MELISSA BLOCK, HOST:
This is ALL THING CONSIDERED from NPR News. I'm Melissa Block.
AUDIE CORNISH, HOST:
And I'm Audie Cornish. Defense Secretary Chuck Hagel has ordered the Pentagon to revamp its procedures for finding and identifying the bodies of lost service members. That's tens of thousands of Americans who died in World War II, Korea or Vietnam, but whose bodies were never returned. Hagel ordered a team to examine the agencies that handle that mission. And this next report is about what that team will find.
For months NPR has investigated the process in partnership with the independent news organization ProPublica. NPR's Kelly McEvers found the agencies to be slow and inefficient and using methods that many scientists say are outdated.
KELLY MCEVERS, BYLINE: To understand how these agencies work we focused on the lead agency, the Joint POW/MIA Accounting Command, otherwise known as JPAC. It's housed in a collection of ten low-slung buildings on a U.S. military base in Hawaii. JPAC's mission is this: find, identify and return servicemen whose bodies never made it home from war. And sometimes their process is straight forward.
First, says supervising anthropologist Denise To, a team of scientists travels to a site to look for remains. A jungle in Vietnam or Cambodia, a battlefield in Korea or...
DENISE TO: ...a beautiful area in Belgium, absolutely stunning rolling green hills. But there's war going on and a plane crash there. And a lot of individuals lost their lives there.
MCEVERS: Then whatever's found at the site is brought back to JPAC. The non-biological stuff comes to this lab. Pieces of watches. Is that a lighter? That looks like some kind of jewel piece, piece of a bracelet, another buckle.
The biological stuff, the actual remains of the soldier go to the main lab downstairs. So this is the main JPAC lab. And just to describe it what we're looking at is basically just like a series of tables, one, two - what, 15 tables, yeah? And what's laying out on each table are remains, mostly bones. Some of them look to me, a layperson, like almost a complete skeleton. Others, this one that's right in front of us, it's fragments, pieces. In some places almost like dust. We walk up to one of the tables. Forensic anthropologist Marin Pilloud is getting to work on a case. So you're opening a box.
MARIN PILLOUD: Yes. This is a mummified foot. These were found in ice, so they're pretty well preserved.
MCEVERS: The case is a plane crash from the time of the Korean War. The remains of 19 servicemen were found in a glacier.
PILLOUD: Actually I have several guys in soft tissue just over there.
MCEVERS: This is one of the straightforward cases. The serial number on the plane led JPAC to a roster of who was on it. Still, Pilloud says it's taken her a year and a half just to sort the remains, test them for DNA and get those test results back.
PILLOUD: So right now they're separated by bones that I was able to refit back together. And then these tags all represent different DNA samples.
MCEVERS: It gets even more complicated if it's something like a jungle battlefield in the South Pacific where hundreds of marines might've died on one beach. Tom Holland is the director of the main lab at JPAC. He personally signs off on every identification. He describes the process as a funnel. First, you make a list of all the people who could've been on that beach. Then, you narrow it down.
TOM HOLLAND: It may be 100 people, 200 people, 500 people. That's fine. That's how you start the process. But then you use artifacts, you use the archaeological analysis, you use the anthropology, you use the DNA. And the idea is that you get down to one individual.
MCEVERS: The thing is, JPAC ID'd 60 people last year, even though Congress has mandated them to start ID'ing 200 each year by 2015. We calculated the U.S. spends $1.3 million on average for every ID. Critics say that's just too slow and too costly. So what's the problem? NPR and ProPublica spent the last five months exploring that question from the lab in Hawaii to battlefields in the Philippines to the halls of the Pentagon to frustrated families in Texas, Illinois and Maine.
What we found was a disjointed system handled by a jumble of disconnected federal agencies, memos on a single subject that went back and forth for years, straightforward lab tests that took months to complete and scientists who use an outdated approach that other missing persons labs abandoned decades ago. So let's go back to that funnel again, that narrowing-down process lab director Tom Holland described.
HOLLAND: It may be 100 people, 200 people, 500 people. That's fine. The idea is that you get down to one individual.
MCEVERS: Our reporting found JPAC uses historical records and forensic anthropology, basically measuring bones and creating biological profiles, to narrow it down. Then they use DNA to confirm their findings. That puts JPAC at odds with prominent missing persons labs around the world. They use a DNA-lead system that works like this.
ED HUFFINE: You DNA test all the bodies, you DNA test all the family references. And DNA says this particular body is a member of this family.
MCEVERS: That's biochemist Ed Huffine. He used to work at the Army's DNA lab. In 1999 he went to Tuzla, Bosnia, to help identify 4,000 sets of remains that had been found in dozens of mass graves after the Srebrenica Massacre four years before. He found they were using the classic anthropology approach, measuring bones and teeth, trying to match them against records.
HUFFINE: Only seven out of more than 4,000 had been able to be identified.
MCEVERS: So Huffine and others developed the DNA-led method. That means they took the DNA of the missing and put it in a database, then went to the families of the missing and put their DNA in another database. Then they compared the two databases until they found matches.
HUFFINE: I would go to family meetings and explain, here's what DNA can do. Here's how it can potentially identify missing loved ones.
MCEVERS: Families gave samples. The number of matches skyrocketed.
HUFFINE: Within a several-year period, Bosnia was making more than 400 IDs per month from the DNA-led system.
MCEVERS: As of today, 85 percent of the victims have been matched to their relatives by DNA. And some of the remains have been returned. Now the DNA-led approach is a technique that's routinely used by scientists who are looking to identify remains, the FBI, Argentinean and Guatemalan scientists working to ID victims of the so-called dirty wars, and authorities after 9/11 and Hurricane Katrina.
One of the last holdouts? JPAC. They still use the classic method. JPAC says that's because it's different with American servicemen. First, JPAC won't just match a leg bone to a relative and call it an ID. They want to return as complete a set of remains as possible. And, says JPAC lab director Tom Holland, you can't just go digging unknown soldiers up from cemeteries unless you know you're going to get it right.
HOLLAND: If I start exhuming large numbers that I can't identify, at some point somebody in Washington is going to look at this and say, they're now only identifying half of what they're exhuming.
MCEVERS: Digging up American remains only to store them in a lab, he says, is considered disrespectful by most military officials in Washington.
HOLLAND: We are talking about war dead. We are talking about men who this country sent in harm's way. And they're owed a modicum of respect.
MCEVERS: Holland is so concerned with getting it right, he only approves 4 percent of the requests to dig up remains. In nearly two decades as JPAC's lab director, Holland has never had an individual ID overturned. Still, former JPAC employees told us this aversion to risk is one main reason JPAC makes so few IDs each year. One way to build a faster DNA-led process, several scientists told us, would be to house JPAC's DNA work in one place, or at least under one chain of command.
As it stands now, our reporting found the DNA work is done at six different agencies, in Hawaii, Delaware, Kentucky, Tennessee, Texas and Virginia. Congress first started noticing the problems at JPAC and at sister agencies last summer. The government office of accountability found what it called leadership weaknesses and a messy bureaucracy.
SENATOR CLAIRE MCCASKILL: We are here today to review the Department of Defense's management of POW/MIA accounting.
MCEVERS: At this hearing on Capitol Hill last summer, Senator Claire McCaskill, a Democrat from Missouri, held up a jumbled confusing looking organizational chart of all of the branches of all the different agencies tasked with recovering servicemen who died in past wars.
MCCASKILL: Look at that. I mean, is it any wonder that this is a mess?
MCEVERS: Even the acronyms and names are seemingly endless: JPAC, DPMO, AFDIL, LESL, Army Casualty, Navy Casualty, Air Force Casualty, Marines Casualty, many of them falling under different parts of the Pentagon. JPAC officials told us each layer of bureaucracy is necessary, that they take the utmost care with every case. They say returning soldiers who died in past wars is what defines us as a nation.
In an interview, Senator McCaskill said that sentiment is not enough.
MCCASKILL: This is a subject matter that everyone is so warm to. It's a cause that everyone believes in so deeply that I think people have not asked the tough questions or provided the kind of oversight that this effort really demands. So for 20 years, there has been dysfunction, disarray and that's what we're trying to get to the bottom of.
MCEVERS: Now the Pentagon is doing just that. Defense Secretary Chuck Hagel has ordered a team to come up with a plan to maximize the number of IDs, improve communication with families and streamline the bureaucracy. DNA scientist Ed Huffine says higher-ups can reorganize all they want, but if JPAC doesn't change its approach...
HUFFINE: The system will still fail.
MCEVERS: Lisa Phillips knows all about how the system can fail a family. Her great Uncle Joe died in a POW camp in Burma during World War II. His remains were later recovered and put on a plane to a U.S. base in India.
LISA PHILLIPS: The aircraft went down, 55 minutes from destination and it has not been found since.
MCEVERS: Phillips says Joe's wife waited for his body to come home until the day she died in 1980.
PHILLIPS: She had said that she would never move out of Connecticut because she wanted the government to know where she lived when they brought her beloved Joe home.
MCEVERS: In 2002, Phillips went to JPAC. They didn't take up her case so Phillips and other families hired a private investigator to find the crash site. They tracked down the relatives of the 53 men whose remains were on Uncle Joe's plane and got them to collect and send in their DNA. She says not only did JPAC and the other agencies not ask for the DNA, but when the families went to submit their samples, they found it was nearly impossible to figure out how. They started with the Pentagon's prisoner of war missing personnel office website.
PHILLIPS: Then you have to click on the little American flag, then you have to click on World War II, then you have to go search for DNA.
MCEVERS: Which took Phillips even further down the rabbit hole, all this to give JPAC a tool most scientists say they should be seeking themselves, ideally in a public nationwide campaign. Uncle Joe's plane still hasn't been found. JPAC did go search one site in India, but it was the wrong plane. Phillips says she won't stop until he is found.
PHILLIPS: I will never give up. I've already threatened that. I said don't even think when I'm dead. I said my kids will go look.
MCEVERS: Phillips says if major change doesn't happen at JPAC and its sister agencies, agencies that have promised to leave no one behind, she and the thousands of other families who are looking for their lost relatives will have to continue the search on their own. Kelly McEvers, NPR News.
BLOCK: This investigation was reported in partnership with the independent news agency ProPublica. NPR and ProPublica were granted rare access to JPAC's central identification lab. You can see photos from the lab on our website, NPR.org.
NPR transcripts are created on a rush deadline by a contractor for NPR, and accuracy and availability may vary. This text may not be in its final form and may be updated or revised in the future. Please be aware that the authoritative record of NPR's programming is the audio.