Sharing a loss before anyone called it insurance

The oldest form of insurance is not a contract but an understanding: if your boat is lost, we will make it good between us, on the understanding that you would do the same. Groups of shipowners, of householders in the same street, of farmers with stock in the same district, arrived at this arrangement repeatedly, and it is genuinely different from a modern policy in one respect. Nothing is paid until a loss happens. The obligation is a promise to contribute when called upon.

The call and the premium

That distinction, between a fund built by calls after the event and a fund built by premiums before it, is the main axis of the whole history. Calling after the event is simple and cheap, requires no reserve, and cannot become insolvent in the ordinary sense, because it does not hold obligations it has not yet funded. Its weakness is severe: the members most likely to be unable to pay a call are precisely the ones affected by the event that caused it, and a loss large enough to matter arrives when everybody is least able to help.

Collecting in advance solves that and creates every subsequent complication. Once money is held against future losses, someone must decide how much is enough, which requires an estimate of losses that have not happened yet. Once there is a fund, there is a question of who owns it. Once there is a question of ownership, there is a question of what happens on withdrawal, on dissolution and on a surplus. The apparatus of insurance follows from the decision to pre-fund.

Fire, and why the mark was on the wall

Fire cover developed its own physical apparatus for a practical reason. An insurer of buildings in a dense town has an obvious interest in the fire actually being put out, and in an era when firefighting was organised by the insurers themselves, a brigade arriving at a burning building needed to know within seconds whether it was one of theirs. Hence the metal mark fixed to the front of the building. It is often read as advertising; it was primarily identification.

The same logic produced the first serious risk classification. Insurers of buildings began distinguishing construction, use and proximity, because a thatched workshop next to a chandler's is not the same risk as a brick house in a wide street. Once losses are being recorded and classified by cause, the fund has data, and the possibility of pricing individually rather than sharing equally appears. That possibility is where mutual insurance starts to become a technical business.

What the mutual form preserved

A mutual insurer has no shareholders, so its surplus belongs to the people it insures, and is returned as reduced contributions, as a distribution, or simply as a larger reserve held on their behalf. That produces a genuinely different disposition. There is no pressure to grow for the benefit of an owner who is not a member, and a very direct interest in avoiding losses, since every loss is paid by the same people who would otherwise keep the money.

It also produces the mutual's characteristic difficulty, which is capital. A body that cannot issue shares can only build reserves out of retained surplus, slowly. That is manageable for a fund whose risks are small, frequent and local, and it becomes an increasingly hard constraint as risks grow larger and more correlated. Much of the later reorganisation of mutual insurers turns on that single arithmetic fact rather than on any change of principle.