Time and again, programs to modernize Defense Department record-keeping have fallen prey to bureaucratic rivalry, resistance to change and a lack of consequences for failure.
ALEXANDRIA, Virginia – The U.S. Air Force had great expectations for the Expeditionary Combat Support System when it launched the project in 2005. This accountants’ silver bullet, the Air Force predicted a year later, “will fundamentally revolutionize the way the Air Force provides logistics support.”
The new computer-based logistics technology would replace 420 obsolete, inefficient and largely incompatible “legacy” systems with a single, unified means of tracking the hardware of warfare. And it would be done for a mere $1.5 billion, combining three off-the-shelf products from Oracle Corp and modifying them only enough so that they could work together.
Seven years and $1.03 billion taxpayer dollars later, the Air Force announced in November 2012 that it was killing the project. ECSS had yielded “negligible” value and was “no longer a viable option,” the Air Force said. It would have taken an estimated $1.1 billion more to turn it into a system that could perform about one-quarter of its originally planned tasks, and couldn’t be fielded until 2020.
An August 28, 2013, report on the project, commissioned by an undersecretary of defense, filled in more of the blanks. The original promise of ECSS “was an exaggeration not founded on any true analysis,” it said. The plan was “ambiguous”; the Air Force failed to determine what ECSS would replace and what it would need to succeed.
That seven-year exercise in waste was not an anomaly. It was the norm for the U.S. Defense Department’s effort in recent years to upgrade the way it keeps track of money, supplies and people. Burdened with thousands of old, error-filled record-keeping systems – estimates range from 2,100 to more than 5,000 of them – the Pentagon is unable to account for itself, and thus for roughly half of all congressionally approved annual federal spending.
To fix that, the Defense Department has launched 20 or more projects to build modern business-management systems since the late 1990s. At least five were subsequently killed as complete failures after billions of dollars were spent on them. Nine projects now under way or already implemented carry an estimated total cost of $13.9 billion to build and operate, according to the Defense Department comptroller’s office. All of those in use can’t do everything they were supposed to do and are hooked to legacy systems they were supposed to replace.
The Defense Department inspector general said in a 2012 report that just six of these so-called Enterprise Resource Planning projects under way had racked up cost overruns of $8 billion and delays ranging from 1.5 to 12.5 years. With each failure, a pattern emerges: An off-the-shelf product with a proven track record in the private sector is chosen and then modified to the point where it doesn’t work properly.
“On every single one of the ERPs, they go out and customize the shit out of it to make it do what the legacy system did the same way the legacy system did it,” said Mike Young, a former Air Force logistics official and now a consultant on defense logistics and accounting.
Reuters has found that success is likely to remain elusive unless the Pentagon can change the way it goes about fixing its accounting problems. Interviews with scores of current and former defense officials, contractors and Pentagon watchers, as well as a review of dozens of reports by oversight agencies, show that the Pentagon is continually thwarted by a lack of accountability for failures, rivalry among and within various branches of the department, resistance to change, and an incentive to spend.
With its efforts to build reliable accounting systems in disarray, the Pentagon isn’t likely to meet a congressionally mandated 2017 deadline to be audit-ready. All other federal agencies are audited annually, in accordance with a 1990 law, and with rare exceptions, they pass every year. The Pentagon alone has never been audited, leaving roughly $8.5 trillion in taxpayer dollars unaccounted for since 1996, the first year it was supposed to be audited.
In previous installments of this series, Reuters has exposed the staggering costs and harmful effects of the Defense Department’s chronic accounting dysfunction. Persistent pay errors hound soldiers, sapping troop morale, while an impenetrable tangle of logistics and personnel systems can hinder commanders’ ability to know who and what are available for deployment. And the lack of reliable accounts – Pentagon staff routinely insert billions of dollars a year of false accounting entries to cover missing information – conceals huge sums lost to waste, fraud and mismanagement.
In response to questions about the Pentagon’s modernization efforts and the findings of this Reuters investigation, the office of Undersecretary of Defense Robert Hale, the Pentagon’s comptroller, emailed a written statement that said: “I note with disappointment that these articles misrepresent the efforts of a group of hard-working government workers who, despite furloughs and sequester and turmoil, have successfully provided financial services during two wars.”
The Pentagon has for years kept lousy books with impunity.
The 2009 law requiring the Defense Department to be audit-ready by 2017 provides for no penalties if it misses the deadline. Senators Tom Coburn, an Oklahoma Republican, and Joe Manchin, a West Virginia Democrat, introduced legislation earlier this year that would, among other things, limit new weapons programs, if the Pentagon misses the target. The bill has attracted co-sponsors, but otherwise has gone nowhere in the Senate.
From 1995 through 2002, Senator Charles Grassley pushed through an amendment to the annual defense appropriations bill requiring the Pentagon to account for its expenditures by following one seemingly simple procedure: match each payment to the expense it covered. The order was ignored, and Grassley gave up. “The goal was for the practice to become self-sustaining,” Grassley said in an email to Reuters. “It was wishful thinking.”
Rivalry and turf issues among and within each of the military services also thwart comprehensive fixes to the bookkeeping mess. Each branch has insisted on building from scratch its own systems for basic accounting, logistics and personnel, roughly tripling costs. The Army, Navy and Air Force also routinely disregard department-wide standards and rules imposed by the secretary of defense’s office in order to preserve their own ways of doing things.
That’s what happened to the Defense Integrated Human Resource System, which was intended to replace the scores of payroll and personnel systems that cause so many pay errors. Competing demands from military services ultimately rendered the system useless, and it was killed in 2010 after sucking up $1 billion.
High turnover takes a toll, too, as Pentagon personnel are shuffled into new jobs every few years, reinforcing what people involved in many projects said is a lack of personal investment in successful outcomes.
In 2009 – when the Air Force was four years into it ill-fated ECSS project – Jamie Morin became Air Force comptroller, succeeding John H. Gibson II, who now is vice president of the defense-support division of aircraft maker Beechcraft Corp. As the service’s top financial official, Morin would have been a primary user of ECSS. Less than a year after the project’s collapse, the Obama administration nominated Morin to head the Pentagon’s Office of Cost Assessment and Program Evaluation.
At Morin’s October 2013 confirmation hearing before the Senate Armed Services Committee, none of the members asked him about ECSS or another troubled Air Force modernization project under his watch, the Defense Enterprise Accounting and Management System. In a written response to questions from Reuters, Morin said: “In testimony and reports, we have indicated that the Air Force is on a well-designed, albeit aggressive path toward meeting … audit readiness goals.”
The Air Force replaced a brigadier general and a civilian executive after the ECSS project was killed, an Air Force spokeswoman said, without providing further details. But at the time, it blamed the failure mainly on primary contractor Computer Sciences Corp, saying the Falls Church, Virginia, company lacked the necessary capabilities.
David Scott Norton, an accounting systems specialist who worked for Computer Sciences Corp on the project, disputed that, saying the problem was high turnover. Contractor personnel “who talked to the client [the Air Force] in the beginning didn’t implement the system,” he said. Air Force personnel, too, were “always moving in and out. That just doesn’t work. You needed a dedicated team.”
The Pentagon’s inefficient method of pursuing efficiency has been on full display in the Army, which, among other efforts, has been building three separate new systems to handle accounting.
It launched the Logistics Modernization Program, or LMP, in 1998 with Computer Sciences Corp as the primary contractor. The Global Combat Support System – Army, or GCSS-A, began in 1997 and used Northrop Grumman as contractor. And the General Fund Enterprise Business System, meant to be the Army’s new central accounting system, began in 2005, using Accenture as contractor.
The three projects, each overseen by different agencies within the Army, with three different primary contractors, at different times bought licenses to use the same off-the-shelf software package, SAP’s Enterprise Resource Planning package. Each team then modified the software to create its own version to fit specific needs without making sure they worked together, people involved in the projects said.
In 2008, as work on all three projects was under way, the Army office that oversees acquisition of information systems issued a report, obtained by Reuters, faulting the Army for building a “fragmented portfolio of ERP systems that have developed along independent paths. … The Army cannot trace its business processes from factory to foxhole” without incurring additional “huge integration and operational costs.”
The report recommended that the Army halt work and consolidate the three systems. Doing so, it said, would save between 25% and 50% of the estimated $4.7 billion construction and operating costs of the three separate systems.
Backers of each project objected, according to people involved, and the report’s recommendations were ignored.
Kristyn Jones, deputy assistant secretary of the Army for financial information management, said that if the Army were to start over, “we probably would have chosen a different path.” She said one of the biggest challenges was getting workers to adapt to it. She and others involved in building GFEBS said many of the thousands of Army workers who would use the new system referred to their required training sessions as entering “the valley of despair.”
Gary Winkler, the Army program executive who oversaw acquisition of computer systems for the Army from 2007 through 2011, said developing the three projects separately, without coordination, hurt all of them. They “all go through these stovepipe approval processes without considering what is going on to the left or the right or behind,” he said.
As a result, the systems had to be linked through a costly network of pipelines – 282 such pipelines among the new and legacy systems, according to the Government Accountability Office, the investigative arm of Congress. “If you draw all of the connections, it looks like a bowl of spaghetti,” Winkler said. The systems must be continually tweaked because when one element requires a software update, for example, all the pipelines and other linked systems have to be updated, too.
When the Logistics Modernization Program, intended to streamline supply lines and better manage inventory, was eventually fielded in 2010, it was deficient in so many ways that the Army had to add an “Increment 2,” which won’t be ready until September 2016. This has increased the projected cost of building and operating the LMP to $4 billion from $2.6 billion, according to Army figures.
“The current system does not support certain critical requirements, including enabling the Army to generate auditable financial statements by fiscal year 2017,” the GAO said in a November 2013 report.
The second Army project, GCSS-A, had an original completion date of late 2015; late 2017 is now the target, for an estimated total building and operating cost of $4.2 billion, compared with the originally projected $3.9 billion.
Lastly, GFEBS, after an investment of more than $760 million, was fully fielded around the world in 2012, one year behind schedule. The Army touts it as a success, saying on its website: “The system is transforming the way the Army does business by enhancing the information available for leaders and managers across the Army.”
But this success is limited. GFEBS can’t track basic transactions – for example, payment of an electricity bill for an Army installation. To do that, it relies on legacy systems it was meant to replace. And because some legacy systems are unable to communicate with GFEBS, operators of those systems have had to revert to manually preparing spreadsheets to pass on data from thousands of Army posts.
NOT EXACTLY SEXY
Bookkeeping has never been a priority for the military. “They don’t train contracting officers or disbursement officials at West Point,” said a former senior Pentagon official who was involved in modernization efforts.
For years, winning the Cold War was the primary directive, with little consideration for cost. More recently, wars in Iraq and Afghanistan overshadowed concerns about rising defense spending, bad bookkeeping and attendant waste.
John Hamre became Defense Department comptroller in 1993, three years after Congress passed the law requiring that the Pentagon be audit-ready by 1996. He didn’t think upgrading accounting systems was vital to the Pentagon’s ability to fulfill its mission. “Would I like a better accounting system? Absolutely,” said Hamre, who left his post in 1997 and now is chairman of the Pentagon’s Defense Policy Board and chief executive of the Center for Strategic and International Studies, a bipartisan, nonprofit think tank in Washington, D.C. But, he said, “we’re getting military missions done every day. We just don’t use accounting for that.”
Many of the people interviewed for this series involved at all levels of the Pentagon’s accounting modernization program said that until recently, lack of interest or attention from the very top – from secretaries of defense and the civilian secretaries of the individual military services – has meant that no one steps in to impose order and consistency.
“You cannot just throw money at an ERP system and expect it to work unless somebody at the top says ‘You’re going to work together, and you’re going to get it done,’ ” said Norton, the former Computer Sciences Corp employee.
President Barack Obama’s three defense secretaries to date – Robert Gates, Leon Panetta and, now, Chuck Hagel – have spoken out strongly both publicly and inside the Defense Department about the importance of meeting audit-readiness deadlines. Panetta in 2011 set an even tighter deadline, ordering that a major portion of the department’s books be audit-ready by 2014. Officials have since acknowledged that the deadline won’t be met, and that the department plans to conduct an audit more limited in scope than what Panetta ordered.
The secretary of defense is empowered by law to order the military services to clean up their bookkeeping, adhere to Defense Department accounting rules and hew to common standards for building new accounting systems. But he doesn’t control the purse-strings – Congress does – so he “can’t say to the military services, ‘You can’t have the money’ ” if you don’t make this work, said Richard Loeb, a former Office of Management and Budget official and now an adjunct professor of government contracting and fiscal law at the University of Baltimore law school.
Congress has enabled the Pentagon’s institutional bias against change. “I think we do, all of us, bear some share of responsibility – myself included,” said Representative Robert E. Andrews, a Democrat from New Jersey on the House Armed Services Committee. “If I were to go home and start explaining that I was proposing a bill related to getting the Defense Department accounting systems to work right and the importance of accounting, they’d be asleep about half way through the first sentence.”
NUTS AND BOLTS
The Defense Logistics Agency, which buys, stores and distributes supplies for the U.S. military, has built a $2 billion-plus accounting system to make itself audit-ready.
It’s a failure.
In a March 2013 report, the Defense Department inspector general said the so-called Enterprise Business System was so compromised by fundamental errors in its construction that its data couldn’t be used to produce the standard financial statements required for an audit. Fixing the system to meet the Pentagon’s basic accounting requirements “would be cost prohibitive,” the report said. The DLA is using it anyway.
Fixing logistics is important – both to save money and to ensure that supplies are on hand when needed. The private sector realized that a long time ago. In recent decades, companies like Wal-Mart Stores Inc and Best Buy Co have raised supply-chain management to a precision science that saves huge sums.
By contrast, the Pentagon’s fragmented logistics systems “have contributed to longer lead times, excess inventory and stockpiling, duplicative activities and systems, inadequate performance measurements, and increased costs,” according to a 2011 report by the Defense Business Board, a group of business leaders that advises the secretary of defense’s office.
The board found that the department uses more than 1,000 separate logistics systems, and that in 2010, logistics cost $210 billion, or about 30% of that year’s defense budget.
Department-wide, duties are shared between the DLA, handling ordering and storage of supplies, and the U.S. Transportation Command, or Transcom, which handles delivery. Each has its own administration and computer systems, and each of the military services operates its own depot maintenance, supply, delivery and logistics accounting systems.
The Defense Business Board recommended combining all of the various logistics systems to “achieve significant budget savings, allowing the Department to preserve funds for force structure and the modernization of military capabilities.”
No action was taken on the board’s recommendations. “I have nothing for you on this,” William Urban, a spokesman for the secretary of defense’s office, said in an email response to questions about why the recommendations were not adopted.
Defense Secretary Hagel and other top Defense Department officials have argued that the impact of the budget sequester – automatic across-the-board spending cuts written into the 2011 congressional budget agreement – would be disastrous for the nation’s defense capabilities if allowed to continue. Congress heard: The latest budget deal would restore a big chunk of the cuts that would have occurred in 2014 and 2015.
But lack of reliable numbers on how the Pentagon spends the money it receives undercuts arguments for protecting higher spending levels. “You have the (military) service chiefs pissing and moaning on the Hill because they have to take the sequester,” said Franklin Spinney, a former senior analyst in the Pentagon who has written extensively on Pentagon weapons acquisition and spending priorities. But “they don’t have a clue what that’s going to cost.”
There is no doubt that bad bookkeeping conceals movements of money that in some instances are illegal. The Antideficiency Act of 1884 forbids anyone to commit U.S. funds to purposes not explicitly approved by Congress – a way to prevent federal officials from writing government checks for anything they want. The law includes civil penalties – officials responsible for violations can be demoted or dismissed – and also provides for criminal prosecution.
But because the Pentagon has never been audited, it is impossible to determine the frequency or extent of violations. In its annual Antideficiency Act Report for 2012, the GAO reported that the U.S. Special Operations Command illegally diverted more than $136 million over six years to pay for a helicopter development project. The Special Operations Command concluded that there had been “no willful or knowing intent on the part of the responsible parties” to break the law, though it issued a “letter of admonishment” to a senior civilian employee.
THE RIGHT WAY
The Pentagon isn’t incapable of fixing its broken business operations. For decades, the Mechanization of Contract Administration Services, or MOCAS (pronounced “moh-KAZZ”), which handles payment of nearly all of the Defense Department’s most complicated contracts, caused some of the department’s biggest headaches.
It made erroneous payments and spewed out inaccurate data. Much of this was because contract data, long strings of alphanumeric code, had to be entered by hand. Typos were rife. As much as 30 percent of the transactions it processed had to be redone, often by hand. The Defense Finance and Accounting Service, which operates MOCAS, said the system currently handles $3.4 trillion in active contracts.
In 1994, the Pentagon launched the Standard Procurement System to rationalize procurement by replacing MOCAS and at least 10 other contract payment systems. By 2002, the new system was years behind schedule, and the estimated cost of building and operating it had risen to $3.7 billion from $3 billion. When tested, it crashed. The Pentagon killed it.
Faced with having to rely on MOCAS indefinitely, the Defense Finance and Accounting Service decided to tweak it. Funding for an upgrade was limited, so DFAS relied heavily on its own workers, who knew MOCAS inside and out. It also hired as consultants retirees who had spent much of their careers working with the system.
MOCAS runs on COBOL, one of the earliest computer languages, and it lacks many of the options and preferences that make modern applications like Microsoft Word so easy to use. The MOCAS update team built an add-on to make MOCAS more user-friendly. It acquired another add-on that allowed electronic invoices and other data from contractors to flow directly into MOCAS. By eliminating paper invoices, they reduced the error rate to a tiny percentage.
The fixes worked so well that in 2008, the Defense Contract Management Agency, one of MOCAS’s main users, posted a “Happy 50th Birthday” wish to it on the agency’s website.
According to users, the 55-year-old system now handles tasks with lightning speed and an extremely low error rate.
But that success is hollow. The Pentagon remains saddled with other contract-payment systems. And MOCAS is not an accounting system; it must transmit data back to scores of accounting systems across the department for each transaction to be entered in ledgers. Many of the accounting systems are old and require manual data entry, so entry of transactions may lag behind MOCAS by a month.
Piecemeal fixes that don’t address overarching dysfunction reduce even further the chances the Pentagon will be audit-ready by 2017. Worse, said Charlie Murphy, a longtime member of Senator Grassley’s staff, the Defense Department will have added billions of dollars in new technology that failed to deliver. “Supposedly,” Murphy said, “there’s a maestro that’s going to make them all play like a symphony.”
(Edited by John Blanton)