Salem Press

Back More

Article Citation
Save this article
Print this Article/Save as PDF
Email this article
Copy Permalink to this Article to Clipboard✓

Hollerith’s Tabulating System

The tabulating machines Herman Hollerith provided the 1890 census had already developed beyond his original idea. His cards provided spaces into which the operator could use Hollerith’s key punch to enter forty discrete items. Clipping one corner of the cards prevented their being entered upside down or backward. Once punched, the cards went to the tabulator machine operator, who used a hand-operated press. The upper portion of the press had retractable pins in each position where holes could be punched in the card. Wherever the press found a hole, the pins completed a circuit that moved a dial by one unit. The machine was not limited to counting only forty facts. Hollerith used electrical wiring on the back of the tabulator to connect two or more categories that were registered on separate dials. The first run of the cards produced seventy possible combinations of data.

Hollerith expanded the analytic capacity of his machine by using a sorting box divided into compartments. As the tabulator counted the seventy combinations, operators chose cards with certain characteristics, perhaps race or sex or citizenship, and placed them into the compartments to be run through the machine later. In 1890, the cards were run through the tabulator seven more times to complete the tables of correlations the census director desired. For the 1900 census, Hollerith provided a vertical sorter that automatically dropped the cards into the chosen boxes. The 1900 tabulator also incorporated an adding machine, permitting it to add quantities as well as count units.

In adapting his machine for commercial and industrial use, Hollerith had to change his cards to record numbers, rather than the specific items of information the census required. To do so, he modified his card to use the entire surface, not just the edges. He divided the card into vertical columns, each having the numerals from 0 to 9; groups of columns could be assigned as needed to record relevant numbers. In the New York Central Railroad’s case, the first five columns of the card registered weights of shipments; additional columns recorded other data or prices. Hollerith’s system could now be adjusted to suit various industrial and commercial enterprises.

Machines were custom-built to suit the needs of the user. Until 1906, the electrical connections needed to correlate data required hand-soldering to process new combinations of data, a time-consuming process. Hollerith automated the changes by providing a plug board, permitting easy switching of the wires, effectively creating a process of physical programming that prefigured electronic programming of computers.

Great Lives from History: Inventors & Inventions

Herman Hollerith

by Milton Berman

American engineer

Hollerith invented tabulating machines using key punches to record data on cards, along with counting, sorting, and recording devices to process the data. After successful use in the 1890 census, his machines proved invaluable to governments and businesses all over the world.

Primary fields: Business management; industrial technology; mechanical engineering

Primary invention: Census tabulating machine

Early Life

Herman Hollerith’s father fled Germany after the failure of the 1848 revolution and died when his son was seven. Hollerith stubbornly resisted attending public schools and was tutored privately by a Lutheran minister. He entered the City College of New York in 1875, then transferred to the engineering program at the Columbia University School of Mines, graduating with distinction in 1879. When one of his professors became chief special agent of the U.S. Census, he invited nineteen-year-old Hollerith to accompany him to Washington.

Herman Hollerith’s tabulating machine, patented in 1889 and used to compile the 1890 census. His machines accomplished in one year what would have taken nearly ten years of hand tabulation

ph_hollerith-herman.jpg

Getty Images

Hollerith worked at compiling statistics on manufacturing. On the side, he computed life tables as a favor for John Shaw Billings, director of vital statistics at the Census. Hollerith credited Billings with starting him on the task that would consume the rest of his life by casually saying that there ought to be a machine to carry out the mechanical work of tabulating population and other statistics. Hand-counting the steadily increasing population took so long that there were fears that future counts might not be completed until after the succeeding one began, rendering the census useless.

In 1882, Hollerith accepted a position at the Massachusetts Institute of Technology teaching mechanical engineering, but he discovered that he disliked teaching and left after one year. During his time in Boston, Hollerith considered various solutions to the problem Billings had posed. Appointed assistant patent examiner at the United States Patent Office in May, 1883, he once again left his position after one year. Hollerith, who always preferred to be in control, then opened an office as an expert consultant on patents. Hollerith’s knowledge of patent law and procedures would prove invaluable in his career as an inventor.

Hollerith became engaged in 1885, but his fiancé died of typhus the following year. He would always thereafter be extremely solicitous about his family’s health. He married in 1890 and fathered six children; Hollerith’s three sons followed him into engineering careers.

Life’s Work

Hollerith applied for his first patent for his census machine on September 24, 1884. It was granted on January 8, 1889, along with two other patents describing improvements on his first design. The basic idea was to pass nonconducting paper between electrically charged pins and a metal plate or drum. Electricity flowing through holes in the paper representing coded information activated electromagnetic counters, permitting rapid tabulation of data.

Hollerith first planned to use paper tape but soon decided that tape would make retrieval of single pieces of information a tedious and time-consuming task. Realizing cards could quickly be reread, he settled on cards the size of a dollar bill, which would fit easily available storage boxes. Holes punched on all four sides of the card with a handheld conductor’s punch recorded the raw data. When punching cards by hand proved too slow and hard on the operator, Hollerith added a mechanical key punch, modeled after the typewriter; pressing keys punched appropriate holes in the card.

The director of the Census arranged test runs of Hollerith’s machine tabulating mortality statistics in Baltimore in 1887, and later in New York City and New Jersey. In direct competition with two other machines, Hollerith’s proved clearly superior and was chosen to compile the 1890 census. His invention was estimated to have saved the Census $5 million in labor costs. It completed in three months the simple head count that would have taken over two years by hand. In the next three years, Hollerith’s machine produced, for the first time anywhere, a complete analysis of the census data. The Census preferred to rent machines rather than purchase equipment that would stay idle for most of the decade. Hollerith liked the idea: Since he owned the machines, he could prevent anyone from altering his invention. Leasing became the standard way Hollerith placed his machines.

His success with the 1890 census won Hollerith immediate international fame. The Franklin Institute of Philadelphia hailed his machine as the outstanding invention of 1890. Hollerith was invited to speak at the Royal Statistical Society in London in 1894 and International Institute of Statistics in 1895. His machines were used for censuses by Canada, Norway, and Austria in 1891, and subsequently in Russia, France, Germany, and Great Britain.

Hollerith modified his machine to accommodate the needs of an agricultural census and also sought to adapt the tabulators to the world of business, since this would provide a steadier source of income than census work. He custom-made his machines to suit the needs of railroad companies and for use in commercial bookkeeping, management control, and sales and cost analysis. In 1896, Hollerith incorporated his business as the Tabulating Machine Company, holding 502 of 1,000 common shares. To ensure card quality, Hollerith required that customers use only those his company manufactured. Profits from sale of cards exceeded income from machine leases.

In 1911, Hollerith agreed to sell his company to Charles R. Flint, who specialized in combining smaller companies into larger entities. Flint combined the tabulating company with a maker of computing scales and a manufacturer of time clocks into the Computing-Tabulating-Recording Company (CTR). Hollerith received $1,210,500 for his majority stock holdings. However, he was not ready to retire and insisted on a contract as a consultant for ten years at $20,000 per year, granting him a veto on any changes to his machines and stating that he was not to be subject to orders from anyone.

His veto power proved an irritant to Thomas J. Watson when he was hired to run CTR in 1914. Watson wanted to push development of improved tabulators, but despite careful diplomatic handling of Hollerith, sometimes found the stubborn inventor blocking changes needed to match competitors’ machines. Watson would be embarrassed after World War II, when it became clear that the semiautonomous German subsidiary Hollerith established in 1910 was run by ardent Nazis who enthusiastically adapted the machines to aid management of the Holocaust. Hollerith retired in 1921. In 1924, Watson renamed Hollerith’s company as International Business Machines (IBM).

Hollerith continued to invent after selling his company, mostly designing improvements to his tabulator. His last patent was granted in 1919; at the time of his death, he held thirty-one U.S. patents, as well as patents from eight foreign countries. However, Hollerith never patented his punch card with its clipped corner, since he did not think it was an original idea. In the 1920’s, his attention focused on his Georgetown home, his farm, and a series of yachts. Hollerith died of a heart attack eight years after retiring.

Impact

Hollerith’s tabulator made it possible for the U.S. Census to complete future decennial head counts in time for the constitutionally mandated reapportionment of the House of Representatives. Counting by hand could not have kept up with the rapidly increasing American population. As Hollerith improved his machines, it became possible not only to count but also to correlate data. One could now easily determine how many families were headed by two parents, how many children were in each family, and what language they spoke, or any other combination of desired characteristics. The superintendent of the 1890 census boasted that for the first time it had been possible to produce the most complicated tables as easily as the simplest, providing invaluable information for governments, social scientists, and industry.

The improved machines proved particularly useful to managers of the increasingly large and complicated American business enterprises. After the New York Central Railroad used tabulators in 1896 to keep track of freight shipments, railway companies discovered even more ways the machines helped in running day-to-day operations and providing easily retrievable business information. Hollerith’s cost-accounting machines transformed the fields of commercial bookkeeping and industrial accounting. Department stores depended upon the machines for sales analysis. Insurance companies used hundreds of thousands of cards to keep track of their policies. Hollerith’s machines expanded female employment opportunities, since the devices were as easy to use as a typewriter.

For the first time, business as well as government could process massive amounts of information efficiently, economically, and in time to be of use. Until replaced by computers, for which Hollerith’s tabulators were a significant forerunner, his machines would be indispensable for data processing.

Citation Types

Type

Format

MLA 9th

Berman, Milton. "Herman Hollerith." Great Lives from History: Inventors & Inventions, edited by Robert F. Gorman, Salem Press, 2008. Salem Online, online.salempress.com/articleDetails.do?articleName=GLI_1183.

APA 7th

Berman, M. (2008). Herman Hollerith. In R. F. Gorman (Ed.), Great Lives from History: Inventors & Inventions. Salem Press. online.salempress.com.

CMOS 17th

Berman, Milton. "Herman Hollerith." Edited by Robert F. Gorman. Great Lives from History: Inventors & Inventions. Hackensack: Salem Press, 2008. Accessed December 14, 2025. online.salempress.com.

Table of Contents