The "Productivity Paradox" Requires a Human Solution

By Gary Chapman

November 18, 1996

Copyright 1996, The Los Angeles Times

In the late 1980s, economic researchers discovered what has came to be called the "productivity paradox." Despite an immense investment in information technologies -- over a trillion dollars since the beginning of the PC revolution in about 1980 -- productivity growth in the U.S. has been either stagnant or weak. And growing productivity is what contributes, more than anything else, to expanding opportunity and a better material life.

Over the last few years, some evidence has emerged that the productivity paradox is receding as companies undertake the fundamental changes needed to get the most out of technology. At the same time, however, some experts are beginning to worry that one of the hottest new trends in computing -- stripped-down PCs known as "network computers" -- may make the problem worse.

The productivity paradox has been described best in a readable and important book, recently released in paperback, titled "The Trouble With Computers: Usefulness, Usability and Productivity" (MIT Press) by Thomas K. Landauer, a professor of psychology at the University of Colorado.

Says Landauer: "[T]he information revolution has yet to produce vast and obvious economic benefits or bring widespread and major improvements in the quality of life." There are clear effects of computerization, such as the speeding up of production and consumption and the ubiquity of computers, but wages and productivity have stalled during the years we've spent most heavily on information technologies.

Why is this? Landauer and others have offered several explanations.

In recent years, managers have discovered that to gain advantage from computers in the workplace, they have to "re-engineer" the company to match their business with the capabilities of computers, rather than simply computerizing traditional methods. Some people call this "complementarity" -- investments in capital equipment, especially computers, need to be accompanied by a rethinking of the job process, employee roles and organizational hierarchy.

The so-called flattening of organizations -- pushing authority down to lower levels wherever possible -- and giving workers more autonomy and the tools they need to do their jobs have helped produce gains in productivity that can be related to computerization.

Landauer acknowledges that some firms seem to be "getting it," but he rejects the notion that the productivity paradox is a thing of the past. He critiques some of the rosier studies in his book, and he warns that the success stories "are not yet gains everyone experiences or can expect."

Three weeks ago, STB Accounting Systems of San Rafael, Calif., released a study showing that the typical PC user spends 43% of his or her time on the machine "futzing" with the computer -- loading or changing software, organizing the hard disk, tweaking the interface or trying out new features of applications.

This same study, also sponsored by the Gartner Group, a computer consulting firm, estimated that the actual cost of a PC to an employer is about $13,000 per machine, only 21% of which is the capital expense of the hardware and software. The rest of the cost is eaten up by administration, technical support and "futzing."

Managers already aware of this predicament, if not the actual figures, have started to take seriously the idea of returning to centrally managed computer resources, when "dumb" terminals connected to mainframes or minicomputers were run by data-processing managers.

Thus the recent surge of enthusiasm among many in the corporate computing world for low-cost network computers that would access programs and data on computer networks via the Internet. Why not do away with expensive desktop machines and all that futzing? Why buy Pentium or PowerPC workstations for people who do mostly word processing?

But the real problem with this idea is that centralization of computer management is likely to undo the progress made in integrating computers into productive workplaces.

Harvard professor Shoshana Zuboff, who may have started the movement of reconceptualizing work with her 1988 book, "In the Age of the Smart Machine," recently wrote: "The paradise of shared knowledge and a more egalitarian working environment just isn't happening. Knowledge isn't really shared because management doesn't want to share authority and power."

Landauer's entire list of solutions to the productivity paradox seems to fly in the face of the network computer concept. He recommends "UCD, UCD, UCD," as he puts it -- "user-centered design, user-centered development, user-centered deployment."

Without users in control of the tools they work with on the job, says Landauer, we'll continue to see a mismatch between capital investment and productivity growth. That's bad for everyone.

Zuboff has observed that consultants understand "how much of their sustained employment they owe to the fact that few managers actually know what goes on in their workplaces." When misunderstandings and poor work practices are coded into software -- as is likely to happen when a centralized computer resource manager tries to produce software to please all employees at once -- that can drive everyone up the wall.

Solutions to the productivity paradox are more complicated than we thought. A technical fix is not available. Remedies involve turning engineers into amateur sociologists, anthropologists into technical experts and workers into self-managers. The problems are human and organizational, not primarily technical.

Unfortunately, it would be characteristic of modern American business to try to solve a "bottom line" problem with a short-term fix, like reasserting control over workplace computers and information. But we'd continue to pay the price for this in our economy. The balance of power in the workplace is everyone's concern.

Gary Chapman is director of The 21st Century Project at the University of Texas at Austin. He can be reached at