Different Methods of Tabulating Census Data

YEAR: 1889
ITEM: Periodical
COMPANY: Department of the Interior,
    Census Office
COUNTRY: U.S.A.
IN OUR COLLECTION: Yes

Herman Hollerith's Electric Tabulating System is generally regarded as the beginning of electronic computing in this country. This document, with the typically long governmental title "Report of the Commission Appointed By The Honorable Superintendent Of The Census. Different Methods of Tabulating Census Data." was the basis upon which the contract for the 1890 Census was awarded to Hollerith.

The Commission tested three different methods and Hollerith's was by far the best. After running a test consisting of data from four districts in St. Louis (raw data taken from the previous census), the Hollerith method proved to be considerably faster in both the transcribing phase and in the tabulating phase. (see figures below)

All three methods used individual cards, each of which represented one person. Hollerith's method was to punch holes in the cards to store the information and then to read and tabulate the cards using his Electric Tabulating machine. A second method, offered by Mr. Hunt, involved writing information on cards using colored inks and a third method, offered by Mr. Pidgin, involved writing information on colored cards. Both of those methods used hand sorting to tabulate the data. The test results were as follows:

TRANSCRIBING PHASE:	TABULATING PHASE:
Hollerith's method -- 72 hours 27 minutes	Hollerith's method -- 5 hours 28 minutes
Pidgin's method -- 110 hours 56 minutes	Pidgin's method -- 44 hours 41 minutes
Hunt's method -- 144 hours 25 minutes	Hunt's method -- 55 hours 22 minutes

Related Items

Get Our Newsletter