Sharing under scarcity

Why do the peoples living with the least reliable food supply give away the most of what they catch?

The intuition is that generosity is a luxury: give away your surplus when you have one, hold tight when you do not. Foraging societies invert it. A Hadza or Ache hunter who brings down a large animal keeps a small fraction for his own household and watches the rest go out across the camp, under some of the strongest norms anyone has recorded. Before explaining that, it is worth checking the question's own wording — because the driver turns out not to be scarcity.

Look at what varies within a single camp. Roots, nuts and berries are gathered in fairly reliable quantities, day after day, in small divisible amounts. They are shared — but mostly within the household, and nobody makes a fuss. Large game is shared across the whole camp under intense social pressure. Same people, same day, radically different rules. So the explanation cannot be a general disposition to generosity. It attaches to the type of food.

What distinguishes them? Not calories, and not scarcity. Large game arrives rarely, unpredictably, in enormous indivisible packages, and rots within a day or two in the heat. Gathered food arrives predictably, in portions, and can be got again tomorrow. The word we want is variance.

Now put a number on it, because it is more extreme than most people picture. Among the Hadza, big game is brought back on something like three of every hundred hunter-days. What does that mean for one man feeding one family on his own returns? He fails on ninety-seven days in a hundred. No amount of skill rescues that — it is a strategy with a high mean and a catastrophic distribution.

Now let twenty hunters put their catch into one pot. Each still fails on ninety-seven days in a hundred, but the chance that all twenty fail on the same day is roughly fifty-fifty, so the camp eats meat about half of all days; at thirty hunters, around three days in five. Nothing has been created. The same meat is eaten by the same people; only its distribution over days has changed, and that change is the entire benefit.

Which reframes what the giving is. What was the hunter's alternative? He cannot store a large carcass — it will spoil, and he cannot carry it. The meat he hands out has almost no value to him tomorrow. What he gets back is a claim on other people's future kills: he has converted a perishable asset into a durable one. That is not charity. It is the only banking available.

This is why the norms bite. A claim on the future is worth something only if others honour it, so what needs enforcing is not giving but not refusing. In many of these societies the transfer is initiated by the asker, and the unthinkable act is to say no. Alongside that sit deliberate mechanisms for stripping status out of success: among the Ju/'hoansi a hunter's kill is belittled by everyone present, and the meat is credited to the owner of the arrow that struck it rather than the man who shot it. Both prevent the pooling arrangement from hardening into a hierarchy of creditors and debtors.

Is risk-pooling the whole account? Honestly, no, and this is actively argued. One rival reading is that a man standing over a carcass simply cannot defend it: the tenth kilogram is worth far less to him than the first, and less than the cost of fighting off the person asking, so sharing is what tolerated scrounging looks like from outside. Another holds that big-game hunting is a poor way to feed a family and a good way to advertise one's quality, the returns coming as status and allies rather than meat repaid in kind. The evidence does not cleanly separate these, since they predict much the same behaviour.

The question's premise needs a small correction, then. It is not the least reliable food supply in the sense of the poorest; it is the most variable one. A society can be materially poor and share little if its returns are steady, and reasonably well-off yet share intensely if its windfalls are erratic and perishable. What predicts sharing is the shape of the distribution, not its mean.

It is worth separating this from two neighbouring ideas. It is not the anthropology of the gift, where a transfer creates a personal debt whose deferred repayment is the point and the relationship is the good — here the debt is diffuse, the ledger barely kept, and what is bought is smoothness of consumption. Nor is it insurance in the actuarial sense, where a pool works because independent risks cancel and a firm holds capital against the remainder. Here there is no firm, no capital and no contract, and the risks are far from independent: a drought hits everyone at once, which is exactly the failure these arrangements cannot cover, and why exchange partners in distant camps — whose bad years do not coincide with yours — are cultivated so carefully.

• What happens to these arrangements when a storable, hoardable good such as grain or cash enters the economy.
• Why exchange partnerships are maintained with distant camps whose fortunes are uncorrelated with one's own.

Every term the collection defines is gathered in the glossary.