The Secrecy Charade

A long-time researcher on top-secret projects reveals why our secrecy system doesn't work.


The Soviet Union has secrecy; the United States is playing at it. It is not clear that the use of longterm secrecy for the United States makes any sense. It is clearer that the extent to which we are now enforcing secrecy is exaggerated, and it is completely obvious that merely playing at secrecy has no value at all.

The Executive Order proposed by President Reagan this past April (see sidebar, p. 44), supposedly makes an attempt to give our secrecy regulations a more consistent and effective footing. But widespread criticism of the order includes the charge that it unwisely broadens the government's power to keep information secret.

The goal of changes in our secrecy laws should be to bring about an even stricter policy on a level of less stringency—that is, better protection of far less classified information. Such a policy, by the very fact that it would try to keep fewer secrets, could be more effective in areas that really matter.

The simplest solution would be to abolish secrecy altogether. Unfortunately, this cannot be done. If we were to publish the prospective routes of our nuclear submarines, the sense of having such submarines would disappear.

On the other hand, long-term secrecy is not apt to work. The Soviet Union as a totalitarian police state has placed greatest emphasis on intelligence work, and such work is unhampered by the requirements of civil rights or even common humanitarian values. Furthermore, intelligence personnel in the Soviet Union are among the most influential. Therefore, the secret service attracts much of the best talent in a country where the end is taken to justify the means. One can understand why long-range secrecy has never worked in democracies. Such an attitude is abhorrent to people who enjoy and value freedom in their daily lives.

An assessment of the security system in our nation today is made more difficult because our intelligence community has an interest in appearing to be efficient. The fact that our secrets are being lost is itself a secret that is never admitted except in cases of obvious and indubitable proof.

One way to evaluate what and how much we ought to keep secret is to ask the question: From whom do we wish to keep secrets? from the Soviets? from terrorists? from our allies? from our own people?

My firm belief is that it is useless to try to keep secrets from the Soviets for longer than one year. Therefore, I propose that classified material should be automatically and completely declassified after that period. There may be a few exceptions, but they should be exceedingly few and of clearly demonstrable importance. What one might do to limit their number is to require a presidential order to create such an exception. The president would have to be advised of the reasons why these secrets needed to be kept for a longer period but would assume personal responsibility for their continuance as a secret. This would happen in few cases. The present situation—where we are drowning in Secret, Top Secret, Special Access, For Eyes Only, and similar types of information—would be radically and beneficially reversed.

At the same time, the greatly limited secrecy that would prevail in the new situation should and could be rigorously enforced. Violations should be prosecuted as most serious criminal offenses. Today, we have the peculiar situation where we have innumerable secrets and probably not a single convicted violator. In other words, we have a system of unenforceable laws that lowers respect for laws.

The effect on the Soviet Union of obtaining the vast majority of our classified information after a year would be that that nation could save some expense on intelligence operations. We, on the other hand, might well be better able to keep our most important secrets for a year, a feat that we perform with questionable success today. The effect in other respects would be equally beneficial for us.

Aside from the Soviet Union, there is the matter of the availability of our secrets, particularly about weapons, to terrorists. It should be noted that, with very few exceptions, no country that undertook the construction of a nuclear bomb was seriously hampered in its efforts by the existence of secrecy. The myth that terrorists can make an atomic bomb from stolen material in a garage is widely accepted. It is supported in the public mind by the circumstance that a Princeton student could make suggestions about a nuclear bomb that were not nonsense. What is overlooked is the fact that in order to produce a bomb, a group of people who must be highly gifted and of impressive expertise would have to spend $20–$30 million in their "garage." This would be a fairly conspicuous operation.

These statements can be generalized to the troubling question of industrial secrecy. We are well aware that the Soviets are obtaining advanced technology directly or indirectly from ourselves and our allies, partly by purchase of certain items and partly by outright espionage. It seems to me that the right answer is not further complication of industrial interaction with our friends and trading partners, but rather fostering, through openness, more rapid development with which the Soviet Union would have trouble keeping pace. Our secrecy is useless in military areas where the Soviets are probably ahead of us. They not only know all our secrets but probably they know secrets that we will discover in the next three years. In high-technology industries, we are ahead, but efforts to conceal our progress slow us down. On the whole, secrecy seems counterproductive.

The most serious disadvantage is that we keep our own people in ignorance and in a state of ever-increasing confusion. In a totalitarian regime, people don't participate in government, and therefore they do not have a "need to know." The Soviet Union is paying for its policy of secrecy by the lesser involvement of people. This obvious disadvantage of a police state leads to less and less intelligent contributions by junior scientists who are not necessarily the least ingenious.

In a democracy, national policy is most obviously and deeply influenced by public opinion. Therefore, the public has an urgent "need to know." Even if secrets would not interfere with public understanding of the real issues of defense, it would still remain hard to inform our own people sufficiently and accurately enough that the average opinion of the layman would be sufficiently enlightened to arrive at the right decisions. The greatest evil in secrecy is that it encourages ignorance and reinforces emotionally based decisions.

An obvious example is connected with atomic weapons. The secrets of fission and fusion bombs are several decades old. They are known to at least a million individuals. To imagine that they are secret is folly. If these secrets were generally known, it would be much easier to make it clear to the public that while nuclear war would be terrible, it will not lead to extinction of mankind. The small residue of uncertainty in that statement is connected with the dangers inherent in any radical change in the history of the world.

The danger is greater that a man like Colonel Qaddafi could secretly produce a few atomic bombs and deliver them to selected friendly terrorists. Indeed, proliferation of nuclear weapons is disturbing and quite possibly unavoidable, but secret proliferation of atomic weapons bringing about the situation suggested above would be even more dangerous. This sort of proliferation would be more easily avoided if the United States, having renounced long-term secrecy, would turn to enforced long-term openness in as great a part of the world as possible. I am not saying that greatly reducing our secrecy practices will provide a radical cure for an extra-dangerous headache. All we can hope for in our difficult fight with terrorists and their associates is a better chance to fight an obvious and growing peril.

The rest of the problem is more straightforward and happily connected with a real potential improvement of the present situation. Today, the most obvious and important division between various kinds of regimes can be defined by asking which countries pursue stability and steady improvement of the livelihood of all people, and which choose to embark on radical changes in the political institutions of the world. Those for stability and against any major conflict are our natural allies. This alliance is not based on temporary expedience but on our long-range goals; such associations are almost certain to endure. Allies are also necessary in order to obtain strength and to reduce the danger of war, which is becoming more acute because of the destabilizing and increasing influence of the Soviet Union.

Under these conditions, it is clearly desirable to share our secrets with our allies, thereby increasing the effectiveness of our cooperation. The result, again, may be that the Soviets can penetrate our secrecy with somewhat greater ease. If we recognize that most secrets cannot be kept beyond a year and make arrangements with our allies for joint efforts in keeping secrets for the required short time, the overall result may be better, not only in increasing cooperation but also in enforcing short-term secrecy. The paradoxical effect could be that the Soviet Union may find it harder to obtain secrets under the new regulations than under the present system.

What is even more important, secrecy has prevented a general explanation of the recent efforts by which our MAD policy of Mutual Assured Destruction can and should be replaced by the development and deployment of new defensive weapons. These protective weapons are apt to be atomic in nature, but they would play the role of a shield rather than a sword. Today, such a statement seems unacceptable. It is apt to remain in that status unless rules of secrecy are appropriately changed so that more detailed explanations can be offered.

In 1969 when antiballistic missile (ABM) systems first received widespread attention, I mentioned this point in a television debate I had with Richard N. Goodwin. I suggested that we proceed with research on an ABM system and reassess its value at the end of one year. Goodwin replied that allowing the program to start would be like offering the devil a finger: one would end up losing an arm. I suggested that when the review would be held after a year, people would be able to judge the question more intelligently if the interim publications on such systems included more detailed facts—if information could be declassified. Goodwin replied that he had been Jack Kennedy's speechwriter for some years, and that the American people knew all that they needed to know. I mentioned my surprise that, as a liberal, he did not have a greater understanding of the need for openness. Goodwin asked if I were questioning his credentials, adding, "I am as good a liberal as you are." No doubt this is true, but it is equally true that many liberals seem to be unaware of the importance of openness. (The reversal of publicly perceived roles is echoed today in discussions of defense policy. The hawks are arguing for better protection of the population; the doves are arguing for more and bigger weapons.)

One of the main differences between democracies and totalitarian states is that a scientist whose business is openness finds it onerous to work under the restrictions of secrecy in a democracy. In a police state, such feelings of revulsion matter little. In the United States and allied countries during the prelude to World War II, the eagerness for defense work (such as that done on radar and code-breaking computers) saved the free world. The presence or absence of this attitude could easily make the difference between the survival and the extinction of our way of life.

The question of secrecy and openness is not a simple one. Oversimplified solutions, including the one proposed above, will prove to be insufficient. In arguing for openness, I would not like indiscriminately to invite all technically competent people into our laboratories. However, secrets that can be given away in a few sentences will never survive for a long period. By comparison, the transfer of more-complicated information can be made difficult rather easily. Industry has done so for long periods of time. What I am talking about can be compared to the difference between a law of nature and the nebulous thing called know-how. The former should never be a secret (at least not due to human volition); the latter, ill-defined as it is, has been successfully limited by industries for considerable periods.

In the end, the distinction between whether some information should be kept secret or released depends on the exercise of an art. It should be recognized that this art will not be pursued in an effective manner in a free society unless the guiding rules are made as reasonable as possible and unless people recognize the reasonableness of the rules and therefore cooperate willingly in their execution.

I have emphasized that in exercising secrecy, time is the essential element. What is necessary and desirable in the short term becomes impractical and even absurd on a long-term basis. Another less clearly defined guideline may be the distinction between rigid secrecy and flexible privacy. In the longer run, the tricks of any trade can be limited to the chosen practitioners—the general facts cannot be.

That uncertainty remains in the exercise of limiting information is an unfortunate fact. That the present state of affairs is in great need of change is even more obvious. Perhaps the best example is an incident that occurred in the not very distant past when American security officers attempted to prevent a Soviet visitor from informing Americans who did not possess proper clearance about an advance made in the Soviet Union. This Russian secret was in fact well known in American secret laboratories and furthermore could have been easily guessed by the rest of the international scientific community.

As long as we live in a world full of political hazards, the problem of secrecy will remain. But there can be no question that our present procedures could be improved. To my mind, it is completely clear that improvement lies in the direction of fewer secrets and more understanding.

Nuclear physicist Edward Teller was a key member of the Manhattan Project during World War II and later worked at Los Alamos on the development of thermonuclear weapons. He is the author of numerous books on the public policy aspects of scientific and technological issues and today is a resident scholar at the Hoover Institution at Stanford University.


For a period of about 25 years, until the advent of the Reagan administration, the federal government made fitful progress in the direction of declassifying government information and making it available to the public. Since the election of Ronald Reagan, however, that trend has been reversed dramatically.

Perhaps the most important measure it has taken to extend the range of governmental secrecy has been Executive Order 12356, signed by the president in March. According to Allen Adler, legislative counsel of the Center for National Security Studies in Washington, the order's message is unmistakable: "To the sensible bureaucrat, it says that more information should be classified and for longer periods of time. When in doubt, classify."

Adler notes that there are three significant changes EO 12356 makes from the Carter administration classification policy. First, federal officials effectively no longer have the discretion they were given previously to decide whether to classify information. Now, if there is "reasonable doubt," they must classify information pending a review process. Second, it eliminates requirements that government officials consider the public's interest in disclosure before classifying information and that classifications be based on identifiable potential damage to the national security. Third, it permits the reclassifying of information that had been declassified if "the information may reasonably be recovered."

One of the most comprehensive reviews of EO 12356 so far has been conducted by the House Government Operations Committee. It released a report in August that blasted the administration for the secrecy surrounding the drafting of the executive order, the reasons offered by the administration for the changes it was making, and the order's failure to address the real problem—overclassification. The report noted that the White House made only minimal efforts to consult with Congress and the public before adopting the order: the White House took a year to draft the order, yet Congress had only three weeks to comment on it, and it was never released for public comment. (Although executive orders do not need congressional approval to take effect, the Carter administration had published a draft of its security classification rules and actively sought public comments before implementing them.)

Adler emphasized that the actual classification practices of any administration are mediated by the federal bureaucracy, so they bear only a rough relationship to stated policy. Still, there are disquieting indications of the direction the Reagan administration is headed—at the very least regarding the reclassifying of information—in an incident involving author V. James Bamford.

In 1979, Bamford requested and obtained 250 pages of documents on the National Security Agency (NSA) for a book he was writing called PUZZLE PALACE: A REPORT ON AMERICA'S MOST SECRET AGENCY, which is scheduled to be published later this year. The documents concerned a Justice Department investigation in 1975 into reported illegal spying by the National Security Agency and the Central Intelligence Agency, an NSA operation that placed noted anti–Vietnam War protesters under surveillance, and agency monitoring of international telex communications in possible violation of the 1934 Communications Act.

In the fall of last year, the Reagan administration decided to reclassify the information that Bamford had obtained nearly two years earlier (and some of which he had already given to a reporter for the SUNDAY TIMES of London). Gerald A. Schroeder of the Justice Department wrote Bamford demanding the documents. "You are currently in possession of classified information.…It is therefore your duty and obligation as a United States citizen to return this information to the Department of Justice." Bamford refused: the foolishness of reclassifying information that had already been seen by reporters and publishers in this country and abroad apparently eluded the Justice Department.

The Bamford episode is somewhat reminiscent of the censoring policy of Captain Yossarian in CATCH-22, who had to review letters of enlisted men. "To break the monotony, he invented games. Death to all modifiers, he declared one day, and out of every letter that passed through his hand went every adverb and every article.…One time he blacked out all but the salutation 'Dear Mary' from a letter, and at the bottom he wrote, 'I yearn for you tragically. A.T. Tappman, Chaplain, U.S. Army.'"

—Paul Gordon