Graham Allison revisited

Graham Allison, whom I did not know personally at the time, was one of the biggest influences on my graduate education at Harvard and my subsequent career.  Then a graduate student in Government, he belonged to an informal symposium called the May Group, after the historian Ernest May, who became my own dissertation adviser.  Another key figure in the group was the political scientist Richard Neustadt, Allison's adviser, who became a dear friend of mine.  Allison's masterpiece, Essence of Decision, which appeared just as I began graduate school in 1971, tried to summarize what the May group had learned about how governments made policy, both in theory and in fact.  To do so, he drew on what was then known about the Cuban Missile Crisis, with respect to decision making in both Moscow and Washington.

Allison built the book around three different models of governmental behavior.  The first, the rational actor model, was in a sense the most critical, because it represented what most people tended to think about governments, but quickly became, in the context of the book, a straw man.  Borrowing, perhaps, from microeconomics, historians and political scientists tended to assume that government leaders knew what they wanted and found the most efficient way to try to get it.  The actions of governments all traced back to a single plan.  Rational actor models dominated discussions of nuclear strategy in those days, as well as of more general analyses of Communist behavior.  They also dominated the argument over US Cold War strategy, on both the traditional anti-Communist and the newer revisionist sides.

Moving to Soviet and American moves in the missile crisis, Allison showed that a second model--Model II, the organizational process model--did a lot more to explain what the two sides actually did during the crisis. Model II focused on bureaucratic routines, or SOPs, which determined what parts of governments did, and how they did it.  US intelligence had recognized Soviet missile bases in Cuba because they looked exactly like similar bases in Europe.  Confronted with a new threat from Cuba, the Pentagon simply handed the President their existing war plan, combining sustained air strikes with a full scale invasion.  The Navy followed SOPs for blockades after the President ordered a quarantine.  I became convinced that organizational routines hold the key to understanding almost any large institution in the modern world--and I learned that those routines increasingly come to reflect the interests of the organization itself, rather than a broader rational goal which the organization was originally founded to pursue.  That insight has rarely let me down.

Model III, the governmental politics model, fit neatly on top of Model II.  The senior officials in the American government who met daily in the famous ExCom during the missile crisis represented bureaucracies--with the exception of the President and his brother the Attorney General, whose own bureaucracy played no role--but also brought their own views to the table.  Their critical decisions grew out of the interplay of their views and personalities, and who favored a certain course of action became just as important as what the course of action was.  Many years later, Ernest May and Philip Zelikow laid that whole process bare in their remarkable book, The Kennedy Tapes, which published transcripts of almost all the Excom meetings. Allison used Robert Kennedy's memoir 13 Days to show how an alliance of the President, RFK, Robert McNamara, and a few others had managed to avoid war, but he did not know that on the critical Saturday of the crisis, an even smaller group had agreed that Robert Kennedy would assure the Soviet Ambassador that the US would shortly remove its nuclear missiles in Turkey if the Soviets agreed to remove theirs from Cuba. McGeorge Bundy, the National Security adviser, revealed that in the late 1980s in his book, Danger and Survival.

Models II and III provided me with a framework for my dissertation on relations among Germany, Britain, and France on one side, and the new states of Eastern Europe on the other, which became Economic Diplomacy and the Origins of the First World War.  Neither then nor in later books such as American Tragedy and No End Save Victory did I ever use Allison's models explicitly, but I always tried to answer the kinds of questions that he had asked.  Meanwhile, from the beginning of my teaching career to the end, I assigned Essence of Decision whenever I could. The students in my freshman seminar on the origins of the First World War read it halfway through the semester and a couple of hundred students read it (or at least, were supposed to) as the reading period assignment in my lecture course on international politics, 1870-1970.  And I know that many of them, like me, never saw the world the same way again.

Surrounded as I was in the 1970s by people who had become familiar with Allison and gotten used to putting his ideas into practice, I thought that they would become a permanent part of the academic landscape.  I was wrong.  By 1990, although he remained at the JFK School of Government as a professor and dean, his book had become just another shooting star that had flashed across the sky.  Political scientists became more dedicated to rational actor models than ever, with rare exceptions.  Actually learning about what bureaucracies did, and why, was too much work.  Historians meanwhile lost interest in the real workings of governments completely.  All this makes me feel very lucky to have been exposed to him when I was.

The other day, one of my former Harvard students asked me on Facebook whether Allison could explain the Trump administration.  I gave it some thought.  Trump has undoubtedly introduced a whole new style of governing into American history, one that Allison did not specifically address. The deviations from Allison's models that he represents, however, still reveal a lot about his administration and its significance.

Trump, as I have said many times, simply cannot be viewed as a rational actor.  He lacks the attention span to absorb enough facts to make an informed decision about anything.  I have in the last few weeks been at work on a very narrowly focused history of the American presidency, and I have been struck, reading the annual messages of Presidents from Washington to (so far) Andrew Jackson, how frequently they specifically refer to the need to keep passion in check, and the critical importance of a well-educated citizenry in a democracy.  Trump on the other hand has literally nothing to offer but strong emotions--love for his followers, hatred of his enemies, and a boundless (but thereby illusory) self-confidence.  He rarely measures what he does against actual results--if has done it, it must be good.

Meanwhile, Trump for more than three years has been either subverting or ignoring the bureaucracies that he inherited.  While some of them have functioned without interference, his appointees have transformed others.  The EPA now tries to promote pollution of many kinds.  The Justice Department under William Barr transforms the guilty into the innocent, and vice versa. Homeland Security and ICE have new missions and new leadership.   Most notably, the whole foreign policy making structure centered on the National Security Council appears to have broken down.  Trump does not even consult it on major foreign policy decisions, and he has delegated enormous authority to his son-in-law Jared Kushner, who seems to operate on his own.

Since models I and II (see above) play little role in Trump administration policy-making, model III reigns supreme.  Policy emerges from unequal battles among individuals, in which Trump concentrates on remaining supreme.  He has fired anyone of any independence of mind whom he has appointed and is now almost entirely surrounded by sycophants who will accept his own view of reality.  Drs. Fauci and Birx, tasked with huge responsibilities in the midst of a pandemic, obviously understand that they must try to do their jobs without contradicting him in public.  The Trump presidency has offered an opportunity to men like Mike Pompeo and William Barr, who share his arrogance and hatred and have won their way into positions of enormous power by doing his bidding.  To find an analogy for this kind of system in western history, one would have to go back at least to early modern Europe, when monarchs and noblemen ruled the world without reference to the needs of their peoples.  Even within that context, however, Trump would stand out as a disastrous ruler.

Allison, thus, still allows us to understand how our government is working, even though our government has abandoned the principles and policies of 60 years ago.  How our nation has abandoned them, at least to the extent that he could be elected and reshape the government in his image in the first place, is a subject for another day.  The increasing supremacy of emotion, however--never more on display that in the last two weeks--has in my opinion a great deal to do with it.