Census files in Suitland, Md., circa 1949. Computers were first used to help process the data in 1951.
It's that time of decade again. On Jan. 25, U.S. census workers began knocking on doors in Noorvik, Alaska, the first stop in an epic attempt to count everyone in America. Article 1, Section 2 of the U.S. Constitution calls for an "actual Enumeration" of the population every 10 years in order to determine how many Representatives each state gets in the House. The survey has also collected data on occupations, education and housing, among other subjects. The first Census, in 1790, was mainly a head count of free, white, draft-eligible men. Later queries were sometimes absurdly specific: in 1850, data collectors were instructed to "ascertain if there be any person in the family deaf, dumb, idiotic, blind, insane, or pauper." The 1870 Census distinguished between farmers and "farm laborers" and between housekeepers and those just "keeping house." (Enumerators were also instructed to "use the word huckster in all cases where it applies.") Until the Civil War, surveys differentiated free people from slaves, who had historically counted as three-fifths of a person.
Race has long been a muddled matter: 1890 classifications included mulatto, quadroon and octoroon, Chinese and Japanese. In 1930, Mexican was listed. The 2010 survey has caused a stir with the inclusion of Negro in addition to black and African American.
Despite the evolution of data gathering, miscounts have occurred, particularly among the urban poor. Democrats tend to say sampling--the extrapolation of data from smaller groups--is more accurate, but Republicans, suspicious of overcounting in left-leaning areas, argue that the Constitution's use of the word actual mandates a nose count. Getting it right is important: in addition to its role in doling out congressional seats, the Census influences the allocation of more than $400 billion in federal funds that affect the lives of some 300 million Americans. How many, exactly? It'll tell us that too.