Over Our Heads

Should Seattle worry about rogue missile attacks, or the multibillion-dollar program being deployed to stop them?

Gaze out on Everett’s Port Gardner Bay and imagine a 250-foot-high sphere, something like an astronomical observatory or a mutant golf ball, perched on the edge of a floating 400-foot-long rig. The mystery object, which could really start going up as soon as next year, is called the Sea-Based X-Band Radar, and it’s a crucial component of President George W. Bush’s Ballistic Missile Defense program.

The Pentagon will decide this August whether this vision will come true. The other sites in competition are in Hawaii, the Marshall Islands, California, and Alaska. A few environmental groups have filed protests, noting the damage that might be done by such powerful radar beams. The Pentagon’s environmental impact statement denies this claim. The big ball will be in port for less than half the year. (Most of the time, it will be on station out in the Pacific Ocean.) While in port, the radar will be turned on for 20 to 60 minutes a day. At 150 meters (the closest anyone can come to the dock at those times), the effect will be the same as standing two inches from a microwave oven.

The larger debate, which has been heard far less, involves the program itself. Missile defensethe ability to shoot down an enemy’s nuclear missiles after they’re launched and before they land and explodeis a dream that nuclear war planners have been conjuring for nearly half a century. But no president has taken it more seriously or brought it closer to reality than Bushnot even Ronald Reagan, whose utopian vision of a nuclear shield inspired the derisory nickname “Star Wars.”

The military budget that Congress passed earlier this year allocates $9.1 billion to missile defense, a 20 percent hike over the amount approved last year and three times the annual levels spent at the peak of Reagan’s heyday.

Bush also did what many conservatives had only fantasized about doing for 30 yearshe pulled out of the 1972 ABM Treaty, which, besides forming the centerpiece of U.S.-Russian arms-control agreements, banned the deployment (as well as sharply restricted the research and development) of anti-ballistic-missile weapons. Last December, when he ordered Defense Secretary Donald Rumsfeld to start deploying a system, he called the move “an essential element of our broader effort . . . to meet the new threats we face,” and added that “defending the American people against these new threats is my highest priority.”


Bush’s program is far more elaborate than any previous incarnation of the concept. If ever completed (and there is much doubt about this), the system would be divided into three segments, each consisting of hundreds or thousands of parts, all acting in tandem.

The first segment of the program involves “boost-phase intercept.” Satellite-based radars detect a missile being launched. Then lasers, fired from other satellites or from a specially equipped Boeing 747, shoot the missile down in the three to five minutes that it rises through the sky, before it clears the atmosphere.

Second is the “midcourse intercept,” the 20-minute phase when the enemy’s missile arcs through outer space. In the Pentagon’s most widely discussed scenarioa North Korean missile aimed at America’s West Coastthis would be the phase when it crosses the Pacific Ocean. Or, on the threat boards of an earlier era, it was when Soviet missiles flew over the North Pole, then down across Canada, toward any number of possible “aim points.” In this phase, various devicesspace-based, ship-based, air-based, and ground-based “kill vehicles,” all guided by highly sophisticated radarswould try shooting the missile down before it gets close to our shores.

Third is the “terminal defense,” the final 30 seconds when the missile’s warhead plunges back down into the atmosphere toward the target, and when U.S. anti-missile missiles make a last-ditch effort to destroy the weapon before it explodes.

The boost-phase intercept is, in some ways, the easiest. As a rocket ascends toward space, it moves slowly, straight up, with engines burningabout as vulnerable a target as might be imagined in this realm. In another way, though, it’s the hardest phaseyou have to know where the missile is being launched, and the sensors and lasers need to be orbiting overhead at all times.

Terminal defense might be thought a fairly straightforward task. Computerized radars extrapolate from launch data where the warhead will be landing; fire-control systems then automatically aim the weapons in that direction; and bang! you’ve got him. As we learned from the first Gulf War, though, it was very hard for Patriot air-defense missiles to shoot down Iraqi Scuds (postmortem analysis indicated only a 9 percent success rate). As we learned in the second Gulf War, the computers in even the newest-generation Patriots have a tendency to mistake British war planes for Scuds. (They shot down a British Tornado and an American F18, and would have downed an American F16 as well, except that the F16 pilot, seeing that his plane was being “painted” by the Patriot’s radar, fired an anti-radar missile at the missile batteries and destroyed them in the nick of time.) And Scuds are much slower, bulkier, and shorter-ranged than the warheads to be shot down in Bush’s multilayered missile defense system. (Nor are we talking apples and oranges here: The new Patriots just used in Gulf War II, the PAC-3s, are to be among the interceptors in this system’s terminal-defense phase.)

The midcourse intercept, the phase between launch and re-entry, is where the X-Band Radar comes in. This radar consists of 81,000 transmit-receive modules, and all of them will be working overtime. It must seek, detect, and track enemy missiles as they’re arcing in outer space. It must also discriminate between the missiles and all the other flotsam floating up therenot just satellites and meteorites, but decoys and debris that the enemy might have programmed the missiles themselves to release in flight. The radar must transmit this data to the fire-control systems, which will assign weapons to the targets. Then it must assess whether the weapons fired actually destroyed the missileand, if they didn’t, start the process again.

This last stepassessing success or failure of shootdownis harder than it might seem. The Bush system’s interceptors are called “kinetic kill vehicles.” They don’t explode. Rather, they destroy the target by smashing into it, a task that involves closing in on it at 4.6 miles per second (16,560 mph) 144 miles above the Earth. The radar operates at X-band (extremely high) frequencies because it needs to detect what’s going on way up there very, very precisely.

Finally, all these phases must be integrated into a highly centralized and automated command-control system. The boost-phase sensors must alert the later-phase components that a missile is on its way. The X-Band Radar must not only perform its part of the mission but relay data to the terminal-defense radars, so they can do theirs.


If you have ever tried to debug a new software program, you must be rolling your eyes by now, dismissing this missile defense business as wildly impossible. You would not be alone. For several decades, many scientistsincluding several inclined to favor the idea of such a programhave dismissed the real thing because the technical obstacles are just too enormous. Among the latest and most interesting skeptics to make his qualms public is Thomas P. Christie, current director of the Defense Department’s own Office of Operational Test and Evaluation (OT&E). In his annual report, released last February with little fanfare, Christie made the following observations in his chapter on the missile defense program. The system, he wrote, “has yet to demonstrate significant operational capability.” The tests conducted to date have “suffered from the lack of production-representative test articles and test infrastructure limitations.” In other words, the mock warheads that the system’s interceptors have been shooting downand even then with so-so scoresdo not resemble the warheads that a real enemy would launch. Their velocities are slower, their trajectories are different. Often, to make the test easier, the mock warhead has beamed an extra-bright “signature” so the radar can spot it more readily. There have been no tests in which the system tries to shoot down more than one mock warhead, and there have been no tests involving realistic decoys.

The issue of decoys is a crucial one. In a September 1999 National Intelligence Estimate, the CIA concluded that any nation capable of building ballistic missiles “would also develop various responses to U.S. defenses,” including such “readily available technology” as decoys, chaff, or wrapping warheads in radar- absorbent material.

Anticipating this possibility, the missile defense program’s managers decided in 1997 to conduct a test in which the mock warhead would be surrounded by nine or 10 decoys, all of them with properties similar to the warhead. By 1998, they had changed their minds. The revised order specified that there would be just three decoys. In 1999, the bar was lowered again: One decoy would be enough, and it could be a large balloon.

Philip Coyle, Christie’s predecessor who was in the OT&E office at the time, recounted this sequence in his report that year, noting that the balloon in question was “very dissimilar” from the target warhead and that, therefore, the missile defense radar “can easily discriminate” between the two. In other words, the test was rigged.

In some quarters, the OT&E office is viewed as having hypercritical standards. Its mandate is to ask the tough questions, to insist on stringent scenarios. But a close reading of testimony from the missile defense agency (MDA), the Pentagon shop that runs the missile defense program, reveals the same cautions.

Buried in the MDA’s budget documents last February is this sentence: “There is no final or fixed Missile Defense architecture.” (“Architecture” in this sense means what it does in common usagea design, a blueprint.) At a press conference a couple months earlier, on Dec. 17, 2002, Maj. Gen. Ronald Kadosh, the MDA’s commander, had gone further still. “The final architecture is not knowable today,” he said, “because we have a lot more research and development to do.” (Note: It’s not just unknown, it’s unknowable.)


Nor, for the same reason, does anyone know how much a viable missile defense system will cost, or how long it will take to develop, build, and test one. The MDA’s budget book contains a timetable, showing certain key elements of the multilayered system just beginning to come into development as late as 2010and the book doesn’t pretend to make predictions beyond that year.

Gen. Kadosh, to his credit, has always been unextravagant in his claims. Last June, in hearings before the House Armed Services Committee, Kadosh said, “I cannot overemphasize the importance of controlling our expectations and persevering through the hard times [note: He even admits the tests are going through “hard times”] as we develop and field a system as complex as missile defense.” He defined the program’s “test philosophy” as adding “step-by-step complexities over time. It is a walk-before-you-run, learn-as-you-go developmental approach.”

And yet, last Dec. 17 (the same day as Kadosh’s “unknowable” press conference), Bush ordered Rumsfeld to begin deployment of the missile defense program, with 10 Boeing-made ground-based anti-missile interceptors to be fielded by the fall of 2004, and another 10 by 2005 or 2006, at Fort Greeley Air Force Base, Alaska, along the trajectory that missiles fired from North Korea would follow.

These anti-missile missiles would be used primarily for expanded testing. But, at the same press conference, J.D. Crouch, the assistant secretary of defense for international security, said they would also provide “a very modest interceptor inventory . . . a useful defensive capability” (adding “but one that, you know, has limits”). Rumsfeld himself, who has long been an advocate of missile defenses, made the point more directly. Last February, speaking of a looming nuclear-missile threat from North Korea, he said, “We need to get something out there.”

Rumsfeld made this comment to justify a request that he had just filed rather quietly with the congressional Armed Services Committees, asking them to exempt missile defense from the law requiring all weapons to undergo operational tests before being deployed. The Senate committee’s ranking Democrat, Carl Levin of Michigan, objected, noting that this law had been designed “to prevent the production and fielding of a weapon system that didn’t work right.”

After some press coverage of this exchange, Rumsfeld retracted his request. But he is moving ahead as planned, simply altering the language, referring to deployments as “test beds.” We’re not producing or fielding weapons, he seems to be saying. We’re just testing them. Similarly, the budget for missile defense is labeled “Research, Development, and Testing,” not “Procurement.”

The program’s defenders argue that technological progress will solve its current problems, and to some extent they might be right. Certainly the ABM Treaty inhibited serious testing. The treaty banned not only the deployment of anti-missile weapons but also the testing of such weapons at sea or in space. With the treaty discarded, systems can now be tested over more realistic ranges, trajectories, and speeds.


However, some of a missile defense system’s limitations are inherent to the mission, ultimately immune to technology’s rapid pace.

As far back as 1958, when the U.S. Army was spending billions on the nation’s first anti-missile system, called Nike-Zeus, a panel of Pentagon scientistsall of them hawks who enthusiastically favored the idea of the programconcluded that the system simply would not work against a dedicated attack, mainly because of its vulnerability to decoys and chaff. In 1970, when Richard Nixon was pushing the Safeguard ABM system, a top-secret memo from his science advisers warned that it “would be obsolete within three to four years after it is first deployed,” due mainly to the enemy’s ability to develop such countermeasures as decoys and chaff. Nixon signed the ABM Treaty with Leonid Brezhnev not out of dovishness or some adherence to arms-control theory, but rather out of recognition that ABMs were not a cost-effective defense.

Bush’s missile defense program, after several more years and several more tens of billions of dollars, mightmightbe able to shoot down one nuclear missile. However, what if the enemy fires two missiles? All bets are off. There is not a single testnot even a rigged testthat indicates a system could deal with more than one target at a time. Nor is any such test currently planned.

Here’s another possibility. We spend hundreds of billions of dollars (and no one doubts that is what it will take) to build such a system. The thing works reasonably well in tests. It maintains a certain level of reliability and readiness. What might an enemy do to defeat such a defense? Maneuver around it. Rather than launch a ballistic missile from its homeland (the most expensive kind of threat to pose), a clever enemy might launch a low-flying cruise missile from a barge near an American coast. Or it might import a bomb in a truck, a boat, or a suitcase. Bush is spending billions for missile defense but peanuts for homeland defense. Rumsfeld is focused more on putting an X-Band Radar in Port Gardner Bay than in providing port security. It may well be that they’re focusing on the wrong nuclear threat.

Fred Kaplan writes the “War Stories” column for Slate magazine, and is the author of The Wizards of Armageddon.