"East Asian Regional Security Futures: Theater Missile Defense Implications"
The United Nations University, Tokyo, Japan, June 24-5, 2000


by Michael O'Hanlon

Missile defense has emerged as probably the most important defense issue in the 2000 presidential campaign, and one of the most difficult issues in current U.S. diplomacy.  It may also be the first security matter that the next American president needs to face, given the timeline on which the existing Pentagon national missile defense program has been placed and the need to determine how it will be reconciled with the 1972 U.S.-Russia Anti- Ballistic Missile Treaty, which presently bans any such nationwide defenses.

This essay reviews the state of various U.S. programs, including those theater-missile defense (TMD) systems designed to counter short-range or medium-range missiles, and national missile defense (NMD) against threats to the American homeland, with considerable focus on the implications of missile defense for the East Asian theater.  It concludes with several observations, informed partly by the China-Japan- U.S. discussion in Tokyo of June 24-25, 2000 sponsored by the Nautilus Institute and United Nations University at which I was a participant. 


Since the Gulf War, the United States has significantly improved its  only existing missile defense system, the Patriot.  Patriot's radar now has greater range and can track more objects simultaneously.  Starting in 2001, the Pentagon is to deploy a further-improved version of the Patriot.   That is later than initially expected—another reminder of the difficulty of developing even relatively simple missile defense systems— but the new Patriot is likely to work with high effectiveness.  It will have the ability to identify warheads so that it will not be fooled by an enemy's use of simple decoys or the breakup of a missile's body during atmospheric reentry (as early vintages of Patriot were in Desert Storm).  It will also possess a new "hit-to-kill" interceptor missile that achieved a completely successful test in early 1999.  Whereas the existing Patriot system, known as PAC-2, can defend an area with a radius of some 10 to 15 kilometers, the new PAC-3 will triple that coverage.  The PAC-3 interceptor has its own self-contained radar for homing in on a target.  It also features 180 small thrusters for fine steering in the final phases of intercept (earlier Patriots have fins for steering and blast-fragmentation warheads). (1) 

The Pentagon is also continuing to develop a low-altitude theater  defense based on Navy "Aegis-class" ships that uses a modified form of the Standard antiaircraft missile.  Known simply as the Navy area defense system, it is designed to have a coverage zone somewhat larger than that of the Patriot PAC-3.  Recent tests to validate the capabilities of the system's individuals components--the missile and the ship radars--have been successful. (2)  The Navy hopes to deploy this system by 2003 and to install it on at least 40 ships.

Less advanced are other programs designed to provide defense against theater missiles over regions of a few hundred kilometers' width—a much more difficult task than lower-tier local defense, but also one that must be mastered if one is to protect large regions affordably.  The key programs are known as THAAD (for Theater High- Altitude Area Defense) and the Navy Theater Wide system (or NTW).  They have often been in the news because of their testing difficulties.  However, THAAD's problems were more the result of shoddy workmanship than the viability of the hit-to-kill concept. (3)  In any event, THAAD finally scored a direct hit during a test in June of 1999 and another hit in August of that year.  The interceptions, using THAAD's infrared seeker to guide the  missile's final approach, occurred at a much higher altitude and greater distance from base than Patriot is capable of. (4)  Whichever of these programs, THAAD or NTW, turns out to advance more rapidly is to be fielded in 2007, the other sometime thereafter—though THAAD enthusiasts, only recently on the ropes after six straight test failures, have been talking about speeding the program up in light of their pair of successes in 1999.

A final possible TMD system is the airborne laser program (ABL), which would weaken an offensive rocket's outer  surface, causing it to rupture while burning.  In other words, it would destroy the missile in its boost phase.  In theory it could work against any missile, medium-range or long-range, with a boost phase long enough to give the ABL time to shoot at it. 

The ABL system would work as follows.  Two aircraft, each carrying a laser capable of firing 20 shots before returning to the ground for more chemical fuel, would operate near the likely launch points of enemy missiles and above the clouds at about 40,000 feet altitude.  In principle, it could be deployable by the end of the decade as well—or at least so the Pentagon hopes.  Specifically, the laser is supposed to be flight-tested in 2003, and if that test is successful to enter into engineering manufacturing and development the next year; initial operating capability would be in 2007 and the full capability of seven aircraft would be available in 2009.  The ABL program schedule seems likely to slip, however, given the fact that high-powered lasers are a fundamentally new type of technology.  The program remains essentially in the laboratory experimental stage today, a point emphasized by the Pentagon's director for operational test and evaluation Philip Coyle and director of defense research and engineering Hans Mark.(5)  For example, even though progress has been made in dealing with the effects of atmospheric turbulence on the laser beam (by deforming the laser mirror to compensate for that turbulence), the Pentagon still does not have a sufficiently powerful laser of the size and weight needed to operate aboard an aircraft. 

If and when the ABL is available, it would likely have a capability against any ballistic missile within a range of several hundred kilometers.  That would give this system the same advantage that other boost-phase defenses would have in not threatening the Russian or Chinese deterrent.  However, it is not clear if countermeasures, such as reflective coatings on offensive missile bodies, could defeat such a laser.  And again, most fundamentally, it is not yet clear that the basic concept of an ABL will prove workable within the next decade.(6)  For that reason, while vigorous research and development of the laser technology makes sense, it is not considered an option for next- generation TMD or first-generation NMD in the remainder of this paper.(7)


In 1997, the Clinton Administration decided to commit to development of a light national missile defense system— in other words, a system capable of defending U.S. territory against long-range missiles fired from different continents, rather than defending deployed U.S. forces or allies against short-range and medium-range missiles launched from nearby locations. 

The Clinton administration intended for the U.S. NMD system to be based at a single site in North Dakota or Alaska, and to make at least a provisional decision about whether to deploy the system in 2000.  It originally hoped to have such a defense system deployed by 2003, but later recognized that the earliest practical data for initial deployment would be 2005.  That date was itself subsequently pushed back, to 2006, in an August, 2000 announcement by President Clinton.  The Clinton Administration's initial NMD capability for 2006 might be enough to shoot down as many as a few warheads, perhaps up to five.  The administration hopes to have deployed a full 100 interceptors by 2007 or 2008 that could shoot down tens of warheads from North Korea; this capability would be termed "C1."  However, the system would only be able to intercept a few warheads from the Middle East, since its radar coverage in that direction is to be far less comprehensive, and since the interceptor missiles are to be based in Alaska, a long ways away from the trajectories of Mideast missiles aimed at East Coast cities.  More thorough coverage against Mideast threats would come with the so- called C2 and C3 capabilities, including additional sensors as well as interceptor missiles based in North Dakota to complement those in Alaska.

The immediate political lineage of the Clinton Administration's strategic defense plans dates to the Republican takeover of Congress.  True to their 1994 Contract with America, they proceeded to write a defense authorization bill mandating a national missile defense by 2003.  President Clinton vetoed the bill, but then sought to co-opt the issue from the Republicans.  The Clinton Administration devised its "3+3" program for developing a system over three years and deciding in 2000 whether to deploy the system over the following three years.  To make the program more realistic, the Pentagon's May 1997 Quadrennial Defense Review added an additional $2.3 billion to it, doubling planned costs over the six-year period.(8) 

Support for NMD built up more steam in 1998, when both Iran and North Korea surprised the intelligence community with missile tests that showed rapid progress in their respective missile programs.  Iran tested the Shahab 3, with a range of some 1,300 kilometers, and continued to develop the Shahab 4 with an estimated range of at least 2,000 kilometers.  North Korea tested the Taepo Dong 1 with a range of some 2,000 kilometers.  The Taepo Dong 1 is a three-stage rocket, as intercontinental ballistic missiles using standard rocket fuel and materials must be. Its first two stages worked successfully before the third failed (after having successfully separated from the second stage).  The overall accomplishment was nonetheless surprising and disconcerting to U.S. intelligence.  North Korea also kept developing the Taepo Dong 2, a rocket four times as large and with a range of 4,000 to 6,000 kilometers.(9)  A range of 4,000 kilometers would put the Aleutian Islands within reach; a range of 6,000 kilometers would put most of Alaska and the small western islands of Hawaii at risk.(10)  The Taepo Dong 2 might even be able to deliver a nuclear-sized payload to parts of the continental United States if North Korea can succeed in adding a workable third stage to it, as in fact it tried with the Taepo Dong 1 in 1998.(11)  However, the technical challenges to making the Taepo Dong 2 work, and to building a small enough nuclear warhead to fly atop it, are significant. 

Since 1999, the Clinton Administration has explicitly allocated money in its future years defense program to deploy NMD.  It added $6.6 billion to its defense plan for the years 2000 through 2005 for that purpose (making for a grand total of $10.5 billion for NMD over that period, including research and development costs as well).(12)  Estimated total acquisition costs have since risen to $12.7 billion through 2005 (13) -- and to a grand total of around $20 billion (in constant 2000 dollars) including costs before and after the 2000-2005 period.(14)  Even if costs increase by 50 percent in the end, as is normal for high- technology weaponry, they would hardly be enormous by comparison with Pentagon fighter, submarine, and destroyer programs expected to run into the several tens of billions each.  Nonetheless, they would be  significant for a Pentagon that already needs to spend at least $20 billion a year more on weapons in the years ahead, due to the growing age of equipment purchased in large amounts in the Reagan years.(15)  Even in an era of surpluses, such large defense spending increases may not prove practical, as evidenced by the modest amounts of money both Governor Bush and Vice President Gore promised for the Pentagon during the 2000 presidential campaign.  So while missile defenses may not be inordinately expensive in an absolute sense, they are not cheap either, and the Pentagon will likely face real opportunity costs in any decision to deploy them.

In addition to the interceptor missiles, plans call for a new "x-band" radar (with relatively high frequency and thus relatively high accuracy), as well as upgrades to various U.S. early-warning radars around the world, to detect incoming threats and guide interceptors to them.  Altogether, this deployment plan is deemed "C1" by the Pentagon. 

The C2 capability would retain 100 deployed interceptor missiles, but add three more x-band radars, interceptor missile upgrades, and expanded communications infrastructure to share data between various sensors.  It might also include the advanced space-based infrared surveillance (SBIRS-Low) satellite constellation to the ground-based radars used with the C1 capability.  The basic goal would be to make the system more robust against Mideast threats, considered likely around 2010 (particularly from Iran), and to make it more capable against decoys and other countermeasures by allowing more precise imaging of both warheads and decoys.  It would be deployable by 2010, and cost an additional $5 billion for radar and communications upgrades plus another $10 billion for a satellite constellation known as SBIRS-Low (that is, space-based infrared sensor, low altitude).(16)

The C1 and C2 systems would fall within ABM Treaty guidelines allowing as many as 100 long-range interceptors to be based at a single site.  However, they would violate the ABM Treaty in its current form, since that treaty does not permit a territorial defense of any kind or size.  The single base of interceptors is supposed to defend only the nation's capital or an ICBM missile field.  The ABM Treaty would need to be revised or abandoned to permit deployment of the proposed NMD system.(17) 


Is the Clinton Administration plan sound, in light of the statements even from Pentagon officials that its NMD program is "ambitious," "risky," and perhaps even rushed? 

It seems likely that someday, hit-to-kill technology against an identifiable incoming reentry vehicle flying a clear trajectory will work.  Successful tests of the Patriot PAC-3, THAAD, and the national missile defense system were conducted in 1999—a notable accomplishment even if the NMD test was somewhat jury-rigged and somewhat lucky. An NMD test failure occurred in January 2000 when infrared sensors on the kill vehicle lost the target in the last few seconds before anticipated intercept, due to the fact that water droplets frozen in the kill vehicle's sensors blocked the flow of coolant to the sensor and thus prevented it from reaching the low temperatures necessary for operation.  However, such problems are to be expected in high-technology development programs, and do not suggest that the technology itself is unworkable—at least against a simple threat.(18) 

However, the 2005 schedule has now been proven to be unrealistic, and even the new target date of 2006 will be difficult to meet.  The actual booster to be used in the system will not be tested for the first time until 2001, and a production-quality version of that booster not before 2003.  As noted, only 3 of 19 scheduled intercept attempts were completed before the deployment decision was reached; two of those failed, meaning that other tests will have to accomplish much of what they were supposed to.  Moreover, even the total number of flight tests planned is modest for a major missile system; although the MX was tested only 19 times in the R&D phase, all other major strategic missile programs were each tested at least 25 times, the Tomahawk cruise missile was tested 74 times, and the Patriot—perhaps the most comparable, given its mission and its integration of missiles with an advanced radar— was tested 114 times.(19)  Technologically speaking, this program is rushed, probably imprudently so. 

Nor is there any sign of it being less rushed as time goes on.  In early 2000, the Pentagon's director of operational test and evaluation, Philip Coyle, predicted that the current missile defense development schedule will not hold.  Describing the deployment as possible even by 2006 is quite ambitious, even though it reflects the second postponement of the system's initial operating capability. (20)


An even greater challenge, and one that will not be easily solved by a simple scheduling change in the NMD program, is to discriminate advanced countermeasures from actual warheads.  This may not always be easy even within the atmosphere, but it is particularly difficult in the exoatmosphere, where NMD systems now under development would have to work.(21)  In such regions of outer space, air resistance will not have had a chance to separate out the generally lighter decoys from the heavier warheads (as it would be able to do for the Patriot and other TMD systems which operate within the atmosphere).(22)  In outer space, even extremely light decoys would fly the same trajectory as true warheads, so speed could not be used to distinguish the real from the fake.  To mimic the infrared heat signature of a warhead, thereby fooling sensors that measure temperature, decoys could be equipped with small heat generators, perhaps weighing only a pound.  To fool radars or imaging infrared sensors, warheads and decoys alike could be placed inside radar-reflective balloons that would make it impossible to see their interiors.(23)   Decoys could also be spun by small motors so that the balloons surrounding them rotated at the same speed as real warheads, in case the defense's radar was sensitive enough to pick up such motion. 

Such countermeasures would doom a national missile defense of the type now under consideration and development in the United States.  Although the Ballistic Missile Defense Organization claims to know how to deal with decoys, its argument is unpersuasive for those that would have the characteristics mentioned above.  For example, it cites radars that can resolve the details of an object's shape and the polarization of its radar return.(24)  But such radars are physically incapable of telling a radar-reflective balloon enclosing a spinning warhead from one containing a spinning decoy, since both objects would give the same radar returns. They would also doom other exoatmospheric interceptors such as the Navy Theater Wide system, whether operated in an NMD or a TMD mode.(25)

The news is not entirely bleak.  It is not trivial to develop good decoy technology, including the means to dispense decoys in space.  After all, the superpowers did not develop MIRV technology for releasing several independently targeted warheads from a single missile until they had had long-range ICBMs missiles for a decade or so.  (MIRV technology is more complex than decoy technology, but there are important parallels.)  Defenders of the Clinton Administration's system often note that Britain had considerable difficulty in developing ballistic-missile countermeasures with its so-called Chevaline system.  However, that system, designed to help warheads penetrate an endoatmospheric or terminal defense, was far more complex than what North Korea, Iran, Iraq, or some other country would need to defeat the planned U.S. system.  Simple decoys are just fine in outer space; far more complex  decoys are needed to mimic heavy warheads within the Earth's atmosphere.  Unfortunately for the United States, it is in outer space where its planned system would have to detect, discriminate, and destroy enemy missile warheads.

That said, it is nonetheless not entirely trivial to develop decoys that work properly.  It is not particularly hard for a technologically sophisticated country, but nor is it easy—and it may take a country such as North Korea quite a while to perfect the needed countermeasures.  Specifically, it is likely to require some realistic flight testing to make such decoys work, and a state without the resources or diplomatic space to test very much may not succeed in any limited period of time.(26)  Although countermeasures would be unnecessary for an attack using many bomblets filled with chemical or biological agents, such attacks would be less dangerous than those with nuclear warheads.  Chemical weapons are intrinsically less lethal than nuclear or biological agents.  Biological agents are most lethal when distributed over a wide area by a device like a crop duster.  They also tend to be most effective when the targeted country does not realize it has been attacked with biological agents until victims begin to show symptoms of disease, at which point it would already be too late to treat most of them effectively.  Ballistic missiles are therefore less than ideal means of delivering such agents.(27) 

All in all, a light nationwide defense based on exoatmospheric missiles will have serious limitations.  But it may still provide some capability against the type of threat a North Korea or Iran could develop in the next decade or so— assuming that China and Russia do not retaliate against a U.S. NMD deployment by transferring countermeasure technology to such states, which the U.S. intelligence community fears they might in fact do.(28)  It also provides the advantage of providing some type of defense capability against both Mideastern and Korean threats, and  the political reliability and physical security of being based on U.S. soil.  For these reasons, the Clinton Administration's plan to build such a defense is not without merit—even if its exclusive reliance on this system, and its preference to rush to build it in Alaska, are probably mistaken.


To make it much tougher for an enemy to defeat NMD with fairly simple countermeasures, the United States could develop interceptor missiles for boost-phase defense.  They would be relatively small but extremely fast-burn rockets that would be fired very quickly after an enemy launch was detected, catching up with the enemy ICBM while it was still in its burn or boost phase within the atmosphere.(29)  At that point, the enemy  ICBM—essentially a large, full gas tank—would be highly vulnerable and easy to see.  It would not be trivial to hit, since it would be accelerating, making its trajectory hard to predict.  But a fast interceptor missile with high acceleration could overcome that challenge. 

While still burning, an ICBM would also not have had the chance to dispense decoys or countermeasures, since they would not yet be up to the speeds needed for intercontinental trajectories, and would be rapidly slowed down if released within the atmosphere in any case.  While an advanced enemy could build fast-burn ICBMs to counter such a defense system, these types of ICBMs are much harder to develop than current missiles owned by the likes of North Korea, Iraq, and Iran.(30)

One drawback to the boost-phase defense concept is that, by hitting the rocket rather than the (nuclear) warhead, it would not necessarily destroy the latter.  The warhead could then continue onward, possibly detonating—and, most likely, at least scattering radioactive material where it landed.  However, since the rocket would have been hit before completing its boost phase, the warhead would almost certainly not have enough speed to reach densely populated parts of North America, and would most likely land in arctic waters or the tundra of Alaska or Canada.  While far from ideal, chances are low that anyone would be killed as a result, and chances are almost zero that any appreciable number of individuals would lose their lives.  By the standards of nuclear war, that would be a very good outcome.

Such defense systems could be deployed near the Korean peninsula, Middle East, and other potential trouble spots provided appropriate basing on land or at sea was available.  Because the U.S. intelligence community considers a potential North Korean threat most imminent—a judgment reaffirmed by its 1998 Taepo Dong I launch—the boost- phase concept would most likely be based in Northeast Asia initially, either on barges in the Sea of Japan or on land in the general vicinity of Vladivostok, Russia.  Deployments focused on the Middle East could follow, perhaps in 2010 or thereafter (though there is some chance that they might be considered sooner, depending on the evolution of the Iranian and Iraqi intermediate-range missile threats to Europe).

Given the size of Iran, defending against a missile launch by that country would require two boost-phase interceptor bases.  One would be north of Iran—in the Caspian Sea, Turkmenistan, or possibly Kazakhstan or Uzbekistan.  Such basing could be tough to arrange, meaning that boost-phase defense against Iran may not work, or may require airborne interceptors that would be usable only if the United States first established air superiority as part of a major military operation.  The other base would be below Iran, in the Persian Gulf, Sea of Oman, or possibly Oman, Saudi Arabia, or the United Arab Emirates.  Iraqi missiles might be defended against by a base in eastern Turkey, which could also defend the United States (and large parts of Europe) against launches from certain parts of Iran.(31) 

The limited geographic scope of boost-phases defenses are at once one of the concept's greatest strengths and also its greatest weakness.  Basing a defense on foreign territory, especially that of a non-ally, would raise questions about its dependability in wartime.  Land-based boost- phase systems in particular would be difficult to move if new threats developed.  Sea-based boost-phase systems would not be useful against missiles from all potential threats, since not all are near international or friendly waters.  But missiles launched from the interior of a large country such as Russia or China could not be intercepted with boost-phase defenses—making it likely that Moscow and Beijing would find such a defense system less threatening than the current Clinton system.

Such a boost-phase system does appear within reach technologically,  particularly against North Korean or Iraqi threats—quite possibly on roughly the time horizon of the Clinton administration's planned C1 capability, and almost certainly as fast as the C2 system could be fielded.  It would require a new interceptor of extremely high speed, but that would be the main technical challenge, and a fairly straightforward (if significant) one at that.  Missiles like the Trident II or D5 missile were built within a decade, and this boost-phase missile likely could be as well (though the challenges would admittedly be different in its case, relating more to speed than to precise guidance).(32)  In fact, even in the early 1980s a major U.S. defense contractor was proposing a missile that could reach ICBM-range speeds of 7 kilometers per second in just 50 seconds of boosting, in contrast to boost times of 70 to 100 seconds commonly foreseen for the slightly- faster boost-phase interceptors.(33)

A boost-phase defense would not require a sophisticated sensor network on a par with what is required for the Clinton administration's program.  In fact, its main infrared seeker would have such a hot target to home on that it could use relatively inexpensive, simple, short-wavelength devices rather than the long-wavelength IR seeker needed on the exoatmospheric interceptor system (to say nothing of the long-range radars needed by the latter system).(34) 

Nor would a boost-phase defense likely be as expensive as the current NMD program now in the works.  The rockets would probably be more expensive than those envisioned for the Clinton administration's planned system (about $18 million apiece, according to the latest estimates of the Congressional Budget Office).  But they would probably not be on a par with the most expensive ICBMs and SLBMs ever built, which are much larger systems.  (The MX, for example, cost more than $100 million per copy,  or about six times the cost of an interceptor in the planned administration system).(35)  But the rest of the NMD technology—which accounts for about two-thirds of the costs of the administration's C1 proposal, and three-fourths of the cost of C3 (36)—would be far simpler for boost-phase defense.

This boost-phase defense concept would also require modifications to the ABM Treaty.  But it seems likely that in the end, such modifications should be less troubling to Moscow (and Beijing) than the planned system, since the defense would not work against missiles launched from the interior of Asia.  It could in theory be reconfigured to work against a missile after it had left the atmosphere, but that would require better sensors as well as testing.(37)   Barges would also have a hard time avoiding attack by cruise missile or submarine if deployed near a large, military advanced couintry.  Russian President Vladimir Putin appears to agree with these arguments, if his proposals shortly after his June, 2000 Moscow summit with President Clinton are a fair and sincere indication.  However, as of this writing it is too soon to judge his true views on the subject, given the sketchiness with which he initially proposed them.(38) 

The coverage zone of the boost-phase defense would be about 1,000 kilometers beyond the location of the interceptor missile base, since the interceptors would have only two to three minutes to make their intercepts. (They might be launched a minute or minute and a half after the enemy ICBM was fired, and would accelerate for 70 to 100 seconds before cruising at roughly 8 to 9 kilometers per second thereafter).(39)  Missiles launched from central, western, and southern China or most of Russia and headed over the North Pole would thus be beyond the range of interceptors based near the Korean peninsula, since the ICBMs would have completed their boost phase before interceptors could reach them.   In theory, these boost-phase defenses could be used to shoot down Chinese shorter- range missiles headed at Taiwan or Japan.  But they would be very expensive devices to use in such missions, and would almost certainly not be produced or deployed in sufficient numbers to constitute a serious counter to a PRC missile force numbering in the several hundreds of weapons.  In addition, while they might work in a TMD mode against intermediate-range missiles, they would not work against short-range missiles, which have burn times far too short to allow boost-phase interceptors to reach them while they are still burning.


To create a "thicker" nationwide defense than considered above, the United States could either simply increase the numbers of interceptors and radars and other components of the systems proposed above, or consider alternative approaches.  Three ideas are considered here: an expansion of the Clinton Administration's plan, perhaps even beyond the "C3 capability" now envisioned by the administration for 2011 or so; a mid-course concept using Navy surface combatants with Aegis radars and the Navy Theater Wide interceptor; and the Bush Administration's "GPALS" concept of the early 1990s. 

None of these concepts makes sense today.  By likely provoking strong counterreactions in Moscow, Beijing, and elsewhere, they could worsen nuclear security rather than help it, and lead to a proliferation of countermeasures technology that would likely negate the benefits of mid- course defense systems (like that planned by the Clinton administration) even against countries like North Korea that may not be able to develop and deploy countermeasures themselves in the near to medium-term future.


As one possible system, the Clinton Administration's successor might expand current U.S. plans for NMD further.  For example, the United States could deploy missile defense sites in both North Dakota and Alaska, winding up with a total of 250 interceptor missiles, as well as nine advanced missile-tracking radars on U.S. and allied territory. Setting up two interceptor missile sites in this way is often described as the "C3" option or capability and envisioned for deployment by 2011 by the Pentagon.(40)  It might cost $10 billion more than the C2 option, roughly speaking—for a total of about $50 billion in acquisition costs, all told.(41)

If concerned about the sea-based ballistic missile threat, the United States might also add one or two defensive bases or special, dedicated ships along each coast, for a total of up to a half dozen sites.  Even if it expanded the number of sites further in that way, it would not necessarily need to increase the number of associated interceptor missiles— though it would need new, southward facing radars. 


The Navy Theater Wide defense system would use the Aegis radar deployed on about 60 U.S. Navy cruisers and  destroyers as its engagement radar, and use a modified form of the Standard Missile as its interceptor, with a hit-to- kill final stage atop it.

Of all the current U.S. theater missile defense (TMD) programs, the navy theater wide system is the only one that raises any ABM Treaty compliance issues—and that has any real potential ABM capability.  The treaty permits all theater missile defenses without restriction, but does not clearly define the demarcation point between theater and strategic missile defenses.   In 1997, the United States and Russia reached an accord that defined as TMD any system using interceptors that do not exceed speeds of 3 kilometers per second, and that are not tested against incoming warheads with speeds greater than 5 kilometers per second or ranges greater than 3,500 kilometers.  The demarcation agreement has not been ratified by the Senate, and probably will not be— but it may wind up having some influence over policy in any case, at least temporarily.  The NTW program is to stay below the two latter thresholds, but its interceptor has a maximum speed greater than 3 kilometers per second—in fact, greater than 4 kilometers per second— making its status somewhat ambiguous.(42)   As Ted Postol points out, its speed is sufficient to raise concerns that, if tied in with advanced sensors, it could have some national missile defense capability.(43)  It would in effect have an NMD capability for defending U.S. allies in East Asia or, in most cases Europe, from potential foes in North Korea or the Middle East, respectively.

Seeing this NMD potential as a virtue, in 1999 a group convened by the Heritage Foundation proposed using Aegis- equipped cruisers to form a nationwide defense capability.  It would, in their eyes, have been a near-term way to provide relatively "thick" nationwide defense relatively inexpensively.  (Heritage had been proposing similar ideas in less detail since roughly 1996.) The 1999 commission advocated buying about 650 interceptors to deploy on 22 cruisers (more than half of which would presumably be in U.S. ports or coastal waters at any time, given normal Navy ship rotation schedules), as well as possibly a barge in the Great Lakes or a site in North Dakota.  These interceptors would be tested against long-range missile warheads to be sure they would work against such high- speed threats.  Their radar systems would be interlinked with other sensors such as large early-warning radars and satellites, including the SBIRS-Low (formerly Brilliant Eyes) system.  The Heritage group put out an estimate of some $3 billion for this system.(44) 

The Pentagon's Ballistic Missile Defense Organization did not agree with the Heritage cost analysis, later estimating a total cost of $16 billion to $19 billion for the system.  That included the costs of dedicated ships (since defenses could be badly needed in wartime—when most Navy Aegis warships could be deployed), as well as upgraded missiles with enough range, maneuverability, and nuclear hardness to provide reliable nationwide defense.  (The Pentagon claimed that the planned NTW missiles would not do the NMD job adequately.)(45)  However, the Pentagon did concede that a Navy Theater-Wide system might do a good job of complementing a ground-based system by covering coastal areas better than ground sites in Alaska or North Dakota could, and simply by providing additional radars and interceptors.(46)


Another large national missile defense was President George Bush's proposed Global Protection Against Limited Strikes (GPALS) system.  That design included 1,000 space-based "brilliant pebbles" that would intercept outgoing ICBMs early in their flight (ideally in boost phase) as well as 750 ground-based interceptors at a total of six sites.  Depending on how an enemy attacked, such a defense might be capable of shooting down 100 to 200 warheads with a "leakage" of no more than several warheads getting through.  A defense of this size could be needed to defend against a rogue Russian launch that involved all the warheads on a given submarine or within a given ICBM field. 

GPALS would be very expensive, perhaps approaching $100 billion.(47)   Given its dependence on interceptors that would work in the exoatmospheric, or mid-course, phase of an enemy missile's flight, it would also run the same risks of being virtually useless against a missile force armed with good countermeasures that a much smaller, cheaper system such as the Clinton Administration's would face.  Finally, its space-based elements might not be available as quickly as the above two systems.


The above is a briefing on the state of U.S. missile defense programs, in which I have attempted to describe the technologies and the political debates without interjecting many personal opinions or policy preferences.  But I would like to add several subjective views at this point, most of which I expressed in Tokyo during the trilateral meetings at U.N. University.

The case for missile defense is a complicated one, but in the end I support both TMD and NMD.  However, the latter program needs to be handled extremely carefully, in light of possible repercussions for the U.S.-Russia relationship, for the safety and security of the Russian nuclear deterrent, for the cohesion of NATO, and for the stability of East Asia.  If done carelessly, it is possible to deploy NMD in a way that would make the world, and the United States, less secure rather than more secure.  The Clinton administration's proposed national missile defense system, featuring 100 interceptors to be deployed in Alaska by 2007 or shortly thereafter with the potential to destroy up to a couple dozen enemy warheads, is neither very promising technologically nor ideal for U.S. relations with allies and neutral countries. 

U.S. defense planners, including myself, are generally truly concerned about the North Korean missile threat.  In particular the 100 or so Nodong missiles that could reach Japan would give Pyongyang its only reliable, rapid way to threaten that country in a future crisis or conflict.  North Korea might use that threat in an attempt to convince Tokyo not to allow the United States access to bases on Japanese territory.  This threat might work best if the missiles carried conventional warheads.  If North Korea threatened to attack Japan with weapons of mass destruction, or went so far as actually to do so, it would be subject to U.S. retaliation with nuclear weapons—as well as a likely overthrow of the regime in Pyongyang, the arrest of its leaders, and their trial as war criminals.

Alternatively, in the course of a war provoked by North Korea in which the United States and South Korea had decided to overthrow the DPRK  regime, Pyongyang might even threaten Japan with missile carrying chemical or biological agent or a nuclear warhead (if it really had a nuclear warhead small and robust enough to place on a missile—something that is unlikely but not clearly impossible for North Korea).  At that point, its leaders might feel they had little to lose, and see such a threat as their only remaining hope of negotiating a peace that left them in power (or at worst in asylum).  Of course, if North Korea had an ICBM, it could also directly threaten the United States.  But it could try to paralyze the United States by threatening its allies as well.

Both of these possibilities appear credible to most American defense planners.  Many Chinese would admittedly reach a different conclusion about the seriousness of the risk.  But having lost 28 soldiers to an Iraqi SCUD missile in the 1991 Persian Gulf War, and having watched Israel be bombarded with SCUDs as well, Americans are highly concerned about this threat.  We all hope for the best given the détente process now unfolding in Korea, but do not consider it prudent to count on that process succeeding.  North Korea already has a missile capability against South Korea and Japan, so we should proceed to develop good TMD and deploy it in East Asia promptly.  North Korea is developing an ICBM, so we too should develop NMD—though we may not need to deploy it as quickly as now planned if the situation on the peninsula continues to improve.

As for the Taiwan Strait situation, in my judgment it is understandable that China feels it needs a certain coercive leverage over Taipei.  Absent a real invasion capability, missiles provide China a promising means of gaining that leverage.  However, like most Americans I am also unwilling to accept that China, particularly with its current style of government, should have a large military advantage against Taiwan.  The United States has broader reasons to support Taiwan as well.  It has an historical interest in the island's security, dating back half a century; were it to desert Taiwan now, that would have an impact on the credibility of its commitments around the world.  The United States also has an interest in discouraging Taiwan from developing nuclear weapons by limiting its feeling of insecurity. 

This set of factors makes for a very complicated policy context.  It does not argue against TMD for the United States.  Washington cannot be expected to leave its forces vulnerable to Chinese attack if it does become involved in any future conflict in the region.

Washington should, however, use great care in handling any transfers of further TMD to Taiwan.  Limited, lower- technology defenses make more sense than Aegis destroyers, for example.  One does not wish to embolden Taiwan to act provocatively by a major missile defense sale that makes it feel more secure than it really is, and leads it to act with a feeling of impunity in its dealings with China.  Such an argument for restraint in TMD sales to Taiwan will be unlikely to prevail in the United States, however, if China keeps up its major missile buildup near the strait.

What about NMD in the context of U.S.-China security relations?  China's concern about even a limited U.S. system is understandable, given the modest size of its deterrent.  Even though it has always exercised restraint in its nuclear policies, its modest intercontinental force is a symbol of its great-power status and a deterrent of last resort against the remote possibility of aggressive attack by a foreign power, so it is only natural that Beijing would not wish to lose this deterrent.  That said, one cannot expect the United States to leave itself defenseless against North Korean, Iranian, or Iraqi missiles to satisfy these Chinese concerns, which have less to do with Beijing's real security worries than with perceptions of its international stature. 

Some Americans, most of them Republican, would also like a robust  defense against China if they could have it.  Concerned about Taiwan's security, these conservatives—perhaps constituting around a quarter of the overall American Congress and public—would be delighted to have a perfect defense against the entirety of the PRC nuclear force, whatever its size.  At one level, their concern is understandable, given China's frequent threats against Taiwan, its strong-armed techniques at Tiananmen and in Tibet, and its occasional elliptical threats against the United States itself.  On the other hand, such American aspirations are almost surely unrealistic, given China's ability to build up its force, develop midcourse decoys, and if necessary consider other means of delivering weapons of mass destruction against the U.S. homeland.

In the end, I believe the United States should build a light NMD.  Preferably, the defense should emphasize boost- phase interceptors that would have little capability against Chinese or Russian missiles.  A modest number of midcourse interceptors, like those now envisioned under the Clinton Administration proposal, might be considered as a way to handle accidental or unauthorized launch of at most a few missiles.  But midcourse interceptors should not be heavily relied upon, since even a country like North Korea may be capable of developing decoys that could defeat them. 

Given the improvement in inter-Korean relations, the desirability of working out an accord with Russia on NMD, and the need to consult further with friends and allies before deploying defenses, I favor delaying any NMD deployment decision at least until 2002 and probably longer.  I also support most types of TMD, though the technical capability of NTW to overcome countermeasures puts its viability in some doubt, and my current preference is not to sell NTW to Taiwan.

As for China, it is only natural to expect it to modernize and increase the size of its long-range nuclear arsenal somewhat.  But my expectation is that China would be wiser to develop exoatmospheric decoys, to defeat both NTW and the currently planned U.S. NMD system if necessary, than to plan on defeating these defenses simply through large-scale saturation attacks.


1. Bradley Graham, "Army Hit in New Mexico Test Said to Bode Well for Missile Defense," Washington Post, March 16, 1999, p. 7; James Glanz, "Missile Defense Rides Again," Science, April 16, 1999, p. 417; David Hughes, "Patriot PAC-3 Upgrade Aimed at Multiple Threats," Aviation Week and Space Technology, February 24, 1997, pp. 59-61. 

2. Robert Holzer, "U.S. Navy Missile Defense Effort Wins Support," Defense News, July 26, 1999, p. 3. 

3. Bradley Graham, "Low-Tech Flaws Stall High-Altitude Defense," Washington Post, July 27, 1998, p. 1. 

4. The first intercept occurred at White Sands Missile Range, New Mexico, at an altitude of just under 60 miles, with both target and interceptor traveling at about 1.5 kilometers per second on impact.  The second occurred a few miles higher, but at considerably higher interceptor speed (2.5 kilometers per second), and against a reentry vehicle descending from 300 kilometers' altitude.  See "World News Roundup," Aviation Week and Space Technology, June 14, 1999, p. 56; Robert Wall, "THAAD at Crossroads After Intercept," Aviation Week and Space Technology, August 9, 1999, pp. 29-31. 

5. Kerry Gildea, "Theater missile defense programs in trouble, top Pentagon official warns," Aerospace Daily, May 25, 1999; Geoffrey E. Forden, "The Airborne Laser," IEEE Spectrum (September 1997); John Donnelly, "Basis for Pentagon Approval of Airborne Laser 'Questionable,'" Defense Week, March 15, 1999, p. 1. 

6. General Accounting Office, Defense Acquisitions:  DoD Efforts to Develop Laser Weapons for Theater Defense, GAO/NSIAD-99-50 (March 1999).

7.  The space-based laser program is probably at least another decade in the future; see General Accounting Office, Defense Acquisitions:  DoD Efforts to Develop Laser Weapons for Theater Defense, p. 20.

8. Craig Cerniello, "QDR Supports Nuclear Status Quo, Adds Billions More to NMD Program," Arms Control Today (May 1997), p. 29. 

9.  Richard L. Garwin, "A Defense that will not Defend," Washington Quarterly, vol. 23, no. 3 (Summer 2000), pp. 213-224.

10. Zakheim, "Old Rivalries, New Arsenals," IEEE Spectrum, pp. 30-31. 

11.  George J. Tenet, Director, Central Intelligence, before the U.S. Senate Committee on Armed Services, February 2, 1999; Richard L. Garwin, "The Wrong Plan," Bulletin of the Atomic Scientists, vol. 56, no. 2 (March/April 2000), pp. 36-41. 

12.  William S. Cohen, "FY 2000 Defense Budget: Briefing Slides," Department of Defense, February 1999. 

13.  Roberto Suro, "Missile Sensor Failed in Test's Final Seconds, Data Indicate," Washington Post, January 20, 2000, p. 4. 

14.  Geoffrey Forden, Budgetary and Technical Implications of the Administration's Plan for National Missile Defense (Washington, D.C.:  Congressional Budget Office, April 2000), p. 10.  The administration and CBO cost estimates differ by only about $2 billion, or just over 10 percent.

15.  See Michael O'Hanlon, "Defense and Foreign Policy:  Time to End the Budget Cuts," in Henry J. Aaron and Robert D. Reischauer, eds., Setting National Priorities:  The 2000 Election and Beyond (Washington, D.C.:  Brookings, 1999), pp. 37-72.

16.  Forden, Budgetary and Technical Implications, pp. 5-17; Briefing by The Honorable Walter B. Slocombe, Under Secretary of Defense for Policy, "U.S. Limited National Missile Defense Program," Harvard-CSIS Ballistic Missile Defense Conference, Cambridge, Mass., May 11, 2000.

17.  Testimony of the Honorable Walter B. Slocombe to the House Armed Services Committee, October 13, 1999.

18.  Roberto Suro, "Missile Sensor Failed in Test's Final Seconds, Data Indicate," Washington Post, January 20, 2000, p. 4; Bradley Graham, "U.S. Anti-Missile Test Is Latest in String of Successes," Washington Post, October 4, 1999, p. 1; Richard L. Garwin, "A Defense that will not Defend," Washington Quarterly, vol. 23, no. 3 (Summer 2000), pp. 213- 224.

19.  Statement of Kadish before the Senate Armed Services Committee, February 28, 2000, p. 6; Forden, Budgetary and Technical Implications, p. 25.

20.  Tony Capaccio, "Pentagon's Top Tester Warns on Missile Defense Time Line," Defense Week, June 5, 2000, p. 1.

21.  George N. Lewis and Theodore A. Postol, "Future challenges to ballistic missile defense," IEEE Spectrum (September 1997), pp. 60-68.

22.  See Welch et. al., "Report of the Panel on Reducing Risk in Ballistic Missile Defense Flight Test Programs," p. 56; Elaine M. Grossman, "Rumsfeld Commission Member Sticks to Guns on Opposing Defenses," Inside the Pentagon, July 30, 1998, pp. 19-20. 

23.  David C. Wright, Testimony on the technical readiness of national missile defenses before the U.S. Senate Committee on Foreign Relations, May 4, 1999; Richard L. Garwin, testimony before same committee on same date. 

24.  See Lieutenant General Ron Kadish, "Clearing the Fog:  Eliminating Misconceptions in the Debate about Deploying a Limited National Missile Defense System," Armed Forces Journal International (June 2000), p. 59.

25.  See Union of Concerned Scientists and MIT Security Studies Program, Countermeasures (April 2000).

26.  Union of Concerned Scientists and MIT Security Studies Program, Countermeasures, pp. 145-148.

27.  Defense Science Board 1998 Summer Study Task Force, Joint Operations Superiority in the 21st Century (Washington, D.C.: Department of Defense, 1998), pp. 97-100; Office of Technology Assessment, Proliferation of Weapons of Mass Destruction (Washington, D.C., 1993), p. 52. 

28.  Bob Drogin and Tyler Marshall, "Missile Shield's Destabilizing Potential Cited," Los Angeles Times, May 19, 2000, p. 1.

29.  See Richard L. Garwin, "The wrong plan," Bulletin of the Atomic Scientists, vol. 56, no. 2 (March/April 2000), pp. 36-41); Richard L. Garwin, "A Defense that will not Defend," Washington Quarterly.  Garwin estimates the weight of such a boost-phase interceptor, which need not carry a heavy payload, at 14 tons; by contrast, existing U.S. ICBMs and SLBMs generally weigh 30 to 100 tons. 

30.  Theodore A. Postol, "A Russian-US Boost-Phase Defense to Defend Russia and the US from Postulated Rogue- State ICBMs," Massachusetts Institute of Technology, Cambridge, Mass., briefing paper presented at Carnegie Endowment for International Peace, Washington, D.C., October 12, 1999.

31.  Presentation by Richard Garwin, Harvard-CSIS Ballistic Missile Defense Conference, Cambridge, Mass., May 11, 2000.

32.  See for example Thomas B. Cochran, William M. Arkin, and Milton M. Hoenig, Nuclear Weapons Databook, Volume I:  U.S. Nuclear Forces and Capabilities (Cambridge, Mass.:  Ballinger Publishing Co., 1984), p. 145.

33.  The Martin-Marietta Corporation (now part of Lockheed Martin) prepared such a study for the Fletcher Panel, an official advisory group evaluating technologies for national missile defense for the Pentagon.  See John Tirman, ed., The Fallacy of Star Wars (New York:  Vintage Books, 1984), p. 62.

34.  Theodore Postol, "Hitting Them Where It Works," Foreign Policy, no. 117 (Winter 1999-2000), pp. 132-133.

35.  Congressional Budget Office, The START Treaty and Beyond (October 1991), p. 139; Geoffrey Forden, Budgetary and Technical Implications of the Administration's Plan for National Missile Defense (April 2000), p. 12.

36.  Geoffrey Forden, Budgetary and Technical Implications of the Administration's Plan for National Missile Defense (April 2000), p. 10.

37. Richard L. Garwin, "Effectiveness of Proposed National Missile Defense Against ICBMs from North Korea," March 17, 1999, available at www.fas.org/rlg. 

38.  Alessandra Stanley, "Putin Goes to Rome to Promote Russian Arms Control Alternative," New York Times, June 6, 2000, p. 1. 

39.  Theodore Postol, "Hitting Them Where It Works," Foreign Policy, no. 117 (Winter 1999-2000), pp. 132-133.

40.  Lieutenant General John Costello, Army Space and Missile Defense Command, "National and Theater Missile Defense," briefing slides presented at the Association of the U.S. Army, Arlington, Va., May 17, 1999, available at www.smdc.army.mil.

41.  Charles V. Pena and Barbara Conry, "National Missile Defense:  Examining the Options," CATO Policy Analysis No. 337, Cato Institute, Washington, D.C., March 1999, p. 19.

42. Lisbeth Gronlund, Letter to the Editor, Arms Control Today (June/July 1998), p. 36; John Pike, "Ballistic Missile Defense: Is the U.S. 'Rushing to Failure?'" Arms Control Today (April 1998), p. 10. 

43.  See the chapter of which Postol was primary author in Harold A. Feiveson, ed., The Nuclear Turning Point (Washington, D.C.: Brookings, 1999), pp. 86-89. 

44.  Ambassador Henry Cooper and the Heritage Foundation's Commission on Missile Defense, Defending America:  A Plan to Meet the Urgent missile Threat (Washington, D.C.:  Heritage Foundation, 1999), pp. 36-58.

45.  Walter Pincus, "Estimate Skyrockets for Expanding Navy's Ship-Based Missile Defense," Washington Post, March 5, 1999, p. 4; Statement of Lieutenant General Lester Lyles, Director, Ballistic Missile Defense Organization, before the Subcommittee on Strategic Forces, Senate Committee on Armed Services, February 24, 1999, pp. 11-14.

46.  See Ballistic Missile Defense Organization, "Summary of Report to Congress on Utility of Sea-Based Assets to National Missile Defense," Department of Defense, Washington, D.C., June 1, 1999.

47. David Mosher and Michael O?Hanlon, The START Treaty and Beyond (Washington, D.C.: Congressional Budget Office, 1991), pp. 69-70, 167- 171.