Second of two articles

Until 1998, it was an article of faith for the U.S. intelligence community that no potentially hostile country -- apart from Russia or China -- would pose a long-range missile threat to the United States before 2010, at the earliest.

Scarcely a year later, CIA analysts were saying something entirely different. They predicted that North Korea, one of the world's last surviving hard-line Communist states, could test an intercontinental ballistic missile capable of hitting U.S. territory "at any time." According to a September 1999 intelligence forecast, Iran could test such a missile "in the next few years."

This abrupt shift in thinking was prompted, in part, by a series of troubling events, including missile tests in North Korea and Iran, nuclear tests in India and Pakistan, and reports of Russian scientists selling their services to the highest bidder.

But there is also evidence that the new intelligence forecasts were the result of something else: a concerted campaign by the Republican-dominated Congress, supported by Israel, to focus attention on the leakage of missile technology from Russia to Iran. The government of then-Prime Minister Binyamin Netanyahu feared that Israel could soon become a target of Iranian missiles. Congressional Republicans wanted to build public support for a national missile defense system.

"It was the largest turnaround ever in the history of the [intelligence] agency, and I was part of making it happen," said Rep. Curt Weldon (R-Pa.), a leading critic of what he often called the Clinton administration's "misguided" approach to Russia in the late 1990s. Weldon, a champion of missile defense, was openly scornful of pre-1999 CIA estimates of the missile threat from states such as Iran and North Korea.

Weldon and other conservatives said the intelligence shift was a necessary corrective to what they viewed as politically skewed intelligence forecasts during the Clinton years. They were particularly upset by a 1995 national intelligence estimate that flatly stated that "no country, other than the major declared nuclear powers, will develop or otherwise acquire a ballistic missile in the next 15 years that could threaten the contiguous 48 states and Canada."

By contrast, Joseph Cirincione, director of the nonproliferation program at the Carnegie Endowment for International Peace, said the 1995 intelligence estimate "holds up pretty well in hindsight." He accused Weldon and other Republicans of mounting a "conscious political strategy" to attack the intelligence assessment because "it stood in the way of a passionate belief in missile defense." As a result, he said, the intelligence process has become politicized.

"Intelligence analysts have learned to give the Congress what they want, while preserving the integrity of the analysis," said Cirincione, a former Democratic staffer on Capitol Hill. "What happens is that you get assessments that include all possible worst cases."

CIA officials argue that the post-1998 estimates are the result of "improved tradecraft." They say the agency reviewed its procedures following publication of a 1998 report on the ballistic missile threat by a bipartisan commission headed by a former (and future) defense secretary, Donald H. Rumsfeld, and began to consult a wide range of independent experts from industry and academia.

Some consumers of intelligence within the government say the shifting forecasts of the ballistic missile threat are a case study of how an ostensibly objective intelligence process can be buffeted by conflicting political pressures, from home and abroad.

"Nobody believes the CIA estimates," said a longtime counter-proliferation expert from another government department. Another analyst said that "nuances" tend to get taken out of the estimates as they proceed up the bureaucratic ladder. "The job of the CIA is to warn, but they never back down from previous warnings," the analyst said.

A Congressman's Battle

The argument over the 1995 intelligence estimate got underway even before its publication. According to Capitol Hill sources, the Clinton administration leaked details of the still-secret document to congressional Democrats, who used it to argue the case against missile defense.

As chairman of the House Armed Services Committee's subcommittee on military research and development, Weldon assumed responsibility for countering the Democratic offensive. He did so by staging a dramatic showdown with a CIA analyst, David Osias, who had been dispatched to Capitol Hill to give him and other committee members a briefing on the intelligence finding in a secure fourth-floor conference room.

Weldon said he "went ballistic" after Osias insisted that there would be no hostile missile threat to the continental United States for at least 15 years. "I said, 'Do you mean to tell me that the unrest in Russia represents no additional threat?' " The congressman said he was also furious that the CIA study excluded Alaska and Hawaii from its threat assessment. (A North Korean missile would have to travel nearly 6,000 miles to hit California, but only 3,700 miles to hit Alaska.)

"This is over, this is [expletive], this is a politicized process," Weldon recalled yelling, before bringing down the gavel on the closed-door session. Intelligence sources confirmed that Osias was subjected to a severe grilling at the secret hearing.

The debate over the 1995 estimate coincided with an aggressive Israeli campaign to alert the Clinton administration to what Netanyahu advisers saw as a growing missile threat from Iran, a radical Islamic state that has often threatened to destroy Israel. Israel had information that Iran was working on a scaled-up Soviet Scud missile, known as the Shahab-3, that would theoretically be able to hit Tel Aviv from launching pads in western Iran.

Israel had intelligence that Russian missile experts were traveling to Tehran and giving advice to the Iranians. Former Israeli officials said they were greeted with skepticism from Clinton administration officials who were reluctant to strain relations with Russian President Boris Yeltsin, then seen in Washington as the symbol of Moscow's fledgling democracy.

"It was as if the Americans did not want to know the facts, or the facts were too embarrassing for them to confront," said Uzi Arad, a former intelligence adviser to Netanyahu.

The Israeli allegations of technology transfers between Moscow and Tehran became the basis of a series of congressional hearings in 1997 and 1998, and Republican calls for economic sanctions against any state that provided missile technology to Iran. The confrontation came to a head in June 1998 when the Republican-dominated Congress passed the Iran Missile Proliferation Sanctions Act, which would have imposed mandatory sanctions on any country selling missile technology to Iran. The legislation was promptly vetoed by President Bill Clinton.

Administration officials scrambled to enlist Israeli support to get Congress to back down and accept a diluted version of the legislation, rather than override the president's veto.

"The administration's rationale was that it was up to the Israelis to get the genie back into the bottle since they had let it out in the first place," said a Republican staffer, describing how Israel persuaded House Speaker Newt Gingrich (R-Ga.) to go along with White House wishes.

Against their own judgment, congressional conservatives allowed the presidential veto to stand. But they would soon acquire fresh ammunition in their campaign against the 1995 intelligence estimate.

'Mass Conversion' at CIA

The first serious attempt by congressional Republicans to persuade the CIA to revise its estimate of the long-range missile threat ended in failure. A blue-ribbon panel headed by former CIA director Robert M. Gates reported to Congress in December 1996 that the technical case against "rogue states" acquiring intercontinental ballistic missiles, or ICBMs, in the foreseeable future was even "stronger" than that presented in 1995.

Unhappy with the conclusions of the Gates committee, Congress appointed a new commission, this one headed by Rumsfeld. The Rumsfeld report -- delivered in July 1998 -- turned out to be much more alarmist than the 1995 estimate. The Rumsfeld report predicted that a rogue state would be able to "inflict major destruction" on the United States "within about five years" of a decision to develop an ICBM. For several of those years, the report added, "the U.S. might not be aware that such a decision had been made."

According to commission members, the five-year estimate was based largely on briefings from missile engineers at major U.S. defense contractors, including Lockheed Martin and Boeing. The commission asked the American rocket builders how long it would take them to build an ICBM, from the starting point of a Third World country such as Iran. "The answer was five years or less than five years," recalled Barry Blechman, chairman of the Henry L. Stimson Center, a research organization in Washington.

Rumsfeld accused the CIA of committing the sin of "mirror-imaging" in its earlier estimates: the notion that "just because it took us 10, 12 years to do something, it is likely to take others that long or longer." In fact, he insisted, it would probably take other countries less time to develop ICBMs, as much of the relevant information was already available.

By framing the question to U.S. missile experts in a novel way, the Rumsfeld Commission avoided the mistake of assuming that a country such as North Korea would necessarily follow the same path to an ICBM as the United States or Russia. But critics argue that the commission might have fallen into a mirror-imaging trap of its own: the assumption that an isolated Third World country has the same access to missile components and missile technology as a major U.S. defense contractor.

"I don't believe that the Rumsfeld Commission made a serious analysis of the industrial base needed to develop long-range missiles," said Theodore A. Postol, a professor of missile technology and national security policy at the Massachusetts Institute of Technology. "Of course, American contractors will tell you that building an ICBM is easy. But these are people who live in an incredibly rich industrial environment. A Third World country faces an entirely different set of problems."

The political impact of the Rumsfeld report was strengthened by the fact that its conclusions were unanimous. The Democrats had been allowed to appoint three members of the nine-member panel, so it was difficult for them to argue that the report was politically tainted. The Democrats' experts included Blechman and Richard L. Garwin, a leading nuclear physicist strongly opposed to missile defense on technical and scientific grounds.

Garwin and Blechman said they were struck by the way in which countries such as North Korea, Iran and Pakistan were pooling their resources and taking advantage of existing know-how. Since the beginning of the Bush administration last year, and Rumsfeld's reappointment as defense secretary, the conclusions of the Rumsfeld Commission have been elevated to quasi-doctrinal status within the government, according to several officials.

"Nobody dares say a word against Rumsfeld, at least in public," said one government nonproliferation expert. Another spoke of a "mass conversion" within the CIA, even among analysts who were predicting something entirely different just a few years before.

Even so, the Rumsfeld Commission's conclusions remain highly controversial, even within the government. The State Department's intelligence unit, the Bureau of Intelligence and Research, has long taken a less alarmist view of North Korean and Iranian capabilities than has the CIA or the Pentagon. A new national intelligence estimate on the missile threat, issued this month, publicly enshrined the State Department's dissenting views for the first time, officials said, even though the declassified version referred only to an unnamed "agency."

The new estimate also acknowledged what outside experts have long maintained: Rogue states or terrorist groups are unlikely to use missiles as their method of choice for delivering weapons of mass destruction. The estimate said that "covert delivery methods," such as a ship or a civilian airplane, were cheaper and more reliable than ballistic missiles.

The idea that a country such as Iran or even Libya could be well on its way to deploying an ICBM -- as opposed to a short- or medium-range missile -- without the United States knowing about it strikes many outside experts as absurd.

"Iran is trying to achieve a credible regional capability," said a French Foreign Ministry official. "A more long-range program is a matter of speculation. They say they want a satellite launch capability, but we don't see them putting much effort into it."

Rumsfeld Commission members continue to defend their conclusions. "We never said anything about states deploying ICBMs by a certain date, we just said they would be capable of doing it," said Blechman, who notes that the U.S. intelligence community failed to predict a huge Soviet nuclear arms buildup in the 1960s under Leonid Brezhnev.

'Golf Ball of Destruction'

The Rumsfeld Commission report led U.S. intelligence experts to reconsider the very nature of an intercontinental-range ballistic missile. Before 1998, they had thought of ICBMs as similar to the weapons possessed by the United States and the Soviet Union: sophisticated, powerful and highly accurate missiles that could be hidden in silos and launched at a moment's notice.

The new definition of an ICBM, as conceived by the Rumsfeld Commission and embraced by the CIA, covers virtually any rocket capable of landing a warhead, however small, somewhere on U.S. soil, at least in theory. Under the new definition, North Korea already has ICBM capability against the United States because it possesses a rocket that could, conceivably, land a tiny warhead somewhere in Alaska.

The North Korean Taepodong-1 is a three-stage rocket, tested for the first (and so far only) time on Aug. 31, 1998, soon after the release of the Rumsfeld report. According to U.S. intelligence officials, the first stage was a No Dong, the North Korean version of a scaled-up Scud B. The second stage was the North Korean equivalent of a Scud B. The third stage consisted of a small, solid-fuel rocket probably acquired from Pakistan or China, carrying little more than a radio transmitter.

This unwieldy contraption was the missile equivalent of the "Hail Mary pass," according to David Wright, a senior analyst for the Union of Concerned Scientists. Although the third stage exploded and the missile flew no more than 1,000 miles, the launch demonstrated that North Korea might soon have the capability of putting a satellite into orbit.

The elastic nature of what exactly constitutes an ICBM has caused some skeptics within the government to joke about what they call "the golf ball of destruction."

"There is an idea out there that if you can land anything on American territory, the result will be vast devastation. That is simply not true," said a government expert at odds with the official CIA line.

The skeptics argue that the North Koreans have a long way to go before their missiles pose a real threat to the United States. First, they have to develop a rocket that actually works. Second, they need a warhead that will not burn up when it reenters Earth's atmosphere. Third, they have to develop rockets powerful enough to deliver a militarily significant payload. And fourth, they need to mate the missile to a nuclear or biological warhead.

Many experts believe that a chemical or biological attack on the United States using a crude Taepodong-type rocket can be excluded because such weapons must be delivered with a high degree of precision to be effective. Nuclear warheads can be less accurate, but are much heavier than chemical or biological weapons, meaning that they would require more powerful rockets to reach the United States.

According to CIA estimates, a two-stage Taepodong-2, which is now under development by North Korea, could deliver a payload of several hundred kilograms to Alaska or Hawaii. This would probably be sufficient to deliver a biological warhead, but not enough for an unsophisticated nuclear weapon.

A larger question is whether 1950s Soviet Scud technology, of the kind now widely available in the Third World, can serve as a basic building block for ICBMs. Russian scientists cite their own experience in arguing that a Scud cannot be upgraded to an ICBM. To develop missiles that could reach the United States, the Soviets moved to systems with more powerful propellants and vastly improved guidance systems.

"There are certain things you can do to improve the Scud, such as lightening the airframe, installing new turbopumps and clustering engines, but you quickly run into limitations," said Timur Kadyshev, a missile expert at the Moscow Institute for Physics and Technology. "At some point, you need to switch to better technology."

Until recently, this was also the view of most CIA experts. Before 1998, CIA officials routinely argued that North Korea would have to adopt an entirely new propulsion system to achieve ICBM capability, the development of which could easily be detected by U.S. technical means. The agency's position now is that a similar result can be achieved by clustering engines and adding extra stages.

Russian experts say that, while this may be feasible in theory, the addition of each new engine increases the chances of failure. A two-stage Taepodong-2, for example, is believed to consist of four No Dong engines clustered together as the first stage, and a single No Dong as the second stage. Mathematically, such a missile is at least five times as likely to fail as a single, far-from-reliable, No Dong.

Another significant change in CIA methodology has been the abandonment of the long-held view that a lengthy testing period was required before a new missile system could be considered a real threat. According to Robert Walpole, the national intelligence officer responsible for coordinating estimates of missile threats, the willingness of Third World countries to resort to nuclear blackmail has made the accepted wisdom of a five-year gap between testing and deployment obsolete.

When the United States and the Soviet Union deployed missiles, he explained, "they had to be in hard silos so that the other guy could not take them out. But if what you are more interested in doing is threatening the other side, not having a retaliatory launch capability, you don't have to deploy [missiles] in that sense of the term." Many independent experts say they believe that repeated tests are required before a missile can be deployed.

As military weapons, the Taepodong-1 and Taepodong-2 clearly leave much to be desired. Before the rockets can be launched, they have to be assembled next to tall open-air gantries, in full view of U.S. spy satellites and planes flying off North Korea's coast. The process of fueling and completing final checks takes three or four days, according to Charles P. Vick, a missile expert at the Federation of American Scientists.

As a political and propaganda weapon, however, the Taepodong-1 has already proved very effective. The North Koreans "got everybody's attention" with their August 1998 missile test, said Joseph S. Bermudez, a leading expert on North Korean missile programs. "They made America wake up and pay attention to them, which is one of the things they desperately want. They want to be perceived as a powerful nation."

Rep. Curt Weldon (R-Pa.), a missile defense proponent, was skeptical of CIA missile threat assessments.Donald H. Rumsfeld, now defense secretary, led a commission that in 1998 warned of greater danger.