Book recommendations (4): The Strategy of Conflict

Thomas Schelling won the Nobel Prize in Economics in 2005 together with Robert Aumann for work on game theory. Aumann’s work is in the traditionally mathematical area of game theory. Schelling’s is much more accessible and his main insights are to be found in “The Strategy of Conflict”, published in 1960.

Like much of the early game theory work, the book mainly concerns nuclear war, which is the ultimate zero-sum game (well, negative sum really). But Schelling explores the subject in a way that has a very wide range of applications to social life. In particular he analyses the psychology of threats in a fascinating way and shows how people can often converge on a solution (such as not blowing up the world) by use of shared cultural information, based on “focal points”.

For example, if you wanted to meet a friend in New York City but somehow had forgotten to specify a time and place, you might well converge on 12 noon at the top of the Empire State Building. Why? Because you would each put yourself in the other’s shoes and imagine times and places that would be “salient”, something that is hard to define without having a lot shared information. You would not choose odd times like 3.42pm and you would not choose obscure or random locations. You might not get it right but you could radically shrink the set of possible times and places to a very few.  In this way apparently impossible or computationally vast problems can be solved in practice, and the world emerges as better coordinated than we might expect.

There are many excellent books on game theory, which has been developed in both a rigorous mathematical direction and as a set of more heuristic parables, just-so stories and illustrative examples. Some of Schelling’s insights, expressed with virtually no maths, have since been formalised and made more precise. But many of his insights are very hard to turn into maths because of their very ambiguity.

The fact that is I am writing this now is a result of the most important fact in modern human history, namely that the USA and USSR didn’t quite go over the brink to Mutual Assured Destruction, though there were at least two very near misses. Schelling wrote when the outcome to the Cold War was unknown. Now we can read his book with a bit more detachment and appreciate the wider insights into how people and institutions made up of people, interact in ways that mostly make sense and mostly produce more order rather than less. This readable book is one of those that make you appreciate the power of a truly clever mind working productively on a problem of real interest.

Quite separately, Schelling is also famous for an early (1969) simulation in the field of what is now known as agent based economics (or more generally agent based social science, using computer simulations). He showed that even slight preferences for living next to neighbours of the same ethnicity can lead to very stark racial segregation over time.  Using a simple chequer board with coins, he showed that each time a household moved to follow its weak preference, the board became increasingly polarised. This was an illustration of the power of positive feedback. It didn’t prove that racial polarisation was caused by this process but showed that it didn’t require strongly racist attitudes in society for cities nonetheless to become highly racially divided.

Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.