Following the 1880 census, the Census Bureau was collecting more data than it could tabulate. As a result, the agency held a competition in 1888 to find a more efficient method to process and tabulate data. Contestants were asked to process 1880 census data from four areas in St Louis, MO. Whoever captured and processed the data fastest would win a contract for the 1890 census.

Three contestants accepted the Census Bureau's challenge. The first two contestants captured the data in 144.5 hours and 100.5 hours. The third contestant, a former Census Bureau employee named Herman Hollerith, completed the data capture process in 72.5 hours.

Next, the contestants had to prove that their designs could prepare data for tabulation (i.e., by age category, race, gender, etc.). Two contestants required 44.5 hours and 55.5 hours. Hollerith astounded Census Bureau officials by completing the task in just 5.5 hours!

Herman Hollerith's impressive results earned him the contract to process and tabulate 1890 census data. Modified versions of his technology would continue to be used at the Census Bureau until replaced by computers in the 1950s.

Learn more about Herman Hollerith and his inventions at our January 2016 Webpage commemorating his 1889 patent for electronic tabulation.

Herman Hollerith's tabulator consisted of electrically-operated components that captured and processed census data by "reading" holes on paper punch cards. The primary components of the system are explained below.

To begin tabulating data, census information had to be transferred from the census schedules to paper punch cards using gang punches and pantographs. Using this equipment, Census Bureau clerks "punched" each card to represent specific data on the census schedule.

Census Bureau clerks using pantographs could prepare approximately 500 cards per day. To operate the mechanism, the operator positioned the punching stylus over the desired hole in a punch card template. Each hole in the template corresponded to a specific demographic category. Pressing the stylus into the template created a punched hole in the paper card that was read by the Hollerith tabulator's card reader.

Each Hollerith tabulator was equipped with a card reading station. The manually-operated card reader consisted of two hinged plates operated by a lever (similar to a waffle iron). Clerks opened the reader and positioned a punched card between the plates. Upon closing the plates, spring-loaded metal pins in the upper plate passed through the punched data holes in the cards, through the bottom plate, and into wells of mercury beneath. Pins that passed through the punch card completed an electrical circuit when contacting the mercury below. The completed circuit energized the magnetic dials on the Hollerith tabulator and advanced the counting hands. Upon completion of the electrical circuit (signaled by the ringing of a bell), the clerk transcribed the data indicated by the dial hands.

The 1890 Hollerith tabulators consisted of 40 data-recording dials. Each dial represented a different data item collected during the census. The electrical impulses received as the reader's pins passed through the card into the mercury advanced the hands on the dials corresponding to the data contained on the punch card (i.e., responses to inquiries about race, gender, citizenship, age, etc). When the bell signaled the card had been read, the operator recorded the data on the dials, opened the card reader, removed the punch cards, and reset the dials.

A sorting table was positioned next to each tabulator. After registering the punch card data on the dials, the sorter specified which drawer the operator should place the card. The clerk opened the reader, placed the punch card in the designated sorter drawer, reset the dials, and positioned a new card to repeat the process.

An experienced tabulator clerk could process 80 punch cards per minute.