FAS | Government Secrecy | July 2000 News ||| Index | Search | Join FAS

Chronicle of Higher Education
July 7, 2000

Critics of Los Alamos Ignore a Flawed Policy on Secrets

by Steven Aftergood

The root cause of the recurring security failures at Los Alamos National Laboratory, we are told, involves a clash between the culture of security and the culture of scientific inquiry.

Los Alamos’s security has been under scrutiny since last year, when Wen Ho Lee, a former scientist there, was accused of improperly downloading and copying materials from a classified computer network. When, more recently, two computer hard drives containing nuclear secrets went missing and then mysteriously reappeared, the spotlight on the issue became even more intense.

“The scientific culture of the weapons laboratories complicates, perhaps even undermines, the ability of the Department [of Energy] to consistently implement its security procedures,” according to a report of the President’s Foreign Intelligence Advisory Board.

“The scientific and academic community has been disdainful of those who are emphasizing the importance of security, and that is a deeply rooted culture that we need to extirpate,” said Senator Richard Bryan, a Democrat of Nevada.

High on the list for possible extirpation are the contracts held by the University of California to manage Los Alamos, Lawrence Livermore, and Ernest O. Lawrence Berkeley National Laboratories. “I am deeply concerned,” said Senator Jeff Sessions, a Republican of Alabama, “that we have contracted the security and development of our most vital national secrets to a state university. I do not any longer have confidence in it.” Energy Secretary Bill Richardson has said that he may terminate the University of California contract, but that hasn’t deterred other senators from calling, with varying degrees of self-righteousness and pomposity, for Richardson’s resignation.

The notion that scientists and security officers are “oil and water” -- or, as Senator Pat Roberts, a Republican of Kansas, put it, “sheep and cattle” -- is a convenient formulation that is nevertheless misleading. It does not properly describe the particular realities of nuclear weapons science. It also obscures other, far more important challenges to the security of the nation’s most sensitive secrets.

As a general statement, it is undoubtedly true that science encourages and depends on openness. It is by publication of research that scientists announce their discoveries to the world. It is through peer review that the canons of scientific research and the quality of publication are upheld. It is through the cross-fertilization of ideas among diverse disciplines that new vistas for scientific exploration are opened.

Secrecy by definition impedes each of these essential facets of the scientific enterprise. But granting the importance of openness for scientific and technological progress, it does not follow that scientists can’t keep a secret.

In fact, the story of nuclear-weapons development begins with scientific secrecy. At the dawn of the nuclear age, in 1939, before the government or the military ever got involved in nuclear research, scientists led by Leo Szilard initiated their own nuclear-secrecy policy which they imposed on themselves. Given the wartime environment, they declined to publish certain experimental results related to the fission of uranium and the production of plutonium.

The great nuclear physicist Enrico Fermi went so far as to write that "Secrecy was not started by generals, was not started by security officers, but was started by physicists."

The scientists who work in our nuclear weapons programs have long since made their peace with the need for secrecy in their research. After all, if a significant fraction of nuclear weapons scientists couldn’t resist blabbing about their work, all of our nuclear secrets would have long ago been disclosed. Yet plenty of secrets remain. Those scientists who cannot abide the sometimes-oppressive security environment probably would never have applied for a security clearance and signed a non-disclosure agreement. Those who nevertheless did so would soon have moved on to work in a more-congenial atmosphere.

If there is a "clash of cultures" at the weapons laboratories today, it doesn't revolve around legitimate applications of secrecy. No protests are heard about the need for classification of nuclear-weapons secrets, or about the general propriety of background investigations and security clearances.

Where there are protests, they concern what is perceived to be an arbitrary and unwarranted assertion of authority. The most frequently cited example is the widespread opposition to polygraph testing. Energy Secretary Richardson bemoaned the fact that “onehalf of the [Los Alamos] X Division members ... signed a petition opposing polygraphs.” But they were well within their rights to do so.

The polygraph is something of a fetish in the world of intelligence, where polygraph testing is an indispensable rite of passage, at least from a sociological point of view. But the validity of polygraph testing for general screening of employees has never been scientifically demonstrated. (In contrast, there is some evidence of polygraph utility in incident-specific investigations.) That is one reason that polygraph testing as a condition of employment is prohibited by law in the private sector.

George P. Shultz, secretary of state during the Reagan administration, famously threatened to resign during the Reagan Administration, rather than undergo polygraph testing. He was no scientist, nor was he indifferent to the requirements of national security. Rather, like many others, he found the polygraph to be intrusive and degrading. It is also prone to error.

Hypothetically, there are two ways to absolutely eliminate any future security violations at the nation’s nuclear laboratories.

One way would be to gather all of the nuclear secrets into a vault and to seal the vault permanently shut. That way they could never be removed from secure control-- or used.

Another way would be to publish all of the secrets on the World Wide Web. Once that was accomplished, then by definition it would no longer be possible for anyone to steal those secrets.

Since those options are impractical, it is necessary to accept the fact that there can be no absolute security. The best one can aim for is to manage the security risks, keeping them to a reasonable minimum, while fostering quality work and limiting costs.

Unfortunately, the current Congress is hooked on absolute risk avoidance, which is the enemy of good security policy. Last year, for example, Congress adopted legislation requiring the Energy Department to conduct a new review of hundreds of millions of pages of documents at the National Archives. The documents had already been declassified, but representatives wanted them searched for stray nuclear-weapons information that may have been inadvertently disclosed. This was a poor use of security resources, especially since no new funds were appropriated to carry out the requirement.

Meanwhile, as the DOE security czar, General Eugene E. Habiger, noted recently, Congress allocated a mere $10 million last year out of the $65 million that the department had requested for security upgrades in the nuclear-weapons complex. In view of the recent consequences of inadequate security, perhaps outraged members of Congress will now call for their own resignations.


If absolute security for nuclear secrets is out of reach, what then should be done? The way forward, charted several years ago by then-Energy Secretary Hazel O’Leary but never fully implemented, is to tailor the application of security through a combination of declassification and increased classification.

O’Leary has been celebrated by some, vilified by others, for her Openness Initiative and her ambitious declassification program. She is less well known, by supporters and opponents alike, for advocating higher classification in certain sensitive areas.

In particular, she began a Fundamental Classification Policy Review, which sought to establish a rational, updated foundation for the classification of nuclear weapons information. That review, conducted by government scientists and military officers, endorsed the declassification of various categories of information, but also called for increasing the classification of other categories of information, from "Secret Restricted Data" to "Top Secret Restricted Data." The 137 categories recommended for upgrades are classified, but are likely to include details on arming and disarming nuclear weapons.

That approach became known as the Higher Fences Initiative, because it envisioned placing higher security “fences” around select categories of highly sensitive nuclear-weapons information, while relaxing or eliminating controls on information of lesser sensitivity. In that way, finite security resources could be brought to bear in the most efficient way.

If such higher fences had been in place, the hard drives at Los Alamos would probably not have gone missing, since they would almost certainly have been bumped up to Top Secret, and Top Secret material -- unlike “merely” Secret material -- is rigorously accounted for.

Although both Congress and Secretary Richardson seem blissfully unaware of the fact, the Energy Department has actually been pushing the Higher Fences Initiative since 1997. Its efforts, however, have been consistently blocked by the Defense Department, whose concurrence is required.

In December, Pentagon officials wrote to DOE that the costs of instituting the Higher Fences Initiative would be “substantial," because it would entail upgrading of security clearances for personnel to handle the newly Top Secret information, construction of new secure facilities, and so forth.

There is a delicate balance to be struck here between security, financial costs, and ease of use. Maybe it is possible to reconcile the positions of the Energy and Defense Departments. Unfortunately, the charged political environment in Congress is not conducive to reconciliation.

Nonetheless, the basic principle of "higher fences" remains sound. By focusing security on the most sensitive secrets and relaxing security on everything else, it should be possible to turn a vast, intractable problem into an entirely manageable task.

Secretary Richardson has asked his agency to provide him with recommendations on the future role of the University of California by early this month. But it will make little difference who administers the weapons labs until more rational and realistic security policies are put in place.

Steven Aftergood is director of the Project on Government Secrecy at the Federation of American Scientists, in Washington.

FAS | Government Secrecy | July 2000 News ||| Index | Search | Join FAS