Another Munich crisis? Understanding the limits of policymaking by analogy

We see it often enough. A democracy deals with an authoritarian state, and those who oppose concessions cite the lesson of Munich 1938: make none to dictators; take a firm stand.
And so we hear those voices today as US President Donald Trump proposes that Ukraine sacrifice territory for peace with Russia.
This is not to say that those who cite the lesson of Munich are always wrong, or even wrong in this case. It’s just that they may not always be right. More generally, we need to understand the limits of historical analogies. Use them with caution.
Historical reasoning, as former US defense secretary Ashton Carter put it, is ‘the dominant mental methodology of real policymakers’. The use of historical analogy has deep roots in the way humans think, to the point where it’s more unreasonable to expect it not to influence policymaking. Political scientists identify four sources of analogies—study, tradition, memory, and experience.
It is well-documented that president John F Kennedy’s approach to the Cuban Missile Crisis was informed by Barbara Tuchman’s account of the outbreak of World War I. This deeper study is also exceptional; given the demands of policy work, officials are rarely able to develop historical awareness beyond what they passively develop through cultural or institutional tradition.
This is why a small number of well-worn analogies get the most airtime. Few who refer to the Munich Agreement know it in fine detail. Even fewer have first-hand experience with it, as Anthony Eden did when confronting General Nasser in 1956. Yet personal experience and vicarious memory are by far the most common sources of analogy.
But historical analogies are persuasive: a study published in February shows that appeal to historical analogy in public messaging can swing opinion in favour of a given foreign policy action by about 7 percent. It’s not that we misuse historical analogies in a motivated or biased way, but rather that historical analogies are typically used by those who are already motivated reasoners.
Also, foreign policy issues generally aren’t compatible with scientific method, let alone mathematical certainty. What we refer to as ‘international order’ is a way of coping with this by artificially introducing regularity and predictability into a domain that, by nature, has none. Successes have here been modest at the best of times, and in moments of crisis, history is all that’s left to inform our actions. Thus, as order frays and crisis becomes persistent, analogies—to Athens and Sparta, to 1930s Europe, to the Cold War—become more relevant.
But the flaw in reasoning by historical analogy is more intrinsic. What historian Yuen Foong Khong called analogical explanation—drawing analogies with historical examples to grasp and evaluate new situations and forecast likely outcomes of action—has the same logical structure, and the same weaknesses, whether it’s being used to decide where to take a new date for dinner or if a country should invade Iraq.
Consider how critics have drawn on the Munich analogy in recent weeks. If, they analogise, the Trump administration allows Russia’s seizure of eastern Ukraine as the Allies allowed Nazi Germany’s seizure of the Sudetenland, then Russia’s assurances of ‘peace in our time’ are no more trustworthy than Germany’s.
The circumstances of Ukraine today and Czechoslovakia in September 1938 resemble each other on some points but differ on countless others. To draw the analogy is to decide that determining factors in the present are held in common with the past, while the points of difference are unimportant: that Nazi Germany did not possess nuclear weapons, for instance.
This view is defensible, but that isn’t the point. The point is that this view is, on its face, much harder to defend than the view that we are today simply in a Munich-like situation in Ukraine.
This leads to a core insight from Richard Neustadt and Ernest May. Policymakers are going to analogise from history regardless, so we must improve how they do it, and try to avoid disaster.
To Neustadt and May, there’s no substitute for historical awareness, and using a wider range of historical analogies often leads to more productive consideration of an issue than fixating on a single, hackneyed one. But those who use these analogies must understand that, in doing so, they are committing to definite claims about what is and isn’t causally effective, morally relevant, or both, in the course of history. Without this understanding, any number of historical analogies could be misused.
A more disciplined approach can be as complex as convening a panel of historians to give advice, or as simple as asking: what new evidence would I need to consider this historical analogy implausible?