DoD Battles IT Bureaucracy

By Elizabeth Wasserman

DoD Battles IT Bureaucracy



Focus On | e-Government
Business Goal | To become more responsive at less cost
Technology Issue | Hundreds of old, duplicative and incompatible legacy systems that work poorly, if at all
Management Challenge | Entrenched bureaucracy, changing budgets, political infighting, cultural resistance and revolving leadership
Results So Far | Mixed, or pending

The Pentagon's surface scars—its charred outer perimeter, its damaged roof, the gaping hole left by American Airlines Flight 77 when it slammed into the building on Sept. 11 at an estimated 400 mph—are no longer visible, though cranes are still positioned on the western front of this five-sided icon of American military might.

In the past year, rebuilding the Pentagon has become a symbol of the administration's resolve to win a new kind of war, with 21st-century weapons systems that rely more heavily on information technology than did the tanks and dumb bombs of the Cold War. A key part of the strategy is a battle being waged within the walls of the Pentagon, to topple organizational barriers that have long stood in the way of meaningful change: old technologies, incompatible systems and a Defense Department bureaucracy so entrenched that it can, at times, keep the Air Force from communicating with the Navy and the Army from sharing data with the Marines.

Ironically, it was on the afternoon of Sept. 10, 2001, less than 24 hours before the first planes hit the World Trade Center, that Defense Secretary Donald Rumsfeld first declared his war on the Pentagon's bureaucracy. At the time, Rumsfeld's initiative was viewed internally as yet another in a long line of peacetime initiatives waged over the years to cut costs, reduce staff and dole out lucrative technology contracts to favored firms in the defense establishment. "In this building," Rumsfeld told several hundred stone-faced Defense Department bureaucrats seated in the agency's auditorium, "despite this era of scarce resources taxed by mounting threats, money disappears into duplicative duties and bloated bureaucracy—not because of greed, but of gridlock." Some 673 different and uncoordinated financial systems inside the DoD, for example, had been making it impossible to track $2.3 trillion in financial transactions. "Innovation is stifled," Rumsfeld continued, "not by ill intent, but by institutional inertia." A new idea from an officer, he said, had to climb 17 levels of bureaucracy to get to the Defense Secretary's desk.

Then Rumsfeld, a former CEO of drugmaker G.D. Searle & Co. and of DoD contractor General Instrument Corp., said the agency's computer systems had become so complex and incompatible that information could not be shared from one floor of the Pentagon to the next, except in paper form. "The technology revolution has transformed organizations across the private sector," Rumsfeld told the group, "but not ours, not fully, not yet. We are, as they say, tangled in our anchor chain."

The mission now, says Rumsfeld: Use information technology not only to build smarter weaponry, but also to modernize and streamline front- and back-office operations to speed the response times of the Army, Navy, Air Force and Marines to new types of enemy threats. In the parlance of the military establishment, the Pentagon's new technology-heavy approach to war in the post–Sept. 11 world is known as C4IST, which stands for Command, Control, Communications, Computers and Intelligence Systems Technology—the heart of so-called "network-centric" warfare. The idea is to have real-time information about where everything is and where it's all going, even to know how much the enemy knows, thus permitting more efficient and innovative defense operations in wartime and in peace.

Funding is squarely behind it: All told, this seemingly monolithic agency spends nearly a billion tax dollars every day; its budget before Sept. 11, at $328 billion, was the largest among government agencies. It continues to be the largest, with $379 billion being sought for 2003. Further ahead, the budget is proposed to rise to $470 billion. Says Jay Farrar, a former Marine who is vice president for the Center for Strategic and International Studies, a Washington, D.C.-based think tank: "At the Pentagon, a lot of technology has been bought to upgrade weapons, but traditionally, not enough energy, attention and money has been put toward upgrading and streamlining the level of back-office technology. At the end of the day, the thinking had been: 'Do you put money out on the battlefield where you save lives, or do you put it in the office building to streamline a process?'"

Now, DoD is starting to do both. Even skeptical government watchdogs have taken note that Rumsfeld set aside $100 million in this year's DoD budget for financial modernization and a study aimed at reforming the Pentagon's IT architecture. At press time, President George W. Bush was proposing a 17 percent increase, or $53 billion, in federal IT spending. If approved in late September, the Department of Defense will get half, or $26.4 billion (see "Fighting Back"). "The events of Sept. 11 and the homeland security crisis have fundamentally changed how the government is thinking about its missions now," says Payton Smith, an analyst with Input, Inc., an IT market research firm in Chantilly, Va. "Sept. 11 gave the DoD and all federal agencies a new sense of urgency for making these things actually happen."

Meanwhile, the Pentagon's CIO, John Stenbit, a 30-year technology veteran and former defense systems expert at defense contractor TRW Corp., has put on hold troubled IT programs and has instituted strict new management controls, at one point requiring a progress review of the Navy Marine Corps Intranet before giving the project more funding. Further, new committees, councils and boards have been created to control spending and modernize business practices. Previously, there were few, if any, such agencywide checks and balances—or attempts at cross-agency business cooperation. "Power today," President Bush said months after the attacks, "is defined not by size anymore, but by mobility and swiftness." In that vein, the push to start transforming the DoD's creaky information backbone into a real-time tracking, coordination and support tool for a new-age military establishment and its operations—"is part of an entirely new approach to how the DoD now thinks about the rules of war," says Stenbit.

Technology Minefield

Technology Minefield

The challenge is formidable, and it's not certain whether Stenbit can pull it off: This is by no means the first attempt to torpedo the Pentagon's entrenched back-office operations. In the late 1980s, a modernization plan, called the Corporate Information Management Initiative, was supposed to replace 2,000 duplicative systems and save billions by streamlining operations. Instead, it wound up costing $20 billion and produced few results, according to David Walker, the comptroller general of the General Accounting Office, the watchdog arm of Congress.

The danger, of course—whether in e-government or private industry—is that by adding new technology, managers also run the risk of creating whole new layers of red tape. Further, trying to forge quick and fundamental change in a bureaucracy built by trillions of dollars over decades on millions of lines of outdated or misunderstood computer code is analogous to trying to make a U-turn in an aircraft carrier.

As in the private sector, politics, legions of rules and regulations and the need for multiple levels of project review—a tangle of red tape reflected in the scandals in the 1980s over $640 toilet seat covers and $400 hammers—have often doomed even the best of intentions, and have delayed and lengthened projects of any kind to the point of cost overruns and obsolescence. Legacy systems abound. The Defense Logistics Agency, the arm of the DoD that supplies military forces with food, clothing, fuel and $15 billion in annual supplies, has COBOL systems so old that components are no longer supported by the computer maker.

Further, cultural roadblocks can be daunting. The average 1.7-year tenure of the department's top political appointees hinders long-term planning and management accountability. IT operations have tended to be developed piecemeal, with a lack of agencywide performance measures and incentives for change, thanks to turf wars between the service branches.

Consider the Pentagon's Standard Procurement System, a $3.7 billion digital purchasing project that was supposed to replace 76 different legacy systems and automate the process by which the DoD's various branches and departments buy everything from weapons to shaving cream. When American Management Systems Inc. won the contract in 1997, the system was scheduled to be up and running by March 2000. But by September 2001, only about half of DoD's 43,000 users were online. Stenbit froze the project this summer. "People were installing technology without regard to deadlines and budgets," says Randy Hite, who heads up technology reviews for the GAO. "Government invests in technology to provide business value…but this wasn't happening." The GAO, in a report, added: "DoD's management of SPS is a lesson in how not to justify, make and monitor the implementation of information technology investment decisions."

To be sure, the Pentagon has a longstanding credibility problem when it comes to talk of reform. But even the agency's staunchest critics say Sept. 11 has, in some ways, put the agency onto a new course. "I think [DoD] learned something after Sept. 11. They have the momentum now. The stars are aligned, and change is much more likely to happen now than it would have been a year ago," says Lawrence Korb, a former assistant defense secretary during the Reagan administration and now a defense strategy analyst with the Council on Foreign Relations, a Washington, D.C.-based think tank. Still, Korb adds, change won't come quickly. He estimates the technology transformation could take at least a decade to complete.

Priming the Pipeline

Priming the Pipeline

CIO Stenbit's strategy is both huge and simple: Deploy an IT modernization plan for the Pentagon that reflects the struggles many companies in private industry face—the need to centralize financial controls and communications while keeping some decision-making decentralized. For starters, Stenbit wants to maximize the benefits of existing IT projects, minimize project delays and speed up IT acquisitions. He's created a Rapid Improvement Team to oversee a dozen pilot projects aimed at finding ways to cut, from years down to 18 months, the time it takes to get technology approved and put into the hands of the soldiers, sailors, airmen and technology workers who need it. Says Air Force CIO John Gilligan: "In wartime, we can cut through the bureaucracy without sacrificing quality and still provide rigorous oversight. But if something's not quite on that wartime priority list, it can take years to process the requirement, get the funding, do the contracting action and get a system installed."

Gilligan says the panel has already been making changes, such as giving the branches more autonomy to make decisions, streamlining the review processes, and posting periodic reviews of IT purchases and installations of intranets, instead of printing out scores of paper copies for individuals with oversight responsibility to circulate for weeks, if not months or years, at a time.

Also since Sept. 11, at the behest of Stenbit, each of the DoD's three armed service departments—the Army, the Air Force, and the Navy-Marine Corps—have been working to overhaul their IT systems, in some cases for the first time in more than a decade. Changes have been as simple as giving recruits an e-mail address before they get their first buzz cut, and as complex as making plans to digitize supply, weapons procurement and order systems. The Pentagon is also creating mammoth intranets, serving more personnel—1.4 million active duty men and women in uniform, another million in the National Guard and Reserves, and some 659,000 civilian employees—than any private enterprise in the U.S.

And to better manage supply chains, there are at least seven different enterprise resource planning systems being deployed to update technology that, in some cases, dates to the 1950s and, in at least one case, is based on 80-column punch cards. "Because of the pace of technology," Stenbit says, "I'm not a big believer in centralized planning of IT. I am much more of an impressionistic, van Gogh–like person believing that we need to let technology lead us, continually, into areas unique to the targets at hand."

Case in point: The Navy has four pilot projects under way. The Naval Sea Systems Command has awarded a $36 million contract to IBM Corp. to set up a regional maintenance system as part of a $347 pilot program. The Space and Naval Warfare Systems Command went live with its SAP system to track business operations last July under a $58 million program that will replace 35 legacy systems. The Naval Air Systems Command had a one-year, $9 million pilot with KPMG Consulting Inc. to manage its E-2C Hawkeye program, part of a $192 million ERP pilot program. The Army awarded a 10-year, $680 million contract to Computer Sciences Corp. to convert two legacy systems in the supply chain to ERP systems.

But Stenbit's approach hasn't come without criticism. Last November, the DoD's Inspector General criticized the plethora of ERP programs under development and raised concerns that the military services are developing at least $2 billion worth of financial management systems in a hodgepodge fashion before they know for certain that these systems will be compatible with the DoD's other standard finance systems. Stenbit's solution, for now: Regularly convene the various ERP program managers to make sure their systems can talk with each other. Meanwhile, efforts to gun down cultural silos are also under way. Says the GAO's Hite: "Without a department-wide plan, people try to do things to control just their little part of the world. [The DoD] needs to put in place a blueprint for the department as a whole."

Another big-ticket item on Stenbit's fix-it list is purchasing. Before the DoD can transform the armed forces, he says, it will have to change the way it buys weapons and supplies. The current system was developed to cope with the relative certainties of the Cold War, but it isn't flexible enough to meet the fast-changing demands of the more volatile new war on terrorism. Now, for example, when the supply officer aboard a U.S. Navy aircraft carrier at sea needs to order new parts for a plane, he or she must contact the mother ship in the fleet, which then communicates back to its shore-based command, which in turn must notify the Navy Supply Systems Command, the $6 billion-a-year organization that moves supplies and spare parts to keep the Navy's ships stocked. Each of these steps involves different, disparate technology systems—some involving COBOL systems.

To combat this, NAVSUP is testing a state-of-the-art SAP system that may one day manage all Navy supplies of ammunition, food and parts worldwide. Navy Commander Tom Gerstner, director of change management and training for NAVSUP, estimates there could be up to $4.5 billion in purchasing savings by 2007, simply by getting rid of some mainframes, reducing the time needed to reconcile disparate databases and cutting the cost of inventory.

Speed gains are also expected. During a recent test, Navy officials followed a spare part through the current system, tracking it through 35 separate physical moves and 29 different systems of paper logbooks before it was delivered. "In the new environment, we will have one computer system, and instead of taking 30 days to deliver a part, it should only take 15," Gerstner says.

By far, the DoD's biggest ERP project to reform purchasing is in the Defense Logistics Agency, the arm of the DoD that supplies everything from jeep tires to uniforms to helicopter spare parts. With more than $15 billion in sales annually, the DLA would rank around 125 on the Fortune 500—between Sun Microsystems Inc. and The Coca-Cola Co. David Falvey, the DLA's program executive officer, says DoD officials looked at a host of private companies, from Hewlett-Packard Co. to Dell Computer Corp. to Florida Power & Light, before deciding that ERP would be the cornerstone of an estimated $750 million business systems modernization program through 2005. "We need to have more material shipped directly from the supplier to the troops, to cut out the middle step and all the costs related to holding that inventory," Falvey said of the program, which went online in July.

Stenbit is also a big advocate of outsourcing—in sharp contrast to military tradition, where proprietary code and systems unique to the DoD were developed, partly for security reasons. Stenbit hopes one day to replace much of this with commercially developed systems and software. Not only would that save money, Stenbit says, but it would let the DoD keep better pace with new technologies. "The idea of the military industrial complex that Eisenhower talked about was government telling defense companies what to make," says Stephen Gale, director of the Organizational Dynamics program at the University of Pennsylvania. Now, says Gale, new technology available on the open markets—from smartcard technology to intranets—drive what the Pentagon does.

Small Victories

Small Victories

Mark Forman "In federal government today, we have at least 6,600 systems, because we have 6,600 forms that have to be put online. Somebody else told me it's really 26,000. So we may have 6,600 forms and 26,000 computer systems. That's a lot. We've only got 32 lines of business. How do you go from that to tens of thousands of systems, 33 million-plus Web pages and 23,000-plus Web sites? We've got a redundancy problem. There's just no doubt about it."
Mark Forman, White House director, IT and e-government; and director, federal CIO Council

Are any of Stenbit's efforts bearing fruit? Just ask Maj. Alan Dye, stationed some 1,500 miles from Washington, D.C., at Fort Hood, a U.S. Army base in central Texas. It used to take Dye weeks, if not months, to apply for a promotion: He'd have to write a letter to the Army personnel office, request a copy of his records, then wait days or weeks to review his data on microfiche file and reader. This time around, though, all Dye had to do was to log onto his PC, download his file from the Army's new $31.9 million intranet portal, Army Knowledge Online, and print it out. The whole process took seconds. The portal, a key component of the Army Knowledge Management project, is available to 1.2 million active and retired service personnel. It offers Web-based e-mail, news, banking information and some 8,000 links to classified and unclassified data, depending on one's log-in code. And the payoffs in savings so far have been enormous. The Army Personnel Command, for example, has been able to cut the number of network servers from more than 4,000 to just 43, saving the Army $28 million, says Cathy Michaliga, AKM director.

And that's not all. Now, servicemen and women are told via the Web when they need a dental checkup to qualify for deployment. Fast access to the health status of soldiers is a must for commanders in the event of a defense crisis: Each soldier must first have a clean bill of health. During the Gulf War, it could take up to a week for a commander to clear each recruit; AKO crunches that time to no more than one hour.

The Air Force is also developing its own intranet, consolidating hundreds of disparate legacy systems at 110 bases—from San Antonio to Okinawa—into one master Web access point. That enables pilots to, say, update the repair status of their jet fighters, check the latest weather forecasts and find directions to the local base commissary. The Navy and the Marines, meanwhile, are building an intranet to replace 200 different shore-based networks by 2008. Under a unique outsourcing contract for $6.9 billion awarded in October, the Navy Marine Corps Intranet will be owned and operated by Electronic Data Systems Corp. and will be leased back by the Navy in a way that requires it to pay for intranet service just as it does electricity.

The NMCI project, Stenbit says, is being closely watched by other government agencies as a potential model for government outsourcing. Says Navy Deputy CIO David Wennergren: "What I want to buy is performance rather than infrastructure." Even the GAO's Hite applauds the project's potential. "It's like a Sears Roebuck catalog," he says. "You order the service you need and pay as you go."

Down the road in Stenbit's push for what he calls a more "network-centric" organization is a plan—not unlike those in business everywhere—to get the Pentagon's back-office technologies to inform front-office strategies—in this case, the commanders and the generals. One of those projects involves using satellite communications to coordinate airborne intelligence and reconnaissance with troops on the ground or at sea. The system, according to Brig. Gen. Stephen Ferrell, the DoD's Space Architect, would use space lasers to zap information between commanders-in-chief and the different military services.

Wiring that up not only helps execute military strategy, it also helps keep track of who is where. Stenbit's Defense Integrated Military Human Resources System would replace 88 different personnel and payroll systems and consolidate the tracking of military personnel from basic training to retirement under one system. In Afghanistan, for example, each time troops were deployed, the joint task force commander had trouble keeping the troop records straight. Under each commander were a mix of people from the Marine Corps, Army, Navy and Air Force. But each of those branches had a different personnel system and needed different IT personnel to coordinate records for payroll and in case of injury. Other agencywide projects include computer systems that secure classified communications, book travel and provide reimbursement.

Can it all work? It's a lot to keep track of, and IT—whether in the Pentagon or throughout government—suffers a credibility problem on Capitol Hill: Some estimate the number of redundant computer systems in operation by the federal government, as a whole, at 6,600; others estimate closer to 26,000, according to Mark Forman, the Bush Administration's IT czar. There's also a stormy history of federal IT projects gone awry. The "poster child" of wasted IT effort: the Internal Revenue Service's Tax Systems Modernization program, which cost more than $3 billion before being killed in 1997.

But one thing has become clear in the past 12 months, at least at the Pentagon: Fixing IT systems isn't just a low-priority peacetime mantra left over from the closing days of the Cold War, when the enemy was more predictable and low-tech, and the battleground, geographically isolated. Now the threat is digital as well as land-based, triggering the need for fundamental change.

In a speech to students at National Defense University in January, Rumsfeld told students worried about another terrorist attack on American soil that all the high-tech weapons in the world won't transform the armed services "unless we also transform the way we think, the way we train, the way we exercise and the way we fight." And, he might have added, the way the Pentagon does business.

Elizabeth Wasserman is a Washington, D.C.-based writer. Formerly, she was Washington Bureau Chief for The Industry Standard.

This article was originally published on 08-19-2002