


Ahlburg, D. A. and C. J. DeVita. "New Realities of the American Family." Population Bulletin 47, 2 (1992).

"The Changing American Family." New York Times (May 18, 2001): A18.

Lewin, Tamar. "Confusion Ensued After Census Report on Two-parent Families." New York Times (April 21, 2001): A16.

Schmitt, Eric. "For First Time, Nuclear Families Drop Below 25% of Households." New York Times (May 15, 2001): A1.

South Dakota Department of Labor. "From Inkwell to Internet: The History of the US Census." Labor Market Information Center: Labor Bulletin (May 2002): 1-4.

Web sites

U.S. Census Bureau. 〈〉 (accessed July 22, 2006).