IT'S AN ARTICLE OF FAITH that the technology of the information age will breed winners and losers: Technology- rich nations will outgrow technology-poor ones; companies with powerfully managed resources compete more effectively than the computational laggards; and upper-class children with access to personal computers should academically outperform poor children with no access.

Less obvious, however, is that there will be winners and losers in the most information-intensive organization in the world: the bureaucracy of the federal government. Some federal agencies and departments may be doomed to an impotent existence of technological obsolescence; others seem certain to bask in the advantages that clever exploitation of information technology can provide.


*Winner: An agency that has a sophisticated database system like the Pentagon's that enables it to show recalcitrant congressmen just how much defense money flows into each of their districts.

*Loser: An agency like the IRS whose new large central computer systems have been a disaster from start to finish, delaying millions of refund checks, affecting the buying plans of millions of Americans for months, helping to undermine taxpayers' faith in both the fairness and effectiveness of a tax system based on trust.

*Winner: Ironically, also an agency like the IRS, which plans to purchase at least 20,000 lap-sized computers that should enable it to perform audits on the spot, in a company's office, with a speed and precision impossible just three years ago, increasing collection abilities substantially.

*Loser: The Commerce Department's International Trade Administration, virtually unautomated and overwhelmed with export licenses waiting to be processed.

This is prompting screams from American companies who say the delay is costing them overseas business; Japan has highly automated and thus far faster export-license processing. ITA's poor data-processing capabilities have hurt it in its turf battle with the Pentagon for control over export licenses. To add insult to injury, the Pentagon is offering Commerce access to its computer networks for export-license processing.

*Winner: The Congressional Budget Office because of its ability to use desk-top computers to answer in hours questions from congressional committees that used to take a week. One of the most difficult questions in Washington is how much a proposed piece of legislation is likely to cost. CBO's personal computer network and the answers it can cough out gives CBO a higher profile on the Hill.

*Loser: The Social Service Administration because its ancient computer systems can't adequately cope with changes in calculations and regulations. Recently, it took the SSA half a year to make adjustments to pension payments that were by law intended to go into effect in three months. These glitches have prompted greater Congressional "micromanagement" of the SSA.

It's not just agencies that will ascend or decline as a result of the new technologies. The careers of individual federal employes within them will be affected. Examples:

*Winner: Any bureaucrat who can use his personal computer to tap into his agency's central database to instantaneously find out just what raw numbers and assumptions his rival is using to bolster support for a pet program. With this information, he will be in a superior position to challenge that project.

*Loser: Any bureaucrat who lacks immediate access to his agency's databases but discovers that he needs to present three new budget and employment-level scenarios to his boss before lunch tomorrow.

*Winner: Any bureaucrat who can program his computer to actively monitor key statistical indicators in his agency as an early warning system to anticipate and avoid problems. A program could be written, for example, that would warn Defense Department auditors instantly if the Air Force were paying more than 20 percent more than agreed- upon budgets for peripheral items such as toilet seats or coffee pots.

*Loser: Any bureaucrat who is denied access to central data bases because his boss doesn't trust his peons with the means for such high-powered second-guessing.

This disproportionate impact of new computer and communications technologies on individual federal employes and their agencies does't lurk in the way the government's $20-billion-a-year data-processing budget is divvied up. It's in the nature of the technology itself -- particularly personal computers.

After years of procurement delay, Washington's bureaucracy is sprouting a new nervous system. In five years, asserts one former General Services Administration official, all of the nation's more than 1 million federal white collar workers will probably have either a personal computer or an intelligent terminal on their desks.

The information-intensiveness of our bureaucracy, the way bits and pieces of federal data can be wielded to create constituencies or sabatoge programs and the increasing importance of being able to measure cost-effectiveness or conceal its absence, arguably make the personal computer the most intriguing new technology on the Washington scene since the photocopier.

If you buy the notion that knowledge can be power, the personal computer can be everything from a useful bureaucratic tool to a terrorist device.

It's analogous to the introduction of the telephone. Picture how the bureaucracy might have evolved if some agencies had said that all employes should have a phone on their desks while others -- more cautious or worried about unauthorized use of phones -- permitted only top employes to have phones. Lower-grade workers in the latter case would have to share phones -- one phone for every five people, for example, and that should only be on a party-line basis.

Clearly, the result would be a mess. Interoffice communications would be cumbersome. Communications with the outside world for some agencies would be a function of mail correspondence rather than instantaneous discussion.

The same holds for computing technology -- the other part of the government's technological nervous system.

The skilled agencies will use the technology to help identify opportunities and build new constituencies and support. They will have immediate access to key information instead of waiting the days or weeks for the data-processing department to catch up to its backlog of requests. Analysts will have the ability to churn out reports until the printer bins overflow.

A congressman wants to know what would happen if an agency regulation were completely eliminated? A personal computer spreadsheet model using minor variations of the same basic assumptions can generate a stack of printouts a yard high overnight. Individuals can do in hours what it took committees days to create.

That's the personalization of computing; that's the decentralization of access that has so many large-computer data-processing bureaucrats worried while program analysts lick their lips. Duelling computers. My PC- generated analysis can whip your PC-generated analysis.

But that's only the first part of personal computing's impact. Computers communicate. The can be linked into networks just as telephones are.

What happens when a congressional oversight committee insists on a direct link to an agency's computer? That cuts out the middleman of laborious oral and written testimony and taps right into the agency's information nerve center.

A committee can examine, day to day, week to week, the key agency indicators that it most cares about. An Armed Services subcommittee can track the procurement costs of a weapons system; the committee can inspect for cost overruns on a weekly or even daily basis.

A labor committee can track Labor Department efforts to track minority set- asides. A commerce committee can monitor Small Business Adminstration loan patterns.

Don't ignore the importance of the immediacy of data access. In the bureaucracy, there's a huge difference between knowing something on a weekly basis and learning what happened last quarter.

Don't ignore the fact that, as personal computers grow more sophisticated, they can be programmed to track variances or turning points in key indicators automatically. Think of how such a system might have been used to identify the way the Divad antiaircraft gun's poor performance record didn't match the billions spent on it.

In effect, there are now forces for the decentralization of computing power for monitoring and oversight that was unimaginable but a decade ago. Indeed, a worst case scenario might have an one agency's computer tracked by dozens of personal computers ranging from the OMB to a sister agency to a competing agency to the relevant House and Senate subcommittees.

Technically speaking, the bureaucracy's automation efforts fall under under the 1980 Federal Paperwork Reduction Act and the Office of Managment and Budget is responsible for overseeing agency efforts such as naming "chief information officers" to spearhead such computerization. The agencies have been granted a lot of discretion in what data-processing equipment they get and what they do with it.

But one can't legislate behavior. The diffusion of technology from agency to agency varies widely. Some of that is due to simple technological illiteracy. But much of it is due to concerns that decentralizing the ability to get information is synonomous with decentralizing power. As a rule, agencies don't like to give up power -- sometimes not even to their employes.

As a result, the proliferation of personal computers will probably merely exaggerate the existing disparities within the federal government. Agencies that are inefficient and unresponsive now, will probably be inefficient and unresponsive after computerization, only in a more embarrassingly obvious fashion. And in that portion of the government that works because clever people make timely phone calls to the right people to exchange information, we will see computer-to-computer networks performing similar functions.

In an era when budgets are being slashed and reductions in force are the rule, agencies that successfully manage the new technology will probably have an edge. Agencies mired in poor management practices will probably just automate them and fall further behind in their efforts to serve their constitutencies.

In a town like Washington, where what you know still is almost as important as who you know, the bottom line may well be that the federal bureaucracy could become the living illustration of the oldest joke in computerdom: To err is human, to really foul things up takes a computer.