Testimony of Gary Milhollin
Director, Wisconsin Project on Nuclear Arms Control
Before the House Committee on Government Operations
Subcommittee on Commerce, Consumer, and Monetary Affairs
September 21, 1990
I am pleased to have this opportunity to address the Subcommittee on Commerce, Consumer and Monetary Affairs on the subject of United States export controls.
I am a member of the University of Wisconsin Law School faculty and director of the Wisconsin Project on Nuclear Arms Control in Washington, D.C., a project devoted to slowing the spread of nuclear weapons to developing countries.
I would like to begin by saying that our present conflict with Iraq is unprecedented in the following sense: it is the first time that U.S. forces have been used to confront the effects of arms proliferation in a developing country. Iraq’s great potential for building weapons of mass destruction is one of the reasons why U.S. troops are now poised for war in the Gulf. If war comes, and Western “guests” still shield Iraq’s arms factories, the West will be forced to bomb its own citizens to destroy its own exports. An attack on these installations will also guarantee a bloody, full-scale land war and even more U.S. casualties–many from Iraqi chemical weapons.
I believe that the West has fallen into this situation because of ineffective export control policies. Although other countries have contributed more to Iraq’s war machine than the United States, our exports have clearly been imprudent. I would like to discuss three cases which show how the inadequacy of U.S. export controls have contributed to the problem the United States now faces in Iraq.
Recent cases
Rocket Casings for Brazil
Last week seven large rocket motor casings were shipped out of Chicago to Brazil. The casings were heat-treated by a Chicago firm so that they could withstand the stress of the first launch of Brazil’s largest rocket, the VLS.
The VLS will enable Brazil for the first time to launch a satellite and to build a strategic missile. As a missile the VLS will have a range of over 2000 miles with a payload of 500 kilograms, the presumed weight of a first-generation nuclear missile warhead.
The VLS is being tested by CTA, the research arm of the Brazilian Air Force. The man most responsible for developing the VLS is General Hugo Piva, the former head of CTA, who is now in charge of a Brazilian rocket technology team in Iraq. The team, composed of former employees of CTA and other Brazilian companies, is helping Iraq improve the performance of the SCUD missiles now aimed at U.S. troops–missiles that probably carry chemical warheads.
The team is also helping Iraq develop its new Al-Abid space launcher, first tested last December. Brazil’s VLS and Iraq’s Al-Abid share the distinction of using exactly the same configuration of rocket motors–five motors grouped together in a first stage, with two more single rocket motors stacked on each other as the second and third stages. If the Al-Abid works, it will give Iraq the ability to launch spy satellites and move Iraq closer to having a strategic long-range missile.
It is absolutely clear that the Brazilians will pass on whatever they learn about rocketry to Iraq. Brazil has converted every one of its space rockets into a missile for export, and Iraq has been a preferred customer. The U.S. heat treatment of the VLS casings will directly contribute to strategic missile proliferation in both Brazil and Iraq.
The heat-treatment of such an item is a service contained on the U.S. Munitions List. Therefore it requires an export license from the State Department. It is now clear that State granted the license in error. Since Brazil is a notorious proliferation risk–it rejects the Nuclear Non-Proliferation Treaty and has secret missile and nuclear weapon development programs–it is difficult to imagine how any part of the U.S. government could intelligently agree to assist Brazil in such a venture. The fact that State did agree shows that something is seriously wrong with the export review process.
Why did State agree? Last Friday, before the Joint Economic Committee, State said that the heat treatment was “consistent with U.S. policy” and was “not controlled.” It is, of course, controlled by the munitions list or Brazil would not have needed an export license. State Department licensing seems to be a sort of low-level labelling, in which someone who doesn’t know anything about missiles compares applications to a checklist. This is done in a special atmosphere, created by people who want to maintain good relations with Brazil, just as they wanted to maintain good relations with Iraq until the invasion of Kuwait.
As soon as the case got to the Pentagon–to people who knew something about missiles–the license was suspended.
After the mistake was discovered, the State Department overrode the Pentagon’s objections and pushed the export through. Instead, State should have admitted its mistake, apologized to Brazil, and offered compensation for the casings, which had already been heat treated.
By sending the casings out, State undermined U.S. credibility as a member of the Missile Technology Control Regime (MTCR). Missile casings are on the MTCR’s list of items which the members are not supposed to export. We put great pressure on France to prevent the sale of French rocket motors to Brazil, although France said they would be used only to build a space launcher. The United States argued that there was no difference between a big space launcher and a big missile, and that the French sale would undermine the MTCR. We must now admit that there is also no difference between a big rocket casing and a big missile casing, and that our heat treatment of Brazil’s missile casings also undermined the MTCR.
Nuclear Furnaces for Iraq:
On July 19th, the White House blocked the sale of a “skull” furnace to Iraq by the Consarc Corporation of New Jersey. This high-performance furnace can melt titanium for missile nose cones and other critical missile parts and might be able to melt plutonium and uranium for nuclear bomb cores. The skull furnace was to be accompanied by three other furnaces: an electron beam furnace from Consarc, and furnaces for vacuum induction and heat treatment from Consarc’s British subsidiary.
Used together, the four furnaces would have formed a powerful production line, far exceeding Iraq’s needs for its stated end use of manufacturing artificial limbs for victims of the Gulf War. According to U.S. officials, Iraq would have had a “Cadillac” production line for atomic bomb and ballistic missile parts, even better than the facilities at American nuclear weapons labs. The White House intervened at the last minute, after the furnaces were crated and ready for shipment.
The White House had to take this drastic step because U.S. export law changed on July 1, 1990. On that date Cocom (the Coordinating Committee on Multilateral Export Controls, composed of Japan, Australia, and all NATO countries except Iceland) dropped export controls on thirty categories of equipment, almost all of which can be used to make nuclear weapons or long-range missiles. Skull furnaces were among the items decontrolled by both Cocom and the United States.
The furnaces were stopped through good fortune. The manufacturer, Consarc, notified the Commerce Department in 1989 of its intention to sell Iraq a high-performance furnace that could aid a nuclear program. The Commerce Department told Consarc–mistakenly, it appears–that there was no need to apply for an export license. In June, just before the furnaces were to be shipped, the State Department contacted the Customs Service, which put a twenty-day hold on the shipment. In July, before the hold expired, the Commerce Department gave Consarc a letter requiring Consarc to get an export license.
Because the furnaces were dropped from the U.S. Commodity Control List on July 1, the only remaining ground for requiring an export license is Section 778.3 of the Export Administration Regulations. This section obliges the exporter to get a license if he “knows or has reason to know” that the commodity “will” be used in “fabricating … nuclear weapons.” Thus to block the shipment the U.S. government had to declare that Iraq was going to use the furnace to make atomic bombs.
The declaration was an affront to Iraq, which claimed that the furnace had only civilian applications. Making this affront was the price that the White House chose to pay in order to stop the export. If Pakistan or Israel tries to buy the same furnace next month, the diplomatic price will be higher. Both of these countries have strong ties to the United States. Making last-minute nuclear accusations is not the best way to handle export cases.
Nor is it an adequate way to handle them. The U.S. government only knew about the furnaces through Consarc’s earlier application for an export license. If Iraq should order another skull furnace next week from someone else, no license requirement will apply. The furnace is no longer on the export control list. Without a license application, the government will not know about the order, and will not be able to block it by giving the exporter a letter notifying him of Iraq’s nuclear intentions, the procedure required under Section 778.3.
Pakistan, India, South Africa, or any other country trying to make nuclear weapons and missiles can also buy the furnaces directly from Consarc, without applying for a license or providing any notice to the U.S. government. Consarc has not received notice that any of these countries might put its furnaces to nuclear weapons use, so no license for such sales would be required, even under Section 778.3.
Consarc, in fact, now has an application from Romania for furnaces comparable to those that Iraq ordered. If Consarc fills this order, there will be a fairly obvious risk that the furnaces will not stay in Romania.
In view of the White House’s action in the Consarc case, the U.S. decision to decontrol the furnaces seems to have been a mistake. If the Bush administration believes, as it clearly does, that the furnaces would help Iraq make nuclear bombs and missiles, the Commerce Department should not have dropped the furnaces from the export control list.
Supercomputers for Brazil:
In addition to these two cases, another is nearing resolution now. The State and Commerce Departments want to approve the export of an I.B.M. supercomputer to Brazil. Supercomputers were invented to design nuclear weapons for the U.S. arsenal and are the single most powerful tool for designing nuclear weapons and ballistic missiles. The supercomputer would be sold to Embraer, the Brazilian aircraft manufacturer.
Embraer’s stated use for the supercomputer is aircraft design. However, the programs for calculating air flows around aircraft noses and wings are essentially the same as those for calculating the forces acting on missile noses and fins, and are closely similar to those for modeling nuclear explosions. Embraer is located next door to and exchanges personnel with CTA, Brazil’s Aerospace Technology Center. CTA’s scientists, who have converted all of Brazil’s space rockets into missiles, will have access to Embraer’s supercomputer and could use it for military purposes.
Furthermore, Embraer and CTA have both contributed personnel to the Brazilian team in Baghdad that is helping Iraq extend the range of its SCUD missiles. These high-tech mercenaries will have direct access to Embraer’s supercomputer and could share its calculations with their Iraqi customers. For the sake of making a sale, the State and Commerce Departments are willing to transfer a tool to Brazil that could enhance Iraq’s ability to threaten U.S. and allied forces in the Middle East with ballistic missiles, possibly carrying chemical warheads.
In addition to calculating the forces acting on a missile in flight, supercomputers can simulate the implosive shock wave that detonates a nuclear weapon, calculate the multiplication of neutrons in a chain reaction, and model the nuclear fusion reaction in a thermonuclear explosion. CTA is an integral part of Brazil’s “parallel” nuclear program and has enriched uranium nearly to nuclear weapons grade in Brazil. Iraq is also working to master this process, so there is the risk that Iraqi weapons designers could receive the data on nuclear explosions generated by Brazil’s supercomputer through CTA. The I.B.M. supercomputer could thus help design the Iraqi bomb as well as Iraqi missiles.
The Commerce Department’s regulations require that a country seeking to purchase a U.S. supercomputer should have good “nonproliferation credentials.” The applicant should be a party to the Nuclear Nonproliferation Treaty, have opened all of its nuclear activities to international inspection, have a nuclear trade agreement with the United States, and be generally cooperative on nuclear non-proliferation policy matters. Brazil meets none of these criteria, and is a major military exporter to Iraq and other countries. In other words, Brazil is exactly the kind of country that makes export controls necessary. However, Commerce and State nevertheless think Brazil should receive a supercomputer. If such transfers can take place, it is fair to ask why we have export controls at all.
Other U.S. sales:
These cases now seem to be the rule rather than the exception. Over the past six years, the United States has exported–to Iraq alone–such dual-use technologies as high-speed oscilloscopes and mainframe computers that can be used for missile design, and millions of dollars’ worth of electronic devices that can be used for chemical testing and the production of chemical weapons. Most recently it was revealed that the United States licensed the export to Iraq of image enhancing equipment with aerial reconnaissance and missile targeting applications.
Iraq rightfully is the center of concern today, but military programs in other Third World countries are also progressing, and if we do not find an effective way to manage this problem we will have to mount sequels to Operation Desert Shield to undo what American suppliers have done elsewhere.
Flaws in the System
Our export control system is breaking down for two reasons: first, because the wrong people are in charge of it, and second, because it is secret.
It has frequently been said that there is a conflict between the Commerce Department’s duty to promote exports and its duty to regulate them–that Commerce has conflicting missions in the export field. The Consarc case illustrates this problem. Although Consarc explicitly told Commerce last year that the skull furnace could aid Iraq’s nuclear program, Commerce nevertheless decided that no license was necessary.
The furnace was not an isolated piece of equipment. It was intended to fit into a production line with other furnaces that together could mass-produce atomic bomb and missile parts. To understand the implications of Iraq’s purchase, one had to understand how the furnaces fit together into a system and know the current status of Iraq’s nuclear weapon and missile production efforts. To understand all of that, one had to consult experts at the Pentagon, the Department of Energy, the State Department, and the Arms Control and Disarmament Agency. Commerce apparently failed to do so.
The furnace was only a single element of Iraq’s overall procurement effort. Iraq has fielded a worldwide team dedicated to acquiring everything it needs to mass-produce nuclear weapons, chemical weapons and long-range missiles. Iraq will try to buy the components of this war machine from different suppliers, one at a time, hoping that the totality of its efforts will not be understood. There is no hope of stopping Iraq without referring individual export cases to the U.S. agencies charged with tracking Iraq’s purchasing and development efforts.
I believe that the responsibility for licensing exports of dual-use items should be removed from the Commerce Department and given either to an independent regulatory agency such as the Nuclear Regulatory Commission, or to some other department, such as Defense or State, that has no export promotion function. It is essential to recognize that the number of dual-use items on the control list is small; that well over 90% of the applications to export them are granted; and that the value of the few applications which are denied is tiny compared to the overall value of U.S. foreign trade. The only real significance of these items is strategic, not economic. The place to decide strategic questions is not the Commerce Department, which is only concerned with the economic aspects of trade. The proper place to do so is an agency that specializes in security questions.
The best-known example of a federal agency that tried to promote and regulate at the same time is the old Atomic Energy Commission, which had the job of both promoting and regulating nuclear energy until 1974, when the functions were split. The Nuclear Regulatory Commission now regulates; the Department of Energy promotes. The regulatory process gained credibility and effectiveness from this separation.
The other important lesson we can draw from the is the great benefit of making decisions in public. All of NRC’s export decisions are made on the public record and in the light of day. This is the main reason why we are not hearing horror stories about U.S. nuclear exports to Iraq. Hardly anyone would want to defend such a transaction in public. Notwithstanding the NRC’s openness, our nuclear industry still seems to compete effectively on the international market.
The Commerce Department’s process is secret. Neither Congress not the public is permitted to examine in the open the record of what Commerce has sent to Iraq over the last five years. Cases come into public view only when someone inside the government becomes angry enough to leak them to the press. This is true despite the fact that all the dual-use exports that Commerce licenses are for civilian commodities restricted to peaceful use.
Congress should now require Commerce to disclose what the United States has sent Iraq over the last five years. Without that knowledge, we do not know what our military forces may be facing there, and don’t have any facts upon which to judge the licensing process. Commerce could publish the data easily by releasing the annual summaries of the licenses granted for dual-use items on the Commodity Control List. These annual summaries already exist in a database. They could be published by pushing a button. They would tell Congress and the public exactly what high-technology U.S. exports we now face in Iraq, and what we may face in the future as Iraq makes use of the exports. There is no excuse for not doing this. All of the exports were of civilian products, and all of the transactions have been completed. There is no risk that pending transactions will be revealed.
In the future, Commerce should be required to publish quarterly summaries of cases decided, so that Congress and the public can see what sort of exports are being approved. This light on the process would go a long way toward solving our dual-use export problem. Today, only the exporters know what cases are pending, and only the exporters’ voices are heard by the licensing officers when decisions must be made. The public and Congress are frozen out of the process.
Scope of International Export Controls
The final topic I would like to discuss is the recent decision to reduce the export controls applied through Cocom. As I have already stated, the United States dropped export controls on skull furnaces on July 1. Among the other sensitive commodities dropped from the list are:
Item (by old ECCN number) Comments
1075 Spin forming machines U.S. officials tried to prevent Iraq from getting these machines from Germany. They are used to make uranium gas centrifuges for converting natural uranium to nuclear weapon material
1129 Vacuum pumps Last year, U.S. officials seized vacuum pumps that Iraq was trying to import from the U.S. without a license. They are used to pump fragile uranium gas through centrifuges
1635 Maraging steel In 1987, U.S. officials arrested a Pakistani in Philadelphia for trying to smuggle maraging steel out of the U.S. Maraging steel is used to make the thin metal walls of uranium gas centrifuges
In addition to the above items, there is a second group that the United States has dropped off the licensing list for Eastern Europe, but has kept on the list for developing countries with nuclear and missile ambitions. These include the nuclear weapon triggers called krytrons that Iraq tried to smuggle out of the United States in March.
Iraq cannot buy these triggers directly from the United States without a license. But Iraq can order them through Romania. It is perfectly legal to ship U.S. triggers in large quantities to Romania (or Hungary or Czechoslovakia) without a license. This means that there is no obligation to notify the U.S. government of the shipment, ask the buyer for an end use statement or restrict the triggers’ re-sale. Thus the triggers can go to Iraq through Romania without breaking any laws. This gap in the U.S. control system is another mistake. There is no point in barring Iraq from buying directly what it can legally buy indirectly. The only apparent effect of such a system is to enrich Eastern Europe through brokers’ fees.
Following are some of the items in this category:
Item (by old ECCN number) Comment
1312 Isostatic presses Uses high temperature and pressure to press plutonium to the exact size needed for fission bomb cores, and can form “carbon-carbon” for missile nose cones and nozzles. Lower-performance presses are now cleared for Eastern Europe
1541 Cathode ray tubes Used as displays for oscilloscopes (item 1584 below)
1542 Cold cathode tubes Includes the krytrons Iraq tried to smuggle out of the United States in March, and triggered spark gaps, both of which can trigger the rapid electric discharge that detonates a nuclear explosion
1559 Hydrogen thyratrons Same nuclear triggering function as cold cathode tubes–larger than krytrons, better than triggered spark gaps
1584 High-speed oscilloscopes Can process the rapid data from nuclear tests, help develop missile guidance systems, and sort the data from missile flight tests
A few of the items that Cocom dropped are still on the U.S. export list for all destinations, meaning that they require a license even for Eastern Europe. These are mainly items useful for making long-range missiles. Iraq therefore cannot buy these items directly from the United States without a license, or order them through Eastern Europe without a license.
The problem is that Iraq may be able to buy them from other members of Cocom without a license. The Cocom list is the only basis for export control in most European members of Cocom. Although Cocom is only supposed to deny technology to Communist countries, most of the West Europeans have not distinguished between keeping technology away from the Warsaw Pact and keeping it away from the Third World. Thus, when an item falls off the Cocom list, it simply drops out of these countries’ export licensing systems. This is true of the United Kingdom, Italy, Spain, and undoubtedly of other members of Cocom.
The UK, for example, was holding up Consarc’s two furnaces in Scotland before the U.N. embargo against Iraq took effect. without the embargo, however, U.K. officials would not seem to have any legal basis for blocking the shipment even in light of its probable nuclear application.
Some of our Cocom partners adhere to the Missile Technology Control Regime, a seven-country accord including the United States whose members agree not to export long-range rockets or the means to make them. U.S. officials have asked the other MTCR members to refrain from exporting the decontrolled items. But in view of the existing disputes among the members over what the regime covers, it is not clear what effect these U.S. requests will have. For example, the other members are not likely to regard Eastern European countries as missile proliferation threats. This will mean that Iraq can order these items from other Cocom members through Romania.
The items in this category appear to be limited to the following two, which the Commerce Department designated as being of special concern for missile proliferation in March, but which were deleted from the Cocom list in June:
Item (by old ECCN number) Comments
1518 Telemetering and Used in missile guidance
telecontrol equipment to receive data from missile flight tests and to guide pilotless aircraft and missiles
1587 Quartz crystals Useful in radars, electronic warfare and lasers
By the end of 1990, the entire “industrial list” of dual-use items will be dropped. In its place will be a much smaller “core group” of items restricted to eight specific categories. Unfortunately, the categories do not seem to include several sensitive items that the United States has tried for years to keep away from proliferant countries. By the end of 1990, these items will probably be available without a license from other Cocom members, even if the United States decides to retain them on the U.S. control list. The items include:
Item (by old ECCN number) Comments
1312 Isostatic presses Same use as described above. Even high-performance presses will be dropped from control by 1991
1357 Filament winding Can produce special fibers for machines the bodies of uranium gas centrifuges (used to produce nuclear weapon material) and for the casings of rocket motors
1553 Flash x-ray devices For years, the Swedes have been criticized for selling these to India, Israel, Pakistan and South Africa, all of which wanted them to study the implosive shock wave that detonates fission bombs
1362 Vibration testing The United States recently
equipment denied India’s request to buy this “shake and bake” equipment, used to test the ability of nuclear warheads and missiles to withstand the forces encountered in use
1585 High-speed cameras “Streak cameras” can follow the shock wave that detonates fission bombs, and can help develop rocket motors