Data Processing and Technological Change: The Post Office Savings Bank, 1861-1930 MARTIN CAMPBELL-KELLY In 1861 the British government established a national savings bank, using its network of post offices as feeder branches. The Post Office Savings Bank (POSB) provided high-security, nationwide sav ings facilities for a relatively affluent and mobile working class, for whom local savings banks and thrift clubs were no longer practical. By 1890, a large centralized data-processing operation had been es tablished in which more than sixteen hundred clerks maintained nearly five million accounts opened in more than ten thousand post offices. The bank remained completely unmechanized until the 1920s, by which time it had a head office with four thousand clerks, still processing transactions entirely by hand using Victorian fixedleafledgers . Large-scale mechanization was finally achieved between 1926 and 1930, using card-based records and bookkeeping ma chines. This article describes the evolution of the savings bank’s informa tion system and discusses the reasons for the slow adoption of the new information technologies. There were both rational and irratio nal motives for the resistance to office machinery, embedded in a culture of institutional fossilization and hostility to newfangled in ventions. Victorian Data Processing In England, several large-scale offices employing hundreds of clerks were established during the period 1840-70. Victorian data processing in England had two defining characteristics: specializa tion of clerical function and the absence of mechanization. Unlike Dr. Campbell-Kelly is reader in computer science at the University of Warwick, England. The research described in this article was undertaken while he was Simon Senior Research Fellow in the Center for the History of Science, Technology, and Medicine at the University of Manchester, 1993-94. It was first presented at the SHOT Annual Meeting, Washington, D.C., October 1993.© 1998 by the Society for the History of Technology. All rights reserved. 0040-165X/98/3901-0001$02.00 1 2 Martin Campbell-Kelly TABLE 1 Commercial Occupations, 1841-1901 Male Female 1841 ..................... +94,000 1,000 1851 .................. 91,000 — 1861 .................. 130,000 2,000 1871 ..................... 212,000 5,000 1881 ..................... 352,000 15,000 1891 ..................... 449,000 20,000 1901 ..................... 597,000 27,000 Source.—B. R. Mitchell and P. Deane, Abstract of British Historical Statistics (Cambridge 1962), p. 60. (■Includes railway clerks, classified under transport in later censuses. the clerk in the small office, who could perform a wide range of tasks, workers in the large-scale office were specialists—ledger clerks, casting clerks, correspondence clerks, writers, copyists, and so on. The workers were unsupported by any significant mechaniza tion, because the offices were established long before the availability of office machines, such as typewriters and adding machines, and predated the invention of loose-leaf and card ledgers. Information processing was achieved solely by means of handwritten entries in ledgers, hand copying of documents, the filling out of forms, and handwritten correspondence. The only important data-processing innovations of the period were rudimentary carbon copying and the use of preprinted forms. The rapid growth of office work in the United Kingdom in the 19th century is indicated in table 1. Between 1861 and 1881 the num ber of people in “commercial occupations” increased by 180 per cent, and the proportion of the workforce employed in these occu pations more than doubled.1 By contrast, in the same period in the United States the proportion of the labor force in the information sector grew by only 12 percent; it was not until the period 1880-1900 that the first twenty-year doubling of information workers occurred, when the information sector grew by 96 percent.2 In the United States, the rapid growth of office work from 1880 to 1900 coincided with the office machine boom, and it has been assumed generally ’There is no separate census tabulation for clerks, but workers in commercial occupations included about 70 percent clerks, the remainder being shop workers, salesclerks, and miscellaneous nonmanual workers. 2J. R. Beniger, The Control Revolution (Cambridge, Mass., 1986), p. 24. See also M. U. Porat, The Information Economy: Definition and Measurement (Washington, D.C., 1977). The Post Office Savings Bank, 1861-1930 3 that the one could not have...