One of the earliest uses of punched card tabulating machines, the forerunner of modern computers, was the US Census, as described by Wikipedia:
[Herman] Hollerith built machines under contract for the US Census Bureau, which used them to tabulate the 1890 census in only one year. The 1880 census had taken eight years. [...] In 1911, four corporations, including Hollerith’s firm, merged to form the Computing Tabulating Recording Corporation (CTR). Under the presidency of Thomas J. Watson, it was renamed IBM in 1924.
Logically, then, one might expect that will all the advances in computer technology that have happened since then, computers would be used at every step of the 2010 census. Ironically, it may not happen that way. CNN reports that a $596 million contract for handheld computers and support software for the census may have more than a billion dollars in cost overruns, and still may not deliver a working solution. The 2010 census data collection may instead be done using pencil and paper.
How did they do the data collection for the 2000 census?
The suboptimal way to take the 2010 US census
One of the earliest uses of punched card tabulating machines, the forerunner of modern computers, was the US Census, as described by Wikipedia:
Logically, then, one might expect that will all the advances in computer technology that have happened since then, computers would be used at every step of the 2010 census. Ironically, it may not happen that way. CNN reports that a $596 million contract for handheld computers and support software for the census may have more than a billion dollars in cost overruns, and still may not deliver a working solution. The 2010 census data collection may instead be done using pencil and paper.
How did they do the data collection for the 2000 census?