Pastoralists: displacement of people who were never in one place

A census fixes people to parcels on a date. Pastoral systems have neither fixed people nor parcels: the same family uses different areas in different seasons, rights are held by category rather than by individual, and the resource that matters most — water in the dry season — may be a hundred kilometres from where anyone was standing when the enumerators arrived.

What a project actually takes

Rarely a homestead. Usually some combination of: dry season grazing reserve, which is the binding constraint on the whole system; a water point; a migration corridor now blocked by a fence, a canal or a road; a salt lick; and calving or refuge areas used only in bad years.

Losing any of these can render a much larger area unusable, because the value of rangeland lies in the combination of resources reachable across a season rather than in the hectares themselves.[1] Compensating the fenced polygon misses almost all of that.

Taking a water point does not take a water point. It takes every hectare of grazing that was only usable because the water was there.

Why the standard instrument fails here

Three assumptions break at once. That each household has a parcel — pastoral tenure is typically common, layered and negotiated. That presence at enumeration indicates residence — herders may be hundreds of kilometres away for months. And that a single survey round captures use — seasonal resources are invisible in the wrong season.[2]

A survey run once, in one season, against parcels, will conclude that a project area is largely unoccupied. That conclusion has been reached, repeatedly, about rangeland that supports thousands of people.

Enumerating a mobile population

• Survey resources, not only households. Map water points, corridors, seasonal grazing and the times of year each is used, with the users identified as groups where individuals cannot be.
• Work across seasons, or at minimum interview about the full annual cycle rather than about current use.
• Go where the users are, including outside the project area, in the season they are reachable.
• Use the institutions that already govern access — elders, water committees, customary authorities — as the route to identify user groups, while checking that the poorest users are actually represented by them.
• Record livestock as the asset it is, since herd size is the wealth measure and the vulnerability measure in one.

What compensation looks like

Cash to individuals for a collective resource loss fits badly and often fails outright. More useful, in rough order of preference: physical replacement of the specific service — a new borehole for a lost water point, with a maintenance arrangement; negotiated access to alternative grazing, agreed with the communities that hold it rather than announced; crossing points and underpasses designed into linear infrastructure at the places herds actually move; and restocking or veterinary support where herd losses have already occurred.[3]

Community funds are the common default and the weakest option, because pastoral users of a lost resource are frequently not the people who control the receiving institution.

Conflict risk

Displacing pastoral access does not remove the need for it. Herds move somewhere else, onto land held by other people, and the resulting competition is one of the more serious indirect consequences a project can cause — particularly where the receiving area is farmed.

Assessing where displaced herds will go, and engaging those communities before the pressure arrives, belongs in the impact assessment. It is also the argument for locating linear infrastructure and fences with reference to migration routes rather than only to engineering cost.[4]