Overview » Executive Summary » Limits of Prediction: Unanticipated Consequences...

An Example: The Productivity Paradox

Since the 1950s, businesses and other organizations have adopted computers and computer-based tools with the goal of improving productivity — in other words, raising workers' output per hour of work. The United States alone has spent over $4 trillion on information technology and related expenses since 1960; this amounts to an annual average of about 10 percent of US gross domestic product.³³ In the 1960s, automation improved productivity dramatically. But since then, productivity across the United States' economy has grown only slowly. Improved productivity has remained elusive even for organizations which firmly believe that ICTs have helped them. Numerous economists and others³⁴ have studied this so-called "productivity paradox." The phenomenon is widely recognized, though its explanation remains contested.

One key element of the explanation is the phenomenon of unanticipated consequences. Gains achieved through ICT-based improvements are often counterbalanced by increased costs in other areas. Such costs have proven very difficult to anticipate. They tend to take such forms as:

• higher salaries for employees, given their higher required level of technical expertise;
• difficulty in performing related activities, due to bad or overly rigid design and incompatibility between systems;³⁵
• unexpectedly high costs of maintenance, upgrades, and follow-on systems;
• lock-in to specific technological platforms;³⁶
• transfer of data to new technological platforms as old ones become obsolete or otherwise unavailable;³⁷
• long, steep learning curves for complex systems, and
• failure of users to adopt the technology in the ways intended by the designers, due to social conventions, organizational disincentives, inadequate training, or simple inertia.³⁸

The productivity paradox vividly illustrates the enormous significance of unanticipated consequences of technological change.

» Next: Technology: Neutral Tool, or Politically and Socially Active?

33. Landauer, 1995.
34. Landauer, 1995; Strassmann, 1984, 1990, 1997, 1999.
35. Cooper, 1999; Borenstein, 1994.
36. Shapiro, 1999.
37. Brodie, 1995; Rothenberg, 1995; Task Force on Archiving of Digital Information, 1996; Conway, 1996; Dollar, 1999.
38. Orlikowski, 1992; Grudin, 1994; Balter, 2000.