Micromotives and Macrobehavior

by Thomas C. Schelling · 7 ideas

  1. Results Don't Reveal Motives

    What a whole group ends up with often says little about what each person wanted. A traffic jam is made entirely by drivers, and not one of them wants it. Before you judge people by a group outcome, work out which ordinary choices could have produced it.

  2. Equilibrium Isn't Approval

    An equilibrium is just a state where nobody has a reason to change. It can be a bad one: everyone drives because buses run empty, and buses run empty because everyone drives. When things settle, ask if the result is good, not just stable.

  3. Check the Arithmetic

    Some outcomes are fixed by counting, not by behavior. In musical chairs someone is always left standing, and half of any class is always in the bottom half. Before blaming people for a result, check whether the numbers made it unavoidable.

  4. Mild Preferences, Sharp Splits

    If people move only to avoid being a small minority among neighbors, a mixed area can still split into fully separate groups. Each move sets off more moves. Don't read a sharp divide as proof of strong feelings.

  5. Critical Mass

    Many choices depend on how many others make them, like showing up to a weekly meeting. Below some number, people drop out until it dies; above it, it keeps itself going. Push hard to get past that point, and act at the first signs of shrinking.

  6. Small Costs to Others

    Drivers who slow down to look at a crash each lose a few seconds, but together they cost everyone behind them far more. People weigh only their own cost. When small choices add up to a big shared cost, change the setup instead of asking people to care.

  7. Rules People Want

    Players who won't wear a helmet alone, because it puts them at a disadvantage, may gladly vote for a rule that makes everyone wear one. Each wants the safety but not the cost of going first. When people resist a good habit one by one, offer a shared rule.

More on systems