Climate action is a stag hunt

A stag hunt has two stable outcomes. If both hunters cooperate they bring down the stag — the big shared prize (mutual cuts → a livable global climate outcome). If either bails, the cooperators get nothing, so the safe move is to settle for the hare: a small, certain, solo payoff that needs no one else (free-riding — keep growing, let others decarbonize).

Both "both hunt stag" and "both chase hare" are equilibria — once you’re in either, no one wants to move alone. The stag equilibrium is payoff-dominant (better for everyone). The hare equilibrium is risk-dominant (safer if you’re unsure of the other). The whole drama is which one we coordinate on.

Watch the edges into trust. The reinforcing loop (R) is the one we want: mutual cuts that actually get delivered raise trust, which makes Nation A and Nation B more willing to commit, which produces more cuts. Cooperation compounds.

The problem is temptation to free-ride, which drains trust on a balancing path (B): if I can see the upside of defecting, I assume you can too — so I pre-emptively hedge toward the hare. Start trust low (it’s set to 40 in the model) and run the simulation: the free-ride loop wins, trust bleeds out, commitments soften, and the system settles into the risk-dominant hare equilibrium — exactly the stalled diplomacy we keep living through. Nobody had to be a villain. Low trust did it.

A prisoner’s dilemma is grim: even perfect trust doesn’t fix it, because defection pays regardless. A stag hunt is solvable, because the cooperative outcome is already what everyone prefers. You don’t have to change anyone’s payoffs or appeal to altruism. You only have to change beliefs about what the other will do.

That moves the entire problem onto one variable — trust — and trust is something institutions are actually built to manufacture.

Three levers all act on the same node:

• Signaling. Credible, costly, visible commitments (binding domestic law, early unilateral cuts) raise others’ belief that you’ll hunt the stag — pushing the R loop into gear.
• Treaties with verification. Monitoring and reporting let each side watch the other commit in near-real-time, so trust no longer has to be taken on faith.
• Lower the penalty for a failed attempt. The hare wins because a defected-on cooperator loses everything. Shrink that downside — transition finance, technology sharing, ratchet mechanisms that let laggards catch up — and committing to the stag stops feeling reckless.

Notice none of these argue that cutting emissions is good. Everyone already knows that. They argue that the other side will show up — which, in a stag hunt, is the only argument that matters.