Thomas C. Schelling (1921–2016)

there are some interesting situations where seemingly simple individual behaviors aggregate to surprising and potentially disappointing collective results: hockey players given an individual choice will opt for no helmet, making themselves collectively worse off and begging for league-mandated helmets; homeowners with a slight preference to live near people who look like themselves will induce stark segregation that shocks their own consciences. if these examples are familiar it is because this book introduced them.

the style of reasoning used in this book is now so familiar that it was briefly a meme: you could say schelling was the original "it's time for some game theory" guy. but that would be unfair, i think: the meme is mocking a top-down, overly theoretic way of forcing the facts into speculative theories. in this book schelling is sensitive to the facts of concrete situations, working in a bottom-up fashion that starts with humdrum examples taken from life, adding more detailed observations to the model until he brings it to the breaking point.

one pleasure of the book is seeing a systematic thinker at work: when schelling is done modeling a situation, he begins turning the parameter dials to see what other interesting outcomes he can get in theory and considering what concrete situations they might describe. when he's exhausted those possibilities, he lays the various models out and organizes them into logical schema. even there he does not stop, but instead ponders the nature of these schema, their limitations, their general properties: where do these simple accounting identities come from? what are the limits of their application? what kinds are there for closed and open systems? why do they seem obvious only after we have used them for a time?

schelling has a unique mind and i may need to dip into his other books.… (mer)