‘The Trouble with Computers’

To the Editors:

James Fallows, in his review of Bill Gates’s The Road Ahead, discusses my book, The Trouble with Computers: Usefulness, Usability, and Productivity, along with some other books that are less sure than Gates’s that we are headed in the right direction. Being the subject of only two paragraphs of Fallows’s piece, I might not presume to respond. However, his offhand dismissal of the alarming evidence that I report is based on a fallacy that is extremely dangerous to our economic and social well-being. Fallows believes that the long inability of information technology (IT) to raise productivity is a thing of the past. As proof he cites the epidemic of corporate downsizing and rising stock prices.

Unfortunately, these developments are better explained by a very different interpretation: The failure of huge investments in information technology (about five trillion invested, about zero net return) has driven US businesses into a desperate gamble to avert ruin. They are trying to keep profits up by delivering poorer service (voice-menu receptionists, delayed installations) and lower quality (airline food and seat-room) using fewer underpaid and overworked people. More than one Fortune 500 company has been driven to extreme downsizing directly by huge expenditures—hundreds of millions—on computer-based “reengineering” efforts that were either abandoned or ineffective. I have heard of no case of demonstrated corporate-wide efficiency gains that come close to justifying the magnitude of the downsizings at IBM, AT&T, NYNEX, Hughes, Kodak, BellSouth… Meanwhile, highly productive firms like 3M eschew IT excesses and layoffs.

While there were good increases in labor productivity statistics in 1994 and 1995, it is a bit early to declare victory, especially as it is possible to suspect that some of the rise is accounted for by laid-off white-collar workers who are rehired as independent consultants and not counted as employees. Meanwhile, people work for less real pay, are frequently fired or fear for their jobs, feel diminished devotion to their companies, interest in their work, and loyalty to their companions. Fallows appears to have been over-impressed by industry promotions like The Road Ahead, common myths and mystique about computers, and the phony excuse that productivity gains in services go unmeasured, even though healthy progress was measured regularly before the dominance of computers.

A more realistic picture of the downsizing movement is one of frightened top management of efficiency-stagnant industries making devil’s bargains with Wall Street, mortgaging the future real effectiveness of their businesses and the morale of all America for short-term gains in profit and stock prices. The straight road depicted on the cover of Gates’s book appears to be the same one we’ve been on for too long. It does not lead where we want to go. We need more from computers than attractive toys and marketing wizardry, we need powerful tools that greatly increase the efficiency with which people, firms and industries produce goods and services that people want to pay for. Believe it or not (I think the data in my book would convince you), we have had a very few of these so far.

Past technological revolutions—in energy, manufacturing, agriculture, transportation, communication, medicine—have led to higher wages, shorter, more pleasant work hours, and better living. There is compelling evidence that information technology can do this for the service economy. That it hasn’t is due to a peculiar characteristic of the enterprise, the almost universal reliance of buyers on vendor promises—Clifford Stoll has it right when he calls it snake oil—rather than objective assessment of utility. Traditionally trained engineers and managers find it much harder to evaluate mind tools than muscle tools. But there are proven, affordable ways—often called user-centered design—with enormous benefit-to-cost ratios. Computers are wonderfully powerful devices, but their producers and users must learn to harness them more effectively. They won’t if we keep wishing the problems away.

Thomas K. Landauer
Institute of Cognitive Science
University of Colorado at Boulder
Boulder, Colorado

James Fallows replies:

Mr. Landauer has misconstrued my brief but wholly positive allusion to his book. In The Trouble with Computers he very effectively explained why the billions of dollars that companies have invested in computers have had such modest measurable results. Much of the problem, he showed, was caused by a gap between the mentality of software engineers and the needs of realworld business officials, workers, and customers. What seemed “logical” and “intuitive” to the computer designers was often awkward or mysterious to people who were supposed to use the new technology—for example, that the “Alt-F4” key combination would be a “natural” way to turn a computer program off. Therefore, Mr. Landauer argued, computers had been less truly productive for American industries than the huge investment in them would imply. Their benefits would continue to be limited until there was a closer match between the design of computer systems and the realities of the business world.
This analysis strikes me as convincing and consistent with many other findings from the business world (for example, that Japanese manufacturers have kept up their productivity despite investing much less in computers than American firms have). I also agree with Mr. Landauer’s contention in this letter that big US-based firms have not figured out the secret to true, sustainable increases in productivity like those that bolstered American prosperity through the 1950s and 1960s. But as a narrow matter of economic definition, big US firms are becoming more “productive” through their layoffs of the last half-dozen years. IBM, AT&T, Kodak, Boeing, and other firms have laid off tens of thousands of workers while increasing the value of their corporate output. This may or may not be wise long-term strategy, for the reasons Mr. Landauer explains in his letter. But it is indisputably a “productivity” gain in the technical sense of the term—more output with less labor—and is connected to the current economic anxiety about which Mr. Landauer and many other writers are concerned.