Export Controls and Nonproliferation Policy

Export Controls and Nonproliferation Policy

May 1994

OTA-ISS-596NTIS order #PB94-179975

GPO stock #052-003-01371-1

Recommended Citation: U.S. Congress, Office of Technology Assessment, ExportControls and Nonproliferation Policy, OTA- ISS-596 (Washington, DC: U.S. GovernmentPrinting Office, May 1994).

for sale of u.s. government printing officeSuperintendent[ of documents, Mail Stop SSO1p. Washington , DC 20402-9328

ISBN O-1 6-045014-4

Foreword

E xport controls on dual-use goods, technology, and software will con-tinue to be one useful tool in U.S. efforts to stem the proliferation ofweapons of mass destruction and missiles that can deliver them. Ex-port controls cannot completely block such proliferation. However,

they make a proliferant’s task harder; by increasing the cost and the difficultyof weapon programs, they can buy valuable time for broader nonproliferationefforts to take effect.

The benefits of export controls, and their ultimate effectiveness, are hard tomeasure. They depend on several factors, most importantly the degree towhich nations capable of supplying key weapon technologies cooperate incontrolling their supply. Controls also impose costs, which—like the bene-fits—are difficult to quantify. These costs are borne primarily by companiesproducing controlled goods

The difficult task for both Congress and the executive branch is to design anexport control system that serves U.S. security interests but also takes due ac-count of economic interests and fairness to regulated exporters.

The primary purpose of this report is to identify options for enhancing theeffectiveness of export controls in slowing or preventing the spread of capabili-ties to develop and produce weapons of mass destruction. Nevertheless, reduc-ing the burdens of export regulation on U.S. exporters has been a major focus ofdiscussions about revising the Export Administration Act. Therefore, the re-port also examines policy options directed mainly toward the goal of reducingthese burdens, but with special emphasis on their implications for nonprolifer-ation policy.

OTA prepared this report as part of an assessment on the proliferation ofweapons of mass destruction, which was requested by the Senate Committeeon Foreign Relations and the Senate Committee on Governmental Affairs, andendorsed by the House Permanent Select Committee on Intelligence, theHouse Committee on Foreign Affairs, the Senate Committee on Banking,Housing, and Urban Affairs, and the House Committee on Armed Services.

ROGER C. HERDMANDirector

. . .Ill

Advisory Panel

James F. Leonard1

Chairman since 6/1/93Executive DirectorWashington Council on

Non-Proliferation

James E. Goodby2

Chairman through 3/22/93Distinguished Service ProfessorCarnegie-Mellon University

George AnzelonAssociate Division LeaderLawrence Livermore National

Laboratory

Will D. CarpenterChemical Industry Consultant

Lewis A. DunnAsst. Vice PresidentScience Applications International

Corp.

Randall ForsbergExecutive DirectorInstitute for Defense and

Disarmament Studies

Thomas R. FoxDirectorOffice of National Security

TechnologyPacific Northwest Laboratories

Alan R. GoldhammerDirector of Technical AffairsIndustrial Biotechnology

Association

John M. GooginSenior Staff ConsultantMartin Marietta Energy Systems,

Inc.

Robert G. GoughSenior Member, Technical StaffSandia National Laboratories

Elisa D. Harris3

Senior Research AnalystThe Brookings Institution

Geoffrey KempSenior AssociateCarnegie Endowment for

International Peace

Joshua Lederberg4

Rockefeller University

John W. LewisCenter for International Security

and Arms ControlStanford University

Lee W. MercerCorporate Export ManagerDigital Equipment Corp.

Matthew S. MeselsonDepartment of Biochemistry and

Molecular BiologyHarvard University

Stephen M. MeyerCenter for International StudiesMassachusetts Institute of

Technology

Gary MilhollinDirectorWisconsin Project on Nuclear

Amls Control

Marvin M. MillerSenior Research ScientistDepartment of Nuclear

EngineeringMassachusetts Institute of

Technology

Janne E. NolanSenior Fellow in Foreign PolicyThe Brookings Institution

William C. PotterDirectorCenter for Russian and Eurasian

StudiesMonterey Institute of International

Studies

iv

Barbara Hatch Rosenberg Leonard S. SpectorDivision of Natural Sciences Senior AssociateState University of New York at Carnegie Endowment for

Purchase International Peace

Lawrence Scheinman5 Sergio C. TrindadeAssociate Director PresidentPeace Studies Program SE2T International, Ltd.Cornell University

1panel member until June 1, 1993; panelchair after June 1, 1993.

2Re~ign~ t. ~conle Chief U.S. Negotiator

for Safe and Secure Dismantlement of Nu-clear Weapons.

~Re~lgned Jan, 29, 1993 to jt)in Nati(mal Se-

curity Council staff.~Ex-officio;” Member of Technology Assess-

ment Advisory Ct)uncil.‘Resigned Aug. 1.3, 1993 to 13ec(mle C(wn-

selor for N(mproliferati(m in the U.S. De-partment of Energy.

Note: OTA appreciates and is grateful for the valuable assistance and thoughtful critiques provided by the advisory panel members.The panel does not, however, necessarily approve, disapprove, or endorse (his report. OTA assumes full responsiblity for the reportand the accuracy of its contents.

v

Preject Staff

Peter Blair PROJECT STAFFAssistant Director Gerald L. EpsteinIndustry, Commerce, and Project Director

International Security Division

THOMAS H. KARASAlan ShawProgram Director Anthony FainbergInternational Security and Space

Program

coNTRACTOR

ADMINISTRATIVE STAFF

Jacqueline Robinson BoykinOffice Administrator

N. Ellis LewisAdministrative Secretary

Dan Fenstermacher

vi

Additional Reviewers

lain Baird Greg Garcia Richard SeppaDepartment of Commerce American Electronics Association Tektronics, Inc.Washington, DC Washington, DC Beaverton, OR

Richard T. CupittCenter for East-West Trade PolicyThe University of GeorgiaAthens. GA

Zachary DavisCongressional Research ServiceWashington, DC

Paul FreedenbergBaker & Botts, L.L.P.Washington, DC

John Harvey Henry SokolskiCenter for International Security Aletheia Enterprises

and Arms Control Arlington, VAStanford UniversityStanford, CA Kenneth R. Timmerman

The Middle East Data ProjectMichael Pocalyko Kensington, MDOffice of the Secretary of DefenseWashington, DC

Note: OTA appreciates and IS grateful for the valuable assistance and thoughtful critiques provided by the reviewers. The reviewersdo not, however, necessarily approve, disapprove, or endorse this report. OTA assumes full rcsponsibility for the report and theaccuracy of its contents.

vii

c ontents

1 Summary 1Bottom Lines 2Options for Enhancing Export Control

Effectiveness 2Options for Ameliorating Industry Burdens 7

2 Introduction 11Nonproliferation Emphasis 12Export Enhancement Emphasis 13

3 Current United States NonproliferationExport Controls 17

4 Weighing Benefits and Costs 21Assessing Benefits 21Estimating Costs 25

5 Options for Enhancing ExportControls 31Improved Export Control List-Making 31More Effective Licensing Administration 33Improved Enforcement 38Strengthening Multilateral Controls 42

6 Reducing the Burdens on Industry 55Reduce the Numbers and Purposes of Controls 55Eliminate the “Knows, is Informed, or has Reason to

Know” Tests 59End Unilateral Reexport Controls on Exports to

Cooperating Countries 63Streamline the Application Process 64Analyze and Publish the Economic Costs 65

. . .Vlll

APPENDICES

A Estimating the Economic Costs of ExportControls 67

B Complexities of Setting Export ControlThresholds: Computers 71

C Glossary of Abbreviations 77

Index 79

Summary

Export controls on dual-use goods, technology, and soft-ware will continue to be one useful tool in U.S. efforts tostem the proliferation of weapons of mass destruction andmissiles that can deliver them. The effectiveness of export

controls in slowing proliferate ion will vary with the characteristicsof the weapons of concern, the capabilities of the target countriesand programs, the controllability of the designated commoditiesand technology, the degree of international cooperation, and thequality of enforcement. In some circumstances, they may do 1ittleto stem proliferation; in others, they may impose significant ob-stacles and delays in acquiring such weapons. Thus, they may buyimportant time during which policy makers may bring other non-proliferation tools to bear.

The overall benefits to national security of applying exportcontrols come at a price to the companies and industries whoseproducts are controlled. The difficult task for both Congress andthe executive branch is to design an export control system thatserves U.S. security interests but also takes due account of eco-nomic interests and fairness to regulated exporters. The task ismade more difficult by the inherent problems in trying to estimateboth the benefits and the costs of export controls.

This report is a product of OTA’s project on the proliferation ofweapons of mass destruction. As such, its primary purpose is toidentify opt ions for enhancing the effectiveness of export controlsin slowing or preventing the spread of capabilities to develop andproduce those weapons. Nevertheless, reducing the burdens ofexport regulation on U.S. exporters has been a major focus of dis-cussions about revising the Export Administration Act. There-fore, the report also examines policy options directed mainly at

The difficult task for both

Congress and the

executive branch is to

design an export control

system that serves U.S.

security interests but also

takes due account of

economic interests and

fairness to regulated

exporters.

II

2 I Export Controls and Nonproliferation Policy

the latter goal, but with special emphasis on theirimplications for nonproliferation policy.

BOTTOM LINES1.

2.

3.

4.

Several options are available for improving theeffectiveness of nonproliferation export con-trols: some of the options require increased re-

sources and priority for nonproliferation ob-jectives,

■ some would require considerable institution-al change within the U.S. government,

• some would require substantial changes ininternational attitudes toward nonprolifera-tion export controls, and

■ none of the enhancement options is a ‘*magicbullet” that will dramatically alter the pros-pects for stemming proliferation.

Formulating better export control policies re-quires that the U.S. government gather andanalyze better and more complete informa-tion about the actual economic costs ofmaintaining export controls.Assessment of effectiveness and costs of non-proliferation controls should be separated fromthat for controls established for other purposes.Industry concerns about the burdens imposedby export controls could be addressed by theimposition of rigid rules limiting U.S. unilater-al imposition of controls; however, from thepoint of view of nonproliferation policy, itwould be preferable to leave the executivebranch enough discretion to adapt to specificexceptional situations, coupled with:• a general presumption against unilateral con-

trols and• extensive reporting to, and oversight by,

Congress on policy rationale, outcomes, andcosts.

OPTIONS FOR ENHANCING EXPORTCONTROL EFFECTIVENESSI List-MakingIssue: What measures might improve the abil-ity of the U.S. government to identify the ex-

port items, buyers, and end-users that poseproliferation risks?

Several U.S. agencies are involved in settingU.S. export control policies. For various reasons(bureaucratic as well as technical) information isnot shared as systematically among them as itmight be.

Option: Develop shared and improved database.Newer computers, with higher speed and more

memory, allow consideration of new techniquesfor distributing, sifting, and analyzing informa-tion on proliferation problems. Applying suchtechniques within the government, however,would require some changes in bureaucratic pro-cedures as well as some additional resources.

| Licensing AdministrationIssue: What measures might allow the officersreviewing export application licenses to bringthe best and most complete information to bearon their judgments?

Option: Modernize the license-processing data-base.

The computer technologies alluded to in thesection on list-making could be even more useful-ly applied to improve the license application re-view process. Ideally, the interagency computersystem would allow analysts in all reviewingagencies to extract in real time: data about otherprevious or current applications, technical back-ground data on the proposed exports, and currentintelligence or other data about the parties to theproposed transaction.

Issue: How can the external accountability ofthe nonproliferation export licensing processbe improved?

Option: Publish nonproprietary licensing data.Post-licensing publication of data summariz-

ing dual-use license approvals would enhance un-classified research by nongovernmental investi-gators of export-import patterns that mightidentify previously undetected weapon programsor supply networks (see below, in the section onimproving multilateral export controls, for thebenefits of strengthening unclassified analytic ef-forts). Second, publishing licensing information

Chapter 1 Summary 13

might set a precedent for helping to persuade othernations to release comparable information, thuseasing the task of both governments and non-gov-ernmental groups in identifying possible avenuesof proliferation.

Undertaking this policy would require specialcare in protecting legitimate proprietary data fromaccess by exporters’ competitors.

Issue: How can the broadest possible range ofsubstantive, technical, and policy judgment bebrought to bear on licensing decisions?

Referrals of license applications by the Depart-ment of Commerce for review by other agenciesnow takes place according to rules agreed onamong the agencies of jurisdiction. Critics of pastlicensing decisions have argued that, in practice,Commerce inappropriately approved licenses thatother agencies would have blocked if given thechance. Others point out that Commerce actedwithin the laws and higher level policy guidanceof the times.

Option: Formalize interagency review processesfor licenses involving proliferation-controlleditems.

Various advocates have proposed that all mili-tary-relevant license applications be routinely re-ferred to the Defense Department or the ArmsControl and Disarmament Agency, or that all nu-clear-related applications be referred directly to alegislatively (not just administratively) estab-lished Subgroup on Nuclear Export Coordination.

A related issue is the degree of independentpower to be assigned to individual agencies and tointeragency committees. Should each agencyhave a veto over license applications, should inter-agency committees vote by majority rule, orshould Commerce have the power of decision un-less another agency invokes escalation processesto appeal the majority or Commerce’s decision?

Proponents of the strictest possible enforce-ment of export controls argue that the more re-view. the greater the chance of blocking inap-propriate exports. Proponents of a streamlinedreview process argue that too much bureaucracycan delay license decisions to death, even whentheir rejection is not justified.

Two additional considerations might beweighed in this debate. The first is that thorough,multi agency reviews within reasonable periods oftime are feasible if agencies are required to make adecision either by action or by default within aspecified period and if they are given sufficientmanpower and technical resources for license re-views. Second, attempting to stack the deck in li-censing decisions by granting one agency oranother primary jurisdiction is not necessarily apermanent solution to perceived problems. Forexample, the Defense Department in previous ad-ministrations has been less willing to approvesome exports than other departments; in the fall of1993, however, it seems to have been in fullagreement with the Commerce Department thatcurrent thresholds of performance for controlledcomputers were unrealistically low.

I Enforcement of RegulationsIssue: How can the government help exportersmake better evaluations of prospective custom-ers?

Option: Distribute more information on suspectbuyers, users, and programs.

One legislative proposal is for the governmentto publish a regular bulletin to better inform ex-porters about the risks of proliferation and whatexporters can do to help reduce those risks. U.S.companies have in the past provided the gover-nment with important leads about illegitimate buy-ers; increased sharing of government informationwith exporters might enhance the latter abilitiesto help. Dissemination of information by the gov-ernment may sometimes imperil intelligencesources and methods or risk undermining ongoinginvestigations. It also risks the embarrassment,and possibly the injustice, of publication of suspi-cions that turn out to be incorrect. The potentialpayoff from more active industry cooperationwould have to be weighed against such risks.

Issue: How can verification and enforcementactivities be made more effective?

Option: Improve pre-license and post-shipmentchecks.

4 | Export Controls and Nonproliferation Policy

Commerce Department pre-license checks ofpotential buyers (and post-shipment checks on ap-proved licenses) can help identify suspicious cus-tomers. Resources for these checking activitieshave been limited, and the checks conducted havebeen poorly planned and executed. Additional re-sources and top-level attention to developing sys-tematic strategies could make checks a more use-ful tool (although their utility will remain limitedfor many types of exports). Additional resourcesfor Export Administration and Customs enforce-ment activities could also be considered.

Option: Improve Bureau of Export Administra-tion Enforcement Office data resources.

Enforcement officials at Commerce’s Bureauof Export Administration (BXA) have been study-ing various sources of data beyond those availablefrom intelligence and law enforcement agencies tosee if they might help reveal suspicious exportpatterns. For example, U.S. Census data on all thetypes and quantities of items going to a particularcountry might reveal purchasing patterns that sug-gest diversion of imports to a proliferant weaponprogram. Thus far, however, Commerce has nothad the resources to put this sort of analysis intothe context of a larger, more encompassing data-base, of the type described above.

Option: Fully utilize sanction authorities.Current legislation gives the executive branch

a range of economic sanctions (including the im-position of further, noneconomic, sanctions) toapply to foreign “persons” who aid proliferationthrough the sale or transfer of items on U.S. con-trol lists. (The Clinton administration draft ExportAdministration Act (EAA) of 1994 would harmon-ize sanctions for chemical, biological, and mis-sile weapons proliferation, now authorized inother pieces of legislation; the pending State De-partment authorization act for fiscal year 1995institutes similar sanctions for nuclear prolifera-tion.) The actual application of sanctions is left toconsiderable executive branch discretion. One op-tion would be to leave the president less discretionin choosing, deferring, or waiving sanctions.Rigid requirements, however, risk forcing thepresident’s hand in cases where more subtle action

might have a greater effect on nonproliferationgoals. Too much discretion, on the other hand,risks avoiding diffilcult choices and sending inap-propriate messages to those who foster proliferation.

A compromise option would be to permit theflexibility requested in the Clinton administrationdraft EAA, but to accompany it with more explicitprovisions for accountability to Congress aboutthe costs and effectiveness of sanctions imposedor the reasoning behind deferring or waivingthem.

Multilateral Control ArrangementsSince there are very few technologies useful toproliferant weapons programs that the UnitedStates produces uniquely, international coopera-tion among potential suppliers or transshippers isessential to effective export controls.

Issue: How can the United States keep a lowlevel of international consensus on the transferof conventional military technologies from un-dermining current agreements on nonprolifer-ation of weapons of mass destruction?

With the end of the Cold War, the membership,targets, and listed technologies for the Coordinat-ing Committee on Export Controls (COCOM,formerly a Western arrangement for denyingtechnology to Communist nations) are undergo-ing significant changes that must be multilateral ynegotiated. With technologies applicable to weap-ons of mass destruction already addressed in othermultilateral export control regimes, the COCOMsuccessor regime will most likely attempt to regu-late the transfer of technologies for developing ormaking conventional weapons. Consensus will bedifficult to reach, both within the United Statesand among the international participants, aboutwhat technologies should be controlled, and forwhat reasons.

Option: Separate COCOM succession from re-gimes for nonproliferation of weapons of massdestruction.

The nonproliferation regimes dealing withweapons of mass destruction (and missiles), forwhich considerable consensus has already been

Chapter 1 Summary | 5

painstakingly built, should not be mixed into con-troversies over revisions of lists formerly con-trolled by COCOM for other purposes.

Issue: How can coordination among membersof multinational nonproliferation export con-trol regimes be enhanced?

Option: Promote an information sharing net-work.

Communications and information tools cannotsubsitute for a genuine willingness to cooperateamong adherents to export control regimes. Givensuch willingness, however, they could make iteasier to implement cooperation. The UnitedStates has instituted a pilot program for a sharedcomputer network among Nuclear Suppliers’Group (NSG) members (the NSG is a group of na-tions that has agreed to common export controlpolicies for nuclear technologies and dual-usetechnologies applicable to nuclear programs).Such a network would offer a variety of opportu-nities for increased coordination among the nu-clear suppliers. In agreeing to multilateral con-trols on dual-use technologies, the NSG membersalso agreed to avoid undercutting each other’s de-cisions by informing one another when they denyexport license applications for the listed items.Timely dissemination of this information wouldallow each supplier to consider its own export de-cisions in the light of those made by any of the oth-ers. Once refused an export license in one country,a potential buyer would not have a chance to findanother supplier in another country even if thatcountry did not have independent reason for sus-picion about him. License denial information, aswell as some of the other kinds of information de-scribed below, could be especially useful to gov-ernments without the extensive export control in-frastructure and intelligence resources of some ofthe larger members of the NSG.

Option: Extend the NSG database network ideato the other export control regimes.

Such a network could be extended to membersof the Australia Group (chemical and biologicalweapons) and Missile Technology Control Re-gime (MTCR) as well, since there is already alarge overlap in membership among those groupsand the NSG. This step would be most useful incombination with agreements in those regimes toreport export denials, as the NSG members do.Such agreements, however, will not be easy to ob-tain.

Option: Expand international reporting to ap-provals as well as denials.

With this wider range of data about exportswith weapon program potential, all NSG (or otherregime) members would have a better chance ofdiscerning trade patterns that might help identifysuspicious end-users or possible diversion paths.Because of fears of revealing proprietary data ofuse to competitors, however, regime membersmay resist revelation of their approved licenses. 1

Should the United States decide to seek such re-porting, it may need to test that resistance throughthe leadership both of exhortation and of its ownexample. Even the expenditure of considerablediplomatic capital with other regime membersmay not be enough to bring about this degree ofcooperation.

Option: Increase intelligence sharing.Whether by means of a networked database or

through other means of communication, sharingintelligence data about unscrupulous suppli-ers, buying and financing operations, ques-tionable agents, and suspicious end-users is animportant means by which supplier groupscan coordinate their export controls. Shared in-telligence could, for example, help members ofthe NSG make better informed licensing judg-ments by giving them more information abouthow prospective buyers measure up against thecriteria that the NSG has agreed to take into ac-count in licensing decisions.

1AS noted above, some firms might be fearful that Confidential (but still legitimate market information might he revealed to c(~nlw’titers if

all sales were reported. Even If the supplier-group(mp data were not In the public d(m)ain, there w (mid be the p)ssibill[y that participating g(wem-nwn[s would leak in fom)atl(m to their OW n cmm[~ fimls.


In some situations, national intelligence agen-cies having trusted relationships with one anothermay be able to share secret information. Amongstthe large and diverse sets of nations making up thenonproliferation supplier groups, however, thecontinuous, direct sharing of classified informa-tion seems unlikely. What seems more feasible isthe production and dissemination of analysesbased on open sources. It may also be possible todevelop open-source evidence for facts that mightoriginally have been indicated or discovered bysecret means.

An option to consider is to provide governmentsupport for nongovernmental, open-source data-base and analytic projects. One means of supportfor such efforts is to contribute grants or award re-search contracts to the private institutions carry-ing on such projects. Whether the informationshared multilaterally comes direct] y from the U .S.government, or whether it comes from privateU.S. institutions, there is some risk that it will beperceived as a U.S. tool for manipulating intern-ational opinion and decisions to serve unilateralU.S. interests. This risk imposes a need for con-siderable tact and diplomacy in the ways in whichthe United States attempts to persuade other na-tions to act on the information provided.

Option: Support development of former SovietUnion states’ administration of export controls.

The effectiveness of global export controls willbe greatly weakened unless Russia and the otherformer Soviet states join and effectively partici-pate in the full set of western nonproliferation con-trol regimes: NSG, Australia Group, and MTCR.Some progress has been made in this directionwith Russia already in the NSG, vowing to be-come a de facto member of the MTCR, and prom-ising to adhere to Australia Group guidelines. Theother newly independent states should also bebrought into the nonproliferation regimes. Thesenations also need to develop effective export con-trol systems. The United States has offered severalmillion dollars in Nunn-Lugar funds for that pur-pose to each of the four republics retaining Sovietnuclear weapons, but has reached agreement onspending the money only with Belarus. Other re-

publics could probably also make use of financialassistance. U.S. agencies have also been offeringtechnical assistance in export controls to the for-mer Soviet states.

At the Moscow summit in January 1994 Presi-dents Clinton and Yeltsin signed a joint “Memo-randum of Intent” on “Cooperation in the Area ofExport Control,” saying their governments in-tended to cooperate in “any or all” of six areas in-tended to improve nonproliferation export con-trols and that they “may” establish expert workinggroups to carry out their intent. At this writing, it istoo soon to tell whether these actions will be takenor whether they will result in concrete improve-ments in the Russian control system.

Option: Seek greater cooperation from develop-ing countries.

Newly industrializing countries that are notmembers of the established export control groupsare also becoming possible sources for proliferantweapon programs.

In its draft for the EAA of 1994, the Clinton ad-ministration proposed that (individual validated)license-free exports of controlled items could bepermitted to and among members of a multilateralregime. More convenient access to dual-usetechnology items might serve as an incentive forsome developing nations to join supplier regimes.On the other hand, were these nations so well-be-haved in the first place, license approvals prob-ably would have been forthcoming anyway. A dis-advantage to removing validated licenserequirements is that the United States would losethe opportunity to judge on a case-by-case basiswhether the recipient country’s own export con-trols were strong enough to prevent retransfer ofsome items. Instead, it would have to arrive at ageneral judgment to that effect. Of even greaterconcern is that, if the emerging supplier is itself aproliferation threat, it might acquire easier accessto items needed for its own weapon programs,even as it helped control supplies to others.

Other steps aimed at bringing more nations intoexport control cooperation have been proposed.While worth exploring, they may be difficult tosell to some developing nations, who have per-


ceived export controls more as a means of eco-nomic discrimination than as a nonproliferationtool. In attempting to better inform developing na-tions about the purposes and effects of export con-trols, the industrialized countries would have totake care to avoid the appearance of simply dictat-ing their own views of the proliferation problemand how to deal with it.

OPTIONS FOR AMELIORATINGINDUSTRY BURDENSFrom the point of view of the effectiveness of ex-port controls, it is desirable to have exportingcompanies see the system as fair and just, so thatthey will have every incentive to help make thecontrols effective—for example, by reportingpossible illicit purchase attempts. From the pointof view of U.S. competitiveness in internationalmarkets, i t is desirable to place the leastconstraints consistent with national security onexporting firms.

I List MakingIssue: How can the United States protect its ex-porters from competition from firms in coun-tries with less stringent export controls?

Option: Promptly remove controls from itemsthat are availablefrom other countries in similarquality and quantities.

A policy of attempting to control only itemsthat were not available from other sources wouldlead to a shorter list and might result in fewerlosses of business from U.S. companies to foreigncompetitors. Proponents of unilateral export con-trols argue that this is tantamount to knowinglyselling a gun to a criminal just because he mayhave been able to buy it from someone else. Someexporters may feel that they should not be deniedlicenses to sell to such users, on the ground thatsomeone else will anyway. Most, however, wouldnot wish to do business with users trying to buildweapons of mass destruction. It is not the loss ofthese relatively rare sales that exporters fear, butrather that the export licensing process itselfcauses them to lose legitimate business to foreign

competitors at the same time that it fails to keepthe proscribed items out of the hands of prolife-rants.

Other countries may be more willing to controlnew items (or exports of currently controlleditems to newly identified end-users) if the UnitedStates demonstrates its own will to do so first.Thus, proposals to limit U.S. export controls tomultilaterally controlled items have included pro-visions for at least temporary impositions of uni-lateral controls to allow attempts to reach multilat-eral consensus. Putting a legislative limit on thelength of time for which unilateral controls can beimposed does carry a risk: other nations whom theUnited States is trying to persuade to follow suitcan just stall negotiations until the statutory limiton the U.S. controls runs out.

Those in favor of retaining some discretion forthe government to maintain some unilateral con-trols argue that in some cases the United Statesshould set a standard of leadership behavior, what -ever else some other nations might be doing. Intaking a principled stand against assisting thespread of weapons of mass destruction, the UnitedStates may help bolster international normsagainst such proliferation, protect U.S. companiesfrom the embarrassment of being identified withproliferation activities, and possibly win overother supplying nations to its position.

Option: Reduce the size of the export control listto narrow the scope of its purposes.

After the initial reforms of COCOM controlswith the end of the Cold War, the Department ofCommerce (DOC) Office of Export Licensingwent from handling over 100,000-125,000 exportlicense applications a year to about 24,000 in 1992and 25,000 in 1993. Many of the remaining li-cense applications concern items controlled forpurposes other than the nonproliferation of weap-ons of mass destruction. Most of the items remain-ing on the (formerly COCOM) “national security”control lists relate to possible conventional mili-tary applications. Items controlled because theymay be used in making weapons of mass de-struction or missiles are largely the subject of


negotiated international supplier agreements;the option of narrowing the scope of controlsdoes not have much applicability to this area.

The COCOM lists were designed primarily toslow Soviet progress in a broad range of militarytechnologies. The fact that they might also slowthe development of the Soviet civilian economywas seen as, if anything, an additional national se-curity benefit of the regime. COCOM’S originalpurposes became largely (though perhaps not en-tirely) obsolete with the breakup of the SovietUnion.

A new set of goals for controls over dual-usetechnologies related to conventional weapons hasnot yet emerged. Late in 1993, COCOM membersagreed to abolish the organization at the end ofMarch 1994 and to replace it with a successor re-gime. As COCOM formally ended, however, thegoals and procedures of that successor regime re-mained unclear. Some have proposed that theUnited States initiate an explicit new nonprolifer-ation regime aimed at limiting the spread of ad-vanced conventional weapon technologies. Sucha policy, aimed at keeping particular types ofconventional weapons out of reach of many na-tions, would require a different export controlstrategy than one directed at maintaining a West-ern military advantage by restraining the technicaldevelopment of a single large military-industrialcomplex. In the absence of clear-cut opposingblocks of allies, there is bound to be less consen-sus about who should be the targets of such a strat-egy.2 It is therefore likely to be more difficult tosell the strategy multilaterally than it was to per-suade states to participate in the original COCOMregime.

Issue: How can uncertainties and costs stem-ming from “EPCI” rules be reduced?

The Bush administration’s Enhanced Prolifera-tion Control Initiative (EPCI) and certain subse-

quent legislation led to Export AdministrationRegulations requiring individual validated li-censes (IVLS)3 for almost any items that the ex-porter “knows” or “is informed” might be used inany way in a chemical, biological, or missileweapon program. In December 1993 the Com-merce Department issued further guidance speci-fying that the rule would apply to items destinedto be directly employed in such a program. For nu-clear weapon programs, the rule is stronger: a li-cense is required for any item that the exporter‘*knows or has reason to know” will be used di-rectly or indirectly in such a program.

Industry representatives, at least prior to theDecember 1993 clarifications, argued that theEPCI rule unnecessarily hinders their economicperformance by:■

✘

■

●

requiring virtually all exporters to establishcostly programs to find out whether their cus-tomers are involved in a proscribed activity,imposing unilateral controls on U.S. exports ofitems that are likely to be available to prolife-rant programs from foreign sources anyway,burdening honest exporters with regulations,when illicit exporters will not apply for licensesanyway, andbecause of uneven information among export-ing firms, giving honest exporters who arenonetheless ignorant of export control require-ments an unfair economic advantage over theirbetter informed competitors.

Option: Eliminate the EPCI ruleThe Congress could eliminate the EPCI rule by

legally requiring the DOC to consolidate its dual-use or “commercial” export controls into a singlelist that fully enumerates all the products forwhich an export license is required and all thecountries and specific end-users as well. Thiswould greatly simplify the exporting companies’

2mem is more international con5en5u5 about re5tm1ning the general spread of weapons of mass destruction than there is a~)ut n~aintaining

the military superiority in conventional weapon technologies of the advanced industrial nations.

~T() ~)blain ~ IVL, the exp)~er must fi]e an application with the Department of Commerce stating the items to be shipped, their value, me

buyer, and the end-user.


job in deciding whether a license application wasnecessary and whether it was likely to be ap-proved.

Elements of this proposal exist in the current re-gimes. The NSG, the Australia Group, and theMTCR all center on agreed, published lists ofcommodities. On the other hand, the regimes donot require the members to agree in advance onwho all the controlled countries and end-usersmay be. Instead, they provide agreed criteria fordeciding whether an export should go forward.

Publishing lists of all suspect buyers and usershas drawbacks, including risking the compromiseof intelli.gence and law enforcement data. It alsoeliminates the government ability to control ex-ports that pose an imminent proliferation threateven though they are not on a published controllist. These drawbacks must be weighed against theadvantages of having better informed legitimateexporters.

Option: Maintain the EPCI rule, while attempt-ing to assure its fair application.

Defenders of the “knows or has reason toknow” rules argue that exporters who may be trad-ing with a proliferant end-user find it too easy tolook the other way, or to fail to report what theyknow, as long as their own particular export is noton a specific control list. More important, the rulegives the government a safety net by allowing theapplication of export controls when it learns abouta pending transaction that risks helping a weaponprogram, but which is not explicitly covered bythe current Commerce Control List. Finally, manycompanies would themselves prefer not to dealwith end-users developing weapons of mass de-struction, whether their products are critical tothose programs or not. Procedures for the gover-nment to inform them of the character of their buy-ers may well save them from public embarrassmentl a t e r o n .

The DOC’S December 1993 guidance shouldassure U.S. exporting firms that they do not haveto worry that they will be subjected to extraordi-nary demands to probe deeply into the character ofend-users of relative] y innocuous products. Advo-cates of a “knows or is informed rule” point out

that the stronger form of the rule (“has reason toknow”) has existed for some time for nuclear ex-ports and in other legal areas. The judicial systemhas not generally permitted unreasonable inter-pretations of what constitutes a “reason to know.”In practice, no firms have been penalized for hav-ing failed to apply for a license for something thatthey are alleged to have known would be used in abanned project. On the other hand, questionablesales have been prevented by the government in-forming exporters that transactions with certainbuyers would require an IVL.

Option: Change the “knows or is informed” ruleto just an “is informed” rule.

Under this option, the government would notexpect companies to “know their customers” andapply for licenses in dubious cases. It would, how-ever, retain the legal ability to stop risky transac-tions about which it had obtained intelligence byinforming the exporter of a license requirement,even if it could not expect companies to reportthe “red flags” that may indicate suspiciouscustomers.

Option: Maintain the rule, but publish a specificlist of controllable items.

The government would generate a separatecontrol list of products or technologies that, al-though not listed as requiring export licenses,could be “directly used” in proliferant programs.The exporting companies would then be responsi-ble only for knowing or having reason to knowwhether recipients of those particular items wereengaged in illicit activities. The firms, if in doubt,could ask the government for advisory opinionson prospective buyers. The government couldalso make the companies’ job easier by publishingthose advisory opinions about particular end usersso that other firms could be forewarned. This lattermeasure would carry some risk of alerting illicitprocurement agents the the U.S. government Was

aware of their activities.

I Licensing AdministrationIssue: How can exporters be given licensing de-cisions in time to avoid losing sales?


Option: Place statutory limits on licensing proc-essing times.

The Clinton administration’s draft EAA pro-poses assuring that nearly all license applicationswould either be resolved or referred to the Presi-dent within 90 days of filing. For reviewing agen-cies, the default decision is approval if the dead-lines for objections are not met. There seems to beno reason why, with sufficient resources, currentlicense decision deadlines could not be shortenedto the times proposed in the Administration bill,or even less, without diminishing the quality ofanalysis and review that the license applicationsreceive. Doing so, however, is likely to cost addi-tional funds that the executive branch has not re-cently been willing to allocate to export controlmanagement.Issue: How can the economic costs of exportcontrols be given appropriate weight in policyand licensing decisions?

Some U.S. exporters have argued that the gov-ernment imposes export controls without ade-quate consideration of the costs they will imposeon U.S. industries. They have proposed, therefore,that assessment of the costs of controls should bemade an integral part of the export control proc-ess. Costs may include:■

m

■

m

resource and opportunity costs to the govern-ment,sales forgone or denied because of controls,exporters’ administrative costs in complyingwith regulations, andbusiness lost because of licensing delays or cus-tomer perceptions of supplier unreliability.

Current data about the actual costs, direct or in-direct, imposed by export controls on specificU.S. industrial sectors and on individual firms isnot generally available. Estimates used in publicdiscourse are either anecdotal or based on data setsnot well-designed to provide the needed informa-tion.

Option: Require regular economic impact state-ments for export control policies.

The Clinton administration’s draft EAA statesas U.S. policy:

. . . to ensure that U.S. economic interests play akey role in decisions on export controls and totake immediate action to increase the rigor ofeconomic analysis and data available in the de-cisionmaking process.

Such a policy could be reinforced by a require-ment for regular “economic impact” statements toCongress attempting to estimate the overall costsof controls to the U.S. economy as well as theirmore specific costs to certain industries. Such es-timates should help enlighten a debate now featur-ing many claims and counter-claims, but 1ittle realdata.

As desirable as such costs estimates may be,however, it is important to recognize that gather-ing usable data will require overhaul of the currentDOC license-processing computer system and ex-penditures on economic research and extensiveexporter surveys. Even after these efforts aremade, the nature of this particular estimatingproblem will dictate that many uncertainties stillremain.

Introduction

Less-developed nations seeking to produce weapons ofmass destruction (or missiles for delivering them) usuallyneed to import certain equipment, materials, and technol-ogies. The United States and other countries have insti-

tuted export controls on such commodities as a tool of nonprolif-eration policy. OTA’s report on The Proliferation of Weapons ofMass Destruction: Assessing the Risks] concluded that exportcontrols will continue to be a useful nonproliferation tool.They are unlikely to stop a determined proliferant in the long run,but nevertheless may buy important time in the shorter term—time that may be used to bring other nonproliferation tools to bear.

The Export Administration Act (EAA) provides the legislativebasis for U.S. export controls on dual-use items—goods and ser-vices with civilian applications that could in principle be used formilitary purposes.2 In 1994 the EAA, temporarily renewed in1992, will expire and Congress must reauthorize it. Virtually ev-eryone involved in export control matters agrees that a new EAAis overdue. There is less agreement about what the most urgentproblems are and what the best solutions may be. Moreover, thecore issues are likely to remain in contention well beyond pas-sage of a new export control bill.

The initial report of this OTA assessment pointed out that thereare tensions between the goals of effective nonproliferation ex-

2

, . . export controls

are unlikely to stop a

determined proliferant in

the long run, but

nevertheless may buy

important time in the

shorter term.

2Thc Nuclear Non-Proliferation Act of 1978, however, establishes the basis for controlof nuclear-related dual-use items.

I 11


port controls on the one hand and enhanced exportcompetitiveness on the other. When it comes tospecific proposed reforms of the EAA, however,the tradeoffs are not completely clear-cut and di-rect. Some proposed changes in export controllaws and regulations are aimed primarily atimproving their effectiveness. Proponents ofother changes are seeking mainly to reduce theburdens to industry of export controls and toreduce barriers to legitimate exports. Thesetwo objectives are not necessarily contradictory:any given change intended for one purpose mayhinder, have no effect on, or even help pursuit ofthe other.

The most desirable export control measureswould contribute to one or both goals (effective-ness and economic competitiveness) and detractfrom neither. For example, increasing the re-sources available to review export license applica-tions might assure both that the applications arescreened in greater detail (possibly increasing thechances that dangerous exports will be stopped)and that the review process is sped up (reducingthe waiting time for exporters and their potentialcustomers). If an option does detract from onegoal while contributing to another, policy makerswill, ideally, evaluate the tradeoff and choose themore valued goal.

Unfortunately, as will be shown below, thisevaluation does not always lend itself to a clear-cut analysis. As a result, the revision of theEAA is likely to become a focus of strong politi-cal controversy both in the Congress and in theexecutive branch.

NONPROLIFERATION EMPHASISThose whose foremost objective is strengtheningthe nonproliferation regimes tend to stress thebenefits of export controls. Citing Pakistan, Iraq,and others, they point out that in the past, ineffec-tiveness of export controls has resulted not justfrom the inherent deficiencies of the tool, but fromfailures by the United States and other nations toapply it rigorously. They argue that, given the hor-rendous consequences of the spread of weapons ofmass destruction, even partially effective controls

may help avoid catastrophic future costs. (Somewould also argue that we have a moral imperativenot to contribute to proliferation, whatever othersmight be doing.) They also point out that controlsserve some purposes beyond limiting access topotentially dangerous goods and technology. Bycreating a record of what is sold to whom, controlsprovide information to help monitor proliferation.They also serve to indicate a government’s deter-mination to oppose proliferation: a state that de-cries the spread of weapons of mass destruction,but does nothing to prevent its own citizen fromhelping those who seek to acquire them, losescredibility y.

Those stressing the benefits of controls, then,argue that some economic sacrifices (in the formof reduced exports) are worth the price. They saythat if exporters are burdened by controls, the bur-dens should be seen as part of the price of doingbusiness with potentially dangerous commodi-ties. Moreover, some nonproliferation advocatesquestion whether the business lost because of non-proliferation-related controls is in fact very signif-icant to the U.S. economy: few solid figures areavailable to prove that it is. Presented with pro-posals to ease the burdens to industry imposed byexport controls, these advocates are more likelythan others to perceive dangers that those propos-als will decrease effectiveness of controls.

Within the school of thought emphasizing thebenefits of export controls, there is some divisionbetween those who would apply nonproliferationcontrols uniformly toward all potential prolifer-ants and those who advocate singling out “roguenations” that are perceived to pose the greatest im-mediate threats to international stability. Those fa-voring a more universal policy argue that weaponsof mass destruction are dangerous no matterwhich states are acquiring them. Therefore, prolif-eration on the part of states considered friendly tothe United States should be opposed as vigorouslyas that by states thought to be more hostile. First,the international consensus needed to fight prolif-eration is much harder to mobilize in a world ofdouble standards, in which proliferation is toler-ated in some states but not in others. Second,

Chapter2 Introduction | 13

states are not likely to agree on which statesshould be considered particularly unfit to possessweapons of mass destruction, making it hard toreach consensus on measures that should be im-posed against regime violators. One state’s allymay well be another state’s “rogue.” Third, as wasshown in the case of Iran in the late 1970s, or Iraqin the late 1980s, today’s friends can turn into to-morrow’s adversaries. Arsenals amassed when bi-lateral relationships are favorable may pose seri-ous diplomatic and military problems should thepolitical situation change.

In contrast with those favoring a universal ap-proach to nonproliferation strategy, others stressthat such weapons become a concern primarilywhen they are acquired by ‘brogue” or “outlaw”states that are particularly hostile towards theUnited States or to international security. In thisview, "weapons don’t kill, nations do”: it is moreimportant to deal with the particular dangersposed by such nations than it is to enforce globalnonproliferation norms. From a United States’point of view, today’s “rogue’* states include Iran,Iraq, Libya, and North Korea (but not India, Paki-stan, or Israel).3

Since these states generally already havestrained relations with Washington, little wouldbe lost by applying highly coercive policies—such as more restrictive export controls and stron-ger economic sanctions—that threaten to furtherdisrupt ties to the United States. As with the uni-versal nonproliferation approach describedabove, this approach views preventing the prolif-eration of weapons of mass destruction, at least torogue states, to be one of the highest national pri-orities. Unlike the universal approach, however,

this approach would devote less attention to non-proliferation efforts targeted against friendlystates.

Table 2-1 contrasts the export control policiesimplied by these two approaches as well as a thirdapproach, emphasizing enhancement of exports,described below. The views here do not, of course,encompass every individual with a role in the ex-port control debate. Other positions between theones starkly differentiated here are also possibleand likely.

EXPORT ENHANCEMENT EMPHASISThose who worry most about the economic costsof export controls tend to emphasize the ineffec-tiveness of unilateral controls and the spreadingavailability of dual-use technologies. They arelikely to give less credence to arguments about theutility of partially effective controls and the valueof the United States’ continuing to set an exampleof more stringent controls when other suppliersfail to cooperate quickly. They are more likely toperceive measures reducing the burdens to indus-try of export controls as increasing the effective-ness of controls as well. But when a tradeoff doesseem necessary, they argue, the governmentshould make explicit evaluations of not only theforeign policy benefits of controls, but also oftheir economic costs. 4 For further discussion ‘f

the arguments about the costs and benefits of non-proliferation export controls, see chapter 4 and ap-pendix A of this report.

Table 2-2 lists the criteria that a policy satisfy-ing proponents of the contrasting approaches toexport controls would have to meet-criteria that

3President clinton's Assistant for National Security Affairs refers to "backlash states," naming Cuba as well as North korea, Iran, Iraq, and

Libya. See Anthony Lake, “Confronting Backlash States,” Foreign Affairs, vol. 73, N(). 2, March/Apr]l 1994, pp. 45-55.

4Analysts from the Brookings Institution have expressed a third point of view: export controls are increasingly ineffective and should be

drastically reduced, but in their place the United States should try to build an international consensus on achieving greatly increased tran.sparen-(y in international trade and in national industrial activities. This might mean that reduced exp)rt contro/s would be replaced by increased ex-port reporting requirements, plus intensified governmental and other monitoring aimed at exposing proliferant programs to international sanc-tions. See testimony of Janne Nolan and John Steinbruner before the Subcommittee on Economic Policy, Trade, and Environment of the HouseCommittee on Foreign Affairs, on June 9, 1993 and June 23, 1993, respectively. Hovever, hcy(md those two short prescntati(ms, further analy -SIS of this appr(mch dtws not appear to ha~ c been carried (wt or published.


Policy issue—

Priority tononprolifefationpolicies

Universality ofnonproliferationpolicies

Unilateral U.S.export controls

Using exportcontrols aseconomicsanctions

Nonproliferation emphasis —

Few If any foreign policy prioritiesshould be higher

Apply universal rules to all poten-tial proliferants, whatever theircurrent politlcal stance, today’s“safe” nation may become tomor-row’s “rogue”

Are acceptable, even when noteffective in blocking proliferation ”■ Nuclear Nonproliferation Act

requires U.S. not to contributeto proliferant nuclear weaponprograms, whether other sup-plier nations door not,

● U S leadership often neces-sary to win export controlcooperation from other nations

When any dual-use items (including those not normally controlledfor proliferation reasons) ex-ported to a given country havepotential to be diverted to weap-

on programs, all such transfersshould be denied, the economichandicap borne by the target na-tion may help persuade It to endweapon-of-mass-destructionprograms

SOURCE : Office of Technology Assessment, 1994

would maximize both control effectiveness andease of compliance. Although the criteria do notappear to be logically incompatible, finding theappropriate tradeoffs among them is a difficult po-licymaking task.

The primary focus of this report discussion ofpolicy options is on measures that might increasethe effectiveness of U.S. export controls. Whererelevant, however, the possible consequences forexporting companies are also considered. Anadditional set of options deals with reducing the

“Rogue Nation” non-proliferation emphasis— .- —.

Nonproliferatio is the highest priority regarding“rogue nations” thatthreaten regional or glob-al stability: other goalsmay rank higher withfriendly nations

Focus nonproliferationpolicies (and export con-trols) on rogue nations

Same as nonproliferationemphasis, but to be ap-plied selectively to roguenations and theirsuppliers

Silmilar to nonproliferationemphasis, but to be ap-plied only to rogue na-tions, such as thoseIdentified by the UnitedStates as supporters ofInternational terrorism

Export enhancementemphasis

Nonproiferation needs to beweighed against other na-tional objectives, particularlyglobal economic competi-tiveness

May favor either a universal-Ist or a “rogue” approach, aslong as near complete coop-eration among suppliers ex-ists

Unilateral controls are lnef-fective, economically costly,and should be avoided, ex-ceptions may be made If em-bargo against target nationsIS complete, not partial

Broad-based export controlsintended to punish or coercea state are generally ineffectivebut if applied should beas part of an expicit com-plete embargo of the targetnation’s economy, exportcontrol laws should be ap-plied only for their explicitlyauthorized purposes, not asad hoc sanctions

—

burdens export controls may place on exportingcompanies.

Chapter 3 of this report discusses the U.S. ex-port control regime as it was configured early in1994. This configuration was the baseline fromwhich Congress would revise the Export Admin-istration Act, which governs U.S. export controlson dual-use commodities.

Chapter 4 discusses the problem of assessingthe benefits and costs of export control measures.It outlines the factors that determine how effective

Chapter 2 Introduction |15

Regime goal Criterion

Effective control lists

Effective Iicensing administration

Effective enforcement

High degree of multilateral support

Minimum burden on exporters

■

■

■

■

●

●

■

●

●

●

●

●

■

■

■

■

■

■

■

■

●

Timely and thorough I[st-construction process identifies the right goods,technology, and users for controlsPolicy makers have flexibility to adjust to changing circumstancesLicensing process feeds back Information about buyer behavior that wouldbe useful to the Iist-making process

Licensing decisions based on adequate informationLicensing decisions based on the best available judgmentLicensing offices have adequate resourcesLicensing process IS run efficientlyExporters kept well informed of suspect end-users

Adequate investigation and prosecution of exporting violatorsRegular monitoring of end-uses

Agreement among major suppliers on controlled commodities and usersNo undercutting of license denials by other governmentsEffective enforcement

Policies explicitly balance nonproliferation goals and economic competitive-ness goalsCommodities controlled kept to a minimumForeign competitors do not undercut controlsApplicants have access to lists of controlled items, countries, and end-usersLicensing decisions are rapidEnd-user controls are not so onerous as to deter Iegitimate buyersLicensing decisions are consistent, fair, and subject to adequate appealsprocessLicensing process protects proprietary Information that could be useful tocompetitors

SOURCE Off Ice of Technology Assessment 1994

particular export controls may or may not be. OTAhas previously observed that the effectiveness ofexport controls in slowing proliferation will varywith the characteristics of the weapons of concern,the capabilities of the target countries andprograms, the controllability of the designatedcommodities and technology, the degree of in-ternational cooperation, and the quality ofenforcement. In some circumstances, they maydo little to stem proliferation; in others, theymay impose significant obstacles and delays.

Chapter 4 of this report also describes the po-tential costs of imposing export controls andpoints out the difficulty of reliably quantifyingthose costs. OTA also pointed out in its earlier re-port that, besides the costs to the government of

administering an export control system, the af-fected exporting companies must bear the burdensof complying with regulations and the possibleloss of legitimate business to competitors who areless strictly regulated. However, data to reliablyquantify such losses are difficult to find.

Chapter 5 analyzes policy options aimed atmaking nonproliferation export controls more ef-fective. The chapter does not attempt to pro-vide a single set of recommendations reflectingone coherent approach to export control policy.Instead, it analyzes a range of options culledfrom a variety of sources. It categorizes those op-tions according to the phase of the regulatoryprocess each would affect most:


■ the making of lists of controlled goods and Chapter 6 reviews some recent proposals fortechnology as well as target countries and orga- easing the burdens that export control laws andnizations, regulations place on exporting companies. Of par-

■ the administration of export control licensing, ticular interest for this report are the possible con-- the enforcement of laws and regulations, and sequences (positive or negative) of such measures● the engagement of international cooperation in for nonproliferation efforts.

making controls effective.

CurrentUnited States

NonproliferationExport

Controls 3

s lowing the proliferation of weapons of mass destruction isonly one of many purposes for which the United Stateshas established a system of export controls. Table 3-1shows the range of U.S. agencies administering export

controls, the major legislation underpinning those controls, andthe variety of purposes intended for them (as of early spring,1994).1 This report focuses on the most controversial type of ex-port controls, those over dual-use items. The Department of Com-merce administers export licensing of dual-use items under au-thority of the Export Administration Act.

Figure 3-1, presenting the sub-categories on the (Departmentof) Commerce Control List (CCL), shows the range of dual-useexport controls falling under the EAA of 1979.2 The EAA con-tains two broad categories of dual-use item control: “national se-curity” (established by Section 5 of the Act) and “foreign policy”(Section 6). National security controls are primarily those placedon items formerly in the COCOM3 Industrial List, originally in-tended to preserve Western technological superiority by reducingthe flow of advanced dual-use technologies from Western indus-trial nations to the Soviet bloc and other Communist nations.

Despite their obvious national security implications, controlsover items that might be used to make weapons of mass destruc-

I For additional descriptions of the establishing laws, see the first report of this OTA

assessment, proliferation of Weapons of Mass Destruction: Assessing the Risks, OTA-[SC-559 (washington, DC: U.S. Government Printing Office, August 1993).

2In 1992, this law was extended for 18 months, to end in June, 1994.3COCOM was the Coordinating Committee on Multilateral Export controls. It was

abolished in the spring of 1994, probably to be succeeded by another, substantiallychanged, multilateral export control arrangement.

Slowing the proliferation of

weapons of mass

destruction is only one of

many purposes for which

the United States has

established a system of

export controls.

I 17


Agency Legislation

Department of Commerce:Bureau of Export Administration Export Administration Act

Nuclear Non-Proliferation Act

Trading With the Enemy Act

International Emergency EconomicPowers Act

National Defense AuthorizationAct

Department of Defense

Department of Energy

Department of State:Center for Defense Trade

Arms Export Control Act

Atomic Energy Act


Arms Export Control Act

National Defense Authorization Act

Department of Treasury: Trading With the Enemy Act

Off Ice of Foreign Assets Control International Emergency EconomicPowers Act

Nuclear Regulatory Commission Atomic Energy Act


Items or countries

National Security Controls (COCOM)

Nuclear Referral List

Embargoed countries

Iran / Iraq, MTCR

Defense articles and services

Nuclear technical assistance andsubsequent arrangements

Nuclear-related dual-use items

U.S. Munitions List defense articlesand services

Nuclear weapon and designequipment

Chemical weapon agents

Biological weapon agents

MTCR

Various prohibited transactions

Nuclear power generation, nuclearmaterial, and fuel cycle equipmentand technology

Note that this table only identifies the major departments with legislated responsibilities for administering the export control licensing. It does not list allthe many bureaus and interagency groups that play a variety of jurisdictions and roles in managing the whole export control process

SOURCE Office of Technology Assessment, 1994

tion were grouped in the EAA in the Section 6“foreign policy” category. Controlled in this lattergroup are the items on the Australia Group (chem-ical and biological weapons) and Missile Technol-ogy Control Regime (MTCR) lists.4 Nuclear ex-ports are controlled under authority both of theEAA and of the Nuclear Non-Proliferation Act of1978. (See figure 3-2 for further detail on nonpro-

liferation portions of the CCL.) Miscellaneousother controls, not directly related to prolifera-tion, fall under the rubric of foreign policy.Many industry complaints about the economicburdens of export controls have arisen fromcases concerning these other foreign policycontrols, as well as COCOM controls, ratherthan nonproliferation controls.

~e MTCR began as an informal agreement among coordinating countries to control goods relevant m missile manufacture; the UnitedStates implemented these controls administratively. Amendments in 1990 to the EAA, made in a section of the National Defense AuthorizationAct for that year, provided a speeific U.S. legislative basis for missile technology controls.

Chapter 3 Current United States Nonproliferation Export Controls I 19

International Industrial List (CoCom)

Critical militarily-related technologies

Crime Control and Human Rights

Crime control and detection instruments and equipmentcontrolled for human rights purposes

L

Regional Stability

Certain sensor and navigation equipment and technologiescontrolled for regional stabMy reasons

AntiterrorismComplete embargoes to Cuba, North Korea, Iraq, Libya;licenses for various commodities to Syria and Iran

Embargoed Countries

Commerce and Treasury administer embargoes against Hati,Montenegro, North Korea, Serbia, and Cuba

LibyaReexport of U.S. goods from third countries, except for certaincommodities

1[w

SupercomputersExports require safeguards to prohibit access by CoCom-proscribednationals and any users revolved in strategic or proliferation activitiesthat could be detrimental to U.S. interests.

Ld South Africa

Embargo on arms and police equipment

7

Weapons of Mass Destruction and MissilesChemical, biological, nuclear weapons-related technologies andmaterials; missile-related items (See expanded list, fig. 3-2)

The Department of Commerce administers a variety of dual use commodity, technology, and software export controls, on/y someof which relate directly to weapons of mass destruction The Commerce Control List (CCL) IS categorized by some 430 ExportControl Classifications Numbers (ECCNS). The ECCNS do not correspond one for one to sing/e commodity technology orsoftware items m some cases an ECCN covers on/y a sing/e narrow/y defined item, but in many cases multlple related items fallunder the same ECCN. The reason for control of each ECCN category may be single or multlple (e g , for both National Securityand Nuclear Proliferation) but the reason(s) /@cd may m fact apply on/y to some subset of iterns within the category

SOURCE Off Ice of Technology Assessment 1994


Biological (Australia Group)

Items Target countriesCertain dual-use equipment; For biological agents, allbiological agents, related except Canada;technical data, or any other For others, list of countries incommodity that exporter knows Supp. 5, Pt. 778, EARwill be used for BW in listed

Chemical (Australia Group) LItems: Target countries54 chemical precursors and For precursors, all non-related technical data; certain Australia Group countries;other dual use equipment to For others, list of countries inlisted countries; any other Supp. 5, Pt. 778, EARcommodity that exporter knowswill be used for CW in listedcountries

Nuclear Referral List (Nuclear Suppliers Group)

Items Target countriesCommodities or related No published lists, but licensetechnical data that might be reviews closely tied to country’ssignificant for nuclear explosive status as member of Nuclearpurposes or that exporter Non-Proliferation Treaty.knows (or has reason to know)will be used for nuclear-weaponrelated purposes (e.g. makingspecial nuclear materialswithout internationalsafeguards ).

Missile Technology Control Regime

Items Target countriesSpecific items related to use in For listed items, all countriesmissile programs. except Canada.Any other commodities or For others, listed missilerelated technical data that programs, EAR Supp. 6exporter knows will be used in countries, or programs in non-a missile program MTCR that exporter knows to

involve missiles

SOURCE office of Technology Assessment, 1994

WeighingBenefits

and Costs 4

As the first report of this OTA assessment pointed out, therisks attached to the proliferation of weapons of mass de-struction are considerable. Advocates of strong exportcontrols point out that the costs of proliferation may in-

clude thousands or millions of lives, billions of dollars of proper-ty destroyed, or, at a minimum, billions of dollars paid for mili-tary preparations to deter or do battle against owners of suchweapons. Therefore, if export controls could be shown to be ef-fective in preventing proliferation, they might be judged wellworth the economic burdens they might place on the nationaleconomy or individual exporters.

ASSESSING BENEFITSThe great majority of the world’s nations have signed agree-ments ] recognizing that the further spread of nuclear, chemical,and biological weapons would be dangerous to international se-curity, and should be opposed. The signatories to these treatieshave also agreed that those possessing the wherewithal to producesuch weapons should not help other nations do so. The majority ofnations able to supply goods and technology needed for produc-ing the weapons have agreed to control exports from their territo-ries as a nonproliferation measure. Implicit in these agreementsis the belief that export controls on at least some items are auseful nonproliferation tool.

In a world where all the materials, tools, and technology need-ed to develop and produce weapons of mass destruction (or mis-

] I. e., the Nuclear Non-Proliferation Treaty (NPT), the Biological Weapons Conven-tion (BWC), and the Chemical Weapons Convention (CWC).

The majority of nations

able to supply goods

and technology needed

for producing the

weapons have agreed

to control exports from

their territories as a

nonproliferation measure.


siles to deliver them) were readily traded, acquir-ing the weapons would be cheaper and the time toget or develop them shorter. But how much cheap-er and how much shorter? The answer to that ques-tion varies widely from country to country, as wellas from one type of weapon to another. As notedabove, several key factors affect export controleffectiveness: 2

the nature and level of technical sophisticationof the weapon sought,the state of industrial and technical develop-ment of the target countries,the controllability (or degree of general avail-ability) of dual-use items (those with both civil-ian and military applications),the degree of cooperation among all the relevantexporting nations, andthe degree of success in monitoring and enforc-ing of controls by each cooperating nation.

The Technology VariablesThe ability of export controls to block access toneeded goods and technology depends strongly onthe type of weapon being pursued. So too, does theproliferant’s ability to develop alternatives or“work-arounds” to the items it cannot purchaseabroad. In evaluating the effectiveness of exportcontrols, therefore, it is important not to lump allweapons of mass destruction together. s

Nuclear WeaponsExport controls have the best chance of effec-tiveness against nuclear weapon proliferation(compared to that of other types) because the

processes for producing weapon-usable fis-sionable materials are difficult and costly. Paki-stan, for example, had to abandon efforts to pro-duce plutonium when external assistance ended;its uranium enrichment program relied heavily ontheft, smuggling, and black market transactions,frequently in violation of export control laws.4

South Africa, on the other hand, devised a methodof uranium enrichment that relied less on imports(but still received clandestine foreign assis-tance). 5 Although it is easier to design and builda primitive bomb once fissionable material hasbeen acquired than it is to produce the material,more advanced designs (improving on size,weight, and explosive yield) require additional in-fusions of technology.

Chemical WeaponsExport controls can increase the cost and difficul-ty of producing large quantities of high-qualitynerve agents under safe conditions. They may alsohelp keep advanced delivery technologies (e.g.,chemical cluster bomb designs) out of the handsof some nations. Nevertheless, controls are un-likely to block a nation determined to producechemical weapons. As industrialization spreadsto more countries, so will civilian chemicaltechnologies that can be applied to weapon-agentproduction. Moreover, with environmental,health, and safety standards rising around theworld, modem chemical facilities are increasing] yadopting the type of production technology thatformerly had been used only for the most toxiccompounds.

2For further discussion of key technologies for each type of weapon, see U.S. Congress, office of Technology Assessment, Proliferation ofWeapons of Mass Destruction: Assessing (he Risks, OTA-lSC-559 (Washington, DC’: U.S. Government Printing Office, August 1993) and

Tethno/agies Under/)ing Weapons oj’klass Dcsfrwfion, OTA-BP-ISC- I I 5 (Washingt(m, DC: U.S. G(wemment Printing Office, DecemberI 993).

‘Ibid.

4SCC Leonard s. SFcIor With Jacqu~]in~ R. Sml(h, /V//(/ear Arnbi/ions: 7“he .Spread o/”Nw/ear weapons, ] 989-) 990 (B(mkkr. CO: west-

view Press, I 990), pp. 90-91.51 bid., pp. 270-271.

Chapter 4 Weighing Benefits and Costs | 23

To a certain extent, the Chemical WeaponsConvention (CWC) will compensate for the de-clining utility of export controls: the Conven-tion’s verification measures will constitute a kindof “post-shipment end-user check” for trade inprecursor chemicals and chemical manufacturingequipment, since suspicious locations will be sub-ject to challenge inspections. (The Treaty will alsoban the transfer of chemical weapon precursorchemicals to non-CWC parties. )

Biological WeaponsThe basic equipment and raw materials needed togrow biological warfare agents are in widespreaduse for commercial food processing and pharma-ceutical purposes. As knowledge of biotechnolo-gy spreads, so will the ability to produce largeamounts of agent in small facilities. Export con-trols are unlikely to be a strong bar to the ac-quisition of biological weapons.

MissilesSuccessful missile export controls will not pre-vent the countries now suspected of having weap-on-of-mass-destruction programs from findingways to deliver such weapons. These nations havecombat aircraft that could do the job. Some ofthem already have relatively short-range ballisticmissiles. Any of them could also utilize less con-ventional means of delivery. In addition, mostcould probably derive at least simple cruise mis-siles from small airplanes or unmanned aerial ve-hicles. Nevertheless, missile export controls canhelp limit the spread of such advanced missiletechnologies as precision guidance, stagedlong-range ballistic missiles, advanced reentryvehicles, and long-range cruise missiles. A fre-quently cited case of successful export control ac-tion is the blockage of “Condor’* solid-fueled mis-

sile technology from Argentina to the Iraqi “Badr2000” missile program.

I The Cooperation andEnforcement Variables

Only since 1984 for chemical weapons, and1992 for nuclear and biological weapons, haveinternational groups of supplier nationsagreed on a multilateral basis to control the ex-ports of specified dual-use commodities thatmight be used to produce those weapons. Thesegroups include most, but not all, of the major po-tential suppliers of the items in question. (A sig-nificant exception for all three groups is China. )An important immediate task is to gain the coop-eration of the newly independent states of the for-mer Soviet Union.b As industrialization spreads,more countries become potential suppliers and,therefore, potential candidates for membership inthe supplier groups. This can be a complicatedproblem when one of the targets of an export con-trol regime is also a potential supplier.7

Supplier-group export controls can be usefuleven if all possible suppliers do not adhere fully tothem. First, most nations and companies do notwish to contribute to the proliferation of weaponsof mass destruction. The information shared bymultilateral export control groups helps theirmembers identify potential misusers of their prod-ucts, and thereby lets them avoid inadvertent in-volvement in such programs. Second, in an envi-ronment of broad international consensus thatcertain kinds of exports should be controlled,there is a greater chance that pressure of variouskinds can be brought to bear on the few nonpartici-pants to limit or end their offensive behavior.

Even with nominal international agreementon export controls, however, there has beenwide variation in how the controls are inter-preted and enforced by each nation. First, the


criteria for withholding export licenses are subjectto the judgment of each state’s licensing authori-ties.8 For example, Russia, although not a mem-ber of the Missile Technology Control Regime(MTCR), had agreed to abide by its restrictions onthe transfer of rocket-related technology. At thesame time, it planned to sell both hardware andproduction technology for cryogenic rocket mo-tors to India. India and Russia argued that this wastechnology suitable only for space-launch ve-hicles and not applicable to military missiles. TheUnited States, however, argued that the terms ofthe MTCR forbade the transfer of such technologyto a country such as India with a military missileprogram.9 Implementation of the Nuclear Suppli-ers’ Group (NSG) dual-use export control guide-lines (see below) is also up to the discretion ofeach member, but the members do undertake notto undercut transfers refused by other states.

Even if interpretations of export control agree-ments among the participants were always in har-mony, the related laws, implementing regulation,and administrative practices are unlikely to beuniform. National export control systems varywidely in:

■ the degree of detail in legislation and regula-tions,

■ the administrative resources for managing thesystem,

■ the information available to the licensing offi-cials,

D the numbers and skills of customs officials,■ the degree or lack of corruption in administra-

tion and enforcement,= the resources devoted to enforcement opera-

tions, and~ the legal sanctions available for punishing vio-

lators of export control regulations.

Weaknesses in one or more of these factorsoffer opportunities for proliferant organiza-tions to circumvent export controls andsmuggle out or divert at least some of the com-modities they want. Even so, from the standpointof the potential buyer of controlled commodities,controls that are only partially enforced still pres-ent difficulties. First, the buyer has to go to thetrouble and expense of finding a sufficiently un-scrupulous seller. Second, even if some items areavailable in small numbers from such sellers, thebuyer may need larger quantities than backdoordeals can supply. Third, he may not be able to ob-tain the necessary technical services and spareparts to keep his diverted equipment running.Fourth, he may have to resort to costlier methodsof production than if he had full access to intern-ational markets. Fifth, discovery of one or more il-licit transactions may tip off exporting statesabout the program for which the purchase is beingmade, and thence lead to counteractions.

Much of Eurasia now poses new problems inthe harmonization of export control practices.First, in the European Union (EU, formerly theEuropean Economic Communities), the emer-gence of a truly common market in which mostcontrols may not be applied to intra-Union tradewill mean that the strength of export control en-forcement will depend on the weakest links. Thatis, if goods or technology move freely within theUnion from countries with strong export controlmachinery to other countries with weaker enforce-ment, they may leak out of the region to potentialproliferants. As EU negotiations on export con-trols have proceeded, Germany (with the strictestcurrent export control regime) had argued for

81n the COCOM ~eglnle, ~em~m refeme~ proP)sed ex~)fis of listed items tt) an administrative headquafiers, ~d any menl~r could veto a

license approval.

IJu s ~)licynl~ers were nl{)re ~oncemed over the ~)tentia] milita~ uti]ity of some of the technology being transferred to India th~ they. .were over the cryogenic I iquid-fueled motors themselves. In July 1993, Russia agreed to adhere completely with the U.S. interpretation ofMTCR requirements and to withhold the further transfer of rocket motor production technology to the Indian space program. However, much

documentation had previously been shipped. The Russians would also proceed with the sale to India of four of the rocket motors themselves.


common regulations as strict as its own; apparent-ly losing that struggle, it has promoted a rule al-lowing individual EU members to enforce con-trols that are stronger than the Union’sregulations. 10

Meanwhile, to the east, most of the republics ofthe former Soviet Union still lack effective cus-toms controls over their borders. let alone strin-gent export control policies for dual-use technolo-gies. This means that commodities that thenonproliferation supplier groups are trying to con-trol may leak not only from one former Soviet re-public to another, but beyond the former Sovietboundaries as well. 11

In sum, many variables conspire to weaken theeffects of export controls on programs to makeweapons of mass destruction. But to say that ex-port controls are sometimes ineffective is also tosay that they are sometimes effective. Althoughsome would-be proliferant nations may be able towork around many supplier controls, others maylack the resources to do so. Even if it is not pos-sible to estimate the deterrent effect that exportcontrols may have on the calculations of somenations deciding whether to pursue weapons ofmass destruction, it is logical to assume thatthere is such an effect.

For those states that pursue weapons of massdestruction in spite of controls, the costs and de-lays may be important. For example, without ex-port controls on nuclear-weapon related commo-dities, one can easily imagine that South Africacould have built dozens of nuclear weapons ratherthan 6, and that it might have then been more re-luctant to eliminate its arsenal and join the NPT.To take another example, without the barrier, suchas it was, of export controls, Iraq might have builtnuclear weapons before it invaded Kuwait, dra-matically changing the context for operationsDesert Shield and Desert Storm. Denials of for-

eign technology probably helped slow the Argen-tine and Brazilian nuclear programs until thosecountries were ready to join the nonproliferationregime. In short, export controls on some items,even if imperfect, may help buy time that makes acrucial difference.

The cost and delay that export controls imposeon proliferants is probably impossible to quantify,or even to estimate qualitatively. Thus policy-makers confront a dilemma as they contem-plate how to enact and administer an exportcontrol regime: the benefits, while potentiallygreat, are essentially intangible and long-term,and accrue to the nation as a whole; the costs,however, are more palpable and immediate,and are unevenly imposed across a few firmsand industries.

ESTIMATING COSTSLike any regulatory regime, export controls im-pose costs both on the government and on the in-dustries regulated, and those costs can be both di-rect and indirect. For the U.S. government, thedirect costs are those born by the administeringagencies (see table 3-1 ). Since some officials onlyhandle export control issues as part of their work,even the direct costs of export controls to the gov-ernment are hard to estimate. Moreover, sincelimiting the spread of weapons of mass destruc-tion and missiles is only one of many objectivesof the U.S. export control regime, estimatingthe cost of nonproliferation controls alone isalso difficult. The indirect costs of export con-trols to the government may include the follow-ing:

● time and attention of high-level officials drawnaway from other nonproliferation and foreignpolicy issues,

10see H, Mu]ler ~t, ~],, }-r<)nl ~/a(.~ .~)lecl) t. Wh;(t Angel? T-he Ncb$, ~ernl~n E~por; c~nfr~)l P[)ll(y, PRIF Rcj)l)rts No. 32 (Frankfurt am

Mare, Gummy Peace Research Inst]tulc Frankfurt, January 1994), p. 56. The authors point out, however, that pressures fr(ml Gemxm busi-

nesses clalmlng unfa]r disadvantages relative (() their EU c(mlpetitors will result in weakening of Gem~an regulati(ms as well.

I I see foflh~onlino” OTA re~~fl (m [he pr(~lifcra[ion ]mplicati(ms of the breakup {~f the fomler soviet union.5


● diplomatic or economic concessions made inbargaining with other nations for cooperationon export controls, and

■ damage to diplomatic or economic relations en-suing from imposition of export controls onforeign nations or of sanctions against foreignviolators of U.S. export regulations.

As will be pointed out below in the section on“Strengthening Multilateral Controls” (ch. 5),some policy options for increasing multilateralcooperation are likely to meet considerable resis-tance from other members, or from prospectivemembers, of the international export control re-gimes. Pursuit of these measures might cost theUnited States considerable geopolitical capital,and might require that other U.S. goals in dealingwith those nations (e.g., promoting human rightsin China) be subordinated to the nonproliferationgoal.

Officials and journalists from some developingnations have expressed the view that export con-trols are aimed less at preventing proliferationthan at blocking the diffusion of advanced civil i antechnology from industrialized nations to newcompetitors.

12 A biological weapons expert in-

volved in international activities related to the Bi-ological Weapons Convention (B WC) argues thatthis view may become a major obstacle to winningdeveloping nations’ cooperation in possible ef-forts to add verification measures to that agree-men t . ]3

In economic terms, the heaviest price for exportcontrols is paid by the exporting firms whoseproducts are subject to regulation. Industry rep-resentatives testifying before Congress andelsewhere have complained that current U.S.export controls hamper their companies’ ex-port competitiveness in several ways.

First are direct costs. The most obvious directcost is the loss of business that would have beenpermitted in the absence of controls. The gover-nment tells companies that they may not make cer-tain sales because doing so would in some wayharm the national interest. The value of export li-cense applications denied gives some measure ofthis lost business. There are presumably manyother sales, however, that companies do not botherto try to make because they have reason to believethat an export license would be denied.

Another direct cost of export controls is admin-istrative: tracking massive and complex U.S. ex-port regulations and then assuring company com-pliance imposes time, money, and personnelcosts. In part, the regulations are so complex be-cause they spell out so many exceptions to thegeneral rules. Nevertheless, exporting companiesneed to track the rules and exceptions. In somecases, smaller companies may find the burdens sogreat that they forgo exports entirely.

Company compliance problems may be com-plicated by the de facto absence of public identifi-cation of all controlled commodities and end-users. A “knows or is informed” regulationrequires individual validated licenses (IVLS) forany exports that might be “directly employed in”the design, development, acquisition, or use ofmissiles or chemical or biological weapons in acountry listed in one of the supplements to the Ex-port Administration Regulations. An exporterwho has even “reason to know” that items or datamight be used directly or indirectly in a nuclearprogram must also apply for a license. Such rulesat least to some extent shift regulatory and intelli-gence-gathering burdens onto exporting companies.

However, the Department of Commerce(DOC) also offers guidance about what an export-

I zAl[hough [hl~ ~>rcep[lon ,Ilay ~x i~[, it does not appear to be tx)m (mt by the faCtS. ]n 1992, for exampk, the ~ approved 1.43 ‘icenses*

representing potential expwts valued at $319.5 million, for the export of items controlled for chemical or bwlogical weapon proliferation rea-sons; it denied only 24 such licenses, valued at $7.4 million.

13Barbara Ha(ch R{)5en~.ru ~r50na] comnlunica[ion,” Mar. 24, 1994. At present, the BWC has no VerifiCatiOII provisions. Seved na[i(~nsa,have proposed that a regime of compliance monitoring be added to the Convention.


er can reasonably be expected to do to avoid cus-tomers who may be engaged in inappropriate end-uses. Before December 1993, the “know” ruleseemed to apply to any items going to a proliferantend-user. In that month, however, the DOC issueda Guidance statement intended to ease exporterconcerns that the rules might be arbitrarily en-forced. 14 It should be pointed out that no companybefore or since that guidance was issued has actu-all y been penalized for failing to apply for an IVLwhile knowing or having reason to know that theend-user was a suspected proliferant.15

It now remains to be seen whether corporateconcerns about the “know” rule will be assuagedby the new guidance or not. In the past, companieshave also argued that the “is informed” part of therule was unfairly applied, with some companiesbeing informed that certain buyers were unaccept-able and others not being informed; those not in-formed were then left able to make sales fromwhich their competitors were unfairly barred.Commerce officials have acknowledged thisproblem of uneven information and said they areaddressing it.

An IVL entitles the exporter to ship a specifiedquantity of licensed items to a particular destina-tion for a period of up to 2 years. For some itemsand destinations, a company may be able to avoidapplying for an IVL by obtaining a “distributionlicense”

. . . that authorizes exports of certain commodi-ties under an international marketing program,generally to three or more consignees that havebeen approved in advance as foreign distributorsor users. This procedure is a special privilege re-served for firms with a thorough knowledge ofand cxpericncc with the Export AdministrationR c g u l a t i o n s , a n d a n i n t e r n a l c o n t r o l m e c h a n i s m

to assure strict compliance with the require-ments of the license. 16

Although this program may relieve the exporter ofthe need to apply for IVLS in many cases, it doesrequire the license holder to monitor more closelythe behavior of its buyers and of its own com-pliance with regulations. On the other hand, DOCofficials report that some exporters say their com-petitive abilities have actually been strengthenedby the additional information that internal controlmechanisms provide to their decision makers. Forthe purposes of this report, note that distribu-tion licenses, with very few exceptions, do notapply to nonproliferation items; therefore,whatever the cost burdens-or benefit-ofdistribution licenses, they cannot be attributedto nonproliferation controls.

A complete accounting of the direct costs toU.S. industry of compliance with export controlregulations is not available, but some informationis. A 1992 industry-sponsored survey of 42 largeexporters found that 30 to 40 percent of their ex-ports required IVLS, and that these companies av-eraged 24 employees and $1.3 million a year eachon licensing and compliance.

17 One large U.S. ex-

porting firm with $14 billion in annual sales and$4 to $5 billion in annual exports in the early1990s reportedly maintained a 100-person exportlicensing department costing several tens of mil-lions of dollars per year. 1 8

Unfortunate] y, the study’s sponsor, the Nation-al Association of Manufacturers (NAM), did notrelease any details about the study, such as whichfirms were surveyed and what fractions of theirbusiness costs were attributable to export controlcompliance. With 30 to 40 percent of their exportsrequiring licenses, it is clear that these firms were

I fFor fuflhcr dlscussl[)n of [he “~nc)~” and ‘“reason to know’” ru]es, See below, chs. 5 and ~.

I ~Expofl A~nlln15tra(lon Rcgulatl(ms, 15 CFR \ 773.3 (Jan. 1, 1993 ).

17N:it10niil A5soclat10n” ,)f Manufacturers stln ~, as r~plfled In ]n/crn{l//(jni// 7)-cIdc Reporter, ,Aug. 26, i ()~~, p. ~~~ and cl[Cd b) J. Da\id

R]chardst)n, .$~;~n~ L) L’.,S, li~porr 1)~.~inf cnf[lci (Washlngtt)n. Institute for Intc!matlt)nal Ectmomics, 1993), p. 38, fo(~tnotc” 16.

I ~R1chardson, Ibid,, p. 37; the firm was nt)t named.


Total Returned AllType of exports exported Approved w/o action Denied applications

($17.8 B) ($5.3 B) ($0.812 B) ($23.9 B)—.

All goods and services $640.5 B 2.8% 0.8% 0,1 % 3 7%

Industrial supplies and capital goods $282.1 B 6.3% 1.9% 0370 8 5%

Advanced technology $107.1 B 16.6% 4.9% 0 8% 22 3%

NOTE Many license applications returned without action may have been refiled later and been counted again in the “Approved” column The “lndus-rial supplies and capital goods and the “Advanced technology” categories are separately derived and probably contain considerable overlap

SOURCE Department of Commerce and John Sullivan Wilson, “The U S 1982-93 Performance in Advanced Technology Trade”, percentages byOffice of Technology Assessment

not representative of U.S. exporters as whole,since a much smaller fraction of all exports is sub-ject to licensing (see table 4-l). In addition, aswith the few other studies in this field, there is noway to determine what portion of the license activ-ity and its costs could be attributed to nonprolifer-ation export controls as opposed to other kinds.

Finally, even in the absence of export con-trols, companies may find it worthwhile tomonitor the character of their buyers: mostcompanies would prefer not to contribute tothe violation of U.S. and international nonpro-liferation norms, and they certainly do not wantthe bad press that can come from revelations thatthey have done so.

More difficult to measure are the indirectcosts of lost business attributable to exportcontrols. In at least some cases, the export licensereview and approval process seems to have takenso long that potential buyers have sought othersuppliers in other countries who could deliver or-ders more promptly. In other cases, the Office ofExport Licensing approves export only withconditions intended to assure that the items will beused for stated purposes at stated places. One suchcondition is the requirement for a reexport 1icense:the buyer must agree to apply to the U. S. govern-

ment if he wishes to transfer the commodity to athird country. Another condition is sometimesthat the end-user must agree to accept inspectionsby U.S. personnel to assure that the items are be-ing used for the stated purpose.

Rather than accept these conditions, buyersmay seek other suppliers in other countries that donot impose them. Again, in the case of nonprolif-eration controls, other members of the multilateralexport control groups also require permission toreexport. Finally, when buyers either are deterredfrom ordering in the first place or cancel orders be-cause of licensing delays, the would-be exportermay lose not only the initial deal, buy any follow-on orders that might have succeeded it. Althoughit is logical that export losses result from these fac-tors, direct, or even indirect, statistical evidence ishard to come by.

Whatever the burdens of the export control sys-tem on industry, it is important to keep in mind, asnoted at several points above, that only some ex-port controls are imposed for nonproliferation rea-sons (see next section). Several of the mostprominent industry problems have not beenwith controls directly related to the means ofproducing weapons of mass destruction, but

lgFf)r [he resu]ls of one eff(wt to detect U.S. expwt shortfalls [() other COCOM members, sce Richardson, ibid., pp. 102- I o~. me author

concluded that his research findings did not support the hypothesis that trade with COCOM partners was adversely affected.

Chapter 4 Weighing Benefits and Costs [29

with controls imposed for other purposes. Inthese instances, the industry objections have beennot just to the burdens of the licensing process, butto specific license denials. For example, high-speed telecommunications switching equipmentis on the “national security” list of items requiringIVLS. Until the end of March 1994, the U.S. gov-ernment blocked the sale of such equipment toChina on unspecified national security grounds.Manufacturers argued that they were losing mil-lions of dollars in sales of equipment that the Chi-nese would either buy elsewhere or end up makingfor themselves, thereby frustrating the U.S. exportdenial purpose anyway.20

The most ambitious attempt at estimating thelosses attributable to export controls estimatedthat national security export controls on dual-useitems to Communist countries cost between $4.5and $20 bill ion in lost exports in 1989, while tradesanctions against several other countries cost be-tween $2.4 and $3.1 billion.21 The author of thatstudy later estimated that for 1993, these controlsmay have cost $20 billion, and perhaps as much as$30 billion, in U.S. exports a year.22 However,there are many uncertainties in such an analysis;moreover, projecting the findings of that studyinto the future seems questionable, given the endof the Cold War, the decline in Coordinating Com-mittee for Multilateral Export Controls(COCOM), the relaxation of controls on comput-ers and telecommunications equipment, and thecollapse in buying power of the former Commu-nist countries. This method also prov ides no directway of disaggregating the effects of different typesof export controls on the larger economic picture.

For further illustration of the difficulties of esti-mating the economic impact of nonproliferationexport controls, see appendix A.

Beyond the immediate effects on individualcompany profits, reduced (or constrained) exportscan mean a worsened balance of international pay-ments for the U.S. economy. In terms of dollarvolume of exports, however, the potential impactof export controls appears to be relatively small—and declining. Table 4-1 shows that in 1992, thetotal value of U.S. exports for which applicationsfor IVLS were received ($23.9 billion) amountedto about 3.7 percent of all 1992 exports of goodsand services, about 8.5 percent of exports of in-dustrial supplies and capital goods, and about 22.3percent of those of one analyst’s estimate of ad-vanced technology exports. (Note that IVLS aregenerally valid for 2 years, so the value of theitems in a license applied for in a given year doesnot necessarily y correspond to the value of the ship-ments the exporter intends to make during thatyear. On the other hand, other exports, approved inthe prior year, may be shipped during that year. )The dollar value of license applications either de-nied or returned to the applicant without action23

in 1992 represented only about 1 percent of totalU.S. exports in that year, while the value of li-censes actually denied amounted to about one-tenth of 1 percent.

Figure 4-1 shows that both the number of trans-actions for which an IVL was required, as well asthe value of the items represented, has declineddramatically since 1989. This decline is due main-ly to the relaxation of COCOM controls (see be-

20SOIIK” also argue that dtmml c}f U.S.-made ]tems may cause a country to develop Indlgencwsly the lcchn(~h~gy that they could not buyabroad, ha~ ]ng s(~nw potcntlal for creating or fuel mg foreign competitors” f{~r U.S. exporters of those itcms.

~ ] Richardson op. cit., f(wtnotc 17, pp. 96-97. Armmgst countries of prwl iferation concern on the latter list, Iran accounted for the largest

cxp(m sh(wtfalls, cstlmatcd between $1.3 and $1.8 bill l(m.

22 Rlchilrds(~n, “Ec(m(mlic Costs of US Expwt Controls.” Statement before the Subc(mm~ittcc (m Ec(m(m~ic Policy, Trade, and Environ-ment, C(mmut(ce t m Foreign Affairs, U.S. H{mse of Representatives, N(n, 18, 1993.

~ ~Appllcatl(ms may be rctumed to the appllcant wlth[~ut acti(m h-cause they arc incorrectly or Inc(mlpletely filled t~ut, or h>causc the D(Xrequires :iddltl(mal Inf{mnatl(m to make a dcc] sl{m. As [he DOC tal I ics I Icensing act i~ Itlcs, resubmitted appl icatl(ms, If appr(wcd or denied, arectmn[ed iigiiln under th(lse categories. Thcrcforc, [me fhoulct not asiunw that a l]ccnsc returned with(n]t actl(m represents an cxp(wt that is neverappn~~ cd.


1

1

20

00

80

60

40

20

0

.1 16.4 ._—....—.–— I

II 60.7.;

1989 1990 1991 1992 1993

❑ Amount approved Amount denied

+ Licenses processed

80

-J

20

0

The bars in this graph indicate the values of IndividualValidated Licenses approved or denied by the Department ofCommerce m the years shown, whale the line markers (seeright-hand scale) show the combined numbers of licensesapproved or denied. Many license applications each year are‘“Returned Without Act/on” (RWA) because they are improperly

filled out or because more informatlon IS needed Of theapplications RWA, some may not be refiled while others maybe, the latter are then considered to be new applications andmay enter the counts of those approved or denied Note thatlicense approvals are 2-year authorizations to export, and thatexporters do not report to the Commerce Departmentwhether the licenses are fully utilized or not. Therefore, thedollar amounts approved for export do not represent actualvalues of goods shipped in any given year.

SOURCE Department of Commerce, 1993, and Off Ice of TechnologyAssessment 1994

low), which is likely to continue, whateverCOCOM-successor arrangements are negotiated.In the fall of 1993, the DOC announced easing of

controls on computers, which would mean a fur-ther decline in licensing, since computers ac-counted for about $8 billion in individual licensesin 1993 (see app. B on computer export controls).At the end of March, 1994, as COCOM was abol-ished, the DOC announced further relaxation ofcontrols on telecommunications and computingequipment. It estimated that the number of IVLSrequired annually would drop to half of the 1993level.

By the nature of the commodities, technology,and software controlled, the burdens of exportcontrols do fall more heavily on certain high-technology industries (see app. A). To the extentthat the controls of these industries’ exports leadto loss of business, they lead also to the loss ofhigher paying jobs and of tax revenues.24 In addi-tion, if some U.S. high-technology industrieswere to decline, the U.S. military might lose thebenefits of their research and products. In explain-ing a Clinton administration relaxation of exportcontrols on computers, then-Deputy Secretary ofDefense William J. Perry said:

We’re an important customer [of the computerindustry], but we’re no longer a dominant cus-tomer. Basically, our strategy today in comput-ers is to get on the shoulder of the computer in-dustry and take advantage of the developmentswhich are taking place .25

That is, the Defense Department’s ability toembed advanced computer technology in itsweapon systems depends increasingly on ad-vances made first in the commercial sector. ThusSecretary Perry argued that the relaxation of con-trol levels was justified in part because of the stra-tegic benefit of reducing the burden on the indus-try and enhancing its exports.

24JtJhn Sullivan Wilson points out thatW(wkcrs tmlpl(~ycd in hi~h-/echno/o,gy indu.$fries recei~c higher levels of compensation than all (~(her goods-producing businesses, andthe premium paid these workers is growing . . it is clear that, to the extent the United States continues to pursue a trade policy that isfocused (m the opening of’ global markets and trade expansi(m, this will provide for greater employment oppwtuni[ies in relatively bet-ter-paying, high-techn(dogy jobs.

See ‘The U.S. 1982-93 Performance in Advanced Technology Trade,” Cha//cnge, January-February 1994, p. 16. Wilson also points out thatalthough available data indicate that the United States has been doing well in high-technology exports, trade and technology policy makers needbetter data sets than those now collected.

25 WilIiam J. Perry, transcript of Breakfast with Rep(mters, Oct. 15, 1993 (venue not stated).

Enhancing

P■

■

■

the

olicy options for enhancing the effectiveness of nonpro-liferation export controls include measures to improve:

● processes for making lists of controlled items andbuyers,

administration of export license application evaluationsand the enforcement of regulations,the enforcement of laws and regulations, andthe degree of international cooperation on multilateral exportcontrols.

IMPROVED EXPORT CONTROL LIST-MAKINGI Improve Information and AnalysisFormulating and reviewing the contents of export control lists in-volves identifying goods and technologies that could contributeto weapon programs as well as identifying programs and coun-tries of concern.

Maintaining and strengthening intelligence collection andanalysis capabilities are important to identify and track prolifera-tion activities. The U.S. intelligence community has establishedan office for this purpose. To be most effective, however, intelli-gence analysts make full use of information available from otherU.S. government agencies and from open sources. At the sametime, a diverse array of officials and experts outside the intelli-gence community also plays a role in producing the export controllists of goods and target countries. The fullest possible coopera-tion among these players is essential for the government as awhole to develop the most effective policies. No technical fixescan substitute for such cooperation, but, if it exists, several op-tions are available for bettering communication among policy-

ExportControls 5

Since there are very few

technologies useful to

proliferant weapons

programs that the United

States produces uniquely,

international/ cooperation

among potential suppliers

or transshippers is

essential to effective

export controls.

I 31


makers and improving the base of information andanalysis from which they arrive at their decisions.

One option is to create a common databasethrough which all the involved analysts sharethe widest possible range of information avail-able. The Department of Energy (DOE) is creat-ing what might be a prototype for such a system inits “Proliferation Information Networked Sys-tem” (PINS). PINS is intended first to assist DOEin carrying out its nuclear nonproliferation re-sponsibilities within the government (both in pol-icymaking and export license application re-views), but in principle the idea could beexpanded beyond DOE and beyond a focus on nu-clear nonproliferation alone. This classified com-puter network would permit full-text searches andretrievals of information and analyses about cur-rent and past export license actions, nuclear-weap-on-related technologies, foreign countries and endusers, national policies of both the United Statesand other governments, and international agree-ments and policies. This information should helpanalysts better identify countries of possible pro-liferation concern, the types of goods or technolo-gy that proliferant organizations may be trying tobuy, and the international networks of supply thatthey may be using. Such analysis would be usefulboth in developing lists of countries and items tobe controlled and in making decisions aboutwhether to approve particular export license ap-plications.

The DOE PINS is being designed to let usersget access to multiple databases in multiple on-screen windows, comparing and synthesizing in-formation quickly and easily. Newer computers,with higher speed and more memory, allowconsideration of new techniques for sifting andanalyzing information. For example, the De-fense Department’s Advanced Research ProjectsAgency has been sponsoring research on new

ways of searching textual databases for all in-formation relating to a particular subject, ratherthan requiring the user to discover the exact com-bination of keyword searches that will yield there-sults he or she wants. ]

In another example, Australian economist JohnGalloway has developed a system called “NET-MAP,” which lets users integrate and correlatedata from many sources into a single graphic envi-ronment. It creates visual, color-coded representa-tions of connections and patterns among people,organizations, or transactions.2 The developer haslicensed this technology to various companies formany purposes, but it appears to be particularlywell suited to proliferation analysis. For example,data from export license approvals and denials, fi-nancial transactions, customs discoveries and in-vestigations, insurance underwritings, the tradepress, and intelligence sources might be combinedto reveal the kinds of clandestine procurement net-works Iraq used to supply its nuclear weapon pro-gram. For such analysis to be most effective, theanalysts should have access to the full range ofinformation about all types of proliferation.Several of the countries suspected of trying toacquire one type of weapon of mass destructionalso appear to be trying to acquire the others:they may try to use the same procurement sys-tems.

The technology exists to build an interagencynetwork that would expand beyond the DOEPINS and beyond the nuclear nonproliferationmission. Still, those attempting to do so wouldhave to overcome at least three hurdles. First, therates of data flow would be high, so the agenciesusing the system would have to be connected bysecure, high-capacity links, probably fiber opticcables. For example, the bandwidth (i.e., theamount of information that can be sent in a given

I For ~ ~re~~ ~ccount of the ARpA research, See Michael W. Miller, “U.S. Spies Help Scientists pierce Data Jungle,”’ wail .’$[reetJ~14rn~/, JUIY

27, 1993, pp. B I and B8.

‘Citing this system as an example should not be taken as an OTA endorsement. See Clive Davidson, “What Y(mr Database Hides Away,”New’ Scientist, Jan. 9, 1993, pp. 28-31. OTA was also briefed by a U.S. firm, ALTA Analytics, Inc., adapting NETMAP to various governmental

and commercial tasks.

Chapter 5 Options for Enhancing Export Controls | 33

time) connecting the current experimental PINSsites is not great enough to allow the transmissionof much image material from one site to another.The interconnections would not be technicallychallenging, but might be expensive.

A second hurdle is both bureaucratic and finan-cial: all the involved agencies would have to buyand be prepared to maintain similar computerequipment and software. Citing different needsand ways of operating, different agencies oftenresist such coordination. Multiple agencies arealready developing their own, unique prolifer-ation databases for internal use.

Third, although an interagency network couldhandle classified information, there are good rea-sons to keep certain categories of data compart-mented (available only to certain classes of users)as well as classified. The Atomic Energy Act re-quires DOE to protect Restricted Data having todo with nuclear weapons; within that category,some subcategories of information are dispensedonly to those with a particular “need to know” and,for some types of information, whose terminalsare located in specially secured areas. Therefore,great care will have to be taken in finding the levelof detail about nuclear technology to put in the da-tabase that will be useful to all the participants butat the same time does not put nuclear weapon de-sign information at undue risk.

The other category of specially protected in-formation relevant to a nonproliferation networkis intelligence. Ideally, the database would containall the necessary intelligence findings about po-tential proliferant programs without revealing thesources and methods behind those findings. Inpractice, this can be difficult: the very fact that thegovernment possesses some information cansometimes indicate where that information musthave come from.

Computer networks can be designed to allowvarious levels of access to information, dependingon the clearance level of the user or the terminal.This should not bean insurmountable problem for

the PINS or some extended version of it. The larg-er problem is deciding what levels of informationusers need to do their jobs properly.

MORE EFFECTIVE LICENSINGADMINISTRATIONBuilding lists of controlled items, countries, andend-users is just the first step in the administrationof an export control system. The next, equally im-portant, step is the administration of the licensingprocess. Several measures arguably have potentialfor increasing the effectiveness of the licensingprocess.

Improved U.S. Government ComputerSystems for License Evaluation

The Department of Commerce (DOC) now has acomputer system, installed in 1985, that tracks ex-port control license applications.s License appli-cants can submit applications directly by comput-er; alternatively, the Office of Export Licensing(OEL) will digitally scan paper applications intothe system. From then on, the license application,with the supporting information supplied by theapplicant, can be routed electronically through theOEL and to other agencies to which the applica-tion might be referred for review.

There are things that the current systemcannot do that might both speed up and bringbetter information and judgment to bear on li-censing decisions:

allow reviewing analysts of any agency towhich licenses are referred to extract, in realtime (i.e., as the analyst sits at his desk consid-ering an application) data about previous rele-vant decisions or other current applicationsdealing with the same types of commodities,sellers, buyers, or proposed end-users and end-uses;supply, on the same computer screen at thesame time, technical background information,

l~c Sys[erll Is Called th~> “Ex.Fwt Ctmtri)l Aut(muitcd Suppwt System,” (w ECASS


current intelligence, or other information aboutparties to the proposed transaction; andpermit access to all potentially relevant in-formation among all participants (in whateveragency) in the license review process.

The computer technologies described above inthe section on improving the list-building processcould be even more usefully applied to bring thesefeatures to the licensing process. Indeed, as theDOE envisages for its PIN system, the same com-puter network could serve both activities at once.The hurdles to installing such a system that arealso described above would still apply.

Increase Public Accountability ofLicensing Decisions

Today, DOC issues an annual list of licensesgranted for commodities to restricted nations,with summary data about the number of licensesgranted for each type of commodity and the dollarvalue for each type going to each nation. Recentlegislation proposed in the Congress would re-quire that within 6 months after issuance of a li-cense to export any nuclear dual-use item, theSecretary of Commerce would publish the com-modity description, the country destination. theend-use and end-user, the quantity, the date of ap-proval, and the date and method of shipment.4

Speaking in support of this legislation, SenatorGlenn said:

The present system of nondisclosure has led,especially in the case of goods sent to Iraq, to acrisis in public confidence that America has itsown export control house in order. The best wayto restore that confidence and to ensure more ef-fective oversight and accountability is to permitgreater public scrutiny of thecensing data.5

nonproprietary li-

Some analysts have advocated that exportingcompanies should be identified as well:

Congress should now require the CommerceDepartment to publish quarterly summaries ofall dual-use licensing actions . . . The list wouldonly cover licensing actions that have been com-pleted. Pending sales would not be reported.b

Advocates of transparency in licensing deci-sions have been interested primarily in public ac-countability:

Pushing export licensing into the light of daywould encourage the exporters to be honest, en-courage the government to be careful, and allowthe public to find out whether U.S. exports areundermining national security.7

There are other possible benefits from mak-ing the information openly available. First, itwould enhance unclassified analyses by non-gov-ernmental investigators of export-import patternsthat might identify previously undetected weaponprograms or supply networks (see below, in thesection on improving multilateral export controls,for the benefits of strengthening unclassified ana-lytic efforts). Second, it might set a precedent forhelping to persuade other nations to release com-parable information, thus easing the task of bothgovernments and nongovernmental groups inidentifying possible avenues of proliferation.Third, one critic of the current system has arguedthat revealing all licensing decisions (includingdenials and returns without action) would increasethe fairness of the system by letting all sellersknow what types of exports had previously beenapproved. This information, he says, would allowany firm both to predict better whether its own li-cense application is likely to be approved and togive it a basis for appeal of what it believes to be anunfair decision. If, as this author argues, licensing

g~e ‘“Nuclear EXpoII R~organi~alion” Act of 1993,” bill S. 1055 introduced May 27, 1993, Sec. 3 10; the amesponding House bill was H.R.

2359.

Sj{)hn G]enn, ConKres~lona/ Record (May 27, 1993), Daily e~.~ s677~.

~Ga~ Mi]h(~l]in “Licensin~ .Mass Des[ruc[km: U.S. Exp)rts [() Iraq, 1985- 1990, ” manuscript, Wisc(msin ~OJect on Nuclear A~s control,”

June 1991, p. 14.

‘Ibid.


decisions are to be subjected to judicial, as well asadministrative, appeals, such information be-comes all the more necessary.8 Commerce De-partment officials, however, say that each licenseapplication is judged on its own merits under theparticular circumstances; therefore, the experi-ence of previous applicants would not tell the ex-porter much about how his own application wouldfare, either in administrative or judicial review.

There are some object ions to this level of trans-parency in the licensing process. First, companiessubmitting license applications worry about therevelation of proprietary data that would compro-mise competitive advantages. Although some ad-vocates of releasing licensing information arguethat companies have no reason to conceal legiti-mate sales, the question is somewhat more com-plex.9 For some goods or technologies, the factthat certain companies have found (entirely legiti-mate) buyers for particular products could tip offcompetitors to explore markets previously onlyknown to the company applying for the 1icense. Insuch cases, the biggest losers could be the mostsuccessful firms: information about their custom-er bases would be revealed to competitors whohad not yet penetrated the market as well.

In addition, license applicants sometimes sup-ply a considerable amount of detailed data abouttheir products to support their applications; expo-sure of that data and of pricing information couldgive advantages to their competitors for legitimatesales. On the other hand, it should be possible toexclude these more detailed proprietary data fromthe public domain. Finally, since license approv-als are good for 2 years, an approved license maynot correspond to a completed sales agreement;therefore, means would have to be found for pro-

tecting exporters from competitors’ exploitationof information about uncompleted sales. One op-tion would be to require exporters not only to ap-ply for licenses, but to report to the governmentwhen and to what extent the shipments licensedactually took place. This would have the addition-al benefit of providing more complete informationabout international trade patterns in sensitivetechnology.

Publishing licensing data would permitmore external oversight of governmental deci-sions. It would also expose those decisions to thepossibility of politicized second guessing. Out-side observers will question both individual deci-sions and the overall pattern of decisions—that isthe point of public accountability. Those criti-cisms may often be justified; at other times, theywill not. In either case, they will not be made withthe full range of classified and proprietary in-formation available to the decision makers. De-pending on the prevailing political atmosphere,the anticipation of external criticisms (includingthose from Capitol Hill) could lead licensing offi-cers either to be hesitant to approve exports (thusrestraining legitimate business) or reluctant todeny them (thus increasing proliferation risks).

Strengthen Interagency ReviewProcesses

Procedures for referring export license applica-tions to other agencies outside the DOC are meantto assure that those agencies can bring to bear:

■ a broader range of substantive and technicalknowledge and judgment than is available inany single agency, and

‘Ht~ward N. Fent~m, “Reff~m]]ng the Prxwxlurcs of the Expwt Admlnistratt(m Act. A Call for openness :ind Adrnlnlstratii c Duc Process,’”Iewi In[crnotl{mal [,a~% Journal, vol. 27, w Inter 1992. p. 61.

‘) SW, c,g.. Kenneth R. Tlrnmemvm, “Tinw for a N(mProllfcrat](m Agenda.”’ L’je on ,SIIpp/~, w Inter 1993, p. 78, S]rnilari), Mllhf)llln argues:If a company IS ashamed of ha~ Ing sold (me of its pr{xlucts to a clevcl(~plng count~, the company should not have made the sale in the

first place. Repu(ahlc cornpan]cs do not object [() tcll]ng the truth alx)rrt their business. If the sales arc lcg]t}mate, and satisfy the exportcntcrla, there is no reason to keep thcrn hlddcn.

op. cit., footnote 6, p. I 4.


● the other agencies jurisdictional perspectiveson U.S. national security and foreign policy in-terests.

It is also possible that additional analysis bymore than one agency could catch problems thatonly a single-pass review might miss. Some crit-ics of the current administrative arrangementsfor reviewing applications have proposedchanges intended to increase assurances thatthe referral process will block inappropriate li-cense approvals.

For example, one proposal is that all prolifera-tion-relevant applications be automatically re-ferred to the Defense Department, which wouldmanage further referrals and make the final 1icens -ing decision.

10 The basis for this proposal is the

author’s judgment that in the 1980s the Com-merce Department issued numerous export li-censes for Iraq without referring them to the prop-er external agencies. In this author’s view, thereason for these failures is the “conflict betweenthe Commerce Department’s duty to promote ex-ports and its duty to regulate them.”11 It shouldalso be recalled, however, that during the 1980s,high-level U.S. policy was tilting toward Iraq inits war against Iran, and it may have been Admin-istration political judgments—rather than Com-merce Department zeal for export promotion—that led to questionable license approvals.Commerce officials say that during this period,the Department referred applications to the De-fense Department according to mutually agreed-upon procedures.

Another proposal has been to give the ArmsControl and Disarmament Agency (ACDA) astronger role in the export licensing process. Sen-ate Foreign Relations Committee legislation pro-posing this step in 1993 was put on hold at the Ad-ministration’s request until the Administrationcould prepare its proposed revision of the Export

IOMi]ho]]in, op. Cl(., W. 12- 13“

11 Ibid.

Administration Act (EAA). In exchange for thedelay, Administration officials offered the Com-mittee assurances that in the meanwhile theACDA role in dual-use export control review anddecision making would be strengthened.lz

However, attempting to produce policy shiftsby legislating structural changes may not al-ways produce the desired effects. For example,although in the past the Defense Department fa-vored the strictest of export controls, officials atthe highest levels of DOD strongly supported theClinton Administration’s raising of controlthresholds for computers in the fall of 1993.

Another proposal for increasing participationof other agencies is to require the DOC to send in-formation copies to one or more other agencies ofthe licenses it intends to approve but does not in-tend to refer formally to those agencies. Such aprocedure might have two benefits. First, it wouldgive the other agencies additional information touse in their own analysis of international trade pat-terns relevant to proliferation. Second, dependingon how long before actual license approval the in-formation came in, it would give the other agen-cies the opportunity to make the case with DOCthat they should have the opportunity to reviewcertain applications. Even though DOC may bemaking a good-faith judgment that its referral po-licies were consistent with interagency under-standings, differences of interpretation mightarise in particular instances.

Short of removing export licensing manage-ment from the DOC, the interagency reviewprocess for certain types of referral could befurther formalized, with greater authority giv-en to interagency groups. Under current proce-dures, DOC refers Nuclear Referral List items (aswell as transactions involving known nuclear end-users) to DOE, and to other agencies, according torules agreed on between DOC and those agencies.

1 ‘U.S. Congress, Senate, C(mmlittee (m Foreign Relations, The Arms Conrro/ and Nonpro/~era/ion Act oj’/993, Rqx)rt 103-172, Nov. 5,I 993.


If either DOC or DOE believes either that the ap-plication should be denied, or that it should under-go further review, then it is referred to the inter-agency Subgroup on Nuclear ExportCoordination (SNEC, representing the Depart-ments of Defense, Commerce, State, and Energy,as well as the Arms Control and DisarmamentAgency and the Nuclear Regulatory Commis-sion). In calendar year 1993. 740 applicationswere so referred. The SNEC either provides itsunanimous information and advice to DOC orsends the application for higher level review. Theproposed Nuclear Export Reorganization Act of1993 would have legislatively established theSNEC within the National Security Council, re-quired it to review applications to export any itemon the Nuclear Referral List, and give it deciding(rather than just advisory) authority over li-censes.

13 The group would also have been respon-

sible for maintaining the Nuclear Referral List it-self. The purpose of such a role for the SNECwould be to assure that the full panoply of in-formation and expertise available in the govern-ment would be brought to bear on every Iicensingdecision.

Formalizing the interagency review process inthis way would probably impose the costs ofcreating a new bureaucratic unit, complete withstaff and administrative support. It would also re-quire a new computer system able to manage re-cords of discussions and decisions for both Iicens-ing processing and export control targetingpurposes. On the other hand. such a computer sys-tem would probably be useful whether a new bu-reaucratic unit were created or not. If the new unitwere created, the agency personnel already per-forming the licensing review and 1ist-constructionfunctions could be assigned to the SNEC full-timerather than part-time, so the net additional costmight not be high. Another benefit of having a for-erealized, routine, and well-staffed interagency re-

view process could be to shorten the time that li-censing decisions now take. That might helpanswer one of the major industry complaintsabout the current process (see below, ch. 6, p. 64).

I Increase Export Control AwarenessAmong Exporters

Through the Federal Register and through publi-cations of its own, the DOC informs exporters ofexport control regulations and of countries andend-users of particular concern. One proposal isthat the SNEC publish a regular bulletin thatwould expand on such efforts to include informa-tion on

. . . regulations, international agreements, and

other relevant developments [to inform] export-

ers and the general public about the risks of pro-liferation and efforts to reduce or eliminate such

risks. 14

Such a publication could also cover other typesof proliferation besides nuclear. Extensive publi-cation of government information on suspect pro-grams would be one way of enabling exporters tocooperate with nonproliferation efforts. Accord-ing to the U.S. Customs Service, industry is itsbest source of information about illegal acquisi-tion attempts. Thus, it is important not only thatcompanies comply with export regulationsthemselves, but that they report approachesfrom buyers who may be trying to evade theregulations. (The Commerce Department and theCustoms Service already have publicity programsaimed at informing exporters of possible indica-tions of illegal exports. )

To encourage cooperation by U.S. companies,the United States could permit firms to petition forinvestigation of, and possibly sanctions on, for-eign companies that they suspect are undercuttinginternationally agreed export controls. 5 This


process might both help assure U.S. firms thatthey would not have to face unfair competitionand help provide the government with more in-formation on possible avenues of proliferation.

IMPROVED ENFORCEMENT

I Increase Enforcement ResourcesThe DOC Bureau of Export Administration,cooperating with U.S. diplomatic posts abroad,the U.S. Customs Service, and foreign gover-nments, sometimes conducts pre-license checksand post-shipment verifications on the end-usersnamed in export license applications. The re-sources now avail able for conducting these checksare extremely limited. Nor, according to a 1993 re-port by the DOC Inspector General, have they al-ways been administered systematically, efficient-ly, or according to established guidelines. Inparticular, random checks lacked a strategic plan,with stated purposes and priorities. In fiscal year1992, commercial officers at foreign posts con-ducted 568 pre-license checks and 177 post-ship-ment verifications, of these, 65 to 75 percent wererandom checks, while the remainder resulted fromderogatory information.16

Checks and verifications could, first, be sys-tematized. More effective sharing of data andanalysis (as might occur with other options dis-cussed in this report) could help the DOC and oth-er agencies plan a more coherent checking strate-gy. Second, checks could be increased annually ona stepped basis, with each increment of checkingactivity weighed against the number of undesir-able buyers that it revealed. If the checks were ef-fective, one would expect to see an initial jump insuch discoveries, followed in a few years by a de-cline as the threat of discovery deterred more ob-jectionable buyers. The trick would be to find thepoint of declining marginal returns-the incre-ment of expense in checking that did not produce acommensurate increment of deterrence.

Both checks and verifications are best seen asmeans of gathering evidence about the credibilityof buyers before a new or additional license isgranted (or before additional shipments are madeunder an existing license). Checks may help weedout obvious front-company buyers and firms oragencies that have clearly misrepresented theirfunctions. However, unless an item has actuallybeen removed from its intended site, post-ship-ment checks may not easily detect whether itemsare being used for their stated purposes. If some-one is misusing a controlled item at the declaredsite for its legitimate employment, that user isprobably capable of concealing the fact from theU.S. Foreign Commercial Service officers whousually do the checking. If the item has been di-verted elsewhere, little information (except thatthe buyer is untrustworthy) has been obtained; re-trieving the transferred item is unlikely to be anoption.

In addition to pre-license and end-user checks,Commerce Export Administration and Customsenforcement officials conduct investigations,gather evidence, and make arrests. Resources forthese more traditional law-enforcement activitiescould be increased.

I Reinforce the “Know” RuleCurrent regulations require companies to obtainIndividual Validated Licenses (IVLS) for exportsof any items, listed in the Commerce Control Listor not, that they know (or are informed) are des-tined to be “directly employed in” the design, de-velopment, acquisition, or use of missiles orchemical or biological weapons in a country listedin one of the supplements to the Export Adminis-tration Regulations (EAR). The Department alsooffers guidance about what an exporter can rea-sonably be expected to do to avoid customers whomay be engaged in inappropriate end-uses. Anexporter who has even “reason to know” that

I ~~f[ices ~)f lnsP.c[or Genera] :~t the U.S. D~p~rtnlm[s t)f Commerce, Defense, Energy, and sla~c, “The Federal Government Export Li-

ccns]ng Pnmsses for Muniti(ms and Dual-Use Cimml(tiitics: Special [ntcragcncy Review, ” September 1993, pp. A 13-A 17.

I 758 Federal ReK151er, fjIU)29-6f103 ] , h’C. 23, 1993.


items or data might be used in a nuclear programmust also apply for a license. 18

Exporting industries have argued stronglyagainst maintaining this rule; its pros and cons arediscussed further in chapter 6, on options for re-ducing industry burdens. Should it be retained,however, the kind of bulletin described in the sec-tion above could help assure that exporters havesufficient reason to know about risky exports to beheld accountable. The Commerce Department al-ready encourages companies to request advisoryopinions from the government as to whether acontemplated export would subject them to legalsanctions.

Expand Computer Network ResourcesThe same sort of computer network, discussedabove, that could enhance list-making and licens-ing could also assist enforcement personnel inidentifying suspicious transactions. Enforcementofficials at Commerce Bureau of Export Admin-istration (BXA) have been studying varioussources of data beyond those available from intel-ligence and law enforcement agencies to see ifthey might help reveal suspicious export patterns.For example, U.S. Census data on all the types andquantities of items going to a particular countrymight reveal patterns of imports suggesting diver-sion to a proliferant weapon program. Thus far,however, Commerce has not had the resources toput this sort of analysis into the context of a larger,more encompassing database of the type de-scribed above.

I Extend SanctionsU.S. laws provide for penalties against U.S. per-sons (individuals or firms) who violate U.S. ex-port regulations. In recent years, Congress has at-tempted to bring sanctions to bear on others whoaid proliferation as well. In 1991, Congress ex-

tended sanctions to foreign persons whose exportsmaterially contribute to either chemical or biolog-ical weapon programs. The sanctions are bans onU.S. government procurement from those personsand on any United States imports from them. A1990 law also imposes various sanctions on for-eign persons who violate the Missile TechnologyControl Regime (MTCR), including a ban on U.S.imports from those whose exports have substan-tially contributed to a non-MTCR adherent mis-sile program. Similar sanctions could be institutedfor those engaging in illicit nuclear exports.

The Administration draft EAA attempts to con-solidate the sanctions provisions of the currentlaws on chemical and biological weapons prolif-eration and use and on missile technology control.The sanctions section of this draft bill is summa-rized in table 5-1.

The subject of sanctions in export controls is aconfusing one because of the circular relationshipbetween the two: sometimes sanctions are toolsto enforce export controls, and sometimes ex-port controls themselves are the sanctions.Moreover, export controls adopted for one pur-pose are applied as sanctions for another purpose.It is important, therefore, to make judgmentsabout controls and sanctions in the context of theirpurposes. In particular, from the standpoint ofnonproliferation policy, the utility of export con-trols intended primarily to deny access to itemsthat directly contribute to proliferation should notbe judged on the basis of their effectiveness or costin efforts to punish some nations for their supportof international terrorism.

When economic sanctions are applied for anypurpose, they usually pose dilemmas for policymakers. First, they impose costs on the UnitedStates as well as on the target of the sanctions: ex-port bans cost sales to U.S. firms; import banskeep out things that U.S. consumers may want or


Chemical, Biological Weapon (CBW)Action or Missile Proliferation

—- — —

Presidential Any foreign person has knowingly, or with reason to know,determination contributed materially to the efforts of any government,

group, entity, or project to use, design, develop, produce,stockpile, or otherwise acquire chemical or biologicalweapons (or missiles)

■

Mandatorysanctions

Discretionarysanctions

Mandatorysanctions forCBW use

Discretionarysanctions forCBW use

■

●

through the export or transfer of any chemicals, biologicalagents or equipment which may contribute to a chemical orbiological weapons program such as those listed by theAustralia Group (or items listed in the Missile TechnologyControl Regime annex) whether or not of U.S.-origin, orby participating in any financial transaction related to thedescribed activity orby facilitating the described activity

the Australia groupDenial of exports of items controlled by(or the MTCR annex)

Imports of such items from such entities prohibited

In event of CBW proliferation, President may choose any ofthe 11 actions Iisted as mandatory or discretionary for useof CBW (see rows below), such sanctions shall be propor-tionate to the harm the sanctioned behavior has caused tothe national security or nonproliferation interests of theUnited States

1)

2)

All

Chemical, BiologicalWeapon Use

The government of a foreigncountry has used chemicalor biological weapons inviolation of international lawor used lethal chemical or bi-ological weapons against itsown nationals,Within 3 months of the abovedetermination, violation gov-ernment has not■ ceased use,■ provided reliable assur-

ances of non-use in thefuture, and

■ agreed to on-site inspec-tions to verify non-use

of the sanctions listed belowas Mandatory for CBW use

If President makes second de-termination above, he mustimpose at least 3 of the follow-ing 6 listed below as Discre-tionary for CBW use

1) No U S Government procurement for a minimum of 2 years of any kind from or produced by

2)

3)4)5)

1)2)3)4)5)6)

CB-using countryTermination of U S foreign assistance (except urgent humanitarian aid and agriculturalproducts)Termination of U.S. arms salesDenial of U.S. Government credit or other financial aidDenial of national-security sensitive EAA-controlled exports

Oppose loans or other aid by international financial institutionsProhibit any U S bank from making loans or credit except for agricultural productsProhibit U.S. exports to the country of all items except agricultural productsRestrict Importation of articles that are the growth, product, or manufacture of the countryDowngrade or suspend diplomatic relations with the countrySuspend country’s air carriers from engaging in foreign air transportation to or from the U S.

— —


Exceptions

Deferrals 1 President may delay determinations (above) or sanctions to protect ongoing criminal inves-tigations or sensitive Intelligence sources being used to gather further Information on prolif-eration

2 President may delay sanctions for up to 180 days if the U S IS engaged in diplomatic effortsto curtail the sanctioned conduct or obtain sanctions against the person from the govern-ment of jurisdiction over that person If these efforts succeed, U.S. sanctions not required

Sanctions not required in cases ofexport or transfer authorized by, or exports to, a country adhering to the Australia Group or asignatory to the Chemical Weapons Conventiondefense procurement under existing contracts, if the defense articles or services are notreadily available elsewhere, or they are essential to national security under defense co-pro-duction agreementsother Imports under existing contracts, spare parts component parts, information, or technol-ogy essential to U S products or production, routine servicing of products not otherwisereadily available, medical or other humanitarian itemsany transactions subject to the reporting requirements of the National Security Act of 1947performance of prior contracts when barring it not necessary to achieve U S national securityor nonproliferation objects and would be contrary to the national interest

Waivers of President may waive sanctions if he or she determines that a waiver iS Important to the nationalapplication of interests of the U S and notifies Congress not less than 20 days before waiver takes effectsanctions

SOURCE Department of Commerce and Off Ice of Technology Assessment, 1994

that U.S. producers may need.20 Second, they risk

achieving so much distance between the U.S. andthe target state that the latter decides to simplydefy the sanctions and resist all further U.S. influ-ence. Third, if the United States is too far ahead ofthe rest of the international community in impos-ing sanctions, its efforts are likely to be undercutby other nations.

The Administration draft EAA attempts to takeaccount of these dilemmas by granting the presi-dent nearly total discretion in imposing sanctions;essentially, the draft authorizes a wide range ofsanctions, extending up to complete embargo,then adds sufficient deferrals, exceptions, andwaivers to allow him to do nothing if he so de-cides. Broadly speaking, one of two other legisla-tive policies could be adopted:

■ first, 1imit the president ability to defer orwaive one or more sanctions (i.e. mandate themor narrow the exceptions); or

■ second, limit the president’s authority to im-pose sanctions, either in kinds or in duration.

Each of these three legislative approaches—flexibility, mandate, or restriction-has its draw-backs. Granting great flexibility risks that a presi-dent will do nothing when the Congress mightwish that he would do something, or vice-versa.Unconditionally mandating sanctions risks forc-ing the president to take actions in unforeseen cir-cumstances that may be costly but either ineffec-tive or actually detrimental to nonproliferationgoals. Restricting sanctions risks making them

~ON(Jt (rely do sanctltm~ inlp)se C(MIS (m the side Impt)sing the sanctions, but the> fall unevenly on its citizens. Firms that depend on the

c~p{m t~f ct~ntrol led Itcms haf c nl(~rc to I(w than tht)sc that do not, firms that depend on ]mpwts fr[ml the targeted party ha~ c rm~re to lose thanthose who impwt from clsew here. Et cn In a t(~tal enlbargt~ of the target party, particularly if it is an entire nati[m, s(m~e U.S. exporters and impor_-ers w ( mld hwe more than others, depend]ng on the prior patterns of trade he[w een the two c(wntries.


unavailable in circumstances where they might beeffective.

A compromise option would be to permit theflexibility requested in the Administration draftEAA, but to accompany it with more explicit pro-visions for accountability to Congress about thecosts and effectiveness of sanctions imposed. TheAdministration bill already would require assess-ments of economic costs and qualitative estimatesof effectiveness for export controls; presumably,these required reports to Congress would have tobe made when export controls were used as sanc-tions as well as when they were used for their pri-mary purposes of denial. The same kinds of as-sessments could also be required for the othertypes of sanctions listed in table 5-1. Such reports(if their quality were maintained by the demandsof watchful congressional oversight committees)would permit the legislative branch to make inde-pendent judgments on whether executive branchdecisions on the costs and benefits of sanctionswere serving the national interest.

STRENGTHENING MULTILATERALCONTROLSSince there are very few technologies useful toproliferant weapons programs that the UnitedStates produces uniquely, international coop-eration among potential suppliers or trans-shippers is essential to effective export con-trols. The United States has played a key role inthe establishment and operation of the existingmultilateral supplier groups: the Nuclear Suppli-ers Groups (NSG), the Australia Group, theMTCR, the Coordinating Committee on Multilat-eral Export (COCOM), and the COCOM succes-sor. Recent successes include leading the NSG in

1992 to agree to adopt multilateral controls on cer-tain dual-use technologies (see box 5-1 for de-scription of NSG guidelines) and getting Russia in1993 to promise full compliance with the terms ofthe MTCR.21 Additional steps to strengthen mul-tilateral controls are possible.

Keep Conventional andMass-Destruction Weapons onSeparate Tracks

The oldest, most highly coordinated, but also themost contentious, of the supplier-group regimesfor dual-use items was COCOM. COCOM dual-use controls (the “Industrial List”) were intendedprimarily to keep advanced conventional militarytechnologies out of the hands of potential adver-saries of the United States and its allies.22 Differ-ing interpretations of COCOM requirements ledto some disputes between the United States and itsEuropean allies. The administration of nationalsecurity (i.e., COCOM) controls also led to thegreatest complaints of unfairness from U.S. in-dustry.23

With the end of the Cold War, the membership,targets, and listed technologies for any successorarrangement to COCOM (former] y a Western ar-rangement for denying technology to Communistnations, terminated at the end of March 1994) areundergoing significant changes that must be mul-tilaterally negotiated. With technologies applica-ble to weapons of mass destruction already ad-dressed in other multilateral export controlregimes, the COCOM successor regime, ifcreated, will most likely attempt to regulate thetransfer of technologies for developing or makingconventional weapons. Consensus will be diffi-cult to reach, both within the United States and

2] When it adopted new guidelines on dual-use technology transfers, the NSG also adopted a rule, long advt~ated by the United States, thatthe transfer of certain nuclear-related “trigger list” technologies w(mld be conditioned on acceptance by the recipient of IAEA safeguards onany other facilities in the country of the same type to which the technology was being transfemed.

22 COCOM also had a list of nuclear-related technologies, but apparently these had little consequence for U.S. exprt administration be-cause the other nuclear-supplier agreements are rmwe comprehensive. In addition, there was a COCOM list of mil itary equipment, controlled inthe United States under the Am~s Expwt C(mtrol Act.

‘sSee, for example, Th(~mas T. Connelly, “Statement on Behalf of AMT—The Association for Manufacturing Technology-bef(we theSubammlittee on Ec(momic Policy, Trade, and Envir(mment of the House Committee on Foreign Affairs,” Nov. 18, 1993, pp. 6-7.


Supplier should establish export licensing procedures for the transfer of equipment, material, and related

technology identified in the Annex These procedures should include enforcement measures for violations. In

considering whether to authorize such transfers, suppliers should exercise prudence in order to carry out the

Basic Principle and should take relevant factors into account, including

a

b

c

d

e

f

9.

whether the recipient state iS a party to the Nuclear Non- Proliferation Treaty (NPT) or to the Treaty for the

Prohibltion of Nuclear Weapons in Latin America (Treaty of Tlatelolco), or to a similar International legally-

binding nuclear nonproliferation agreement, and has an IAEA safeguards agreement in force applicable to

all its peaceful nuclear activties,

whether any recipient state that iS not party to [the treaties named above] has any [nuclear fuel-cycle facili -

ties] that are operational or being designed or constructed that are not, or will not be subject to IAEA safe-

guards,

whether the [item] to be transferred IS appropriate for the stated end-use and whether that stated end-use iS

appropriate for the end-user,

whether the [item] to be transferred iS to be used in research on, or development, design, manufacture,

construction operation or maintenance of any reprocessing or enrichment facility;

whether government actions, statements, and policies of the recipient state are supportive of nuclear non-

proliferation and whether the recipient state iS in compliance with its international obligations in the field of

nonproliferation

whether the recipients have been engaged in clandestine or Illegal procurement activities and

whether a transfer has not been authorized to the end-user or whether the end-user has diverted for pur-

poses inconsistent with the Guidelines any transfer previously authorized

SOURCE International Atomic Energy Agency INFCIRC/254Rev .1 Part 2 July 1992—

.~mong the international participants, about what ling production of advanced conventional weap-technologies should be controlled tind for whatreasons .24 Therefore, the nonproliferation re-gimes dealing with weapons of mass destruc-tion (and missiles), for which considerable con-sensus has already been painstakingly built,should not be mixed into controversies overCOCOM revisions of technologies controlledfor other purposes. Administration officialshave spoken of moving the emphasis in a succes-sor arrangement from maintaining the West mil-itary technology edge over Communist countriesto limiting the proliferation of technologies enab-

ons to some states not now possessing them. Seetable 5-2 for a comparison of the COCOM andweapons of mass destruction regimes.

A National Academy of Sciences study on ex-port controls proposed either an additional non-proliferation category and regime for convention-al weapons technology, or incorporation ofconventional technologies into one of the existingregimes. 25 The above ar~urnen( favors a separate

negotiating forum for conventional weapontechnologies. Insofar as those overlap with nu-

24FtJr example. st)nw nat itms resist cooperation with nonprt)liferati(m expwt c(mtrol regimes (m the gr(mnds that the Lln]td StatcJ IS scch-lng such C( )n[ro]s Prinlari]y, [t J pro[~ct ][SC[ f frt)nl cconon~ic” ctmlpct]tl(m, alth(mgh th]s IS nt~t true, the iirgument ft~r It is c:isier I(J make In (he caseof cx pm controls” Intended t( J hl(xk the transfer of tcchn(}log]es that might k> usable for a br{)ad range of conk entl ( )nal m i I itary appl lcat Ifms, notjust wca~ms t)f mass dcstructl(m.

2$ Panel (m the Future Dwgn and Inlplmwntat]tm t~f U.S. Namml Sccunt) Ekp{wt C{mtrols, Find/n<q (’onvnon Grcmnd: .!J’. S. Llporf Cw-

tro/.f /n a Ch(Jrrqcd (;/hJ/ L’rr\>IrmnICrrI (W’ashlngttm. DC Nat](mal Acackmy Press. 199 I ), p. 131.


Regime COCOM NSG, AG, MTCR COCOM successor?

Purpose Maintain Western military-tech-nological advantage over tar-get states

Targets States under the control ofCommunist regimes, buyerswho might divert items tosuch states

Scope of Wide range of dual-use com-controls modities, technology, and

software, including those rele-vant to modern industrial de-velopment as well as thosespecifically applicable to de-veloping or producing ad-vanced conventional weap-ons in addition to weapons ofmass destruction

Rationale for Deterrence of Communist ag-international gression by maintain techno-consensus Iogical superiority of allied

over Communist militaryforces

Principle of Consensus all members mustoperation agree to sale of controlled

items

Prevent or slow the spread totarget programs of capabilitiesto develop or produce nuclear,chemical, or biological weap-ons, or missiles

Activities and facilities to devel-op, produce, or otherwise ac-quire weapons of mass de-struction or missiles, buyerswho might divert items to suchactivities

Narrower range of dual-useitems applicable to developingor producing banned weapons

Prevention of threats to interna-tional peace and security frompossession of weapons ofmass destruction by those notalready having them

National discretion guidanceand control lists mutually nego-tiated, but Iicensing decisionsremain at national level

Prevent or slow the spread totarget programs of capabili-ties to develop or produceadvanced conventionalweapons

Rogue nations falling short ofsome standards of interna-tional behavior

Similar to COCOM items

Containment of threats to re-gional or global securityposed by “rogue” or “back-lash” nations

Probably national discretion


clear or missile technologies, they should be cov- should not authorize transfers of the listed items:ered under the latter regimes anyway.

■

Enhancing Nuclear Suppliers’ GroupCoordination

■

In March 1992, the NSG agreed to adopt commonexport controls on a list of nuclear-related dual-use materials, equipment, and technologies. Theyagreed to the “Basic Principle” that suppliers

for use in a non-nuclear-weapon state in anuclear explosive activity or an unsafe-guarded nuclear fuel cycle activity, or

in general, when there is an unacceptablerisk of diversion to such an activity, orwhen the transfers are contrary to the ob-jective of averting the proliferation of nu-clear weapons .26

‘blntematifmal Atomic Energy Agency, INFCIRC/254/Rev. I iPart 1, July 1992, “LCtmlnlunicati(ms Received Fnml Certain Member StatesRegarding Guidelines for the Export of Nuclear Material, Equipment and Technok)gy: Nuclear-Relaled Dual-Use Transfers,” Annex Attach-

ment, p. 2.


The NSG agreed that decisions to approve ex-ports of the items on this list would take into ac-count several relevant factors in deciding whethertransfers were acceptable (see box 5-1 ). One ofthose factors was to be whether a transfer has beenrefused by someone else (i.e., a ““no-undercut”guideline).

One observer has expressed concern about theactual results of this agreement:

The NSG members have agreed to exchangeinformation on non-NSG states’ nuclear pro-grams and dual-technology purchasing activi-=tics, and to meet for consultations at least once ayear. SO long as these arrangements remainloosely specified and relatively uncoordinated,they may not substantially improve the overallquality of the intelligence available to NSGmcmbcrs. 27

This author proposes that the United States. . . should systematically communicate in-formation and share intelligence assessmentswith other NSG members in order to ensure thatmultilateral restrictions on sensitive dual-technologies arc effectively implemented andlists of restricted technologies are properly keptup to date.28

Modern telecommunications and comput-ing technology make it possible to convey muchof this information almost instantaneously.The Department of Energy Office of Arms Con-trol is sponsoring a project on International ExportInformation Sharing, centering on computeriza-tion of information sharing for the NSG agree-ment on controlling exports of dual-use technolo-gies. The types of information to be included inthis database are:■

■

■

export 1icense denials;reference data useful to Nuclear SuppliersGroup members:documents and information related to NSGguide l ines on specifically nuclear-relatedequipment, materials. and technologies: and

■ documents and information related to NSGguidelines on nuclear-related dual-use equip-ment, materials, and technologies.

The database would reside on an internationalcomputer network, with each member state hav-ing an inexpensive terminal linking it to the sys-tem. Besides giving the members access to a com-mon database, the system would also allow themto exchange electronic mail on NSG export con-trol matters. Thus far, 20 NSG members haveagreed to install test terminals for this system, and8 have been emplaced.

Such a network would offer a variety of op-portunities for increased coordination amongthe Nuclear Suppliers. In agreeing to multilater-al controls on dual-use technologies, the NSGmembers also agreed to inform one another whenthey deny export license applications for the listeditems. Timely dissemination of this informationwould allow each supplier to consider its own ex-port decisions in the light of those made by any ofthe others. Once refused an export license in onecountry, a potential buyer would not have a chanceto find another supplier in another country even ifthat country did not have independent reason forsuspicion about him. License denial informa-tion, as well as some of the other kinds of in-formation described below, could be especiallyuseful to governments without the extensive ex-port control infrastructure and intelligence re-sources of some of the larger members of theNSG.

The reference data, documents, and other in-formation in the database would include:

■ official documents. key officials and contactpersons, and various types of supporting in-formation including International Atomic En-ergy Agency (IAEA) information circulars anddata on related international agreements;


copies of other nuclear-proliferation-relevantagreements (e.g., the Nuclear Non-Prolifera-tion Treaty [NPT]) and membership lists;guides to the nuclear fuel cycle, to help ascer-tain the significance of specific equipment ortechnologies; andthe latest information on the control of “TriggerList” items—those directly nuclear-relateditems whose export requires that the buyer sub-mit his facilities that either use, or could use,the items to IAEA safeguards.

Not only would the database provide immedi-ate posting of all denials of 1icenses for transfers ofdual-use materials, equipment, and technology,but it would also constitute a cumulative record ofthe items, suppliers, and proposed buyers in thedenied transactions. Such an organized recordcould help the member governments better identi-fy and act on particular proliferation risks.

In addition to license denial information, thedatabase would include other information on po-tentially risky end-users, such as those with un-safeguarded nuclear activities, or those on variousmembers’ lists of suspected proliferants. It couldalso serve as a funnel for some of the contribu-tions of national intelligence services to themultilateral group. On some occasions, it maybe possible to enter information into such a rela-tively open forum by developing unclassifiedsources to cite for facts first detected by classifiedmeans. (See section below on the utility of sup-porting open-source proliferation analyses. ) Apossible drawback to permitting such contribu-tions to a database would be the risk that inaccu-rate information (intentionally or unintentionallyplaced) would accumulate and be difficult to re-move.

The NSG sharing scheme in principle couldbe expanded by including export license ap-provals as well as denials. With this wider rangeof data about exports with nuclear-weapon pro-

gram potential, all members would have a betterchance of discerning trade patterns that might helpidentify suspicious end-users or possible diver-sion paths. For the reasons cited above with re-spect to the option of the U.S. Commerce Depart-ment publicly reporting license approvals, othermembers of the NSG may resist revelation oftheirs. 29 Should the United States decide to seeksuch reporting, it may need to test that resistancethrough the leadership both of exhortation and ofits own example. Even the expenditure of consid-erable diplomatic capital with other regime mem-bers may not be enough to bring about this degreeof cooperation.

On a separate track, the IAEA has had discus-sions about maintaining a register of all nuclear-related transfers. The most recent agreement wasfor the purely voluntary reporting only of fissilematerial transfers and specially designed nuclearequipment—not dual-use technologies. IAEA of-ficials reported to OTA that compliance even withthat limited agreement has been uneven.

Expanding the NSG Database IdeaThe reference information in the proposed Nu-clear Suppliers Group database would also in-clude the export guidelines of the MTCR and thecontrol list of the Australia Group. Other than fur-nishing up-to-date details about those regimes, thedatabase as now proposed would play no furtherrole in coordinating the suppliers. Nevertheless,the basic mechanisms of the proposed NSG da-tabase could be extended to the AustraliaGroup and the MTCR. This step would be mostuseful in combination with agreements in thoseregimes to report export denials, as the NSG mem-bers do. Such agreements, however, will not beeasy to obtain. Nevertheless, if the political diffi-culties could be overcome, a single proliferationexport-control database seems technically feasi-ble, since there is a high degree of overlapping

~9As mmxt ah~ve, stm~e fim]s might be fearful that c(mfidtmtial (but still Iegilirnate) market mfomlation might be revealed to competitors” ifall sales were reported. Even if the supplier-group data were not in the pub] ic d(mlain, there would be the possibility (hat participating gover-nments w(mld leak infomlati(m to their {nvn cxwntry’s fim~s.


Australia Group Missile Technology(CW and BW) Control Regime

/ “ A r g e n t i n a \ \

/

IcelandNew Zealand

\(Have

\declaredintention toadhere toguidelines:Brazil

Belgium Luxembourg ChinaaCanada Netherlands IsraelDenmark Norway South Africa)Finland Portugal

Germany SwedenSwitzerland Russia

Hungary United Kingdom (MTCR

United States probable)Romania

\BulgariaCzech Republic

Nuclear Suppliers PolandGroup

a China promised 10 adhere to guidelines in 1991, but has not said it would adhere to revised guidlines of 1993

There iS a considerable overlap among the memberships of the three major nonproliferation exportcontrol groups.


membership among these groups (see figure5-1 ).30 Even if perfect overlap were not achievedamong the three suppliers groups, levels of accessto the system could be differentiated by groupmembership. Alternatively, separate databasescould be set up for each group.

Aside from the supplier groups, there are twoother international groupings whose export con-trol systems would benefit from shared data net-works: the European Union (EU) and the newlyindependent states of the former Soviet Union. In1992 the European Community (EC) Commis-

sion reported to the EC Council on a review of theexport control systems of the member states. Itfound important discrepancies among the states.At the end of August 1992, the Commissiondrafted an export-control guideline for adoptionby the Council. This draft included a proposal for:

. . . a system of information transmission andexchange, to include all orders and transactionsof dual-use items, before actual transfers takeplace. An electronic data network is envisionedto build on the insights and information of na-tional agencies and to inform all licensing agen-

Wscc ~onw~ .s. s~’ctor and Vlrg]nla Form “Preventing Weapons Proliferation: Should the Regimes be Combined’?”’ (Muscatinc, 1A: The

Stanley Foundatwn, 1992).


to build on the insights and information of na-tional agencies and to inform all licensing agen-cies immediately about the refusal of license ap-plications. Thus, a binding exclusion of refusedlicenses, in effect in all EC member states,would be in force.31

It now appears that the EU will not adopt suchmeasures in the near future. Should it ever under-take to do so, U.S. experience in developing itsown networks could position it to cooperate in theestablishment of a European Union network. Thatnetwork might, in turn, become a basis for assis-tance to other states or groups of states in estab-lishing their own systems. For example, in thesummer of 1993, Russia, Ukraine, and Belarusdiscussed creation of an economic union; one pro-posal discussed was a common customs and ex-port control system for the group. (See below forfurther discussion of the export control situationin the former Soviet republics. )

Increase Intelligence SharingWhether by means of a networked database orthrough other means of communication, sharingintelligence data about unscrupulous suppli-ers, buying and financing operations, ques-tionable agents, and suspicious end-users is animportant means by which supplier groupscan coordinate their export controls. Shared in-telligence could, for example, help members ofthe NSG make better informed licensing judg-ments by giving them more information abouthow prospective buyers measure up against thecriteria that they have agreed to take into accountin licensing decisions (see above, box 5-1).

The greatest obstacle to sharing intelligencedata is the risk that revealing what an intelligenceagency knows might also reveal how it found out:that sources and methods will be compromised.Recognizing this problem, the CIA’s Non-Prolif-

eration Center is placing increased emphasis on“actionable” intelligence—information that canbe safely revealed when necessary to moveagainst proliferation activities. Enforcement offi-cials at Commerce’s BXA have begun a prolifera-tion database based on open sources, but purelyfor internal use. In principle, such data could beused to help explain to exporters why licenses arebeing denied, to inform companies about what po-tential customers to avoid, or to alert other coun-tries to possible proliferation risks.

In some situations, national intelligence agen-cies having trusted relationships with one anothermay be able to share secret information. Amongstthe large and diverse sets of nations making up thenonproliferation supplier groups, continuous, di-rect sharing of classified information seems un-likely.32 What seems more feasible is the produc-tion and dissemination of analyses based on opensources. It may also be possible to develop open-source evidence for facts that might originallyhave been indicated or discovered by secretmeans.

All information sharing need not take the formof current intelligence. When the supplier groups(NSG, Australia, MTCR) meet, their gover-nments could take the opportunity to send experi-enced export control officials, not just temporarilyassigned diplomats. These officials could be en-couraged to examine comparable problems, ex-change ideas about methods, and discuss actualcase examples that might hold lessons for theircounterparts.

To increase opportunities for multilateralinformation sharing, one option to consider isto provide government support for non-gov-ernmental, open-source database and analyticprojects. Examples of such projects are the Moni-toring Proliferation Threats Project at the Monte-rey Institute of International Studies and the data-

J IHara]d Mu]]er, “me Exp)fl Con[r(}ls ~~ba[e ]n the ‘New’ European C(mlmunity,” Arms Conlro/ 7bday, March 1993, p. 12.

32~e Unl[ed States ~cP)~ed]y did find ways ofshar-lng lntell l~~nce inft~mla(i(m ahmt Iraq with the United Nations Special Comnlission” on

Iraq, but this could be made a more routine practice. For a discussion of possible nati(mal intelligence contributions [o United Nati(ms activities,see Garret J(mes, “lntell igence Suppwt to United Nati(ms Activities, ” U.S. Am]y War College Study Project, Apr. 15, 1993.

— . . —


base of the Wisconsin Project on Nuclear ArmsControl, in Washington. One means of support forsuch efforts is to contribute grants or award re-search contracts to the private institutions carry-ing on such projects. Another would be to shareinformation with them informally: perhaps givingopinions as to which open sources are more or lessreliable, which analyses are more or less conso-nant with government analyses. One analyst sug-gests:

More “’cross-cultural” communication be-tween the governmental and non-governmentalnon-pro] iteration communities would be bene-ficial. With no access to classified informationbut a suspicious attitude toward bureaucratic as-sessments, non-governmental analysts have thepotential to reach fundamentally incorrect con-clusions. Perhaps the non-governmental com-munity should become more tentative in its con-clusions as it demonstrates greater skepticismabout the reliability of sources. On the otherhand, the governmental community may also betoo quick to dismiss public sources. Assigningan individual in the higher echelons of govcrn-m e n t t o s a n i t i z e c l a s s i f i e d i n f o r m a t i o n f o rpublic release without revealing sources andmethods would facilitate cross-cultural commu-nication between these two communities.33

Whether the information shared multilaterallycomes directly from the U.S. government, orwhether it comes from private U.S. institutions,there is some risk that it will be perceived as a U.S.too] for manipulating international opinion anddecisions to serve unilateral U.S. interests. Thisrisk imposes a need for considerable tact and di-plomacy in the ways in which the United States at-tempts to persuade other nations to act on the in-formation provided. Another way to reduce therisk might be to help create and sponsor interna-tional nongovernmental organizations to monitorand analyze proliferation problems. The goalwould be to minimize the perceived control or in-

fluence of any one national government, with thehope that many governments would both contrib-ute help to and utilize the products of such orga-nizations.

The immediate goal of increased intelligenceand other information sharing among gover-nments would be to enhance their export controls.At the same time, greater public informationabout proliferation activities could help mobilizeinternational support for the whole range of non-proliferation policies surveyed in the first reportof this OTA assessment: not only coercive actionsagainst violators of nonproliferation norms, butinternal and external pressures on governments torenounce weapons of mass destruction and adhereto the nonproliferation regimes.

I Support Development of FSUAdministration of Export Controls

The effectiveness of global export controls will begreat] y weakened unless Russia and the other for-mer Soviet states join the full set of western non-proliferation control regimes: NSG, AustraliaGroup, and MTCR. Some progress has been madein this direction with Russia already in the NSG,vowing to become a de facto member of theMTCR, and promising to adhere to AustraliaGroup guidelines. The other newly independentstates should also be brought into the nonprolifer-ation regimes. These nations also need to developeffective export control systems. The UnitedStates has offered several million in Nunn-Lugarfunds for this purpose to each of the four republicsretaining Soviet nuclear weapons, but has reachedagreement on spending the money only with Bela-rus. Other republics could probably also make useof financial assistance. In addition to funding,U.S. agencies have also been offering technicalassistance in export controls to the former Sovietstates.

~~Mark G, McDonouoh ” “Nu~]ear Non.pro]” iferatlon ~oj~~t, Ct~nfercncc (m Strcngthcnlng the N(m-prolifcrati(m Regime: S~l~Cl~d AnalJ-a ,ses. Fred] n.gs, and Reconlnlendatlons,”” manuscnpt, Carnegie End{~w mcnt for lntcmat](mal Peace. Mar, 18-19, 1992, p. 12. For a (iIscussitm of

the ISSUCS raised by the prt)spect of sharing mtcll igencc inft~m~ati(m w ]th an intcmali(~nal (~rganizat](m, we Garret J(mcs, op. cit., footnote 32.


A report containing views of both U.S. andRussian experts observed that Russia has yet todevelop an effective export control system. Over-all, the need is for:

. . . a competent civil authority with the will andcapabilities to enforce the laws, decrees, opera-tion regulation, licensing procedures, and en-forcement practices recently adopted by theGovernment of the Russian Federation.34

Specifically, the problems include:

. . . 1 ) creating an adequate legislative andexecutive basis for the structure as a whole andeach of its institutional bodies; 2) overcomingthe lack of transparency and openness in the ad-ministrative and other non-classified activitiesof enterprises and scientific institutes; 3) instal-ling an effective licensing system in the RussianFederation regarding its rights in both the inter-nal and international arenas, including protec-tion for intellectual property rights; 4) overcom-ing the present ability of Russian enterprises andinstitutions to conclude contracts with foreignbuyers, including contracts for dual use technol-ogies and armaments, which circumvent nation-al authorities in respect to export authorization,registration, and licensing; 5) instituting cus-toms controls and bringing them up to a suffi-cient level of effectiveness, particularly at bor-ders with the neighboring states of the formerrepublics of the USSR.35

Moreover, not only in Russia, but elsewhere aswell:

. . . the establishment of sovereignty in the newstates of the former USSR is unfortunately beingaccompanied by the weakening of legislative,executive, and judicial powers, a rise in crime,and the formation of organized crime syndicateswhich include civil servants. The problem ofnon-proliferation is also exacerbated by the uni-fication of organized crime structures on an in-ternational level .36

Members of the NAS-RAS group argued thatthe United States and Russia should work to har-monize and refine their export control lists. Theyproposed that Russian and American scientistsand engineers work together to identify choke-points for the unwanted export or internal transfer

37 They suggested that the twoof technologies.countries could establish a bilateral laboratorygroup that would work to identify and agree upondangerous dual-use technologies.38 The twocountries might also:

. . . establish a joint data bank group whichwould establish joint lists of restricted technolo-gies and enterprises or “’projects of concern” towhich certain technologies should not be inter-nally transferred or exported.39

At the Moscow summit in January 1994, Presi-dents Clinton and Yeltsin signed a joint “Memo-randum of Intent” on “Cooperation in the Area ofExport Control,” saying their governments in-tended to cooperate in “any or all” of six areas in-tended to improve nonproliferation export con-

3$u.s. Na[lona] Academy of Sciences and Russian Academy of sciences, “ Dual Use Technologies and Export Administration in the Post

Cold War Era” (Washington, DC: National Academy of Sciences, Apr. 1, 1993), p. 9.

3Slbid., p. 14.361 bid., p. 10.371 bid., p. 17.381 bid., p. 17.391 bid., p. 20.


trols and that they ‘*may” establish expert workinggroups to carry out their intent.40 At this writing, itis too soon to tell whether these actions will be tak-en or whether they will result in concrete improve-ments in the Russian control system.

l Seek Greater Cooperation FromDeveloping Countries

The newly independent states of the former SovietUnion are not the only emerging source of com-modities that could contribute to the spread ofweapons of mass destruction. Newly industrializ-ing countries that are not members of the estab-lished export control groups are also becomingpossible suppliers to proliferant weapon pro-grams.

41 invo]ving such nations in multilateral

export control arrangements could have two bene-fits. First, should they establish reasonably effec-tive export control systems, the new supplierswould be less likely to contribute to proliferation.Second, their very membership in the internation-al groups could undermine assertions that the non-proliferation regimes are discriminatory and in-tended to preserve the economic and militaryadvantages of the more prosperous nations. On theother hand, if the emerging supplier is itself a pro-liferation threat, it might acquire easier access toitems it needed for its own weapon programs,even as it helped control supplies to others.

India in particular—but other nations as well—has long argued that the Nuclear Non-Prolifera-tion Treat y discriminates unfairly against non-nu-clear states. Part of its argument is that until allstates give up nuclear weapons, the other statesshould not be forced to give up the nuclear option.But another part of its argument can be summa-rized as follows:

. . . technology export barriers erected on thegrounds of national security are also aimed atretention of Western industrial supremacy andcontrol of the global technology markets.42

When the United States persuaded Russia tostop the transfer of cryogenic rocket motortechnology to India in 1993, a frequent theme inthe Indian press was that the “real” reason for theU.S. action was to prevent commercial competi-tion from the Indian space program.

The Clinton Administration’s proposedchanges in the Missile Technology Control Re-gime are designed in part to respond to such argu-ments. The President announced in his United Na-tions speech on September 27, 1993:

Now, we will seek to strengthen the prin-ciples of the Missile Technology Control Re-gime by transforming it from an agreement ontechnology transfer among just 23 nations to aset of rules that can command universal adher-ence. 43

%c SIX areas were:A. C{mductlng b]lateral and multilateral discussi[ms at the political and technical level (m matters rel:it]ng u) the enhanccrmmt of expm

c(mtrt)l S} stems;B, C(mductmg bilateral c(msultati(ms at the expert and government levels (m (Migatitms relating I{) mm-use of expwl ctmtrt)llcd items f~w

unapprtwcd purp)ses;

C. C(mctuctmg bilateral consultations (m specific multilateral expwt c(mtr(~l regimes and their implcnwntati(m and (m the technical parame-

ters of the items and technol(~g]es c(nered by them;D. Participating In seminars, ctmfcrences, and other multilateral meetings devotul to c(msidering CX.PM control issues,E. Dlscusslng {)ppmunitics to train pcrs(mnel inwdvcd with expwt c{mtr(~l, the work of licensing and cust(m~s agencies, and

F. J(~mt cff~ms to expand c(}{~pcration in the area of export c(mtr(d.‘.’ Text’ of Memorandum (m Export C(mtrols,’” FBIS-SOV-94010, Jan. 141994, p. 20.

~1 See Wllllanl C. po((er, Cd /n[erna[lona/,vl((/c(ir Tr(ulf ond Iyt)nl]rc)l[terotic)n.. The Challcnqe ol’the Emcr,q/rtR Sqy?ller.y (L~~ingt~m, MA;. . c.Lcxingtfm B(NAs. 1990) and The In(crnalionni MI YVIIC Bazaar: ThCI Nat Suppllcrr ,?’et)tork (B(~uldcr, CO w’estvlw press, 1994).

42 Brahnla Chellaney In The Global l)rf~mion o/ Milirory Te<hnolo<q>.. 7’he f’rot. ccdln,q Y (!/ o }iimL.~h[)p held at (he iJni\crsll> of”k+’[ fconsin,

Madwrr, I)ctember 6-8, 199/ (Madis(m, WI Center for Intcmat](mal C[){)perat](m and Sccurit~ Studlcs, Llntkcrs]ty {~f Wlsc(msin), p. 19.

‘~presdent BIII Cllnttm, “A& Jrcss U) the 48th Scssi(m of the United Natl(ms General Assembly,.’ New York, N}’. Sept. 27, 1993.


A White House fact sheet explained the pro-posed bargain with developing nations seeking toimport space launch vehicle technology:

We will support prudent expansion of theMTCR’S membership to include additionalcountries that subscribe to international nonpro-liferation standards, enforce effective exportcontrols and abandon offensive ballistic missileprograms . . . We will continue to retain a strongpresumption of denial against exports to anycountry of complete space launch vehicles ormajor components . . . For MTCR membercountries, we will not encourage new spacelaunch vehicle programs, which raise questionson both nonproliferation and economic viabilitygrounds.

The United States will, however, considerexports of MTCR-controlled items to MTCRmember countries for peaceful space launchprograms on a case-by-case basis. We will re-view whether additional constraints or safe-guards could reduce the risk of misuse of spacelaunch technology .44

Critics of this new policy stress the risks, argu-ing (as the Administration’s statement acknowl-edged) that space launch technology is in someways analogous to plutonium reprocessingtechnology: it is economically unsound and car-ries inherent proliferation risks. First, the nationsthat do manufacture and launch space launch ve-hicles all lose money doing so.

45 Second, space

launch rocket technology is eminently transfer-able to ballistic missile programs. A nation that iscomplying with nonproliferation norms todaycould change its mind tomorrow, and still be inpossession of missile technology; no plausiblesafeguards are likely to change that potential.

Therefore (from this point of view), in the inter-ests both of fostering the economic welfare of de-veloping nations and of limiting missile prolifera-tion, the transfer of rocket technology should notbe used as an incentive to adhere to nonprolifera-tion regimes (for a supporting example, see box5-2).

A contrasting view is that the Administration’schanges on missile export policy do not go farenough. As one analyst has pointed out, NASA isnot “economically viable,” but the United Statesstill supports its own space launch program forother motives. Countries with fledgling space pro-grams are unlikely to be persuaded that these mo-tives are legitimate for the United States (or Rus-sia, China, France, and Japan) but not forthemselves. Nor will they all accept the conceptthat they must forswear missile programs forthemselves while the existing members of theMTCR are entitled to keep theirs.% Given themodest benefits proposed (“case-by-case” consid-eration) and the major concessions asked for (fulladherence to nonproliferation norms), it is notclear that in practice there will be many takers forthe new Administration Policy on the MTCR.

In its draft for the EAA of 1994, the Clinton Ad-ministration proposed providing for (individualvalidated) license-free exports of controlled itemsto and among members of a multilateral regime. Inaddition, under this draft law, nonmembers couldbe granted adjustments in access to controlleditems depending on their adherence to U.S. exportcontrol policies. This more convenient access (un-der either provision) to dual-use technology itemsmight serve as an incentive for some developing

%ffice of the Press Secretary, The White H(mw, “Fact Sheet: N{mproliferati(m and Expmt C(mtrol P~)licy,’” Sept. 27, 1993.

JsHowever, selllng launch sewices to foreign or domestic commercial firms may help defray the costs of fLIlfilIing other govemnlental

purposes, such as national autommly in space-launch capabilities. In the case of Russia, it could be that space launch services could profit be-cause of the sunk costs in space launch infrastructure and vehicles already produced primarily f(w military purposes.

~See ~tatenlent by L(~ra Lunlpe in “ne Administration’s Non-proliferation” and Exp)rt Control Policy, ’’Arms C~nlrO/ Tw@, vol. 23, N(). 9,

November 1993, pp. 12-13.


The September 20, 1993, launch failure of the Indian Polar Satellite Launch Vehicle (PSLV) can be Interpreted

to support the arguments for using export controls to deny launch vehicle technology to new entrants Although

this rocket was to place an Earth remote sensing satellite into a sun-synchronous polar orbit, It could also be

used as an Intercontinental ballistic missile The Indian Space Research Organization (ISRO) obtained key

technology for the second stage, Iiquid-fueled rocket motor from the French Societe Europeene de Propulsion

In the face of export controls, India developed other key technologies Indigenously—e.g., maraging steel and

solild propellant (HTPB) for the first stage motor An Indian journalist concluded before the launch failure that

It IS these and other Instances of organization foresight which saved the launch vehicle program when the U Sembargoed all sales to the ISRO These very same qualities WiII have to be revived in the ISRO if the launch vehicle

program IS to survive the trials ahead 1

These efforts to work around missile technology export controls apparently have not yet been fully success-

ful On its maiden launch, the PSLV suffered a mishap after separation of the second stage (of four) that resulted

in the rest of the vehicle reaching too low an altitude to reach orbit ISRO officials reportedly concluded that the

next PSLV launch would have to be put back 2 years

ISRO officials had reportedly hoped to sell as many as 9 satellite launches on the vehicle between 1996 and

2000, thus brlnglng$100 milllon in business However since the PSLV development program had already cost

$144 milllon over 12 years, and since ISRO had said that it could produce the launchersatacostof$15 milllon

each, it iS not clear when if ever, the project would have produced profits 2 Now that the program has been set

back another 2 years arguments that the space launch business lsan economic Ioser for developing countries

seem even stronger

1 Gopa! Ral The H/nd~/ (Madras) Sept 11 1993 p 8 JPRS-TND-9035 Nov 10 1993 p 322 For reports on the launch failure and on cost estimates see K S Jayaraman, Launch Failure Dents Indias Space Plans Nature,

vol .365 (Sept 30 1993) p 382 and Tm Furmss PSLV FaJure Delays Ind[an Space Plans F/{ght/nternatlonal, Sept 29 1993 p 23

——

nations to adhere to supplier-regime guide] ines.On the other hand, were these nations so well-be-haved in the first place, license approvals prob-ably would have been fort hcoming an yway.47 Theremoval of IVL requirements would probably bewelcomed by U.S. exporters who feel that currentregulations are too burdensome. The disadvan-tage to removing validated license requirements is

that the United States would lose the opportunityto judge on a case-by-case basis whether the recip-ient country’s own export controls were strong

enough to prevent retransfer of some items.Instead, it would have to arrive at a general judg-ment to that effect.

Bringing new suppliers or transshippers intothe established groups controlling exports is agoal that could contribute to nonproliferation,even if it may be difficult to accomplish in somecases. One analyst has suggested that at the 1995NPT renewal conference, the parties to the treatycould formally acknowledge the obligation of allof them, not just the nuclear weapon states, to re-


frain from assisting other states to manufacture oracquire nuclear weapons. The Conference couldthen endorse specific guidelines for national ex-port control laws and procedures.48

One analyst has proposed additional measuresfor fostering increased cooperation from the de-veloping world:

1. the members of the supplier groups could pro-vide statistics on license approvals and denialsto counter perceptions that export controls aredesigned or function to impede economic de-velopment;

2. supplier groups could meet regularly with de-veloping countries that adhere to nonprolifera-tion norms to explain the reasons for nonprolif-eration export control policies and answercomplaints;

3. more ambitiously, the supplier groups could es-tablish a global forum on international technol-ogy transfers and export restraints, seeking a“North-South” consensus on how proliferationcould be constrained while civil developmentis fostered; and

4. supplier nations could bias their developmentaid in favor of nations that comply with non-proliferation and export control regimes.49

In attempting to better inform developing na-tions about the purposes and effects of export con-trols, the industrialized countries would have totake care to avoid the appearance of simply dictat-ing their own views of the proliferation problemand how to deal with it. As noted earlier in thischapter, some nations perceive economic discrim-ination even when the facts suggest otherwise.Considerable diplomacy may be required to gainan open-minded hearing for factual presentations.

Formally conditioning development aid onnonproliferation compliance could also offend de-veloping nations’ sensitivities. International de-velopment assistance programs might have a dif-ficult time politically in deciding what degreesand kinds of proliferation or nonproliferation be-havior by what nations should lead to larger orsmaller aid allocations.50

The United States, for its part, might have diffi-culty reconciling its other foreign aid objectiveswith the nonproliferation objective. It is one thingto reduce assistance as a sanction for certain pro-liferation behavior; it would be another to reallo-cate aid given to some nation for one purpose (say,supporting Israel and Egypt to bolster Middle Eaststability) to some other nation as a reward forcooperation on nonproliferation.

~~.wls Dunn in Haral~ Mu]]er and LCW is A, Dunn, Nuclear E.x]w-t Controls and Sui)ply Side Re,swaint.~: oplwn.ytw Rejim (Southamp-

ttm, UK: Prograrnnw for prt)rm~tlng Nuclear N(mpn)liferati(m, Study Number Four, (lct(ixr 1993), p. 28.

~gHara]d MU]{~r, ibid., pp. I ~- I ~.

~~or a discussion of attaching ~)]icy ctmdili(ms to foreign assistance, see Nicole Ball, ‘“Levers for Plowshares: Using Aid T(J Encourage

Military Reform,” Arm.! Conrro/ Today, vol. 22, N(). 9, November 1992, pp. 11-17.

Reducingthe Burdenson Industry 6

The current U.S. export control system has come understrong criticism from some U.S. industrial associationsand companies. As noted in chapter 4, in the section onestimating the costs of the system, they complain that

many U.S. export controls both fail to produce any meaningfulresults and place unfair burdens on U.S. exporters. From thepoint of view of the effectiveness of export controls, it is desir-able to have exporting companies see the system as fair andjust, so that they will have every incentive to help make thecontrols effective-for example, by reporting possible illicitbuying attempts. From the point of view of U.S. competitive-ness in international markets, it is desirable to place the leastconstraints consistent with national security on exporting firms.

Some measures for reducing the burdens of the system on ex-porters could be carried out without impairing the effectiveness ofcontrols, and it can be argued that some of those measures wouldeven enhance effectiveness. There is inevitable controversy, how-ever, over whether some burden-reducing measures would helpor hinder the effectiveness of controls in slowing proliferation.

REDUCE THE NUMBERS ANDPURPOSES OF CONTROLSExporting industries have been the strongest advocates of severe-ly reducing the numbers of commodities on the Commerce Con-trol List (CCL). The companies in these industries are under-standably concerned about the burdens the export control systemplaces on them compared to companies from other countries.There is a case to be made that limiting controls to a relativelyfew key technologies could enhance their effectiveness. The

There is inevitable

controversy. over

whether some

burden-reducing

measures would help or

hinder the effectiveness

of controls in slowing

proliferation.

I 55


benefits of a much smaller export control listmight include the following:●

●

●

■

feeling less burdened by the system, exportingcompanies might be more enthusiasticallycooperative in helping to see that the remainingcontrolled items do not fall into the wronghands;with the United States arguing for a muchsmaller range of controlled items and a smallerrange of reasons for controlling them, coopera-tion of other nations in export controls might beeasier to obtain;the range of U.S. controls is broad enough thatother countries sometimes suspect commercialmotives to be behind U.S. attempts to enforcecontrols; that reason for resistance could be re-duced with a smaller list; andgovernment administrative and enforcementefforts might be released from nonproductiveattempts to block exports that the buyers willstill find elsewhere.These arguments are most persuasive when ap-

plied to the items controlled by the CoordinatingCommittee on Multilateral Export Controls(COCOM) industrial list, which is being phasedout (but which may be replaced in some form by asuccessor agreement). Most U.S. nonprolifera-tion controls coincide with those already win-nowed by negotiation in the multilateral non-proliferation export control regimes. Thus,controls over items related to weapons of mass de-struction and missiles are the strongest candidatesfor continuation if controls overall are reduced.

I Foreign AvailabilityExporters have argued that if a commodity isavailable from foreign sources that do not havecomparable export controls, U.S. export controls

are useless, since objectionable users can obtainthe items elsewhere and continue unhindered withtheir weapon programs. ’ Proponents of unilateralexport controls argue that this argument is tanta-mount to condoning selling a gun to a criminaljust because he may have been able to buy it fromsomeone else. Some exporters may feel that theyshould not be denied licenses to sell to such users,on the ground that someone else will anyway.Most, however, would not wish to do businesswith users trying to build weapons of mass de-struction. It is not the loss of these relatively raresales that exporters fear, but rather that the exportlicensing process itself causes them to lose legiti-mate business to foreign competitors at the sametime that it fails to keep the proscribed items outof the hands of proliferants. Industry representa-tives cited as an example of this problem the caseof high-performance computers, which have beencontrolled both because of conventional military-related applications and because of their potentialuse in nuclear weapon and missile programs. TheClinton administration announced in September,1993, that it agreed computers no longer could orshould be controlled at previous levels (see be-low).

In the case of such ● ’national security” controls(as opposed to the “foreign policy” controls,which include items of proliferation concern), theExport Administration Act (EAA) requires thegovernment to remove items from the list wheninvestigation shows that they are readily availablefrom foreign sources. In this context, “availabil-ity” means that it is possible to buy the item inquantities and of quality comparable to that avail-able in the United States.

One proposal for export control reform,then, is to make timely employment of the test

I F[)r ~xamp]e, we Frederick p. Waite and M. Roy Goldberg, “Responsible Export Controls or ‘Nets to Catch the wind’?: The commerce

Department’s New U.S. Controls on Exports of Chemical Precursors, Equipment and Technical Data Intended (o Prevent Development ofChemical and Biological Weapons,” Cah’jhrnia Western International LaHI Journal, vol. 22, 1991 -1992: 193-208.

Chapter 6 Reducing the Burdens on Industry | 57

of foreign availability to all items retained onthe CCL.* The Department of Commerce (DOC)would be required to conduct frequent reviews offoreign availability, without exporters having torequest such reviews formally. The United Stateswould remove unilaterally controlled items fromthe CCL. It might propose removal of items frommultilaterally agreed export control lists if its re-view finds them available from outside the multi-lateral regime. An item might be found to be un-available abroad for one of two reasons. First, theU.S. producer might be the only source of supply,and items that could substitute for the controlleditem could not be purchased elsewhere. Second,all, or nearly all, of the principal suppliers mighthave agreed to control their exports of the item inthe same way. A policy of attempting to controlonly items that were not available from othersources would lead to a shorter list and to fewerlosses of business from U.S. companies to foreigncompetitors.

A policy of decontrolling goods or technologythat are available from other countries withoutcontrols could lead to a vicious circle. Achievingmultilateral controls has usually required leader-ship by one nation, most often the United States.Other countries may be more willing to controlnew items (or exports of currently controlleditems to newly identified end-users) if the UnitedStates demonstrates its own will to do so first.Thus, proposals to limit U.S. export controls tomultilaterally controlled items have included pro-visions for at least temporary impositions of uni-lateral controls to allow attempts to reach multilat-eral consensuses Putting a legislative limit on theterm of unilateral controls does carry a risk: other

nations whom the United States is trying to per-suade to follow suit can just stall negotiations un-til the statutory limit on the U.S. controls runs out.Negotiating multilateral controls might then be-come more difficult in the absence of U.S. leader-ship by example.

Another objection to the strict foreign avail-ability requirement is that in some situations theUnited States, for moral reasons, does not want itscitizens to contribute to another nation’s programto acquire weapons of mass destruction, whetherthat prohibition would significantly delay theweapon program or not. Requiring effectivemultilateral export controls as a condition ofU.S. export controls removes the option of set-ting a unilateral standard for U.S. nationals.

Eliminating, or even putting a short time limiton, unilateral controls could also inhibit the use ofexport controls as an indirect form of sanctionsaimed at controlling weapon proliferation. In thecurrently most publicized example, the UnitedStates is denying high-technology exports to Iranas a way of punishing Iran for its apparent pursuitof weapons of mass destruction and its support ofinternational terrorism. Some of the denied ex-ports-most notably jet transport aircraft thatBoeing wanted to sell the Iranian airline—arenominally controlled as a sanction in punishmentof Iran’s support for international terrorism. ButU.S. Secretary of State Warren Christopher hasexplained the actual intent of the U.S. controls isto make Iran

. . . understand that it cannot have normal com-mercial relations and acquire dual-use technolo-gies on the one hand, while trying to developweapons of mass destruction on the other.4

z Congressional testimony and a draft revision of the Export Administration Act by the National Association of Manufacturers stress thisIdea. See E.~pwt Control Rcfi)rm.. A Key to U.S. E.xp{v-( ,7WWYY; Poliqv Rccwnnwndatwns ( Washington, DC: National Association of Manufac-turers, June 1993).

~lbid.

Jwamen Christopher, al a press c(mference in Luxemb(mrg, June 9, 1993, quoted by Elaine Sciolino, “U.S. Asks Europe to Ban Arrns-

Linked Sales U) Iran,” NCM YorL Times, June 10, 1993, p. A-5.


Other countries have shown little inclination togo along with this policy, and the aircraft saleseems likely to go eventually to the European Air-bus Industrie consortium.5

Even when the focus is on control of items thatcould be used for weapons of mass destruction,there is a further disadvantage to a strict require-ment that the items be under rigorous multilateralcontrol. There is an inherent fuzziness in theworkings of export controls; as noted above, theireffectiveness is subject to a wide range of vari-ables. Even when it is not possible to achieve100% agreement and compliance on multilateralcontrols among all possible suppliers, partiallyeffective controls may still be better than noneat all, depending on the financial and technicalresources of the buyer and the state of progressof his weapon program. Therefore, although it isreasonable to have a strong presumption againstunilateral controls, there may be instances wherecontrols that do not have universal support canstill be useful. Decisions for complete decontrolshould be informed by the best possible analysisand intelligence data about current countries ofproliferation concern.

It maybe possible to persuade key suppliers towithhold particular exports in special instances.But it will be harder for the U.S. government topersuade foreign governments to go along in thoseinstances unless it has a legal and regulatory basisfor imposing the same restraints on its own ex-porters, as well as a consistent policy of denyingexports in comparable situations. How long anygiven control is worth pursuing before being giv-en up as a lost cause is hard to specify in advance.An alternative to a fixed (say, 6 month) term for allunilateral, or less than unanimously multilateral,controls would be to establish an explicit processof accountability by officials entrusted with judg-ing just how long an effort makes sense. Such aprocess might, for example, include a periodic as-

sessment of foreign availability for all controlleditems, coupled with an explicit justification toCongress of the rationale behind continued con-trols for goods found to be available outside theUnited States in comparable quantity and quality.

I AlternativesBesides applying a strict foreign availabilitycriterion, another way to reduce the size of theexport control list is to narrow the scope of itspurposes. After the initial reforms of COCOMcontrols with the end of the Cold War, the DOCOffice of Export Licensing went from handlingover 100,000-125,000 export license applicationsa year to about 24,000 in 1992 and 25,000 in 1993.With the end of COCOM and the further relax-ation of controls on computers and telecommu-nications technologies in March 1994, the DOCestimated that license applications would declineby nearly half again.6 Many of the remaining li-cense applications concern items controlled forother purposes than the nonproliferation of weap-ons of mass destruction. Most of the remainingCOCOM or “national security” items relate topossible conventional military applications. TheCOCOM lists were designed primarily to slowSoviet progress in a broad range of militarytechnologies. The fact that they might also slowthe development of the Soviet civilian economywas seen as, if anything, an additional national se-curity benefit of the regime. COCOM’S originalpurposes became largely (though perhaps not en-tirely) obsolete with the breakup of the SovietUnion.

But a new set of goals for controls over dual-use technologies related to conventional weaponshas not yet emerged. Late in 1993, COCOM mem-bers agreed to abolish the organization in thespring of 1994, but to replace it with a successorregime. At this writing, the goals and procedures

su s ~eexp)fi ~ontro]s” on Cefialn (,l.S..supp]led components of Airbus planes may prevent such sales in the shoti run, but substitution of. .European components seems likely in the longer run.

~ornas L. Friedman, “U.S. Ending Curbs (m High-Tech Gear to Cold War Foes,” New York fime~, Mar. 31, 1994, p. D5.

Chapter 6 Reducing the Burdens Industry | 59

of that successor regime remain unclear. Somehave proposed that the United States initiate anexplicit new nonproliferation regime aimed atlimiting the spread of advanced conventionalweapon technologies. Such a policy, aimed atkeeping particular types of weapons out of reachof many nations, would require a different exportcontrol strategy than one directed at restrainingthe technical development of a single large mili-tary-industrial complex. In the absence of clear-cut opposing blocks of allies, there is bound to beless consensus about who should be the targets ofsuch a strategy. It is therefore likely to be more dif-ficult to sell the strategy multilaterally than it wasto persuade states to participate in the originalCOCOM regime.

A third way to reduce the size of export con-trol lists would be to partially substitute re-porting requirements for licensing require-ments as a nonproliferation tool. That is, thegovernment could require firms to report, but notseek a license for, the export of any items from apublished list of goods and technologies. This listwould be compiled from technical analyses of theoverall needs of programs for weapons of massdestruction, not just the most critical items. Theobjective would be to discover constellations ofimports that might serve as indicators of weaponprograms or clandestine acquisition networks. Al-though goods that might contribute to prolifera-tion would still be shipped under this approach,national intelligence organizations or multilateralnonproliferation organizations could then utilizethis information to take action against specificproliferant programs.

Such an export reporting regime would clearlybe most productive if it were multilateral: prolif-erants seeking to conceal their buying patternswould have less opportunity to find alternativesources. The current multilateral export controlregimes (Nuclear Suppliers Group [NSG], Aus-tralia Group, Missile Technology Control Regime

[MTCR], and the COCOM successor) would pro-vide logical frameworks in which to place exportreporting agreements. However, even if theUnited States, one of the world’s larger exporters,were to establish a reporting list unilaterally, thatwould probably significantly assist proliferationanalysts.

An export reporting list would probably belarger than the current export control lists: an itemwould be subject to reporting not just if it couldmake a significant contribution to a weapon pro-gram, but also if it could serve as an indication of aweapon program. Although the numbers ofmanufacturers and transactions would be largerthan those now affected by export controls alone,the burdens would be lessened: fewer exportswould be subject to complex regulations and li-censing delays. On the other hand, as noted earlier,exporters may resist revelation of their approvedlicenses because of fears of revealing proprietarydata of use to competitors.

ELIMINATE THE “KNOWS, IS INFORMED,OR HAS REASON TO KNOW” TESTSThe Bush administration’s Enhanced Prolifera-tion Control Initiative and certain legislation ledto Export Administration Regulations requiringIndividual Validated Licenses (IVLS) for almostany items that the exporter “knows “ might beused in any way in a chemical, biological, or mis-sile weapon program. 7 Late in 1993, the Com-merce Department issued further guidance speci-fying that a license is required if the exporterknows or is informed that an item will be directly

8 for nuclear weaponemployed in such a program.programs, the rule is stronger: a license is requiredfor any item that the exporter “knows or has rea-son to know’” will be used in such a program. In-dustry representatives, at least before the Decem-ber 1993 clarifications, argued that the effect ofthis policy is to require virtually all exporters to

7ne only ~)thcr ~ountnes ~l(h a ‘.~now,le~ge MI” rcgar~]ess of the nature of the conmlodity” are Gcmlmy and Japan.

858 Federa/ Regls/er 68029-6803 I (~c. 23, 1‘3).


establish costly programs to find out whether theircustomers are involved in a proscribed activity. Inthis way, they say, companies are forced to per-form intelligence services for the government.Moreover, the items exported, if available anywayfrom uncontrolled suppliers, will not actually af-fect the outcome of proliferant programs. Mean-while, government licensing and enforcement ef-forts go to monitoring exports to impose unilateralcontrols that do not really make a difference.

Exporting firms opposing this policy have alsoraised three other objections. First, although hon-est exporters will be exposed to liability, criminalfirms will simply not apply for licenses. Second,many honest exporters are, nevertheless, notaware of the sweeping nature of the “know” rule,and therefore simply do not apply for licenses.This fact puts those firms who do apply for li-censes at a competitive disadvantage compared tothose who do not. Third, with respect to the ‘“is in-formed” part of the rule, firms have also com-plained that the government has informed onlysome exporters about bad customers, foreclosingthat business for them while leaving other export-ers free to trade and profit in ignorance with thesame customers. Commerce Department officialshave acknowledged that sometimes firms havebeen informed only selectively about risky cus-tomers; they say they are going to improve that sit-uation.

| Advantages of an All-Inclusive ListIn its draft revision of the EAA, the NationalAssociation of Manufacturers (NAM) proposedbarring the “knows or is informed” rule through arequirement that the United States consolidate itsdual-use or “commercial” export controls into asingle list which full y enumerates all the productsfor which an export license is required and all thecountries and specific end-users as well. Thiswould greatly simplify the exporting companies’

job in deciding whether a license application wasnecessary and whether it was likely to be ap-proved.

Such a published list might also help improveinternational export control coordination. Manycountries lack the information and intelligence re-sources of the United States. One way of sharinginformation about potential suppliers and pro-liferants would be to publish the U.S. lists oftarget programs. Even in the absence of formalexport control coordination mechanisms, the U.S.proscription list could have useful influence. For-eign governments and companies would be in-formed that the United States considered certainfirms, countries, and end-users to be proliferationrisks. The NAM draft bill, however, carries thecoordination a step further: the United Stateswould not maintain commodities or users on itsown list unless it could gain multilateral agree-ment among all the significant suppliers to imposeequivalent controls, and to do so as effectively asthe United States. Under the requirement that alllists be multilateral, publishing the list would benot only beneficial, but essential.9

Elements of this proposal exist in the current re-gimes. The NSG, the Australia Group, and theMTCR all center on agreed, published lists ofcommodities. On the other hand, the regimes donot require the members to agree in advance onwho all the controlled countries and end-usersmay be. Instead, they provide agreed criteria fordeciding whether an export should go forward.

| Drawbacks of an All-Inclusive ListThe United States export regulations concerningmissile-related technologies do identify someend-user programs to which exports are not per-mitted. The United States also publishes a Tableof Denial Orders listing entities barred from re-ceiving licenses to export controlled items. Nev-ertheless, publishing the names of all suspect end-

91n ~&jl[lOn, the NAM bll] Prop)ses [hat n. I icenses be required for trade among adherents to the nlullilateral agreements, while a license

w(mld always be required for export to a m)n-member. The Administration draft EAA proposes the option for license-free zones, but does notrequire them.

Chapter 6 Reducing the Burdens on industry | 61

users could have drawbacks. Using informationbased on clandestine sources or methods of datacollection risks tipping off the observed parties sothat they can reduce or eliminate their vulnerabili-ties to those methods. Moreover, merely identify-ing front companies or illicit transshippers as sus-pect may lead them to change names and locationsor go out of business and reestablish themselvesin another form. Such actions could interfere withongoing investigations, or prevent break-up orprosecution of illegitimate supply networks.Sharing suspicions about prospective buyers alsorisks the embarrassment, and possibly the injus-tice, of dissemination of information that turns outto be incorrect. These risks (of compromising in-telligence and of releasing unprovable suspi-cions), then, must be weighed against the benefitsof giving exporters better information about pro-spective customers. Since these risks are likely tovary with each case, it can be argued that the gov-ernment should have some discretion in publish-ing its concerns about buyers.

Another drawback to publishing complete listsof proscribed firms and countries is that at leastsome are likely to consider their names to havebeen placed there unfairly. Firms or governmentsmay demand either that proof (which might havebeen based on classified intelligence sources) berevealed or that they be removed from the list. Anunsatisfactory response by the U.S. Governmentmight lead to unnecessarily y strained relations withthe objecting foreign governments. Questionsmight also be raised domestically y or internationa-lly about why some target countries are namedwhile others that should be are not.

On the other hand, when a license is denied, thenominal consignee or end-user implicitly receivesinformation that he is “on the 1ist,” whether the listis published or not. (However, if the end-user is ina country with proscribed programs, and the de-nial is justified on that ground, possibly the partic-ular consignee or end-user may not infer that it issuspect and on the proscribed list.)

Transshipper and end-use data available to ex-port control officials may change rapidly, puttinga premium on flexibility and last-minute changesin 1icensing decisions. The NAM draft bill permits‘“emergency” unilateral U.S. controls, providedthat the list is published. It does not, however, ap-pear to allow for any discretion by licensing offi-cials based on last-minute or classified informa-tion.

I Arguments for the “Know” RuleDefenders of the “knows or has reason to know”’rules argue that exporters who may be trading witha proliferant end-user find it too easy to look theother way, or to fail to report what they know, aslong as their own particular export is not on a spe-cific control list. Suppose, for example, thatanother nuclear proliferant chose to follow the ex-ample of Iraq and build calutrons to enrich ura-nium. When a military research establishmentbought parts suitable for use in calutrons, thatmight be an indicator of a nuclear weapon pro-gram; the supplier might realize that, but not feelobligated to inform its own government. The gov-ernment might feel, however, that a) the suppliershould not be aiding a nuclear weapon program(whatever his competitors might do) and b) that itshould report its knowledge of the existence ofsuch a program and of the possibility that calu-trons might be under construction.

Supporters of the “know” rule or (in the case ofnuclear-related items) the “reason to know” rulealso argue that in reality U.S. exporting firms donot have to worry that they will be subjected to ex-traordinary demands to probe deeply into the char-acter of end-users of relatively innocuous prod-ucts. They point out that the stronger form of therule (“has reason to know”) has existed for sometime for nuclear exports and in other legal areas.The judicial system has not generally permittedunreasonable interpretations of what constitutes a‘*reason to know. ” 10 In practice, no firms appear to

I ~SCC Scn. John G]Cnn, “omnibus NLJC]CM PUJ] ifcrat](m c{)ntrx)l Act of ] ~!)~. A section-by-section” ~’scrlpllon.’” ~’(Jn,~ref$fon~J/ ~cf”or~

May 27, 1993), Daily cd.. S6773.


In December 1993, the Department of Commerce provided further guidance to exporters on their responsi -

bilities under the “know” and “reason to know” rules governing applications for exports items not on the Com-

merce Control List that might be going to activities involving design, development, production, stockpiling, or

use of missiles or weapons of mass destruction. Here are excerpts from this” ‘Know Your Customer’ Guidance”

(A.) Decide whether there are “red flags. ” Take into account any abonormal circumstances in a transaction that

indicate that the export may be destined for an Inappropriate end-user or destination Commerce has developedIists of such red flags that are not all-inclusive but are intended to illustrate the types of circumstances that should

cause reasonable suspicion that a transaction WiII violate the EAR [Export Administration Regulations].

(B.) //there are "red/flags, “require. absent ’’redflags"... there is no affirmative duty upon exporters to Inquire,

verify or otherwise “go behind” the customer’s representations. However, when “red flags” are raised information

that comes to your firm, you have a duty to check out the suspicious circumstances and inquire

(C.) Do not se//-b/red. Do not cut off the flow of information that comes to your firm in the normal course of busi-

ness An affirmative policy of steps to avoid “bad reformation” would not insulate a company from liabilty

Employees need to know how to handle “red flags “ Knowledge possessed by an employee of a company can be

inmputed to a firm so as to make it Iiable for a violation This makes it Important for firms to establish clear policies and

effective compliance procedures to ensure that such knowledge about transactions can be evaluated by responsible

senior officials

(D.) Reevaluate all the informaftion after the inquiry... If [the “redflags’’ can be explained or justifled] you may

proceed with the transaction [Otherwise]... you run the risk of having had “knowledge” that would make your action

a violation of the EAR

(E.) Refrain from the transaction disclose the information to BXA[Bureau of Export Administration] and wait...

Industry has an important role to play in preventing exports and reexports contrary to the national security and foreign

policy interests of the United States BXA WiII continue to work in partnership with industry to make this front line of

defense effective, while minimizing the regulatory burden on exporters

As can be seen, the regulations as explained by Commerce do not require firms to initate intelligence opera-

tions. At the same time, they do seem to require a thorough understanding of what “red flags” to look for and a

systematic program of company compliance policies and procedures. Although companies exporting toilet

paper or Iight bulbs would not have to be concerned about their products being directly employed” in prolifera-

tion activites, other companies might have to make intelligent guesses about what combinations of their prod-

ucts and customer red flags should be reported to Commerce.

SOURCE 58 Federal Register 68029-68031 (Dee 23, 1993)

have been penalized for having failed to apply for which risks helping a weapon program, but whicha license for something that they are alleged tohave known would be used in a banned project. Inits December 1993 guidance to exporters, theDOC spelled out in greater detail what is expectedof exporters under the “know” rules. See box 6-1for excerpts from that guidance.

There are arguments in favor of maintaining a“know” rule. First, it gives the government a safe-ty net by allowing the application of export con-trols when it learns about a pending transaction

is not explicitly covered by the current CommerceControl List. Second, it improves the gover-nment’s ability to obtain information about pos-sible weapons proliferation programs by requir-ing firms who come into such information, or whoencounter a “red flag” (the term in Commerce De-partment guidance) that should arouse suspicion,to pass the information along to the government.Third, many companies would themselves prefernot to deal with end-users developing weapons of

Chapter 6 Reducing the Burdens on lndustry | 63

mass destruction, whether their products are criti-cal to those programs or not. Procedures for thegovernment to inform them of the character oftheir buyers may well save them from public em-barrassment later on.

A weakened alternative to the “knows or is in-formed” rule would be a simpler ‘*is informed”rule. Today Germany has a “knows or is in-formed” rule applying to all its dual-use technolo-gy exports, not just those for weapons of mass de-struction. In negotiation with other EuropeanUnion (EU) partners, however, Germany has ap-parently indicated a willingness to settle for the “isinformed” part of the rule for EU regulations, andfor that to apply only to goods destined for pro-grams to produce weapons of mass destructionand missiles (i.e., to hold exporters responsible forapplying for licenses for unlisted goods only whenthe government informs them that they may beutilized in such a program). ’ If the United Stateswere to establish this rule, then at least the gover-nment would retain the legal ability to stop riskytransactions about which it had obtained intelli-gence, even if it could not expect companies to re-port the “red flags.”

Another alternative to subjecting the export ofall commodities to the “knows or is informed”rule would be for the government to generate aseparate control list of products or technologiesthat, although not listed as requiring export li-censes, could be significantly useful in proliferant

programs. (A variation on this idea is presentedabove: there, an expanded list would be subjectonly to reporting requirements, not to licensing.)The exporting companies would then be responsi-ble only for knowing or having reason to knowwhether recipients of those particular items wereengaged in illicit activities. The firms, if in doubt,could ask the government for advisory opinionson prospective buyers. The government couldalso make the companies’ job easier by publishingthose advisory opinions about particular end usersso that other firms could be forewarned. The gov-ernment could further supplement its publishedlists by indirectly assisting private organizationsin developing lists of suspect end users from pub-lic sources.

END UNILATERAL REEXPORTCONTROLS ON EXPORTS TOCOOPERATING COUNTRIESThe United States may require, as a condition ofgranting an export license, that the receiving partyguarantee that it will not reexport the controlleditem to a third country. In the past, some Euro-peans have resented U.S. imposition of reexportcontrols as attempts at extraterritorial enforce-ment of U.S. laws. 2 U.S. exporters have arguedthat when foreign competitors do not require suchreexport assurances, they have a better chance ofmaking sales. If the country of the first user is en-

I I See H. Mul]cr et. a]., Fr<)nl Bla(k .Yhcep I() While Angel? The Ne\t German L-.xpurl Control Policy, PRIF Reports No. 32 (Frankfufl an)

Main, Gem~any: Peace Research Institute Frankfurt, January 1994), p. 54.121n 19~ ~ Na(lOnaI Acaden]y of Sciences study panel de]egatlon repofled aflcr a European fact-finding “ission:

Through(wt Europe there was a str(mg adverse reacti(m to U.S. expmt c(mtrt)l ~)licy, in particular its extraterrittmial aspects. TheEuropeans have major problems with U.S. c(mtrf)ls (m the reexp)rt by any c(mntry of U.S.-origin items. Nearly all the Europeans withwh(m~ the delegati(m met th(mght their country was doing an adequate job of maintaining t~f a d(mwstic ex~mt c(mtrt)l regime. Theyargued, therefore, that U.S. reexport ctmtro]s (m COCOM items were t-x~th unnecessary and an unneeded intrusi(m. In a sense, suchc(mtrols were seen as a threat to nati(mal sovereignty and as driving a wedge between the United States and Eur(}pe.

Panel (m the Future Design and Implementa[i(m of U.S. Nati[mal Security Expwt C(mtn)ls, Finding Common Ground: U.S. E.;porr Confro/.~~n a Chan~ed G/oba/ Eni)ironrnenl (Wash ingt(m, DC: Nati(mal Academy Press, 199 I ), p. 268. See also Jan Htkkema. ‘The European Perspec-tive (m Proliferati(m Expwt C(mtro]s, “’ in Kathleen Bailey and R(hcrt Rudney, eds., Pro/iferafion and Expwf Con[rols (Lanharn, MD: Un Ivers i-ty Press of America, 1993).

On the other hand, J. Da\id Richardson, Sir~ng Up U.S. E.rpor/ 1)~.r~ncenfi~es (Washingt(m: Institute for intemati(mal Ec(momics, 1993),found no statistical evidence that U.S. exports to COCOM partners fell below what (me w(mld have expected with(mt reexpwt controls. in addi -ti(m, DOC officials argued to OTA in late 1993 that, alth(mgh U.S. reexpmt c(mtrols may have led to tensi(ms with COCOM partners in the past,more permissive reexport provisions in the Exp(wt Administrati(m Regulati(ms had since largely addressed the partners” c(mcems.


forcing export controls equivalent to those of theUnited States, then it should not be necessary forthe United States to demand that it be given theright to judge further exports. The problem isgreatest when other countries have not agreed tothe same rules as the United States (for example,in banning the sales of commercial aircraft toIran); or when they have agreed to the same con-trols but are unable or unwilling to enforce themeffectively.

For nonproliferation controls, the problemdoes not appear to be as great. The NSG mem-bers, for example, have agreed that they will all re-quire reexport licenses for the nuclear-relateddual-use items that they export. This could beanother issue, therefore, that is best separatedfrom negotiations over how to revamp COCOMcontrols.

STREAMLINE THE APPLICATIONPROCESSIndustry representatives have complained that thesometimes lengthy decision process for U.S. ex-port controls has placed them at an unfair competi-tive disadvantage with respect to foreign suppli-ers. Although average license processing timesare short, some license decisions are delayed bythe interagency reviews conducted to assure thatsome applications receive the most thorough scru-tiny from all the relevant experts and agencystandpoints. Commerce officials point out that al-though the changes in COCOM requirementshave reduced the annual number of license ap-plications from around 125,000 to around 25,000,the remaining 25,000 are the most difficult to ana-lyze. Defense Department officials argue that con-siderable progress has already been made in short-ening license review times. 3

The Administration EAA draft proposes assur-ing that nearly all license applications would beeither resolved or referred to the President within90 days of filing with the DOC. If no referral toother agencies were required, the license would beapproved, or the applicant notified of DOC’S in-tent to deny it, within 9 days. If the applicationwere referred to other agencies, they would haveto recommend approval or denial within 30 days;if they should fail to act, they would be deemed tohave no objection to the export. If the agencies in-volved disagreed, an interagency committeewould review the case and its chairman wouldmake a recommendation to the Secretary of Com-merce. If one or more agencies objected to thatrecommendation, they could appeal it to a higherlevel interagency process which would either re-solve the dispute or refer it to the President—again, all within the 90-day period that began withDOC’S receipt of the application.

There seems to be no reason why, with suffi-cient resources, current license decision deadlinescould not be shortened to the times proposed in theAdministration bill, or even less, without dimin-ishing the quality of analysis and review that thelicense applications receive. This might be ac-complished by:

■ increasing the personnel needed to process li-censes;

■ streamlining interagency review processes,perhaps by detailing expert personnel to a cen-tral review office where their full-time workwould be 1icense review; or

■ developing the kinds of computer network re-sources described earlier in this report.

These measures would, however, cost addi-tional funds that the executive branch has not re-

I J1n ca]endm year 1993 the average processing time for licenses not referred to other agencies was 10 days; the avemge for referred licenses-,was 49 days; the average for all licenses was 3 I days. The DOC Inspector General reported in 1993 that from Jan. 1 to Sep. 30, 1992, 9,004licenses not referred to other agencies t(x)k an average of 9 days to process; 8,695 others, referred to other agencies, took an average of 50 days.See OffIces of Inspector General at the U.S. Departments of Commerce, Defense, Energy, and State, “The Federal Government’s Export Licens-ing Processes for Munitions and Dual-Use Commodities: Special Interagency Review,” September 1993, p. A-5.

Chapter 6 Reducing the Burdens on Industry | 65

cently been willing to allocate to export controlmanagement.

ANALYZE AND PUBLISH THEECONOMIC COSTSSome U.S. exporters have argued that the gover-nment imposes export controls without adequateconsideration of the costs they will impose onU.S. industries. They have proposed, therefore,that assessment of the costs of controls should bemade an integral part of the export control proc-ess. One analyst suggests that the new Export Ad-ministration Act:

should require timely annual reports on the. . . .quantitative effects of US export controls on USexport competitiveness . . . Such reports shouldinclude sectoral and product detail, and shouldalso attempt to size up effects of export controlson US direct investment and alliances abroadand on foreign direct investment and alliances inthe United States. 14

This analyst suggested that the statistical tech-niques he used to study the question of national se-curity controls on exports to Communist countriescould be applied in such reports. They probablycan be used, but assessing the economic effects ofparticular export control measures would requiremore specific and detailed data than the currentexport control data management system yields.First, analysts would need to be able to breakdown license applications into their individualproduct components and assess the values of eachtype of component affected. (Under the currentsystem, the values of exports affected can only bereported by the total value of the items falling un-

der various Export Control Classification Num-bers, not by the exact descriptions of the items orby the reasons for which they are each controlled).

It would also be desirable to develop a means ofcomparing the types of products controlled withthe categories of products for which the Bureau ofthe Census collects export data. Second, analystswould need some means of assessing the amountsof business forgone because exporters were de-terred by the licensing process from even attempt-ing to make some sales, because the licensingprocess deterred buyers from carrying through or-ders, or because buyers went first to suppliers inother countries with less burdensome controls.Estimates on forgone sales would depend heavilyon exporting firms’ perceptions and judgments;some means would have to be found of compen-sating for possible biases in their perspectives. ] 5

Analysts making economic impact assess-ments of national security (COCOM) export con-trols would also have to conduct surveys of busi-nesses that maintain internal control mechanismsto qualify for distribution licenses (which permitthem to avoid applying for IVLS). The report writ-ers would need information on the costs of main-taining such internal mechanisms and estimates ofthe competitive disadvantages or advantages theymay produce. For nonproliferation controls,though, the costs of qualifying for distribution li-censes do not apply, since such licenses are rarelygranted for those items.

Insofar as export controls help stem prolifera-tion (or achieve other objectives), the costs of go-ing without certain export controls should also begiven weight in assessing the net benefits and

I +j~~ J, David Richardson, .’&~monlic Costs of US Exp(Jrt Cc)ntrOk,” Statement bef(~rc the Subct}rnmi(tee (m Ec(momic Policy, Trade, andEn\ ]r(mment, C(mmllttec (m Foreign Affairs, U.S. H(msc (}f Rcprescntat]\cs, NtJv. 18, 1993 p. 12. A similar pro~~sal for fornlal evaluation ofthe costs {Jf ctmtrt)ls IS found in BcnJanlln H. Flowe, Jr., .’Testln](my bef(m the Subcommittee on Ec(momic Policy, Trade, and Envir(mrnent of

the H(mse C[mm~ittec (m Foreign Affairs,” June 9, 1993, pp. 8-9.

1‘An altcma[l} c to this dircc[ ~n)ptrica] approach wimld be to use the n]eth(~d applied by Richardson, Si;irr~ Up U.S. E.~p~rt Disin~’en[i~’eS,

op. cit. That method in~ olved a) estimating the Iek cl of twerall expwts (t)r, at best, expwts categorized b) the broad Standard International TradeClasslficati(m system) that the United States sh(mld expect to send to other countries depending (m their ]nc(me, p~pulation, and geographicaldistance: and b) estlnlating the sh(mfall fr(m those levels of e~pmts to c(mntries subject I(} c(mtrols. Whatever else the advantages or disadvan-tages of this n~e(hod, i( w III be difficult [() appl~ specifica]]y to n(mpr(~]iferation controls” until gl(~hal trade statistics become available for thespecific :(Mds c(mtr{)lled.


costs of those controls. That is, the potential costsof proliferation taking place should be weighed in.This kind of assessment, though, as noted in thefirst section of this report, is an even more difficulttask. The issue is not merely what the costs ofproliferation would be, but what the probabili-ty of hypothesized proliferation events wouldbe with and without the controls in question.

Some argue further that, at least in the case ofnuclear nonproliferation controls, the national ob-ligation under the Nuclear Non-ProliferationTreaty (NPT) to refrain from helping other nationsacquire nuclear weapons outweighs any likelyeconomic costs of nuclear-related dual-use exportcontrols; therefore, those costs should not be animportant consideration in whether the controlsare maintained or not. This interpretation, how-

ever, has not been subscribed to either by U.S. ad-ministrations or by other NPT members.

On the other hand, a benefit for nonprolifera-tion efforts may result from better U.S. and in-ternational data collection on the economic effectsof some kinds of export controls. Better informa-tion about the actual patterns of trade in prolifera-tion-relevant commodities could lead to a betterunderstanding of the consumption patterns andsupply networks of potential proliferants.

The Clinton administration’s draft EAA statesas U.S. policy:

. . . to ensure that U.S. economic interests play akey role in decisions on export controls and totake immediate action to increase the rigor ofeconomic analysis and data available in the de-cision-making process.

——- ——

Appendix A:Estimating

the Economiccosts of

Export Controls

Discussing Clinton Administration changes in export con-trol policy for computers, then-Deputy Secretary of De-fense William Perry said that the economic burden to ex-porters imposed by controls on computers was ". . . a

significant factor, but I do not know how to quantify it.” 1 This ap-pendix illustrates the difficulties in trying to assign economiccosts of nonproliferation export controls in the U.S. machine toolindustry. First, however, is a discussion of the general difficultiesof finding meaningful data.

DATAThe Department of Commerce (DOC) computer system for man-aging export control application reviews began as a means of sim-ply tracking the status of applications. A weakness of the system

A

is that it is not designed to yield certain kinds of aggregate datathat would help assess the economic impact of controls. The basicunit of record keeping is the license application. After determin-ing whether a given product requires an export 1icense, a companymay need to apply for an Individual Validated License (IVL) toexport the good to a specific buyer. However, a single license ap-plication may cover multiples of the same article, or it may coverseveral types of article, each with its own Export Control Classifi-cation Number. It may also include items that, if they were not tobe shipped with a controlled product, would not require a 1icense.(The Department maintains a “Commerce Control List” that

Then-Deputy Secretary of

Defense William Perry

said that the economic

burden imposed to

exporters by controls on

computers was “. . a

significant factor but /

do not know how to

quantify it. ”

1 wl]l ianl J. pew, trans~rlpt of Breakfas[ with Reptmers, (kt. 15, 1993 (~’~nue not

state(l).167


specifies the kinds of goods and technology thatare subject to export controls. Although somecontrol numbers may be assigned each to a single,narrowly defined product, others may cover abroad range, and may contain either very generaldescriptions or large sublists of commodities.)

The result of this system is that the Commercedatabase can be searched either for numbers of li-cense applications or for aggregate values of pro-posed shipments in licenses containing specificcontrol numbers. But. short of individually ex-amining each license application, it is not possibleto determine the values of specific kinds of ex-ports when several kinds are included in single li-censes. In addition, licenses frequently aregranted on the basis of 2-year forecasts by the ap-plicants. The DOC has no way of knowing wheth-er the licensed transactions actually take place.Given the shrinking development periods and life-cycles of high-technology goods, 2-year licensesmay never be fully utilized.

Complicating matters is the fact that the ExportControl Classification Numbers bear no relationto the ways in which other trade statistics are kept(e.g., the Bureau of the Census’ export and importrecord system2). Thus, it becomes difficult to de-termine the actual portion of a particular industrialsector that is affected by the requirement to applyfor an export license.

Even if such numbers could be determined,however, they do not tell the story of sales notmade either because the buyers chose to shop innations with less cumbersome export restrictionsor because potential sellers chose not to bear thecosts they perceive to be imposed by the system.

CASE STUDY: MACHINE TOOLSMachine tools cut and form metals or other hardmaterials with varying degrees of precision.Sometimes they are used directly in manufactur-ing, and sometimes they are used to make the ma-chines that produce other articles. They are essen-tial to civilian industry, but they have a range ofmilitary industrial applications as well. They areuseful for manufacturing many types of conven-tional weapons and vehicles. They are also usefulfor building nuclear weapons, for manufacturinghigh-speed centrifuges that can enrich uranium togo into nuclear weapons, and for making precisionmissile parts. Numerically controlled (usuallymeaning computer-controlled) machine toolsmeeting certain performance specifications are onthe Commerce Control List (CCL) for both nu-clear and missile nonproliferation reasons. Re-lated computer hardware and software are also onthe list. In addition, some tools not on the list fornonproliferation reasons are there for national se-curity, i.e., Coordinating Committee on Multilat-eral Export Controls (COCOM), reasons.

The U.S. machine tool industry declined dra-matically between the 1970’s and the 1980’s: inconstant 1982 dollars, shipments declined from ahigh of $5.6 billion in 1980 to $2.2 billion in 1992.Thus, if the entire industry were considered to beone corporation, its sales in 1992 would haveranked only 159th in the Fortune 500 list.3 In ex-ports:

■ Total U.S. machine tool exports in 1992 wereslightly over $1 billion. The industry thus de-pended on exports for about 34 percent of its

2The Census Bureau (since 1989) gathers trade statistics using the Hammnized System (HS), which many countries use to facil itate c(mlpar-ison of international trade by commodity for various countries. The classificati(ms of products in the HS bear m) relati(mship either to the ExportC(mtroI Classificati(m Numbers or to the product descriptions on the Commerce Control List.

“’The Fortune 500 Largest U.S. Industrial Corporations,’’” For/une, Apr. 19, 1993, p. 190; with shipments estimated at $3.02 billion in 1992dollars, if the industry were a single corporation it would have ranked between the Berkshire Hathaway company of Omaha and the JeffersonSnmtilt company of St. Louis. The number one c(qx)rati(m, General Motors,” had sales at-xwt 44 times larger.

Appendix A Estimating the Economic Costs of Export Controls | 69

m

■

●

revenues. (The machine tool industries of othermajor producing countries are even more de-pendent on exports. See figure A- 1 for distribu-tion of the world machine tool export market.)Machine tools accounted for about 3.4 percentof U.S. durable goods exported in 1992.Import penetration is high—in 1982 importsaccounted for 26.4 percent of machine toolsconsumed in the United States, but by 1992they were 46.3 percent of consumption (eventhough, for several years beginning in 1987,"vo]untary restraint agreements” between theUnited States and several other machine toolproducing countries helped restrict exports tothis country).In the 1980s the United States consistently im-ported a billion or more dollars per year moremachine tools than it exported. The differencewent down to about $700 million in 1992, butconsumption also declined.

All types of machine tools are not subject to ex-port controls. Those subject to nonproliferationexport controls are primarily computer-control ledtools of relatively high precision. Numericallycontrolled machine tools of all types accountedfor about $304 million, or 36 percent, of the U.S.machine tools exported in 1992, meaning thatthey accounted for about 12 percent of machinetool industry revenues. Table A-1 shows that in1992, the Commerce Department approved 572applications containing over $454 million worthof machine tools controlled for national security(COCOM) or foreign policy reasons. In the samecategories of control, over $7 million in licenseapplications were denied. Recall that IVLS are for2-year periods and that the figures represent pro-posed sales, not actual shipments (nor shipmentsthat would have taken place had a 1icense not beendenied). Moreover, machine tool shipments gen-erally occur 9 to 18 months after orders are placed.Nevertheless, the table suggests that a substantialportion of U.S. machine too] exports require

All others 17°/0

France 2.90/.United Kingdom 3.40/.

Tawan 3.9%. , —

Italy 9.1 0/0

w.80/0

SOURCE Association for Manufacturing Technology 1993 and Officeof Technology Assessment 1994

IVLS. Note, on the other hand, that in the sameyear ( 1992), only two approved applications, val-ued at $1.8 million, were for machine tools con-trolled only for nuclear nonproliferation reasons,and only one application for such an export, val-ued at about $400,000, was denied. As COCOMcontrols are further altered, the impact of exportcontrols on the industry should decline.

In terms of dollar value relative to the GrossNational Product (GNP) or the overall export pic-ture, machine tools are not of great significance; interms of the dollar value of business subjected toindividual export licensing requirements, ma-chine tools constituted about 2.5 percent.

Nevertheless, individual machine tool firmsmay be at risk. They depend on exports to stay inbusiness and to supply revenues for research, de-velopment, and modernization. Since 1985, theUnited States has imported 40 to 50 percent of itsmachine tools. Machine tool industry advocatesargue that theirs is a strategic industry. buildingmachines

. . . essential to our military readiness and ourability to respond quickly and effectively in theevent of a national emergency ...4

~~orllas T, Conne]]) .. S[atenlen[ on B~ha]f of AhfT—Th~ Association” f{~r Manufacturing Tcchnt~lo:j-&>fore” th~ Subccmmlittce (m Ec(J-

mm]ic P~)lJcy, Trade, and Envir{mnwnt of [he House C(mm]ittec (m Foreign Affairs, ” Nov. 18, 1993, p. 2.


Licenses approved Licenses denied

Reasons for (Value $ ( V a l u e $ -

control Number millions) Number millions)

National security 312 309,4 6 6.3(COCOM) only

National security 260 144,6 5 1,1and foreign policy

Foreign policy only 1 0 0 6 0 0Nuclear proliferation 2 1 8 1 0.4

only

Totals 575 455.8 12 7,8

SOURCE Department of Commerce, 1993

If the United States wants to maintain some ele-ments of the U.S. machine tool industry for na-tional security reasons, it may find that exportcontrols that put the industry at a competitive dis-advantage can interfere with that goal. The indus-try is highly dependent on exports for its liveli-hood and its research and development resourcebase; since it is a relatively small industry, busi-ness failure of a few key firms could have a majoreffect on the indigenous supply of advanced ma-chine tools. In addition, industry advocates pointout that U.S. companies are most competitive inthe technological “high-end” products—the onesmost likely to be subject to export controls.

| Costs to U.S. EconomyOn the other hand, of the eight countries thatbought nearly 70 percent of U.S. machine tool ex-

ports in 1992, only two, China and Taiwan (to-gether accounting for about 9 of those 70 percent-age points) were likely to cause any proliferationconcerns and possibly evoke licensing delays. Atpresent, there is no concrete evidence to show thatexport controls in general, let alone the small frac-tion represented by nonproliferation controls,have in fact significantly harmed the industry.Better data, however, might show otherwise andgive policy makers a better notion of whether onetype of control or another places either the indus-try or particular firms at risk.

—

Appendix B:Complexities ofSetting Export

Control Thresholds:Computers B

1993 COMPUTER EXPORT CONTROL CHANGESIn September, 1993, the Clinton Administration announced that itwould:

■ increase the threshold of computer capability above which U.S.licenses to most destinations would be required from 12.5MTOPS (Million Theoretical Operations Per Second)] to themaximum that current Coordinating Committee on Multilat-eral Export Controls (COCOM) agreements would allow, 194MTOPS;

● propose to COCOM partners to raise the multilateral thresholdfurther to 500 MTOPS;

■ propose to raise the definition of a supercomputer (in the bilat-eral control agreement with Japan) from 195 MTOPS to 2,000MTOPS and review and update the requirements for safe-guards on exported supercomputers;

■ expand the availability of distribution 1icenses for computerexports; z and

■ eliminate the control threshold for shipments to COCOM andCOCOM-cooperating countries and increase the threshold for

] MTOPS is the unit of measurement In the Department of Commerce’s standard of“CompJsite Theoretical Pcrf(mnance” by which c(m]putcrs arc compared for expwt con-trol purposes.

2A chstributi(m I Icense allows a company to rmmit(lr its own exports for certain ilcms,prowded it maintains an internal control mechanism and submits I(J periodic government

audits of Its exp)rt rcc(mis. See Expwt Adnllniswatl(m Regulations, 15 Cl% ~ 773.3 (Jan.

1, 1 993).

It is questionable how

significant a role

advanced computation

may play in improving the

designs of a nuclear

proliferant such as Iraq,

Pakistan, or North Korea,

especially in the absence

of nuclear testing.

I 71


shipments to many other destinations up to thesupercomputer level.3

The Department of Commerce (DOC) esti-mated that the first step would free about $30 bil-lion worth of computer exports annually from therequirement to obtain licensing authorization (seefigure B-1 ). That $30 billion constitutes nearly 52percent of the $58 billion worth of computer ex-ports roughly estimated by the DOC to require ei-ther distribution licenses or individual validatedlicenses in fiscal year 1993 (note, however, thatdistribution licenses generally do not cover itemscontrolled for nonproliferation reasons). Thehigher threshold for defining supercomputerswould also freeup about $5 billion worth of com-puter exports annually from requirements forplacing safeguards on their end uses.

110-500 MTOPS

12.5-100 MTOPS $3B00 MTOPS$1 B

1Requiredindividuallicenses

< – - ,

\

v

) Covered bydistribution

licenses

(Various levels)

SOURCE Department of Commerce, 1993

In ensuing negotiations, COCOM partnersagreed only to decontrolling computers below260, not 500 MTOPS, although U.S. officials con-sidered this only an interim step. At the end ofMarch 1994, the DOC announced that individuallicenses would no longer be required for ship-ments of computers up to 1,000 MTOPS to formerCOCOM target countries. (The threshold wouldremain at 500 MTOPS for sales to nations listed inthe Export Administration Regulations as beingof nuclear proliferation concern.4) For the super-computer control agreement, Japan would onlyagree to raising the threshold defining supercom-puters to 1,500, not 2,000 MTOPS.

UTILITY OF COMPUTERS FORDESIGNING WEAPONS OF MASSDESTRUCTIONComputers at the level of today’s high-perfor-mance machines are not now—and never were—an essential technology for designing fairly so-phisticated nuclear weapons .5 Computers cancontribute to weapon design by simulating thecomplex, high-speed physical processes occur-ring in a nuclear weapon. However, they are farfrom being critical tools that will make a differ-ence in whether a country acquires nuclear weap-ons or not. Moreover, they are of most use to stateswith nuclear testing experience, since the calcula-tions performed in weapon simulations are vali-dated with test data.

Advanced weapon designers rely heavily oncomputers, and designers at any level of experi-ence may also wish to use—although do not re-quire—advanced computational capability. Nev-ertheless, the United States, drawing on itsextensive body of nuclear test data, developedhighly advanced nuclear weapons with computers

sTrade ~omotion and coordinating” Committee, U.S. Department of C(mm~erce, A Messagejtir Grow fh in a G/obal EwnomY: US E.vwr(s= US Jobs (Washington, DC: U.S. Department of Commerce, 1993).

qln addltl(~n, c(~nlputers above 6 MTOPS would c(mtinuc to be denied to Iran and Syria, while Cuba, Iraq, and Libya continued to k general-

ly embargoed by the United States.

5For a discussion of the Utl]lty of high-perfomlance computers to a nuclear proliferant, see U.S. Congress, Office of Technology” Assess-

ment, Technologies Underlying Weapons ofMass Destruction. OTA-BP-ISC-115 (Washington, DC: U.S. Government Printing OffIce, Decem-

ber 1993), pp. 125, 150-152.

Appendix B Complexities of Setting Export Control Thresholds: Computers | 73

vastly less capable than today’s high performancemachines. The Soviet Union and China developedtheir nuclear weapons with even less computingpower. 6

High-performance computers are relativelymore important for advanced weapons, includingthermonuclear ones, than for first-generation fis-sion weapons. They can also be useful in the de-sign of ballistic missiles’ and other conventionalmilitary systems. According to a 1986 Depart-ment of Energy Report,

With large-scale computers, we have beenable to improve our designs by optimizing de-sign parameters, while reducing the number ofcostly experiments in the design process. (Testsinvolving high explosives have been reducedfrom 180 tests for a 1 955-vintage weapon tofewer than 5 for today’s weapons because ofcomputation.) 8

Moreover, although non-nuclear tests can pro-vide information on the processes by which a nu-clear explosion is triggered, no laboratory tests(other than computation) can simulate the proc-esses of release of energy from nuclear materials.Therefore, the ability to carry out computer simu-lations can help weapon designers optimize thedesigns they want to test. Lacking adequate com-putational capabilities, the designers of the firstU.S. nuclear weapons had to build in large mar-gins of error, making the weapons much bulkierand heavier than they are today.

A U.S. supercomputer available in the early1980s (the period immediately preceding theDOE report on supercomputer utility) was theCray X-MP, whose peak performance was about235 MFLOPS (Million Floating Point OperationsPer Second—in this case roughly equivalent to the

Commerce Department’s MTOPS). This wasabout half the threshold that the Clinton Adminis-tration proposed in September 1993 to decontrolto most destinations and one-quarter of the March1994 threshold.

LIMITS OF EXPORT CONTROLSIt is questionable how significant a role advancedcomputation may play in improving the designs ofa nuclear proliferant such as Iraq, Pakistan, orNorth Korea, especially in the absence of nucleartesting. A judgment on this question would de-pend on:■

m

m

■

whether and to what extent the proliferant wereable to obtain design information from one ofthe nuclear powers,how far both simulations and weapon designscan be refined in the total absence of actual nu-clear tests,how capable the proliferant is of acquiring andusing the necessary software, andthe minimum practical thresholds of computa-tional capability for carrying out the necessarys emulations.

THE QUESTION OF FOREIGNAVAILABILITYCritics of the Clinton Administration’s relaxationof computer export controls have pointed out thatthe Nuclear Non-Proliferation Treaty (NFT) is alegally binding undertaking”. . . not in any way toassist, encourage, or induce any non-nuclear-weapon State to manufacture . . .“ nuclear weap-ons; in this view, the phrase “in any way” is notconditioned by whether other nations are provid-ing assistance, by whether U.S. firms are losing

6SCC Jack Wt)rlt(m, “’St)me Myths Ahmt High-Perf(mnancc Ctmlputcrs and Their Role in the Design of Nuclear Weapms,” Worlt(m &Associates Tcchmcal Report N(). 32. June 22, 1990, and “ExFn-t C’(mtrols ft~r High-Perf(Jmlance C(mlputers in the 1990s: A Reassessment,”’W’orltfm & Ass(~]ates Techn]cal Rep{wt Nt). 43, Nov. 1, 1993. SW also the tes[imtmy of John Han ey hcf(w the U.S. H(mse of Representatives

Cc~nln]ittec on Science. Space. and Technology, Aug. 13, 1993.

‘See Gary M ilhollin, “’Designing the Third W{)rld B(mlb,” Wi.~con.\in Atadcm) Re\IeM, winter 1990-1991, pp. 15-18.

Xsee w]]] ,anl D. Wilson et a],. .-~e NCC~ for Supercomputcrs In Nuclear w~:l~)ns Design. “ manuscript, U.S. Department of Energy, Officeof M]lltary Applicat](m, January 1986, p. 9.


Proposed new control level

Sun Server

COCOM 1994 level

Silicon Graphics workstation

DEC workstation (Alpha)

Apple PowerPC (80MHz)

Pentium PC

Previous control level

486-66 MHz

386-33 MHz PC

260

4~ ~o 100 200 300 400 500

CTp (millIon theoreticaloperations per second)

SOURCE Department of Commerce, 1993, Apple Computer, 1994,

and Off Ice of Technology Assessment, 1994

legitimate exports because of NPT compliance, orby the degree of importance of the assistance.9

The U.S. Nuclear Non-Proliferation Act of 1978(Section 309) does specify that the Department ofCommerce should control ‘*. . . all export items[other than those licensed by the Nuclear Regula-tory Commission] which could be, if used for pur-poses other than those for which the export is in-tended, of significance for nuclear explosivepurposes.” The definition of “significance” is notgiven, but clearly nuclear weapon designerswould rather have computers than not, and wouldrather have more computing power than less.

Proponents of the computer decontrols arguethat the potential effectiveness of controls shouldbe taken into account. Although computers abovethe thresholds previously controlled by the UnitedStates may be useful to proliferant nations, they

are increasingly available from non-U.S. sources.Despite continuing to control supercomputers inpart because of apparent nuclear proliferationrisks, the United States was unable to persuade theother members of the international Nuclear Sup-pliers Group to place them on the Group’s list ofmultilaterally controlled dual-use technologies.

Although COCOM did control computersabove the 195 MTOPS level, Administration offi-cials judged that agreement on this thresholdcould not be sustained as COCOM underwent fur-ther post-Cold-War revision. Figure B-2 showsthe Composite Theoretical Performance (CTP) ofseveral U.S.-made computers, for which the cen-tral processing units have become or soon will be-come widely available throughout the world. A1992 Commerce Department study of foreignavailability of computers showed that machinesexceeding the 12.5 MTOPS threshold were avail-able from Brazil, China, Hong Kong, India, SouthKorea, Singapore, and Taiwan—none of whichwas a member of COCOM. Machines exceeding60 MTOPS were available from Hong Kong, In-dia, and Taiwan. The report predicted that widelyavailable first-generation workstations based onthe newest microprocessors would have CTP val-ues ranging from 50 to 194 MTOPS. In general,advanced microprocessor chips are not con-trolled, and would be very difficult to control be-cause of their small size, low cost, and vast con-sumer distribution.

Not only are higher performance central proc-essing units becoming more widely available, butpersonal computers and work stations can be net-worked to process data in parallel, allowing themto exceed the performance of any element in thenetwork. The hardware and software for doing sois widely available and not difficult to use. 10 Onthe other hand, some kinds of simulations may notbe amenable to parallel processing, but instead re-quire direct access by a single central processingunit to a large amount of random access memory.

gApp]ylng this ~trlcture t{) dual-use exPJ~s, however, has not been subscribed to either by U.S. administrations or by other N~ members.

Ioworlton,” .LExP)fi controls” ft)r High-Perfomlance Computers . . ..” ibid.

Appendix B Complexities of Setting Export Control Thresholds: Computers | 75

Depending on the job the weapon designer is try- and end-user controls on American computersing to do, parallel processing may or may not be penalized them while serving no useful purpose.11

useful. Moreover, with the coming widespread availabil-Since high-performance computers are avail- ity of new high-power commercial processors

able from foreign sources and are not essential to such as the Pentium, Alpha, and Power PC, U.S.whether any nation acquires nuclear weapons, computer makers could lose much of the new mar-U.S. companies argued that requiring licensing ket likely to center on those chips.

I I Testimony of Tlnl Dwyer of Sun Mjcrosystcnls, speaking for the American Electr{mics Associati(m at a hearing of the SUhCtmlnlittee on

Ec(mornic Policy, Trade, and Envir(mment of the House Foreign Affairs Committee, June 9, 1993.

BWCBXA

CCLCOCOM

CTP

CwcDOCDOEEAAEAR

ECECCN

EPCI

EUIAEA

Biological Weapons ConventionBureau of Export Administration(Department of Commerce)Commerce Control ListCoordinating Committee onMultilateral Export ControlsComposite Theoretical Per-formanceChemical Weapons ConventionDepartment of CommerceDepartment of EnergyExport Administration ActExport Administration Regula-tionsEuropean CommunityExport Control ClassificationNumberEnhanced Proliferation ControlInitiativeEuropean UnionInternational Atomic EnergyAgency

Appendix C:Glossary of

Abbreviations cISRO

IVLMFLOPS

MTCR

MTOPS

NAM

NPTNSGOEL

PINS

PSLVRWASNEC

Indian Space Research Orga-nizationIndividual Validated LicenseMillion Floating Point OperationsPer SecondMissile Technology ControlRegimeMillion Theoretical Operations PerSecondNational Association ofManufacturersNuclear Non-Proliferation Treat yNuclear Suppliers GroupOffice of Export Licensing,Bureau of Export Administration,Department of CommerceProliferation Information Net-worked SystemPolar Satellite Launch VehicleReturned Without ActionSubgroup on Nuclear ExportCoordination

I n

Index

AAbbreviations glossary, 77ACDA. See Arms Control and Disarmament AgencyAgencies administering U.S. export control laws, 18All-inclusive lists, 8-9,60-61Approaches to dual-use export control policy

export enhancement emphasis, 13-16nonproliferation emphasis, 12-13“rogue” nation nonproliferation emphasis, 13

Arms Control and Disarmament Agency, 3,36Atomic Energy Act, 33Australia Group

current U.S. export controls, 9, 18, 59, 60multilateral export control options, 5, 6, 42, 46,

49

BBelarus, 6,49Benefits of export controls. See also Enhancing ef-

fectiveness of export controlsbackground, 21-22cooperation and enforcement variables, 23-25introduction, 13technology variables, 22-23

Biological weapons, export control effectiveness23

Bureau of Export Administration, 4,38, 39, 48BXA. See Bureau of Export Administration

for,

cCensus Bureau, 65Central Intelligence Agency, 48Checks and verifications, improvement of, 3-4,38Chemical weapons, export control effectiveness for,

22-23Chemical Weapons Convention, 23China, 70Christopher, Warren, 57Clinton, Bill, 6,50Clinton administration draft EAA

ACDA role changes and, 36

developing country cooperation incentives, 6,52-53

economic impact statement option, 65-66licensing administration options, 10,64-65sanctions provisions, 4, 39-42

COCOM. See Coordinating Committee on Multilat-eral Export Controls

Commerce Control List sub-categories, 17-18,55-57

Commerce Departmentcomputer export controls, 30, 72, 74computer system weaknesses, 67-68export licensing role under EAA, 17-18foreign availability question, 73-75IVL guidelines, 59-60“know” rules guidance, 26-27,59-63licensing administration options, 3, 10,33-37,46,

64-65list-making options, 7-9,57-58machine tools industry case study, 69pre-license checks, 4,38

Communist countries, 29. See also specific coun-tries by name

Competitiveness issues. See International competi-tiveness issues

Computer export controls, 30,56,58-59,71-75Coordinating Committee on Multilateral Export

Controlsbackground, 7-8computer export controls, 58-59, 71-72, 74current U.S. export controls, 18economic impact assessments, 65estimating costs of export controls, 29-30multilateral export control options, 4-5, 42-44phaseout of, 8,56reexport controls, 64

Costs of export controls. See also Reducing the bur-dens on industry

economic impact statement option, 10,65-66estimating example, 67-70introduction, 1-2, 11-16to exporting firms, 26-30

I 79


to the U.S. economy, 29,70to the U.S. government, 25-26

Criteria for dual-use export control policy, 13-14Current U.S. nonproliferation export controls,

17-20. See also Benefits of export controls; Costsof export controls; Reducing the burdens on in-dustry

Customs Service, 4,38CWC. See Chemical Weapons Convention

DDefense Department, 3,30,32,36,64Developing countries, 6-7,51-54DOC. See Commerce DepartmentDOE. See Energy DepartmentDual-use controls under the Export Administration

Act. 17-18

EEAA. See Export Administration ActEAR. See Export Administration RegulationsEC. See European CommunityEconomic costs of export controls. See Costs of ex-

port controlsEconomic sanctions, 4,39-42Effectiveness of export controls. See Benefits of

export controls; Enhancing effectiveness of exportcontrols

Energy DepartmentNuclear Referral List procedures, 36-37Office of Arms Control project, 45Proliferation Information Networked System,

32-33,34Enforcement options

export control benefits assessment and, 23-25extend sanctions, 4, 39-42improve pre-license and post-shipment checks,

3-4,38increase export control awareness among ex-

porters, 3,37-38increase information processing resources, 4, 39increase law-enforcement resources, 38reinforce the “know” rule, 38-39

Enhanced Proliferation Control Initiative rules, 8-9,26-27,59-63

Enhancing effectiveness of export controls. See alsoBenefits of export controls

enforcement options, 3-4,37-42introduction, 1-2, 11-16licensing administration options, 2-3,33-37list-making options, 2,31-33multilateral export control options, 4-7, 42-54

EPCI rules. See Enhanced Proliferation Control Ini-tiative rules

EU. See European UnionEurasia, 24European Community, 47. See also European UnionEuropean Union, 24-25,47,63Export Administration Act of 1979, 1, 11, 14, 17-18,

56Export Administration Act of 1994 drafts. See Clin-

ton administration draft EAA; National Associa-tion of Manufacturers, draft revision of the EAA

Export Administration Regulations, 38-39

FForeign availability of commodities, 7,56-58,73-75Foreign Commercial Service, 38Foreign policy export controls, 18,56Former Soviet Union republics, 6,23,25,47-48,

49-51

GGalloway, John, 32Germany, 24-25,63Glenn, John, 34

HHigh-performance computers. See Computer export

controlsHigh-technology industries, 30,56-59,71-75

IIAEA. See International Atomic Energy AgencyIndia, 24,51Individual Validated Licenses

Clinton administration draft EAA and, 6,52-53EPCI rules and, 8,26-27,38-39,59-60machine tool industry case study, 69

Industry burdens from export controls. See Reduc-ing the burdens on industry

Intelligence-sharing increases, 5-6,48-49Interagency review processes, 3,35-37International Atomic Energy Agency, 45-46International competitiveness issues, 7-16,56-58,

64-65,73-75International cooperation. See Multilateral export

control optionsIran, 13,57Iraq, 13,36“Is informed” rules, 9,26-27,63IVLS. See Individual Validated Licenses

JJapan, 72

Index | 81

K“Know” rules, 9,26-27,38-39,61-63

LLibya, 13Licensing administration options

formalize interagency review processes, 3,35-37modernize the license-processing database, 2,

33-34place statutory limits on processing times, 9-10,

64-65publish economic impact statements, 10,65-66publish nonproprietary licensing data, 2-3,34-35seek cooperation from developing countries, 6-7streamline the application process, 64-65

List-making optionscreate databases for list-making, 2, 31-33EPCI rule options, 8-9,59-63foreign availability of commodities and, 7,56-58narrow the scope of list purposes, 7-8, 58-59

MMachine tools industry, 68-70Missile technology, export control effectiveness for,

23Missile Technology Control Regime

current U.S. export controls, 9, 18, 59export controls benefits assessment, 24, 60multilateral export control options, 5, 6, 42, 46,

49,51-52sanctions for violations of, 39

MTCR. See Missile Technology Control RegimeMultiagency reviews, 3,35-37Multilateral export control options

end unilateral reexport controls, 63-64enhance NSG coordination, 5, 44-46expand international reporting to approvals, 5, 46expand the NSG database idea, 5, 46export control benefits assessment and, 23-25foreign availability of commodities and, 56-58increase intelligence sharing, 5-6, 48-49seek greater cooperation from developing coun-

tries, 6-7, 51-54separate COCOM succession from nonprolifera-

tion regimes, 4-5, 42-44support development of FSU administration of

export controls, 6, 49-51

NNAM. See National Association of ManufacturersNational Academy of Sciences, 43National Association of Manufacturers

costs study, 27-28

draft revision of the EAA, 60-61National security export controls, 17-18,29, 56. See

also Coordinating Committee on Multilateral Ex-port Controls

NETMAP, 32North Korea, 13NPT. See Nuclear Non-Proliferation TreatyNSG. See Nuclear Suppliers GroupNuclear Export Reorganization Act of 1993,37Nuclear Non-Proliferation Act of 1978, 18,74Nuclear Non-Proliferation Treaty, 51,53-54,66,

73-74Nuclear Referral List, 36-37Nuclear Suppliers Group

current U.S. export controls, 9, 59, 60dual-use export control guidelines, 24enhancing NSG coordination, 5, 44-46expanding the NSG database idea, 5, 6, 46-48foreign availability question, 74multilateral export control options, 42, 49reexport controls, 64

Nuclear weaponscomputational requirements to design, 72-73export control effectiveness for, 22

0OEL. See Office of Export LicensingOffice of Arms Control, 45Office of Export Licensing, 7-8,33,58

PPerry, William J., 30,67PINS. See Energy Department, Proliferation In-

formation Networked SystemPrivate institutions as intelligence sources, 6,48-49Publication

of economic impact statements, 10, 65-66of information on suspect buyers, users, and

programs, 3,9,37-38of lists of controllable items, 9, 63of nonproprietary licensing data, 2-3, 34-35

RReducing the burdens on industry

economic impact statement option, 10,65-66ending unilateral reexport controls, 63-64EPCI rule options, 8-9,26-27,59-63estimating costs of export controls, 26-30increasing export control awareness among ex-

porters, 37-38introduction, 1-2, 13-16licensing administration options, 9-10,64-66list-making options, 7-9,55-63

Reexport controls, 63-64


Reporting requirements as export controls, 5,9,46,59,63

“Rogue” nations, 13Russia, 6,24,49-51

sSanctions in export controls, 4,39-42SNEC. See Subgroup on Nuclear Export Coordina-

tionSoviet Union. See Former Soviet Union republics;

RussiaState Department, 4Subgroup on Nuclear Export Coordination, 37Supplier-group export controls. See Multilateral ex-

port control options

TTaiwan, 70

Technology variables in assessing export controlbenefits, 22-23

Telecommunications technology controls, 58-59

uUkraine, 48Unilateral imposition of controls, 2,7, 14,56-58.

See also Enhanced Proliferation Control InitiativeRules

vVerifications, improvement of end-use, 3-4,38

wWeapons of mass destruction regimes vs. conven-

tional weapons regime,

YYeltsin, Boris, 6,50

PLEASE PHOTOCOPY - DO NOT CUT PAGE— — — — — — — — — — -

Order Processing Code:

*7394

4-5,42-44

Superintendent of Documents Publications Order Form

❑ YES, please send me the following:

copies of Export Controls and Nonproliferation Policy (92 pages),S/N 052-003 -01371-1 at $5.50 each.

P3Telephone orders (202) 783-3238

(The best time to cull is between 8-9 a.m. EST.)To fax your orders (202) 512-2250

Charge your order. It’s Easy!

The total cost of my order is $ . International customers please add 25%. Prices include regular domesticpostage and handling and are subject to change.

Please Choose Method of Payment:(Company or Personal Name) (Please type or print) ❑ Check Payable to the Superintendent of Documents

(Additional address attention line)❑ GPO Deposit Account C u I n n - n❑ VISA or MasterCard Account

(Street address) I I l l l l l l l l l l l l l l l l ! l l11111 (Credit card expiration date) Thank you for

(City, State, ZIP Code) your order!

(Daytime phone including area code) (Authorizing Signature)

YES NO

(Purchase Order No.) May we make your name\ address available to other mailers? ❑ ❑

Mail To: New Orders, Superintendent of Documents, P.O. Box 371954, Pittsburgh, PA 15250-7954

Export Controls and Nonproliferation Policy

Documents

Transcript of Export Controls and Nonproliferation Policy