Tito's Nuclear Legacy

We must have it

According to Stevan Dedijer, a former director of Vinca, the government of Josip Broz Tito decided in the late 1940s to develop the capability to build nuclear weapons. Dedijer wrote in 1969 that he was recruited to the program in September 1949. He recalled a meeting he had on the morning of January 17, 1950, with Edvard Kardelj, one of Tito's closest collaborators. Kardelj said, “We must have the atomic bomb. We must build it even if it costs us one-half of our income for years.” 3

Although there is little documentary evidence of a military dimension to Yugoslavia's nuclear program before 1953, the wide-ranging and costly nature of the nuclear research initiative is apparent. The cornerstones of the early program were three nuclear research centers. The Institute of Research on the Structure of Matter (subsequently renamed the Boris Kidric Institute of Nuclear Sciences) was founded at Vinca in 1948; the Jozef Stefan Institute, under the direction of Anton Peterlin, was started in Ljubljana, Slovenia, in 1949; and in 1950, the Rudjer Boskovic Institute, under the leadership of Ivan Supek, was founded in Zagreb, Croatia.

OPEN IN VIEWER

May 28, 1999: An aerial view of the Vinca Institute of Nuclear Sciences, taken during NATO's spring bombing campaign.

According to an official history, titled Nuclear Energy in Yugoslavia, more than 1.75 billion dinars (about $35 million at the 1953 exchange rate) was spent on building and operating the institutes between 1948 and 1953. 4 An additional billion dinars may have been expended on uranium ore exploration and production during the same period.

This was an enormous investment for the young Yugoslav state, and it was probably not coincidental that the nuclear program was launched shortly after Tito learned of Stalin's nuclear weapons plans and Yugoslavia was excommunicated from the Soviet-led camp.

Evidence of the weapons orientation of the program is found in a document dated May 25, 1953, “O dva bitna uslova za razvitak atomske en-ergije kod nas” (“On Two Essential Conditions for the Development of Atomic Energy Here”), which was drafted by Stevan Dedijer, Pavle Savic, and Robert J. Walen, a Dutch-born senior scientist at Vinca. Prepared for the top political leadership, it identifies “production of atomic weapons” as the first of two goals for the Yugoslav atomic energy program. (The other goal was the “use of atomic energy in the economy.”)

The authors complained that the weapons effort had been hindered by Yugoslav bureaucracy and the concealment from the scientific leadership of key information regarding the organization of research efforts. It offers a number of specific illustrations about how this policy of concealment, “immeasurably sharper than that of any country, except in the Soviet bloc,” might hamper the timely purchase of 10 tons of heavy water from Norway.

The document concludes with a plea that “for the time being, at least for a limited number of scientists at our Institute, all restrictions on information concerning organization, methods, results, and investments in the field of uranium research must be removed.”

The document leaves no doubt as to the military objective of the nuclear program–a perspective shared by at least one U.S. intelligence assessment from the period. According to a January 23, 1954 report from the U.S. Army attaché in Athens, “the Yugoslavs have commenced a program to produce atomic weapons.” 5

Beginning in 1955, formal responsibility for supervising the Yugoslav nuclear program was vested in the Federal Commission for Nuclear Energy headed by Aleksandar Rankovic. Rankovic was vice president of the Federal Executive Council–and head of the secret police, which made him one of the most influential officials in Yugoslavia.

Before the war, Rankovic had been an apprentice to a peasant tailor, making him an odd choice to head a large scientific program whose purpose was to promote atomic energy in economic development. But if the true purpose of the nuclear program was military, Rankovic's background in the secret police made more sense. And it paralleled the role played by Secret Police Chief Lavrenti Beria in the Soviet bomb program.

During the 1950s, Yugoslavia pursued an ambitious program of nuclear research consistent with the development of a capability to build an atomic bomb. Vinca's Department for Spent Fuel Reprocessing was created in 1956; a heavy water-moderated zero-power critical assembly “RB reactor” was constructed at Vinca in 1958; and in 1959 a Soviet-designed and built 6.5-megawatt heavy water-moderated “RA” research reactor capable of using uranium fuel enriched to 80 percent uranium 235 was commissioned at Vinca. The heavy water and highly enriched uranium for the reactors were provided by Moscow. 6

Although research was heavily tilted in favor of plutonium, Yugoslav scientists at both the Rudjer Boskovic Institute and Vinca also studied uranium enrichment. Vinca's Laboratory of Physical Chemistry used a calutron for that purpose, and scientists in Zagreb used a small cyclotron to research electromagnetic isotope separation techniques.

The Norwegian connection

Yugoslavia's closest foreign collaborator in the nuclear sector in the early 1950s was Norway, a country with an advanced nuclear research program, initially driven in part by its own military considerations.

The leader of the Norwegian nuclear research program, Gunnar Randers, visited Vinca in 1952; that year, the Yugoslav scientist Dragoslav Popo-vic began a two-year residence at Ran-ders's Institute for Nuclear Energy Research in Kjeller, Norway.

Correspondence in 1953 between Randers and other Norwegian nuclear scientists and Dedijer (by then the director of Vinca) indicates that Yugoslavia was especially interested in the chemical extraction of plutonium from irradiated fuel. Several scientists from Vinca spent years at Kjeller researching the process. 7

By mid-1953, Dedijer had a draft agreement with the Norwegian firm Norsk Hydro-Elektrisk Kvaelstofak-tieselskab to purchase 10 tons of heavy water. (Yugoslavia also may have succeeded in smuggling a quantity of highly enriched uranium from Kjeller to Vinca in 1953.) 8

Although Vinca chose not to consummate the heavy water deal, opting for a less expensive purchase from the Soviet Union, the institute became increasingly involved with Norway on work in plutonium reprocessing. This is reflected in a variety of scientific publications by Yugoslav authors working at Kjeller from the mid-1950s to the early 1960s, as well as by intensive negotiations to procure a plutonium reprocessing plant for Vinca. 9

In a letter to Randers dated October 22, 1960, Yugoslav Undersecretary of State Slobodan Nakicenovic indicates that the plant was needed for reprocessing spent fuel from the research reactor at Vinca. According to Naki-cenovic, the reprocessing plant “should have a capacity of at least 1.5 tons of uranium per year to meet the requirements of the RA reactor under a forced regime.”

Nakicenovic suggested that if the plant were modeled after the one at Kjeller, its capacity might be increased to 10 tons per year as “a reserve for a possible increase in power of the RA reactor or for fuel from future new reactors.” 10

Subsequent correspondence between Yugoslav and Norwegian officials indicates that an agreement was reached between Vinca and the firm Noratom to build a reprocessing laboratory at Vinca based on the Norwegian plant at Kjeller–but on a larger scale. The engineering blueprints for the plant were delivered to Yugoslavia in 1962, and in 1966 Norway shipped gram quantities of high-grade plutonium to Vinca.

By this time, however, it appears that Tito had put his nuclear weapon aspirations on hold, and the Norwegian plant was never built. Nevertheless, a laboratory-scale reprocessing facility, equipped with four hot cells and using the Purex process to treat uranium metal fuel from the RA reactor, was in operation at Vinca by 1966. 11

India's wake-up call

Why Tito deactivated the nuclear weapons program sometime in the early 1960s is a puzzle. Perhaps he believed that economic development and civilian nuclear power deserved more attention than weapons. Or perhaps it was because the weapons program was costly and little progress had been made. The diminished Soviet threat to Yugoslavia may have been a factor.

As one of the principal contenders for leadership of the global non-aligned movement, and as a public champion of nuclear disarmament, Tito may have decided that Yugoslavia's position would be adversely affected if its nuclear weapons aspirations became known.

Whatever the reasons for the policy shift, it proved temporary. The catalyst for the program's revival came on May 18, 1974, when India–Yugoslavia's rival for leadership of the non-aligned movement–conducted its first nuclear test.

Less than a month later, in early June, the heads of the major Yugoslav nuclear research institutes, as well as senior representatives of the armed forces and the military intelligence service, were summoned to the headquarters of the Yugoslav National Army. Co-author Slaus, then acting director of the Rudjer Boskovic Institute, was among those who received an urgent summons to the top-secret meeting. Fewer than 20 individuals attended the half-day event.

As was the custom under Tito, the meeting was ethnically diverse–several Slovenes, a Macedonian, a Montenegrin, two Croatians, and a number of Serbs. All participants signed non-disclosure documents, and they were told not to discuss the subject of the meeting with the republic-level ministers of science.

The participants were instructed by military representatives to utilize a dramatically expanded nuclear power program as a cover for a parallel military effort. Institute directors were asked to provide a list of names of scientists at their institutions who should be involved.

The next important meeting about which we are aware was held at Mo-rovic, one of Tito's retreats in Vojvodina, between December 23 and December 28, 1974. At this slightly larger gathering, participants were once again told that Yugoslavia planned to develop nuclear weapons. Much of the discussion was devoted to determining which civilian energy programs could best conceal military activities. At that time, the most substantial nuclear project involved the construction of a nuclear power plant by Westinghouse in Krsko, Slovenia.

While the weapons program was secret, in retrospect one can find hints of a reorientation in Yugoslavia's nuclear policy. At the 1975 Nuclear Non-Proliferation Treaty Review Conference, for instance, Yugoslavia strongly criticized the lack of international cooperation in the use of nuclear energy for peaceful purposes as promised by Article IV. It also condemned the nuclear weapons states for not fulfilling their obligations under the treaty to discontinue the nuclear arms race and to stop nuclear testing. Both views were shared by many delegates.

Although Yugoslavia did not block approval of the final declaration of the conference, it denounced the declaration for not reflecting the deliberations and positions stated at the conference. In a letter to the president of the conference, the Yugoslav representative said that his government “finds itself in a position to re-examine its attitude toward the treaty and to draw corresponding conclusions.” 12

Later, at the thirtieth session of the U.N. General Assembly, Yugoslavia said that if the nuclear powers continued to preserve their weapons monopoly, it was “not realistic to expect a number of countries to be prepared to refrain voluntarily and over the long term from acquiring modern and sophisticated weapons.” 13

In December 1975, an article in the government newspaper Borba was more specific, suggesting that as part of its general defense concept Yugoslavia might have to “reconsider its attitude toward the question of nuclear weapons.” 14 This view was echoed by a senior military official in March 1977, who indicated that although Yugoslavia did not possess an atomic bomb it might be forced to reconsider abstinence unless there were a change in the behavior of the “club of nuclear power.” 15

It is unclear how the United States regarded Yugoslavia's nuclear weapons ambitions in the mid-to late 1970s. A study by the Mitre Corporation sponsored by the Ford Foundation included Yugoslavia in a list of “insecure” nuclear threshold states–states with the technical capability to develop nuclear weapons and the security motivations to seriously contemplate such an option. (The other states were Israel, South Africa, South Korea, and Tai-wan.) 16

Shortly after the release of the Mitre report in January 1977, Joseph Nye, a key non-proliferation expert in the Carter administration, testified before Congress that the list of “imminent nuclear weapons states” was “quite a reasonable estimate.” 17

At about the same time, the Carter administration attempted to revise the nuclear fuel supply arrangements that had been concluded between the IAEA and Yugoslavia and sought to impose controls over the spent fuel from the Krsko nuclear power station.

Dual tracks

Because of his opposition to the plans for the military nuclear effort and the appointment of a new director of Rud-jer Boskovic Institute, co-author Slaus withdrew from the project after the December 1974 meeting. Another coauthor, Miljanic, did not join the program until January 1981. As a consequence, we have no first-hand information about the course of nuclear weapons research in Yugoslavia between 1975 and 1981.

Based on the state of the program in 1981, however, it is reasonable to conclude that relatively few resources were invested in weapons research and development during the period. What little work took place seems to have involved information collection, a decision as to the type of weapon to pursue (a plutonium-based implosion bomb), and elaboration and development of the civilian cover for the program. It was not until 1982–some two years after Tito's death–that Adm. Branko Mamula, the newly appointed secretary of defense, began to forcefully promote the weapons program.

OPEN IN VIEWER

November 30, 1953: Old comrades-in-arms celebrate the tenth anniversary of the Communist government. From left: Aleksander Rankovic, Tito, Milovan Djilas, Edward Kardelj. Rankovic was head of the secret police—and oversaw the nuclear program.

During most of Mamula's tenure, which lasted until 1988, Yugoslavia maintained two parallel nuclear programs: a dedicated nuclear weapons effort known as Program A, and a civilian nuclear energy project referred to as Program B. Although the two programs formally had different objectives, administrative bodies, and funding sources, they were closely linked.

While Program B pursued 11 projects ostensibly for peaceful purposes, its major activities actually included the design of a plutonium production reactor, uranium metal production, development of an expanded plutonium reprocessing capability, heavy water production, and the design and construction of a fast experimental assembly. 18

All of these projects were closely related to the needs of Program A. In fact, Program A assumed that the fissile material for a bomb would be provided by Program B. Among the major projects launched under Program A were: the design and development of the chemical high explosive package for an implosion-type fission bomb; production of the nuclear explosive components for the bomb, including the neutron source to initiate the chain reaction; design and performance of experiments with all components, except the nuclear explosive; computer modeling of nuclear processes for different bomb configurations and material composition; and exploratory studies of different aspects of underground nuclear testing.

Principal operational responsibility for overseeing Programs A and B resided with the nuclear division of the Military Technical Institute, headed by Col. Martin Sajnkar. Sajnkar, who had a graduate degree in chemistry from Zagreb University, was the closest approximation in the Yugoslav nuclear weapons program of the Manhattan Project's Gen. Leslie Groves.

Most of the nuclear work directly related to the weapons program was concentrated in the Belgrade area. Three institutes were involved: Vinca, the Institute of Physics, and the Military Technical Institute (MTI). Although each institute had specific areas of expertise and contractual obligations, much of the work involved teams of specialists drawn from all three.

While most of the work for Program A was concentrated in Belgrade, the Rudjer Boskovic Institute in Zagreb and the Jozef Stefan Institute in Ljubljana also had links to the military program. Our knowledge of activities in Ljubljana is sketchy. Although individual scientists at Jozef Stefan may have worked on MTI or Vinca subcontracts, the administration of the institute appears to have exploited both Program A and B primarily for the acquisition of equipment for non-military program objectives.

In Zagreb, Program A work was narrowly focused on the development of neutron sources for a bomb. A team of less than 10 physicists, electroengi-neers, and technicians from Rudjer Boskovic was directly involved in the project. Only two members knew the actual purpose of the work.

Initially, the Zagreb group was asked to learn how to design and produce a neutron source for a Nagasaki-like plutonium-implosion bomb. This led the team to examine a polonium-beryllium neutron source. Subsequently, alternative sources (deuterium-deuterium and deuterium-tritium) were analyzed. The work was mainly exploratory and no final decision was made on the selection of the neutron source. Although considerable work was done on the design of the sources, a prototype was never produced.

To house the secret work on Program A, a new building was built at Rudjer Boskovic. Although it was the site for work on nuclear waste–the Program B project and the civilian cover–it also housed investigations into possible neutron sources for the initiation of a fission explosion.

The Program A activities at Zagreb were the subject of the only detailed published report about the nuclear weapons work undertaken by Yugoslavia in the 1980s. A January 1988 article in the German magazine Der Spiegel reported that Rudjer Bos-kovic, with financial assistance from the Energoinvest Company of Sarajevo, had been engaged in reprocessing radioactive waste. 19 Although the article was mistaken about the nature of weapons work being done at Zagreb, it correctly pointed to the front role played by Energoinvest.

The most sensitive weapons-related projects under Program B were based in Belgrade. They included work at Vinca on the design of a plutonium production reactor (innocuously referred to as an experimental research reactor), fabrication of uranium metal fuel, plutonium reprocessing, and design and construction of a zero-power fast breeder reactor.

Principal responsibility for developing a heavy water production capability, however, was given to the Institute for Thermal and Nuclear Energy, an affiliate of Energoinvest. Overall responsibility for organizing and managing Program B activities was transferred from federal authorities to Energoinvest, the largest single commercial enterprise in Bosnia-Herzegovina. (Energoinvest had a work force of about 50,000; it was best known internationally in the nuclear field as the producer of steam separators for Soviet RBMK reactors.)

A March 1995 NUKLIN document reported that the “Program for Mastering the Technologies of the Nuclear Fuel Cycle,” of which Program B was a large component, would require “220 investigator years annually” until 1990. The cost of the program for that period was estimated to be 21.768 billion dinars (approximately $95 million). 20

Lack of urgency

On July 7, 1987, a Serbian holiday, coauthor Miljanic was invited to a meeting at MTI headquarters where it was announced that the presidency of Yugoslavia had decided that Program A would end. In the dozen years since that announcement, very little information has become available about the reasons for the program's halt. Major questions remain about the progress made before closure, the reasons for the lack of progress toward a weapons capability, and the extent to which remnants of the weapons programs remain in a potentially useful form.

When the nuclear weapons program was revived in 1974, the security threat posed by the Soviet Union was less acute than at the outset of the early weapons program. The threat, however, was still perceived as real because of the Soviet invasion of Czechoslovakia and the persistence of the “Brezhnev Doctrine” of limited sovereignty for socialist countries. In addition, an elderly Tito was acutely aware of the increased international and domestic threats Yugoslavia would face after his death.

Both he and senior members of the armed forces may have regarded nuclear weapons as a means to enhance deterrence vis-à-vis Moscow (and the United States) in a post-Tito Yugoslavia, as well as to deter Belgrade's immediate neighbors should they harbor irredentist aims.

As the leader of a small, developing, and morally righteous state in pursuit of its proper place in the international pecking order, Tito may have anticipated gaining prestige with a successful weapons program. Indeed, prestige, especially within the non-aligned movement, probably became the determining factor in Yugoslavia's decision to reactivate its weapons program.

This tendency would have been reinforced in the mid-1970s by the increasingly active efforts of some of the major nuclear-supplier states to tighten regulations on nuclear commerce.

Despite the pressures to go nuclear and the formal decision to do so in 1974, the Yugoslav weapons program was never characterized by a sense of urgency. This was reflected in the uneven quality of the scientists recruited for the program, the modest target dates set for the program, and the absence of serious consequences for those failing to meet targets.

Although the pace of work intensified after Tito's death and the rise in influence of the military, even the enthusiastic support of Defense Secretary Mamula failed to make the program comparable to those undertaken by India, Pakistan, Israel, South Africa, or even Iraq.

Unlike those countries, Yugoslavia's program was not primarily driven by pressing security concerns, and the country had paltry financial resources to devote to it. The effort also suffered from a distinct lack of enthusiasm for Program A objectives on the part of many of Yugoslavia's best nuclear scientists. (However, many of these scientists were adept at siphoning off Program A funds to procure laboratory equipment for non-weapons purposes.)

In addition, the Yugoslav weapons effort was impeded by a stultifying bureaucracy, as well as by the differing interests pursued by individual republics.

Future deterrent?

We are not in a position to accurately assess the overall progress Yugoslavia had made in acquiring a nuclear weapons capability when Program A was terminated in 1987 or when Program B came to a halt in 1990, following the adoption of a federal law in 1989 prohibiting the construction of additional nuclear power stations. 21

But regardless of the precise point Yugoslavia occupied in 1987 on the path toward nuclear weapons, arguably its most valuable and enduring asset is its highly skilled personnel. Dozens of Yugoslav physicists, chemists, and engineers obtained decades-long experience in a broad range of activities related to most aspects of the nuclear fuel cycle. A smaller but significant number also acquired expertise in nuclear weapons design, computer modeling of nuclear weapons processes, and the design and manufacture of chemical high explosives for use in a fission bomb.

In addition to its experienced work force, Yugoslavia's greatest weapons asset today is its nearly 50 kilograms of fresh, weapons-grade uranium fuel and 10 kilograms of low-irradiated highly enriched uranium. The fresh fuel, enriched to 80 percent uranium 235, was provided by the Soviet Union in the 1970s and 1980s, and consists of 5,056 fuel elements in sealed aluminum containers.

Vinca also hosts a storage pool containing spent fuel elements that, if reprocessed, could yield more than five kilograms of plutonium. All of this material is under IAEA safeguards.

However, the security of the highly enriched uranium at Vinca remains suspect because of lax physical protection, the potential for theft by criminal or terrorist groups, and the prospect for state-sanctioned diversion or seizure of the material.

The problem of safeguarding fissile material at Vinca was first publicized in 1997 in articles by Mark Hibbs in Nuclear Fuel. 22 Officials at Vinca, he wrote, had expressed concern about the security of the highly enriched uranium in light of growing economic and political unrest in Serbia. This concern was reportedly communicated to U.S. and IAEA officials along with a request for assistance in removing it from the country.

According to Hibbs, the request–which Yugoslav officials deny having made–came after the IAEA made security improvements in mid-1996. The installation of an electronic surveillance system by the IAEA, however, is unlikely to have significantly improved security at a site deficient in modern physical protection, material control, and accountancy measures. (Some Yugoslav observers report that the only visible signs of physical security at Vinca are “one policeman guarding the gate at the entrance and a barbed-wire fence.”) 23

Although the “insider threat” has diminished because of the lack of routine activity involving the highly enriched uranium, Vinca may present an attractive target for terrorists with grievances against the Milosevic regime, heavily armed and well-financed criminal groups, or rogue military or militia elements.

But the most plausible diversion scenario is the one least contemplated by U.S. government officials, who generally remain confident in the adequacy of IAEA and national safeguards. This scenario entails a decision by an isolated Milosevic government to withdraw from the Nuclear Non-Proliferation Tr eaty, sever its safeguards agreement with the IAEA, and revive the nuclear weapons program.

A variant of this scenario, which differs primarily in timing, would involve an effort to divert highly enriched uranium from IAEA safeguards before abandoning the cover of the treaty.

As a consequence of the Serbian campaign in Kosovo and the resulting conflict with NATO, Yugoslavia emerged from the war as an international pariah state. As such, it resembles other past international outcasts such as Israel, Ta iwan, pre-1993 South Africa, North Korea, and Iraq, all of which pursued nuclear weapons.

The more isolated the regime and the more hostile the international environment, the less relevant are global norms regarding weapons of mass destruction. Isolated regimes also are inclined to discount the political costs of violating international taboos.

Further, Yugoslavia's military leadership, who found itself in combat against an overwhelmingly superior NATO adversary last spring, may be looking favorably at nuclear weapons as a future deterrent.

Even without pressure from his own military, Milosevic may well decide there would be broad-based domestic support for an independent nuclear force, the pursuit of which could also divert attention from a variety of domestic and foreign policy failures.

Next steps

Although Yugoslavia's profile increasingly resembles that of past prolifera-tors, the Milosevic regime has not publicly indicated an interest in reviving the Yugoslav nuclear weapons program. Indeed, it continues to receive IAEA inspectors. The country's NPT status, however, remains murky. Belgrade resists formally acceding to the NPT, arguing that it should be accepted as the sole successor to the Socialist Republic of Yugoslavia. (Slovenia, Croatia, Bosnia-Herzegovina, and Macedonia joined the NPT after achieving independence.) This stance prevented Yugoslavia from participating in the 1995 NPT Review and Extension Conference and in the post-1995 NPT review process.

Belgrade's posture does not appear to have produced much anxiety among Western governments, which have not attached high priority to removing the highly enriched uranium from Vinca. Similarly, many non-governmental proliferation experts regard Yugoslavia as an unlikely candidate for nuclear weapons proliferation.

The lack of concern is based upon the current downsized nuclear program at Vinca, the presence of IAEA safeguards, the absence of known international nuclear weapons procurement activities, and confidence that intelligence would detect changes in Yugoslav nuclear policy.

But relatively little weight is given to the experience Yugoslav scientists have acquired in weapons-related research; the potential for Yugoslavia to extend its prior collaboration with Iraq in some military areas to the nuclear sector; possible access to Russian nuclear material, equipment, and technical know-how; and the growing security and political incentives for Yugoslavia to “go nuclear.”

We hope the history of Yugoslavia's covert nuclear weapons activities is well known to the Western intelligence community and prompts the West to reconsider its relaxed approach to Belgrade's nuclear future. But the record of intelligence forecasts of proliferation developments (or, in some instances, slow responsiveness to accurate forecasts) unfortunately does not inspire much confidence. In any event, it would be prudent for the international community to undertake immediate actions.

First, there is an urgent need to remove the fresh weapons-grade material from Yugoslavia. This could best be accomplished by a joint U.S.-Russian initiative, perhaps also involving the IAEA, in which Washington would subsidize the “buy-back” by Moscow of the Soviet-origin fissile material at Vinca.

Moscow seldom has been responsive to this kind of material repatriation, but the Russian Ministry of Atomic Energy (Minatom) repeatedly has indicated to senior Energy Department officials its readiness to participate in the removal of the uranium from Yugoslavia.

But as is often the case, the main obstacle in rapidly seizing this nonprolif-eration opportunity is a bureaucratic dispute among different agencies within the U.S. government as to how much to pay for the operation, which agency would absorb the cost, and whether or not it should be made part of a larger deal to repatriate Soviet-origin material in other countries.

Unless action is taken quickly to overcome bureaucratic inertia and devise a means to implement the buy-back proposal, a golden nonproliferation opportunity will be missed. Already, there are signs that Yugoslav officials, aware of preliminary U.S.-Rus-sian-IAEA deliberations, will make a case for retaining the highly enriched uranium to support the restart of Vin-ca's long mothballed research reactor.

Removing this small but proliferation-significant quantity of uranium from Vinca would be a low-cost, high-return measure that would impede the development of nuclear weapons in Yugoslavia, if Milosevic decided to go that route. But by itself, removal would not prevent Yugoslavia from reviving its nuclear program, which in the past emphasized plutonium.

Indeed, as long as Yugoslavia behaves like–and is treated as–an international pariah, its decision-makers may well conclude that more is to be gained than lost in pursuing a nuclear weapons option.

It will not be easy to alter this decision-making calculus, and resolution of the dilemma may require a change in the Belgrade government. A recognition by the United States of the need to think beyond technical solutions to Yugoslavia's nuclear future, however, is also a necessary condition for crafting a nonproliferation policy with a reasonable prospect for success.

An effective nonproliferation policy should take account of Belgrade's siege mentality, offer meaningful inducements for good behavior as well as penalties for noncompliance, and provide compelling reasons why adherence to a strong international non-proliferation regime is in Yugoslavia's self-interest.