Disclosure of nuclear numbers reveals high accuracy of unauthorized tracking

By Associated Press
Thursday, May 6, 2010; B03

Shortly after nuclear weapons sleuths Tom Cochran and Bill Arkin published their unauthorized estimate of the size of the U.S. arsenal in 1984, they got a call from alarmed U.S. officials.

"They called us over and wanted to know where we got the numbers," Cochran recalls of a time when almost everything about history's deadliest weapon -- including how many the United States possessed -- was classified secret. It was a culture of secrecy born during the Cold War out of a belief that nuclear candor could be dangerous.

America's official nuclear silence ended Monday when the Obama administration not only disclosed the number of U.S. nuclear weapons available for use in wartime -- 5,113 as of Sept. 30 -- but surprised many by also publishing weapons totals for each year dating to 1962. (Data from before 1962 were released in 1993.)

Secretary of State Hillary Rodham Clinton said the disclosures prove that the United States is doing its part for disarmament.

The move also puts pressure on Russia, the world's other main nuclear power, to provide similar accounting as the two sides consider following up their new START arms reduction treaty with a more comprehensive accord.

The U.S. disclosure didn't just strike a blow for transparency in the global debate over nuclear weapons. It revealed the remarkable accuracy of the effort by Cochran, Arkin and others, who labored for 30 years to pull back the veil that surrounded these weapons.

"We were pretty close all along -- sometimes right on the button," said Robert S. Norris, another of the nuclear number detectives.

Their most recent estimate was 5,100, or 13 weapons fewer than the number the government published Monday.

The number reflects the nuclear-tipped missiles and bombs that U.S. forces have available for wartime use. It does not include about 4,500 retired warheads that are in storage awaiting dismantlement.

Many official nuclear secrets remain, although they arguably are secret in name only. It is widely known, for example -- but not officially confirmed -- that building a bomb takes more than eight pounds of plutonium.

The administration has made no public commitment to publishing updated stockpile numbers in the future but said the logic of disclosing the number implies that it will be updated annually.

Norris, author of a biography of Leslie R. Groves, the Army general who led the Manhattan Project that built the world's first atomic bomb in 1945, said the cocoon of secrecy that enveloped the U.S. nuclear program grew out of a belief that the survival of the nation was at stake.

"It was always: 'The Russians are coming. Don't bother us,' " about openness, Norris said.

In the foreword to the government's official account of the bomb project, published after the August 1945 attacks on Hiroshima and Nagasaki, Groves offered a warning for anyone who might feel a need to ask for details beyond those in the official book.

"Persons disclosing or securing additional information by any means whatsoever without authorization are subject to severe penalties under the Espionage Act," Groves wrote.

In 1984, Cochran and Arkin produced the first volume of their encyclopedic Nuclear Weapons Databook, compiled from mining a mountain of responses to Freedom of Information Act requests and from years of scouring congressional hearing transcripts, unclassified government reports, budgets and other open records.

It was, in the view of McGeorge Bundy, who served as national security adviser to President John F. Kennedy, a historic victory for "private clarification" over "government obfuscation." Bundy praised the 340-page work in a New York Times Book Review article in March 1984.

"This volume," he wrote, "contains more facts about the past, present and future of such forces than have ever been put in one place before." No doubt that is why Cochran and Arkin got the phone call from the Energy Department once their book hit the stores.

"They couldn't believe we had the numbers. They thought there was a leak in the system," Cochran said. He and Norris work for the Natural Resources Defense Council, an environmental group.

In 1984, the nuclear detectives estimated that the United States that year had 26,000 weapons -- although they subsequently refined the figure to 23,717.

On Monday, they learned the actual total was 23,459, just 258 off the mark.

At other times, they were farther from the mark.

From 2004 to 2007, their numbers were off by about 2,000 warheads.

Hans Kristensen, who worked with Cochran and Norris, said that he and his cohorts sometimes heard useful "whispers" from certain officials but that for the most part, they relied on simple number crunching and a nose for nuance.

"There was no 'Deep Throat' here, no inside source who has been leaking information," he said.

© 2010 The Washington Post Company