Dual-Use Technologies and Export Control in the Post-Cold War Era (1994)

John R. Harvey, Ph.D.

Center for International Security and Arms Control

Stanford University

Notes of briefing given to joint meeting of the U.S. National Academy of Sciences/ Russian Academy of Sciences working group on dual-use technology export controls held in Washington on May 26-29, 1992

The original purpose of the session is to address issues related to item #9 of the Protocol of the Soviet-American Meeting on Dual-Use Technologies and Conversion (signed in Moscow 20 December 1991). Specifically, item #9 directs the working group to examine the:

Needless to say, there is a strong overlap of this issue with item #4 of the protocol, namely:

My thinking on these matters, however, has evolved to something somewhat broader than the original subject. Specifically, the trend in the U.S., and very likely in the former Soviet Union and the newly democratic states of Eastern Europe, will be towards either a closer coupling, or continued close coupling, of the commercial and military technology and industrial bases. As I will try to argue, this trend will make it much more difficult to isolate, via verifiable end use controls, confidence building measures, and associated instruments, the commercial from the military application of dual-use technologies.

• The trend towards closer coupling of the commercial and defense tech bases in the U.S. and Russia.

• The current Department of Commerce approach to controlling export of dual-use items including:

  • products/commodities

  • technologies to produce products and technical data

  • human resources

• Issues in cooperative verification of end use controls

• An illustrative example

  • MTCR high-leverage dual-use technologies

  • Comments on the Russian sale of cryogenic liquid motors to India

• DoD Budget has declined 35% since peak spending years of Reagan first term

• Major weapons system procurement has declined at a much more dramatic rate:

  • $130B spent on procurement in 1985

  • $50B projected for 1997

• DoD Science and Technology (S&T) funding has been level so far, but:

  • overall budgets are declining

  • as procurements decline, defense contractor IR&D will correspondingly decline so that total S&T investment will decrease

• To maintain excellence in defense technology in the current budget climate, DoD strategy must emphasize greater efficiency in defense R&D. Figures 1 and 2, and tables 1 and 2 illustrate some of these points. This will include:

  • significantly decreasing, if not eliminating, the use of milspecs

  • relaxing requirements for companies regarding technical data rights

  • reducing onerous auditing and accounting requirements on contractors

• But, most importantly, DoD must take advantage of the great overlap between key commercial technologies and military critical technologies. DoD must:

  • exploit the commercial base for non-defense unique technologies

  • focus its technology development on key defense-unique technologies.

• Thus, the U.S. approach will be to much closer coupling of the defense and commercial technology bases.

• In Russia the system is different: the defense technology and industrial base has worked and is dominant; the commercial base, in many cases an offshoot of the defense base, has not succeeded in any measure in producing affordable and quality consumer goods for the Russian people.

• As Russia moves to convert (in effect transforming its military technology and industrial bases to commercial use) it has stated that it will use arms sales as a mechanism to finance this process. It, therefore, is likely to maintain or even strengthen the close coupling between defense and commercial technology bases.

• This tight coupling is in the exact opposite direction from what we would desire in promoting transfer of dual-use items and technologies to the East. Specifically, end use controls that try to isolate a dual-use item from exploitation by the military will be much less likely to succeed if the defense and commercial bases are one in the same or close to it.

  • Verification will also be much more difficult.

• We will have the same problem if advanced Third World states evolve in this way.

Figure 2 DoD Science and Technology Funding

Perhaps more important for U.S. maintenance of its technological "edge" in military capability is how spending on defense science and technology (i.e. the 6.1 basic research and 6.2 applied research programs, also called the defense technology base, and the 6.3A advanced technology development programs) will evolve in the future. If spending on the Strategic Defense Initiative (SDI) is excluded, these programs, constituting about 2% of the DOD budget, represent defense R&D that is not directed to a specific weapons system but which advances our understanding of, and experience with, key technologies that can lead to new defense capabilities.

As shown in Figure 2, when SDI is excluded, it may be seen that DOD science and technology funding has not shared in the defense spending increases of the early to mid-1980s. Indeed, the current 5-year projection suggests flat or slightly declining funding, in constant dollars, for these programs. When industry IR&D funds (which are tied to procurements) are counted, the prospects increase for significant overall reductions in defense R&D investment. The next chart shows the strong overlap between the national and DOD critical technologies lists.

• First, what are the principal concerns about transfers of dual-use items (including products, technologies to produce products, technical data, and human resources) which have both military and civilian application?

  • concern about diversion of an item to military program of recipient state

  • concern about recipient state reshipping item to undesirable state

• How are dual-use exports currently controlled?

  • Commerce has line responsibility, other agencies advise

  • Commerce issues licenses (general and validated)

• Possible outcomes of Commerce Department action on particular license request:

  • export prohibited to stated recipient

  • export permitted without restriction

  • export permitted with end use controls

    • recipient identifies end use

    • agrees not to divert or retransfer without authorization

    • agrees to inspections

• Compliance assessment of end use controls involves:

  • prelicense and post-shipment checks (may be random or routine)

  • checks conducted by commercial attaches in embassy, Department of Commerce ''flying squads''

• Enforcement involves sanctions and/or criminal penalties

  • Commerce's enforcement function is woefully undermanned for what Congress has asked it to do

• Which states are members of the "club" attempting to control technology transfer? Which countries would be the targets? The options for the members seem to be

  • the United States and Russia

  • COCOM and the former Warsaw Pact

  • CSCE (but this excludes Japan, Taiwan, South Korea, PRC)

  • UN (in this case both "members" and "targets" belong)

• As membership widens it becomes more and more difficult to agree on "targets."

• What are the goals of a regime?

  • to prohibit direct transfers of high-leverage dual-use items to some states or entities within states

  • to secure effectively verifiable end use guarantees against diversion, unauthorized retransfer

  • to create confidence that obligations are being lived up to,

  • are there other goals?

• How should the U.S., Western Europe and Japan "hedge its bets" against the remote possibility of a resurgent, militaristic neo-Soviet state arising from civil unrest in Russia and one or more of the new republics?

• What are criteria for verification?

• want to detect militarily significant diversion in sufficient time to react

• react means:

• mitigate effects of diversion

• redress any emerging instabilities resulting from diversion

• There will be limited resources for verification and control. This suggests that the export control regime prioritize the technologies of concern for specific states. E.g., (in order of decreasing priority):

• nuclear/chemical/biological of paramount concern

• dual-use technologies related to advanced delivery systems

• dual-use technologies related to PGMS,

• technologies related to other types of munitions

• It may also be useful to prioritize recipient states as to the level of concern regarding diversion. For example, if the U.S. were deciding (in order of increasing concern):

• U.S. allies

• benign neutral, non-aligned (NNA) (e.g. 'responsible' state, signed NPT/CWC/BWC, etc.)

• other NNA

• non-signers of NPT

• adversary of U.S. client or ally

• terrorist state

• adversary able to directly threaten U.S.

• But decisions on "targets" will most likely be multilateral. In any case, the pros and cons of the use of discriminatory policies directed against "countries of concern" should be addressed.

• In applying any policy, the issue of a country to absorb advanced technologies applicable to military systems (e.g., used in modification or reverse engineering of existing systems) should also be addressed.

• Other criteria for verification:

• will verification be cost effective

• will verification work? issues in this regard are:

• can key dual-use items be acquired elsewhere?

• how pervasive is the technology?

• are there large numbers of suppliers?

• will controls/verification degrade competitive advantage of U.S. firms?

• What are the mechanisms for verification?

• intelligence and NTM (e.g., could KGB and CIA cooperate?)

• cooperative measures among suppliers and between suppliers and recipients

• transparency and openness

• data exchange

• registry of weapons sales

• registry of transfers of dual-use items

• but, key question is how to avoid placing U.S. firms at competitive disadvantage

• inspections

• bilateral, between supplier and recipient

• multilateral

• target inspections on high-leverage items

The following five charts illustrate in more detail some of the previous discussion as it pertains to the MTCR. Specifically, in an earlier paper, we analyzed the ability of key proliferants for indigenous production of ballistic missile systems and associated technologies. The level of skill required was divided into four categories: (1) no indigenous capability, (2) the capability to modify existing systems obtained from abroad, (3) the capability to reverse engineer existing systems and (4) the ability to develop solid-propellant boosters along the lines of the U.S. ICBM program of the 1960s.

We found that many states of concern already possess the indigenous capabilities associated with points (2) and (3) above and, in ten years, many more could be at the level associated with point (4). Thus, export controls will have limited utility in restraining development in these states. On the other hand, some states have no indigenous capabilities and even if key technologies were provided they would be unable to absorb them in a missile program.

We did identify key missile-unique and dual-use technologies required for production and prioritized them. We found that human resources acquired from abroad was the single most helpful factor in providing states a capability for indigenous production.

An interesting example is the recent sale by Russia of cryrogenic liquid motor systems and production technologies to India's space launch program. The U.S. has complained about the sale because of the commonality of liquid rocket motors for space launch and military systems, and imposed sanctions on Glavcosmos and the Indian missile research organization. Three items are noteworthy in this regard: (1) Gennadiy Burbulis, advisor to Yeltsin, called for close international supervision over the sale but this did little to reassure the U.S., (2) by providing hard currency to a Russian high-technology endeavor, the sale would advance two compelling U.S. foreign policy goals: it would have a positive effect on the chaotic Russian economy and it would reduce the incentives for key scientists and engineers to sell their skills elsewhere, and (3) U.S. cooperation with India in providing jet engines for India's light strike aircraft may actually play a larger role in increasing India's ability to deliver ordnance than any missile program.

• No indigenous capabilities

  • Libya, Syria, Saudi Arabia, Yemen

• Capability to modify Scud-like systems, little else

  • Egypt, Iraq, Iran, Pakistan

• Capability to reverse engineer Scuds, make changes and produce solid propellant short-range missiles

  • North and South Korea, South Africa?, Argentina, Brazil

• Advanced capability (near early-60s U.S. capability)

  • India, Israel, Taiwan?

• No indigenous capabilities

  • Libya, Syria, Saudi Arabia, Yemen

• Capability to modify Scud-like systems, little else

  • Iraq?

• Capability to reverse engineer Scuds, make changes and produce solid propellant short-range missiles

  • Egypt, Iran, Argentina, Pakistan?

• Advanced capability (near early-60s U.S. capability)

  • Israel, India, Taiwan, North and South Korea, South Africa, Brazil

• MTCR cannot prevent a number of states from acquiring ballistic missiles

  • Scud-like missiles can be modified, reverse-engineered

  • short-range, solid propellant systems are simple

• Simple controls on systems will restrain some states

• In some cases, qualitative improvements can be inhibited

  • accuracy less than about 0.3% of range

  • range greater than 1000 km

• Controls can decrease reliability and increase costs

• Experienced engineers, scientists, and technicians

• emphasize propulsion, G+C

• also management, systems integration

• Missile manufacturing facilities

• Complete missile systems and subsystems

• Missile guidance equipment; firmware and software

• Large solid propellant mixers (300 gallons and larger)

• Rocket motor test stands

• Cooperative missile development (e.g. SA-2 booster)

• Space Launch Vehicles, components, and technologies

• Inertial Navigation Systems; gyroscopes and accelerometers

• Materials; e.g. carbon-carbon, polybutadienes

• Precision and Numerically Controlled Machining equipment

• Supercomputers and finite element codes

• Metal rolling and forging equipment