More fundamentally, it is the rules, the system, that should be changed, in pursuit not merely of abstract principles of openness but of the realistic benefits of genuine democracy. These potential benefits, foreseen by the drafters of the Constitution and the First Amendment but demonstrated most dramatically in the last fifty years, include the more likely avoidance or termination of execrable, unjustified, immoral, dangerous or catastrophic courses of action: policies of a kind we have actually suffered, that could not have gained or maintained political support from an aware and adequately informed public because these policies did not remotely deserve it and because in some cases they continue to endanger national and human survival.
Precisely in light of that experience, the burden of proof in our democracy should shift to lie upon keeping secrets from the public rather than on revealing them. Freedom of governmental information should be the rule, the standard and norm, not the occasional result of a protracted and contested process. And in the nuclear era, the survival of a large part of humanity may depend on the spread of democracy far more open than exists in any current nuclear weapons state, starting in our own country.
The creation of a "Doomsday Machine" first in the United States and then in the Soviet Union -- nuclear forces on hair-trigger alert capable of destroying most life on earth by producing nuclear winter -- and the persistence of these Machines subject to false alarms and accidental detonation a generation after the end of the Cold War has always depended on public ignorance of unconscionable risks, maintained by dangerously effective secrecy systems on both sides.
The point here is not to make the unrealistic claim that no secrets are compatible with or justified and even required under a democracy. Rather it is to recognize that the existing, past and current system and culture and practice of secret-keeping is dangerous and destructive in its effects, massively and urgently so. It needs something far beyond reform and regulation, a reconsideration and institutional overhaul that amounts to a replacement by a secrecy system almost unrecognizably different in its scale and democratic accountability for the secret-keeping that must remain.
As someone who has read hundreds of thousands of pages of classified documents, I would estimate that if a very large random sample of the billions of pages now classified were examined by a panel of well-informed and experienced individuals-not currently employed by the Executive branch or the (largely co-opted) Intelligence Committees of Congress nor hoping soon to be so employed -- they would judge that much less than 1% of such material that is more than two or three years old deserves to be classified any longer, if it ever did, by a reasonable application of the system's own criteria.
That was the expert estimate forty years ago of one of the most experienced security authorities in the Pentagon, William F. Florence, who had drafted many of the Department of Defense regulations on classification. He testified as an expert witness in Congressional hearings and in my trial that at most 5% of classified material actually satisfied the official criteria of potential relevance to national security (which he had played a major role in formulating) at the moment of original classification; and that perhaps 1/2 of 1% continued to justify protection after two or three years. vi
That would be all the more true if the criteria for classification -- as they should, and as was directed for a period in the Carter administration -- balanced hypothetical or possible dangers from informing an adversary against a strong general need and right of the public to know.
This is to be contrasted with frequent estimates in critical hearings or studies on classification that as much as a quarter or perhaps "50%" of information currently still classified is "over-classified," and no longer requires special protection and handling if it ever did. In the eyes of Florence (and me) such presumptively shocking judgments are under-estimates by an order of magnitude, perhaps two. True "over-classification" may be ten to a hundred times greater in scope.
There has been an actual large-scale test of these contrasting estimates, whose implications seem never to have been noted in these discussions. The Pentagon Papers which I released consisted of thousands of top secret pages, which were not just a random sample of material ranging from 3 to 26 years old but which represented, in a high proportion of cases, the most "sensitive" and closely-held of top secret documents: crucial national intelligence estimates, JCS recommendations and plans, secret presidential decisions and discussions.
Most of the documents that were selected to be quoted in full or in large part in the studies were not classified top secret inadvertently, carelessly, thoughtlessly, or because they happened to derive from other, more-properly-classified documents or were stapled together with such. They were precisely what any careful staffer or commander would have assigned top secret status and stamp-marks, by the standards of the bureaucracy.
Yet in two years of close search of these documents while I was being prosecuted, the government was never able to identify a line or a paragraph that could convince a judge or a jury that national security had been impaired by my revealing it. In other words, my own prior judgment that no single sentence or page would have that effect, nor all of them together, was confirmed by the government's inability to show otherwise. The percentage of the documents from 1945 to 1968 that was properly classified as of 1971 proved by this actual test to be not 50% or 1% or .05% but exactly 0. And, to repeat, this was not a random selection of top secrets but a highly selected, though huge, collection of "sensitive" documents.
Another way of judging the actual scale of wrongful-not merely "unnecessary'-withholding of information from the public is to look comprehensively at the library of documents that have emerged eventually from safes through the long and tortuous process of FOIA requests, often twenty and thirty years after being written. Look in particular at the voluminous files of declassified documents, by subjects, in the website of the National Security Archive of George Washington University, http://www.gwu.edu/~nsarchiv.
Note when the documents originated and when they were finally released (often after years of appeals). And ask yourself: When should this information have been available to journalists, scholars, Congress, the public? For what fraction of this interval could a solid case be sustained that the value of this information for the functioning of democracy -- to understand and evaluate and if necessary change policy and performance, to uphold domestic and international law, to hold officials accountable, to exercise public sovereignty over foreign policy-was outweighed by a need to conceal it from foreign powers? The same questions can be asked of the more comprehensive but still-further-delayed publications of the series Foreign Relations of the United States (FRUS). Oddly, I've never seen this critical exercise suggested, let alone carried out.
What can be asked of these bulk disclosures from FOIA and FRUS should also be asked of every leaked document revealed belatedly in the press. "When should we have known this? By what specious reasoning, and why in reality, was it withheld so long? Who lied to us about this, how were we misled, and why? What interests were served? What does this tell us about our real foreign policy, and the proper role of Congress, the issue of war powers and the possibility of democratic control of foreign policy?"
And finally: "Is there not a need for Congress to investigate the nature and dangers of a secrecy system so prone to abuse and dangerous to democracy?" Historically, all these questions should have been asked in Congressional hearings and by investigative reporters and scholars soon after the Pentagon Papers became public. They were not.
At the same time, the almost exclusive critical focus on "over-classification" or "unnecessary classification" is misleading in two important respects. First is the much-quoted, mistaken proposition, expressed by Justice Potter Stewart in his Pentagon Papers decision, that "When everything is secret, nothing is secret." In other words, indiscriminate use of the secrecy stamps supposedly breeds carelessness and near-contempt for the secrecy constraints, resulting in too many leaks, or the president's alleged inability to keep any secrets reliably or very long out of the newspapers.