Vous êtes sur la page 1sur 13

The President and Fellows of Harvard College

"The Digital Hand": How Information Technology Changed the Way Industries Worked in the United States Author(s): James W. Cortada Reviewed work(s): Source: The Business History Review, Vol. 80, No. 4 (Winter, 2006), pp. 755-766 Published by: The President and Fellows of Harvard College Stable URL: http://www.jstor.org/stable/25097268 . Accessed: 30/07/2012 15:21
Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at . http://www.jstor.org/page/info/about/policies/terms.jsp

.
JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship. For more information about JSTOR, please contact support@jstor.org.

The President and Fellows of Harvard College is collaborating with JSTOR to digitize, preserve and extend access to The Business History Review.

http://www.jstor.org

Author's

Note

James W. Cortada

The Digital Hand: How Information Way Industries Technology Changed the Worked in theUnited States
any major business activity at the end of the twentieth cen could be done without computing, often linked to some form tury Hardly In my three-volume book, The Digital Hand, I of telecommunications. have surveyed the phenomenal growth of computing and have explored its impact on a wide array of industries, from the 1950s onward. The first volume was published in 2003, the second in 2006, and the third will come out in 2008, all under Oxford's imprint. This essay outlines my experience ofwriting and researching the book and describes some ofmy conclusions. I chose the title for the volumes to echo Alfred D. Chandler Jr.'s idea that organized business practices and structures provide a "visible hand" of rational operations in a modern economy. At least that was how I conceived of the project initially. As I worked my way up to the present, however, my research findings caused my views to diverge in creasingly from Chandler's description of the corporation. By the time I reached the end of the twentieth century, the corporate world described in The Digital Hand had come to look more like the networked and dif fuse society described by Manuel Castells than the highly integrated corporations of the precomputer era. Although managers used informa tion technology to carry out their missions?as described by Chandler, Peter Drucker, and others?as a result, and often unintentionally, they began transforming their companies.1 Given that computer applications tend to vary by industry, rather than by firm, I felt that the project required an examination of the expe riences of a variety of industries, one at a time, documenting exactly
research on the history and manage Society and Culture, 3 vols. (Oxford,

JAMES W. CORTADA is a historian who conducts ment of information technology. He works at IBM. 1 Manuel Castells, The Information Age: Economy, 1996-98). Business Review 80 (Winter History and Fellows of Harvard College. 2006):

dent

755-766.

2006

by The

Presi

James W. Cortada

/ 756

troleum, other transportation industries, wholesalers, newspapers, book publishing, and television? And what about the public sector, including state and local governments, and higher education? They were all ex tensive users of information technology (IT) and played important roles in the American economy. There was the problem of what to do about the new industries that the computer itselfmade possible, such as the software industry, the disk-drive industry, the PC industry, manufac turers of telecommunications products, such as cell phones and routers, and Internet service providers. Finally, I felt that the book had to ad and economic activity spawned by the In dress the new businesses ternet: e-Bay, for example, became almost an industry unto itself, and Microsoft had been a "New Economy" business icon for three decades, sharing celebrity status with Apple, Cisco, and SAP?all firms that owed their existence to the computer. I also believed that I needed to study enough industries to ensure

how computers came to be used and their effects on enterprises. I ini tially thought that selecting a half-dozen industries would be sufficient to accomplish the task, just as James L. McKenney and his colleagues had done in the early 1990s in their book Waves of Change.2 Upon reading their book, however, it became obvious to me that they had barely scratched the surface of the story, despite their good work. Early candidates for investigation included the automotive, retail, banking, insurance, and brokerage industries. But what about other chemical, pe large industries, such as railroads, telecommunications,

that patterns of technological, business, and economic behavior could be identified, documented, and placed in the context of late-twentieth century business practice. That set of objectives required looking at more than a handful of representative industries. My objective was to study enough industries, so that I could generalize about roughly 70 percent to 80 percent ofwhat all of them were doing, which meant ulti mately examining anywhere from thirty-six to forty-five industries and their use of computer technology. Both historians and executives influ enced the research agenda for this project. There were many models of The excellent historical scholarship to follow, Chandler's being only one.3 interests of senior executives proved to be another matter. During

2 James L. McKenny, Waves through Information Tech of Change: Business Evolution nology (Boston, 1995). 31 have been influenced as well by the work of David E. Nye, who recently summarized his ideas in Technology Matters (Cambridge, Mass., 2006), and even more so by Thomas P. inWestern Society, of Power: Electrification Hughes, particularly by his classic Networks on diffusion (Baltimore, 1983). Everett M. Rogers, Diffusion of Innovations, 1880-1930 strategies, was also an influence, especially the fourth and fifth editions of his book (New (New York, 1988). York, 1995, 2003), as was Eric von Hippel, The Sources of Innovation

The Digital Hand / 757


the three decades that I had worked with executives in the IT industry, and with many line executives (customers), I was always asked ques tions about best practices and about the role of prior experiences. Members of the Charles Babbage Foundation, which had established to the Charles Babbage Institute (CBI) at the University ofMinnesota were to who had ask executives of the computing, gathered history study basic questions about the history of this class of technology. Their issues also guided my work. In the end, I organized the book as follows: volume one describes all the industries thatmanufactured and sold physical items.4 The sec ond surveys those, such as the financial, telecommunications, media, and entertainment industries, that sell services.5 The third focuses on public-sector institutions, such as defense, tax collection, law enforce ment, specific agencies, state and local government, school grades K-12, and higher education.6 There were four main research questions. For which applications did enterprises use computers and why? To what extent were comput ers used from one industry to another? What were the results, or conse How did com quences, of using computing and telecommunications? puting change the organization industries? With these questions and overall functioning of firms and inmind, I began collecting material.

The notion that firms were "tribal" in their behavior, leading them to identifywith a specific industry, became the central organizing prin ciple for the study. An enormous amount of research material was available, almost all of it contemporaneous. To address managerial and operational issues, I turned to publications from the national industry associations themselves: weekly and monthly magazines that described many issues of the day, including the role of IT; industry conference proceedings that provided case studies and lessons learned; and indus tryseminars that articulated the benefits and operational details of adopt often promoted the use of computing and ing computing. Associations on the number of employees performing various conducted surveys tasks assisted by computing. To understand how many firms and industries used specific types
4James W. theWork ofAmerican Cortada, The Digital Hand: How Computers Changed and Retail Industries (New York, 2004). Manufacturing, Transportation, 5James W. theWork ofAmerican Cortada, The Digital Hand: How Computers Changed and Entertainment Industries (New York, 2006). Financial, Telecommunications, Media, 6 James W. Cortada, The Digital Hand: How Computers Changed theWork ofAmerican Public Sector Industries (New York, forthcoming).

James W. Cortada

/ 758

studies carried structure, and use of various technologies. Specialized out by the BLS on dozens of industries from the 1960s to the 1980s were particularly valuable. Many consulting and marketing firms also

of equipment, software, or applications, I relied on a large body of gov ernment publications, which included materials produced by the Bu reau of Labor Statistics (BLS) and Census Bureau data on industry size,

the Charles Babbage Institute. Individual associations and companies, such as the archives of the New York Stock Exchange and the Bank of America, also kept IT records. Each industry's experience with computing over the past half century had to be described, requiring, in effect, nearly forty industry descriptions. Recognizing each industry's unique characteristics became critical. Thus, insurance firms were different frommanufacturing com panies, software from banking, and so forth. So the first step was to understand what types of firms made up the membership of an indus try.Usually I could establish this by finding the national industry asso ciations. These associations listed members, defined their business roles that and segments, and often published weekly or monthly magazines served as a forum for airing industry issues and documented uses of computers over time. Then I compiled a list of published sources that described the history and contemporary business issues of an industry. economic analyses, and included articles in industry magazines, I think tanks. studied these sources to iden consultants and reports by an in what which firms industry, they produced, and tify participated the source of their profits, and to track their evolution from about 1950 to the present. That effort defined the context and described the back ground for the study of the use of computing. Reconstruction of the role of computers became the next step, re

tain additional information about computing's role in industry. Finally, research centers like the Rand Corporation and the Brookings Institu tion were also enlightening. To fill in the details on specific applications at the firm level, I turned to collections at the IBM Archives, corporate records for Remington at the Hagley Museum and Library, BankAmerica, and Rand/Univac the Burroughs and Control Data Corporation (CDC) records housed at

conducted studies throughout the period on the use and deployment of specific IT equipment and software. Many of these studies are stored in archives or were published as articles in contemporary business jour nals. The records of various congressional committee proceedings con

These

quiring scrutiny of the contemporary literature on how to use comput ers (and why) in a particular industry. Associations, academic presses, and trade-book publishers churned out dozens of books and thousands of articles on this topic, which appeared in bookstores, industry publi

The Digital Hand / 759


In addition to cations, technical journals, and academic publications. use to these works advocated the adoption how computers, explaining in of technology. Articles on computer use appeared simultaneously that others were grap dozens of industries, with no acknowledgement pling with the same issues.

I began the project at the end of the 1980s as a long process of col lecting materials on a wide range of industries, recognizing that some of these, such as agriculture or real estate, would not make it into the final study. Every week I read industrymagazines, themajority ofwhich were not indexed, requiring each issue to be perused, and I collected

back to the 1960s. Most economic white papers and many publications from think tanks could only be found on government and private Web sites. Because Web sites come and go, itwas essential to make elec tronic or hard copies of the posted documents and to state in endnotes the date when they were last accessed online. Researchers working on similar projects will have to be adept at using both the printed text and the Internet. Finding material often required the help of several types of individ uals. Because academic libraries rarely kept older publications on com puters, online book dealers became the single most important source for contemporary books on such matters as computing in police work, use of IT in classrooms, and the role of telecommunications in banking,

photocopies of all relevant articles. This process of collecting materials by industry consumed most of the 1990s and continues unabated, even after completion of the third volume. The sources I relied on for this study reflected the broader reality of our working lives, in that some were in paper form, while others were already posted on theWeb. Most of the industry publications and aca demic articles were still in bound paper volumes. Many government publications were either only available on the Internet or were posted there in addition to hard copy, including official reports, some dating

Industry associations varied in their usefulness. LOMA?Life Office a rich library ofmaterials inAtlanta and Management Association?has a staff eager to find information and to help reproduce articles from the

preserved old computer-industry conference proceedings, as they had begun discussing the role of digital data and information management as early as the 1950s.

all published decades ago. CBI had gathered many of the IT industry conference proceedings as well. Itwas a pleasant surprise to find that the librarians at schools of library science had extensively collected and

James W. Cortada

/ 760

dozens of seminars and conferences the Association had hosted over the years. On the other hand, the American Banking Association (ABA)? one of the oldest in the United States and perhaps themost effective in a blow to scholarship persuading itsmembers to use computers?dealt when itdecided in the 1990s to dispose of its library,which had been in existence since the nineteenth century. The beginning and concluding chapters of each volume of The Dig italHand had to show some relation to one another: a challenge, since theywere written several years apart. Each volume was produced in se quence. However, all ofwhat eventually became volume one and a third of volume two were written before the publisher agreed to handle the project.

Key Findings of theStudy


A set of findings began to emerge as I worked, suggesting some broad historical patterns that transcended individual companies and industries. For one thing, industries did not operate in isolation but learned from each other. Firms operated similarly across multiple in dustries, whether in their accounting practices, in their shared manage ment values, or in theirmanagement of supply chains. Everyone within an industry used computers in the same way to perform the same func tions. Thus, when bankers initiated ATM services, they seemed to do so all at once within a short period of time. They often voiced the same

cost justification?and rarely initiated audits to rationale?quantifiable ascertain whether they had received the anticipated benefits. Overwhelmingly across all industries in each decade, management adopted IT in order to improve internal business operations and to lower operating costs. While much of the literature advocating com puter use focused on expanding marketing and distribution and selling more goods and services, managers treated those activities as welcome, but nonessential, benefits that were not required to justify the cost of projects. Companies preferred to implement computer use in small, in cremental ways that fit the size of their budgets and minimized the risk alike. Firms and industries incorpo of failure to firm and managers rated computing on a continuous basis. For example, they would add functions to existing software tools as new releases of the software were issued, such as moving from using a clerk to take down a customer's order over the telephone to having the customer submit the order via the company's Web site. Adoption of IT applications occurred in tandem with the fashion able managerial practices of the day: management by objectives in the 1960s; process and quality improvements in the 1970s and 1980s; and

The Digital Hand / 761


process reengineering and networking in the 1990s. Accounting and fi nance and what became known at the end of the 1980s as supply chains were themost nearly universal. Both uses warrant historians' attention across multiple firms and industries. As the technology made it pos sible to linkwork activities, these two classes of use became integrated in an incremental fashion across the decades, so that by the end of the 1980s, suppliers, producers, and customers had become more inter connected. Thus, for instance, whereas an automotive manufacturer in the 1960s would post a production schedule on its computer to deter mine the number of tires to order from a supplier, fifteen years later that same manufacturer would share the file with the tire supplier, in sisting that the supplier determine what types of tires were needed at individual plants on specific dates. By the 1990s, the suppliers were even participating in the design of the automotive firm's new products. Thus many vendors had become involved in the design of new vehicles'
components.

the technology itself. its users Furthermore, as an application or technology matured, and uses became more alike, regardless of industry. For example, using an ATM became a similar experience no matter the bank, just as word processing systems became more standardized regardless of the brand of software. The experience of IT therefore mirrored the shift to a changed way of working and living that historians had documented earlier for automobiles, electricity, or even supermarkets. Early adopt ers enjoyed momentary strategic economic benefits as a use became

Across most industries, managers embraced speed, shed costs, and reduced theirworkforces as they increasingly relied on computing. They added applications to a growing list of uses over time, such that by the 1970s in some industries, and about a decade later in others, firms had achieved a level of interdependence thatmade it impossible for them to return to practices that predated the arrival of the computer. Thus busi nesses and whole industries had moved to a post-Ford, postindustrial age style of operation that was not yet fully defined. In other words, they had moved to a way of doing work thatwas profoundly affected by

bolic contemporary press so often called it, but in reality companies were undergoing an incremental evolution in their use of computers. While "back-office" automation was normally and widely imple

standard in their industry. The period of time between first adoption and ubiquitous use ranged from three to seven years, and the speed of adoption hardly varied throughout the second half of the century. As the consequences of technological change accumulated, organizations changed incrementally. A review of a particular use or firm might lead one to the conclusion that there had been a "revolution," as the hyper

James W. Cortada

/ 762

mented in an incremental fashion, some transaction-oriented applica tions were revolutionary at the time of their initial introduction. Exam ples were the online airline reservation system SABRE, ATMs, and even the use of barcodes in the grocery industry. However, the advantages of novel initiatives disappeared quickly as other members of an industry adopted the systems and altered them. With the passage of time in an industry, an application became less of a differentiating factor and con ferred fewer benefits on the original early adopter. Nearly all the indus tries that I looked at underwent this experience, an exception being the U.S. military with itsmost advanced weapons systems.7 In the 1980s, as economists began debating whether computers improved a nation's productivity, managers of firms did not hesitate to acquire ever more amounts of IT and telecommunications. Apparently

they knew something that the economists did not. At the most basic level, the technology delivered the performance that these managers wanted, or needed. Over time, they relied so much on IT that itbecame part of the normal operations of industry. Hence, a "digital style" was emerging inwhich managers became accustomed to using information housed on computers. While late adopters of a new use tended to enjoy fewer competitive or economic benefits, the earliest adopters encoun tered more problems in implementing a novel application. IT adoption thus illustrates the path-dependency phenomenon. By the 1990s, vari ous surveys of "best practices" made it possible to quantify the eco nomic and operational benefits of IT. Over time, these studies revealed that outsourcing could be rapidly accomplished by leveraging existing technical infrastructures, such as the Internet, and preexisting online

soning in deciding to acquire specific technologies. Banking, for example, bases its operations on industry influence. The features common to banks would include active, influential indus try associations, extensive sharing of data (e.g., for check processing), multiple regulatory practices and requirements, and deployment ofmany technical standards that are shared and embraced by all members of
71was thinking ofmissiles and atomic bombs, but even the latter changed and later other countries, acquired them. as the Soviets,

applications. A key finding of the project that I uncovered was the resemblance of industries to ecosystems, which led me to conclude that historians and other scholars should look at the industries themselves, and not just at individual firms. Common patterns of IT adoption varied widely between industries, but they nonetheless showed homogenous uses of technology. Thus itwas possible to create industry matrices that de scribed their overall style of operations, their use of IT, and their rea

The Digital Hand / 763


the industry.These features require themanager, firm, or vendor of banks or similar industries to pay close attention to regulators, work closely with their industry associations, and even collaborate with other mem bers of their industry to accomplish important reforms. Banking, there fore, relies heavily on task forces, which are important sources of inno vation that are often overlooked by historians of technology or business,

thuswarranting further study. Often banks and similar groups developed new industry-specific applications of IT, such as the modern check in banking, bar codes in retailing, and CAD/CAM software inmanufactur ing. In short, some of themost important IT innovations were often not conceived by the vendors of computers, software, and telecommunica tions, but rather by industrymembers themselves, who requested ven dors and universities to develop new applications or technologies. As I proceeded with the project, new questions surfaced that are still unanswered. How well do the findings apply across all industries, both in the United States and in other countries? For most U.S. indus tries, they hold up well. However, we cannot assume that the retail in dustry has operated in the same manner in France or Japan as U.S. re

tailers. We know that manufacturing generally functioned in the same way, but that banking, for instance, did not. So it is not clear that we can generalize about activities in other countries based solely on the ev idence provided about the United States. It is certain, however, that American corporations imposed their uses of IT on their employees in other nations and increasingly on suppliers outside U.S. borders. Thus industries dominated by American firms, such as petroleum, software,

semiconductors, and software, had to conform to American methods, but those in telecommunications, finance, transportation, agriculture, or government had not necessarily followed suit. Can research on other countries' industries match the degree and intensity of our studies of American industries? Business and economic historians consulted for this project normally answered no. But such a research agenda may be Western Europe, Japan, Australia, and New Zealand, which feasible for have long histories of extensive IT use and a propensity to organize and communicate among themselves at the industry level. A second genera tion of young historians working in Europe is beginning to explore pos sible areas of study in the business history of IT. I suspect, however, that theywill be severely hindered by the lack ofmajor collections of ar chival and contemporaneous publications inEurope. Such a project would have to be tested on an industry-by-industry basis, probably even one country at a time,much theway Chandler researched Scale and Scope.8
8 Alfred D. Chandler bridge, Mass., 1990). Jr., Scale and Scope: The Dynamics of Industrial Capitalism (Cam

James W. Cortada

/ 764

I originally anticipated that the technological characteristics of the machines and software would require more attention than the question of how the technology came to be so widely deployed. However, I was issues required more surprised to realize that business and managerial A than technical discussions. space study concentrating on a particular use of examine its technical issues, as re industry's computing might

companies, industry niches, and services that still confuse "themarket" as new technologies and products appear. Some of these, like cell phones, which might be more accurately classified as consumer products, seem not to belong to the telecommunications industry at all. Another realization that came as a surprise tome while Iwrote these books was the degree towhich theAmerican economy had remained oli

tinct products, and tailors the size and proportion of its staff and fund ing to its own needs. Retail and commercial banking are quite different too, as are state and federal government agencies. Even contemporary textbooks that describe the activities of firms in any industry often proved to be inadequate or had to undergo multiple editions as functions evolved over time. Telecommunications represented themost extreme structure since its case, changed radically in 1984 with the breakup of the nation's AT&T, nearly monopolistic telephone service provider. Its an unleashed enormous, breakup complex, diverse explosion of new

cent scholars have demonstrated.9 An essential task that I undertook was understanding and reveal ing the ways inwhich an industrymade money. Learning about insur ance companies, for example, meant uncovering the sources of their in come and outflows and tracking their allocations of employees and assets, which was not always easy. Each type of insurance, such as life, property and casualty, and health, has its own cost structures and dis

gopolistic throughout the second half of the twentieth century. Invari ably, a half-dozen firms have dominated their industry, as occurred in and pe the movie industry, or even, as in the case of pharmaceuticals a stranglehold on their portion of the econ troleum, have maintained omy. In these industries, the largest firms often were the most innova tive early adopters of a new information technology or use. They could most afford to experiment, and once they had proved the value of a cer tain digital application, they had the resources to blanket their industry with the new offering. Although many small firms experimented and often were touted in the trade press as the next wave of the future, they frequently either failed to succeed (as happened with so many dot-com
Yates, Structuring the Information Age (Baltimore, 2005); Arthur L. Norberg, at Eckert-Mauchly and Management and Commerce: A Study of Technology Computers and Remington Rand, 1946-1957 Research Associates, Computer Company, Engineering (Cambridge, Mass., 2005). 9 Joanne

The Digital Hand / 765


firms) or were simply swallowed up by the larger ones, a routine event in the pharmaceutical and software industries. One could argue that computing reinforced oligopolistic economic behavior, because the tech nology made it possible for larger firms to absorb rivals and integrate them quickly into their operations, a common occurrence in the life insurance industry after 1980 and in some portions of the telecommu nications industry at the end of the century. This study confirms that until the chronology of an industry is ex amined,

it is not possible to appreciate fully the pervasive influence of regulators in the evolution of IT inAmerica. Regulators determined the timing of new business offerings, such as online banking and brokerage services. Their control over the direction of telecommunications was evi

dent when the FCC, in collaboration with the Congress, formulated many Act of 1996. Related to the of the elements of the Telecommunications influence of regulators was the role played by Congress, particularly in the context of themyriad laws governing opportunities in banking, var I found that regulators ious wireless media, and telecommunications. balanced the needs of markets against state-mandated imperatives, much along the lines described, for example, by Lawrence Lessig. Addi tionally, however, regulators wanted to promote the use of specific tech nologies, such as the Internet by schools and on-line commerce in the 1990s, and thus behaved much as Frank Webster describes.10 My own work reinforces Lessig's views, particularly in the second and third vol umes, where I describe how specific regulatory practices have directly affected the actions of telecommunications and education in kindergar ten through the twelfth grade.

though research is also being directed toward public institutions.11 It has become increasingly difficult to recall a time when computers were not integral toAmerican business.
Lessig, Code: Version 2.0 (New York, 2006); Information Society, 2nd ed. (Oxford, 2005). 11 The 2006 E-Readiness (London, 2006). Rankings 10 Lawrence Frank Webster, Theories of the

Deployment of computing across the American economy has been pervasive. On the strength of the evidence, one can reasonably con clude that computing is now ubiquitous in both the public and private sectors. Often discussions within business circles today are less about new applications and more about how to replace aging, expensive IT in frastructures. A recent survey conducted by IBM and the Economist In telligence Unit confirmed that consumers rather than the private sector constitute the frontier for new computer uses in the United States, al

James W. Cortada

/ 766

the files on which it is based will be deposited at the Charles where theywill be organized by industry and within Institute, Babbage I hope this will industries, often according to period or application. a on start to In to twenty others fast achieve their work. fifteen help years, nothing would satisfy me more than the news that The Digital Hand had become woefully out of date. late 2007,

can possibly serve as a template for con While The Digital Hand on further business studies issues, it could just as easily lead ducting other investigators to conjure up different approaches. By no means does my work cover the subject. Every sentence that I wrote felt like a headline, demanding the backup paragraphs, pages, or chapters in the detail that historians expect but that I could not always provide, be cause the publication had to be limited to three volumes. So there is room formany more related studies. When this project is completed in

Vous aimerez peut-être aussi