Representative Democracy Cannot Bind Its Future Self

In a debate I watched some time ago, the streamer Destiny made a point in passing that I have not been able to stop thinking about: the United States has broken its agreements with Iran several times, and this teaches other states that American commitments are unreliable.

The specific case is arguable in both directions. The theoretical point underneath it is more interesting, and I have not seen it discussed much.

The Structural Problem

Every few years, the government of a representative democracy is replaced. And the replacement is, in a meaningful sense, a different government — free to repudiate its predecessor's commitments, and frequently elected on a promise to do exactly that.

This creates a credibility problem that has nothing to do with anyone's honesty. A negotiator across the table from a democratic state is not really negotiating with that state. He is negotiating with one administration, whose authority to bind its successors is limited and whose successors may have campaigned against the very agreement being signed.

An autocracy can promise things a democracy structurally cannot. Not because autocrats are more honest — obviously they are not — but because an autocrat's promise is backed by a continuity of power that a democratic promise is not. When the counterparty will still be there in fifteen years, a fifteen-year commitment means something.

Why the Usual Consolations Fall Short

There are standard replies, and they are weaker than they look.

"Treaties are binding in international law." They are, and states break them, and the enforcement mechanism is largely reputational. That is precisely the resource being depleted.

"Reputation disciplines states." It does, but reputation attaches to the country while the decision is made by an administration that will be gone before the reputational cost fully lands. The incentives of the decider and the bearer of the cost are not aligned — the same structural defect that appears everywhere else in this system.

"Democracies are actually more reliable, because their commitments are publicly debated and institutionally embedded." This is the strongest reply and it has real force. A treaty ratified by a legislature after public argument is harder to reverse than an autocrat's whim, and there is empirical work suggesting democracies do honour agreements at high rates. I take this seriously.

But notice the scope. It applies best to agreements that have become part of the domestic settlement — where reversal would be costly for internal reasons. It applies worst to exactly the agreements that matter most in a crisis: recent ones, contested ones, ones associated with a single administration, ones a challenger can campaign against. The cases where binding is most valuable are the cases where the binding is weakest.

What Would Actually Help

If the diagnosis is right, the remedy is to move commitments out of the space controlled by a single administration.

Supermajority ratification does some of this. So does embedding a commitment in domestic law with entrenchment provisions, or in an institution with independent standing.

But the more direct answer is to have the commitment made by the body that does not turn over: the public itself. A treaty approved by referendum is not the property of an administration and cannot be disowned as one administration's error. A successor who wants out has to go back to the same body that approved it and argue the case — which is a real constraint, and is the sort of constraint that makes a promise worth something to the party receiving it.

That is not a complete solution; publics change their minds too, though more slowly and less abruptly than governments do. But it moves the commitment to the most stable body available, rather than the least.

I offer this as a genuinely open question rather than a settled conclusion. It is possible that the credibility cost is real but small, and worth paying for everything else democracy provides. I would want to see the numbers before believing either way — but the mechanism seems real, and I have not seen a good account of why it should not matter.