Threshold Model of Collective Behaviour.

01Overview

Granovetter's threshold model dissolves the question 'why do crowds riot?' into a more precise one: what is the distribution of individual tolerances? Every person has a threshold — the number of others they need to see acting before they will act — and collective behaviour ignites only if that chain is unbroken. This small reframing explains the field's hardest observation: identical crowds can behave completely differently under near-identical conditions, because a one-person change in the threshold distribution can determine whether protest spreads or nothing happens at all. The model also explains why the worst violence often begins with the most moderate actors: the first mover merely legitimises second movers, and the chain runs through citizens, not instigators.

02Key theorists

Granovetter's 1978 paper in the American Journal of Sociology introduced the model through a vivid hypothetical about a crowd deciding whether to riot, showing that an equilibrium can rest on the last few hesitant joiners and collapse on a small change in the starting conditions. Schelling had earlier formalised tipping in neighbourhood segregation — small preferences for like-neighbours generating stark segregation — with the same mathematics. Kuran's 1989 Public Choice paper applied thresholds to revolutions, explaining the notorious surprise of overnight regime collapse: once thresholds untangle, the cascade of participation can run through a society faster than any intelligence service can track.

03How it works

1. 01Individual thresholds: from zero (the committed) through small integers (followers) to the very high (the crowd arrives at the door)
2. 02Chain connectivity: ignition requires an unbroken sequence — a gap between modest and moderate actors stalls everything
3. 03Equilibrium dependence: outcomes depend on the whole distribution, so small changes to key individuals can flip the aggregate
4. 04Overnight surprise: when thresholds are hidden (preference falsification), societies look stable until suddenly they don't
5. 05Moderates matter: the decisive actors are rarely the radicals but the first ordinary people who join
6. 06Preference cascade: participation itself reveals private opposition, lowering everyone else's threshold in turn

04Where it shows up

Protests that ignite or fizzle in near-identical conditions; market panics where a single institution's decision, not sentiment, starts the run; juries and committees tipping into decisions no member preferred; school and workplace cultures where one visible defection re-orders everything.

05The propaganda link

Regimes spend their lives policing thresholds: removing the first movers, punishing early joiners, keeping the radicals visible and isolated — because the Party understood that the crowd's problem is never motive but sequence. The system that wins the middle wins the turnout.

06How to resist

• Study the threshold, not the mood: who needs what number before moving?
• Protect the first joiners — they are the load-bearing people, not the wild ones
• Beware serial decisions masquerading as independent ones; each visible yes lowers the next person's threshold
• For organisers: engineer the chain knowingly, and expect it to be fragile at every link
• For the governed: note when a system works hard to keep the chain broken