The Boring Middle

Most conflicts do not end the way the last three lessons describe. They do not end in a fall, or a victory, or a moment anyone can name.

They end in a document, produced after years of meetings, most of which failed.

Nobody Wins Anymore

Through most of the twentieth century, civil wars ended with one side defeating the other. Since about 1990 that has changed, and negotiated settlements have become the more common ending.

This is not because people got kinder. It is because a war usually reaches a point where neither side can win and both sides can see it. Researchers describe that condition as a stalemate that hurts. It is the least uplifting reason imaginable for peace, and it is the ordinary one.

What follows is talking. Enormous quantities of talking, in hotel conference rooms, over years, conducted largely by people you have never heard of.

The Problem At The Center

Every one of these negotiations has the same knot in it, and you already know what it is.

Two armed sides are asked to disarm. Whoever goes first is defenseless if the other one lies. Both know this, so both hesitate, and both hesitations confirm the other's suspicion. That is Lesson 4.2 with rifles already in the room.

Nothing about mutual goodwill solves it, and the parties usually do not have any. What solves it is the fourth item from that lesson, somebody above both of them. A third party who monitors, verifies, holds weapons in storage, sits between the forces, and can be blamed. Researchers who have studied this find that outside peacekeeping forces measurably reduce the chance a war restarts, and the effect is one of the better supported findings in this field.

That is the whole reason an international force with an unimpressive mandate and a reputation for accomplishing nothing is often accomplishing the only thing that matters, which is making it safe to be the one who disarms.

How Long This Takes

Longer than you think, and with more failure than the story ever includes.

In Northern Ireland, the agreement most people can name was signed in 1998. Its core terms closely resembled a deal that had collapsed twenty five years earlier. After the signing, the actual decommissioning of weapons took another seven years, through repeated crises, suspended institutions, and negotiations about the sequence of steps that were themselves negotiations about who would be humiliated.

In Colombia, formal talks with the country's largest armed group ran for four years. The agreement was signed and then rejected by the public in a referendum by a very narrow margin. It was renegotiated and passed through the legislature instead, and implementation has been contested every year since.

In Mozambique, a fifteen year war ended in 1992 through a process run partly by a small religious community with no state behind it, and the settlement held.

The pattern is that these processes fail several times before they work, and the failures are not evidence that the approach is wrong. They are the normal shape of it. Almost everyone involved is unknown, and the work consists of drafting, verifying, scheduling, and sitting through the same argument for the ninth time.

Who Breaks It

There is a specific danger in these processes and it is worth naming, because it explains events that otherwise look insane.

Every settlement creates people who do worse under it than they were doing before. A commander whose authority came from the fighting becomes a minor official. A faction that was included in the war is left out of the deal. A network that made money from the disorder loses the disorder.

Those people have a strong interest in wrecking the agreement, and the cheapest way to wreck one is a single well timed atrocity. It does not require much. One attack at the right moment can pull both leaderships back to positions they had already left, because each has to prove to its own side that it will not be taken advantage of.

This is why negotiators spend so much time on things that look like details. Who is invited. What happens to mid level commanders. Whether there is a schedule that survives an incident. An agreement that cannot absorb one bad week is not an agreement, and building one that can is most of the job.

What Settlements Buy And What They Cost

Power sharing is the usual instrument. Guaranteed seats, quotas in the army and the police, regional autonomy, a share of the revenue.

It works, in the sense that it stops the killing more reliably than the alternatives. And here is the price, stated plainly.

Power sharing hands positions to people who obtained their leverage by killing. It often freezes the identity divisions it was built around, by writing them into the constitution as permanent categories. It can lock a country into a politics organized around the war for decades after the war.

Which brings back something from Lesson 2.1. Peace is not quiet. A settlement can end the fighting and preserve almost every injustice that caused it, and produce a state that functions poorly and cannot reform itself. That is still usually better than the war. It is not the same thing as the conflict being resolved, and treating a signature as a resolution is how the next round gets missed.

Most peace is not an event. It is a maintenance job with no completion date, done by people whose names do not appear anywhere.