The reason a walk-away must be credible to work

Stating that you will leave the table is easy and free, which is exactly why it rarely moves anyone. The other side is not evaluating your sentence. They are estimating what happens to you if the deal collapses, and that estimate is built from things you said long before the threat.

What everyone sees

A walk-away is treated as a matter of resolve: hold the line, do not blink, sound like you mean it. Delivery does matter at the margin. But it operates on a variable the other side has already largely set, because their belief about your alternatives was formed from information you leaked across the whole negotiation.

What is actually happening

Schelling analysed commitment as the central problem of bargaining: a threat is effective when the threatener has given up the option not to execute it, so credibility is manufactured by removing your own freedom rather than by asserting determination. Crawford’s treatment of disagreement in bargaining formalises the corollary, showing how impasse arises when commitments are made that cannot be retracted without cost. Both imply the same practical test. A walk-away is credible when carrying it out would be reasonable given your visible alternatives, and no amount of firmness substitutes for that, because the other side is modelling your incentives rather than your mood.

Why it stays hidden

The asymmetry hides because the threatener experiences their own certainty vividly. They know they would leave, so the threat feels strong. The other side has no access to that internal state and reads only the observable position, which is why the same words land differently depending on facts established weeks earlier.