The reason the last five per cent takes half the time

Eleven of twelve clauses are agreed in two sessions. The twelfth — about who is named in the announcement — takes six weeks, three calls that end badly and an intervention from someone senior on both sides. It is worth nothing in money.

What everyone sees

Remaining items are assumed to be representative of the whole, so five per cent of the issues should take about five per cent of the effort. That assumes a random sample. What remains is the opposite of random: it is precisely the subset that survived every attempt to trade it away.

What is actually happening

Fisher and Ury’s framework separates positions from interests and distinguishes issues that admit of exchange from issues where the dispute concerns standing rather than substance — the latter cannot be split because dividing them concedes the point. Northrup’s analysis of identity in conflict describes the mechanism directly: once an issue is bound to a party’s sense of who they are, concession threatens self-definition rather than resources, and parties will accept material loss rather than make it. Sequencing then guarantees the concentration. Negotiators begin with tractable items to build momentum, which is sound practice, and the effect is that every identity-bound item is deferred to the end and arrives together with no easy items left to trade against.

Why it stays hidden

The concentration hides because the agenda records size, not type. A clause about naming looks smaller on the page than a clause about payment terms, and progress is reported as items closed. Nothing in the tracking distinguishes an issue that can be split from one that cannot, so the end of the list looks like a short task.