FAS

Pentagon Manual Urges Precision in Classifying Information

04.08.13 | 3 min read | Text by Steven Aftergood

The Department of Defense is not particularly concerned with “openness” in the abstract, but it is strongly motivated to conserve resources and reduce discretionary expenditures.  That imperative dictates the discriminating use of national security secrecy — at least in theory — because of the costs incurred by classification.

“Precise classification guidance is prerequisite to effective and efficient information security and assures that security resources are expended to protect only that which truly warrants protection in the interests of national security,” according to a newly reissued Department of Defense manual that provides Instructions for Developing Security Classification Guides (DoD Manual 5200.45, April 2, 2013).

The Manual presents a framework for classifying information, beginning with a series of questions for determining whether information is eligible for classification in the first place.  (“Can the unauthorized disclosure of the information reasonably be expected to cause identifiable or describable damage to the national security? If the answer is no, the information cannot be classified.”)

Then it details the considerations that may arise in the classification of information concerning military hardware, military operations, intelligence and foreign relations.  Exactly which aspects of a hardware system provide the U.S. with a technological advantage?  Precisely which operational information requires protection in order for the operation to succeed?

The Manual discourages broad, sweeping classification of information.  Through an extended questionnaire for classifiers, it seeks “to systematically bound and refine the scope of the analysis needed to determine which items warrant protection through security classification.”

As detailed and helpful as this guidance is, it cannot decisively resolve all classification questions in advance.  “The outcomes specified in the flow chart are not absolute; judgment must be applied in all cases,” the Manual says.

Some decisions regarding what to classify are easy.  Information about “a foreign official speaking in a highly critical manner of his own government’s policy” or “suggesting how pressure might effectively be brought to bear on another part of his own government” would typically be classified.  Likewise, the fact of ongoing intelligence cooperation between the United States and a country “with which the United States is not allied should always be classified.” Also, “intelligence identifying a sensitive source or method should always be classified.”

On the other hand, the Manual says, classification can be dispensed with in certain areas.  “Intelligence that reveals the identity of a conventional source or method [i.e., one that is not “sensitive”] normally does not require classification.”

And “in general, [intelligence budget] resource information should not be classified unless it reveals some aspect of the intelligence mission, and its revelation would jeopardize the effectiveness of a particular function. An example of classifiable resource information is the intelligence contingency fund.”

At this point, the realization will dawn on some readers that this Manual, which was originally issued in 1999, does not accurately describe — or effectively regulate — DoD classification policy.

It is simply not the case that “in general, resource information [is] not classified unless it reveals some aspect of the intelligence mission.”  To the contrary, the Department’s default position is that intelligence resource information is classified.  That is why the individual budget total of each DoD intelligence agency — and not simply intelligence contingency funds — is not disclosed.  And beyond the total budget for the Military Intelligence Program (MIP), “no other MIP budget figures… will be released,” DoD said last October, “as they remain classified for national security reasons.”

Nor does the Pentagon clearly distinguish between sensitive intelligence sources and methods, which are classified, and “conventional” intelligence sources and methods, which are not.  Doing so sounds like a good idea, but it is hard to detect any sign of it in Pentagon practice.

The Manual notes that all classification guidance should be reviewed at least every five years by the original classification authority that issued it.  But it fails to mention that all guidance is also subject to a broader periodic assessment known as the Fundamental Classification Guidance Review.  It is through such a broader review that changes in classification policy are more likely to come about.