U.S. Mounts High-Stakes Computer Reform Effort
As it has dished out more than $25 billion a year on computer systems, the federal government has appeared guided by this motto: “Buying yesterday’s technology tomorrow.”
But after decades of failing performance, in which billions of tax dollars have been squandered and the public welfare sometimes jeopardized, the government has embarked on an intensive and large-scale reform.
Outside the theater of national politics, far-reaching legislative and regulatory changes have been enacted to improve the way agencies, Congress and the courts buy information technology.
The high-stakes effort will determine whether the government can rise above past inadequacies in serving the public and, more important, live up to taxpayer expectations that a leaner government can provide better services.
If the reforms work, the Treasury stands to reap a good portion of the $170 billion owed by tax cheats every year; retirees will get shortchanged less often because of glitches at the Social Security Administration; passengers on commercial flights will be safer; and the government will spend less on clerical help and more on what propels American society forward.
Experts are encouraged by the breadth of the reforms, which include four major pieces of legislation that free bureaucrats from cumbersome regulations that once spelled more cost and slower decisions on every computer purchase.
After years of neglect, technology spending is getting a higher management priority under the legislative reforms, and for the first time the computer chiefs of every major federal agency are coordinating their programs and sharing their expertise.
“They are radically rewriting the rules and regulations,” said Robert A. Dornan, an analyst at Federal Sources Inc., a market research firm in McLean, Va. “No more micromanagement. I have been in this business for 30 years and have never seen anything like what is going on today.”
Despite good intentions, however, the government is confronting a long, hard climb. Success will be judged by whether the government can catch up to private industry. And by some measures, the gap between the public and corporate worlds is continuing to grow wider.
Federal Executives Still Defecting
Talented federal executives, who have managed to overcome long odds and establish successful records, continue to defect from the government. After years of rapid growth, spending on information technology is leveling off and some agencies are being starved, making it more difficult to know whether the reforms are working.
The government’s problems are so deep, and prior efforts to improve performance have yielded such slim results, that many experts--even proponents of the reforms--are cautious about the future.
“Anybody who says they are sure the reforms will work isn’t being honest,” said Steven Kelman, a Harvard University professor who is serving as the Clinton administration’s chief of federal procurement policy and spearheading many of the reforms. “Nobody can be sure.”
Kelman nonetheless has high hopes. “We are not trying untested ideas,” he said. “We are using ideas that have worked in the private sector.”
But Kelman’s implicit assumption--that government can emulate the behavior of major corporations and achieve the same results--is still unproven.
Ultimately, when a corporation performs incompetently in adopting new technology, the marketplace dispenses swift justice and puts it out of business. But there is no such invisible hand to get rid of an incompetent government agency.
Lacking a mechanism to purge failure, the government can only rely on regulatory policy to promote success. But regulations always change more slowly than technologies and markets, and outdated regulation is a key reason for the government’s sorry state in the first place.
In the 1960s, the greatest fear was that IBM so dominated the market for mainframe computers that the government risked being left beholden to what was then viewed as a classic monopoly.
As a result, Congress passed the 1965 Brooks Act, which set up a system aimed at ensuring that competition was fair and that low bidders won contracts. The General Services Administration was made the overseer of the entire federal establishment’s computer purchases--vastly more authority than it could intelligently handle.
By the 1990s, the dominance of IBM and mainframes was on the wane. The Brooks system became a yoke holding the government back from working closely with the best new high-technology companies.
“What we saw since 1965 were more and more layers of procedures to ensure fair competition,” said Dornan, the analyst. “The only thing that seemed to matter was fair and open competition.”
The Brooks Act system also became a legal quagmire. Losing companies in contract competitions routinely filed protests that forced the GSA to issue stop-work orders. In many cases, the government paid off losers to get them to drop their protests.
“It was straight extortion,” Kelman said.
In the meantime, the government computer market became dominated by defense contractors and others that hardly seem like household names: PRC, BTG, CDSI, Computer Sciences, Sysorex International, Science Applications International and Lockheed Martin. Even Boeing, best known for making passenger jets, got a piece of the action.
Long-dominant IBM sold its division that specialized in government programs. And such personal computer companies as Compaq and Apple opted to sell computers to the government largely through resellers that charge a commission. Other high-tech companies are often reluctant to deal at all with the government’s red tape.
Buying a computer became such a monumental challenge for government agencies that some tried to wrap everything into single modernization projects that took years to complete.
According to a 1994 Senate report, private-sector organizations take an average of 13 months between the time they identify the need for a computer and the delivery date. In the government, the average is four years.
So that systems are not obsolete before they are delivered, bureaucrats have attempted to leapfrog technology. The process has resulted in massive cost overruns, long delays and quirky technology far outside the commercial mainstream.
“The procurement system was such a nightmare that government managers wanted to experience that nightmare only once,” Kelman said. “Agencies were led to do grand designs.”
4 Key Pieces of Legislation
The quiet revolution to undo the Brooks system began in 1993 and has gained momentum. Congress has enacted these four key pieces of legislation:
* The Information Technology Management Reform Act, which revokes the Brooks Act and removes the GSA as the overseer of information technology. The new act also requires every large federal agency to appoint a chief information officer at a senior level.
* The Federal Acquisition Streamlining Act, which cuts red tape for small items, including personal computers, by allowing federal employees to use credit cards for purchases under $2,500 and simplified procedures for items under $100,000.
* The Federal Acquisition Reform Act, which eliminates a wide range of regulations governing big purchases. It essentially says the government went too far with fairness in competitions and frees bureaucrats from having to select low bidders.
* The Government Performance and Results Act, a watershed that for the first time requires agencies to measure the results they obtain from investments in information technology. Since the necessary yardsticks do not exist, the act gives agencies until 1999 to comply.
What motivated these changes was a collective realization in Washington that the federal government had a disaster on its hands, a conclusion stated in so many words in the first of the four laws:
“Poor planning and program management and an overburdened acquisition process have resulted in the American taxpayers not getting their money’s worth from the expenditure of $200 billion on information systems during the decade preceding the enactment of this act.”
“What we have done,” said Bill Greenwalt, an aide to Sen. William S. Cohen (R-Maine), who drafted much of the legislation and was recently nominated by President Clinton to be Defense secretary, “is go from a system that assured failure to one that gives us an opportunity to succeed. We have created an experiment.”
Acts Intended to Provide Freedom
The intent is to eliminate federal regulations that hamstring government agencies, provide freedom to bureaucrats and demand accountability when programs fail, according to John Koskinen, deputy director for management of the White House’s Office of Management and Budget.
“If the government wants to provide better service with fewer people, it will only be able to do that with better information technology,” Koskinen said. “Thirty or 40 years ago, the federal government was an innovator, and I don’t see why it can’t happen again. We care how we are doing in government. The key thing is that we are measuring performance and we care about it.”
Under the reforms, coordination among federal agencies is also increasing. A council of chief information officers from across government meets monthly and a review board has been authorized to investigate special problems, according to Koskinen.
Instead of fair competition, the emphasis now is on best value. Agencies don’t always award their contracts to the low bidder; instead, they select contractors based on past performance.
Rather than massive projects that attempt to modernize all at once, the emphasis is on a piecemeal approach to updating technology. And rather than issue a large contract to one firm, so-called “task order” contracts are being issued to a group of companies, which then battle for work orders.
Everybody is seeking speedy results. At the Defense Department, for example, a command and control system for operations in Bosnia was engineered in two weeks and put in place 10 weeks later, said Emmett Paige Jr., the Pentagon’s chief information officer.
“We believe we can be improved by this law and we want to see it succeed,” Paige said.
But Koskinen’s and Paige’s enthusiasm is not universally shared. Critics say the pendulum is swinging too far toward fast and loose decisions, opening the way for big mistakes and fraud.
“The existing contractors love the reforms, because tens of billions of dollars of opportunities are going underground,” said Dornan, the market analyst. “Nobody is going to know about these opportunities, except the company that gets the contract. There are going to be horror stories, guaranteed. Some guy is going to land a contract because his brother-in-law is the procurement authority. The question is, will Congress keep its hands off when this happens?”
Paradoxically, government attempts to reduce regulations often result in more complexity, not less. The Defense Department, for example, has flow charts of mind-boggling complexity to map out how simpler procedures for buying computers will fit into all its other regulations.
Much of the language in the new laws, moreover, simply exhorts government officials to do better, and many of the provisions that supposedly simplify acquisition rules appear vague.
Critics also worry that expectations about the reforms are inflated. Chris Hoenig, an expert in federal computer systems at Congress’ General Accounting Office, said the government is improving its use of information technology so much more slowly than the private sector that just not falling further behind seems practically out of reach.
“We are never going to change the government,” Hoenig said.
Undercutting government reform, the GAO said, is a serious lack of talent. “The skill is not commensurate with what the government is trying to accomplish,” Hoenig said. “Federal agencies are not identifying the skills they need and they can’t afford to pay for the skills they need.”
Talented government executives, who are high on frustration but low on compensation, often leave. In the last year, such highly regarded executives as Lloyd Mosemann of the Air Force, Renato DiPentima of the Social Security Administration, Cynthia Kendall of the Defense Department, John Lynn of the National Aeronautics and Space Administration and Henry Philcox of the Internal Revenue Service have opted out.
“A lot of the institutional knowledge is leaving the government and moving to the private sector,” said Linda Renfro, president of Lockheed Martin Information Support Services, a major information technology contractor at the Environmental Protection Agency and the Energy Department. “The learning curve is going to increase.”
The push to cut federal payrolls is accelerating the brain drain. One measure of the government’s lack of in-house knowledge is its reliance on outside contractors. Just 22 cents of every dollar spent on federal information technology goes toward government employee salaries. In private industry, by contrast, employee salaries account for an average of 50 cents of each dollar in costs.
Like many executives in the private sector, Renfro is concerned about the rush to reform. “We won’t know for two to four years whether what we have done will fix the problem,” she said. “I am not sure we understand the effects of everything we are doing.”
For even the best talent, the sheer size of the federal government makes technology reform exceptionally demanding.
“You can never make the case that government can outperform the private sector,” Kelman, the OMB executive, acknowledged. “It is unrealistic to expect as high an overall performance as we get from the average Fortune 500 company. The rosy scenario is that it is going to be somewhat lower.”
(BEGIN TEXT OF INFOBOX / INFOGRAPHIC)
About This Series
The Times is running a three-part series examining the long history of failure in the federal government’s information technology programs. Government operations generally depend on computers and software systems purchased in the 1960s that are badly antiquated and seriously undermine government efficiency.
SUNDAY: The $300-billion investment in information technology that has flopped, jeopardizing the nation’s welfare, eroding public safety and squandering untold billions of dollars. The problems extend to virtually every agency, court and the Congress.
MONDAY: The Internal Revenue Service’s troubled attempts to modernize its computer systems, which have led to the expenditure of $4 billion with marginal results so far in the agency’s ability to track down tax cheats. Another organization, the Defense Mapping Agency, spent a decade developing a $2.6-billion computer system that was obsolete by the time it was completed.
TODAY: The intensive effort to reform the system, which includes four key pieces of legislation enacted to streamline cumbersome federal regulations and raise the priority of information technology in the government.
(BEGIN TEXT OF INFOBOX / INFOGRAPHIC)
Legislating Technology
With little fanfare, Congress has undertaken major legislation in recent years to overhaul the federal government’s system for acquiring and managing information technology. The changes are aimed mainly at reducing regulations that in the past have added costs and delays to virtually every program to update computer technology. Though the reforms are wide-ranging, not all experts are convinced that they will solve the problem. The key pieces of legislation are:
[1] The Information Technology Management Reform Act, passes in 1996, eliminates many of the regulations that have guided computer purchases in the federal government during the last 30 years. The law revokes the Brooks Act, a 1960s law that made the General Services Administration the overseer of information technology for much of the federal government. The new law also requires that every large federal agency appoint a chief information officer at a senior level, and it creates a government-wide council of chief information officers.
[2] The Federal Acquisition Streamlining Act, passed in 1994, cuts red tape for purchasing small items, including personal computers and small software systems. It allows federal employees to use government credit cards for purchases of less than $100,000. Government buyers can now get the same prices as ordinary consumers, after years in which federal agencies paid more, despite their clout as big buyers.
[3] The Federal Acquisition Reform Act, passed in 1996, eliminates a wide range of regulations for big purchases. In the past, federal rules made fair competition and low prices the paramount goals, which blocked agencies from closely collaborating with suppliers before issuing a contract. The act essentially says the government went too far with fairness in competitions and frees bureaucrats from selecting only low bidders.
[4] The Government Performance and Results Act, passed in 1993, is a watershed law that for the first time requires federal agencies to measure their results from investments in information, technology, among other things. Historically, the agencies threw computers at many tasks with little evidence that the investments yielded a gain in efficiency or a reduction in operating costs. Often, the computer systems actually increased operating costs. Because the government lacks the management systems to measure productivity, the act gives agencies until 1999 to comply.
