The first section of the dissertation, “Experts and Machines in Corporate Administration, 1917-1958” examines the first groups of would-be managerial technicians during the period before their alliance with computer technology. Chapter two discusses the only group to have received an appreciable amount of historical attention: the scientific office management movement. Its relationship with technology was more complex, and more ambivalent, than previously supposed. These office management reformers argued that office machines would improve efficiency only when installed as part of a fundamental, centralized and technocratic approach to office work. In practice, however, they generally failed to win the managerial recognition they sought. Their claim to unique expertise in systematization was instead usurped by the promises of office equipment to provide a quick, easy fix.
By the 1950s, several new groups of managerial technicians were striving to assert control over the creation of administrative systems. Chapter three focuses on the self-proclaimed “systems men” of the Systems and Procedures Association. While the systems men took many of their ideas and techniques from office management movement, they eschewed the direct supervision of clerical work. Instead, they aspired to senior staff management positions heading “systems” departments in the divisional or corporate offices of large firms. They attempted to turn a mass of individual specialties, such as forms design, file systematization and manual writing into the core of a new profession. The systems men are contrasted here with the better known, if little understood, attempts of scientifically trained operations research specialists to infiltrate corporate management during the same decade. Their story illuminates the opportunities created by the widespread adoption of the multidivisional corporate form during the 1950s, and shows that relations between line and staff management remained the subject of heated debate.
The immediate post-WWII years saw a huge increase in the use of punched card machines for administrative tasks such as bookkeeping, billing and payroll. Chapter four is the first detailed historical account of the institutional context, practices and occupational identities associated with administrative punched card work. By the 1950s, punched card operators and supervisors, by most traditional standards skilled craft workers and foremen, were trying to use their technical expertise to win professional status and managerial recognition. Inspired by corporate accountants, to whom most punched card supervisors were ultimately responsible, they tried to establish themselves as “machine accountants.” Together, the machine accountants, systems men, and operations researchers show three distinct approaches to the construction of managerial expertise on technical foundations. Their assumptions and idea about what was, and was not, managerial were very different.
Section two, “Creating Data Processing, 1954-1958” documents the arrival of the electronic computer in the American corporation. Chapter five begins with an exploration of the bold claims made for electronics in general managerial literature, and then examines challenges and opportunities the computer posed for the each of the groups of managerial technicians previously introduced. Because the computer was so unfamiliar, contemporaries spent a great deal of time trying to explain why a computer might be useful, how to evaluate its potential within a particular business and what the benefits and perils of computer use might be. I build on this literature to document the process by which computers were sold and bought, focusing particularly on the role of feasibility study groups.
The construction of a new corporate institution, the data processing department, around the computer is the subject of chapter six. Data processing combined punched card work with elements of systems and procedures work, while introducing the new task of programming. This chapter looks at the ways in which early computers were actually used, reconstructing the tasks to which they were applied, the jobs created around them and the organizational and physical location of the new department. While the managerial literature promoted the computer as a revolutionary electronic technology, in practice data processing represented a gradual evolution of earlier punched card work. Within data processing, pay and prestige ascended from key punch work (data entry), to machine operation, to programming, to systems analysis, to supervision, to management. In the historiography of computing, this demonstrates that historians must approach the history of computer programming in the context of particular social settings (such as the data processing department, the laboratory, and the software tools team), rather than trying to generalize about the craft and identities of programmers in general.
Section three, “From Data To Information: 1959-1975” charts the gradual evolution of data processing theory and practice over the period. Chapter six begins with an examination of the applications to which computers were put over this period, and the relative pay of the different data processing jobs (operator, programmer, and so on). The applications changed only slowly from administrative to operational jobs, while pay and prestige within data processing changed very little. Yet the new technologies of second- and third- “generation” computers had been expected to bring rapid change to the practices and applications of corporate computing. These included including real-time operation, random access storage (the disk drive), operating systems, and high level programming languages. But while these new machines were supposed to provide the technical basis for on-line, managerial applications and for the consolidation of many smaller computers into a single large system, in fact the transition to the new machines was quite wrenching even without major shifts in the kind of work performed. By looking closely at the actual reception, limitations and potential of these technologies in the context of corporate data processing, I explain this apparent contradiction.
The “totally integrated management information system,” by far the most discussed corporate computing concept of the 1960s, is the subject of chapter seven. Firms tried to build enormous computer systems, into which they would place information on every aspect of their operations, and from which would flow exactly the information (including models and simulations) required by each manager. The idea was eagerly promoted by computer manufacturers, consultants, systems men, ambitious data processing managers, and operations research specialists. It promised to blur the lines between the technical and the managerial, and thus finally provide a firm platform on which the authority of the managerial technician could rest. In the process, they constructed the modern concept of the “information system” and first identified the computer as “information technology.” While impossible to construct, these grand systems profoundly shaped managerial expectations of computer technology, supported the expansion of computing operations, and set goals that have been pursued to this day. The concept was also entangled with the Cold War systems ideology of cybernetics and “total systems” thinking, and so the chapter includes an examination of the ties between this idea and the prominent military application of computer technology to produce semi-automated command and control systems.
The practices and institutions of corporate computing changed more slowly, as shown in chapter nine. By the early 1970s, data processing managers faced more criticism than ever for their failure to keep to budgets and deadlines, or to provide a reasonable quality of service. Yet they retained their faith that they deserved, and would receive, institutional elevation close to the top of the organizational chart. This chapter focuses on the ideas advanced to deal with the operational problems of data processing, including schemes to charge users for services provided and to outsource data processing operations. It also presents the early history of packaged software, another apparently revolutionary technology, from the viewpoint of data processing practice. It concludes with a history of the strangely persistent prediction that the death of the corporate application programmer was imminent, as programming chores were handed over to some new kind of technological fix. While this idea was current from the 1950s to the 1990s, the number of application programmers rose unabated.
Section four, “Professionalism in Administrative Computing, 1959-1975” spans three further chapters. This section looks beyond the focus on technological practices and corporate institutions of the previous section, and focuses instead on the role of professional associations in the creation of occupational identities among corporate computing staff. Chapter ten examines the creation of data processing as a professional identity, through the efforts of the National Machine Accounting Association (NMAA) to remake itself for the computer era. After renaming itself the Data Processing Management Association (DPMA) it embarked on an ambitious program of professionalization and education, including the creation of a certification program for data processing managers. Using the association’s remarkably detailed archival holdings, I show that these efforts foundered on the its internal contradictions: the association was unable to define a coherent body of data processing knowledge, or to retain the support of its existing punched-card oriented members while attracting the support of the better educated new staff drawn into corporate computing.
Chapter eleven looks at a parallel story, the attempts of a small but influential group of computer industry figures to shape a far broader professional identity, as “computer people,” in which data processing would be united with scientific and technical computing work, computer science and even computer industry sales work. I dub this the “pan-computing professionalism” movement. Most of its key supporters had scientific backgrounds and had worked at one time for the RAND Corporation or Southern Californian aerospace firms. During the 1960s they enjoyed some success in setting up new institutions, including the American Federation of Information Processing Societies (AFIPS) and a popular special interest group within the academically inclined Association for Computing Machinery (ACM). Their story, and an account of their interactions with the often suspicious DPMA, illuminate both the alternate identities offered to data processing staff and the reasons for the general lack of contact between data processing and scientific/technical/academic computing.
Chapter twelve extends the story told in the two previous chapters into the 1970s and beyond. Changes during the early 1970s saw the forces of pan-computing professionalism gain the presidency of the ACM, and the installation of new and more open minded leadership at the DPMA. The two sides moved toward collaboration in several areas, including a joint certification scheme and collaboration within AFIPS. Yet many tensions remained, and in the end they failed (both separately and together) to construct a real professional identity for the corporate computing field. The section concludes with a brief survey of more recent developments in corporate computing professionalism.
The final section, “Beyond Data Processing” gives a summary of the most important developments during the period 1975-2000, placing these more recent events in the context of the remainder of the dissertation. Chapter thirteen documents several important changes in the technologies and identities of corporate computing during the post-1975 period. A new professional identity, software engineering, developed as an alternative way of combining technical and managerial expertise in the development of corporate computing. Around the same time, the new importance of “data base management system” software accompanied an ideological shift (based on an evolution of earlier MIS thinking) toward a conception of the corporate computer department as a “data resource function.” Another important development was the collapse of the traditional monopoly over computing resources held by the data processing department, as networks, minicomputers and microcomputers proliferated. Important trends in this direction included development of office automation during the 1970s and of the idea of “end-user computing” associated with the personal computers of the 1980s.
Chapter fourteen explores the rise of the concept of the Chief Information Officer (CIO), the title accorded to most senior corporate computing managers today. Though presented during the mid-1980s as a radical new idea, most of the claims made for the CIO had been seen before for managers of data processing and MIS. Proponents hoped that the CIO would be responsible for all corporate information, rather than just for the management of computer systems. The CIO concept was built on the idea, first presented during the 1970s, that information was a corporate resource comparable to money, in need of a senior corporate level officer to husband it. The chapter also investigates general shifts in the corporate computing workforce, and in the position and role of the corporate computer department during the 1980s and 1990s. It concludes with a brief discussion of the increasingly important role of consulting firms in combining managerial and technical expertise, and of the impact of the recent Internet bubble on the management of corporate computing.
A concluding chapter restates some of the most important findings of the dissertation and explores their historiographic implications. I pay particular attention to the historical and historiographic implications of this story for our understanding of the social history of information. I also discuss some of the avenues for future work revealed by this dissertation.
Summary of the managerial movements considered in this dissertation, their associations, key techniques, technologies and distinguishing features.