A Flight Test Ban as a Tool for Curbing Ballistic Missile Proliferation ⁽¹⁾

INTRODUCTION

Limitations on the testing of United States and Soviet ballistic missiles were considered throughout most of the history of the cold war (see Appendix C). Explicit and implicit restrictions were eventually adopted in the ABM, SALT II, INF, and START treaties. In this paper, I ask whether missile flight test limits (or a ban) are a useful tool to slow (or halt) missile proliferation, and whether such limitations are feasible in the near term. Several issues concerning the feasibility of flight test bans are considered here:

How effective would a flight test ban be in limiting ballistic missile development?

Could a ban be verified with a high degree of confidence?

What complications would be introduced by flights of space launch vehicles?

What would be the `costs' of a comprehensive flight test ban to the great powers, in terms of their force modernization plans and need for reliability testing?

Are the United States and Russia (perhaps also Europe and China) sufficiently concerned about developing country missiles and each others' missiles that they would be willing to give up testing? Are countries of real concern likely to develop long-range missiles in the next several decades?

Would agreement by the declared nuclear powers to forgo missile tests convince third world countries that they too should forgo developing ballistic missiles, given the military force inequality that would remain?

What would be the implications of less than universal adherence?

I finish up by exploring the various possible configurations flight test limitations could take.

FLIGHT TESTING

Flight testing a ballistic missile is not simply a matter of firing off a missile and watching through binoculars. The United States' two long-range test facilities (one on the East coast and one on the West) are very expensive and complex. The estimated cost of replacing the US Army's Kwajalein Atoll facility (an integral part of the Western Test Range, where all US intercontinental-range ballistic missiles (ICBMs) and ballistic missile defenses are tested), is $2 billion.⁽²⁾ The fiscal year 1994 funding request for operating and modestly upgrading the facility at Kwajalein was $166.6 million (down $10.2 million from the year before), and total employment is about 3,000.⁽³⁾

The United States' Eastern Test Range, where submarine launched ballistic missiles (SLBMs) are flight tested and where intermediate-range ballistic missiles (IRBMs) were tested before they were outlawed in the 1987 Intermediate-range Nuclear Forces (INF) Treaty, is also very technologically advanced and expensive. The complex includes IRBM and SLBM launch pads, tracking and telemetry stations in Florida and Caribbean islands, and a down-range instrumented terminal area at Ascension Island.⁽⁴⁾ The facility was upgraded in the mid-1980s in anticipation of the current Trident II D5 missile test program.⁽⁵⁾

The Soviet Union possessed similar flight testing ranges, with launch centers primarily at the Tyuratum Cosmodrome in Baikonur, Kazakhstan and secondarily at the Plesetsk Cosmodrome in Northern Russia. Appendix A contains a preliminary listing of global flight testing and space launch facilities.

The US Flight Testing Model. A ballistic missile lofts its payload to altitude and velocity in a powered boost phase, and then releases it to continue on an unpowered, unguided course. The key components of a ballistic missile are the propulsion, guidance and control systems, and the warhead or re-entry vehicle. Building a ballistic missile---perfecting and integrating these components---especially in a long or intercontinental-range ballistic missile, is a complex and daunting technological task.⁽⁶⁾

The basic design of all ballistic missiles was already defined in 1942, in the German V-2 missile. Nonetheless, materials, manufacturing and instrumentation used in missiles have improved greatly in the ensuing half century. Advances such as longer range through multi-staging, high-speed/low-drag reentry vehicles, multiple independently-targeted reentry vehicles (MIRVs), solid-fuel, and successive generations of inertial guidance have all relied heavily on flight testing. The need for some flight testing in the development of any complete, modern missile system is indisputable. As Farooq Hussain (a test ban skeptic) concedes: "[C]ertain problems---such a those associated with the prediction of ballistic trajectory bias, MIRV manoeuvering and warhead re-entry into the atmosphere---can only be resolved confidently by actual flight tests."⁽⁷⁾

The US Navy and Air Force put their missiles through an elaborate testing sequence. First, they conduct technology/component tests through supplementary flight testing (SFT) of new components on old boosters. Midgetman and Trident II components, for example, were tested aboard Minuteman ICBMs. The services rely heavily on SFT to develop reentry vehicles and guidance technology.

Next are research and development tests, carried out under idealized conditions: Engineers and technical contractors fire missiles from a launch pad (rather than silos or subs) on days when the weather conditions are most suitable. Later in the R&D process more realistic conditions are used. Typically, twenty to thirty flight tests of this sort are conducted for a new design.⁽⁸⁾

Then early production line models undergo a series of initial operational tests (or phase one operational tests) under more realistic launch conditions. These tests are used to estimate system reliability and accuracy. Thirty or forty flights are usually needed to achieve the level of confidence desired by nuclear war planners.

Next come so called `Demonstration and Shakedown Operational' (DASO) tests. Most of the tests in this category are of SLBMs, with test firings from each submarine; these are intended as much to test the sub and the crew as the missile. (For Air Force missiles, DASO tests follow R&D and precede initial operational tests.)

Follow-on tests (FOT---also called phase two operational tests) are carried out at lower rates over the life of the missile system: 1) to detect deterioration over time; 2) to check out modifications; 3) to maintain crew training and readiness; and 4) to maintain confidence and to display system performance for deterrent effect. The number of FOTs has been around six per year for ICBMs and more for SLBMs.^(9),(10) In addition, there is a regular program of aging and surveillance testing, using X-ray and other inspection techniques, static firing of stages, and testing of subsystems.

Third World Missile Testing. The procurement route, range and sophistication, mission and payload of the delivery vehicle all dictate particular flight testing patterns. For several reasons, developing country testing programs are not nearly as sophisticated or extensive as that of the United States.^(11),(12) A primary limiting factor is cost. A testing infrastructure is expensive,⁽¹³⁾ and so are the missiles expended in tests. Many developing countries' missile inventories are wholly imported, and it is increasingly difficult to find resupply because of the emerging norm against missile exports. A meaningful test program could easily deplete the limited missile supply of a developing country.

Second, given that the vast majority of developing country ballistic missile systems have been imported, flight testing is less necessary. Thirteen countries have imported the Soviet Scud-B.⁽¹⁴⁾ The Scud is a simple, proven design, based originally on the V-2. It does not require tight tolerances in its manufacture and handling and, therefore, perhaps a purchasing country could deploy it with little or no testing.⁽¹⁵⁾

^,(16) Similarly, Saudi Arabia purchased an estimated 50 CSS-2s from China. No operational flight tests of this 2,400 km range missile from Saudi soil have been reported.

In addition, most developing country missiles are of short range and conventionally armed. Possessor countries have used them as deterrents, or as counter-city weapons of terror. Neither mission requires extensive testing to achieve the high degree of accuracy and reliability needed for counter-silo nuclear weapons.

Because of the paucity of testing by newly ballistic-missile-capable countries, some analysts have asserted that testing restrictions would be of little utility in curbing missile proliferation.^(17),(18) However, the third world missile development or upgrade programs of greatest concern---those aimed at achieving accurate inertial guidance, solid fuel and multi-staging---must flight test. And, in fact, those few developing countries that are pursuing long range missile or space launch capabilities (Brazil, India and Israel) have serious, methodical flight testing programs, albeit with fewer flights and at less cost than for the superpower programs.

For example, the Indian military has tested its Prithvi⁽¹⁹⁾ missile twelve times in the past four years (see Appendix B). The first test occurred on 25 February 1988 at the Indian Space Research Organization's SHAR Centre (on Sriharikota Island). After five more tests there, Prithvi has since been tested from the military's Chandipur interim test range in Orissa, most recently in late November 1993.^(20),(21) The Indian military will likely deploy Prithvi during 1994. India is putting its longer range ballistic missile, the Agni, and its space launch vehicles through a similar steady progression of tests.

Israel has the most highly developed defense industry in the Middle East and the most advanced military missile production capability outside of the former Soviet Union, United States, France and China. During the 1970s-1980s Israel developed an improved version of its imported Jericho missile, dubbed the Jericho II. (Many reports claim Israel is developing two separate missile systems, the Jericho II with a 800 km range and the Jericho IIB with an extended 1,300 km range.) The missile has successfully flown several times (see Appendix B). Most recently, on 14 September 1989, Israel fired a Jericho II 1,300 km into the Mediterranean. In addition to tests in Israel, two long-range tests of the Jericho II are believed to have occurred in South Africa during 1989-1990.^(22),(23) Reportedly, about 50 Jericho II missiles have been deployed.⁽²⁴⁾ Israel has also twice tested successfully an indigenously developed space launch vehicle, the Shavit.

As the Indian and Israeli military establishments know, zero flight testing of a missile under development will result in zero confidence that the system is functional. Moreover, achieving an acceptable degree of confidence in the reliability of a system, and characterizing its accuracy and performance under varying conditions, requires operational flight testing, the amount of which varies with the amount of information, and statistical confidence in that information, one desires to have.

Although media reports often refer to the `improved accuracy' of third world missiles, without a significant and highly visible testing program, such claims must be treated with skepticism. The measure of accuracy, circular error probable (CEP)⁽²⁵⁾, cannot be determined by a single test; CEP can only be estimated by firing a substantial number of missiles at predetermined aim points. Accuracy can be compromised by subtle imperfections in machining, calibration or system engineering, and most developing countries do not have or do not produce missiles in quantities sufficient to support testing at the rates required to assess progress in the refinement of guidance systems, or even to iron out all the bugs and glitches that may cause catastrophic failure.^(26),(27)

The absence of testing, therefore, may be considered prima facie evidence that some alleged missile capabilities are non-existent. Developing country missile programs have often been exaggerated for political reasons---both by the alleged proliferators (for reasons of deterrence or prestige) and by developed countries (to justify certain military programs and to support arms sales to allies). A clear distinction must be made between the `real' developing country missile programs (like those in Israel and India) and those chimeric programs which appear to lack rigorous (or in some cases apparently any) testing.

Most notable in this regard, perhaps, are exaggerated, or at least unproven, claims of North Korean ballistic missile prowess. The Pyongyang government imported Scud missiles from Egypt in the mid-1970s and began producing its own version of the 300 km missile in 1987. In the late 1980s and early 1990s, a number of sources reported that North Korea was refining the Scud to increase its range to 600 km, and to improve its guidance.⁽²⁸⁾ According to the current Director of Central Intelligence, James Woolsey, now North Korea "is developing and actively marketing a new, 1,000 kilometer-range missile,"⁽²⁹⁾ called the Nodong I. The missile appears to have been successfully flight tested once (in May 1993 to a range of 400-500 km) and unsuccessfully tested once (see Appendix B). As it is allegedly a further modification of the Scud,⁽³⁰⁾ few tests might be needed, but more than one flight test would be expected for a serious weapons program. Various press reports claim that the system is either in late development and will be deployed and exported to other countries shortly or that it already has been. (Following on the heels of the May 1990 test launch, press reports claimed that North Korea intended to extend the range of this missile to 1,300 km.⁽³¹⁾) The functionality, let alone reliability and accuracy, of North Korean-made missiles is uncertain, at best. According to one report, Scud missiles manufactured by North Korea and shipped to Iran in the early 1990s were inoperable, and Iran returned the missiles.⁽³²⁾

Similarly, allegations of missile production by several countries in the Middle East are not supported by available information on flight tests. Whether this is because tests are not occurring, are not being observed or not being reported publicly is difficult to determine. What is clear, is that in the past few years the US military has diverted increasing intelligence assets to cover developing countries and regions considered dangerous.^(33),(34)

ALTERNATIVES TO TESTING

If flight testing were limited or prevented by some control regime, and if suppliers refused to transfer full-up ballistic missile systems, then could a developing country---or even a developed country---achieve an operational ballistic missile capability?

The number of flight tests needed to develop and obtain confidence in a ballistic missile is decreasing over time. (See figure 1, which shows flight testing trends for various US ICBMs.) This decline is due to the accumulation of knowledge and the development of improved alternative techniques for evaluating missile systems. Better instrumentation and analysis methods are applied to static firings of the rocket motors; and more sophisticated simulations of launch and flight are applied to workouts of the guidance and control systems. In addition, continually increasing computational capabilities streamline the development process by aiding design, development and evaluation.

Political considerations have also driven down the numbers and restricted the possibilities for testing. For example, overflights of the continental United States are problematic, and a static testing procedure called Simulated Electronic Launch of Minuteman was introduced to compensate partially for the lack of flight tests fired from active-duty silos.⁽³⁵⁾

Intangible factors such as morale (in turn, affected by the threat perception) of the missile work force, organizational structure and even culture may affect the number of tests needed. A common cause of failure in complex technical projects is poor coordination and communication between teams which produce different subsystems that must work together. A better-educated and better-equipped work force may also be better able to circumvent difficult technical problems and minimize testing.

The transfer of knowledge from non-military space activities to military missiles may also reduce flight testing requirements. But it is easy to overstate the adaptability of civilian space technology to military requirements. Farooq Hussain states that, "The development of very reliable launchers, for both satellites and manned spacecraft, with a minimum number of flight tests has always been the philosophy of NASA....The United States' aerospace industry has now learned the methods by which very high reliability can be attained essentially without a flight-test requirement or with only a nominal one."⁽³⁶⁾ The trick, according to Hussain, is to "introduce a large amount of redundancy in back-up systems." However, this "very high reliability" is purchased only at a cost that would be prohibitive for most military systems. Redundancy means dead weight, and a multiplication of system cost and complexity. A typical commercial launch is carried out only under the best weather conditions, and only after the rocket has been examined by an army of technicians, and a committee of engineers has given the word `Go.' Moreover, the continuing series of unmanned launches, using rockets with long histories such as Atlas, Titan, and Delta, is itself a de facto testing program from which information is obtained to improve the rockets and their operational use. Major upgrades of these systems have been accompanied by an expensive and embarrassing series of failures. Nor has the manned space program been without its disasters and learning curves. After the tragedy of Apollo 1, a series of unmanned launches of the Saturn vehicles was carried out before manned flight resumed; and even the space shuttle, with all its redundancy and sensors monitoring every component, has experienced one catastrophe and many mishaps.

Robert Sherman notes that "The history of missile development is replete with examples of new missiles and new technologies which performed well in computer simulation and ground testing, but which revealed unpredicted---and probably unpredictable---fatal defects in flight testing."⁽³⁷⁾ Of the eight new strategic missiles first tested in the 1980s (MX, Trident II, Pershing II, SS-24, SS-25, SS-N-20, SS-N-23 and an SS-18 follow-on), all but two failed their first flight tests. The MX missile's inertial guidance system performed "brilliantly" in early development tests, but its accuracy fell off when the production team took over production of the missiles from the development team, according to Sherman.

A recently revealed report on Soviet ICBM development demonstrates the requirement for extensive static testing and operational flight testing. Apparently the Soviets had persistent difficulties with hydro-thermodynamic instabilities in their liquid-fuel engines. Production-model engines that worked most of the time would unpredictably exhibit these instabilities. Sometimes the engines failed catastrophically. No systematic differences between the engines that worked and those that failed could be detected, and despite the efforts of hundreds of scientists, the effects could not be reproduced in the lab. Ad hoc solutions were sometimes found, but the scientists could not obtain any systematic control over the phenomena. According to the author of the report, a former Soviet rocket engineer, this problem explains why the Soviets relied for so long on clusters of small, inefficient engines on their booster rockets, instead of moving to the larger, more efficient engines used by the United States. Their solution was to mount sensors in the engines that would shut them down at the first sign of trouble, then the rocket would fly on using the remaining engines.

This example clearly illustrates the kinds of difficulties that may arise unexpectedly in the development of missile systems, and may go undetected in the absence of a rigorous testing program. As it turned out, the clustering fix masked a systematic flaw. During the winter of 1965-6, the Strategic Rocket Forces undertook a standard test of a deployed strategic missile. The nuclear warhead was replaced with a dummy, and the missile was transported to a space facility for the launch toward the Pacific. The operational missile, which had been serially produced in the thousands, exploded on the launch pad. Hydrodynamic instabilities in the fuel feed lines caused the explosion. The oscillations turned out to be associated with a narrow range of air temperatures, around -30 degrees C. The engine had passed ignition tests at 40, 30, 20, 10, 0, -10, -20 and -40 degrees C, but had never been tested at -30.⁽³⁸⁾

If no flight testing is permitted, then every new weapon or component risks catastrophic failure with a high probability.

INCENTIVES FOR THIRD WORLD COUNTRIES

The present response by the developed world to the spread of missile technology is an export control/proliferation management regime, combined with a `technical fix' in the form of anti-missile systems. The Missile Technology Control Regime (MTCR), initiated in 1987 with seven members, has grown to include 25, mostly Western industrialized countries.⁽³⁹⁾ MTCR members and adherents pledge to abide by common export guidelines on missile-relevant technologies and missiles themselves. Although a factsheet on the MTCR issued by the US government in 1987 said that the guidelines "are not designed to impede national space programs or international cooperation in such programs as long as such programs could not contribute to nuclear weapons delivery systems,"⁽⁴⁰⁾ the regime seems to have acquired the goal of doing precisely that. By definition, any space launch vehicle could contribute to a ballistic missile that conceivably could deliver a nuclear payload.

At the same time, many countries of the North, alarmed about the perceived spread of missile capabilities, are developing or purchasing anti-missile systems. The US government plans to spend $3.2 billion in Fiscal Year 1995 to develop ballistic missile defenses, and the United States and Russia are aggressively marketing their tactical anti-missile missiles to countries in the Middle East and East Asia.

Unfortunately, the MTCR and missile defenses do not address the security concerns of the `proliferators.' These policies have therefore failed to eliminate the demand for missiles, borne of regional political tension and local arms races.

Janne Nolan has argued that even if the big military powers agreed to forgo missile flight testing in order to curb missile proliferation, developing countries would not find this offer all that compelling, given the disparities that would remain in the size and capabilities of military arsenals.⁽⁴¹⁾ Yet, assuming non-transfer of further missiles, developing countries would gain some palpable security benefits through a flight test ban. Superpowers like to think of ballistic missiles primarily as deterrents. But in the developing world, missiles have been used recently and extensively against cities, mainly as weapons of terror and attrition.⁽⁴²⁾ A testing ban would immediately improve the security environment of many countries by halting costly and destabilizing regional missile races. If the ban extended down to missiles with a range of 100 km/500 kg payload, then the benefits would be even greater, as deployed short-range (yet often `strategic') systems would be `rusted out' over time.

Such a regime would not equalize the global military imbalance. In particular, the United States and Russia would retain massive conventional superiority as well as massive nuclear capabilities. But in time, these arsenals would also dwindle through lack of operational readiness training, gradual loss of confidence, and (hopefully) eventual strategic irrelevance.

The entire world would benefit by decreasing the chance of accidental or intentional nuclear war. A flight test ban should also alleviate the perceived need for anti-missile systems, lessening global tensions and freeing up vast resources that would be spent to develop and deploy such systems.

Certain developing countries would also be relieved of anxiety about the United States and Russia re-targeting ICBMs on them. A Pentagon report leaked to the press in January 1992 suggested that in the post cold war world, "every reasonable adversary"---some presumably in the developing world---should be targeted with nuclear and non-nuclear strategic weapons.⁽⁴³⁾ As part of its recent `counter-proliferation' initiative, the US Department of Defense is reportedly considering fitting some Trident II D5 missiles with small nuclear weapons,⁽⁴⁴⁾ as well as with conventional warheads. Rear Admiral Thomas Ryan, the director of the US Navy's submarine warfare division, argues that the latter is needed as a credible deterrent against certain developing countries.⁽⁴⁵⁾ A long-range, kinetic energy penetrator is intended to destroy underground command and communication bunkers of potential (third world) adversaries. This wild plan appears to be driven by the search for a new mission for the D5, which was to have been targeted primarily on the hardened SS-18 silos, now scheduled to be eliminated under START II.⁽⁴⁶⁾ The bunker-busting mission would require accuracy of 5-7 meters, which could not be achieved without testing;⁽⁴⁷⁾ on 18 November 1993 the Navy conducted a classified test of a D5 missile equipped with at least two conventional warheads from a Trident submarine off the Florida coast.⁽⁴⁸⁾

US Navy officials are touting a conventional SLBM as an `anti-proliferation weapon.' But it is likely that this current (and disturbingly recurrent) talk of converting ICBMs or SLBMs to engage third world targets from intercontinental range will motivate developing countries to pursue their own long range missile development. In addition, the development of ultra-high accuracy needed for conventional SLBMs could destabilize the US-Russian nuclear relationship and re-energize the qualitative nuclear arms race.

GREAT POWER ISSUES

Undoubtedly, flight testing restrictions would hamper and even make impossible the spread of long range missile capability. But the major military powers may be unwilling to forgo testing to achieve this end. Nolan states that a flight test ban (FTB) "has never been considered seriously by [the five declared nuclear states'] governments," and that "the notion that the superpowers and the NATO allies would abandon missile flight testing in the hopes of persuading third world countries to follow suit lacks credibility."⁽⁴⁹⁾ In this section, therefore, I analyze the crucial issue of the interests of current ICBM powers.

Russia. Some Russian legislators perceive START II to be unfavorable to Moscow. Others worry that the costs of the treaty are too high. They estimate that it will cost Russia $5 billion to comply with START II and to reconfigure its forces, including producing new ICBMs not prohibited by the treaty.⁽⁵⁰⁾ The SS-25 is the only ICBM currently produced in Russia.⁽⁵¹⁾ US intelligence sources reportedly expect that Russia will flight test and deploy one or two follow-ons to the SS-25 and to the SS-N-20 SLBM sometime later this decade.⁽⁵²⁾ The Russian press reported in early April 1993 that the Moscow Thermo-Engineering Institute and the Dnepropetrovsk `Yuzhnoye' Science and Production Association are developing a new "multipurpose" ICBM. According to the report, flight tests will begin in 1994.⁽⁵³⁾

According to CIA Director Woolsey, "Russia's willingness to fulfill START II requirements will depend, in part, on its ability to modernize its remaining forces to make them viable into the next century and to ensure that it remains a strategic super-power."⁽⁵⁴⁾ On the other hand, some $1.2 billion in so-called `Nunn-Lugar' monies are being provided to assist in the denuclearization of the country, but continued strategic modernization is undercutting political support in the US Congress for more assistance.

A converted SS-25 underwent an initial test launch as a space launch vehicle from the Plesetsk Cosmodrome on 25 March 1993. The commercial launcher is known as the `START I'.<sup>(55)</sup> In addition, a Washington, DC-based group of investors, headed by former Chairman of the Joint Chiefs Admiral Thomas Moorer, has also contracted with a Russian company to develop a mobile commercial satellite launcher. Their so-called `Surf' vehicle would incorporate elements from deactivated SS-N-23 and SS-N-20 SLBMs and would be launched from mobile floating platforms. The first demonstration launch is anticipated in 1994, if the project receives approval from the US government.⁽⁵⁶⁾ Several other planned conversions are underway.⁽⁵⁷⁾ Such schemes would complicate a ballistic missile flight test ban, but they do not present insurmountable obstacles.

United States. A flight test ban would preclude the continuation by the United States and Russia of the race for exotic first strike weapons,⁽⁵⁸⁾ such as high-accuracy usable capabilities, defenses, depressed trajectory^(59),(60)/short time of flight weapons, maneuvering reentry vehicles (MaRVs) and precision-guided RVs.

A ban would also erode the reliability and confidence essential to first-strike planning.⁽⁶¹⁾ As Robert Sherman has argued, "statistical analysis can demonstrate that deterrence and stability are highest when strategic missiles on both sides are `semi-rusted'---that is, when only about 30-70 percent can be expected to work properly....Less-than-perfect reliability discourages aggression more than it impairs deterrence."⁽⁶²⁾ However, in spite of this common-sense logic, supporters of an FTB will collide head-on (as do supporters of a nuclear test ban) with proponents of weapons safety and reliability testing. In response to Sherman, Walter B. Slocombe argued that reliability problems would "breed concerns about massive undetected problems with the force, which would foster pressures for early use or, more probably, for abandoning the flight test ban.... Implicitly, a ballistic missile flight test ban reflects the view...that nuclear weapons modernization itself is the chief source of danger."^(63),(64) It is difficult to know what Slocombe means by "pressures for early use," but one need not be wedded to the view that modernization is the wellspring of crisis instability to agree that it can be a source of danger. Moreover, it is hard to envisage any arms control agreement which does not pose the "danger" that it will someday be broken.

With respect to operational reliability, a chart at the recent authorization hearing for the Trident II program noted that "Accuracy, reliability and safety can only be objectively evaluated, verified and predicted by flight test....The ability to detect and correct an unacceptable degradation in accuracy, reliability or safety is required."⁽⁶⁵⁾ This statement confuses the information that is obtained in early development and evaluative test series with that obtained later in continuing operational tests. The latter are not conducted at a rate sufficient to detect the changes that might be expected in accuracy or in reliability rates. A sudden and unexpected degradation probably would remain undetected for a considerable period if testing were relied upon to reveal it. But in fact, the primary guard against such deterioration is inspection and non-flight testing of the operational missiles.⁽⁶⁶⁾

Farooq Hussain argues that an FTB would impede many "desirable technological improvements to existing systems which contribute to increased survivability."⁽⁶⁷⁾ Contributions to survivability, either against first-strike attack or against ballistic missile defense systems, do contribute, in principle, to stability. But only improvements in survivability against defenses depend critically on flight testing.⁽⁶⁸⁾ Thus if defenses remain limited under the ABM Treaty, we need not fear an erosion of crisis stability due to an FTB.

Finally, some in the US defense establishment maintain that an end to testing would

be detrimental to US security, in that it would lead to greater uncertainty about other countries' missile capabilities. During flight tests, missiles transmit a stream of electronic data on the missile's performance to monitors on the ground. Interception of this data, called telemetry, reveals the capabilities of the missile. Denial of telemetry and of visual information through a testing ban, it is said, would lead to worst case threat analyses, which would spur arms races.⁽⁶⁹⁾ However, worst case analyses prevail anyway, without testing restrictions. The trade off here is one of curtailing missile development through a negotiated test ban, or allowing development to continue simply so that it can be monitored.

There are strong political interests in the United States committed to continuing ballistic missile testing. Continued testing is driven by: 1) bureaucratic self-maintenance of the US `missile-lab complex'; 2) the market provided by testing for the US aerospace industry; 3) the quest to develop missile defenses; and 4) the need to maintain superiority over British and French nuclear forces (to justify US leadership of the NATO alliance) and over Russian and Chinese countervailing forces.⁽⁷⁰⁾ Consequently, the United States has conducted 20 to 30 ICBM launches annually in recent years (see table 1). Similar pressures---especially pressures to maintain missile industry jobs---are at play in Russia.

The Trident II (D5), which began launch pad flight testing in 1987, is currently the only strategic ballistic missile in production in the United States. After 19 developmental tests and nine `performance evaluation missile' tests, the missiles are currently being tested in each of the D5 capable nuclear submarines (SSBN). On 20 August 1993, the eleventh Trident II DASO test took place, this one launched from the USS Nebraska.⁽⁷¹⁾ The Navy has slated 35 missiles for DASO tests.⁽⁷²⁾ The fiscal year 1994 US defense budget request reflects the reduction in planned procurement of D5 missiles (down from 779 to 428) and reduced operational testing plans.⁽⁷³⁾ Because of the success of the Trident II test program to date, the Navy ended the CINC Evaluation Test (CET) program early and initiated a reduced follow-on CET. The program provides data to evaluate continually the performance of the system through the design service life (at least 30 years) of the D5 capable Trident subs. The Navy has slated 138 Trident II missiles for CET and follow-on CET.⁽⁷⁴⁾

In addition, the US Air Force is urging guidance and propulsion system replacement programs for the Minuteman III, which, after the year 2003, will be the only remaining land-based ICBMs in the United States.⁽⁷⁵⁾

France. In February 1993 M-4 missiles^(76),(77) were retrofitted onto the last of five French strategic nuclear submarines. Each of the five submarines can launch 16 M-4 SLBMs, each of which carries up to six 150 kt warheads.⁽⁷⁸⁾ France plans to have operational a new generation submarine (Triomphant class) by mid 1995. These will carry 16 M-45 missiles each. Unlike the M-4, M-45 missiles carry electronic counter-measures and penetration aids. France plans, in turn, to replace the M-45 missiles with 8-9,000 km range M-5 missiles in the 21st century.⁽⁷⁹⁾ France has no ICBMs, but has developed short and intermediate range land-based ballistic missiles. France tests its IRBMs and SLBMs into the North Atlantic, with a down-range tracking station on Saint Maria Island in the Azores.⁽⁸⁰⁾

United Kingdom. Britain neither builds nor tests its own ICBMs, but rather procures Trident II (D5) missiles from the United States and tests its missiles and subs on the US Eastern Test Range. The United Kingdom is reportedly slated to purchase 67 D5 missiles (including test missiles and spares) for four Trident class subs during the next 30 years.⁽⁸¹⁾

China. China is developing new ICBMs and SLBMs. Deployment of the DF-31 (ICBM) and the JL-2 (SLBM) are scheduled for mid- to late 1990s, and an even more ambitious 12,000 km range mobile missile, the DF-41 is on the drawing board.⁽⁸²⁾ In addition, China is working on improved guidance and MIRVs.⁽⁸³⁾ China flight tests missiles occasionally into the Pacific, to near the Solomon Islands, and into the Yellow Sea and Indian Ocean.

POSSIBLE CONFIGURATIONS

Comprehensive flight test ban (CFTB). A comprehensive flight test ban would be more effective in impeding missile development than would any partial measure, and it would be more attractive politically to developing country missile aspirants. A CFTB would level the playing field between them and the great powers in terms of permitted and non-permitted missile-related activities. Furthermore, a CFTB would be far easier to verify than existing arms control undertakings (like, for example, the Chemical Weapons Convention and even the Nuclear Non-Proliferation Treaty). The development of a new ballistic missile system cannot be kept secret. As then CIA Director William Webster acknowledged in May 1989, "The status of missile development programs is less difficult to track than nuclear weapons development. New missile systems must be tested thoroughly and in the open...."⁽⁸⁴⁾ Flight testing is unavoidably observable---and becomes more easily observable the longer the range of the missile. US early warning satellites can reliably determine whether missiles are or are not being flight tested.^(85),(86)

The use of possible ballistic missile components in space launch vehicles to circumvent a flight test ban would complicate such a regime.⁽⁸⁷⁾

^,(88) However, even if some component testing could not be prevented, the lack of complete system tests would result in low confidence in missile reliability. As Sherman notes: "War planners are aware of the numerous instances in which components worked perfectly by themselves but, when flight tested together, revealed disastrous incompatibilities that otherwise would have been undiscovered."⁽⁸⁹⁾

Moreover, interception of non-encrypted telemetry signals could expose military-related upgrades on ostensible space launcher flights. Non-encryption of telemetry should be a staple of any FTB regime. Violation of this principle might provide early indication of intention to break out of a missile flight ban treaty.

Undoubtedly, a CFTB would involve tradeoffs between arms control effectiveness and non-interference with space activities. If flight testing restrictions were too lax, they might be ineffectual, allowing the transfer of improvements from the civilian sector to the military. If restrictions were too severe, they might impede civil space programs. In order to build the strongest possible wall between ballistic missile tests and space flights, Robert Sherman has suggested the following guideposts at each stage of flight.⁽⁹⁰⁾

Reentry: High-speed reentry, radar-emitting reentry vehicles and terminal maneuvers could be prohibited. Ballistic missile reentry vehicles approach or impact the earth at many times the speed of sound. Accuracy would diminish if they were slower and spent more time in the atmosphere. High-speed reentry is not used in space programs, however, because "it would be bad for reusable payloads and worse for astronaut morale." In addition, legally permissible re-entry angles could be defined to distinguish between legitimate space booster rockets and ballistic missiles and between satellite/shuttle/spacecraft re-entries and weapons payload re-entry vehicles.⁽⁹¹⁾

Warhead separation phase: The weights and profiles of existing reentry vehicles could be catalogued, and the release of objects sharing the weight and velocity change of missile reentry vehicles could then be banned.

Boost stage: Each party to the flight test ban would list the length, diameter and total impulse of every missile boost stage it deploys; flights of these devices could be prohibited. Where boosters are identical to space launch vehicles, the space boosters must be displayed for inspection, counting and tagging. Tagged boosters would be granted an exemption from the test ban, provided they were not flown on a missile trajectory. When the tagged boosters were expended, all new boosters would have to be verifiably different.

All US ICBMs and the more modern of the Soviet ICBMs use solid-fuel rocket engines. Solid propellants are more stable and storable than are liquid fuels, making them more militarily useful. Through such a cataloguing and tagging system as Sherman proposes, new space launch vehicles could be required to utilize non-storable liquid fuel engines. This could be verified during flight by infrared sensors, which can determime the chemical composition of rocket propellant by its thermal signature.^(92),(93) there is some precedent for this in forced renunciation of use of Pu in energy reactors??--Jon Wolfstahl.

Guidance-systems: Ensuring that guidance being tested on a space shuttle or space launch vehicle is not intended for an ICBM is the most formidable challenge. To deal with this, Sherman recommends internal inspection of missiles and space vehicles.

Partial test ban. Several possible configurations for regional bans or other partial FTB measures exist.

Numerical testing ceilings. A regime could be negotiated that permitted only X number of tests per year. An annual quota of perhaps five or six flight tests (similar to a proposal by President Carter in 1977) would allow the major military powers to maintain confidence in the reliability of their arsenals while slowing modernization and development. Such a ceiling was proposed by Sidney Drell and Theodore Ralston in 1985. They calculated that a fifty percent cut in tests (from the average twelve per year to six) would cause a major delay in achieving confidence in accuracy enhancements.⁽⁹⁴⁾

Hussain notes that the Soviet development philosophy resulted in flight testing more often than did the United States,⁽⁹⁵⁾ a fact which could make a numerical ceiling more difficult to negotiate. However, with the radical improvement in relations between the former Soviet Union and the United States, missile modernization is already slowing. On the other hand, since most third world countries undertake only a small number of tests annually, a test ceiling would do little to prevent continued development and proliferation.

The ceiling could be designed to allow only operational/reliability tests, preventing testing of innovations. Verification of a ban on development tests would be complicated, as it would have to ensure that incremental improvements (for example, guidance improvements) were not being clandestinely flight tested on surrogate missile launchers. The broadcast of un-encrypted telemetry would help verify against such cheating.

A delivery range-delimited ban. The first question is of what range? Many of the current developing country missiles of concern are short range systems (100-300 km). Missiles of similar range in the US arsenal (like the Army Tactical Missile System) are considered tactical battlefield missiles. The effects on Russian short range systems must also be factored in to their likely acceptance or non-acceptance of such a regime.⁽⁹⁶⁾ An agreement to ban tests of such short range systems would be more difficult, but not impossible, to verify.

The United States and Soviet Union agreed in 1987 to renounce their intermediate-range ballistic missiles, those with a range from 500-5,500 km. The INF Treaty also prohibits the signatories from flight testing systems of this class. Former ACDA officials Kenneth Adelman and Kathleen Bailey have promoted the idea of internationalizing the INF Treaty as a possible approach to curtailing third world missile proliferation.⁽⁹⁷⁾ However, nearly all of the systems currently deployed by developing countries would fall below the 500-5,500 km range covered by INF. The ubiquitous Scud-B, for example, would not be included, nor would the SS-21, Lance, or Jericho I systems. In the Middle East, only the Israeli Jericho II, the Saudi CSS-2, the Indian Agni (under development), North Korean missiles under development and Iraqi missiles (which are now being destroyed anyway) would be covered. A ban on testing systems of these delivery ranges would be a meaningful step, but a regime that left their adversaries' missiles in place, and would not permit testing of their own systems, would likely be unacceptable to the Israelis and Saudis.

Such a regime would also leave open the possibility, however slim, of third world countries jumping ahead to missiles with delivery ranges above the 5,000 km INF ceiling, approaching ranges that could strike the continental United States.⁽⁹⁸⁾ Further, since some developing-country ballistic missile programs are driven, in part, by arms races or tension with ICBM-possessing countries (for example, India's concern with China), such a non-inclusive regime might lack support.

Since there is a direct relationship between missile payload and range, the possibility of downloading payload to achieve a greater than permissible range must also be factored in to any range-delimited test ban.

Flight test free zone (FTFZ). A negotiated FTFZ already exists, in the demilitarized Antarctic, and several other regions (Latin America, most of Africa) are de facto FTFZs. In the 1970s the Campaign for a Nuclear Free and Independent Pacific attempted to incorporate an FTFZ into the 1986 South Pacific Nuclear-Free Zone Treaty.⁽⁹⁹⁾ Australia quashed this effort.

A geographically-delimited approach has several advantages for curbing missile proliferation. In particular, it is not dependent on gaining the agreement of all, or nearly all, of the states currently deploying or developing ballistic missiles.⁽¹⁰⁰⁾ Only some subset of this group would be required to go along. A principal difficulty with this approach, however, is that many of the regional arms races overlap each other. For example, a regional FTFZ to include Pakistan and India would probably need to include China; and a Middle East FTFZ might spill over to include Pakistan and India.

The lack of a negotiating history between many of the regional adversaries engaged in missile races may make global or regional FTFZs premature. It might make more sense at the outset to engage regional adversaries in confidence-building exercises and to foster regional peace processes and reconciliation. Conversely, a missile flight test ban is one of the more meaningful and readily verifiable arms control measures imaginable. It is conceivable that regional adversaries---especially those that have alliance partners outside the region---might find a regional FTFZ to be a productive diplomatic strategy to build confidence within the region.

In the Middle East, some tentative steps toward a regional FTFZ were broached by the United States through separate talks with the Egyptian and Israeli governments in the late 1980s. Under discussion, reportedly, were small confidence building steps such as advance notice of missile flight tests and possibly `no first use' pledges that could lay the groundwork for farther-reaching steps in the future.⁽¹⁰¹⁾ In his post-Gulf War Middle East arms control initiative of 29 May 1991, President Bush called for a halt to further acquisition, production and testing of ballistic missiles of any range by states in the region, leading to "the ultimate elimination of such missiles from their arsenals."⁽¹⁰²⁾

This proposal would appear to be in the interests of all countries of the region: The citizens of Israel, Iran, Iraq and Saudi Arabia have all been threatened and attacked by ballistic missiles in the past five years. Moreover, an FTFZ would meet both Israel's and the Arab states' arms control interests. In general, Israel advocates limits on conventional arms transfers to the region, while the Arab states prefer to deal with unconventional weapons first, conventional arms later. Because of their historical use as delivery vehicles for nuclear payloads, and because of their relationship to conventional air force capabilities, ballistic missiles fall into a grey area and so missile disarmament might be acceptable to both sides as a first step. Indeed, Egyptian President Hosni Mubarak has vigorously endorsed a plan for a zone free of weapons of mass destruction in the Middle East, which calls on all Middle Eastern countries to announce their commitment to "deal effectively and honestly with matters involving the delivery systems of various weapons of mass destruction."⁽¹⁰³⁾

A regional testing regime that permitted flight testing only of already-deployed systems would probably be unacceptable to regional actors that felt at a disadvantage (that is, did not then deploy missiles). Conversely, a regime that allowed all countries in a region to develop missiles up to the longest range missile deployed in the region (for example, to the range of the CSS-2 or Jericho II/Shavit in the Mideast), thereby permitting development of long-range missiles by non-allies like Iran or Libya, would be unacceptable to the United States. A total, regional missile flight test ban seems the most likely to be accepted. A central question is whether regional flight test bans could be agreed without a global ban, or at least without superpower participation in the form, for example, of no-first use guarantees.

An FTFZ approach might also have to account for tests conducted by the parties on the territory of another country, outside the region. This has occurred quite often in third world missile development (for example, Israel apparently tested in South Africa; Iraq reportedly tested missiles in Mauritania; and Iran reportedly has recently prepared to flight test missiles in Sudan).

An RV test ban or ban on testing new MIRVs. While this ban would be useful in the great power context (including China), it would not apply to most developing countries. However, it might be desirable to lock all potential long range missile countries into such a ban preemptively to preclude MIRV development.

START II bans, among other things, flight-testing MIRVed ICBMs after 1 January 2003. Director of Central Intelligence James Woolsey told the US Senate in summer 1993, "We will be able to monitor the ban on MIRVed ICBMs...both by tracking the elimination of launchers for MIRVed ICBMs and by analyzing the data from flight tests of new missiles."⁽¹⁰⁴⁾

Ban certain flight trajectories. The most obvious candidate here would be a ban on flight testing missiles (especially SLBMs) on a depressed trajectory (DT).⁽¹⁰⁵⁾ Again, while this would be a useful step in the great power context, it is irrelevant in the near future to developing country missile proliferation. Since such a trajectory would have no overlap with space flights, it would be easy to verify a DT ban.

En route to achieving flight limitation regimes, several confidence inspiring measures could be undertaken. During the cold war, the United States and Soviet Union pre-notified each other of missile tests and broadcast warnings to mariners of the expected area of missile impact (see Appendix C). These practices could be expanded to include any country firing missiles into international waters or overflying another country. States in a given region could also pledge that missiles will be tested on non-provactive flight paths, away from adversaries' land mass or other assets.

CONCLUSIONS

With the cold war ended, security analysts have identified ballistic missile and nuclear weapons proliferation as the leading threat facing the United States now and in the coming years. The US defense establishment perceives the threat from third world missile development programs to be serious enough to warrant an outlay of several billions of dollars this year to develop technical countermeasures and `non-proliferation' programs.

In reality, the scale of the missile proliferation threat has sometimes been exaggerated---by proponents of missile defenses and by `proliferators' themselves. Many of the countries often cited as being a source of concern actually have very limited indigenous programs. Many undertake little or no flight testing, which may be less necessary for imported arsenals of proven and primitive systems like the Soviet-manufactured Scud-B. But missile flight testing is essential to achieve any degree of confidence that a ballistic missile system under development will work as intended.

A few third world countries are steadily developing substantive space launch vehicles and long-range ballistic missiles. India and Israel (and to a lesser stage of development, Brazil) have missile development programs, demonstrable through serial flight testing. Israel and India are not particularly politically worrisome to the United States, but both have nuclear weapons. Eventual development and deployment of nuclear tipped (or possibly nuclear tipped) ICBMs by them would have far-reaching implications.

In addition, China continues to develop more advanced strategic nuclear weapons. In the 1990s, China is expected to deploy three new ICBMs/SLBMs, as well as its first MIRVed missiles. This eventuality would also be globally destabilizing. The perilous political fate of pro-Western politicians in Russia increases the desirability of a ballistic missile flight test ban.

Such a ban would impede US Navy and Air Force plans, as well as those of the other declared nuclear powers. The main question relevant to the establishment of a global FTB regime is whether the United States believes that non-proliferation benefits accrued from an end to testing outweigh the bureaucratic imperatives, and psychological needs to continue testing.

Once a system has been tested adequately, operational reliability can be assured to some degree with methods other than flight testing. Certainly, a global and total FTB would freeze existing ballistic missile developments, and gradually erode those holdings over time. In order to ensure that clandestine development of ballistic missiles was not occurring under the guise of space launcher tests, some special provisions would have to be made. However, an FTB would be more easily verified through satellite and aircraft reconnaissance than any other arms control agreement imaginable.

Agreement to an FTB in the near term would also demonstrate the commitment to nuclear arms reduction which the superpowers pledged as an inducement to countries to sign on to the Nuclear Non-Proliferation Treaty in 1969. This year NPT adherents will decide whether and for how long to extend the Treaty.

Appendix A: Ballistic Missile Flight Test Ranges/Space Launch Test Facilities

BM=ballistic missile flight test facilities. SLV=space launch site or flight test facility.

Argentina: Argentine press reported in 1989 that the Condor II had recently been flight tested in Patagonia.⁽¹⁰⁶⁾

Australia: BM---Woomera. Run by Defense Science Technology Organization, Woomera reportedly has the longest recovery range in Western world. Its instrumented range covers 200 km² and its full range is 800 km long.⁽¹⁰⁷⁾

Brazil: SLV---Barreira do Inferno was the first major launch facility in Brazil; Alcantara was built to launch the VLS but also has launch pads for Sonda III and Sonda IV.⁽¹⁰⁸⁾

China: SLV---Jiuquan, Xichang, Taiyuan.

CIS: SLV/BM---Tyuratum Cosmodrome at Baikonur, Kazakhstan; Plesetsk Cosmodrome, Russia. ICBM tests from Tyuratum fly in a north-easterly direction toward the Kamchatka Peninsula impact zone. In the east, missiles are launched from Plesetsk, the former-USSR's northernmost launch facility and now Russia's main test and launch facility.

Egypt: Heliopolis?? This was the site of Egyptian missile development in the 1960s.⁽¹⁰⁹⁾

France: SLV (Arianespace)---Kourou, French Guiana. BM---Toulon.

India: SLV---SHAR Centre (Shriharikota); BM---Chandipur (Orissa State, east coast)

Israel: SLV/BM---Palmachim Air Force Base (south of Tel Aviv).

Japan: SLV---Kagoshima.

North Korea: BM---Nodongjagu??

Pakistan: SLV/BM---Somniami Bay??

South Africa: SLV/BM---Armiston (near Overberg, in Cape Province).

United States: BM---Vandenberg AFB (California), Cape Canaveral (Florida); White Sands Army Missile Test Range (New Mexico). SLV---Kennedy Space Center (Florida), Vandenberg AFB (California), Wallops Island (Virginia)⁽¹¹⁰⁾.

General source: "Launch Vehicles: Operational Satellite Launcher Directory," Flight International, 7-13 April 1993, pp. 37-41.

Appendix B: Flight Testing of Selected Developing Country Ballistic Missiles/Space Launchers

System Date Launch Site Impact Site/Other Info.