Census History

1790: First Census conducted by Secretary of State Thomas Jefferson, who reported two sets of numbers to President Washington. One set, written in black ink, was the official set of population numbers, which came to about 3.9 million. Jefferson also included a second set of numbers in red ink, which he characterized as representing a closer approximation of the actual number of people, even though they had not all been counted. Jefferson (and Washington) believed the true population was closer to 4.0 million. President Washington used the first Presidential veto on the apportionment bill because he did not agree with the formula used to distribute seats in the House of Representatives among the states.
1869: Rep. James Garfield was accused of trying to politicize the census when he proposed that the Census be organized by congressional district rather than the territories of the U. S. Marshals. At that time, the Marshals were appointed by the Senate, which objected to the proposal.
1870: Historians have said that a mistake in distributing Electoral College seats based on the 1870 census gave the presidency to Rutherford B. Hayes rather than Samuel J. Tilden.
1890: After spending eight years tallying by hand the results of the 1880 census, Census Bureau employees invented the punch card machine for the 1890 Census. The introduction of the punch card made the census more efficient and allowed earlier release of the data; however, it also introduced a new source of error in the census as the data were transcribed from the form to the punch card.
1902: Permanent Census Office created in the Department of Commerce and by the end of the 19th Century, professional enumerators had completely replaced U.S. Marshals as the primary census agents.
1910: Theodore Roosevelt vetoed the census bill because it did not make census enumerators part of the civil service but rather kept them as political appointees.
1920: Congress, faced with a census that showed a shift in the balance of power from rural areas to urban areas, called the numbers inaccurate. Some suggested that conducting the census in the winter caused an undercount in rural areas. Congress let the decade go by without reapportioning the House of Representatives.
1940: When more young men showed up for military service than predicted by the census, the Census Bureau began to study the undercount in the census. Census Bureau introduced its "short form" questionnaire for the majority of the population, using the "long form" set of more detailed questions for only a sample of the population. Prior censuses had required all residents to answer all questions.
1960: Following the 1950 census, noted statistician W. Edwards Demming and his colleagues concluded that the use of enumerators going door to door introduced error into the census, and that a system where people filled out the form themselves (self-enumeration) would be more accurate. As a result, in 1960, the Census Bureau began to collect the census forms by mail, and by 1970 most people were counted by mail, not by going door to door. Of course, counting people by mail was criticized when it was introduced.
1970: Both sampling and statistical procedures were used to add persons to the 1970 census. The 1970 census included about 4.9 million persons who were added on the basis of various statistical procedures, including sampling. For example, a recheck of a sample of housing units labeled vacant, revealed that about 11.4 percent of the housing units originally classified as vacant were occupied.
1980: The U.S. Department of Justice, under President Carter, issued a memo saying that the use of sampling is both constitutional and legal. The census used a statistical procedure called "imputation" to add 762,000 persons into the census count, to correct a mistake in labeling a number of housing units as vacant. This resulted in the shift of one congressional seat from Indiana to Florida. The State of Indiana sued unsuccessfully, with the courts upholding the use of statistical procedures in the census.
1991: Census Director Barbara Everitt Bryant recommended that results of an accuracy-check survey be used to adjust the 1990 census to correct measured undercounts and overcounts. Secretary Robert Mosbacher rejected the recommendation, and the U.S. Supreme Court upheld his right to do so in 1996.
1999: In a lawsuit filed by Republicans in the U.S. House of Representatives, the Supreme Court ruled that statistical sampling methods cannot be used to determine the population data used for congressional apportionment, citing a provision of the Census Act (Title 13, U.S.C.). The Court did not address the constitutionality of using statistical sampling methods to derive the population count.
2000: Census 2000 is considered the nation’s largest peacetime mobilization of personnel and resources, employing 860,000 temporary workers during peak operations. For the first time, the Census Bureau hired a well-known firm to develop a nationwide advertising campaign and created a partnership program to urge people to fill out census forms, which increased the mail response rates for the first time since 1970.