Re:Paper is a pet project where I get to read interesting papers from the past and share my thoughts/impressions. It's not too dissimilar to book reviews, except for papers. Or paper-like things (a talk that gets published, an essay from an academic, etc.)
For 003, I have The Political Uses of Madness from Daniel Ellsberg.
I call it the Madman Theory, Bob. I want the North Vietnamese to believe that I’ve reached the point that I might do anything to stop the war. We’ll just slip the word to them that… “Nixon is obsessed about Communism. We can’t restrain him when he is angry—and he has his hand on the nuclear button”—and Ho Chi Minh himself will be in Paris in two days begging for peace. — Richard Nixon.
With Trump's election (both in 2016 and 2020) as the head of the United States, strategists were quick to revisit the madman theory, first propounded in a lecture by Daniel Ellsberg, called The Political Uses of Madness.
I happened to chance upon this through Annie Jacobsen's The Pentagon's Brain, a book about DARPA's history, legacy and future. Ellsberg makes an appearance in the Vietnam war-era of DARPA, which involved a lot of RAND analysts and researchers.
Ellsberg wrote this before he was hired by RAND, a few years before US's involvement in Vietnam scaled up.
In 1959, when this lecture was written and delivered, the Cold War poles (US and USSR) had figured out the logistics of manufacturing, storing and delivering world-annihilating hydrogen bombs and were rapidly stockpiling to the teeth and more. There was no longer a first-strike advantage; a nuclear war, no matter who started it, would end in a potential total annihilation of most humans in those two countries and more. The deterrence was merely mutually-assured, total destruction.
Ellsberg considers this world, that Wohlstetter called The Balance of Terror, and ponders over what advantage, if any, could a nuclear-war-threatening power have in a blatantly-suicidal bid.
In the world of the Balance of Terror, will any ruler of a major state be able to persuade his opponents that he wants war; not only that he will risk it, but that he means to choose it? Perhaps not; they would have to think him very mad. But then, in that day they will not need certainty, or anything close to it, to capitulate; and where a little madness could win the world, it seems incautious to assume no madman will ever find authority. A man who was slightly mad—who looked, perhaps, at other things but risks, who wanted some things very much, who could act on impulse—could be "useful" to have in national politics: a man of whom it could be said, in negotiation, "If we don't get these demands, it's likely that he will take over; and then . . .?"
A 1% chance is all you need
In the lecture, Ellsberg posits that a mad, crazy leader has an upper-hand. It's not a big one technically. The threatening power has to convince the opponent (the victim) that there's at least a little more than 1% chance that the crazy threatener will precipitate war. That is all it takes for the victim to back down, acquiesce and give in:
And for the man who is convincingly mad, the risks of commitment may be peculiarly small. For his commitment, perhaps only his, will be plausible; his claim to believe the risks are small, or his willingness to take large risks, will be credible. Against rational opponents, the madman in this world can win; more than that, he can win safely. This puts an unprecedented premium on madness.
If [the victim] assigns more than 1% likelihood to the chance that his opponent will carry out his threat, will maintain his ultimatum, he will give in.
The blackmailer's problem is to convince his opponent that he is more than 1% likely to carry out his threat.
Ellsberg's closest point of reference (and in fact, the only one) is World War II. Hitler, the madman dictator, was a looming presence for most political analysts and Ellsberg, no exception, uses Hitler's example across.
He tells us of the time when Pierre-Étienne Flandin had raised the possibility that Hitler could be stood down:
Chamberlain entered the following remark in his diary: "March 12, talked to Flandin, emphasising that public opinion would not support us in sanctions of any kind. His view is that if a firm front is maintained Germany will yield without war. We cannot accept this as a reliable estimate of a mad Dictator's reaction." ... Prime Minister Baldwin calculated his own critical risk at this juncture as precisely as if he had been reading these lectures. "You may be right," he told Flandin, "but if there is even one chance in a hundred that war would follow from your police operation I have not the right to commit England." The "mad Dictator," they calculated, was at least 1% likely to bring on a war if opposed; and that was enough to deter them.
A reputation for madness
Madness as a strategy is not simple.
In the lecture, Ellsberg lays out the nuances, the preconditions and the strategic add-ons one must adopt to sell the madness.
For instance, madness is not opportunistic or just-in-time; that would convince no one. One has to build a reputation:
...to be unpredictable; to seem "a little" erratic, impulsive, unstable.
... prepare for this long ahead of time, building a reputation for erratic, senseless, schizoid behavior.
Because the stakes are so high, the opponent will most certainly expect caution from the madman for what if an accidental discharge happens?
But in this case the enemy's expectations might be unusually precise; ... when a false move blows up the world, he might expect more than normal caution to avoid accidental discharge. The blackmailer, then, must take special steps to remove this influence.
With the dose of schizoid behaviour, Ellsberg recommends the "tactical use of obscurity". In reference to the Chinese position in Quemoy (Kinmen)1, where the Americans had no strategic response, he writes about how such a non-answer could be misconstrued as indecision unless offset by an obscure threat:
The Chinese (March, 1959) are doing new work in the line of inscrutability. What adversary in this generation is going to forget that they have been spending the last six months, shelling the island of Quemoy every other day, excluding holidays? ... Again, a statement by President Eisenhower in a press conference might have cautioned the Communists against relying too heavily on apparent indecision as indicating total lack of commitment. When asked whether there were any circumstances under which a field commander might be authorized to use nuclear weapons on his initiative, the President's immediate reply was: "I don't know."
"Unassailable" need
Ellsberg's other qualification to the madness theory is that the "blackmailer" must demonstrate an "unassailable" need for what's demanded.
It is a talent to have desires so strong that an opponent will believe you literally willing to gamble your life to achieve them. Thomas Hobbes named this condition, in The Leviathan: "To have stronger, and more vehement Passions for any thing, that is ordinarily seen in others, is that which men call MADNESSE." ...
"all Passions that produce strange and unusuall behaviour, are called by the generall name of Madnesse." (Hobbes)
Here, he invokes Hitler's example:
[Hitler] confronted his opponents who wanted nothing badly, with intense, obsessive wants; he found the status quo they offered him unbearable, to be not merely broken but revenged.
Hitler, Ellsberg says, wanted war. He had an obsessive need for "greatness" that transformed into a lust for military wins. His opponents, the rational leaders of other countries, were explicit about not wanting war.
Hitler threatened war; so did his opponents. But these statesmen, who were not mad, did not disguise the fact that they feared war, that they would pay to avoid violence and risk. "War is a fearful thing," said Chamberlain, two nights before he flew to Munich; "we must be very clear, before we embark on it, that it is really the great issues that are at stake, and that the call to risk everything in their defence . . . is irresistable." Hitler had other payoffs. He was, to begin with, a gambler; he meant to take risks, not avoid them. Measuring his payoffs by his willingness to take risks, the possible gains from any gamble come out with a high value, the possible losses trivial.
Hitler, violence and war were values, not costs; they were preferred means, and more than that, virtually ends in themselves.
To Rauschning he said in 1932: "None of these people any longer want war and greatness. But I want war."
In the end, his lust for military victory rather than diplomatic victory led him to disaster; but earlier, this tension between his conflicting desires for the clever, political success and for the triumph of his war-machine, so evident beneath his impatience, gave him an immense advantage in negotiation.
A purposive madness
Not all kinds of madness pay. The unassailable need has to be backed by a kind of relentless, irrational pursuit by all means:
The blackmailer's behavior must seem purposive, not random: but directed to mad ends ... None but a madman, the victim might believe, would accept a high risk of major war for a small increase in territory, for the difference of a few days in an occupation.
Once again resorting to Hitler's example:
"We shall not capitulate—no, never," he said; it was the summer before he became Chancellor of the Third Reich. "We may be destroyed, but if we are, we shall drag a world with us—a world in flames."
The Sudeten Germans must come into the Reich. Hitler said he "would face any war, even the risk of a world war, for this. Here the limit had been reached where the rest of the world might do what it like, he would not yield one single step." Details did not concern him. "Three hundred Sudetens have been killed, and things of that kind cannot go on . . . I am prepared to risk a world war rather than allow this to drag on."
Years later, Roseanne W. McManus wrote Revisiting the Madman Theory in which she differentiated between the two kinds of madness — situational and dispositional, and paired them with the notions of rational-choices-but-made-with-extreme-preferences (purposive) and irrational decision-making:
...situational extreme preferences constitute the type of perceived madness that is most helpful in coercive bargaining.
The madness is not dispositional or a mere characteristic; it comes from someone who we know is capable of rationality but the choice or preference is extreme.
Ellsberg also makes a couple of other strategic points that I found interesting.
First of these is the effectiveness of an ideology or dogma-driven blackmailer:
When a Communist says that a particular prediction is derived from Lenin, his opponents are forced to concede that he is fairly likely to believe that prediction and act on it—no matter how far removed from reality it may seem. Perhaps the test of a publicly-professed doctrine is not, "Is it true?" but "Is it what I want my opponents to think that I believe?"
and:
Personal dogmas, idiosyncratic conceptions of the world, can serve the same purposes, making particular beliefs both plausible and rigid, impervious to communications from the opponent.
Ideology and doctrine have other benefits to a blackmailer, or for that matter, to a prospective victim. They may make him impervious to counter-threats.
That is, even in the face of imminent, total annihilation, the Russians were ready to stand their ground because:
Capitalism will be destroyed; communism will survive, even if the present communists perish; and this would be victory; this is all that matters.
Coupled with this is the natural inclination to reject the opponent's offer (of peace) by downplaying the risks:
... he may deprecate the "badness" of conflict, relative to other outcomes: in particular, relative to the opponent's offer. ... Typically, for instance, both the union leader and the management representative will claim that they have no fear of a strike, that they have "little to lose" by it, are perfectly prepared to last it out, and that they expect the eventual outcome of a strike to be favorable to them...
Ellsberg also factors in the effectiveness of eradicating all communication lines for if one were rational enough to keep the channels open, the threat of influence (and buckling) is huge:
The more accessible the victim is to communication, the more open to rational influence, the more vulnerable he is to last-minute "briefings" such as Goering delivered to his friend, Sir Nevile Henderson, after the first meeting between Chamberlain and Hitler at Berchtesgaden. If there were war, Goering informed him, one thing, at least, was quite certain; there would be "little of London left standing."
The only way to discourage such communications is to cultivate psychosis at the decision-making levels. If ignorance is good, madness is better. But both are risky; ... When to believe your enemy's threats or warnings just a little is to capitulate or to abandon your blackmail, there are great temptations to become totally blind and deaf;
And with Hitler, Ellsberg writes:
[He was] deaf to their counterthreats. No one could reach him, teach him new values or beliefs, crack his confidence by magnifying risks.
Ellsberg, and later McManus, make a case for strategic advantages of the madman theory. McManus, for example, considers a host of leaders who have exhibited varying degrees of madness: Khrushchev2, Saddam Hussein, Gaddafi. Of course, no Nixon or Bush.
However, with Trumpian failures, Samuel Seitz and Caitlin Talmadge re-examined the theory in their The Predictable Hazards of Unpredictability: Why Madman Behavior Doesn't Work. They write that history says otherwise; madman theory just doesn't work:
The historical record, both before Trump’s presidency and during it, demonstrates that madman tactics typically fail to strengthen deterrence or generate bargaining leverage for three reasons.
The first:
... signaling extreme preferences (e.g., all-or-nothing demands for massive unilateral concessions to the US position) and/or a willingness to bear high costs to the point that it deviates from normal consequence-based decision-making (e.g., threats to start a war that will kill many people on both sides) is actually quite difficult.
And, secondly:
even when other countries clearly receive a madman’s message (or tweet), they often do not view it as credible for various reasons.
And finally:
even when such demands and threats are deemed credible, they do not necessarily induce the desired behavior from the targeted country, because madmen have a hard time giving believable assurances of behaving in the future3.
Nevertheless, Ellsberg's paper/lecture (in combination with the whole series, Art of Coercion) is a fascinating light into the mechanics of how cold and calculating political realism is, and how academics and military analysts like him tackled the idea of an upper-hand in a world that virtually guaranteed none.
*
1 PRC bombed Quemoy (Kinmen) islands (which belong to Taiwan/ROC) during the Second Taiwan Strait crisis. Beijing used a weird strategy of bombing on odd-numbered days and pausing on even-numbered ones; a total of ~500,000 shells were dropped. PRC could not win the island. The metal from shells eventually became a local resource.
2, 3 Here, an interesting anecdote. Ellsberg quotes Khrushchev as having said, in 1959, "The difference between Munich and now is that I am not Adolf Hitler." Khrushchev was referring to the fact that when western powers accepted Hitler's Munich demand (The Munich Agreement) where Hitler said that would be the last territorial claim, he did not stick to his word. Khrushchev, in post-war Berlin, had given an ultimatum to the Allies to vacate and demilitarise Berlin or risk a war with the Soviets, and the Allies thought giving-in to this demand had no guarantee that the Soviets would stop demanding more. To which Khrushchev said, "I am not Adolf Hitler." The Allies did not demilitarise, leading straight to the Berlin Wall saga.