Secrecy & Leaks

An Approach to Ameliorating Press-IC Tensions Over Classified Information

By Benjamin Wittes
Friday, May 22, 2015, 2:55 PM

I've been thinking about the exchange over the past couple of weeks---much of which took place on Lawfare---between the New York Times and the intelligence community over the naming of CIA undercover officers in a Times story. (A brief recap in links: here are Bob Litt's original comments, the 20 former intelligence officers' letter, Jack's interview with Dean Baquet, my comments in response, Mark Mazzetti's comments, and Jack's comments.)

I want to float an idea for a mechanism that might ameliorate tensions over this sort of issue in the future. It won't eliminate those tensions, which are inherent in the relationship between a government that has legitimate secrets to keep and a press that rightly wants to report on the activities of government, but it might give the public a lens through which to view individual disputes, and it might create a means through the which the two sides can better and more fully communicate in high-stakes situations.

The basic problem when the press has an undoubtedly newsworthy item that involves legitimately sensitive information is two-fold: the government will tend to err on the side of overstating the sensitivity of the information, because it has to protect against the biggest risk, and the government often cannot disclose to the newspaper the full reasons for its concerns.

Reporters and editors don't have clearances. The stories often involve leaks that will require investigations, after all. So the government's claims are often unnuanced---as Baquet's interview makes clear: The sky is always falling, and the details are always scant. And then, after publication, the sky does not fall, at least not visibly to the press. So the government's concerns seem alarmist. Some of that is because some government claims are, in fact, undisciplined and alarmist.

But at least from the government's perspective, there's more to the story than that. Damage may have,  in fact, occurred that---once again---you're not going to share with the New York Times. And even if it didn't happen, the claim was never that this particular bad thing would certainly happen; it was that publication would create an undue risk that it might happen.

While the ultimate tension is, as I say, unresolvable, one aspect of it certainly can be addressed: the impaired communications between the two parties that hamper the government's ability fully to discuss concerns based on classified materials.

There's a simple mechanism that might improve those communications: a cleared intermediary, chosen by the media organization, who can hear a full presentation by the government on its concerns. The key is that this person would be (a) nominated by the press organization, not chosen by the government, (b) cleared at the appropriate level by the government, and (c) asked to make independent judgments as to whether government requests to withhold information are merited based on all facts the government and the press organization consider relevant.

The use of this mechanism would be entirely voluntary on the part of the press, and the press organization would not be obliged to follow the recommendation of the intermediary. So the mechanism would impose no obligation on the part of the media---save one: If the news organization chose to make use of the mechanism, it would oblige itself to publish the recommendation of the intermediary in the relevant story or stories.

This would have several effects: If the intermediary concluded that the government's concerns were unpersuasive---as, for example, Jack did of this recent incident---it would powerfully validate the newspaper's publication decision to have someone who knew the full scope of the government's concerns backing that decision. If, on the other hand, the intermediary backed a government request to withhold material and the media organization chose to go ahead and publish anyway, it would have to inform readers that it done so over the advice of its own hand-chosen ombudsman. And it would thus make it far easier for future Bob Litts to argue that the paper had "disgraced itself."

This mechanism would also create a means for the government to report damage after the fact in a way that it will not do now. If, following the publication of a sensitive story, something bad happened, the agency might inform the cleared intermediary of its belief that the two events are connected, address any questions about the causal link, and let the intermediary report to the newspaper in general terms his or her conclusions---in general, non-classified terms---as to whether consequences were reasonably attributable to publication.

Such a system would have, I want to emphasize, relatively limited application. The new wave of publications devoted to revealing material because it's sensitive will find any system that risks inhibiting publication to be of negative value. And even among press entities that take national security concerns seriously, it would probably slow things down, as it would add a layer of communication to a system that already requires two bureaucracies to interface with one another. And some in the intelligence community will certainly object to giving highly sensitive material to the media---even if to a cleared representative for the limited purpose of facilitating this sort of negotiation.

Still, I suspect there would be value to a system that allows the government to make a fuller case and allows the press to receive it. At a minimum, such a system would facilitate better communications---and would let the readership know that these communications had take place.