More Rules for Launching Nukes Won't Make U.S. Safer
“We have taken every step man can devise,” said President Lyndon Johnson in 1964, “to insure that neither a madman nor a malfunction could trigger nuclear war.” Apparently lots of people are starting to doubt that this is true. The latest entrant is an op-ed article this week in the New York Times. The authors, Jeffrey Bader and Jonathan D. Pollack, call for legislation requiring that “a small group of officials” give unanimous consent before the president can use nuclear weapons, at least if the U.S. is not itself under attack.
Reading their thoughtful argument made me feel old. Some 30 years ago, the Federation of American Scientists offered a nearly identical proposal, differing only on the membership of the committee. I have taught about this and related questions for decades in my course on the ethics of war, and written about it more than once. 1
I quite understand that the often erratic posturing of the current president leads many to worry about how he will exercise his authority as commander in chief. But I am not sure that proposals to legislate around him are either constitutional or practical, and I worry that they represent an effort to solve the wrong problem.
Nobody likes the idea that an individual acting alone might start what could be the planet’s last war, but decades of worry have yet to come up with a good way of ensuring that it won’t happen. In the mid-1940s, Congress gave control of nuclear weapons to the president, in part to enable a swift response to a surprise attack, and in part to ensure that the military could not act without civilian authorization. 2 During the Cold War, presidents routinely “pre-delegated” their launch power to various military commanders (in case the chain of command was destroyed), but that practice was ended by President Bill Clinton.
Bader and Pollack are correct that the checks on the president's ability to order a launch are nowadays few. In theory, the secretary of defense, as part of the National Command Authority, must authenticate the commander-in-chief's order before the military acts. This is not a veto but a confirmation that the order is actually from the president. True, the secretary might also consider it his job to keep the commander in chief in check, as James Schlesinger reportedly did in the waning days of Richard Nixon’s presidency, but it’s not at all clear that he could pull it off. (One scenario: The president tries to order an attack. The defense secretary refuses to confirm the order. The president fires him on the spot. The deputy then succeeds to the position and confirms the order.)
Congress certainly could (but certainly won’t) alter the National Command Authority structure. I am skeptical that final authority on release of nuclear weapons could constitutionally be taken out of the president’s hands, but the issue would not arise until a security crisis did. In other words, there would be no opportunity to test the constitutionality in a court.
Still, whatever the difficulties of imposing new legal obstacles to the president's launch authority, informal checks certainly exist. For example, as the political scientist Paul Bracken has pointed out, the decision to place the responsibility to warn of an incoming attack in the Air Force rather than the Central Intelligence Agency was itself meant to reduce the likelihood that the U.S. would develop a “presidential nuclear force,” by which he means “a highly centralized nuclear force that is exclusively under political control.” 3
This is no small point. The president might well give the order, but there is no button to push. Multiple institutional layers, in quite different ways, would have to act before the order was carried out. Would they all go along? Possibly, but we can’t be sure. Neither military officers nor national security staff are robots. Should the president order an unprovoked nuclear attack, they might push back. In that situation, argued the political scientist Scott D. Sagan in the run-up to the 2016 election, “a constitutional crisis would be more likely than a prompt following of rules regarding succession and command authority.”
And there’s another check: human nature. Back in the 1950s, the federal government ran a mass relocation drill. Thousands of essential employees were swiftly moved to locations remote from Washington, as they would have been in the event of a nuclear attack. The exercise went reasonably well, but in the aftermath, many of those who had been relocated to other parts of the country confessed frankly that had the emergency been real, they would have refused to leave and died with their families.
The Times piece is motivated by the fear that an erratic president will simply up and launch when there is no threat justifying the use of nuclear force. But that is exactly where the existence of various institutional checks would seem to make the launch order less likely to succeed. 4
The more difficult problem is the one that has preoccupied both strategists and academics since the U.S. lost its monopoly on nuclear weapons: how the commander in chief responds to that 3 a.m. phone call that says some other country seems to have launched missiles. The president probably will have no more than four or five minutes to make a decision, leaving essentially no time to consult anyone else. That is when the fate of the world will rest on the ability of a single individual to put aside stress, passion and fear, and make a rational decision. No legislation will alter that.
It’s odd that every time a tough issue arises, our first instinct is to grab for some new laws. If we only sic some rules on the problem, we seem to think, then we’ve made progress. This is nonsense. Sometimes rules make things worse. Sometimes they give the appearance of a solution when they have solved nothing. Adding bureaucratic layers to the decision whether to launch a nuclear weapon sounds wonderful in the abstract, but nobody knows what would happen in practice. What we do know is that we would be flying in the face of what game theorists have long taught: The more you tell an adversary about what you won’t do, the more the adversary is emboldened.
As I’ve taught worried students for over a generation now, the only real way to restrict the authority of a president to launch a war is to elect someone you trust.
This column does not necessarily reflect the opinion of the editorial board or Bloomberg LP and its owners.
Originally, the warheads were physically separated from launch vehicles and bombs.
See Chapter 2 of this volume.
Perhaps we would do better to spend our worry on the risk of accidental launch of a U.S. nuclear weapon. As Eric Schlosser has chillingly detailed, the chances are significantly greater than we like to think.
To contact the editor responsible for this story:
Stacey Shick at email@example.com