Herman Hollerith

Introduction

Herman Hollerith was born on February 29, 1860, in Buffalo, New York, and died on November 17, 1929, in Washington, DC. He was the inventor of the first patented mechanized punched-card system, the technological foundation for the computing industry.[1] He established a company to pursue the innovation based on census processing in the United States and several foreign countries, including Russia, Norway, and France. He licensed the technology to other firms in Austria-Hungary, Great Britain, and Germany. Hollerith revolutionized the technology used for general statistics and accounts processing by private businesses as well. He eventually sold his company to a conglomerate in 1911 which eventually renamed itself the International Business Machines Corporation (IBM) in 1924. IBM pursued several lines of business, but until 1962 punched cards remained its main business in the United States.[2] Hollerith’s inventions and innovations provided the business foundation for IBM’s prosperity throughout its early years.

Four years after Hollerith’s death in 1929, IBM’s German subsidiary, DeutscheHollerith Maschinen Gesellschaft mit begrenzter Haftung (shortened as Dehomag), resurrected Herman Hollerith’s ethnic heritage as a key element in its successful strategy to become accepted as a loyal German company in the Nazi state’s autarky (self-sufficiency) schemes for war planning. Hollerith’s story epitomizes the entwinement of German and American relations as well as the misuse of ethnic identity. This component of his legacy contrasts sharply with his life in the United States, during which he barely used or identified with his German lineage and made few contacts with his European relatives. In all likelihood he spoke German with his parents, but his surviving business correspondence was exclusively in English and his children do not appear to have learned German. His business networks did not rely on a German-American community, unlike so many other immigrant entrepreneurs.

Family and Ethnic Background

Herman Hollerith was born to German immigrant parents. His father, (Johann) Georg Hollerith, was born in Grossfischlingen, Rhineland-Palatinate on September 18, 1808. He became a Lutheran priest and later a teacher of Greek and Latin at the Gymnasium (high school) at Speyer, Rhineland-Palatinate, in southwest Germany on the Rhine. According to family narratives, his father was a free thinker who got caught in the revolution of 1848 in the German states. He left with his family for the United States just after the Prussian suppression of the revolution and settled in Buffalo, New York. He had married his second wife, Franceska Brunn, on June 14, 1846 in Speyer.[3] Franceska herself was born in Speyer on May 30, 1818.[4] Georg had one child in Germany by his first wife and six children in America by Franceska. Herman was their last child. According to family narratives, Herman’s parents lived on the money from his father’s teaching and managing their own large farm holdings. It is not known whether they brought significant savings from Germany. His father died in an accident on March 9, 1869, when Herman was nine years old. Shortly thereafter, the Hollerith family moved to New York City. Herman’s mother made custom hats for socially prominent women, which, combined with farm income, provided for the family.[5]

At the age of 14, Hollerith began attending the College of the City of New York. (In the United States in the late nineteenth century, “colleges” were often more akin to upper-level secondary schools than to universities.) After a year and a half, he was admitted into the School of Mines at Columbia College (today Columbia University) for more specialized training, and studied mechanical engineering.[6] He graduated from Columbia in September 1879, at the age of 19, having met the graduation requirements solidly but without receiving any academic honors.[7] Education at Columbia placed a high emphasis on school subjects, and less weight on shop experience, which became a significant factor in Hollerith's career.[8] Although his inventions would eventually apply electricity as a prominent feature, he attended no classes or had formal training in electricity. Electrical engineering was only born with Thomas A. Edison's invention of the incandescent lamp and the building of the Pearl Street electric power system in New York City in 1882; it was first taught at Columbia in 1889, some ten years after Hollerith's graduation.[9]

Business Development

Shortly after graduation, Herman Hollerith was recruited as a special agent to join the Census Office, the organization established to process the 1880 census (until 1900 every U.S. census was organized and processed by a new, temporary organization; the permanent Bureau of the Census was not established until 1902). His participation came about through William P. Trowbridge, one of his professors at the Columbia School of Mines, who was already a special agent for the census. Hollerith joined another student and several instructors to perform the short-term lived job of compiling the census. The absence of a permanent institution meant the network of individuals with professional census expertise scattered widely after each census. The invitation offered a young graduate the possibility to get acquainted with various members of the network, which was soon to be dispersed across the country. Hollerith's assignment was to work on statistics relating to manufacturing.[10]

In 1881, the Census Office of 1880 started to downsize because of an appropriation shortage.[11] Hollerith became an instructor in mechanical engineering at the Massachusetts Institute of Technology (MIT) in Boston, but left MIT the following year to become an assistant patent examiner in the United States Patent Office in Washington.[12] In 1884, however, Hollerith resigned from the Patent Office and embarked on his main career as an independent inventor and entrepreneur that would last until 1911.

Later in 1884, Hollerith filed a patent application on his first statistics processing device, which used a continuous roll of paper and consisted of a mechanically-operated punch and an electrically-operated reader.[13] At that time, he expected to work on censuses for several states in 1885, but the job offers he expected did not materialize.[14] Instead, he began working as the manager of the Mallinckrodt Brake Company in St. Louis, Missouri.[15] This company developed electric attachments to help activate the standard pneumatic railway brakes, and Hollerith’s research eventually resulted in four additional patents.[16] He also filed a patent for a machine to corrugate metal tubing to be used for expansion joints in steam pipes.[17] However, none of Hollerith's innovations were quickly commercialized, except those for census processing. He resumed developing his census processing machine, and in 1887 moved back to New York.[18]

During his residence in St. Louis, Hollerith developed his first census-processing system by adding two major improvements. He replaced the paper strip with single cards, introducing the unit-punched card that came to symbolize his system, and he designed a tabulator, an adding machine that accumulated the information recorded on sets of punched cards and displayed the results, which was a real innovation. Hollerith’s first punched-card system could only count; future tabulators also printed the results. The machine was designed to process returned census data of individuals, the largest job in a census, which was difficult to handle because of the absence of organizational knowledge from the previous census. From 1886 to 1889 Hollerith’s punched-card system was tested by processing mortality statistics for the cities of Baltimore (1886), Jersey City (1886) and New York City (1889), and the processing of sickness statistics at the Office of the Surgeon General of the Army (1889). For the New York City and Surgeon General jobs, the tabulators were rented, not sold, and subsequently large punched-card machines were rented as well.[19] IBM continued this practice until it was compelled to sell tabulators, in 1956, after a significant anti-trust case. Hollerith also used these tests to formulate a business model whereby punched cards and machines were produced through subcontracting to third-party companies. Hollerith focused on designing, selling, and maintaining the equipment. This business model remained until Hollerith sold his company in 1911.

In a competition in September 1889, the U.S. office of the 1890 census selected Hollerith’s first punched-card system for processing its population census from the three candidates offered. His punched-card system was based upon three principles: first, representing the information of each census unit, such as the individual person, on a single card or slip; second, mechanizing tabulation; and third, electrically-activated card reading. The successful use of punched-card processing for the 1890 population census was reported widely among statistical communities in the United States and abroad. But his business had weak foundations because it was based on one large user, the temporary 1890 Census Office, and it had little capital. When the census job was completed, the rented machines were returned and the income ceased. Hollerith searched for additional census processing jobs in the United States and abroad, but after 1894 he did not have substantial revenue and started to look for commercial customers.

In 1894 Hollerith was approached by J. Shirley Eaton of the New York Central railroad, who proposed to use punched cards to generate and audit railway statistics. Hollerith developed an application and presented it to the railway in late 1894.[20] It was based on a new punched-card layout and a new adding tabulator, because the job required more functions than the count-based population census. The card was the same size as Hollerith’s card for the 1890 population census, approximately the size of a one-dollar bill, but now information was punched in 24 columns. Originally the punched card conceptualized each punching position individually. Now each column was printed with the numerals zero through nine from top to bottom. It was an ingenious and enduring arrangement. After working for nearly a year to get the assignment, Hollerith set up a trial installation in 1895. A few months later, the railway discontinued the trial because the adding tabulator was not reliable.[21]

However, Hollerith got a second chance and within a few months, he devised a new adding tabulator for the same job, which New York Central accepted after a second trial run. The new adding tabulator proved a success and the railroad became a customer starting in 1897.[22] Hollerith introduced an even longer card than the 1890 population census card.[23] The new card had 36 columns, instead of 24 columns, and could hold more information. In 1907, the number of columns on this card grew to 45, and in 1928 to 80. (The 80-column card was called the “IBM card” and became its standard punch-card format.)

Meanwhile, Hollerith also began to have success abroad. In December 1896, Hollerith had been awarded a contract to supply punched-card machines for the first nationwide Russian census in 1897. Previously he had done business in his own name, but he incorporated the company as the Tabulating Machine Company (TMC) in 1896. Hollerith sold his patent rights and business assets to the company and gained a controlling 50.2 percent of the shares. Ten of Hollerith’s friends bought shares, and the company raised approximately $12,000 in cash (or $330,000 in 2010 dollars). The shareholders were old friends who trusted Hollerith and his punched-card system.[24]

In 1902, the ad hoc 1900 Census Office became the permanent Bureau of the Census by Congressional legislation. This changed the relationship with Hollerith, and he became a major supplier. Until 1904 the Census Bureau had rented tabulators under the conditions set up by Hollerith. When the contract came up for renewal for the fiscal year 1904–05, the Census Bureau secured a reduction in the cost. Hollerith was against the concession, but he was overruled by his board of directors.[25] In early 1905 the Census Bureau got an appropriation for work to develop tabulating machinery based upon Hollerith’s early patents, which would expire the next year. Simultaneously, the Census Bureau tried to establish a contract with Hollerith for further use of his tabulating machines for the fiscal year 1905–06. Hollerith made it a condition in the new contract that the government would abstain from any experimental work on tabulating machines, but the Census Bureau refused. Instead, it withdrew its offer as well as sending back the rented Tabulating Machine Company machines.[26] It was not until the late 1920s that TMC once again supplied the U.S. Census Bureau.

Thanks to his corporate clients, the loss of the census contract was a setback for Hollerith but not a disaster because his business continued to increase in size. From 1901 to 1903 he worked hard to attract private industry. His correspondence paints the picture of Hollerith as his company's sole salesman, an impression substantiated by other sources. In 1901 he circulated brochures to several life insurance companies.[27] He designed punched-card applications for wage administration, sales analysis, and cost accounting to attract new customers.[28] In 1902–1903 Hollerith designed a sales analysis application for Marshall Field’s, the department store, and he developed a cost-accounting system and sold it to Taft-Peirce Company of Woonsocket, Rhode Island, and to the Bethlehem Steel Company.[29] (Taft-Peirce was a subcontractor that built punched-card machines for TMC and was owned by Hollerith from 1901 to 1915. In 1903 the New York Edison Company started using punched cards to produce sales analyses.[30] In 1905 the Pennsylvania Railroad also started using punched cards.[31] Thus by 1907 several large private industrial producers were customers: Marshall Field’s, Eastman Kodak, National Tube, American Sheet & Tin Plate Company, Pennsylvania Steel, Western Electric, and Yale & Towne. In addition, Hollerith continued to negotiate with Simmons Hardware, the H.J. Heinz Company, Regal Shoe, U.S. Steel, and several other railroads.[32]

Hollerith’s new strategy changed TMC from focusing on one large customer to having many customers, which challenged Hollerith’s perception of punched-card systems and changed his business operations. For processing the census, he had built, marketed, and maintained three different punched-card systems with 24, 26, and 36 columns. His expansion and leasing of punched-card machines compelled him to reuse his existing machines and curtailed innovations, which now became complex with the many business customers. The new punched-card jobs during 1903 to 1906 were based on machines in production in 1902. Thus, Hollerith developed a new series of punched-card machines, produced during and after 1907 to become his second coherent system, which included one improved line of punching, sorting, and adding tabulation machines. For instance, the 1890 census used the first system of machines and punch-cards. But the 1890 tabulators only could count and were manually operated. Thereafter Hollerith introduced several new features, including a column-based card, a keyboard punch, an adding tabulator, an automatic card reader, plug-board programming, and a sorter. Punched-card processing now included several separate machines that constituted an integrated system. They needed to be built for the same punched-card standard or size. By 1907, he based his second punched-card system on these innovative features and the 45-column punched card. The adding tabulator facilitated processing of general statistics, which required addition.

Hollerith was a drawing board engineer with limited personal shop experience. He always relied on shop engineers’ practical experience to develop his designs into final products. In his new second machine series, shop engineers played a crucial role in creating the machines. They were required to make the new machines simple to produce and easy to maintain. Hollerith's original punched-card system from 1890 had no moving parts; but now his more advanced sorter and tabulator did. This made design and production even more demanding. Eugene A. Ford of the Taft-Peirce Company had played a critical role in designing Hollerith's new keyboard punch that became a key element of his second punched-card system. The design was less complicated than predecessors and the punch-operators could perforate faster. Recognising this, in 1905 Hollerith engaged Ford as an inventor, but Hollerith did not establish a research and development department. In the years around 1900, the first American industrial research laboratories were established at the American Bell Telephone Company and the General Electric Company.[33] Hollerith, however, did not follow this model. He was conservative and TMC was a small company. Ford remained a single independent consultant-inventor, just as Hollerith remained. Ford's primary workplace was in Uxbridge, Massachusetts, far from the TMC in Washington, as well as from the firm producing Hollerith's machines.[34]

Foreign Business

Hollerith’s foreign business derived from a combination of his work to attract foreign costumers and foreigners approaching him unsolicited. In the spring of 1889, while he was struggling to get the order to process the United States 1890 census, he traveled to Europe and visited Berlin and Paris to find additional customers for his census machine. He exhibited his tabulator in Berlin and in Paris at the great exhibition commemorating the centenary of the Great French Revolution.

He had prepared for his visit by filing patent applications, as he wanted to protect foreign markets for his invention. These enabled a potential low cost strategy by producing machines at home and then rely on patent protection in the export markets. But a foreign patent was neither an easy nor a sufficient tool; for example, Austria and France required local production to make a patent legal.[35] In addition, punched-card machines required qualified support and maintenance, which was Hollerith’s main responsibility for the tabulators as they were rented. When installations eventually became scattered over a wider area, providing repair and maintenance services as quickly as possible necessitated the establishment of a support organisation. Thus, when the foreign contracts emerged, Hollerith began to have support problems, as illustrated by his business with Austria, Norway, Russia, and France, where he had to choose between trained locals and expatriate Americans.

In 1889 Hollerith returned empty-handed from his tour to France and Germany, but his visit and references in American periodicals to the punched-card processing of the United States census in 1890 sparked the production of an Austrian clone and later orders from Norway, France and Russia. Theodor Heinrich Otto Hermann Schäffler had a precision machine-making shop in Vienna, Austria. In 1890, he requested and received a licence to produce the Hollerith punched-card system for the processing of the Austrian census of 1890, which was a success.[36]

In Norway, Anders Nicolai Kiær, the head of the national statistics office (Det statistiske Centralbureau), met Hollerith at international statistics conferences in 1891 and 1893.[37] Hollerith offered Kiær a free tabulator for testing in Norway, and traveled with the machine to Norway in 1894 to personally assemble it. A local Norwegian engineering office took care of maintenance. After an initial positive test, the Norwegian statistical office bought the tabulator the following year for a price of $1,100 (4,000 Norwegian kroner; $30,000 in 2010 prices).[38] The 1890 Hollerith tabulator was used to process the Norwegian census of 1900 and several small assignments during the early 1900s. For the processing of the 1910 census, a set of Hollerith’s improved 1907 standardised punched-card machines was rented from Dehomag, the German Hollerith agency.[39]

In the 1890s, Russia also planned its first general census, carried out in 1897. The Russian Central Statistical Committee contacted Hollerith as early as 1891. After years of negotiations, in 1896 Hollerith obtained an order for equipment to process the census, which had the same magnitude as a U.S. census. All the machines were produced in America and assembled in St. Petersburg by a Western Electric agent, who also provided local maintenance.

In 1896 the national statistical office of France approached Hollerith to organize this new business through the Library Bureau, which had a location in Paris. Founded in 1876, the Bureau was an offshoot of the American Library Association in the United States. By 1896, the company supplied a variety of office machinery and equipment. The Library Bureau got the contract to supply punched-card equipment for processing a portion of the French census of 1896. A French mechanical engineer undertook maintenance in Paris.[40] However, the punched-card business was not that easy. The Library Bureau obtained no additional customers abroad and Hollerith terminated the contract in 1899.

In Europe Hollerith had success with census statistics in Russia, which conducted its first national census in 1897, and in counties using punched cards to improve details of their census statistics (Austria, France and Norway). Great Britain, Germany and other countries with well-established national statistical organizations did not use Hollerith’s first punched-card system, because of its marginal advantage compared with manual processing. In Britain, Hollerith was introduced to the statistical community through the offices of Robert P. Porter, a naturalized United States citizen born in England, who headed the United States census of 1890. In 1894, Hollerith lectured at the British Royal Statistical Society about his system.[41] The next year, he tried, in vain, to form an English stock company based on British capital to sell and support his census-processing system in the British Isles. Porter moved back to his native England in 1901 and started organizing a TMC agency, but it was not until 1904 that his group was able to raise sufficient capital to establish the agency, which they then incorporated as The Tabulator, Limited. The new company experienced a prolonged start-up period, partly due to initial problems handling the shillings and pence of the sterling currency. (Until 1971, sterling currency consisted of twelve pence to a shilling and 20 shillings to a pound, and the decimal-based Hollerith tabulators could not add sterling currency figures.) In 1907 the British agency was reincorporated as the British Tabulating Machine Company.[42]

Finally, the German TMC agency grew in 1911 out of Hollerith’s new systematic strategy to build sales organisations on the European continent. In 1910 he had contracted with Robert Neil Williams, an American engineer, to establish companies on the continent to sell his machines and manage his patents in the various countries. Williams already had engineering offices in Berlin and Paris, and he started his assignment in Germany. In his search for customers there he met Willy Heidinger, who was director of a small company that served as the German distributor for Elliott-Fisher, an American company that manufactured office equipment used for printing and addressing billing statements. Heidinger established a German agency for TMC in 1910 under the name Deutsche Hollerith Machinen Gesellschaft mit begrenzter Haftung (“German Hollerith Machine Company with limited liability,” abridged to Dehomag). The company was capitalized at 300,000 Reichsmarks (or $71,000 in contemporary dollars; $1.7 million in 2010 dollars).[43] Its business was based on processing national, state, and municipal statistics and operational statistics in large companies.

The Computing Tabulating Recording Company

Herman Hollerith and his directors sold the Tabulating Machine Company (TMC) in 1911 to a conglomerate established by entrepreneur Charles Ranlett Flint. Flint was a prominent promoter of mergers.[44] Flint’s offer for TMC was difficult to reject, because he was willing to pay $2.3 million ($55 million in 2010 dollars). According to the family’s narrative, Hollerith claimed failing health as a reason to accept the offer.[45] More likely he and the board of directors sold TMC because of the large amount offered, combined with Hollerith realizing that TMC was growing fast and already had reached a size and complexity that made it hard for him to remain the company’s sole decision-maker, his preferred management style.

Flint merged TMC with three other producers of diverse mechanical equipment and he named the new conglomerate the Computing Tabulating Recording Company (CTR).[46] The new company was headquartered in New York City, and Hollerith’s management force moved there from Washington.[47] After the merger, Hollerith resigned as general manager and later he left the board, as he did not care to attend meetings.[48] By selling and resigning as general manager, Hollerith freed himself from managerial responsibilities so that he could concentrate on designing future machinery development, which he always regarded as the more interesting opportunity. He stayed with the company as a consulting engineer for another decade. For the first three years after the merger, until 1914, this position made him one of the most influential people in CTR, as all proposed changes in the design of machines were to be submitted to him for approval before being offered to customers.[49] During these years, Hollerith made important improvements to his system and filed eight additional patents, including two applications (filed in 1914), which were shaped into the patent on “automatic group control,” issued seventeen years later, in 1931. This patent became a key to controlling the American punched-card industry until 1948.[50] His main collaborator remained Eugene A. Ford, who continued his development work in Massachusetts as an independent consulting engineer.[51] The shop work took place at another CTR company headed by local engineers. They even built their own machinery, for example, a new machine for perforating cards.[52]

In 1914 CTR hired Thomas J. Watson Sr. as its general manager. Watson came to personify the success of IBM as an American-based multinational concern from the 1930s through the 1950. He was president of the company from 1915 to 1956. Watson integrated the four companies constituting TMC, established the company’s famously efficient sales force, and created a research and development organization. Within a few months Watson persuaded Eugene Ford to move from Massachusetts to New York City to establish the first department for punched-card machine development. Soon a model shop was equipped and ten model makers were hired.[53] Hollerith never traveled to New York City to contribute to the development efforts, but for several years, he contributed through extensive correspondence. He and Watson had difficulties communicating.[54]

In 1911 Hollerith earned $1,210,500 ($29 million in 2010 prices) by selling his shares in TMC, and received a ten-year contract with the new company providing him an annual salary of $20,000 ($470,000 in 2010 prices), but it also included a non-compete clause that prevented him from establishing a rival company. The contract expired in 1921 and was not extended as Hollerith was no longer active with the company. Selling TMC made Hollerith a millionaire and he engaged in a life of leisure, which did not include operating in a large organization or making any substantial attempt to start a new business. In 1911 he acquired property in the Georgetown neighborhood of Washington, where he built a mansion. Two years earlier, he had acquired 230 acres of farmland in Tidewater Virginia, where he built a farm and raised pedigree Guernsey cattle, mostly for leisure. Hollerith died at his Georgetown home on November 17, 1929.[55]

Social Status

Herman Hollerith was born in America to immigrant parents. His father died when he was nine years old, and his mother ran the household while he grew up. He had five older siblings and a half-sibling who do not appear to have played a significant role in his life after he left his home, according to the history recorded in his letters and his children’s preserved narratives.

Hollerith appears to have explicitly referred to his ethnic identity as a second-generation German immigrant on only one occasion. He noted in a letter to his wife during travel in Italy in 1894 that “Mrs. Meyer is german, [sic] so I got along very well with her,” perhaps indicating some vague reflection about his ethnicity.[56] He had received high grades in German language (and mediocre ones in French) while at Columbia College.[57] Probably this was an outcome of speaking German with his parents. In his several travels to Europe from 1889 onward, his letters to his family do not refer to any language problems in Germany or elsewhere.

On September 15, 1890, as his punched-card business was beginning to produce substantial revenue, he married Lucia Beverly Talcott.[58] (She was born in Mexico on December 3, 1865, and she died in Washington on August 4, 1944.[59]) Both of her parents had been born in America. However, according to family narrative, she had visited an aunt in Austria just before she first met Hollerith in 1889.[60] They had six children: Lucia Beverly (b. 1891), Herman (b.1892), Charles (b. 1893), Nannie Talcott (b. 1898), Richard (b. 1901), and Virginia (b. 1902).[61] The children did not play any role in Hollerith’s punched-card business. The oldest child was nineteen years old when her father sold the business. His children do not appear to have learned any German, nor are they recorded as having had German-related activities or networks.

His biographer, Geoffrey Austrian, suggests that his choice of the engineering profession might have been inspired by an uncle on his mother’s side of the family who owned Brunn’s Carriage Manufacturing in Buffalo, the same city he lived in during his childhood. Other members of his mother’s family were locksmiths.[62] Stories of these relatives might have inspired Hollerith’s interest in studying mechanical engineering, but it does not explain his preference for electrical-based designs.

After Hollerith’s studies at Columbia between 1876 and 1879, he established professional networks and a social network in Washington. However, these networks and relationships do not suggest a particularly German network. As noted earlier, it was his college professor, William P. Trowbridge, who helped him to get a job at the 1880 Census Office.[63] This job, in turn, introduced him to the statistics community and supplied ideas that he used in developing his first punched-card system. At the census office,Hollerith met John Shaw Billings, who subsequently played a key role in modernizing Hollerith’s punched-card system and finding locations to test it.[64] While at the census office, Hollerith also found time to be a member of the Potomac Boat Club.[65] Again Hollerith utilised the census network in 1882, when he obtained a job as instructor in mechanical engineering at MIT in Boston through its president, Francis A. Walker, who had been census director from 1879 to 1882.

In 1885 Hollerith relocated to a job as manager at the Mallinckrodt Brake Company in St. Louis, Missouri. His biographer, Geoffrey Austrian, argues that his new employer, Henry Flad, was a cousin of his mother.[66] However, the German records of Flad’s birth and Hollerith’s mother’s baptism do not substantiate this.[67] Immigration records show that Henry Flad immigrated to the United States at the same time as Hollerith’s parents, and that they came from the same region of Germany, but it is not possible to substantiate a preexisting relationship that might have led to Hollerith’s employment in St. Louis.

In 1889 Hollerith started establishing business in Europe by filing patents in several countries and traveling to Europe to find additional customers for his census machine. But he did not seem to use these trips to find inspiration for additional capabilities or applications for his products. He was looking for a larger scale of business, not a larger scope or new ideas. Hollerith considered punched cards as a population census processing tool, and he saw the international statistical community as an extension of the United States’ statistical community, where punched cards originated. He sought out various national European statistical offices just as he tried to reach out to public statistics offices in the United States. In 1889, he exhibited his system at the Great Exhibition in Paris and also in Berlin, but there are no preserved details explaining why he chose Berlin as a demonstration site.[68] As noted above, he probably approached the Prussian statistical office while he was in Berlin but met with a lack of interest.

Resurrection

Yet Hollerith’s death in 1929 was not the end of his story. In 1933, Dehomag resurrected him in Germany and deliberately (re)spelled his first name with a double “n” (Hermann), adopting the German tradition. By this time, the Computing Tabulating Recording Company had been renamed IBM and Dehomag had become a subsidiary of the American company, with 90 percent of its shares owned by IBM and 10 percent by the company’s German founder, Willy Heidinger.[69] In the initial years of the Great Depression, Germany’s right-wing governments introduced import controls to attempt to control the country’s balance-of-payments issues. When the Nazis assumed power in January 1933, they further strengthened this focus, causing reductions in the import of manufactured goods.[70] These Nazi objectives made Dehomag vulnerable, due to its overwhelming degree of American control and the fact that it imported nearly all of the tabulating machines which in turn it leased to German customers.

Dehomag and IBM chose to focus on the possibilities of the German market, now under the rule of Adolf Hitler, which was difficult in the Nazi state rapidly developed a policy of industrial self-sufficiency (autarky) that further hampered Dehomag’s opportunities as a foreign-owned company with a business based on imported machines. To maintain its position, Dehomag began manufacturing punched-card machines in Berlin using technology that had been developed by its own personnel. There are no records indicating that IBM disagreed with this strategy. Dehomag worked hard to convince the government and Nazi party officials of the company’s German-ness. One key element of this was to invoke “Hermann” Hollerith. He was born in the United States and his parents were German emigrants. Dehomag asserted that “Hermann” Hollerith was German and referred to him as a “German American” and mentioned several of his German relatives. This discursive resurrection presented punched cards as an example of German ingenuity, and configured IBM and Dehomag as symbols of German heritage. Dehomag’s strategy enabled the company to avoid conflict with the German Ministry for Trade and Industry (Reichswirtschaftsministerium) that controlled foreign trade. The ministry probably saw that Dehomag was the only German company with expertise and capability to produce punched cards essential for the ministry’s statistics production. Dehomag enjoyed steadily mounting business during the German rearmament of the 1930s and during the Second World War. Dehomag-punched cards played a key role in the production of operational statistics in the German rearmament and warfare after the early 1930s. However, recent discussion on IBM, Dehomag, and Nazi Germany has included claims that punched cards were used to locate and persecute Jews in the Holocaust. This has not been substantiated.[71]

Conclusion

Herman Hollerith invented the first mechanized punched-card system and established and incorporated a business to pursue the innovation based on census processing in the United States, but his innovations branched out quickly to private businesses. His business was designing, selling and maintaining punched-card equipment, not its manufacturing. This accomplishment distinguished Hollerith from many individual inventors and innovators of his age. After 1907 he improved his original punched-card system to process general business statistics in response to private industry demand. The new equipment was a success and his company grew quickly. In 1911 he sold his company to a conglomerate, renamed IBM in 1924. The punch-card and tabulation machine business remained the core part of IBM until the advent of electronic computing until 1962 (which still rested on punch cards for storing information).

He was very entrepreneurial as a student and in his early career. He applied good engineering ideas to business problems and showed great capability in converting his ideas into innovations and establishing his own company, but his entrepreneurial scope remained the one-man managed company. When his company attained greater scope around 1900, he first reduced managerial complexity through standardization. When the standardized company gained volume, calling for additional managers, he sold it, in 1911, and his involvement diminished further as the dynamic Thomas J. Watson, Sr. took hold of the company in 1914. The sale of his company made Hollerith affluent and, from then onward, his main focus was a life of leisure at his Georgetown estate and his Tidewater Virginia farm.

How was Hollerith’s choice of education and his subsequent career influenced by being a second generation German immigrant? Hollerith might have been influenced by the handicraft traditions of his mother’s family, but his education and reading focused on engineering, not shop work. He took advantage of the training opportunities offered in the last quarter of the nineteenth century and used his educational career to build a useful professional network, but we know little of his motivations for selecting contacts and in most cases we can only reason from small pieces of circumstantial evidence. Overall, his German ethnicity or networks do not appear to have been a major driver in his career or his life, while they were also nonetheless not completely irrelevant. It is perhaps especially ironic, then, that he came to public prominence as a second-generation German-American only after his death, when his heritage was used as a component of a money-making strategy premised on cooperation with the Nazi state.

Notes