Press Box

Too Many Secrets, Says Secrecy Czar

J. William Leonard frets about the breakdown of the classification system.

In a little-noticed mid-June speech, secrecy czar J. William Leonard fretted over signs of a breakdown of the classification system for national security.

Leonard heads the Information Security Oversight Office, the National Archives branch that develops classification and declassification policies at the behest of the president. In his talk, given at a classification training seminar, Leonard complains that the system has lost touch with the “basics”: Some agencies don’t know how much information they classify; they don’t know whether they’re classifying more than they once did or less; they don’t know whether they’re classifying too much or too little; and they don’t know whether they’re classifying material for too long a period or too short.

This quality-control breakdown has resulted in agencies classifying too much information and, in some cases, classifying information that by law shouldn’t be stamped “secret” in the first place. Implying that the government classified the reports of prisoner abuse at Abu Ghraib to cover up the scandal, Leonard says:

[I]n no case can information be classified in order to conceal violations of law or to prevent embarrassment to a person, organization or agency. … Specifically, “exactly from whom are we keeping the information secret?” In the case of detainee abuse, we are obviously not keeping it secret from the detainees—they experience the abuse and interrogation techniques first hand. And I assume we do not expect them to sign a nondisclosure agreement upon their release from custody based upon the premise that they had been exposed to classified information when they are subjected to abusive techniques.

Don’t mistake Leonard for an ACLU firebrand: As Steven Aftergood of the Federation of American Scientists (whose excellent listserv alerted me to the speech) puts it, Leonard “is not an ‘openness’ advocate or a free-lance critic of government secrecy.” He’s a career security professional who deplores the leaks of classified material to the press.

Leonard attributes what he calls an “epidemic” of leaks to the press to the dysfunctional classification system, which has recently taken to using the war as an “excuse to disregard the basics of the security classification system.” Leaks are coming out of the “highest levels of our government” (the Valerie Plame affair); a former Cabinet secretary is alleged to have handed off classified material to a book author for publication, and the classification machine is operating so poorly down at Guantanamo Bay that a chaplain was publicly charged with pilfering secrets on his computer and then released.

“The problem [Leonard] has identified is that the currency of classification is being devalued by questionable, sometimes suspiciously self-serving secrecy actions,” writes Aftergood in e-mail. “This produces an erosion of security discipline, which in turn fosters an environment in which leaks are more likely to come about. The net result is bad security policy and bad public policy.”

Because leaks of classified information make for such great headlines, journalists rarely give much thought to why something was leaked or why it was classified in the first place. Leonard’s speech encourages us to look for the important story behind every leaked classified-info story and ask these questions: Why was the information classified in the first place? Who or what was served by its classification—some self-interested bureaucracy or our national interest? (Think Abu Ghraib.) Who was served by the leak? Who was damaged? (Think Valerie Plame.) Who is served by declassification delays?

The secrecy czar has spoken. But who’s listening? According to Nexis, nobody. I couldn’t find a single story about the speech. Maybe he should have leaked it to the press instead of posting it on the Web.


See this In These Times piece by Aftergood, in which he gets Leonard to agree to investigate the decision to classify the Abu Ghraib report. Send e-mail to (E-mail may be quoted by name unless the writer stipulates otherwise.)