Ten years after the start of their wholesale invasion of the U.S. workplace, there's a nagging little question about personal computers that won't go away -- are they worth all the money we're spending on them?
That is, are computers, especially desk-top ones, really helping the organizations that buy them become more profitable? More efficient? More productive?
Few doubt that's the case when the machines are used to automate some clearly definable task, such as to keep track of a bank's deposits or to run a company's payroll system.
But when it comes to the white-collar service sector, where tens of billions of dollars are being spent each year putting machines on nearly every desk-top, there's a growing body of evidence to suggest a "productivity gap" between investment and payback.
That notion is a surprising one in a culture that automatically equates computerization with progress.
Yet a slew of researchers who have looked closely at the numbers, such as Commerce Department statistics, can't find consistent links between computer investments and increases in organizational productivity -- despite the marketing claims of computer companies, which are always eager to find buyers for the ever more powerful systems.
"Those of us who use the machines think that a revolution has occurred," said Gary Loveman, a Harvard Business School professor who has studied the matter. "Yet it's hard to find evidence for that, either directly or indirectly."
A decade after IBM's introduction of its first Personal Computer, the current recession is forcing many businesses to curb their enthusiasm for the machines. Now, that appetite might be lessened further by this new research, which suggests that U.S. businesses may have succumbed to a juggernaut of dazzling, addictive technology that in some cases may not be able to justify itself.
For example, Ernst Berndt, an MIT economist, said his studies of 20 manufacturing sectors not only showed no link between computer investments and either productivity or profitability, but also suggested that some sectors were actually spending too much on computers.
That's in contrast to some other corporate expenses, he said, such as those for research and development, whose cumulative benefits can be spotted easily.
And Stephen S. Roach, a Morgan Stanley economist, has long argued that U.S. white-collar productivity continues to decline, despite the tens of millions of personal computers that have been brought in to help.
With computer expenditures now the biggest capital expense at many firms, pushing aside other investments, the question of their actual role in productivity has become crucial. The lack of conclusive answers has prompted the Computer Science and Telecommunications Board at the National Academy of Sciences begin an 18-month study of the topic.
It's a question that computer manufacturers themselves clearly take seriously, as evidenced by the number that helped fund the $200,000 study, which will bring together academicians, business people, labor representatives and others.
While some may find it remarkable that society is only now undertaking so basic an inquiry, Marjory Blumenthal, the executive director of the study, said Americans have a long history of being bewitched by technology but less inclined to ask fundamental questions about it.
Why aren't the billions of dollars spent on computers, especially desk-top machines, showing up in the numbers? Researchers give three possible explanations.
* One is that the problem is not in the machines but in the data; that computers are making significant contributions, though in ways we can't yet quantify. This may be especially true in the white-collar service sector, where "output" is a nebulous concept to which it is difficult to attach a number.
A related argument is that computers are transforming businesses in ways that have a clear business value but that don't show up in the government indexes, such as sales figures and employment levels, used to calculate productivity. For example, Ms. Blumenthal said, they may speed product development time or allow firms to pay more attention to customers.
* The second explanation is that we are still too inexperienced with computers to know how to use them most effectively. Proponents of this idea hold out the hope that new generations of more powerful software or new forms of computer networking will allow machines to live up to their still-unrealized potential.
* The third explanation for the productivity gap takes a far more radical perspective and faults computers themselves, especially desk-top machines brought into organizations for no clearly defined purpose.
These critics worry that the machines become seductive "time wells," for example, luring workers into spending untold hours on intricate spreadsheets or elaborate memos and presentations that are of doubtful real value.