This story was originally published by The Center for Public Integrity, which is a nonprofit, nonpartisan investigative news organization in Washington, D.C.
Shortly after 11 a.m. local time, a U.S. ballistic missile target loaded with a mock nuclear warhead blasted off from Narrow Cape, a low-lying coastal area of Alaska's Kodiak Island. A network of radars from Alaska to California tracked the target, watching for the release of metal chaff, Mylar or aluminum balloons, or other objects like those that North Korean missiles might use to fool U.S. defenses.
This simulated attack on the United States on Dec. 5, 2008 was the first time massive sea- and ground-based defenses would try to penetrate the decoys or countermeasures that might be used to hide a warhead in the near-vacuum of space. As the Pentagon had wanted, a rocket interceptor launched from a silo at California's Vandenberg Air Force Base destroyed the warhead and the radar network performed well, prompting officials to declare the test a success in a press release the same day.
But the real test of U.S. defenses against the countermeasures that North Korean missiles might eventually carry -- the primary objective of that exercise, which was estimated to cost taxpayers between $200 million and $300 million -- never happened. The target malfunctioned and failed to release them.
For years, the public's focus on the nation's nearly $10 billion-a-year missile defense program has been on whether American interceptors can hit incoming ballistic missiles and protect the country and its allies, a feat often likened to hitting a speeding bullet with a bullet. More than $90 billion has been spent since 2002 to develop the means to target incoming threats and intercept them, but without much demonstrated success.
Less attention has been paid to the targets used in U.S. missile defense testing, which have failed or malfunctioned at an alarming rate since the 2002 inception of the Missile Defense Agency, which oversees all the development, procurement and testing programs. In the last five years, target problems occurred in two of the last three intercept tests of ground-based interceptors -- such as those already deployed to Alaska and California -- and in two of the last seven tests of the Army's mobile Terminal High Altitude Area Defense interceptors.
An investigation by the independent Government Accountability Office in 2008 found that 7 percent of the targets launched from 2002-2005 also had problems, a rate that more than doubled to 16 percent from 2006-2007.
Target problems have driven up costs, with GAO estimating the cost of the most recent ground-based interceptor tests at $230 million apiece. In total, the Pentagon is now spending roughly a half-billion dollars a year on targets, and another half-billion a year on testing.
And they have caused significant disruptions to testing schedules, often pushing back critical intercept tests by a year or more. The December 2008 test to see if missile defenses could distinguish between decoys and a warhead has yet to be repeated, undermining claims by both military and elected officials that U.S. missile defenses are capable and effective in protecting the homeland or U.S. troops overseas from a future missile attack by North Korea or Iran.
Those countries will be able to field threatening missiles during the next decade, the National Academy of Sciences told Congress in a report last September, adding that "at some point, countermeasures of various kinds should be expected." Defense officials should expect any weaknesses to be exploited, observed Tom Collina, an analyst with the Arms Control Association. "In a real missile attack North Korea could be expected to use decoys and countermeasures that US defenses would not be able to handle," he predicted.
Risky Test Set for Later This Year
The risk of another major development setback looms this fall when the military plans its first test of the missile defense system intended for Europe. In one of the most complex such experiments the Pentagon has ever attempted, two different interceptor systems will be used to try to defeat a near-simultaneous attack by two air-launched extended medium-range ballistic missiles. Originally meant to involve three different interceptor systems in a raid by up to five missile threats, the test was scaled back due to budget sequestration cuts, two congressional sources said. Unofficial estimates put the cost of the original test at more than $500 million.
A team of GAO investigators that has long pressed for reform in the MDA's targets program recently issued a warning about the Pentagon's plan to to use a new class of air-launched target missiles in this complex test without separately flight testing one of them first. "Using these new targets puts this major test at risk of not being able to obtain key information should the targets not perform as expected," Cristina Chaplain, GAO's director of acquisition and sourcing management, told the Senate Armed Services Committee at a May 9 hearing.
The manufacturer of these new targets, Lockheed Martin, disagrees. Noting that on May 13 it had successfully dropped a prototype out the cargo bay of a C-17 transport plane, it says the target missile is ready for the big test later this year. But the version that was dropped lacked an engine, so the test did not satisfy the GAO.
Richard Lehner, the Missile Defense Agency spokesman, said the agency's scrutiny of key target components and its "proven quality control processes" give officials "the confidence necessary ... to plan for and launch targets for the first time as part of a system-level flight test." Lehner also reiterated the Pentagon's official response to the GAO that any decision to perform a flight test of the new targets "must be balanced against cost, schedule, and programmatic impacts."
Chaplain and her colleagues, including her assistant for missile defense, David Best, and her boss, Paul Francis, have been using their audits and congressional testimony to try to get MDA to resolve the target problems and stop relying on high-risk strategies in which major purchases of targets, interceptors and other hardware are made before all the design and engineering bugs have been worked out.
"Since its inception, MDA has been operating in an environment of tight time frames for delivering capabilities -- first with a presidential directive in 2002 and then with a presidential announcement in 2009 on U.S. missile defenses in Europe," Chaplain told the senators. Budget constraints "have already necessitated tough trade-off decisions and will require additional steps to reduce acquisition risk," she added.
Besides the continuing pressure to meet development and deployment deadlines, there have been instances of poor execution by contractors, she and her colleagues say, as well as difficulties building an inventory of targets that do not have aged components, such as rocket motors from surplus Trident or Polaris submarine missiles.
National Security investigations in your inbox: Sign up for The Center for Public Integrity's Watchdog email
The High Cost of Failure
The stakes are higher as more missile defense elements -- sensors, interceptors and targets -- are added to increase the complexity and realism of the tests. "These are exceptionally expensive tests," Chaplain said in an interview, raising the possibility that well over $200 million will be wasted anytime one of them fails.
The GAO team has found that the missile defense programs most affected by target problems have been the Ground-based Midcourse Defense system and the mobile THAAD system. The first is the sole system tasked with protecting the United States against a North Korean missile attack, with 26 interceptors deployed to Alaska and four to California. The THAAD is a critical piece of land- and sea-based defenses scheduled to be deployed to Europe beginning in 2015.
The GMD system has not had a successful interceptor test since 2008, partly due to target issues, but some lawmakers are calling for its additional deployment in the eastern United States, anyway. And in March, amid North Korean saber rattling, the Obama administration announced a $1 billion plan to add 14 more of the GMD interceptors in Alaska by 2017.
A third system, the ship-based Aegis Ballistic Missile Defense, has fared better with targets and testing, although a key 2008 intercept test was postponed for three years due to problems with a Lockheed Martin LV-2 intermediate-range target missile and other issues, according to auditors. A March 2009 test had trouble with two refurbished Lance missile targets when both fell short of their expected trajectory, causing the Aegis BMD system not to fire one of its interceptors, GAO and MDA officials reported.
GAO also has documented major setbacks to missile defense programs due to other target issues, including inventory shortages and production delays of newly designed targets. The THAAD program was forced to postpone planned flight tests in fiscal year 2009 due to a lack of available targets, delays that GAO said cost about $201 million. A shortage of targets in 2007 prevented the ground-based system from achieving its primary test objectives that year and kept the Army from testing its radar systems.
Several analysts agreed that MDA's efforts to save money have backfired when target-related troubles surfaced. Philip Coyle, a former director of operational test and evaluation at the Pentagon from 1994 to 2001, called it a management, not a contractor issue. "If MDA told the contractors to test their targets adequately, and paid them for it, the contractors would be happy to do that," he said.
Missiles are sometimes thrown into a test against an interceptor without having been flown as targets beforehand to see how they will behave, mainly to save money, said George N. Lewis, a physicist and missile defense specialist at Cornell University's Judith Reppy Institute for Peace and Conflict Studies.
But if a test fails due to a poorly performing target, more money must be spent to buy another target missile and plan and execute another test. "My impression is that it's one of those things where you try to do something cheaply that ends up costing you a lot of money," Lewis said.
Consider what happened on Jan. 31, 2010 after a 45-foot Lockheed Martin missile target was launched from the Kwajalein Atoll in the Marshall Islands and soared toward the edge of the atmosphere. After its booster engine, an old solid-fuel motor from a Trident submarine missile, finished its burn, ground controllers rotated the missile slightly, said Lewis, who has written a detailed analysis of the test with Theodore A. Postol, professor of Science, Technology and International Security at the Massachusetts Institute of Technology.
As the missile turned, the spent rocket stage began "chuffing," Lewis said, spewing chunks of unburned fuel and insulator material, varying in size from less than an inch to 6-8 inches or larger, each creating unexpected radar signals that confused a sea-based radar defense system. The radar failed to identify the warhead, and so an interceptor fired from a silo at Vandenberg could not hit its target.
After that test, which the Pentagon said cost $150 million, missile defense officials took contractors to task for chronic lapses in quality control. "I'm not going to name names today, but I'm going to tell you we continue to be disappointed in the quality that we are receiving from our prime contractors and their [subcontractors] -- very, very disappointed," David Altwegg, then-MDA executive director, told reporters after the test.
But GAO's Chaplain told Congress that MDA never subjected the target that had failed to a "risk reduction flight test."
"While the target ... was successfully flown in that flight test, aspects of its performance were not properly understood and lack of modeling data prior to that test contributed to significant delays in the test program," she said in an April 2012 report.
Lewis said in an interview that debris fallout was not unusual for a Trident C4 motor that was 25-35 years old. Trident missiles had been launched many times over the years, "but had they flown it as a target, they probably would have found out about" the chuffing, he said.
This target-related test failure came less than two months after a target built by Coleman Aerospace, a unit of L-3 Communications, embarrassed missile defense officials when it had trouble during its release from a C-17 transport plane. The launch of a THAAD interceptor had to be aborted when the target's motor failed to ignite once the missile cleared the plane's cargo bay.
"We all sat there and watched the target fall into the water," MDA's Altwegg told reporters after the Dec. 11, 2009 test, which cost $41.2 million. He said the target was found to have a "big-time quality problem."
Its failure led to a delay of the planned test, cancellation of five tests scheduled for fiscal year 2010, and "hundreds of millions of dollars" being spent to develop and acquire new medium-range air-launched targets, GAO said.
The failure also prompted the MDA to suspend Coleman Aerospace for a year due to quality-control issues.
A Decade of Management Problems
The GAO, which investigated the MDA's target procurement program in detail in 2008, traced such performance problems and the rising costs of targets and testing to difficulties the agency had overseeing a long-term contract awarded to Denver-based Lockheed Martin Space Systems in December 2003. The military had decided to abandon its piecemeal purchase of targets and have Lockheed Martin act as a so-called lead systems integrator, charged with developing and producing short, medium and intermediate-range ballistic target missiles for use against all of its missile defense systems.
The idea was to use common components for all the targets, reduce the time needed to produce them, and cut costs. Existing targets in the military's inventory had little in common, varying in size, shape and the age of their components -- including some rocket engines from submarine missiles over 40 years old, GAO said.
But GAO faulted missile defense officials for not doing a thorough cost analysis or evaluating all alternatives before embarking on their plan. By the time Paul Francis, then-GAO's director of acquisition and sourcing management, wrote leaders of the House and Senate defense committees in September 2008 to report the findings of his investigation, the total cost of the target procurement program had ballooned to $1 billion or more, with $553 million already spent.
None of the new targets had been delivered, forcing the MDA to use older targets much longer than planned, emptying its inventory of certain kinds of targets and putting the increasingly complex tests at risk of failure.
"Early in the development of [the procurement program], MDA underestimated the technical and design challenges involved in the development of a new target family," Francis told lawmakers. "By May of 2006, MDA recognized that the funding set aside for [target] development was no longer adequate," he added.
The cost of each target jumped from the $4.5 million to $8.5 million the agency paid in 2002-2006 to an estimated $32 million to $65 million in 2008-2010, Francis said. As a result, the agency's plan would yield fewer targets at higher costs, he said.
The strategy "has not gone as planned," wrote Francis, who added that Lockheed Martin chose to reuse surplus missile components for some of its targets. "The availability of targets for flight tests continues to be problematic, and as a result the scope of the flight test program has been reduced to better match available targets," he said.
Work on all but one of the new classes of target missiles was cancelled in June 2008, partly due to the unexpectedly high costs. The MDA responded by promising to make "threat-representative targets available on schedule and within the funding allotted."
In early 2009, Lt. Gen. Patrick O'Reilly, then the MDA director, publicly acknowledged the gravity of the availability and reliability problems, as well as the rising costs and schedule delays. He promised a new strategy, and began awarding separate contracts for four classes of targets, GAO's Chaplain said.
In fiscal year 2011, the MDA received 11 targets, all of which performed as expected, she said. In addition, the agency awarded a competitively bid contract to Orbital Sciences Corp. to produce eight targets by 2015.
But Chaplain noted that, even under the agency's new approach, an early attempt to award a competitive contract was cancelled after the agency received bids that were more expensive than it anticipated. The agency also continues to rely heavily on Lockheed Martin to produce some of its targets, she said.
In an interview, Chaplain said that even though the number of companies able to build ballistic missiles is quite small, GAO "would still like to see more competition in the procurements to maximize the potential for savings."
Coyle, the Pentagon's former testing chief, sees target failures as part of a larger problem with the testing of the ground-based missile defense system, which he said has gotten worse over time. He pointed to MDA data showing a decline in the rate of testing and the rate of success over the years, with three successful intercept tests out of eight since December 2002, and only one out of three since the Dec. 5, 2008 test that failed to release decoys.
"The performance of systems undergoing engineering development is supposed to get better with time, not worse," said Coyle, a Pentagon veteran who specialized in overseeing such efforts. "If you count [the 2008 countermeasures test] as a failure, then the record since Dec. 5, 2008 is zero out of three," he added. "Zero in five years!"