Judgment Calls   

How top editors decide whether to publish national security stories based on classified information

The outcry over decisions by major newspapers to disclose the Bush administration's secret monitoring of international banking transactions was fast and furious.

Although the Wall Street Journal, the Los Angeles Times and the Washington Post also published articles describing efforts to trace the financial records of suspected terrorists, the New York Times broke the story on the Web and bore the brunt of the outrage. The administration had asked the New York Times and L.A. Times not to publish. But both papers ultimately decided to anyway, posting their pieces the evening of June 22 and publishing them on page one the following day.

The clash between the First Amendment's guarantee of a free press and the patriotic duty to protect American lives and uphold national security puts the media in an uncomfortable position. For the second time in six months, the New York Times had infuriated the administration by exposing a secret program in the war on terror. The piece followed a December 16 story disclosing the National Security Agency's warrantless eavesdropping inside the United States. In that instance, too, the administration had pleaded with the Times to withhold publication. But after delaying for more than a year to conduct additional reporting, the Times published the article--and won a Pulitzer Prize for it.

On June 26, President Bush condemned the global banking records story. "Congress was briefed," he said, answering questions from reporters. "And what we did was fully authorized under the law. And the disclosure of this program is disgraceful. We're at war with a bunch of people who want to hurt the United States of America, and for people to leak that program, and for a newspaper to publish it, does great harm to the United States of America."

Vice President Dick Cheney weighed in more pointedly the next day. "Some in the press, in particular the New York Times, have made the job of defending against further terrorist attacks more difficult by insisting on publishing detailed information about vital national security programs," Cheney said in a June 27 speech. He called the Times' Pulitzer for its NSA story "a disgrace."

Those statements were positively temperate compared with the reaction among administration allies. Sen. Jim Bunning, a Kentucky Republican, accused the New York Times of "treason." Rep. Peter King, a New York Republican who chairs the House Committee on Homeland Security, asked the U.S. attorney general to launch a criminal investigation of the paper. On June 29, the House of Representatives voted 227-to-183 to condemn the publication of classified information and to urge news organizations' cooperation in the war on terror. In early August, Sen. Kit Bond, a Missouri Republican, introduced a bill that would criminalize the unauthorized disclosure of classified information.

The punditry was even more vociferous. On June 28, San Francisco talk show host Melanie Morgan told the San Francisco Chronicle that New York Times Executive Editor Bill Keller should be jailed for treason for approving publication of the banking records story. "If he were to be tried and convicted of treason, yes, I would have no problem with him being sent to the gas chamber," Morgan told the Chronicle. "It is about revealing classified secrets in the time of war. And the media has got to take responsibility for revealing classified information that is putting American lives at risk."

In all likelihood, the most biting media retort to the sorts of accusations leveled by Morgan and others came 11 years earlier, in the 1995 autobiography of Ben Bradlee. In "A Good Life," the former Washington Post executive editor wryly observed: "Editors--and reporters, and especially owners--don't like to be accused of giving aid and comfort to the enemy, even when they know it not to be true. It riles the kooks and the woolly hats of this world, and results in a great deal of ill-tempered and unnecessary correspondence."

In his inimitable style, Bradlee fired back at his critics. ("Dear Asshole," he began in response to a letter labeling him "UnAmerican.") But in this gentler age of transparency and accountability, not to mention heightened public contempt for the media, editors have largely swallowed their frustration. They have chosen to answer passionate--often virulent--denunciations by explaining the painstaking path that national security stories travel to publication.

Keller and Los Angeles Times Editor Dean Baquet took the unusual step of publishing a joint op-ed piece illuminating their decisions on the banking records stories. "There is no magic formula, no neat metric for either the public's interest or the dangers of publishing sensitive information," the editors wrote. "We make our best judgment."

In interviews with AJR, top editors at the newspapers involved in recent high-profile skirmishes, as well as other authorities on national security reporting, explained how stories based on classified information are reported and how editors decide whether to publish in the face of administration objections. (AJR also requested interviews with Cheney and White House Press Secretary Tony Snow; both declined.)

The editors described a deliberative process guided by intense questioning, a willingness to pull back on technical details that could endanger lives and, ultimately, a commitment to informing the public about the government's use of power.

As Keller and Baquet wrote in their joint piece, "The process begins with reporting. Sensitive stories do not fall into our hands. They may begin with a tip from a source who has a grievance or a guilty conscience, but those tips are just the beginning of long, painstaking work."

Washington Post Executive Editor Leonard Downie Jr. says national security articles seldom develop the way the public tends to assume: Not since Daniel Ellsberg leaked the Pentagon Papers to the New York Times in 1971 can he recall an instance in which a single person supplied an entire story to a journalist. Instead, a beat reporter such as the Post's Dana Priest, who won a Pulitzer for her November story exposing the CIA's secret prison system of "black sites" for hiding and interrogating some al Qaeda captives, "gradually unearths the outlines of a story."

Keller concurs that there's a "sort of mythology that some disgruntled bureaucrat calls a reporter cold and says, 'Meet me in a parking garage, and I will hand over a box of files on this secret program.' It pretty much never happens that way." Reporters may hear something that's "the most shadowy notion of what's going on" and stitch information together from multiple places. "That's part of the puzzle you struggle with when you're deciding whether to publish," he says. "Do you trust those people to tell you when it's safe to publish when they may not be in a position to really know?"

As a reporter on these stories, "you want to know everything you can know because even if you withhold certain details, you're going to write a better story if you have all of that in your head," says Steve Coll, a staff writer for The New Yorker, former Washington Post managing editor and author of the 2005 Pulitzer Prize-winning book "Ghost Wars" detailing the secret history of the CIA's role in Afghanistan. "You're in the fog all the time, always groping..you're always concerned that you're missing something."

Occasionally reporters "get lucky, and you have sources that are sitting on top of what you're interested in and allow you to see the whole of it," Coll says. "But often sources only have partial access, or even if they have fuller access, they're only willing to talk about part of what they know... Particularly when the subject matter is surrounded by both classification concerns and political debate, it's as treacherous as it gets. The only way to attack that problem is through reporting, and it means also trying to surround a subject from as many angles as possible."

Often that means seeking viewpoints from foreign as well as domestic sources. "Even if I have great American sources, I want non-American perspectives to make sure there isn't some angle of vision that I'm failing to internalize," Coll says. But even that does not always save a reporter from errors: In the run-up to the war in Iraq, Britain and France also believed in the existence of weapons of mass destruction. The United Nations, however, "did have it more or less right. To the extent that any American paper did document the limits of WMD..it was clear that did come from the U.N."

As a reporter works to confirm information, an initial tip can unravel partially or entirely. On June 13, the L.A. Times' Washington bureau was told about the banking records by a source who said, "'Here's the broad outlines of the story, and it's competitive, the New York Times is working on the story,'" recounts Douglas Frantz, the L.A. Times managing editor who oversees newsgathering. Some of that initial guidance turned out to be wrong. The tipster said the FBI was using administrative subpoenas to go to the Belgium-based cooperative known as Swift and gain access to the records. "As it turned out, it was in fact the Treasury Department."

Reporters are trying to pry out sensitive information during a period of intense anxiety. "I haven't heard direct threats, but there are occasional reminders from administration officials that, in their view, the Espionage Act may apply to newspapers that publish classified information of any kind," says Doyle McManus, the L.A. Times Washington bureau chief. Those warnings, combined with high-profile setbacks for journalists' ability to protect sources in federal courts, do not intimidate reporters, McManus says. "It does make them worry about the well-being of their sources. They are hearing this concern from their sources, and they are taking precautions to try to protect their sources, keeping sources' names out of written notes, for example, taking more care with telephone or e-mail communications."

As reporters dig, they sift through often-intriguing details about particular programs. Investigating the secret prisons story, the Post's Priest "came across a lot of information about covert antiterrorism projects that right from the outset it was clear to us that details here and there would not be important to readers" but would either be "injurious to Americans potentially or could damage these programs potentially," Downie says. "As we gradually shape what we think the story's going to be, most often senior officials will talk to the reporter about concerns and ask them to raise them with us."

Downie says government officials are involved "very, very early on" in the process, "particularly officials that are doing their jobs well. Reporters are talking to them, asking them questions. They can see stories taking shape." Sometimes an official raises a specific concern that may resolve itself as the reporting proceeds, if, for example, a technical detail about an operation turns out to have little relevance to the story or is clearly of interest only to someone who wants to harm the country. Government officials also may drop a request to withhold publication.

Especially with national security stories, the Post "sort of had a standing rule that we wanted to hear about concerns of these kinds if they were raised during the reporting," Coll says. The editors first tried to "understand what [the officials] were worried about, or if it was a concern they were voicing on their own. Then we'd ask the reporter to tell whoever was raising the question to kick it up their chain and determine whether the government side did have a concern or somebody down in the middle ranks was being overly cautious... You sort of challenge them" by asking, "'Are you speaking for your boss? And if you are speaking for your boss, would you mind putting your boss on the phone at some point?'"

If publication concerns reflect the government's position, there's a "tacit protocol," Keller says. "It's one thing for the flack at the department to tell the reporter, 'You shouldn't publish this,' but if the administration is really serious, somebody fairly senior on their side contacts somebody fairly senior on our side."

On rare occasions, the president himself weighs in. Keller, Times Publisher Arthur Sulzberger Jr. and Washington bureau chief Philip Taubman met with Bush in the Oval Office December 5 about the NSA eavesdropping story--the only time in Keller's experience that a president has personally asked the Times not to publish a story.

Sometimes senior editors and government officials meet alone, although Keller, Downie and Baquet all push for the inclusion of the lead reporters in off-the-record meetings with government officials. "I don't like this idea that somehow one head of something calls another head of something, and the reporter whose job it is to write the story is not included in the conversation," Baquet says. "I like the reporter in the room. I think that's better for everybody. They know the story better. They know the nuance."

Keller says having reporters present often disconcerts administration officials. They are "concerned that even if it's off the record, it's very hard for reporters to resist the urge to shape the story based on what they've heard: 'Gee, I had it a little bit off, maybe I better go fix it.'" Asked if he would change a story if he learned of inaccuracies, Keller replies, "You've given your word of honor that what you hear in these briefings will be off the record, and we've tried to abide by that. Speaking hypothetically, if it led me to believe that an element of a story was wrong, I would leave that out, but I don't think I'd feel it was fair to insert the correct version without going back to the administration."

Reporters brief editors carefully before the meetings. If they cannot attend, "We talk to the reporter about everything that took place," Downie says. He also consults the Post's general counsel, although it's rare for the paper's lawyers to attend the meetings.

Downie tries to evaluate whether the information would really endanger lives, or whether officials are seeking to withhold publication for reasons of policy, partisanship or embarrassment. He says most senior government officials comprehend the media's accountability role. "We understand the territory in which we're talking. We're usually very specific" about the story's content, as are government officials in describing what they see as the dangers of publication.

When senior administration officials raised concerns about the consequences of naming Eastern European countries participating in the secret prisons program, "We listened carefully," Downie says. The story's purpose was "accomplished without having to name the countries. Very, very seldom do we decide not to publish a story at all. But quite often we will leave out specific details, technical details, location details that would put lives or programs in jeopardy unnecessarily."

R. James Woolsey, President Clinton's CIA director from 1993 to 1995, twice approached senior media leaders "because a particular fact that one of their reporters had been asking about, if revealed, would have seriously put at risk a source or a method" of intelligence gathering. On one occasion he brought an expert to help make his case; the other time he went by himself to explain the problem privately. "In each case, they said the story doesn't depend on this fact, and thanks for letting us know, and they ran the story without the fact," Woolsey says.

Keller weighs questions such as: How fully do the journalists trust their sources? How far do they trust the government officials urging them not to publish? How secret is the secret? Can the story be told in a way that minimizes the risk? "'Explain to me why publication of the story in your view will jeopardize national security,'" he asks government officials, and "you question them, and then you question them some more. Usually by that point, reporters have a pretty good idea what the arguments are on both sides. Sometimes you do more reporting."

That's what happened with the NSA wiretapping article. "After meeting with senior administration officials to hear their concerns, the newspaper delayed publication for a year to conduct additional reporting," James Risen and Eric Lichtblau wrote in their December 16 article. (That time frame was somewhat imprecise; on August 13, Times Public Editor Byron Calame quoted Keller as saying internal discussions about the story had been "dragging on for weeks" before the November 2, 2004, election.)

As his reporters dug deeper, Keller says it became clear to him that the story centered not on methods of intelligence gathering but on the "use and alleged abuse of executive power. Between 2004 and 2005, the whole kind of framework of the story changed. Written as a story about executive power, it was much easier to focus on the question of legality, the question of the FISA [Foreign Intelligence Surveillance Act] Court and so on, as opposed to the technical details of the program. It wasn't just that we left technical details out, it was that the framework and the whole thrust of the story had shifted. If you read a draft of the story and what we published, you would see not just that we shaved out specifics, but it had become less of a kind of whiz-bang spy technology story, but more a story about warrants or the absence of warrants."

Additional sources provided a more nuanced understanding of objections to the government's eavesdropping. "We knew there were people who were aware of the program who thought it was illegal," Keller says. "We were not aware those misgivings existed in all three branches of government and that they existed in the minds of people whose motives were nonpartisan and impartial, people who could not be characterized as anti-Bush or anti-war-on-terror. We got some additional witnesses who had a lot of credibility."

Regarding the Swift financial records story, Keller says the administration essentially made two arguments about the central dangers to security. Then-Treasury Secretary John Snow emphasized that disclosure would cause international banking executives to withdraw their support of the program. Officials also argued it would alert the terrorists and prompt them to modify their behavior.

"We tried as best we could to report out those two claims," Keller says. "In the case of the embarrassment argument that it will spook the bankers, on the other side was the fact that this monitoring was taking place under a subpoena, and they had what they regarded as clear legal authority, an effective program and safeguards including an outside auditing firm. It seemed to us that if those things were true, there was unlikely to be any pressure on the bankers to withdraw their support."

In a June 25 letter that Keller sent to readers who had written him about the story, he said the second argument "was made in a half-hearted way. It has been widely reported--indeed, trumpeted by the Treasury Department--that the U.S. makes every effort to track international financing of terror. Terror financiers know this, which is why they have already moved as much as they can to cruder methods. But they also continue to use the international banking system, because it is immeasurably more efficient than toting suitcases of cash."

In his own letter to Keller the next day, Snow retorted: "The fact that your editors believe themselves to be qualified to assess how terrorists are moving money betrays a breathtaking arrogance and a deep misunderstanding of this program and how it works." By citing "public interest" as a reason for publication, Snow wrote, "the paper has given itself free license to expose any covert activity that it happens to learn of--even those that are legally grounded, responsibly administered, independently overseen, and highly effective."

Whenever the government makes a request to withhold information, "we go into the discussion feeling like there's got to be a powerful reason not to publish," the L.A. Times' Baquet says. "If the government has a compelling case that a life is in danger, that's an easy call. You don't publish." But if "the government's going to cite the war on terror every time a newspaper's going to publish a secret," Baquet says, that "makes us more skeptical."

The Bush administration has aggressively invoked national security in justifying secrecy and classification of information. On August 18, the National Security Archive, a nonprofit research library based at George Washington University, reported the administration had begun to classify long-available numbers of U.S. nuclear missiles during the Cold War, blacking out information on previously public documents.

"You do your best to check it out," Baquet says of national security claims. "You do all the reporting you can, and then you have to make a judgment, and it's really difficult." But he adds: "Our overriding job, the most important job we have, is to cover the government and its use of power and authority."

In their joint piece, Baquet and Keller quoted from Justice Hugo Black's concurrence in the U.S. Supreme Court's June 30, 1971, Pentagon Papers decision: "The Government's power to censor the press was abolished so that the press would remain forever free to censure the Government," Black wrote. "The press was protected so that it could bare the secrets of the government and inform the people."

The 6-to-3 Pentagon Papers decision was a watershed moment in media law because justices prohibited the government from suppressing the secret Vietnam War history. As that case illustrates so dramatically, clashes between the government and the press over printing national security secrets are long-standing.

Mark Feldstein, director of the journalism program at George Washington University, posits what he calls a "slightly exaggerated" theory about publishing such secrets: Historically, mainstream papers have been reluctant to print classified information and criticize the government when an administration is popular.

When President Johnson declared in August 1964 that North Vietnam had attacked U.S. destroyers in the Gulf of Tonkin, "there was very little scrutiny at the time of what later proved to be fairly bogus information, but as the Vietnam War's popularity declined and the public turned against it, there was more critical reporting," says Feldstein, who teaches media history. In the Iraq war, "it was only as we started getting bogged down in the war that you started seeing more critical reporting, more disclosure in the way of classified information."

Feldstein believes the press, like other institutions, is influenced by political currents. "It's easier now for the Times and the Post to take on the Bush administration when its poll numbers are down and it's not as popular. You are seeing the press emboldened here now, not just because of the heroism of the media, although those are great stories and they should be commended for doing it, and not just because there are more dissidents in the government, but because Bush is weaker, and it's easier to take him on than after September 11, when he was riding high."

One frequently cited case of national security damage inflicted by the press is the Chicago Tribune's June 7, 1942, front-page story on the Battle of Midway, headlined "Navy Had Word of Jap Plan to Strike At Sea." Such knowledge could only have been gleaned if the U.S. had broken Japanese naval code, a fiercely protected secret. President Roosevelt threatened to try Col. Robert McCormick, the isolationist Tribune publisher, for treason. A grand jury was convened, but the Navy, fearful that a trial would heighten publicity, did not cooperate, and the grand jury refused to indict. The Japanese, who didn't change their code, apparently missed the Tribune's blunder.

A more recent case of purported harm involves news media reports about Osama bin Laden's use of a satellite phone. In December, President Bush asserted that publication of a government leak alerted bin Laden to U.S. monitoring of his conversations, causing him to discard the phone. That claim, which referred to an August 21, 1998, Washington Times story, was made first by two former Clinton administration officials in a 2002 book, "The Age of Sacred Terror," validated by the 9/11 commission and repeated by Bush, who chastised the media for "revealing sources, methods and what we use the information for."

But the leak story "appears to be an urban myth," the Washington Post's Glenn Kessler wrote in a December 22 story. "The al Qaeda leader's communication to aides via satellite phone had already been reported in 1996--and the source of the information was another government, the Taliban, which ruled Afghanistan at the time. The second time a news organization reported on the satellite phone, the source was bin Laden himself... It was not until Sept. 7, 1998--after bin Laden apparently stopped using his phone--that a newspaper reported that the United States had intercepted his phone calls and obtained his voiceprint."

Thomas Powers, an intelligence expert and author most recently of "Intelligence Wars: American Secret History From Hitler to al-Qaeda," says, "You could ransack the literature and not come up with three examples" where publication has harmed national security. He calls such claims "wildly overblown," adding, "No government since World War II has demonstrated that it can be trusted in that area. They constantly and repeatedly hide behind national security arguments in order to do dumb or doubtful things."

David Wise, a Washington-based author who has written extensively on the intelligence community, agrees: "It's a phony business, the whole secrecy racket. It's a racket designed to allow political leaders to maintain information control."

Wise's work includes "The Invisible Government," a groundbreaking 1964 examination of the intelligence agencies that he coauthored with Thomas B. Ross. He says there are rare instances when publishing might jeopardize lives, and cites the 1931 Supreme Court case Near vs. Minnesota in cautioning against disclosure of sailing dates of troopships.

But only twice in his 48-year career has Wise agreed to withhold publication at the government's request. In 1961, he was the White House correspondent for the now-defunct New York Herald Tribune. He learned that President Kennedy and Soviet Premier Nikita Khrushchev were about to announce the release of the two surviving crew members imprisoned after their Air Force RB-47 reconnaissance plane was shot down over the Barents Sea. On Kennedy's behalf, White House Press Secretary Pierre Salinger asked Wise not to publish the story because he said premature disclosure might jeopardize the deal.

"Because human life was involved, I recommended to my editors that we not break the news," Wise recalls. "It was a time of great tension between the Soviet Union and the United States. I was a 31-year-old reporter, and I'm thinking, 'Do I want to blow up the world tonight?' More than that, I didn't want these guys to rot in prison for years to come. We didn't get a Pulitzer Prize. We got a nice telegram to my publisher from Kennedy."

The second incident occurred when Wise and Ross were researching their 1962 book, "The U-2 Affair," about the Soviet capture of U.S. pilot Francis Gary Powers. They learned that U-2 spy planes were flying over Cuba. Attorney General Robert F. Kennedy asked them not to use that and they complied, in part because it wasn't relevant to their narrative. "Later, when [a] U-2 spotted the missiles in Cuba, we were glad we had not printed that fact," Wise says.

In general, though, it's "the job of the press to print the story if they get the story," he says. "If something is leaked at a high level as a matter of policy or with the approval of senior officials, how can they then turn around and say you can't publish a secret if you find out on your own? It's a double standard." He praises the recent national security stories, adding, "I'm glad that some of them have been recognized with Pulitzer Prizes."

Bradlee, now the Post's vice president at large, says of government warnings, "National security sounds so awesome. But the things that they call national security overwhelmingly have nothing to do with national security." Because the possibility exists, though, editors owe government officials a fair hearing and careful consideration of their arguments. "It's terrible bad PR to just tell them to go screw, excuse the expression," Bradlee says. "You've got not only to appear to do it, but do it."

He did so at length with "Ivy Bells," a U.S. operation to intercept Soviet underwater cable communications. Post reporter Bob Woodward learned of the interception in 1985, but Bradlee refused to print anything until 1986, after the disclosure that the operation had been betrayed to the Soviets by NSA analyst Ronald Pelton. At that point, if "the Soviets knew all about Ivy Bells, why shouldn't the American public know about it?" Bradlee wrote in his autobiography. The story detailed "an intelligence triumph in the operation itself, and an intelligence disaster in Pelton's defection."

Bradlee had roughly 20 conversations with top intelligence officials before the story went to press. President Reagan called Post CEO Katharine Graham, and CIA Director William Casey threatened prosecution if the Post proceeded. As Bradlee deliberated, even showing officials various drafts of the piece, NBC News broke the story. Recounting the lessons of the case, Bradlee wrote, "Once it was certain that the Russians knew everything about Ivy Bells, there was no issue of national security."

Bradlee says he "wouldn't rethink anything" regarding publication of the recent national security disclosures. But Gabriel Schoenfeld, a senior editor at Commentary magazine, draws a distinction between Bradlee's handling of Ivy Bells and the New York Times' approach to the NSA and Swift stories. "They did not go ahead with that story," Schoenfeld says of the Post. "They did listen to what the administration had to say. I think generally they behaved more responsibly than the Times."

Schoenfeld wrote a March piece asking, "Has the New York Times Violated the Espionage Act?" and believes the Times should be prosecuted for publishing its NSA piece. "I think there are obvious costs and it would be in many ways shocking, but I think it's very important to establish that we have rule of law in this country and newspapers are not above it," he says, adding that editors should not be the final arbiters of whether to disclose classified information. "This is a private corporation. We wouldn't want Halliburton to play this role, and we don't want the New York Times to play it. Too much is at stake." While Schoenfeld doesn't see the Swift story as a clear legal violation, he calls it "an utterly reckless act."

Still, Schoenfeld says that national security stories involve tough calls, and disclosure of classified information occasionally could serve the public interest. He wonders "if the courts ultimately find that the NSA program is illegal, how badly would that undermine my argument? I mean, obviously, the Times played a crucial role in bringing this to light." On August 17, a federal judge ruled the program unconstitutional; the Justice Department immediately appealed.

In retrospect, government officials at times have regretted keeping sensitive information out of print. The New York Times downplayed an advance story about the 1961 Bay of Pigs invasion because of national security concerns. After the disastrous raid, President Kennedy publicly lambasted the press for "indiscriminate and premature reporting," Susan E. Tifft and Alex S. Jones recount in "The Trust: The Private and Powerful Family Behind The New York Times." But Kennedy privately told the Times' managing editor, "Maybe if you had printed more about the operation, you would have saved us from a colossal mistake."

The government also has used pleas not to publish a story to control the release of information on its own terms. Jack Nelson, who spent more than two decades as the L.A. Times Washington bureau chief, recalls investigating a story about the Glomar Explorer, a recovery ship that Howard Hughes' company built for the CIA in a secret effort to raise a Russian submarine that had sunk near Hawaii.

In 1975, as Nelson was preparing his story on the incident, then-CIA director William Colby visited Nelson's office to urge him not to publish. "My response to him was that I thought it was an important story," Nelson says. "I could not see any national security danger," and he knew a number of other people already were aware of the episode.

Colby flew to Los Angeles to talk to then-Times Editor Bill Thomas, who decided to hold the piece: "He actually put it in a safe," Nelson recalls. Colby then returned to Washington and stopped by all the major news outlets in the city, informing each one of the attempt to raise the submarine and begging them not to disclose it because of national security risks.

A few days later, Nelson was in the Class Reunion, then a popular Washington bar, when Colby called to say he could tell his editors in Los Angeles to run the story; columnist Jack Anderson had broken it on the radio that evening. Because Colby had "gone around telling everybody, he got pretty good coverage," Nelson says. "Papers' editorials commended the CIA for the way they were able to raise this Russian submarine."

Describing their calculus for weighing whether to publish, editors cite an imprecise combination of case-by-case evaluations, a balancing test of public interest versus national security, and, ultimately, some reliance on their gut. "It starts with instinct," Bradlee says.

The Post's Downie notes that publication quandaries are by no means unique to national security stories. Editors constantly decide to hold or kill stories for myriad reasons: "It's not ready; it's too long; it's not interesting enough; it's not accurate yet." When confronting murky national security questions, he says, "The burden is on us to become as well-informed as we can be."

Paul E. Steiger, managing editor of the Wall Street Journal, declined an interview request but provided answers to written questions. "The presumption is that we will publish what we have learned through our reporting, if we think a story is newsworthy," he wrote. "But we will always listen to specific concerns raised by responsible government officials and weigh them seriously before reaching any final decision on publishing."

USA Today Editor Ken Paulson says, "News organizations serve an important watchdog role, and their mission is to scrutinize what the government does, particularly if government conduct raises constitutional questions. That said, you have to take a close and careful look at what you're about to publish and assess whether there's anything in your story that could significantly undermine national security."

His paper reported May 11 that the NSA was compiling a database of domestic phone call records with help from several leading telecommunications companies. The Times already had exposed the NSA's eavesdropping, so "any potential terrorists were already aware that the government was monitoring phone traffic. The revelation that phone records--pure data, and not wiretapping--were available to the government didn't seem to us to be endangering national security in any way."

Paulson declined to comment on whether the administration asked his paper not to publish. (On June 30, after post-publication denials from BellSouth and Verizon, USA Today reported that some lawmakers say cooperation by the companies was not as extensive as the paper initially reported, and it could not confirm that those two companies contracted with the NSA to provide bulk calling records.)

The New Yorker's Coll describes the decision-making process this way: "There are certain principles behind each decision, but each case is decided first from the ground up, from the facts. You weigh and evaluate the facts in the context of these broader principles and also precedent."

The facts, of course, form the crux of an editor's decision, and outsiders who aren't privy to off-the-record conversations between news organizations and government officials can't easily evaluate them. Thoughtful people also can reach different conclusions about particular sets of facts, as the Swift banking records story demonstrates.

"Why tell terrorists how you are tracking their money?" asks former CIA Director Jim Woolsey. "I don't understand what publishing that was supposed to accomplish, why the press would consider itself a watchdog for [a program] tracking terrorist funds that, as far as I could tell, had nothing illegal or nonconsensual" about it.

In debating whether to go with national security and terrorism-related stories, Woolsey suggests three factors to guide publication decisions: First, "is there some oversight mechanism" in place? That oversight, he says, may be briefings even to a small subset of Congress. "Is there abuse?" Watergate, he notes, grew out of a burglary of the Democratic National Committee; no equivalent abuse appears to have existed with the NSA's wiretapping program. Finally, is the program being exposed "important in collecting intelligence on terrorist threats to the country?"

Marvin Kalb, a senior fellow at Harvard University's Joan Shorenstein Center on the Press, Politics and Public Policy, describes the Swift story as "one of those 52-48 decisions; it was not a 90-10 call." Kalb, who coedited the 2003 book "The Media and the War on Terrorism," says he would have published the story but focused more on "the lack of accountability for major administration action and decisions, particularly a lack of accountability as exercised by Congress."

Steven Aftergood, director of the Project on Government Secrecy at the Federation of American Scientists, was quoted in the L.A. Times' Swift story as saying the program "boils down to a question of oversight... Their policy is, 'Trust us,' and that may not be good enough anymore."

In an interview, Aftergood described publication of the Swift piece as an even harder decision than other recent cases because "not only was there no crime, no allegation of misconduct, but arguably this is one of the smarter steps we're taking to combat terrorism, and anything that jeopardizes that would be supremely regrettable." He adds: "Now that some time has passed, I don't think there is any evidence of damage. Some of the most important facts of the matter were in the public domain."

Keller concedes the Swift story was a more difficult judgment in some ways, but he felt an enormous surveillance program without congressional oversight or substantial congressional briefing was an important issue. "It's probably easier for the public to get its head around a story like this when a chorus of voices is saying it's illegal or unconstitutional. Whether or not this one was a harder call, it was maybe a harder sell," he says. "It's easier to explain exposing something that a lot of people view as scandalous. That's not the only reason you would expose a secret program. Our job is to tell people how government is waging the war on terror."

Ultimately, the L.A. Times' Baquet felt the government did not make a powerful case to withhold publication. "The reason people think we did something wrong is because they think that our job is to work with the government in the war on terrorism," he says. "That's not our job. Our job is to cover the government."

Baquet is willing to explain his decision, but he hopes he's able to do so without appearing defensive or repentant. "Sometimes I worry that when we explain ourselves that people misunderstand that we were in pain or anguished," he says. "I want to make it clear as the dust settles that I am completely confident in publishing that stuff. I'm not apologetic about publishing the Swift story."

His one regret: "I wish I'd gotten my story up before the New York Times did."