A strange thing happened when Christopher DeLara filed for disability benefits after his tour in Iraq: The U.S. Army said it had no records showing he had ever been overseas.
DeLara had searing memories of his combat experiences. A friend bled to death before his eyes.
He saw an insurgent shoot his commander in the head.
And, most hauntingly, he recalled firing at an Iraqi boy who had attacked his convoy.
The Army said it could find no field records documenting any of these incidents.
DeLara appealed, fighting for five years before a judge accepted the testimony of an officer in his unit. By then he had divorced, was briefly homeless and had sought solace in drugs and alcohol.
DeLara's case is part of a much larger problem that has plagued the U.S. military since the 1990 Gulf War: a failure to create and maintain the types of field records that have documented American conflicts since the Revolutionary War.
A joint investigation by ProPublica and The Seattle Times has found that the recordkeeping breakdown was especially acute in the early years of the Iraq war, when insurgents deployed improvised bombs with devastating effects on U.S. soldiers.
The military has also lost or destroyed records from Afghanistan, according to officials and previously undisclosed documents.
The loss of field records — after-action write-ups, intelligence reports and other day-to-day accounts from the war zones — has far-reaching implications.
It has complicated efforts by soldiers like DeLara to claim benefits. And it makes it harder for military strategists to learn the lessons from Iraq and Afghanistan, two of the nation's most protracted wars.
Military officers and historians say field records provide the granular details that, when woven together, tell larger stories hidden from participants in the day-to-day confusion of combat.
The Army says it has taken steps to improve handling of records — including better training and more emphasis from top commanders. But officials familiar with the problem said the missing material may never be retrieved.
"I can't even start to describe the dimensions of the problem," said Conrad C. Crane, director of the U.S. Army's Military History Institute. "I fear we're never really going to know clearly what happened in Iraq and Afghanistan because we don't have the records."
The Army, with its dominant presence in both theaters, has the biggest deficiencies. But the U.S. Central Command in Iraq (Centcom), which had overall authority, also lost records, according to reports and other documents obtained by ProPublica under the Freedom of Information Act.
In Baghdad, Centcom and the Army disagreed about which was responsible for keeping records. There was confusion about whether classified field records could be transported back to the units' headquarters in the United States. As a result, some units were instructed to erase computer hard drives when they rotated home, destroying the records that had been stored on them.
Through 2008, dozens of Army units deployed in Iraq and Afghanistan either had no field records or lacked sufficient reports for a unit history, according to Army summaries obtained by ProPublica. DeLara's outfit, the 1st Cavalry Division, was among the units lacking adequate records during his 2004 to 2005 deployment.
Recordkeeping was so poor in Afghanistan from 2004 to 2007 that "very few Operation ENDURING FREEDOM records were saved anywhere, either for historians' use, or for the services' documentary needs for unit heritage, or for the increasing challenge with documenting Post Traumatic Stress Disorder (PTSD)," according to an Army report from 2009.
Entire brigades deployed from 2003 to 2008 could not produce any field records, documents from the U.S. Army Center of Military History show.
The Pentagon was put on notice as early as 2005 that Army units weren't turning in records for storage to a central computer system created after a similar recordkeeping debacle in the 1990-91 Gulf War.
In that war, a lack of field records forced the Army to spend years and millions of dollars to reconstruct the locations of troops who may have been exposed to toxic plumes that were among the suspected causes of Gulf War Syndrome.
At the outset of the Iraq war, military commanders tried to avoid repeating that mistake, ordering units to preserve all historical records.
But the Army botched the job. Despite new guidelines issued in 2008 to safeguard records, some units still purged them. The next summer, the Washington National Guard's 81st Brigade Combat Team in Iraq was ordered to erase hard drives before leaving them for replacement troops to use, said a Guard spokesman, Capt. Keith Kosik.
Historians had complained about lax recordkeeping for years with little result.
"We were just on our knees begging for the Army to do something about it," said Dr. Reina Pennington, a Professor at Norwich University in Vermont who chaired the Army's Historical Advisory Committee. "It's the kind of thing that everyone nods about and agrees it's a problem but doesn't do anything about."
Critical reports from Pennington's committee went up to three different secretaries of the Army, including John McHugh, the current secretary. McHugh's office did not respond to interview requests. His predecessor, Peter Geren, said he was never told about the extent of the problem.
"I'm disappointed I didn't know about it," Geren said.
In an initial response to questions from ProPublica and the Times, the Army did not acknowledge that any field reports had been lost or destroyed. In a subsequent email, Maj. Christopher Kasker, an Army spokesman, said, "The matter of records management is of great concern to the Army; it is an issue we have acknowledged and are working to correct and improve."
Missing field records aren't necessarily an obstacle for benefit claims. The Department of Veterans Affairs also looks for medical and personnel records, which can be enough. The VA has also relaxed rules for proving post-traumatic stress to reduce the need for the detailed documentation of field reports.
But even the VA concedes that unit records are helpful. And assembling a disability case from witness statements can take much more time, said Gen. Peter Chiarelli, the retired Army vice chief of staff who worked to combat suicides and improve treatment of soldiers with PTSD and brain injuries.
"You would always love to have that operational record available to document an explosion, but there are other ways," Chiarelli said. "You can provide witness statements from others who were in that explosion. But it's going to be more difficult."
After reviewing findings of the ProPublica-Times investigation, Sen. Patty Murray, D-Wash., who chairs the Senate Committee on Veterans' Affairs, asked Defense Secretary Leon Panetta to report on efforts to find and collect field records.
"Iraq and Afghanistan veterans who are unable to document the location and functions of their military units could face the same type of problems experienced by Cold War veterans exposed to radiation, Vietnam era veterans exposed to herbicides and Gulf War veterans exposed to various environmental hazards," Murray said in a statement.
Already, thousands of veterans have reported respiratory problems and other health effects after exposure to toxic fumes from huge burn pits that were commonly used to dispose of garbage in Iraq and Afghanistan.
DeLara remains embittered about the five years he spent waiting for his disability claim. In an interview at his home in Tennessee, he pointed to Army discharge papers showing he'd received the Global War on Terrorism Expeditionary Medal, awarded for service in Iraq and Afghanistan.
Next to that were blank spaces where his deployment dates should have been.
"If they'd had the records in the first place, and all the after-action reports," DeLara said, "this never would have stretched on as long as it did."
A Desperate Search for Records
The Army is required to produce records of its actions in war. Today, most units keep them on computers, and a 4,000-soldier brigade can churn out impressive volumes — roughly 500 gigabytes in a yearlong tour, or the digital equivalent of 445 books, each 200 pages long.
Field records include reports about fighting, casualties, intelligence activities, prisoners, battle damage and more, complete with pictures and maps. They do not include personnel or medical records, which are kept separately, or "sigact" reports — short daily dispatches on significant activities, some of which were provided to news organizations by WikiLeaks in 2010.
By mid-2007, amid alarms from historians that combat units weren't turning in records after their deployments, the Army launched an effort to collect and inventory what it could find.
Army historians were dispatched on a base-by-base search worldwide. A summary of their findings shows that at least 15 brigades serving in the Iraq war at various times from 2003 to 2008 had no records on hand. The same was true for at least five brigades deployed to Afghanistan.
Records were so scarce for another 62 units that served in Iraq and 10 in Afghanistan that they were written up as "some records, but not enough to write an adequate Army history." This group included most of the units deployed during the first four years of the Afghanistan war.
The outreach effort by the Army was highly unusual. "We were sending people to where they were being demobilized," said Robert J. Dalessandro, executive director at the Army's Center of Military History. "We even said ... 'Look we'll come to you' — that's how desperate we got."
As word of missing records circulated, the Joint Chiefs of Staff became worried enough to order a top-level delegation of records managers from each service branch to Baghdad in April 2010 for an inspection that included recordkeeping by U.S. Central Command.
Centcom coordinated action among service branches in the theater. Among other things, Centcom's records included Pentagon orders, joint-service actions, fratricide investigations and intelligence reviews, with some records from Army units occasionally captured in the mix.
After five days, the team concluded that the "volume, location, size and format of USF-1 records was unknown," referring to the acronym for combined Iraq forces. The team's report to the chiefs cited "large gaps in records collections ... the failure to capture significant operational and historical" materials and a "poorly managed" effort to preserve records that were on hand.
In a separate, more detailed memo, two of the team's members from the National Archives and Records Administration went further.
"With the exception of the Army Corps of Engineers, none of the offices visited have responsibly managed their records," they wrote. "Staff reported knowledge of only the recently created and filed records and knew little of the records created prior to their deployments, including email. ... It is unclear the extent to which records exist prior to 2006."
Part of the problem was disagreement and lack of coordination about who was responsible for certain records, including investigations into casualties and accidents, according to Michael Carlson, one of the two archivists.
"The Army would say it's Centcom's responsibility to capture after-action reports because it's a Centcom-led operation. Centcom would say it's an Army responsibility because they created their own records," Carlson said in an interview. "So there's finger-pointing ... and thus records are lost."
Nearly a year after the U.S. pullout from Iraq, Centcom said it still is trying to index 47 terabytes of records for storage, or some 54 million pages of documents. It's not clear if those include anything recovered after a 2008 computer crash the Baghdad team termed "catastrophic."
Lt. Col. Donald Walker, an Air Force officer who took over as Centcom records manager in 2009, acknowledged that there was confusion about responsibility and confirmed that that some Centcom records may have been lost. In part, he blamed computer problems and the competing demands of wartime.
"Something just had to fall off the plate, there was so much going on," said Walker, who worked out of Centcom's Tampa, Fla., headquarters but was among the Baghdad inspectors.
Rather than risk letting classified information fall into the wrong hands, some commanders appeared to buck the orders to preserve records. One Army presentation asserts that in 2005, V Corps, which oversaw all Army units then in Iraq, ordered units to wipe hard drives clean or physically destroy them before redeploying to the States.
"They did not maintain the electronic files. They just purged the servers," according to the Military History Institute's Crane, who said he heard similar accounts from more than a dozen veteran officers in classes at the Army War College.
The orders directing Washington National Guard's 81st Brigade to erase hard drives before leaving Iraq came "from on high," according to unit spokesman Kosik, who said he confirmed the erasures with a senior Guard officer with first-hand knowledge. He said the orders came from outside the Washington Guard.
"There was a lot of confidential information, and they were not allowed to take it out of theater," said Kosik. "All that was wiped clean before they came home. ... It was part of their 'to-do' list before leaving country."
Steven A. Raho III, the Army's top records manager, said in an interview that he couldn't estimate what, if any, records might be missing. But Raho said his agency wasn't responsible for collecting records, only for storing them in the Army's central records system when individual units handed them over.
Units are not required to do so, he emphasized. "All's I know is we have some and units have some," Raho said.
As a test, ProPublica filed Freedom of Information Act (FOIA) requests for a month's worth of field records from four units deployed in Iraq in 2003 and 2004. The requests went to Raho's Records Management and Declassification Agency, which forwarded them to each unit.
One brigade — the 2nd Combat Brigade of the 82nd Airborne Division — did not respond, but FOIA officers from the three others said they searched and could find no responsive records.
"I don't know where any Iraq operational records are," said Daniel C. Smith, a privacy act officer at Fort Carson, Colo., who handled the request for the 2nd Brigade of the 2nd Infantry Division. "I've never been able to find out where they went."
At Fort Riley, Kan., FOIA officer Tuanna Jeffery looked for records from the 1st Battalion, 41st Infantry Regiment, 1st Armored Division. "Prior to and upon the inactivation of the unit on March 15, 2008, that unit had turned in absolutely no records," she responded.
In a follow-up email, Jeffery said the entire 1st Armored Division did not turn in any field records through 2008.
'They Couldn't Find It'
Chris DeLara is not the type of soldier to wear his heart on his sleeve, but the 1st Cavalry Division's shoulder patch is tattooed on his right forearm in a swirling piece of body art. Beneath it are the words: "Baghdad, Iraq."
DeLara, 38, grew up in Albany, N.Y., never dreaming he might someday fight a war. Now, his tour in 2004 and 2005 haunts his every day. Since winning his appeal in March 2011, he is classified as fully disabled by post-traumatic stress and cannot work. He was awarded a stipend of about $30,000 a year and has moved near Knoxville, Tenn., where he recently bought a modest house.
Getting to a stable point wasn't easy.
DeLara was an administrative specialist, essentially a personnel clerk. But he was repeatedly pulled out of his scrivener's life for missions as a roof gunner on convoys. It was a time of insurgency and exploding factional violence in Baghdad.
"They told us, 'This may be your job, but guess what? You're going to be doing everything,'" he said. "We had many hats. You go to combat, your job is secondary. Combat is first."
DeLara did not want to discuss his combat experiences, but they are described in part by a judge in the Board of Veterans' Appeals ruling that approved his PTSD claim.
In the years after his deployment, DeLara told psychiatrists and others who treated him at various times that two of his friends were killed in an insurgent attack on his convoy, and that he was unable to stop one of them from bleeding to death from a ruptured artery.
He said that one his commanders was shot in the head in front of him by insurgents, and reported that he had killed an Iraqi youth who had tried to attack his convoy after it was stopped because of a roadside bomb, according to the judge's summary.
After his return in 2005, DeLara was diagnosed several times with PTSD or its symptoms, according to VA exam records cited by the appeals judge. He drank and used drugs even though he'd abstained from them in the Army. In 2006, he overdosed on prescription drugs.
DeLara said he lived for a time in a shelter for troubled vets. He and his wife eventually divorced, but he credits her for helping him fight for his claim when he might have given up.
They first applied for a PTSD benefit in 2006, DeLara said. A denial came the next year because his separation document, called a DD-214, did not list any dates of overseas deployment, he said.
"They couldn't find it. Well my ex-wife, she being as persistent as she is, we started pulling all the stuff" to send to the VA, he said. DeLara dug out the movement order sending his unit to Iraq and the brigade roster with his name on it. He added descriptions of his combat experiences. "Basically what it was, I needed to provide proof," he said.
But he was denied again, this time because the VA said his symptoms were of bipolar disorder, not PTSD. DeLara said he appealed but got a letter saying there was insufficient evidence that he'd experienced combat stress. The VA told him that it had "no records, none whatsoever" of his time in combat, DeLara said.
"We basically put the whole packet together from scratch again," DeLara said. This time, he tracked down his former company commander, who was incensed about the VA denials and provided a letter confirming an incident in which DeLara came under enemy fire. Still, two years went by before DeLara received word that his appeal was set for a hearing in January 2011.
Although the judge found in his favor, the ruling notes that, in June 2008, the center responsible for locating his records "made a formal finding of a lack of information to corroborate a stressor for service connection for PTSD." The center even looked a second time but still came up empty-handed.
DeLara said he still can't believe it. "I had dates and everything" in the supporting material he and his ex-wife sent to the VA, he said. "The simple fact is that nobody filled out after-action reports," DeLara said. "There was no record of it."
Asked how often a search for unit records comes up empty, officials at the VA said they didn't know — the agency doesn't track that statistic. A VA spokesperson said missing field records are not a major factor delaying veterans' claims, however. And some veterans' advocates agree.
"As long as an officer or a buddy who witnessed the event is willing to sign a notarized statement, that's good," said John Waterbrook, who advises vets on disability issues in Walla Walla, Wash.
In 2009, as DeLara was refiling his case, veterans' groups complained to Congress that soldiers serving as clerks or mechanics unfairly faced a higher burden of proof for PTSD than those with an obvious combat role, even though they faced the same dangers in wars with no front lines.
The VA relaxed its rules the next year, so that a vet's account of combat stress is proof enough if a VA medical examiner agrees. But while the change helps, it hasn't sped up claims or made field records less valuable, said Richard Dumancas, the American Legion's deputy director of claims.
Field records can come into play for other injuries. Take the case of Chief Warrant Officer 3 Lorenzo Campbell, a 53-year-old soldier with the Washington Guard who filed a disability claim resulting from a 2004 injury in Iraq.
During a rocket attack, Campbell banged his knee on a concrete bumper after jumping out of a Humvee to find cover. He saw a doctor, but there was no record in his medical files. His knee gradually deteriorated, and he now wears a brace and is unable to run.
Campbell said he tried to get records of the rocket attack from the state Guard but was told they were classified and left on computers in Iraq. He said he offered a letter from another soldier testifying to the incident and swore out a statement himself, but it didn't suffice.
"I tried to keep fighting it," he said. "They kept writing me saying they need more information, they need more information."
Campbell said his disability claim took four years to be approved — a delay that could have been shortened had the records been available. "If you have no records," he said, "you can be fighting for five or six years and still not prevail."
Tradition Eroded, Warnings Brushed Aside
Military recordkeeping has been the cornerstone of the nation's war history for centuries. From the founding of the republic through the Vietnam War, recordkeeping was a disciplined part of military life, one that ensured that detailed accounts of the fighting were available to historians and veterans alike.
The records can hold untold stories that can surface decades after a conflict.
The massacre of civilians by U.S. forces at No Gun Ri, South Korea, in July 1950 came to full national attention only in 1999, nearly 50 years after the fact. Journalists at The Associated Press, working in part with military field records, uncovered the extent of the tragedy. Later, other reporters used the records to show that one purported witness wasn't really present.
By the Gulf War, however, what had been a long tradition of keeping accurate, comprehensive field records had begun to erode. Old-style paper recordkeeping was giving way to computers. And Army clerks had been reduced in number, leaving officers to take care of records work.
According to the Army's "Commander's Guide to Operational Records and Data Collection," published in 2009, the problem became evident months after the end of Desert Storm, when vets began reporting fatigue, skin disease, weight loss and other unexplained health conditions.
"When the Army began investigating this rash of symptoms, its first thought was to try and establish a pattern of those affected: What units were they in? Where were they located? What operations were they engaged in?" the guide says. "The answers provided by investigators were: 'We don't know. We didn't keep our records.'"
Afterward, the Army created Raho's records agency and a central records system. As the war on terror began, however, inspections and penalties for recordkeeping at the command level had largely fallen by the wayside, according to Army documents and interviews with officers who helped search for Gulf War records.
Robert Wright, a retired Army historian, said training broke down. "They fight as they train, and they never were trained," he said.
On March 28, 2003, then-Deputy Secretary of Defense Paul Wolfowitz ordered retention of all records in the Iraq war. Military records, he wrote, "are of enduring significance for U.S. and world history and have been indispensable for rendering complete, accurate and objective accountings of the government's activities to the American people."
But in the combat zones, there were other priorities.
Kelly Howard served as operations officer to Army Gen. George W. Casey Jr., who was in charge of the Iraq war from 2004 to 2007. Her primary job was archiving Casey's papers, a task that had been ignored until her arrival in 2006. Casey stored them in a foot locker, among other places.
"The reason so many things got lost ... is because so many people at higher levels weren't requiring it," Howard said, referring to systematic recordkeeping. "You do what your boss wants you to do. It's not that anyone said, 'No, I don't care about that.' It's just so many other things were important."
Alarms mounted at about the same time as DeLara finished his Baghdad tour.
In 2005, the Army's Historical Advisory Committee learned that Raho's agency had not "received any records from units deployed in Afghanistan & Iraq."
This came as a shock. Members of the group include a mix of civilian historians and officials from the Army War College and Center of Military History.
"So we go through the whole meeting," said Richard Davis, senior historian at the National Museum of the U.S. Army. "So I ask the records manager point blank. I said, 'How many records have been retired from overseas by U.S. Army units?' And the answer was zero.
"By late October the records management people here in Washington had received not a single document from Afghanistan or Iraq," Davis said. "At that point all the historians looked at each other and said 'Holy shit! '"
Minutes from the committee's 2006 meeting quote Raho as saying, "Our problems are that the training for Army personnel is incomplete, the responses are uneven, and the records themselves are either incomplete or nonexistent."
Another member suggested writing a book. "As an institutional history, I think it's a great idea," responded historian Pennington, then the committee's chairwoman. "'Losing History': It's a topic that merits visibility and study."
The committee included regular warnings about a broken recordkeeping system in its annual reports to the secretary of the Army.
The 2006 report to Secretary Francis J. Harvey said Raho had described "major problems" in records collection, including "the lack of centralized control of data collection, the destruction of records without evaluation, and inadequate communications between Army units and records collection personnel."
Raho, the report said, "observed that 17 to 23 percent of all Iraq/Afghanistan War veterans will suffer from various forms of PTSD. ... Without strong and immediate action to remedy present shortcomings, the Army's ability to substantiate veteran disability claims will be degraded seriously, with potentially highly troublesome and expensive consequences."
In its 2008 report, the committee said: "Units are losing their own history. This will create a snowball effect, resulting in problems with awards and heritage activities in the future."
Pennington signed the report, adding a personal comment: "After six years of service on DAHAC, and now as its chair, I am frankly discouraged by the frequency with which DAHAC has expressed some of the same concerns, and how little progress has been made on some issues."
Then-Secretary Geren's office responded with a thank-you letter under his signature. But Geren said in an interview that he was not personally informed about missing records, despite his March 31, 2009, letter. "I'm confident it was not brought to my attention."
When McHugh, the current secretary, arrived in 2009, he received a committee report reiterating that the system was broken and pleading for resources to fix it. "This has been requested every year since 1997," the report said.
"It's probably the most serious problem historians have ever had," Pennington said in an interview. "I honestly don't know how we're going to be writing records-based history in 20 to 30 years." Typically, field records remain classified for two to three decades after a war, then are transferred to the National Archives.
Although committee members felt unheard, wheels had slowly begun moving in the Army. In 2007, Raho's agency and the Center of Military History launched the outreach project that discovered the historians were right: Scores of units did not have the records they should.
Because Raho did not have enough staff, the Center of Military History provided detachments for the search. For more than two years they collected field reports, turning up about 5.5 terabytes' worth.
Some additional records have dribbled in since: Dalessandro, the center's director, said one brigade of the 1st Armored Division handed over field records from its 2007 Iraq deployment. It's possible that more might be found from other units, but historians say the chances fade with each year.
Burn Pits: The New Agent Orange?
The demand for the field records isn't likely to abate as members of Congress ratchet up pressure to investigate exposure to burn pits.
Veterans' groups say the long-term health impacts could be similar to those of herbicides in Vietnam. Rep. Michael Michaud of Maine, ranking Democrat on the House Veterans' Affairs Subcommittee on Health, said missing field records "could have consequences for veterans for years to come."
In September, the House passed the Open Burn Pit Registry Act to track veterans with symptoms and find out where they were exposed and for how long. A similar measure is pending in the Senate. The VA currently runs registries for Agent Orange and Gulf War Syndrome, and last year the Institute of Medicine said more research is needed.
Some veterans' advocates say field records could provide critical.
"It's going to be very hard to connect individuals without the field records," said Dan Sullivan, director of the Sgt. Thomas Sullivan Center, a nonprofit named after his brother, an Iraq vet who died from mysterious health complications.
"It would strike me that they are very important."
Versions of this story will be published by The Seattle Times and Stars and Stripes.
Peter Sleeth is a veteran investigative reporter who covered the Iraq war for The Oregonian and helped the paper win a Pulitzer Prize in 2007 for breaking news. Now freelancing, his most recent piece for the Oregon Historical Quarterly is a profile of progressive-era activist Tom Burns.
Hal Bernton has been a staff reporter for The Seattle Times since 2000. He has covered military and veterans affairs, reporting from Iraq in 2003 and from Afghanistan in 2009 and this fall. Among other things, Bernton has reported on veterans' health issues, post-traumatic stress and, recently, improvised explosive devices.
Please follow Military & Defense on Twitter and Facebook.