Intelligence Analysts Have Misjudged Nuclear Threats Since Day One, Ex-CIA Official Says

WASHINGTON -- The U.S. intelligence community's track record in assessing nuclear developments abroad is a spotty one and the failures go back to Soviet-watching analysts in 1949, the former CIA deputy director for arms control said yesterday (see GSN, Feb. 28).

"The intelligence community will always fall well short of perfection," Torrey Froscher said during a discussion here at the Monterey Institute of International Studies.

Froscher, who once managed the agency's weapons proliferation analysis, said predictions and estimates have been hampered by analysts' reluctance to rethink assumptions and inadequate amounts of good data to analyze.

The case of prewar Iraq is "obviously the poster child for the limitations of intelligence," Froscher said.

Prior to the first Gulf War, U.S. intelligence underestimated the level of Saddam Hussein's unconventional weapons programs.  It then overestimated the same programs in the walkup to the U.S.-led invasion of 2003 (see GSN, March 12).

"Though the errors were in opposite directions the critical factor arguably in each case was the same:  the absence of enough concrete specific information to dislodge strongly held preconceptions," he said.

The intelligence community was getting it wrong as early as 1949 as the United States sought to determine when the Soviet Union would go nuclear and challenge its status as the only atomic-armed nation, Froscher said.

The best U.S. estimates suggested the first Soviet nuclear test would occur no sooner than a year later, and was most likely four years away.  Instead, the Soviets detonated an atomic weapon in August of that year.

"Why were they so far off?" he asked.

Analysts were both struggling with a paucity of information, what Froscher called "limited tidbits from various sources," and their own belief that the Soviet Union would not have enough uranium ore for some time to test a weapon.

The Soviet Union had a limited indigenous supply of high-grade uranium ore, and the United States and the United Kingdom were both engaged in an attempt to corner the global market on the material -- essentially shutting the Soviets out.

"This was a highly secret effort and it wasn't widely known," he said.  Still, the analysts were aware of the effort and probably let the assumption that it was working affect their judgments, Froscher said.  "Today we might say these analysts were hampered by their preconceptions or mindset about how the Soviets might proceed."

The U.S. intelligence community stumbled again when China conducted its first nuclear test in 1964.  The estimate of the test's timing was pretty good, "but all the specific assessments that underlay that assessment were wrong," Froscher said.

Analysts believed China would detonate a plutonium device because the communist nation was not ready to perform uranium enrichment, he said. Also, both the United States and the Soviet Union had first tested a plutonium weapon.

They were wrong, and China tested a uranium device.  The U.S. analysts had misjudged the progress of the Chinese enrichment facility, Froscher said.  "Why were they so wrong?"

He did note intelligence successes in helping persuade Libya to give up its nuclear program (see GSN, July 27, 2006) and providing information about the nuclear black market associated with Pakistani scientist Abdul Qadeer Khan (see GSN, March 2).

Froscher declined to answer questions about any current issues related to U.S. assessments of foreign weapons programs.  U.S. officials currently peg Iran as being three to eight years away from possessing a nuclear weapon, and recent comments from a U.S. intelligence official downgraded U.S. confidence in the continued existence of a North Korean uranium-based weapon program.

Allegations of the North Korean uranium program led to the collapse of the Clinton-era Agreed Framework in 2002.  North Korea then restarted operations at its Yongbyon nuclear reactor.

What Froscher would say about U.S. assessments is that analysts are forced to make estimates and judgments with the information they have.

"It's a balancing act," he said.   "From the analysts' perspective you can't afford to soft pedal a warning because you're concerned what the impact might be on policy.  But on the other hand you want to make sure all the possible alternative explanations of a particular piece of information are understood as well."

March 14, 2007
About

WASHINGTON -- The U.S. intelligence community's track record in assessing nuclear developments abroad is a spotty one and the failures go back to Soviet-watching analysts in 1949, the former CIA deputy director for arms control said yesterday (see GSN, Feb. 28).