More Than One Household

Thirteen children, three houses, and one piece of paper on a table.

The first wife had two children. The second had four. The third, who is much younger and whom the older ones have never entirely accepted, had seven. Everything the man owned is on the paper: four properties, a transport business, two accounts, a farm.

Somebody proposes that the estate be divided into three, one part for each house. This is not a manoeuvre. It is how the eldest son saw his own father’s estate divided, and he says so.

Somebody else proposes that it be divided into thirteen, one part for each child. This is not a manoeuvre either. It is what the youngest house has assumed since the burial, and it is also — the room is told, correctly — how it is done in many places.

Both proposals are lawful. Under the first, a child of the first house takes three and a half times what a child of the third house takes. Under the second, they take the same. The estate has not changed. Nobody has been dishonest. Nothing in the room has been decided, and everything has.

(A constructed illustration. The pattern is ordinary; the persons are not real.)

The question is not what the estate is worth. It is whether it divides among houses or among children.

Customary marriage in most Nigerian communities is polygynous, and Islamic personal law permits plural marriage within its own limits. Neither is a deviation requiring explanation; both are ordinary legal arrangements under which a very large number of Nigerian estates fall to be distributed.

What distinguishes such an estate from a monogamous one is not scale. It is that a second unit of account appears. In a monogamous family the only unit below the estate is the individual child. In a polygynous family there is an intermediate body — the house, meaning a wife and her children as a group — and it is a real institution with its own internal seniority, its own compound or wing of the compound, its own economy, and frequently its own view of the estate.

Once that intermediate unit exists, every division must answer a prior question: at which level does the estate split. If it splits at the level of the house, the number of children in each house is irrelevant and a small house is advantaged. If it splits at the level of the child, the house is irrelevant and a large house is advantaged.

This is not a subtlety. It is the whole of the dispute in most polygynous estates, and it is usually conducted without anybody naming it, because each side is using a rule so familiar to them that they take it for the rule.

A related feature causes more friction than it should, and it is worth naming here because it recurs in every act below. Polygynous families carry two orders of seniority at once, and they do not agree. There is the seniority of the houses, fixed by the order of the marriages, and there is the seniority of the children, fixed by their dates of birth across all the houses together. The eldest child of the family is frequently not the eldest child of the first house, and the eldest son of the first house is frequently not the eldest son of the family. Whichever ordering a custom uses for a particular purpose — the identification of the head, the right to speak, the allocation of a particular property — the other ordering is available to whoever is disadvantaged by it, and will be produced.

Everything else in this essay follows from that single structural fact.

Four methods of distribution operate in Nigeria. Which one applies is not a matter of preference; it is fixed by the system that governs the estate, and only within one of the four is there any element of choice.

Where the applicable customary system permits both Method One and Method Two, the procedural question becomes the substantive one, and it is worth more than any individual asset in the estate.

Customary systems differ in how the choice is made. In some it belongs to the head of the family or to the eldest surviving child; in others the ordinary rule is per house unless those entitled agree otherwise; in others the ordinary rule is per head. The applicable rule is the rule of the particular custom, proved as a fact in the ordinary way, by those who know it.

It is worth saying how such a question is actually resolved when it reaches a court, because families assume it is a matter of opinion and it is not. A custom is a question of fact. It is established by the evidence of those who know it — elders, chiefs, members of the community — or, where it has been so frequently established in the courts as to be notorious, by judicial notice. That means a family asserting per-house division must be prepared to prove that this is the rule of their community, and a family asserting per-head must do the same. Neither assertion carries weight because it is made confidently, and both are testable long before anybody goes to court.

Two practical observations follow, and they are the reason this act exists.

The first is that whoever convenes the meeting frames the method, and the method decides the estate. A family that spends its energy arguing about the valuation of the transport business while conceding the arithmetic without noticing has lost before it has begun. The first question at any distribution meeting should be which method is being applied and on whose authority.

The second is that the choice, once acted upon, becomes very hard to reopen. Properties are allocated, occupied, improved and sold on. A challenge brought four years later is a challenge to a settled state of affairs in which third parties now have interests. The window in which the arithmetic is genuinely open is short, and it closes quietly.

A polygynous estate is where the whole of this cluster meets at once, and four collisions recur.

Daughters are included. A method of distribution that counts only sons, whether per house or per head, does not survive the Constitution, and the Supreme Court so held in Ukeje v Ukeje (2014). A distribution conducted on the old basis is not merely unfair; it is one a court will set aside.

The unrecognised household appears. Alongside the recognised houses there is very often a woman with children and no valid marriage in any form. She has no claim as a wife under any of the four methods, and her children have the same claim as every other child. The distinction is set out in Sibling 01 and it produces the same painful shape here as everywhere else in this cluster.

The family house does not divide. Where the principal asset is family property rather than the deceased’s own, no method of distribution touches it, because there is nothing to distribute — the holding passes as a unit and can only be dealt with by the concurrence described in Sibling 04. Families frequently allocate a family house between houses at a distribution meeting, and the allocation has no legal effect whatever.

A predeceased wife does not remove her house. Where a wife died before her husband, her children remain a house for the purposes of a per-house division. The suggestion that a house ends with the wife who founded it is a common one in these meetings and is not, in the systems that use the method, correct.

A rule of division is worth more than any asset. It decides every asset at once, and it is settled before anybody mentions money.

Everything above is the law of intestacy. It is what happens when a man leaves several households and no instructions, and the striking thing about it is that he was entitled to give the instructions and did not.

A will fixes the arithmetic. A testator may provide per house, per head, or on any basis he chooses within the limits of the applicable law — which for a person governed by Islamic personal law is a real and defined limit, and which everywhere leaves the family house and other property he could not dispose of outside his gift. What a will cannot do is stay private: it is published on probate, and in a multi-household estate that means every house reads the full arrangement at once.

A trust fixes the arithmetic and controls the timing. It holds property outside the estate, provides for named people whether or not their household is recognised, appoints someone other than a family member to administer, and can hold assets that would otherwise have to be broken up to be shared. Where a family will not be able to convene, this is the instrument that removes the need for it to convene. Sibling 09 of this cluster takes it apart, and the seven-step architecture in the set-up cluster is the build.

Lifetime transfers do the most definitive work of all and are the ordinary answer for a house that a particular household actually lives in. The funding sub-cluster’s essay on land explains why the deed alone is not the transfer.

Two further points are specific to this essay. The first is that a father who intends per-house treatment should say so, and a father who intends per-head treatment should say so, because in the absence of a statement the choice passes to whoever runs the meeting. The second is that the reasons belong in a letter of wishes rather than in the instrument — in a multi-household family the perceived motive behind an unequal provision does more damage than the inequality, and a document in his own voice explaining what he did is the only thing that answers it after he can no longer be asked.

Where a family expects to face this repeatedly across generations, the mechanism rather than the single instrument is the answer. The family constitution essay in the family governance cluster sets out how a family establishes a membership roll, a forum and a decision rule — which is exactly what the room in the cold open did not have.

He left three houses and no arithmetic. The arithmetic was chosen anyway, by the person who spoke first.