January-February 2004

Science and Security


To the Editor:

The debate over security and biological research in theĀ September-October issue emphasized many points, but failed to stress that the costs to the developed countries of security precautions that slow research heavily favor the terrorists. Practically every writer emphasized the number and importance of foreign researchers in the sciences to the productivity of the American scientific enterprise. However, no one appears to have noted that increasing numbers of papers, even in U.S. journals, come from laboratories not located in the United States. It is highly unlikely that the United States could long restrict knowledge unilaterally; of the authors, Wallerstein did note this point.

Further, terrorists are likely to be more restricted by lack of materials than lack of knowledge; well-known diseases are sufficient to cause major damage. How-ever, losses to the developed countries, especially the United States, from attempts to restrict knowledge are likely to be huge. With biological knowledge developed only slowly, new disease cures are likely to be delayed. This makes the "traditional" diseases available to the terrorists that much more effective.

In deciding where to set security boundaries, we should also remember that new diseases will appear in any case. Even without terrorists, diseases evolve, and antibiotics lose their effectiveness. Slowing the development of new antibiotics, or other means of dealing with evolving threats and natural new diseases, could lead to worse epidemics, and more deaths, than anything terrorists are likely to be able to accomplish. In a sense, a political reaction to terror threats is more like an autoimmune disease in the body politic than a rational response to the actual situation. It is good to prevent a highly infectious new microbe that produces high mortality from getting into general circulation, but doing so may be impossible without costing more in terms of other deaths. It is necessary to make a reasonable estimate of the relative likelihood of the risks, and this has evidently not been done.

Michael E. Green
(Physical Chemistry)
City College, City University of New York