The Limits of Safety: Organizations, Accidents, and Nuclear Weapons

Environmental tragedies such as Chernobyl, the Exxon Valdez, and Bhopal remind us that catastrophic accidents are always possible in a modern world full of hazardous technologies. Yet, the safety record appears to be extraordinarily good with nuclear weapons, the most dangerous technology of all. This safety record has led scholars, policy-makers, and the public alike to believe that nuclear weapons can serve as a safe and secure deterrent into the foreseeable future in the post-Cold War era. In this provocative and path-breaking book, Scott Sagan challenges such optimistic beliefs. Sagan's painstaking research into formerly classified archives penetrates the veil of safety that has surrounded U.S. nuclear weapons operations. Guided by theories of reliability in complex organizations, Sagan has uncovered a hidden history of frightening "close calls" to disaster: lost nuclear-armed bombers fly into the Russian warning net, Air Force officers tamper with missiles to be able to launch them without orders, B-52 bombers crash with thermonuclear weapons aboard and then vanish from the official histories, an unstable pilot deliberately turns on the two arming switches on his aircraft's nuclear bombs, and false warnings during the Cuban missile crisis lead pilots and radar operators to believe that the United States is under nuclear attack. Incomprehension, political maneuvering, and even cover-ups have limited what we have learned from these dangerous incidents, and Sagan maintains that many hidden bugs in the system remain. While the risk of deliberate nuclear war has been reduced with the end of the Cold War, the risk of serious accidents, even accidental war, remains unacceptably high. Theinheritance of nuclear missiles by Soviet successor states, the continuing spread of the bomb to developing nations, and misplaced confidence in the safety of our own arsenal should produce deep concerns. Unless we radically change the posture of our nuclear arsenal, over the long run, when we least expect it, a serious accident will occur. The key factors that scholars believe lead to high organizational reliability - redundant back-up systems, personnel discipline, and trial-and-error learning - have not produced a safe nuclear arsenal. This book therefore challenges our beliefs, not only about nuclear weapons safety, but also about our ability to control the many other hazardous technologies on which modern society is based.